Multinational Enterprises in Asian Development 9781847201027, 1847201024

Citation preview

Multinational Enterprises in Asian Development

Other books by the author Saving, Investment and Growth in India [with Kunal Sen] (2003) Crisis and Recovery in Malaysia: The Role of Capital Controls (2003) Growth, Employment and Migration in Mainland Southeast Asia: Structural Change in the Greater Mekong Economies [with Chris Manning and Piyasiri Wickramasekara] (2000) Liberalization and Industrial Transformation: Sri Lanka in International Perspective [with Sarath Rajapatirana] (2000) Structural Change and International Migration in East Asia: Adjusting to Labour Scarcity [with Chris Manning] (1999) Trade Policy Issues in Asian Development (1998) Macroeconomic Policies, Crises and Growth in Sri Lanka, 1969–1990 [with Sisira Jayasuriya] (1994) Export Instability and Growth: Problems and Prospects for Developing Countries [with Frank Huynh] (1987) Mudal Saha Mulya Ayathana [Money and Monetary Institutions] (2nd edition, 1986) Edited volumes The Economic Development of South Asia (3-volume reference collection) (2001) Developing Countries in the World Trading System [with Ramesh Adhikari] (2001)

Multinational Enterprises in Asian Development Prema-chandra Athukorala, Professor of Economics, Research School of Pacific and Asian Studies, Australian National University, Australia

Edward Elgar Cheltenham, UK • Northampton, MA, USA

© Prema-chandra Athukorala, 2007 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited Glensanda House Montpellier Parade Cheltenham Glos GL50 1UA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA A catalogue record for this book is available from the British Library Library of Congress Cataloguing in Publication Data Athukorala, Prema-chandra Multinational enterprises in Asian development / Prema-chandra Athukorala. p. cm. Includes bibliographical references and index. 1. International business enterprises—Asia. 2. Asia—Economic policy. I. Title. HD2891.85.A86 2007 338.8
885—dc22 2006023098 ISBN: 978 1 84720 102 7

Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall

For my great mentor, Max Corden

Contents ix x xiii

List of figures List of tables Preface 1

2

3

4

5

Multinational enterprises and developing countries: background and preview The role of MNEs in economic development: changing perceptions Scope and preview Foreign direct investment in developing Asia: trends, patterns and prospects Policy context Trends The ‘China fear’ Industry profile Prospects: a summing up Multinational enterprises and manufacturing for export: emerging patterns and opportunities for latecomers MNEs and manufacturing for exports: a typology Evidence Concluding remarks Appendix 3.1: Data sources on MNE share in manufactured exports Product fragmentation and trade patterns in East Asia Data Trends and patterns of product fragmentation Product fragmentation and trade patterns Conclusion and inferences Multinational firms in crisis and recovery: lessons from the 1997–98 Asian crisis Analytical context Capital flows during the crisis Role of MNE affiliates in adjustment and recovery Concluding remarks vii

1 3 6 12 13 26 35 39 41 45 47 54 67 70 72 74 75 84 97 100 101 102 112 119

viii

6

7

8

9

10

Multinational enterprises in Asian development

Capital inflows and the real exchange rate: foreign direct investment versus short-term capital Trends and patterns of capital inflow Real exchange rate–capital flow nexus: a first look Real exchange rate–capital flow nexus: empirical analysis Conclusion Appendix 6.1: The Australian model Appendix 6.2: Measurement of the real exchange rate Trade orientation and productivity gains from international production Analytical framework Data Productivity patterns Determinants of productivity Conclusion and policy inferences Appendix 7.1: Measurement of total factor productivity growth Multinational enterprises and the globalization of R&D: a study of US-based firms Theoretical framework Trends and patterns of R&D internationalization Determinants of R&D intensity Conclusion Appendix 8.1: Country coverage, variable definition and data sources and supplementary regression results Multinational firms and factor proportions in manufacturing: does parentage matter? Foreign direct investment in Sri Lankan manufacturing Methodology and data Results Conclusion Foreign direct investment in economic transition: the experience of Vietnam Investment climate Trends and patterns of FDI Economic impact Conclusions and policy inferences Appendix 10.1: Export processing zones in Vietnam

References Index

121 124 128 133 140 143 145 147 149 153 156 156 161 166 168 169 171 180 191 194 198 200 208 211 217 219 220 225 234 246 248 250 273

Figures 2.1

FDI inflows to developing countries and developing Asia, 1980–2004 3.1 Share of MNE affiliates and world market share in manufactured exports in selected Asian countries 4.1 World trade in parts and components, 1992–2003 4.2 Parts and components exports by major country groups, 1992–2003 5.1 Net capital flows to Indonesia, Korea, Malaysia, the Philippines and Thailand, 1996q1–2002q1 6.1 Net private capital flows to Asia and Latin America 6.2 Net capital inflows and the real exchange rate for selected Asian and Latin American countries, 1985–2000 6.A1 Capital inflows, the real exchange rate and economic adjustment in a small open economy 9.1 Foreign direct investment in Sri Lanka: value in US$ million and as a percentage of gross domestic fixed capital formation 10.1 Foreign direct investment in Vietnam, 1990–2005

ix

32 57 77 78 106 125 129 144

201 228

Tables 2.1 2.2

Foreign direct investment inflows, 1980–2004 Foreign direct investment inflows as a percentage of gross domestic fixed capital formation, 1980–2004 2.3 FDI flows to China as reported by China and by investing economies, 2000–03 2.4 US direct investment in selected Asian countries, 1994–2002 3.1 A typology of MNE participation in manufacturing for newcomer exporting countries 3.2 MNE affiliates and manufactured exports from selected developing Asian countries: MNE share in total manufactured exports and selected export performance indicators 3.3 Determinants of export market penetration: regression results 3.A1 Data sources on MNE share in manufactured exports 4.1 World trade in parts and components, 1992–2003 4.2 Annual compensation per worker in manufacturing, 1990–98 4.3 Parts and components in manufacturing trade 4.4 Percentage composition of parts and components exports and imports by three-digit SITC categories, 2003 4.5 Direction of parts and components trade 4.6 Share of parts and components content in regional manufacturing trade flows 4.7 Intra-regional trade shares: total manufacturing, parts and components, and final trade, 1992, 1993 and 2003 5.1 Capital flows in Asian crisis countries, 1990–2000 5.2 Percentage change in net capital flows during 1997–98 over 1995–96 5.3 Mergers and acquisitions by foreign firms in Asian crisis countries, 1990–2001, announced value 5.4 US direct investment in Asian crisis countries, 1994–2001 (US$ million) x

27 30 36 38 48

55 66 70 76 80 81 85 89 94

96 104 109 109 111

Tables

Asian crisis countries: foreign direct investment as a percentage of gross domestic investment, 1990–2001 5.6 Exports by majority-owned affiliates of US MNEs as a percentage of total host-country exports in East Asia, 1995–98 5.7 Exports by majority-owned affiliates of US MNEs as a percentage of total sales in East Asia, 1995–98 5.8 Share of employment in US affiliates in total manufacturing employment 5.9 MNE presence and post-crisis performance in Malaysian manufacturing 6.1 Net capital inflows to selected Asian and Latin American countries: 1985–99 6.2 Capital inflow episodes of selected Asian and Latin American countries 6.3 Determinants of the real exchange rate in selected Asian and Latin American countries 6.4 Change in explanatory variables during the capital inflow boom compared with the mean level for the sample period 7.1 Overseas operations of US manufacturing MNEs: estimates of productivity and related indicators 7.2 Determinants of productivity growth: regression results with trade orientation measured by export–sales ratio 7.3 Correlation matrix 7.4 Determinants of productivity growth: regression results with trade orientation measured by Sachs–Warner index and black market premium 8.1 R&D internationalization of US MNEs during 1966–2001 8.2 Industry distribution of R&D expenditure of selected countries, 1990–2001 8.3 Overseas affiliates of US manufacturing MNEs: FDI stock, sales, R&D expenditure and R&D–sales ratio by country/region 8.4 Percentage share of R&D expenditure of US MNE affiliates in total R&D expenditure in host countries 8.5 Determinants of R&D intensity: random-effect GLS estimates 8.6 Summary data on variables used in the regression analysis 8.A1 Country coverage 8.A2 Variable definition and data sources 8.A3 Determinants of R&D intensity: alternative regression results

xi

5.5

112

114 114 115 117 127 133 137 138 157 159 160

162 172 174

175 179 187 188 194 195 197

xii

9.1 9.2 9.3 9.4 9.5 9.6 9.7 10.1 10.2 10.3 10.4 10.5 10.6 10.7 10.8 10.9 10.10 10.11

Multinational enterprises in Asian development

Envisaged investment and employment in firms in commercial operation, 1978–2002 Source country composition of export-oriented firms in commercial operation as at end of 2002 Industry composition of export-oriented manufacturing firms with FDI, 1982, 1995 and 2002 Regression results, 1982 Firm characteristics by parentage, 1982 Regression results, 2002 Firm characteristics by parentage, 2002 World Economic Forum growth competitiveness and macroeconomic environment indices, 2005 Index of Economic Freedom Indicators of ease of doing business: Vietnam in global and regional context, 2005 Vietnam: source-country composition of FDI inflows, 1988–2005 Vietnam: sectoral distribution of cumulative approved investment, 1991, 1995, 2000, 2005 Vietnam: spatial distribution of FDI, 1988–2005 Foreign-invested enterprises in the Vietnamese economy: key indicators, 1995–2003 Vietnam: contribution of foreign-invested enterprises to industrial employment, 2000–03 Export performance of foreign-invested enterprises Commodity composition of exports by foreign-invested enterprises, 1996–2005 Vietnam manufacturing: estimates of productivity growth and related data, 2000–03

203 205 207 212 213 215 216 224 225 226 232 233 235 236 237 241 242 244

Preface Both scholarly and policy interest in the role of multinational enterprises (MNEs) in the contemporary world economy has burgeoned in recent years against a backdrop of growing economic integration of national economies into the global economic system. The data chronicling their global operations have become far more extensive, and the scientific literature, both theoretical and empirical, dissecting their activities has grown much richer. However, the implications of the operations of MNEs for economic development in host developing countries remain elusive. There are many unresolved issues relating to designing policies to regulate and monitor the entry and operations of MNEs as part of the overall national development endeavour. This book aims to fill this gap in the literature by examining some issues central to this policy debate in the light of the experiences of developing countries in Asia. Developing Asia provides a valuable laboratory for the study of these issues, given the long-standing presence of MNEs in many of these countries and the diversity among countries in terms of the stage of development and the timing of policy transition towards greater receptivity to MNE involvement in the national economies. The book begins with an overview chapter which traces the evolution of post-war thinking and paradigm shifts relating to the role of MNEs and foreign direct investment in economic development and describes the structure and contents of the ensuing chapters. The next chapter gives a broadbrush picture of policy reforms and the investment climate in developing Asian economies and examines, from a comparative regional and global perspective, their experiences as hosts to MNEs as reflected in foreign direct investment inflows over the past three decades. The rest of the book is structured thematically, with each chapter providing a self-contained treatment of a selected theme of the contemporary debate on harnessing MNE participation in national development. The issues covered in the chapters include the role of MNEs in manufacturing export expansion; the ongoing process of international product fragmentation and its implications for trade patterns and global integration of developing countries; global research and development activities of MNEs; the relative stability of foreign direct investment compared with other forms of capital flows in the context of international financial crises, and the implications of the xiii

xiv

Multinational enterprises in Asian development

operations of multinational enterprises for the recovery process; the implications of MNE presence for productivity growth in manufacturing, and the role of host country trade policy in conditioning the outcome; and the role of foreign direct investment in economic transition in former centrally planned economies (based on the experience of Vietnam). The core (thematic) chapters follow a common structure encompassing the state of the debate, relevant theory, methodology and data sources, and policy implications of the results, with extensive referencing to the related literature for those desiring to pursue the individual topics further. Two key concerns that guide the empirical analysis throughout are the interconnection between theory and practice and the choice of analytical procedures and tools with a view to getting the maximum out of the available (limited) data. I believe that the book will be of interest to a broad audience, consisting of students, professional economists and policy makers. The economics of MNEs is a popular subject in advanced undergraduate (college) and graduate curricula, in its own right or as a major component of courses in international economics, development economics and international business. While there are a number of excellent textbooks on the subject, there is a dearth of empirical evidence and case study material to supplement the analytics. This book aims to fill the gap. Apart from this pedagogical value, the book will also serve as a valuable reference source for professional economists, and policy makers in developing countries and international development agencies in broadening their understanding of the role of MNEs as an integral part of the international dimensions of development policy. The reader will find this book to be unique amongst the few available policy-oriented books in this area, not only in terms of the subject coverage but also in terms of the effort made to draw upon a variety of hitherto unexploited data sources in studying the issues at hand. Chapters 4, 5, 6, 7 and 9 draw upon my sole or joint contributions to the following journals: Asian Economic Papers (MIT Press); the Australian Economic History Review (Blackwell); The World Economy (Blackwell); Transnational Corporations (United Nations); and the Oxford Bulletin of Economics and Statistics (Blackwell). I thank the publishers for granting copyright clearance. The published material is incorporated in the book with considerable modification, rewriting and expansion in order to avoid overlap as well as to update the data and the literature coverage. It is a pleasure to thank everyone who helped me in this endeavour. Most importantly, I am grateful to my co-authors, Sarath Rajapatirana, Satish Chand and Sisira Jayasuriya, both for fruitful research collaboration over the years and for permission to make use of material from our joint papers in Chapters 6, 7 and 9 respectively. I would also like to express profound gratitude to Peter Drysdale, Chris Findley, Ross Garnaut, Hal Hill, Vijay

Preface

xv

Joshi, Chris Manning, Ross McLeod, Xin Meng, Kunal Sen and Peter Warr for valuable comments on and constructive criticism of various versions of one or more of the individual chapters. A special vote of thanks goes to Max Corden, James Riedel and Tony Thirlwall, and also to Sarath Rajapatirana for valuable advice and encouragement over many years. I have benefited from research assistance provided at various stages by my PhD students, Juthathip Jongwanich, Archanun Kohpaiboon, Nobuaki Yamashita and Tran Quang Tien, who always performed beyond the call of duty. To them, I express my profound appreciation. Edward Elgar Publishing Ltd has been unfailingly helpful; I am indebted to Edward Elgar for nudging me into this project, and to Alexandra O’Connell, Suzanne Mursell and the staff for converting a complicated, unwieldy manuscript into a beautiful book. Finally, my family – Soma, Chintana and Chaturica – deserve my warmest thanks for love, forbearance and unwavering support without which this task would never have been completed. Chintana and Chaturica also deserve thanks for help with preparing the manuscript. Chandra Athukorala Australian National University June 2006

1. Multinational enterprises and developing countries: background and preview Multinational enterprises (MNEs)1 are a major force in the evolving process of economic globalization which is encapsulating and reconfiguring the nature of global economic space. By the early 1970s, many US-based enterprises had already gone some way toward creating a global network, when European and Japanese firms began to take their place alongside the US firms. The global spread of companies from relatively high-income developing countries, in particular the East Asian newly industrializing countries, was an even more recent phenomenon, dating from about the mid-1970s. Today, the world is blanketed with affiliates of firms head-quartered in many different countries. Activities of MNEs have expanded from mining and petroleum industries, the traditional mainstay of their global operations, to manufacturing and services. They have established a strong and ever increasing presence in almost every country, including the former centrally planned economies in Asia and Central and Eastern Europe, linking factor and product markets across the globe. Global operations of MNEs are therefore a key factor that impinges on the designing of national development policy in the context of a rapidly globalizing world economy. Foreign direct investment (FDI)2 – the only ubiquitous quantitative indicator of the scale of MNE activity – grew dramatically from an average annual level of US$59 billion during 1980–84 to US$844 billion during 2000–04, recording an annual compound growth rate of about 5 per cent (compared with a 3.8 per cent rate of growth in world merchandise trade).3 The share of FDI in total private capital flows increased from 22 per cent to 32 per cent between these time points. The bulk of total world FDI inflows is absorbed by developed countries, with the share of developing countries (including transition economies) hovering around an average annual level of 28 per cent during 1980–2004. However, FDI inflows account for a much larger and increasing share of gross domestic fixed capital formation in developing countries compared with developed countries (8.5 per cent and 4.4 per cent respectively during 1980–2004). 1

2

Multinational enterprises in Asian development

Moreover, from about the late 1980s, FDI has continued to be the largest single source of total external finance (private capital inflows
official development assistance (ODA)) for developing countries. During 2000–04, FDI accounted for more than half of total external resource flows to these countries and was considerably larger than ODA.4 The scale of MNE activity is obviously better gauged by direct indicators of their performance, such as production, sales or employment, rather than by looking just at FDI.5 However, such performance indicators are unavailable for about half or more of the affiliates of MNEs, because most countries do not collect information on what their firms do outside their national boundaries (Lipsey 2004). According to available rough estimates (made by extrapolating the available data to the global level), production in foreign affiliates of MNEs (that is, production in enterprises located outside the country of residence of their parent companies) increased persistently from about 4 per cent of world output (GDP) in 1982 to about 10 per cent in 2004, and their share in world merchandise trade rose from 32 per cent to 39 per cent. Total employment in foreign affiliates recorded a three-fold increase (from 19.6 million to 57.4 million) between 1982 and 2003 (UNCTAD 2005, Table 1.3). Data on FDI (or direct performance indicators) tell only part of the story of the impact of multinational firms on host economies. This is because ‘the modern multinational company is primarily a vehicle for the transfer of entrepreneurial talent rather than financial resources’ (Dunning 1992, p. 321). The decision of a firm to set up an affiliate in a given country is underpinned by the desire to reap benefits from its specific advantages (in the form of technology, managerial expertise, marketing know-how and so on), which cannot be effectively leased or sold through ‘arm’s-length’ market dealings with unrelated firms. Thus multinational affiliates bring along with them some firm-specific knowledge in the form of technology and managerial expertise that cannot be effectively leased or purchased on the market by the host country. As part of the parent company’s global network, they also have marketing channels in place and possess experience and expertise in product development and international marketing. Given these firm-specific advantages, MNE affiliates have the potential to contribute directly to the economic growth of the host country through improvement in productivity. In addition to this direct effect, the presence of foreign firms can also impact on the overall growth performance of the host economy through ‘spillover effects’ (positive or negative) on local firms. On the positive side, foreign firms can act as conduits of new technology, managerial practices and marketing know-how for local firms, either through pure demonstration effect or as an integral part of their local operations through employee training and direct dealing with local firms in

Background and preview

3

procuring material inputs and services. Technology diffusion may occur through labour turnover as domestic employees move from foreign to domestic firms. On the negative side, MNE affiliates could well make use of their business prowess to thwart domestic entrepreneurial initiatives through ‘unfair’ competition. They could attract scarce trained domestic manpower at the expense of local firms by paying higher wages and providing better working conductions (Caves 1996, Barba Navaretti and Venables 2004, Blomström and Kokko 1998). A given degree of MNE presence as usually measured by FDI relative to the size of the economy is, however, unlikely to have the same impact on economic performance in all host countries. There is ample theoretical reasoning backed by empirical evidence that the national gains from MNE presence are conditioned by the nature of the domestic policy regime and various resource-endowment related factors such as the stage of human capital development and entrepreneurial advancement.6 Relating to the policy regime, a country with an outward-oriented policy regime has the potential to reap greater benefits than a country whose policy regime has a policy bias in favour of import-substitution production. This is because, in contrast to an import-substitution regime, an export-oriented regime generally encourages MNE activities where the host country has comparative advantages in international production. The growth effects of MNE presence may also depend on the host country’s policies impacting on the performance of domestic private enterprises. A policy regime which discriminates against domestic entrepreneurial initiatives in favour of stateowned enterprises, for example, could obstruct spillover benefits from multinational affiliates. As regards the domestic resource endowment, a host country which is at an advanced stage of human capital and entrepreneurial development is better placed to reap technological spillovers from MNE presence than a country with a lower ranking in terms of these preconditions. Spillovers also depend to a large extent on the effort of local firms to invest in skill development and research and development (R&D) activities.

THE ROLE OF MNES IN ECONOMIC DEVELOPMENT: CHANGING PERCEPTIONS During the first decades or so of the post-war period, there was a general consensus in the economic profession that FDI (or private foreign investment (PFI), as it was called then) was beneficial to developing countries in economic take-off as a complement to foreign aid. Foreign aid could do more to help as far as economic and social overheads were concerned, but

4

Multinational enterprises in Asian development

PFI was more appropriate in mining and manufacturing, areas in which governments had little or no expertise. In these areas, PFI was regarded as the largest potential source of capital and other complementary resources – entrepreneurship, technology, management and marketing – lacking in developing countries. Thus the policy advocacy of the time encouraged developing countries to provide a hospitable climate for foreign investment, not only through removing/reducing regulatory barriers but often through such policy inducements as tax holidays and subsidies (Little 1982, Chapter 8). This receptive attitude towards FDI was rather short lived, however. From about the early 1960s the relationship between the MNEs and the national state became increasingly hostile against the backdrop of the wave of independence movements and the struggle for social and economic transformation in the new nations emerging from the colonial era. The governments of developing countries began to be increasingly concerned with the ‘external’ economic and political effects of foreign investment rather than with its measurable effects on economic growth. This dramatic shift in perception was partly based on various highly published cases of ‘ruthless’ exploitation of natural resources by MNEs and the unacceptable intervention of some MNEs in the political affairs of some countries (Vernon 1971, 1977, 2000). There also emerged a growing scepticism in the economic profession of the day about the impact of FDI, with an increasing number of economists portraying foreign direct investment as basically an exploitative relationship which set developing countries on a dead-end route of ‘dependent capitalism’. In line with the basic thrust of development thinking of the day, which considered the external resource gap as the prime constraint on economic take-off in developing countries, the economic debate on FDI began to be dominated by the fear that MNEs might worsen countries’ balance of payments. A series of studies commissioned by UNCTAD on the subject in the early 1970s (the main findings of which were subsequently published in Lall and Streeten 1977) forcefully argued that MNEs contribute to a worsening balance of payments position of host countries through transfer pricing between affiliates located in different countries; by introducing inappropriate, import-intensive technology; and by financing investment from funds raised locally rather than through actual direct foreign investment, while not generating enough foreign exchange earnings (through exporting) to counterbalance their profit remittances. These studies, which spawned a series of supporting studies by various individual authors,7 provided the rationale for an increasingly restrictive and selective approach to foreign investment approval and monitoring of the activities of MNE affiliates in host countries, including the imposition of export

Background and preview

5

performance requirements and restrictions on profit remittances, and setting up of institutional mechanisms for monitoring pricing practices. Another school of thought that held sway (particularly in Latin America) during this period argued that FDI helps to create local initiatives in the early stages of development but later discourages them by suppressing local technological development and growth of indigenous enterprises (Hirschman 1969). The real national development challenge for host countries to MNEs, according to this view, was ‘how to divest [not how to promote] FDI’. The hostile attitude towards MNEs based on these political and economic considerations triggered a wave of nationalizations of MNE affiliates in many countries. Home countries of MNEs, the USA in particular, responded to these nationalizations with various retaliatory actions, compounding the initial host-country suspicion of MNEs as instruments of imperialism. The cumulative outcome was a massive contraction in foreign investment in many countries in the 1960s and 1970s (Vernon 2000, pp. 67–70). From about the late 1970s, the interaction between the developing countries and MNEs gradually shifted from being largely adversarial and confrontational to being conciliatory and cooperative. Several factors accounted for this change in attitude. First, a growing body of scholarly research served to tone down the strongest of the criticisms of MNEs in developing countries. In particular, the new evidence served to demonstrate that the case against MNEs had greatly exaggerated the difference between indigenous and foreign enterprises and that both types of firms may do economic harm in overprotected markets. MNEs could be effectively engaged as partners, rather than protagonists, in the national wealth-creation process by adopting market-oriented policies which promote efficient operation of both foreign-owned and domestic enterprises. There was indeed mounting evidence that some developing countries which had adopted very liberal policies on MNEs, with no regulation on transfer-pricing and relatively attractive tax rates, did not seem to suffer from the problem. Furthermore, an increasing number of well-reasoned analyses convincingly demonstrated that the early anxiety about transfer-pricing practices of MNEs had been underpinned by misconceptions about what the phenomenon was all about. As a key proponent of the early view subsequently admitted, ‘most of the evidence [of transfer-pricing] came from one industry, pharmaceuticals, and involved the comparison of prices charged on intra-firm transactions with what would be charged by a non-patent observing imitator . . . [and this comparison ignored the fact that] . . . an innovative firm had to charge much more than an imitating one’ (Lall 1983b, p. 13). Secondly, the palpable ideological shift in development thinking away from import-substitution to export-oriented development strategy was

6

Multinational enterprises in Asian development

naturally accompanied by a greater receptivity to MNE participation in domestic production. In this policy transition, the role of MNEs in the introduction of new industries or new products to the host country and the tighter linking of the host country to the world trading system began to attract increased policy attention. This ideological shift received further impetus in the 1990s from the demise of central planning and widespread renaissance of the market economy as the dominant socio-institutional system for resource allocation. Thirdly, the virtual disappearance of commercial bank lending to developing countries in the early 1980s compelled many countries to look for alternative sources of new capital. In a context where official development assistance had persistently lagged behind announced aid commitments as well as the investment needs of the recipient countries, liberalization of restrictions on incoming foreign investment turned out to be a natural policy choice. Finally, the passage of time had itself soothed some of the greatest fears about MNEs and both host countries and multinationals have come some way during the past quarter century in accommodating themselves to the existence of one another. In a related new development, labour unions and other non-governmental organizations in home countries have been learning to use the multinationals as levers for the achievement of new goals, such as preventing international pollution, promoting religious freedom and discouraging the use of child labour. These developments have begun to play an important role in reducing the risk of conflict between MNEs and host countries (Vernon 2000).

SCOPE AND PREVIEW The great confrontation extending over the developing world between the national state and the MNEs is certainly in the past. It is now widely accepted that MNEs have the potential to play an important role in economic development, not only by bringing in new capital but also, and more importantly, by referring modern technology, market know-how and modern management practices to developing countries. This broader consensus by no means implies that the governments in host countries have no role in regulating MNE entry to their countries or in regulating the activities of MNE affiliates. It has merely brought about a shift in policy focus from the early confrontational approach to giving fresh attention to tackling the challenges associated with relying on MNEs as a vehicle for achieving developmental objectives through global integration. Although the general role of MNEs in development is well recognized, there

Background and preview

7

is no consensus on appropriate national policies toward MNEs; the extent and form of government regulations relating to both MNE entry and their domestic operations, and the underlying raison d’être, differ enormously among countries. Improved development outcomes from relying on MNEs as development partners are clearly tied to our ability to cast a fresh look at the implications of the increasing role of MNEs in the global economy for domestic economic governance in host countries. Over the past two decades, there has been a boom in academic research analysing the economics of MNEs, but very few attempts have been made to address key issues in this policy debate. This book, which examines selected issues relating to the MNE–development interface with emphasis on the experiences of developing countries in Asia, is an attempt to fill this knowledge gap. Developing Asia provides an excellent laboratory for studying the selected (and related) issues, given the variety of experiences relating to the nature and extent of MNE involvement and the related policy regimes, both across countries in general and over time within most countries. In each chapter, the Asian experience is examined in the context of the existing literature on the patterns and developmental implications of MNE involvement in the global economy. Following this introductory chapter, Chapter 2 sets the context for the ensuing chapters. It begins with a succinct account of the evolution over the post-war years and the current state of national policies toward MNE participation in developing Asian countries. This is followed by an overview of the comparative performance of the developing Asian countries as hosts to MNEs in a global context, using data on gross FDI inflows as the prime indicator of MNE participation. The chapter also examines the emerging patterns of source-country and industry composition of FDI flows, and makes inferences about prospects for attracting FDI in the context of the contemporary debate on the possible crowding-out effect on other countries in the region of China’s emergence as an attractive location for FDI. Chapter 3 takes a fresh look at the role of MNEs in the expansion of manufacturing exports. The analysis here is motivated by the concern that, given major changes in the investment climate in developing countries and in patterns of international production over the past two decades, evidence from the early experience of the newly industrialized countries (NIEs) in East Asia may send the wrong signals to policy makers in latecomer exporting countries. First a typology of the involvement of MNEs in manufacturing for export is developed, based on the premise that MNEs are not a homogeneous but a finely differentiated instrument of global integration. The typology is then applied to empirical evidence from NIEs and latecomer exporting countries in developing Asia. The evidence suggests that

8

Multinational enterprises in Asian development

the share of MNEs in manufactured exports from all countries has recorded a significant increase from about the mid-1970s and that, contrary to the historically specific experience of Korea and Taiwan (and also Japan), the entry of MNEs is virtually essential for the export success of latecomers. The findings reported in this chapter make a strong case for concomitant liberalization of trade and investment policy regimes in determining gains from export-led industrialization. ‘International product fragmentation’ – the splitting of the production process into discrete activities which are then allocated across countries – has been a key factor in the rapid expansion of MNE involvement in the global economy over the past three decades. Chapter 4 provides fresh estimates of the extent and growing importance of this phenomenon in world manufacturing trade and examines the implications of this phenomenon for global and regional trade patterns, with special emphasis on countries in East Asia, using a new data set culled from the UN trade database. It is found that, while fragmentation-based trade (trade in parts and components) has generally grown faster than total world trade in manufacturing, the degree of dependence of East Asia on this new form of international specialization is proportionately larger than that of North America and Europe. The upshot is that international product fragmentation has made the East Asian growth dynamism increasingly reliant on extra-regional trade, strengthening the case for a global, rather than a regional, approach to trade and investment policymaking. Chapter 5 deals with the current debate on the relative stability of FDI compared with the other forms of capital flows and the role of MNE affiliates in economic adjustment in the context of international financial crises. More specifically, is foreign direct investment more resilient at the onset of an economic crisis and during the subsequent economic collapse in a given host country than other forms of foreign capital inflows? Are affiliates of multinational enterprises in a crisis-hit country better equipped to withstand a crisis and to aid the recovery process by readjusting their investment, production and sales strategies than local firms are? In this chapter, these and related issues are analysed, drawing upon the experiences of East Asian countries in the recent (1997–98) financial crisis. The findings suggest that FDI was indeed a relatively stable source of foreign capital in the crisis context and that MNE affiliates were instrumental in ameliorating the severity of economic collapse and facilitating the recovery process. In Chapter 6, the nexus of real exchange rate and capital inflows is examined, with emphasis on the difference between FDI and other forms of capital flows, through a comparative analysis of the experiences of emerging market economies in Asia and Latin America during the period

Background and preview

9

1985–2000. It is found that the degree of appreciation of the real exchange rate associated with capital inflow is uniformly much smaller in Asian countries than in their counterparts in Latin America, despite the fact that the former experienced far greater foreign capital inflows relative to the size of the economy. The econometric evidence suggests that both the composition of capital flows and the differences in the degree of response of the real exchange rate to capital flows matter in explaining these contrasting experiences. While real exchange rate appreciation is a phenomenon predominantly associated with other (non-FDI) forms of capital inflows, a given level of other capital flows brings about a far greater degree of appreciation of the real exchange rate in Latin America, where the importance of these flows in total capital inflow is also far greater. The effect of trade policy regimes on national economic gains from international production in foreign direct investment receiving countries is an issue of obvious policy relevance and analytical interest, but one on which there has been a dearth of empirical research. Chapter 7 aims to fill this gap by examining the determinants of productivity of international production using a cross section of data on overseas operations of manufacturing affiliates of US MNEs operating in 44 countries. The findings support the proposition that, other things being equal, productivity gains from international production tend to be greater under a more open trade policy regime than under a restrictive regime. There is also evidence of a significant negative effect of a stringent domestic tax regime on efficiency gains from international production. Chapter 8 examines patterns and determinants of overseas R&D expenditure of US-based manufacturing MNEs using a new panel data set over the period 1990–2001. It is found that inter-country differences in the R&D intensity of operation of US MNE affiliates are fundamentally determined by the domestic market size, overall R&D capability and cost of hiring R&D personnel. There is modest statistical support for the hypothesis that geographical distance has a positive impact on the R&D intensity of MNE affiliates. The impact of domestic market orientation of affiliates on R&D propensity varies among countries according to their stage of global economic integration. Intellectual property protection seems to matter largely for mature economies with complementary endowments. There is no evidence to suggest that financial incentives have a significant impact on intercountry differences in R&D intensity when controlled for other relevant variables. Nor is there a statistically significant relationship between the size of the capital stock of MNEs and the R&D intensity of their operation across countries. Overall, our findings serve as a caution against hostcountry governments paying too much attention to turning MNEs’ affiliates into technology creators as part of their foreign direct investment policy.

10

Multinational enterprises in Asian development

Chapter 9 deals with an important, yet hitherto sparsely studied, issue central to the fledgling literature on multinational enterprises based in developing countries, namely whether they bring in more appropriate technology to host developing countries compared with their developed-country counterparts. Using hitherto unexploited firm-level data relating to Sri Lankan manufacturing, the chapter examines the differences in capital intensity of production between affiliates of developing-country and developed-country MNEs and indigenous private firms by applying a methodology which avoids some of the major drawbacks of the few previous studies. In a simple comparison of average capital per worker, affiliates of developed-country MNEs generally do exhibit a higher degree of capital intensity than those of their developing-country counterparts as well as indigenous private firms. However, when controlled for firm attributes other than parentage that can affect capital intensity, it is found that the link between parentage and capital intensity is industry specific. There are marked differences in capital intensity between the two types of foreign affiliates in the textile and wearing apparel industries. However, no such differences could be observed in other industries. All in all, the results support the view that developingcountry MNEs adopt more appropriate technology only in those industries where the range of technological possibilities is wide enough to enable significantly different techniques of production to be utilized. Finally, Chapter 10 examines the role of MNEs and foreign direct investment in the process of economic transition from ‘the plan to market’ through a case study of the Vietnamese experience. It surveys the evolution of FDI policy in Vietnam in the context of overall policy reforms and the current state of the investment climate, and examines the experience of attracting FDI from a comparative regional and global perspective. This is followed by an analysis of the impact of the operations of MNE affiliates on economic performance. The Vietnamese experience provides ample support for the view that both the rate of FDI involvement in the economy and the national developmental gains from FDI depend crucially on the level of economic transition as reflected in the extent of privatization/restructuring of state-owned enterprises, market-based decision making and the creation of a legal and institutional framework for foreign and private domestic investment.

NOTES 1. In line with usual practice in this area of study, the multinational enterprise (MNE) is defined here as an enterprise that owns and controls business ventures in more than two countries (including its home country).

Background and preview

11

2. FDI is ‘the category of international investment that reflects the objective of a resident entity in one economy to obtain lasting interest in an enterprise resident in another country’ (IMF 1993, p. 86, emphasis added). Here the resident entity is the direct investor (the MNE) and the enterprise is the direct investment enterprise (the foreign affiliate). The lasting interest implies the existence of a long-term relationship between the MNE and the affiliate and a significant degree of influence by the former on the management of the latter. It is this lasting interest that distinguishes direct investment from portfolio investment and other forms of international capital flows such as foreign aid and commercial bank lending. 3. Data reported in this paragraph are from the UNCTAD World Investment Report database (www.unctad.org). 4. Unlike ODA, FDI is concentrated in a handful of developing countries. However, in recent years FDI inflows have surpassed ODA in the bulk of developing countries, including the least developed countries (LDCs) (UNCTAD 2005). 5. Limitations of FDI as an indicator of the activity of MNE affiliates in host (investment receiving) countries are discussed in Chapter 2. 6. Lipsey (2004) provides an extensive survey of the related literature. 7. For an extensive survey of this literature, see Helleiner (1989).

2. Foreign direct investment in developing Asia: trends, patterns and prospects The purpose of this chapter is to provide the context for the ensuing chapters. It begins with an overview of the evolution and the current state of national policies toward MNE participation in Asian developing countries.1 This is followed by a survey of the global spread of multinational enterprises (MNEs) as measured by foreign direct investment (FDI), with emphasis on the relative position of developing Asian countries as hosts to MNEs, focusing in turn on trends in total FDI inflows against the backdrop of global trends, and emerging patterns of source-country and industry composition of these flows. FDI is the commonly used and readily available measure of global operations of MNEs. In addition to measurement problems involved in balance of payments records of FDI (to be discussed below), this measure has many defects as an indicator of the activity of MNE affiliates in their host countries (UNCTAD 2001, Lipsey 2004). For instance, the parent company may take a minority ownership position in a foreign subsidiary because of ownership restrictions imposed by the host country or for other strategic reasons and yet control the activities of the latter, given its command over production technology, management practices and marketing channels. The actual activity of a given affiliate may not necessarily be in the same industry as that to which FDI inflows are ‘officially’ related, or even in the same host country. Moreover, given scale economies of operation and differences in capital intensity across various industries and sectors of production, a given level of FDI could well imply vastly differing scales of MNE involvement among countries. Given these limitations, direct performance indicators such as sales, output or employment of MNE affiliates are needed to get a better understanding of their operations in host countries. Unfortunately such indicators are not available for most countries, including a large number of developed countries.

12

Foreign direct investment in developing Asia

13

POLICY CONTEXT The term ‘foreign investment regime’ (or policies toward MNE participation in the national economy) is used here in the broader sense to cover rules governing foreign investment and specific incentives offered to investors. Foreign investment regimes of Asian countries have gone through several changes in the post-war period. As an integral part of a palpable shift away from import-substitution towards export-oriented development strategy, the policy regimes have become more liberal over time, not only in the sense of reduced bureaucratic impediments to the entry of foreign firms but also in that the sectors open to foreign investment have been expanded. From about the late 1980s, if there is a concern about FDI it is more likely to have been about insufficiency, in contrast to earlier concerns about an excessive foreign presence. There are, however, significant differences among Asian countries in terms of the nature of their commitment to FDI promotion and timing of related policy shifts during the precrisis era. Hong Kong Hong Kong is in a class of its own in terms of its long-standing laissez-fare approach to foreign direct investment. Historically, Hong Kong evolved economically as an entrepot for South China and politically as a colony of Great Britain. Consequently, it is unique for its well-established tradition of free trade and investment (Lin and Mok 1985). This policy stance has so far remained intact in spite of Hong Kong becoming a Special Administrative Region of China in 1991. Business licensing does not exist except for public utilities and banking. The Hong Kong firms and foreignowned firms operate side by side without discrimination or control. There are across-the-board low income taxes for all investors. There are no special tax incentives for foreign investors. Korea and Taiwan Korea and Taiwan (like Japan) have historically adopted a more cautious approach to FDI as a means of developing indigenous technological capability. During the import-substitution phase of industrialization during the post-war period up to about the late 1960s, foreign investment was welcomed into the light manufacturing export sector, but the overall policy stance towards foreign investment continued to be passive and highly selective. The first attempt by the Korean government to provide a legal basis for the attraction of foreign capital was made in January 1960, through the

14

Multinational enterprises in Asian development

enactment of the Foreign Capital Inducement Promotion Act. This stipulated equality of treatment and even some tax incentives. In the first Five Year Plan, begun in 1962, the government realized the importance of foreign capital, and thus adopted more concrete measures to encourage FDI inflows, but only if they would not hurt domestically owned firms. FDI approval guidelines and procedures were further streamlined in 1973 with a view to providing local firms with breathing space at the formative stage of entrepreneurial development. The most important change was that increased emphasis was placed on encouraging the setting up of joint ventures rather than wholly foreign-owned firms. Industry-specific guidelines relating to project eligibility, foreign equity share and investment scale were also adopted and specific industries not open to foreign investors were clearly demarcated. These heavy-handed criteria remained the backbone of Korean FDI policy until the late 1970s (Koo 1985). There was a notable move toward FDI liberalization from the early 1980s, following the failure of the Heavy and Chemical Industry Promotion Plan of the 1970s. Investment approval guidelines were considerably liberalized in September 1980, allowing FDI in many new areas, permitting firms to be majority- or wholly-owned by foreign investors in many additional cases and reducing the minimum amount of investment. In December 1989, various performance requirements imposed on foreigninvested firms, such as requirements relating to local content, export and technology transfer, were abolished. In December 1996, when Korea joined the OECD, a new Act on Foreign Direct Investment and Foreign Capital Inducement replaced the Foreign Capital Inducement Act. The new act aimed to provide the setting for a shift in FDI policy away from the historic ‘promotion’ role towards a more flexible ‘facilitating’ role. These initiatives notwithstanding, Korea’s overall stance towards FDI continued to be quite restrictive by the Southeast Asian standards (Kim and Hwang 2000). Following the onset of the financial crisis in 1997, the Korean government departed radically from its traditional closed economy approach to FDI. As part of its crisis management policy, the Korean government embarked on a more active promotion of FDI. In November 1998, as part of the reform program agreed with the IMF, the government enacted the Foreign Investment Promotion Act, with a view to creating an investororiented policy environment. The act provided for streamlining foreign investment approval procedures, expanding investment incentives, and the establishment of an institutional framework for investor relations. There was also full-fledged liberalization in the areas of cross-border mergers and acquisitions and foreign land ownership. In June 1998, the government bound its commitment to liberalization of financial services agreed with the OECD as part of its commitment to the World Trade Organization (WTO),

Foreign direct investment in developing Asia

15

thus contributing to strengthening investor confidence in the new FDI liberalization drive. Under the new Foreign Investment Promotion Act, foreign investment restrictions were eased across a wide range of sectors, including banks, other financial institutions and activities, real estate and land, and stateowned utilities targeted for privatization. The most significant initiatives included permitting foreigners to engage in deep-sea foreign transport; increasing the foreign ownership ceiling on publishing newspapers to 30 per cent and periodicals to 50 per cent; increasing the permitted equity ownership by foreigners of Korean telephone service provision from 33 to 49 per cent; allowing foreign financial institutions to participate in mergers and acquisition of domestic financial institutions on equal principles; permitting foreign financial institutions to establish subsidiaries; allowing foreign participation in merchant banks up to 100 per cent; and abolishing restrictions on foreign ownership of land and real estate properties on the basis of national treatment. In Taiwan, during the first decade of post-colonial development (1951 to 1960), the government discouraged FDI in order to promote a new indigenous industrial entrepreneurial class. In 1952 the government promulgated legislations to encourage investment in productive enterprises by overseas Chinese and Chinese residents in Hong Kong and Macao. These investors were permitted to invest in any industry – real estate and services industries in particular. From about the early 1960s, the policy regime became more receptive to non-Chinese foreign investors. But, as in Korea, the government played a very active role in directing investment into selected sectors/ industries in line with national developmental priorities (Ranis and Schive 1985). In order to engage foreign firms in the export promotion drive in a more liberal policy setting, the first export processing zone (EPZ), in the vicinity of Kaohsiung, Taiwan’s largest harbour, was established in 1966. Two more were added five years later (Schive 1990). However, promoting nationally owned firms, private and public, in both industry and services continued to be the general thrust of the industrial policy of Taiwan until the mid-1980s. Competition from foreign investors was structured in such a way as to strengthen national firms through technology transfer and ‘spillovers’ (Amsden and Chu 2003). The Taiwanese FDI policy has become increasingly liberal as part of market liberalization reforms beginning in 1986. Among other measures, liberalization included relaxing and finally removing foreign equity limits on joint ventures, opening markets to new foreign entrants in virtually all manufacturing industries. However, until recently, the government continued to provide local firms with a head start over foreign firms in key services such as banking and telecommunications (Amsden and Chu 2003).

16

Multinational enterprises in Asian development

Restrictions on foreign investment in services were later selectively lifted as part of the reform process leading to Taiwan becoming a WTO member in 2002. ASEAN Countries Among the member countries of the Association of South East Asian Nations (ASEAN), Singapore has throughout maintained a solid reputation for providing the most favourable FDI regime in the region (and one of the most favourable in the world) (Chia 1985, Hughes and You 1969). The entry of FDI to Singapore is largely unrestricted and there is no need for formal approval of investment projects. There is no specific foreign investment law. Nor are there restrictions on repatriation of funds or regulations concerning foreign ownership. Remaining restrictions on FDI in telecommunications and legal services were lifted in 2000 as part of liberalization commitments under the World Trade Organization. The main form of encouragement for foreign and domestic investors alike remains the ‘pioneer status’ introduced by the Pioneer Industries Ordinance Act of 1959. Pioneer status carries tax holidays of five to ten years, which may be extended depending upon possible additional income generation due to the expansion of activities. Special benefits are available for firms which have a high share in manufacturing exports and/or bring in new technologies. The liberal FDI regime in Singapore is an integral part of an overall economic policy regime which has maintained an impressive track record in meeting other prerequisites of a good investment environment (such as high-quality infrastructure and financial services, market-friendly labour market practices, and transparent/expedient legal procedures) (Huff 1994). In the 1960s and 1970s, policy towards FDI in the other four original ASEAN member countries (Indonesia, Malaysia, Thailand and the Philippines) remained ambivalent, alternating between a national distrust of foreign firms and the hope that new foreign investment could provide the technology and capital for rapid industrialization (Lindblad 1998, Chapter 5). FDI policy was thus characterized by a mix of incentives and restrictions, with the balance between the two varying among countries and over time depending on the strength of prevailing anti-FDI sentiment (which was particularly strong in the Philippines and Indonesia). The investment regime of Malaysia, which remained relatively liberal by regional standards in the 1960s, became much more restrictive following the introduction of bumiputra ownership and employment requirements under the New Economic Policy (NEP) (Athukorala and Menon 1996). The other countries gradually began to introduce pro-active FDI policies,

Foreign direct investment in developing Asia

17

including offering various fiscal incentives. But foreign participation remained generally limited to minority ownership. Only pioneer status would qualify for full foreign ownership during the initial stage of operations. Unlike in India, however, there were no restrictions on procurement of foreign technology and the repatriation of profit and capital. From the late 1980s, as an integral part of a palpable shift away from import-substitution towards export-oriented development strategy, FDI policy regimes in these countries have become increasingly liberal. By the mid-1990s, full foreign ownership of firms producing for export was a common feature and ownership restrictions had become increasingly liberal even for firms involved in domestic-market-oriented production. The requirement for registering with the Board of Investment had been abolished in Thailand and it was applied only to firms with 40 per cent of foreign equity in the Philippines. In other countries formal approval remained in force, but the approval procedures had become much simpler and less time consuming. All countries had moved to the so-called ‘negative list’ approach to investment approval, which involves governments explicitly listing those activities closed to FDI, with the implication that any activity not so listed is open. However, the governments continued to restrict foreign participation in such services as media, real estate, energy and utilities. These significant moves towards providing greater opportunities for global integration through FDI occurred in a general economic setting which became increasingly conducive to private sector participation in the growth and development process. The investment environment had become increasingly favourable for a number of reasons, including political stability and policy continuity, an impressive record in maintaining macroeconomic stability, and a proven track record in meeting infrastructure requirements for rapid growth. Beneath these general trends, there were of course considerable differences among countries. By the mid-1990s, Malaysia was generally considered to have the most favourable foreign investment climate in the region (after Singapore), followed by Thailand, Indonesia and the Philippines in that order (Lindblad 1998, Hill 2004). Following the onset of the financial crisis in 1997, as part of the crisis management policy, the governments of the four crisis-affected ASEAN countries (Indonesia, Malaysia, the Philippines and Thailand) embarked on a more active promotion of FDI. In Thailand, nearly all services and manufacturing sectors were opened to FDI and restrictions on FDI in the real estate and financial sectors were considerably relaxed as part of the policies that Thailand agreed to implement in the context of its request for financial support from the IMF (Kohpaiboon 2006). In Indonesia, as part of the reform package agreed upon with the IMF, the government committed itself to promote foreign direct and equity

18

Multinational enterprises in Asian development

investment and to implement a number of measures to increase these flows. In late 1999, the Indonesian Government proposed to reorganize the Investment Board (BKPM) into a new institution under the Coordinating Minister for Economic Affairs, which will focus on investment promotion rather than regulation activities. Implementation of these proposals has been hampered by political turmoil. Measures implemented include significantly narrowing the list of sectors that are closed to foreign investment (in July 1998) and lifting restrictions on foreign investment in wholesale trade. Malaysia has continued to give priority to promoting FDI, despite its radical policy shift in September 1998 (Athukorala 2002a). The newly introduced capital controls were confined to short-term capital flows and aimed at making it harder for short-term portfolio investors to sell their shares and keep the proceeds, and for offshore hedge funds to drive down the currency. With the exception of limits on foreign exchange for foreign travel by Malaysian citizens, there was no retreat from the country’s long-standing commitment to an open trade and investment policy. No new direct controls were imposed on import and export trade. Profit remittances and repatriation of capital by foreign investors continued to be free of control. Immediately following the imposition of capital controls, the Central Bank of Malaysia (Bank Negara Malaysia, BNM) did experiment with new regulatory procedures in this area. But these were swiftly removed in response to protest by these firms. Moreover, some new measures were introduced to further encourage FDI participation in the economy. These included allowing 100 per cent foreign ownership of new investment made before 31 December 2000 in domestic manufacturing regardless of the degree of export orientation; increasing the foreign ownership share in telecommunication projects from 30 to 69 per cent (on condition that the ownership share is brought down to 49 per cent after five years) and in stock-broking companies and the insurance sector from a previous uniform level of 30 per cent to 49 and 51 per cent respectively; and relaxing restrictions on foreign investment in landed property to allow foreigners to purchase all types of properties above RM 250 000 in new projects or projects which are less than 50 per cent completed. In the Philippines, the crisis itself has not resulted in a significant shift in the country’s policy towards FDI. However, the emphasis on the promotion of export-oriented foreign investment, which started in earnest in the late 1980s, seems to have received further impetus following the crisis (Hill 2003). Among the four new ASEAN member countries, in Vietnam significant opening of the economy to FDI was part of the ‘renovation’ (doi moi) reforms initiated in 1986 (Riedel and Comer 1997, Freeman 2004). With a slow and hesitant start in the late 1980s, significant reforms were under-

Foreign direct investment in developing Asia

19

taken in the first half of the 1990s. The reform process lost momentum during 1996–98 partly owing to the East Asian crisis of 1997–98, but partly (perhaps even more so) owing to domestic policy ambivalence and complacency resulting from the success of the initial reforms. There has however been a renewed emphasis on completing the unfinished reform agenda over the past three years.2 In Cambodia the Kampuchea People’s Revolutionary Party (KPRP) government embarked on a market-oriented reform process in 1985. As part of these reforms, the government promulgated a liberal foreign investment code in July 1989 and a National Investment Council was set up in 1991 with the task of reviewing all foreign investment applications. The outcome of these reforms was somewhat lacklustre, however, and perhaps unsurprising given the continuing warfare between KPRP forces and the Khmer Rouge. As an outcome of the UN-led peace process, elections were held in July 1993 and a multi-party democratic government was established in September 1993. The new government set up the Cambodian Investment Board (CIB) under the Council for Development of Cambodia (CDC) to be the ‘one-stop’ service organization responsible for approving foreign investment applications. The new Laws and Regulations on Investment in the Kingdom of Cambodia, passed by the National Assembly on 4 August 1994, set out rules and regulations governing FDI and offered an incentive package which was very generous compared with those in other countries in the region. The foreign investment regime in Cambodia underwent an overhaul in 2003. The revised Law on Investment came into force on 27 September 2005, and represented a major attempt to equalize incentives for foreign and local investors, to achieve greater transparency in incentives provided, and to minimize distortions and delays arising from policy maker discretion. As part of the new reforms, a fast-track procedure has been introduced with the aim of approving investment applications within a 14-day period under the ‘one-stop’ service at the CIB. Seven working groups, which involve both private and public sector participation, have been set up in key sectors to work in collaboration with the CIB to facilitate speedy investment approval, monitoring and promotion. An investor forum, headed by the prime minister, is to be held twice a year as part of the new investment regime. In December 2005, a Sub-Decree was passed to provide the legal framework for setting up special economic zones (SEZs) (which may include general industrial zones and/or export processing zones (EPZs)). The transition to a market-oriented economy in Lao PDR began in 1985 with the announcement of the New Economic Mechanism, a major program of economic reforms. As part of the reform program a Foreign Investment Code was passed in July 1988 and the Foreign Investment

20

Multinational enterprises in Asian development

Management Committee (FIMC) was set up under the direct purview of the prime minister to act as the apex agency that approves, monitors and promotes FDI. At the initial stage, the prime objective of FDI policy in Lao PDR was to engage foreign investor participation in restructuring of state-owned enterprises (SOEs). The Foreign Investment Code was supplanted in July 1994 by the Law on Promotion and Management of Foreign Investment, which was in turn substantially revised in October 2004. Foreign investment is now permitted in joint ventures or fully foreign-owned projects in all business sectors, except in mining and energy projects in which the government contributes to share capital or retains the right to buy a preagreed share of equity. The structure of tax incentives for foreign investors has been designed to take into account the country’s peculiar geography (mountainous terrain) and uneven quality of infrastructure in different parts of the country. For large projects in the mining and energy sectors (which by their very nature tend to be located in remote areas), specific taxation arrangements are negotiated on a case-by-case basis. Finally, in Myanmar the potential for attracting FDI investment has remained subdued because of continuing political problems (Athukorala et al. 2000). China Perhaps the most significant element in economic reforms in China since 1979 has been the opening up of the economy to foreign direct investment. The 1979 Joint Venture Law permitted FDI for the first time since 1949. This landmark legislation codified the right of foreign firms to invest in China and to cooperate with Chinese enterprises in various ways. It defined the nature of joint ventures that would be allowed to invest in China and set the stage for the process of establishing an institutional and administrative framework for monitoring FDI. Several laws aimed at specific issues relating to foreign investment were introduced between 1979 and 1986 (Lardy 2002, Huang 2003, Pomfret 1991). The centrepiece of China’s formal FDI promotion policy has been the special economic zones (SEZs). Chinese authorities chose SEZs as a compromise solution to the problem of introducing foreign investors and their capital participation into China while limiting the political repercussions of opening up. The original inspiration for the SEZs came from the EPZs of East Asia, but they have special Chinese features. The SEZs are much larger than EPZs. Unlike EPZs elsewhere in the region which are typically run by management companies or boards which come under the purview of the central government, SEZs are government units in their own right. In terms of objectives, the SEZs were not just to be vehicles for expanding exports. They were also assigned a central role in the reform process as ‘windows

Foreign direct investment in developing Asia

21

and bridges’ to the outside world, in both directions, and also as ‘economic laboratories’ in which economic policy experiments could be tried out in a geographically restricted area. A striking feature of the FDI regime as it evolved in the 1980s was its dualistic nature. The export-oriented (EO) or export-promotion (EP) regime was generally restricted to foreign firms. During the ensuing years, some of the special provisions to attract FDI which until then had been available only in the SEZs were made much more widely available. At the same time sectors such as retailing, power generation and port development were gradually opened up for foreign investors. Since 1992 significant domestic market access has been given to foreign investors, particularly those who could offer advanced technology. Foreign capital participation in property development (including residential housing) was liberalized in the early 1980s. South Asia During the first three decades of the post-war era, South Asia remained perhaps the most inward-oriented group of countries in the world outside the communist bloc. These countries maintained strict controls on foreign investment in import-substituting manufacturing and other sectors while giving special incentives for export-oriented investment. Technology licensing and joint ownership with local firms were the most preferred forms of foreign investment. Naturally, the latter policies were not to bring about any tangible result, given the anti-export bias in the overall incentive structure. Sri Lanka took the lead in breaking away from the protectionist past, by embarking on a decisive process of economic opening in 1977. Following hesitant and sporadic attempts to dismantle trade barriers in the 1970s, other countries embarked on significant liberalization reforms from the late 1980s. While there are vast inter-country differences in terms of the degree of liberalization achieved during the ensuing years, by the mid-1990s all five countries seemed to have moved into a seemingly irreversible process of economic liberalization and greater receptivity to foreign investment. During the first four decades following independence in 1947, India’s policy with respect to FDI (and other types of foreign capital) remained highly restrictive, as part of a stringent import-substitution industrialization strategy.3 The government did not rule out new foreign investment, but wanted it on its own terms. With a view to minimizing foreign exchange outlay relating to technology acquisition, as far as possible technologies were to be acquired through licensing rather than through FDI. All foreign investment applications were considered on a case-by-case basis and that too within a normal ceiling of 40 per cent of total equity investment. Major

22

Multinational enterprises in Asian development

commercial banks and foreign oil companies were nationalized in the 1970s. The Foreign Exchange Regulation Act of 1973 required firms to dilute their foreign equity holdings to 40 per cent if they wanted to be treated as Indian companies.4 There was severe curtailment of intellectual property rights. For instance, product patents were abolished in industries such as pharmaceuticals and chemicals and the duration of process patents was drastically reduced. The regulatory mechanism governing the entry of MNEs was characterized by an explicit preference for technical collaboration agreements as opposed to FDI; a policy stance dictated by the desire to achieve the (conflicting) twin objectives of minimizing foreign control on business operations and gaining access to foreign technology. In the mid-1960s, India set up two free trade zones (FTZs) to promote export-oriented foreign investment alongside the highly restrictive trade and investment policy regime. But these FTZs never took off for several reasons, such as their relatively limited scale, the government’s general ambivalence about attracting FDI, the unclear and changing packages attached to the zones, and the power of the central government in the regulation of the zones (Bajpai and Sachs 2000, Kumar 1989). In the 1980s, there was some softening of the FDI approval procedure for export-oriented activities and some high-tech industries, but the overall policy stance remained one of the most stringent in the developing world. Foreign investment policy was substantially liberalized at an early stage of the 1991 economic reforms and the process was extended further in the subsequent years (Balasubramanyam and Mhambare 2003). Foreign ownership up to 51 per cent is now permitted in a wide range of industries. Foreign ownership up to 100 per cent is permitted on a case-by-case basis in some designated (priority) areas (pharmaceuticals, airports, suburban development, hotels and tourism, courier services and mass rapid transport systems) and up to 74 per cent in the telecommunication sector. Technology policy has been reformed to give greater recognition to intellectual property rights. Procedures for obtaining permission were also greatly simplified, requiring only registering with the Reserve Bank of India (RBI). Firms are now free to negotiate the terms of technology transfer on the basis of their own commercial judgment and without the need for government approval for hiring of foreign technicians and foreign testing of indigenously developed technology. Despite recent reforms, India’s foreign investment regime still reflects the tension between the traditional aversion to foreign investment and the current recognition of its importance to economic development.5 For example, FDI is still not permitted in pure retailing; global retailers can only participate in India’s retail sector through wholesale trade or by operating retail outlets through local franchises. In apparel and other light

Foreign direct investment in developing Asia

23

consumer-good producing industries, which are important in export expansion and job creation at the current stage of economic development of the country, FDI is limited to 24 per cent of total equity. Restrictions on foreign ownership of land limit the entry of foreign builders and developers into the construction sector (see World Bank 2003b, pp. 55–66 for details). Projects with 51 per cent or more foreign ownership still require a long procedure of government approval. There are also many unresolved problems relating to the overall investment climate (World Bank 2003b). Tariff protection in India is still substantially higher than in most other developing countries, and this continues to block India’s attractiveness as an export platform for labourintensive manufacturing products. While the ‘Licence Raj’ (the infamous industrial licensing policy) has been largely eliminated at the centre, it still survives at the state level, along with a pervasive ‘Inspector Raj’. Private investors require a large number of permissions (for electricity and water supply connections, water supply clearance and so on) from state governments to start business and they also have to interact with the state bureaucracy in the course of day-to-day business.6 Stringent labour laws and high corporate tax rates,7 restrictive labour market practices and a weak bankruptcy framework are other prominent issues. Pakistan has a checkered history of trade liberalization and FDI promotion. Following some trade liberalization attempts in the 1960s, Pakistan qualified for Article VIII status at the IMF in 1970. Even by the mid-1980s there was still a long way to go in lifting QRs and reducing tariffs. From the mid-1980s, controls on foreign investment in manufacturing have diminished sharply, those for the service sector less so. In spite of various bureaucratic controls, the government attitude throughout the 1950s and 1960s was favourable to private investment (Bose 1983, Guisinger and Scully 1991). The FDI regime was more liberal, although there was greater emphasis on joint ventures with minority foreign ownership and technology licensing than on FDI in fully foreignowned ventures. However, supremacy of the state and socialist ideology under a socialist government dominated policy in the 1970s. As a result, a large-scale program of nationalization of key industrial units and widespread control of domestic and foreign trade were instituted. The dismal economic outcome of the interventionist policies eventually paved the way for market-oriented reform. Reforms started slowly in the early 1980s as part of a widespread reform package in conformity with the World Bank conditionality. Removal of restrictions on foreign investment was a major element of the reform program. Full foreign ownership of firms, with full freedom for remittance of profit and investment proceeds, is now allowed in almost all sectors of the economy.

24

Multinational enterprises in Asian development

Following independence in 1971, the Bangladesh government adopted a state-led import-substitution development strategy, which was far more interventionist than that of the united Pakistan. The new government nationalized a large number of industrial enterprises owned by Pakistani entrepreneurs as well as all industrial enterprises with fixed assests exceeding a certain threshold level. The scope of the private sector was limited to small and cottage industry, and foreign investment was allowed only in collaboration with the public sector with minority equity participation. However, existing foreign investments (excluding those belonging to Pakistan) were spared from the sweeping nationalization drive. The socialist-oriented industrial policy of 1973 assigned a very minor role for the private sector, with some investment ceiling on new investment. The next two years were a period of extreme political and economic turmoil. With the change of government in 1975, the strategy of publicsector-led industrialization was abandoned. The new policy reforms, introduced with a view to giving more room to the private sector, included further raising of the investment ceiling and its elimination in 1978, reducing the number of industries reserved for the public sector, streamlining investment approval procedures, and amendment of the constitution to allow denationalization. Further significant changes were introduced in 1980 with the enactment of two pieces of legislation. The Foreign Investment (Promotion and Protection) Act of 1980 provided for the protection and equitable treatment of foreign investment, guarantees against expropriation or nationalization without compensation, and repatriation of invested capital and profits. The Bangladesh Export Processing Zone Authority Act, also passed in 1980, provided for the setting up of three export processing zones during the ensuing decade. The subsequent reforms in the early 1990s included allowing 100 per cent foreign ownership in all foreign investment projects and extending EPZ privileges to all export-oriented projects regardless of their location (Rahman and Bakht 1997, Ali 1999). In Sri Lanka, foreign investment policy remained extremely liberal until the mid-1960s, permitting many MNEs to set up affiliates within Sri Lanka to undertake the domestic production of items previously supplied from their overseas production centres. However, as the import-substitution industrialization strategy was reaching a crisis point by the mid-1960s, the view (which was widely held among development economists at the time) that ‘import-substituting MNEs worsen countries’ balance of payments’ began to dominate Sri Lanka’s policy towards FDI. This view resulted in a dualistic foreign investment policy characterized by stringent restrictions on import-substitution projects and preferential treatment for export-oriented ventures. From about the late 1960s, there was some policy emphasis on the

Foreign direct investment in developing Asia

25

promotion of export-oriented FDI through selective incentives. A White Paper on foreign investment issued in 1966 introduced various tax concessions for export-oriented foreign ventures and relaxed foreign exchange restrictions on the remittance of dividends, interest and profit originating in such ventures. The government’s commitment to the promotion of exportoriented FDI was reaffirmed and further production and tax incentives were introduced by the Five-Year Plan, 1972–77 (Government of Sri Lanka 1972). However, this policy shift in favour of EOFDI occurred in an overall policy and political context which was highly unfavourable to private sector activities in general and to export production in particular. Reflecting the cumulative impact of stringent trade controls, high export taxes and the overvalued exchange rate, the overall incentive structure of the economy was characterized by a significant ‘anti-export bias’ throughout this period (Athukorala and Jayasuriya 1994). As a reaction to the dismal economic outcome of its inward-looking policy, Sri Lanka embarked on an extensive economic liberalization process in 1977, becoming the first country in the South Asian region to do so. Liberalization of the foreign direct investment regime, with a major focus on attracting export-oriented FDI, was a key element of the reform program.8 During the post-war period until the mid-1980s, Nepal continued to pursue import-substitution development strategy in the context of a highly protectionist trade and investment regime. During this period, FDI was allowed only in some selected industries. There were also stringent limits on the share of foreign ownership. There has been a clear policy shift in favour of market-oriented outward-looking development strategy, starting with the Structural Adjustment Policy (SAP) package of 1985. As part of these reforms, Nepal has made substantial changes in its FDI policies as an effective means of promoting private-sector-led growth. The Industrial Policy and Industrial Enterprise Act promulgated in 1987 provided a legal framework to facilitate FDI in medium- and large-scale ventures in all industries with the exception of environment and defence related activities. A new Foreign Investment Promotion Division was created at the Ministry of Industry to act as the central body for the approval and monitoring of FDI projects. FDI was not allowed in small-scale industries, while foreign equity ownership of up to 50 per cent was permitted in medium-sized industries. In large firms with at least 90 per cent of export sales, 100 per cent foreign ownership was allowed. A five-year tax holiday is applicable to export-oriented firms. Subsequent changes to FDI law include expanding the scope of FDI in all industries except defence, cigarettes and alcohol; streamlining the approval procedure; entering into investment protection and agreements to avoid double taxation with major investing countries; reduction of corporate tax rate (in 1997) on domestic-market-oriented FDI projects and FDI

26

Multinational enterprises in Asian development

projects in services to 20 per cent and giving export-oriented firms the option of corporate tax at the rate of 0.5 per cent of export value (FOB) or 8 per cent of profit; and reduction (in 1996) of the minimum fixed asset limit on new firms to Rs 500 million. A new 5 per cent tax on profit remittances by foreign firms was introduced in the 1999/2000 Budget because of balance of payments exigencies (Athukorala and Sharma 2005).

TRENDS This section is based on data compiled from UNCTAD’s World Investment Report, which is now the standard data source in this area. Before proceeding to analyse the data, a cautionary note about the data quality is in order. According to the standard definition, FDI consists of three components. These are equity capital: shares owned by the foreign direct investor (MNE) in its affiliates firms; retained earnings: the MNE’s share (in proportion to direct equity participation) of earnings not distributed as dividends by affiliates or earnings not remitted to the parent company (such retained profits are reinvested by affiliates); and intra-company loans or intra-company debt transactions, referring to short- or long-term borrowing and lending of funds between the parent company and affiliated enterprises. Not all countries record every component of FDI flows. For most countries, data series on FDI capture only equity capital and intercompany debt; the majority of countries do not report data on the second component.9 There is evidence that the ‘retained earning’ component of FDI is positively related with the age of operation of firms in a given country, and that US MNEs have a general tendency to rely more on retained earnings for investment expansion than MNEs from other countries do (Lipsey 2000). Thus this problem of data coverage can lead to a considerable underestimation of the actual magnitude of FDI in a given host country, depending on the history of MNE involvement and the source-country profile of FDI. Even for the components on which data are available, the quality of data varies considerably across countries. For instance, some countries (such as China and Hong Kong) do not make an adequate distinction between portfolio investment and foreign direct investment. For these reasons, comparison of data among countries, and even over time for a given country, should be made with caution. Moreover, the data coverage tends to vary over time in a given country because of changes made to the data recording system.10 Table 2.1 provides data on gross FDI (in current US dollars) to developing Asian countries in a comparative global context. In Table 2.2 the same data are presented as a percentage of gross domestic fixed capital formation

27

128 541 105 019

6

23 515

7934

2854 370 12 197

308 0 156 114 1 34 7082

2620 2978

790

73 406 49 464

12

23 930

8686

1838 6458 6763

211 5 67 80 0 57 2693

772 1610

192

World Developed countries2 European transition economies3 Developing economies2 America and Caribbean Africa Middle East Developing Asia4 South Asia5 Bangladesh India Pakistan Nepal Sri Lanka East Asia and China6 China China, Hong Kong SAR China, Taiwan Province of 1154

16 028 4588

881 7 414 323 5 119 22 542

4322 2366 38 670

18 273

64 024

1572

205 098 139 502

1985–891 1990–941

1980–841 1996

1997

1998

1999

2000

2001

2002

2003

2004

6308

12 101

10 647

1559

37 521 6213

2962 2 2151 719 8 65 46 545

5587 3382 77 717

30 167

76 260

1864

41 726 10 460

3636 14 2525 918 19 133 56 069

2248

45 257 11 368

4962 139 3619 713 23 433 61 823

5790 10 849 5031 3371 90 006 100 925

50 246

222

45 463 14 765

3522 190 2633 507 12 150 65 503

9049 3262 91 459

82 540

117 544 151 746 191 764 186 626

4803

2926

40 319 24 578

3140 180 2168 530 4 201 77 265

11 886 1888 109 695

108 640

232 507

10 492

11 775

12 821

24 106

34 897

89 130

4928

40 715 61 924

3147 280 2319 308 0 173 116 162

4109

46 878 23 777

4113 79 3403 383 21 172 78 611

9627 20 027 3770 7100 141 955 101 483

97 523

1445

52 743 9682

4606 52 3449 823 6 197 67 205

12 994 5691 86 318

50 492

67 526

453

53 505 13 624

1898

60 630 34 035

5463 7153 268 460 4269 5335 534 952 15 10 229 233 71 928 104 890

18 005 18 090 6522 9840 94 755 137 705

46 908

253 179 217 845 155 528 166 337 233 227

9067

341 086 392 922 487 878 701 124 1 092 052 1 396 539 825 925 716 128 632 599 648 146 218 738 234 868 284 013 503 851 849 052 1 134 293 596 305 547 778 442 157 380 022

1995

Foreign direct investment inflows, 1980–2004 (US$ million)

Economy

Table 2.1

28

Share in global inflows (%) Developed countries Transition economies3 Developing economies America Africa Middle East Developing Asia

0.0

18.3

6.2 2.2 0.3 9.5

32.6

11.8 2.5 8.8 9.2

799 11 389 2427 732 3

1413 0 49 1733 360 14

0.0

0 442 1

0 293 0

81.7

4806 2

3860 3

67.4

568

1985–891

116

1980–841

Economy

Korea, Republic of ASEAN7 Brunei Darussalam Cambodia Indonesia Lao People’s Dem. Rep. Malaysia Myanmar Philippines Singapore Thailand Vietnam

(continued)

Table 2.1

8.9 2.1 1.2 18.9

31.2

0.8

68.0

4423 167 942 5181 1990 780

31 1691 23

15 235 7

756

1990–941

8.8 1.6 1.0 22.8

34.5

1.4

64.1

5815 318 1459 11 591 2070 1780

151 4346 88

28 201 583

1250

1995

12.8 1.5 1.3 22.9

38.6

1.6

59.8

7297 581 1520 9493 2338 1803

294 6194 128

30 301 654

2012

1996

15.6 2.2 0.7 20.7

39.3

2.5

58.2

6323 879 1249 13 586 3882 2587

168 4678 86

34 141 702

2640

1997

11.8 1.3 0.5 13.0

26.6

1.5

71.9

2714 684 1752 7472 7492 1700

243 241 45

22 434 573

5040

1998

9.9 1.1 0.2 10.0

21.3

1.0

77.7

3895 304 1725 16 624 6091 1484

232 1865 52

29 289 748

9448

1999

7.0 0.7 0.3 10.2

18.1

0.6

81.2

3788 208 1345 16 485 3350 1289

149 4550 34

22 646 549

8591

2000

10.8 2.4 0.9 12.3

26.4

1.4

72.2

554 192 899 14 122 3886 1300

149 2978 24

18 674 526

3692

2001

7.1 1.8 0.8 12.1

21.7

1.8

76.5

3203 191 1792 5822 947 1200

145 145 25

14 506 1035

2975

2002

7.4 2.8 1.0 15.0

26.3

3.8

69.9

2473 291 347 9331 1952 1450

84 597 19

17 360 2009

3785

2003

10.4 2.8 1.5 21.2

36.0

5.4

58.6

4624 556 469 16 060 1064 1610

131 1023 17

25 658 103

7687

2004

29

3.2 16.1

11.1 20.4

33.7 12.1 1.6 51.9 1.3 30.1

2.0 3.7

1.1 5.3

36.3 7.7 27.0 28.3 0.9 11.3

0.2 5.5

0.3 3.7

25.0 23.8

28.5 6.8 3.7 60.4 1.4 35.2

7.8 7.4

0.4 11.0

31.9 24.0

25.7 4.8 2.9 66.1 2.5 39.6

11.0 8.3

0.9 13.6

27.5 20.0

33.1 3.8 3.3 59.3 2.4 36.9

10.6 7.7

0.9 14.3

23.6 17.8

39.8 5.7 1.8 52.6 2.6 32.2

9.3 7.0

1.0 12.7

24.4 12.0

44.2 4.8 1.7 49.0 1.9 35.1

6.5 3.2

0.5 9.3

17.3 12.6

46.7 5.1 0.8 47.2 1.4 33.2

3.7 2.7

0.3 7.1

16.1 8.9

38.5 3.8 1.5 56.1 1.2 45.9

2.9 1.6

0.2 8.3

21.5 8.6

40.9 9.2 3.3 46.6 1.9 36.1

5.7 2.3

0.5 9.5

33.9 9.3

32.5 8.4 3.7 55.5 3.0 43.2

7.4 2.0

0.6 9.4

32.2 10.4

28.2 10.8 3.9 57.0 3.3 43.2

8.5 2.7

0.9 11.4

26.0 11.0

29.0 7.8 4.2 59.0 3.1 45.0

9.4 4.0

1.1 16.2

Source: Compiled from UNCTAD, World Investment Report (various years).

Notes: 1. Annual average. 2. Based on the United Nations standard classification (under which all Asian countries other than Japan are classified as developing countries). 3. Countries in Central and Eastern Europe. 4. Countries in Asian-Pacific region other than Japan (South Asia + member countries of ASEAN + East Asia and China). 5. Includes Afghanistan, Bhutan, Maldives and Mongolia. 6. Includes Macau. 7. Member countries of the Association of South East Asian Nations.

Share in flows to developing countries (%) America Africa West Asia Developing Asia South Asia East Asia and China China ASEAN

South Asia East Asia and China China ASEAN

30

World Developed economies2 European transition economies3 Developing economies2 America Africa Middle East Developing Asia4 South Asia Bangladesh India Pakistan Nepal Sri Lanka East Asia and China China China, Hong Kong SAR

Economy 3.47 3.55 0.01 3.46 4.83 2.94 0.44 3.68 0.39 0.00 0.27 1.69 0.17 2.34 3.83 2.50 22.87

2.32 2.15

0.01

3.01

3.90 1.50 4.12 2.29 0.34 0.11 0.14 1.21 0.02 3.40 1.65

0.72 14.57

8.84 15.16

6.84 4.84 1.94 6.32 1.02 0.11 0.62 3.31 0.72 4.92 6.03

5.83

3.65

3.92 3.41

15.43 14.38

9.14 5.45 2.63 8.95 2.61 0.02 2.41 6.15 0.86 1.96 8.75

8.21

4.65

5.34 4.50

1980–841 1985–891 1990–94 1995

14.87 21.37

14.16 5.59 3.49 9.46 3.23 0.15 2.87 8.31 1.94 3.67 9.60

9.76

5.22

6.00 4.82

1996

1998

1999

24.75 9.26 2.61 15.53 2.46 2.51 2.27 3.38 0.05 4.18 18.08

16.25

11.40

20.25 21.57

2000

14.92 13.36 11.32 10.33 19.48 29.42 58.61 138.89

18.29 19.37 30.07 9.96 7.70 11.10 2.18 2.03 1.22 10.75 11.56 13.22 4.29 3.06 2.55 1.41 1.83 1.62 3.98 2.90 2.21 7.24 5.57 5.80 2.20 1.21 0.45 11.05 3.69 4.97 10.39 11.96 13.40

11.85 12.52 16.12

10.38 12.46 15.77

7.48 10.87 16.35 5.93 10.34 16.41

1997

9.82

11.34

10.75 11.03

2002

10.54 55.66

10.35 25.81

24.04 15.61 19.11 12.57 5.31 4.09 11.01 8.52 3.20 3.17 0.70 0.43 3.23 2.92 5.16 7.92 2.00 0.57 5.15 5.73 11.99 9.28

14.20

11.83

12.41 11.87

2001

9.13 7.44

2004

9.17 9.07 32.15 71.12

10.29 10.52 11.28 7.34 4.51 6.54 11.70 21.26 3.31 3.80 2.08 3.34 3.21 3.58 3.67 4.66 1.32 0.83 6.49 6.43 8.81 11.39

10.59 14.98

16.85 19.27

9.20 8.78

2003

Table 2.2 Foreign direct investment inflows as a percentage of gross domestic fixed capital formation, 1980–2004

31

3.51 1.27 7.04 0.40 0.09 2.00 0.89 8.67 0.61 5.75 31.71 3.87 0.26

1.22

0.41

5.27 2.48 0.08 1.20 0.04

11.92 0.01 0.38 20.74 2.72 1.90

0.70

2.36 1.05

2.96 1.58

3.40 5.34

0.35 8.37

4.44 6.56

6.79 3.20

7.79 2.33

2.89

Source:

2.68

0.92 4.92

3.89

20.33 6.89 7.71 30.43 4.68 33.47

15.02 10.24 8.87 41.30 3.00 33.77

17.03 15.98 7.83 27.11 3.13 27.78

14.64 24.88 6.21 36.79 7.62 35.10

14.02 21.09 12.71 24.27 29.93 22.91

22.48 7.15 11.87 59.82 23.85 20.13

16.41 3.82 8.37 60.09 12.44 14.99

2.53 2.92 6.36 56.42 14.65 13.87

14.53 2.96 11.95 26.02 3.25 11.59

Compiled from UNCTAD, World Investment Report (various years).

10.79 19.41 4.59 8.94 2.15 2.70 39.92 65.95 6.15 3.07 12.67 12.73

11.03 12.85 12.08 15.20 17.95 23.20 16.08 13.86 10.31 11.41 15.59 1.73 118.01 129.60 139.86 158.68 184.79 132.56 132.17 252.04 474.22 23.63 11.97 34.55 72.35 38.89 65.69 48.74 29.14 27.52 20.96 9.51 11.67 4.35 7.56 9.20 7.66 0.99 6.61 14.26 9.99 0.43 1.56 2.35 8.11 19.13 23.55 17.97 13.94 14.04 7.88 5.47 5.35 3.89 3.17

0.65

2.51

Notes: 1. Annual average. 2. Based on the United Nations standard classification. 3. Central and Eastern Europe. 4. Countries in the Asian-Pacific region other than Japan (South Asia
member countries of ASEAN
East Asia and China).

China, Taiwan Province of Korea, Republic of ASEAN Brunei Darussalam Cambodia Indonesia Lao People’s Dem. Rep. Malaysia Myanmar Philippines Singapore Thailand Vietnam

32

Multinational enterprises in Asian development 300 Developing countries Developing Asia Developing Asia excl. China

FDI inflow, US$ bn

250

200

150

100

50

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

1984

1983

1982

1981

1980

0

Figure 2.1 FDI inflows to developing countries and developing Asia, 1980–2004 (US$ billion) (GDFCF) of each country/region in order to gain a rough idea of the relative economic significance of FDI inflows. Total FDI flows to the regions are compared with total world flows and flows to developing countries in Figure 2.1. Total FDI flows to developing Asia increased sharply from an average annual level of US$5.4 billion during 1980–84 to US$112.4 billion during 2000–04. The share of the region in total FDI flows to developing countries almost doubled (from 28.2 per cent to 54.8 per cent) between these two time points (Table 2.1). As a share of total global flows, the increase was from 9.2 per cent to 13.3 per cent. FDI inflows as a share of GDFCF has been significantly higher than the comparable figure for all developing countries throughout the period 1980–1998, followed by a minor reversal in the pattern during the five years following the Asian financial crisis (Table 2.2). The average FDI–GDFCF ratio for developing Asia for the entire period 1980–2004 was 8.2 per cent, compared with 6.6 per cent for all developing countries and a global average of 7.0 per cent. A notable feature within developing Asia is the dramatic increase in inflows to China. Investment flows to China during the first four years of reforms (1979–83) was rather modest, amounting to about US$2 billion. FDI inflows began to gather momentum in the latter half of the 1980s, but were interrupted by the Tiananmen Square incident in 1989. Then, from 1991, FDI began to increase dramatically. Over the past two decades China has been by far the largest developing country recipient of inward FDI.

Foreign direct investment in developing Asia

33

China’s share in total FDI inflows to developing countries and transition economies increased from 11 per cent during 1985–89 to 29 per cent during 2000–04. For the past five years (2000–04) China has been the second largest recipient of foreign investment in the world, accounting for 7 per cent of total gross inflows (US$50 billion per annum), after the USA (13 per cent of total inflows or US$ 140 billion per annum) (UNCTAD 2005). China’s share in inflows to Asian developing countries increased from 11.4 per cent during 1980–84 to 47.3 per cent during 2000–04, with China accounting for nearly a half of the total increment in FDI inflows to the region during this period. Total FDI flows to the ASEAN countries increased sharply from an average annual level of $3 billion in the second half of the 1980s to nearly $30 billion in 1996, the year before the onset of the recent financial crisis (1997–98). At that time the region accounted for one-fourth of total inflows to developing countries and a little over a third of total inflows to developing Asia (Table 2.1). Singapore, Malaysia, Thailand and Indonesia were among the eight largest recipients of FDI. Flows to the Philippines remained low by international standards, but they increased from $40 million to over a billion by the mid-1990s. The lion’s share of the region’s FDI inflows was absorbed by Singapore – 50 per cent of the regional total in the 1980s and 40 per cent in the 1990s – while Indonesia, Malaysia and Thailand took most of the remainder. In the first half of the 1990s, net capital inflows relative to gross domestic fixed capital (GDFCF) stood at over 30 per cent in Singapore, 19 per cent in Malaysia, 10 per cent in the Philippines, and 4 per cent in Thailand (Table 2.2). The transition economies of Cambodia, Laos and Vietnam recorded healthy FDI inflow figures in the first half of the 1990s, faltering during the latter part of the decade. There was a strong increase of FDI flowing into Vietnam following the opening up to FDI in the late 1980s and early 1990s, followed by a slow-down. In these countries, FDI accounted for much larger shares of GDFCF (30 per cent in Vietnam, 20 per cent in Lao and 24 per cent in Cambodia), reflecting low levels of domestic private investment. Was the continuity of the impressive ASEAN record of attracting FDI broken by the 1997–98 economic crisis? Total inflows to the region have declined persistently from $34 billion in 1997 to $14 billion in 2002 (Table 2.1). However, the post-crisis experiences of individual countries vary substantially. Indonesia has continued to experience contraction in FDI, contributing significantly to the decline in total flows to the region. When the three boom years prior to the onset of the crisis (which were characterized by abnormal investor euphoria) are excluded, there is no discernible break in the trend of FDI inflows to Singapore, Thailand and the

34

Multinational enterprises in Asian development

Philippines. Flows to the Philippines, which were the least affected by the crisis, in fact continued to increase rapidly throughout. Net FDI flows to Malaysia declined from $7.2 billion in 1996 to $6.0 billion in 1997 (a 24 per cent contraction) and have remained virtually flat at that level from about mid-1998, compared with a significant increase in flows to Korea and Thailand (Table 2.1). It could well be that the prolonged period of policy and political uncertainty following the onset of the crisis, and widespread market scepticism about the fate of Malaysia’s unorthodox reform package introduced in September 1998, may have played a role. The two extreme cases of Indonesia (continuous contraction) and the Philippines (continuous increase) clearly suggest the post-crisis decline in FDI inflows to the region was a temporary aberration associated with economic disruption caused by the crisis. Relating to this point, there is also evidence that the decline in FDI after the onset of the crisis was by and large limited to domestic-market-oriented investment and FDI in export-oriented industries continued to increase throughout the period (Athukorala 2003a). As we will see in Chapter 5, interestingly FDI seems to have weathered the crisis far better than domestic private investment in the crisis-hit countries. It is also important to note that the continuation of the crisis-driven decline in FDI inflows to these countries well beyond the economic recovery from the crisis (that is, beyond 2000) was largely a reflection of a large overall decline in global FDI flows during 2000–03 (UNCTAD 2005). Total global FDI inflows declined from US$ 134 billion in 2000 to 83 billion in 2001, 72 billion in 2002 and 63 billion in 2003 and recovered marginally to 65 billion in 2004.11 Total inflows during the four years from 2001 to 2004 were 24 per cent lower than the comparable figure for the preceding four years (1997–2000). Interestingly, FDI inflows to the crisis-affected Asian countries (and to the developing Asian region in general) seem to have shown remarkable resilience to this massive global contraction. The 1990s have seen a marked increase in FDI in India, a trend that represents a clear break from the two preceding decades. The share of FDI in India in total developing country inflows increased from 0.4 per cent in the 1980s to over 1.5 per cent by the first five years of the new millennium. As a share of GDFCF the increase was from less than 0.3 per cent to over 3 per cent between these time points.12 India’s total annual FDI during 2000–2005 amounted to about 7 per cent of that going into China and 19 per cent of that going into the ASEAN countries as a group (Table 2.1). A notable aspect of FDI flows to India is that they have behaved quite independently of the global trends in FDI inflows to developing countries. This pattern clearly suggests that the domestic investment climate (demand-side factors in the investment market) has been the prime mover of investment flows to the country.

Foreign direct investment in developing Asia

35

THE ‘CHINA FEAR’ As we have already observed, FDI inflows to developing Asia from the mid1990s have been dominated by inflows to China. The dramatic growth of FDI inflows to China has been accompanied by a sharp decline in the share of almost every other country in the total regional (as well as global) inflows. These contrasting patterns, coupled with some anecdotal evidence of foreign firms relocating to China,13 have led to serious concern in policy circles in countries in the region (particularly in Southeast Asian countries whose growth dynamism for over two decades has relied heavily on FDI) that ‘competition’ from China has begun to erode their prospect of attracting FDI as a pivotal element of their outer-oriented growth strategy.14 Some of the FDI inflows to China could well have been at the expense of other countries, but there are strong reasons to argue that it would be a mistake to overstate the ‘China factor’. First, there is some controversy over China’s actual FDI inflows (Wei 2000, Pomfret 1991). Part of the reported FDI from Hong Kong (which has accounted for over 40 per cent of total FDI inflows to China over the past ten years) is ‘round-tripping’ capital: capital that originated on the mainland and returned to the mainland disguised as Hong Kong investment to take advantage of tax, tariff and other benefits accorded to foreign-invested firms (this may be 15 per cent of Hong Kong investment). Also, the official Chinese statistics on FDI are believed to contain ‘serious fat’, given the competition among various regions and provinces to demonstrate their superior performance in attracting foreign investors. The comparison of FDI flows to China reported by the official sources with those reported by investing countries in Table 2.3 is consistent with this view. During the period from 2000 to 2003, total FDI flows to China from the eight countries listed in the table, as reported by China, was more than twice the amount reported by these source countries. Secondly, investors from Hong Kong and Taiwan accounted for approximately 66 per cent of China’s total FDI inflows between 1983 and 2004. These flows are presumably driven largely by Chinese ethnic links in addition to the general economic considerations impacting on overseas investment decisions (Huang 2003, Wei 2000, Pomfret 1991). Thus, even if the statistical errors noted above and the official data are taken at face value, it is not realistic to assume that these flows are completely at the expense of other investment locations. Thirdly, data on global investment patterns clearly indicate that the measured decline in the share of ASEAN in total developing country inflows was not entirely due to the increase in flows to China. In fact, inflows to other developing countries (that is, countries other than China

36

Multinational enterprises in Asian development

Table 2.3 FDI flows to China as reported by China and by investing economies, 2000–031 Investing country

As reported by China (US$ million) (1)

As reported by investing country (US$ million) (2)

Percentage difference between (1) and (2) (3)

France Germany Japan Malaysia Netherlands Thailand UK USA

1928 3182 11 454 834 2138 586 3112 14 241

1053 2682 5706 203 600 36 2708 4653

83.1 18.6 100.7 310.8 256.3 1527.8 14.9 206.1

Total2

37 475

17 641

112.4

Notes: 1. Three-year total. 2. Total for the countries listed here. Source: Compiled from UNCTAD (2005), Box Table 1.1.1.

and ASEAN) have increased at a much faster rate, from about 30 per cent of total flows to developing countries to over 53 per cent by 2002, compared with a mild decline in China’s share from 32 per cent to 28 per cent between 1995 and 2002 (Table 2.1). Many of these ‘other developing country’ flows were triggered by liberalization reforms in Eastern Europe, the formation of NAFTA (which prompted massive relocation of production units from North America to Mexico) and regional cooperation initiated in many parts of Latin America. Fourthly, a comparison of FDI inflows to China (a relatively new host country) with those to other countries with a longer history as hosts to MNEs needs to be qualified for possible bias arising from the nature of the available FDI data (reported in the World Investment Report, based on individual countries’ balance of payments records). A well-known limitation of the FDI data for most countries in the region (perhaps all ASEAN countries other than Singapore) and China is that these data do not capture investment financed through retained earnings. At the same time, there is convincing evidence that the relative importance of retained earnings compared with the other two components of FDI (equity capital and intra-company borrowing) is positively related with the duration of MNE involvement in a given host country (Lipsey 2000). This omission is

Foreign direct investment in developing Asia

37

therefore likely to overstate capital inflows to China and understate those to many other countries (in particular the five major ASEAN countries) in the region. The data on US foreign direct investment reported in Table 2.4 (which admittedly appropriately capture retained earnings), in fact, suggest that comparison bias is a real possibility. Note that reinvested earnings account for only 35 per cent of total US FDI in China compared with 57 per cent in the five ASEAN countries. Interestingly, comparison based on the total FDI series reported in the table (which cover all three components of FDI) does not provide any clear pattern of FDI competition between the five countries and China, as far as the US FDI is concerned. Fifthly, migration of some production processes within vertically integrated high-tech industries such as electronics, motor vehicles and cameras to China does not necessarily imply a zero sum game of competing for FDI. Rather, this process opens up opportunities for additional investment in original equipment manufacturing and back-room operations in the ASEAN countries for the Chinese market. For instance, recently Intel Corporation (the world’s largest computer chip maker) simultaneously invested US$200 million in a second semiconductor chip assembly and testing plant in the Central Chinese city of Chengdu (in addition to its US$500 million assembly and testing facility in Shanghai) and at the same time invested US$40 million to expand its design and development activities at its plant in Penang, Malaysia, (and also announced plans to spend $100 million a year on further expansion of research and development activities there).15 More recently Intel signed an agreement with the government of Vietnam to set up a large electronics component assembly plant in Vietnam, as the first step in linking Vietnam to its regional and global operational network.16 The Intel story clearly suggests that the highly publicized cases of MNEs migrating from ASEAN to China may simply reflect only one side of the ongoing process of restructuring international production within the region. It is also consistent with emerging patterns of manufacturing trade in the region; there is clear evidence of rapid expansion of components and parts exports within the broader product category of machinery and transport equipment (Section 7 of the Standard International Trade Classification) from the five major ASEAN countries to China (Athukorala 2005). Trade in parts and components in high-tech industries is dominated by MNEs and makes FDI flows to China and the other countries in the region ‘complementary’ rather than ‘competing’. Finally, as an outcome of dramatic economic transformation over the past two decades, China itself is now becoming a significant overseas investor, predominantly in the other developing countries in the region. Resource-rich countries like Indonesia, Malaysia, Lao PDR and Cambodia have begun to attract ‘resource-seeking’ investors from China. There is also evidence that

38

Multinational enterprises in Asian development

Table 2.4 US direct investment in selected Asian countries, 1994–2002 (US$ million) 1994 1995 1996 1997 1998 1999 2000 India 121 95 16 Equity 71 (D) 19 Intra-company loans 23 (D) 21 Retained earnings 26 54 13 ASEAN 1067 1683 2871 Equity – – – Intra-company loans – – – Retained earnings 1027 1648 2195 Indonesia 72 8 161 Equity 28 (D) (D) Intra-company loans 35 (D) (D) Retained earnings 9 (*) 16 Malaysia 330 631 925 Equity 22 19 (D) Intra-company loans 144 184 (D) Retained earnings 163 427 597 Philippines 172 58 468 Equity 5 4 0 Intra-company loans 97 69 177 Retained earnings 70 123 291 Singapore 112 783 1148 Equity 28 10 141 Intra-company loans 699 201 100 Retained earnings 783 974 1107 Thailand 381 203 169 Equity (D) 2 9 Intra-company loans (D) 79 5 Retained earnings 2 123 184 China 454 341 520 Equity 195 95 64 Intra-company loans 206 38 77 Retained earnings 54 208 379

6 28 2 25 3149 – – 2176 103 0 50 53 451 25 26 400 149 (D) (D) 230 2477 21 1116 1382 31 (D) (D) 111 937 136 251 550

113 147 14 20 358 – – 192 46 0 53 8 318 28 23 322 7 (D) (D) 97 317 88 33 196 412 277 77 213 1115 708 225 182

246 235 60 48 1139 – 825 1503 80 88 27 18 703 4 973 266 119 (*) 254 135 1775 172 248 1356 106 197 181 272 1261 701 64 496

2001

2002

23 195 239 1 125 13 4 52 103 20 19 123 6522 3431 2445 217 – – 2847 – – 3456 1820 1629 31 21 51 0 0 (*) 18 9 39 13 12 12 1369 225 890 330 (D) (D) 363 (D) (D) 676 532 533 798 258 658 2 3 (D) 572 397 (D) 223 136 177 3648 2670 887 205 (D) 64 1648 (D) 115 2205 1029 939 676 815 41 90 (*) 2 246 680 11 339 135 32 1332 1597 517 217 (D) 114 466 (D) 288 649 476 691

Notes: (D) Suppressed to avoid disclosure of data of individual companies. * Zero or negligible. – Total cannot be computed because of suppression of data for one or more countries. Source: Compiled from the US Bureau of Economic Analysis electronic database, http://www.bea.doc.gov/bea/di/diacap_98.htm.

Foreign direct investment in developing Asia

39

rapid growth in wages propelled by rapid economic growth has already begun to erode China’s attractiveness as a low-wage investment location and has begun to entice Chinese firms involved in labour-intensive manufacturing (clothing and footwear in particular) to relocate production in low-wage countries in the region. For instance, Chinese investors are already the largest investors in the garment industry in Cambodia and they have also begun to enter Vietnam. The imposition of punitive trade restrictions by the European Union and the USA on clothing and footwear imports from China in mid-2005 has also begun to act as a powerful push factor in overseas investment by Chinese firms in labour-intensive industries.17

INDUSTRY PROFILE During the first two decades or so of the post-war period, FDI in Taiwan and Korea was predominantly involved in domestic-market-oriented production. In both countries from about the mid-1960s there was a palpable shift in the industry composition of FDI away from early production patterns and toward export-oriented production. From about the late 1980s, FDI has played an important role in the rapid world market penetration of automotive, consumer electronics and electrical goods exports from these countries (Koo 1985, Schive 1990, Amsden and Chu 2003). The traditional colonial pattern of FDI in ASEAN was characterized by a heavy concentration in the primary sector: mining and tropical cash crop production for exports. From the 1960s, there was a gradual shift in the sectoral composition of FDI in favour of manufacturing. From about the early 1980s investment in services tended to increase rapidly, a development which received a further boost from the services sector opening as part of FDI liberalization following the onset of the recent economic crisis. In particular, FDI participation in banking and other financial services has increased through direct acquisitions and mergers of local banks and companies seeking to create internationally competitive units. During 1999–2000, manufacturing accounted for about a third of total FDI in the region, with the primary sector (agriculture and mining) and the tertiary sector (construction and various services) accounting for 28 per cent and 23 per cent respectively (Lindblad 1998, Brooks and Hill 2004). In Singapore, from the beginning, manufacturing FDI was predominantly in ‘efficiency-seeking’ (export-oriented) production, mostly in the electronics industry. In other ASEAN countries, there has been a palpable shift in MNE activities away from ‘market-seeking’ (domestic-marketoriented production) and towards efficiency-seeking production gradually from the mid-1970s, and this process accelerated in the 1990s (Huff 1994,

40

Multinational enterprises in Asian development

Athukorala and Hill 2003). Old-style import-substituting FDI behind tariff barriers is now found only in a few industries, such as automobiles and petrochemicals. As in Singapore, efficiency-seeking FDI in Malaysia and the Philippines has largely concentrated in electronics. In Thailand in recent years FDI has begun to come in a big way into the export-oriented electronics and automotive industries. By contrast, in Indonesia efficiencyseeking FDI has continued to be confined largely to standard labourintensive consumer goods production. In Vietnam, during the first decade following economic opening, FDI was heavily concentrated in domesticmarket-oriented capital-intensive industries (the chemical and automotive industries, in particular) and the construction and services sectors. The period from about the late 1990s has seen a notable expansion of MNE activity in labour-intensive consumer goods production, in particular clothing, footwear and furniture. More recent years have seen some promising signs of MNE entry into component assembly in the electronics and electrical industries, the most notable recent case being the signing of an investment agreement by Intel to set up a large assembly plant in Ho Chi Ming City (already noted). In Cambodia, FDI is heavily concentrated in the export-oriented garments industry. In Lao PDR, hydroelectricity production and extractive industries (mostly gold and tin mining) have been the major attraction to foreign investors, with some investment (mostly by regional investors) in the export-oriented garments industry (ADB 2006). FDI in China was more heavily concentrated in export-oriented countries from the beginning than in Vietnam and other transition economies. Until about the mid-1990s, virtually all of the industrial output of foreigninvested enterprises (FIEs) was exported (Naughton 1996, Figure 2). Since then the share of domestic market sales in total FIE output has gradually expanded in line with the relaxation of investment approval procedures to permit production for the vast domestic market. As foreign firms have begun to benefit from cost advantages arising from scale economies involved in production for the vast domestic market to gain competitiveness in the world market in various consumer durables and electronics products, the efficiency-seeking and market-seeking distinction as applied to FIE operations has become increasingly blurred (Huang 2003). A third of the FDI stock in India at independence (in 1947) was in the primary sector (plantations, mining and oil), a quarter in manufacturing, and the rest in services (mostly trading, construction, transportation and utilities) (Athreye and Kapur 2001, Table 3). From the 1960s, inflows tended to concentrate increasingly in manufacturing (and also there was considerable disinvestment in other sectors). By the late 1980s, manufacturing accounted for about 85 per cent of the FDI stock in the country. Recent trends (based on approval data) point to a noticeable increase in

Foreign direct investment in developing Asia

41

FDI in energy, telecommunication, transportation and various other service sectors.18 Of the total approved FDI during 1991–2002, more than half was in construction and services, with 43 per cent going into manufacturing and the balance to primary sector activities. Within manufacturing, the capital goods sector (basic metal products, machinery and transport equipment) has remained the predominant recipient of FDI. Though India has an enormous supply of low-wage, low-skill manpower which could be used to attract FDI into garments and other simple assembly activities, the overall investment regime has continued to favour foreign investment in heavy, complex activities predominantly focused on the domestic market. There has not been any significant increase in India’s penetration of world markets in industrial products over the past decade despite the increase in FDI (Srinivasan 1998). The only notable exception has been the phenomenal increase in software exports since the mid-1990s (Saxenian 2002).19 An important development in export-oriented MNE production is the rapid expansion in cross-border multi-plant operations within the region. MNE affiliates in high-tech industries located in Singapore and Malaysia, faced with intensifying competition emanating from deregulation and falling trade barriers, have begun to relocate some segments of the production process to neighbouring countries. This process has created a new regional division of labour, based on skill differences, differential factor prices (especially wages) and superior communication facilities. Production is thus dispersed within the region, leading to intensified intra-regional trade in the context of a global industrial network. One important implication of these cross-border production activities, which are primarily focused on the complementarity of national resource endowments, is that promoting FDI is a ‘positive sum game’ in which all players gain from trade in differentiated products (Dobson and Chia 1997, Athukorala 2006a). In Sri Lanka and Bangladesh, much of the recent FDI flows have gone to labour-intensive export industries, mostly garments (Ali 1999, Athukorala and Rajapatirana 2000a). In Sri Lanka, there has also been some foreign investment in electronics and rubber products (see Chapter 9).

PROSPECTS: A SUMMING UP In analysing prospects for attracting FDI, it makes little sense to talk about a single Asian experience with FDI. The region is characterized by great economic diversity among countries, ranging from the highly developed economies of Korea, Taiwan and Singapore to late-reforming low-income countries in South Asia, and to the formerly centrally planned economies of

42

Multinational enterprises in Asian development

Vietnam, Cambodia and Lao which only recently reconnected to the global economy after a long period of economic isolation. There are vast differences among these countries in the structure of their economies, and hence in their patterns of comparative advantage. This suggests that individual countries have their own niches in attracting FDI to be involved in different stages of the production process in vertically integrated global industries. Contrary to the popular perception, China’s emergence as a major investment location is not a ‘zero sum proposition’ from the perspective of the region. Rather, it seems to have added further dynamism to region-wide MNE operations. There is significant potential complementarity of FDI in China and other countries in the region. Migration of some production processes within vertically integrated high-tech industries such as electronics, motor vehicles and cameras to China does not necessarily imply a zero sum game of competing to attract FDI. Rather, it opens up opportunities for original-equipment manufacture and back-room operations in other countries in the region. Even if China continues to be relatively attractive, not all stages of production within vertically integrated global industries are going to move to China. Supply chain managers are reluctant to source all their inputs from any one nation, preferring instead to diversify the risk of exchange rate instability or supply disruptions across countries. There is also evidence that rapid growth in wages has already begun to erode some of the cost advantages of China, owing to its high-speed development. Manufacturing wages in coastal China are now comparable to those of Indonesia and the Philippines, and higher than in Vietnam. Moreover, with rapid growth in China, resource-rich countries in the region like Indonesia, Malaysia, Lao PDR and Cambodia are becoming increasingly attractive investment locations for Chinese firms. India and other South Asian countries have remained under-performers in attracting FDI. India in particular has immense potential for becoming a major host to MNEs. Its greatest asset in this regard is a large, educated English-speaking population that is willing to work for relatively low wages. In spite of its widespread illiteracy, few countries can match India’s combination of low-wage highly skilled workers. The pull of a large established industrial economy like India, despite its current deficiencies and technological gaps, is much greater than that of its smaller, less industrialized neighbours. This is not just because of the potential of its market but also because of the level of local industrial skills and experience, which could provide a fertile basis for the operations of foreign firms if its liberalization process continues. In these circumstances, India could become a major destination of both market-seeking and efficiency-seeking FDI. Its remarkable success in the global software and information technology industries perhaps provides a preview of India’s potential to grow

Foreign direct investment in developing Asia

43

through export-oriented FDI under more liberal trade and investment regimes. The software industry is the unique case in India where restrictions on MNE entry were virtually suspended. This was accompanied by the abolition of quantitative restrictions on imports of computers and peripherals and drastic cuts in import tariffs on these products. This combination of FDI and trade liberalization laid the foundation to make the domestic software industry internationally competitive. Now virtually every major global company in the software industry has a base in India and the entry of MNEs has opened up opportunities for Indian companies to thrive through functional specialization, to develop niche products and services for large clients abroad. As one commentator has put it, ‘the success of foreign investment in the software industry is a measure of the failure of India’s restrictions on foreign investment elsewhere’ (Desai 2002, p. 205).

NOTES 1. 2. 3. 4. 5. 6.

7. 8. 9. 10.

11. 12. 13.

The term ‘developing Asia’ is used here to refer to all countries in the Asian-Pacific region other than Japan, and thus covers South Asia, the ten member countries of ASEAN, Northeast Asia and China. For details on recent policy reforms in Vietnam, see Chapter 10. The cumbersome and complex nature of India’s regulatory framework relating to foreign investment and the related poly regimes have been extensively analysed. See, in particular, Kidron 1965, Kumar 1994 and Athreye and Kapur 2001. In response to this ownership restriction some major MNEs such as IBM and Coca-Cola left the country. Policy makers are of the view that liberalization of FDI should take place in the context of a broader scheme of economic liberalization: otherwise the attraction of new investment will reproduce, and add to, the cost of the past. According to the World Bank’s World Business Environment Survey (2000) managers in India spend 16 per cent of their time in dealing with bureaucracy, compared with 9 per cent in China, 11 per cent in Latin America and 12 per cent in transitional Europe (World Bank 2003b, pp. 98–9). The corporate tax rate for foreign companies is 48 per cent in India, compared with rates in the range of 15 to 30 per cent in East Asia. For details of the reform process, see Chapter 9. For details on the nature and limitations of the World Investment Report on FDI, see UNCTAD 2005, Box 1.1. For instance, the Reserve Bank of India broadened the coverage of its FDI estimation procedure in 2003 (with effect from 2000/1 fiscal year) to include retained earnings. According to the revised data for 2000/1 and 2001/2, on average the new component accounted for about 40 per cent of the total reported FDI figures (Reserve Bank of India 2004). What caused this massive contraction in FDI flows, an unprecedented occurrence since 1970 when the World Investment Report FDI series begins, remains a mystery. The recorded increase in inflows in the past three years over the previous years partly reflects revision made to the estimation method. See Note 10. Yusuf (2003, p. 294, Table 7.3) reports eight cases (involving 12 000 job losses) of relocation of production plants by electronics MNEs from Penang, Malaysia, to China during 1998–2001.

44 14. 15. 16. 17. 18.

19.

Multinational enterprises in Asian development See for instance Freeman and Bartels (2004), Chapter 1 and the works cited therein. Asia Wall Street Journal, 27 August 2003, A1 and A4. The point is based on information obtained from www.intel.com. This point is based on interviews conducted with the officials of the Cambodian Investment Board in March 2006. Data based on investment approvals highly exaggerate the actual (realized) level of FDI in India. For instance during 1991–98 realized investment amounted to only 21.7 per cent of approved investment (Bajpai and Sachs 2000, Table 1). However this is not a serious problem in analysing the sectoral (or source-country) composition of FDI. Software exports from India increased persistently from about $330 million in 1993/94 to over $2650 million by the end of the decade (Saxenian 2002, Table 5.1).

3. Multinational enterprises and manufacturing for export: emerging patterns and opportunities for latecomers* The past three decades have witnessed a profound shift in the relationship between multinational enterprises (MNEs) and developing countries (DCs), as more and more countries have adopted an outward-oriented growth strategy. Affiliates of MNEs, as part of the parent company’s global network, have marketing channels in place, possess experience and expertise in the many complex facets of product development and international marketing, and are well placed to take advantage of inter-country differences in the costs of production. Textured flows of information and knowledge needed for successful market penetration can be more effectively channelled through MNE networks than through arm’s-length trade links. Moreover, MNEs may be better able to resist protectionist pressures and other barriers to market entry in their home countries in such a way as to favour imports from their affiliates. In view of these considerations, enticing export-oriented foreign direct investment (EOFDI) has become an integral element of policy reforms toward export-led industrialization in many developing countries. However, the debate on the role of MNEs in this outward-oriented policy thrust is far from settled. Although the case for trade liberalization is now widely accepted in development policy circles, the case for liberalizing the FDI regime is still debated. Some who favour trade liberalization continue to advocate restriction of or conditions on FDI. This ‘revisionist’ school of thought admits that FDI can play an important role in the transmission of technology, market know-how and modern management practices to developing countries. But it argues for a selective approach to the promotion and screening of FDI, and possibly trade policy, with a view to avoiding potential threats to the development of indigenous entrepreneurial and technological capabilities (Lall 2003, Rodrik 1999, Amsden and Chu 2003). This view has often been reflected in a mismatch between the liberalization of FDI and trade policy regimes in some countries. 1 45

46

Multinational enterprises in Asian development

The empirical underpinning for this revisionist view largely comes from studies conducted in the 1970s, at the formative stage of the export takeoff in the newly industrializing economies (NIEs) in East Asia (for example, Hone 1974, Cohen 1975, Lall and Streeten 1977, Chapter 7, Nayyar 1978). The general inference of these studies was that the export take-off in the East Asian NIEs was predominantly based on local initiatives and ownership; and, at the firm level, transnationality was not an important aid to exporting.2 It was the innovative and selective use of various ‘non-equity’ forms of foreign participation, so the argument went, rather than the direct involvement through FDI, that was the key to NIE success. Contrary to the generally presumed link between the MNE presence and export performance, some studies even inferred that ‘MNEs in Latin American countries, where emphasis on importsubstitution continued to remain the key emphasis of industrialisation, play a much more important role in manufacturing for export than in Asia’ (Nayyar 1978). The purpose of this chapter is to take a fresh look at the role of MNEs in the export of manufactures from developing countries in Asia using more recent and comprehensive data. The chapter is motivated by the concern that, given major changes in patterns of international production over the past two decades, evidence from the early years of export-led industrialization in the East Asian NIEs may send quite inappropriate signals to policy makers in latecomer exporting countries.3 Two major developments are particularly noteworthy. First, an increasing number of firms from some NIEs have become aggressive international investors, and these ‘third world’ MNEs significantly seem to possesses specific competitive advantages over ‘first world’ MNEs in some product areas, particularly where latecomers to export-led industrialization have a comparative advantage in international production. Second, and more importantly, the ‘slicing up of the product chain’ in high-tech industries, involving crossborder relocation of global MNE activities according to the host country’s relative factor endowments, has rapidly gained importance over traditional labour-intensive final goods production as the prime mover of the internationalization of production. The chapter begins with an analytical account of the nature and changing patterns of MNE involvement in production for export in developing countries in the context of the ongoing process of globalization of manufacturing industries. The next section assembles a large body of empirical evidence on the role of MNEs in the export performance of developing host countries in Asia, and analyses it from a comparative perspective. The final section summarizes the key findings and draws out some broader policy implications.

Manufacturing for export

47

MNES AND MANUFACTURING FOR EXPORTS: A TYPOLOGY For the purpose of discussing factors impacting on the decision of MNEs to engage in export production in a given host (investment receiving) developing country, it is important to distinguish between two types of MNE affiliates. These are firms engaged in serving the domestic market (‘marketseeking’ investors) and those engaged in production for the global market (‘efficiency-seeking’ investors) (Table 3.1). When it comes to market-seeking investment in developing countries, the forces explaining the location decisions of MNEs are about the same as those explaining their presence in industrialized countries. The location decision depends primarily on the prevalence in the host country of production opportunities aimed predominantly at meeting domestic demand. Given the scale economies and very small domestic markets in many developing countries, a major (if not the key) determinant of congenial domestic production is restrictions on international trade. As domestic income levels approached industrial country levels, MNEs involved in many production activities aimed at serving both domestic and export markets, but MNE involvement in this area in most developing countries has so far been largely limited to serving the domestic market, and such investment has predominantly been determined by the ‘tariff-jumping’ motive. The socalled ‘life-cycle’ investors (à la Vernon 1962), who expand their production networks globally predominantly on scale-economy considerations, hardly regard low-income countries as attractive investment locations under freetrade conditions. In theory, in certain circumstances, MNE affiliates originally set up to serve local markets could well develop competitive advantage over the years and penetrate markets in other countries without government support (Moran 1998, Bennett and Sharp 1979). But in the real world such cases are rare and limited predominantly, if not solely, to middle-income and upper-middle-income developing countries with sizeable domestic markets.4 In some circumstances it may be possible to entice MNE affiliates which originally entered production to meet local markets to shift to exporting through government intervention. But this is typically more difficult than the encouragement of ‘fresh’ export-oriented investors, since it requires the alteration of the firm’s global production and marketing strategies. A wellknown feature of MNE behaviour is that the parent company strictly controls the performance of its affiliates in the interest of global profit. The export decision of affiliates is, therefore, not simply a matter of responding to domestic export incentives and government directives. Even if importsubstituting MNE affiliates do respond to a host government’s carrot-and-

48

Multinational enterprises in Asian development

Table 3.1 A typology of MNE participation in manufacturing for newcomer exporting countries Product category

1

Exports by marketseeking MNE affiliates – product mix varies depending on the nature of the import-substitution policy regime, domestic market size, export incentives and export performance requirements 2 Efficiency-seeking (export-oriented) production by MNE affiliates 2.1 Resource-based manufacturing – local processing of primary products previously exported in raw state 2.2 Standard consumer goods – clothing, shoes, sporting goods 2.3 Assembly activities within vertically integrated production systems 2.3a Parts and component assembly: parts of electronic and electrical machinery, motor vehicle parts etc. 2.3b Final assembly: computers, cameras, motor vehicles etc.

Production characteristics

Role of MNEs in export expansion

Technology

Factor intensity

Mostly internal to MNEs. Brand names are critical

Mostly capital and skill intensive

Of little importance (and costly)

Diffused

Mostly capital intensive

Of selective importance

Well diffused, but brand names are critical

Labour intensive

Important

Mostly internal to MNEs

Initially labour intensive, but becomes skill intensive as the country moves up the value chain Labour and skill intensive

Extremely important

Mostly internal to MNEs

Important only at an advanced stage of 2.3a and/ or in host countries with large domestic markets

stick approach, there is no guarantee that the final outcome will justify the overall cost involved. Import-substituting production units operating in a small protected market are not usually internationally competitive. Therefore, export incentives have to be introduced and maintained at high levels to generate the anticipated export push. In addition to the related

Manufacturing for export

49

budgetary and institutional constraints, the degrees of freedom available for host countries to resort to such an interventionist policy stance is becoming increasingly limited by the ongoing efforts to enhance the contestability of global markets through international agreements on cross-border investment and competition policies under the World Trade Organization (WTO) and regional trading agreements (RTAs). On the benefit side, there may be little to gain in terms of employment generation because such exports, being simply an extension of import-substitution production, tend to be highly capital intensive (Bhagwati 1996). For these considerations, the present-day discussion on MNE involvement in export-led industrialization in developing countries is focused almost exclusively on ‘efficiency-seeking’ investment (commonly known as export-oriented FDI). The role of MNEs in this sphere is ‘more distinctively a developing-country question’ (Caves 1996, p. 217). Export-oriented FDI is, however, not a homogeneous phenomenon. Rather it is a complicated and finely differentiated means of globalization of production. The opportunities available to a given country in mobilizing FDI in economic growth and development depend on relevant typological characteristics and the investment environment of the country and the changing pattern of international production in the global context. In order to understand the opportunities arising from the interaction of these two factors, it is important to distinguish among three different categories of exportoriented production (Table 3.1): (1) (2) (3)

resource-based manufacturing labour-intensive final consumer goods assembly processes within vertically integrated global production systems.

In the first category, the relevance for a given host country of MNE participation for export expansion depends primarily on the availability of relevant natural resources. Even if resources are available, there are other factors which may render policies designed to entice foreign investors ineffective. For instance, some processing activities, particularly those in the mineral and chemical industries, are characterized by high physical and/or human-capital intensity and may not be economic in a low-income country. A further major deterrent is cascading tariff structures in industrialized countries, which still provide heavy effective protection to domestic processing industries. Insecure property rights in resource-rich developing countries also may act as a deterrent to investors in large, capital-intensive projects. These constraints notwithstanding, there are some product areas where there are significant opportunities for successful export expansion

50

Multinational enterprises in Asian development

through MNE participation. One such product line, which has gained importance over the past two decades for agriculturally resource-rich developing countries, is agro-based processed food, seafood in particular (Athukorala and Jayasuriya 2003). For the typical developing economy, labour-intensive consumables (Category 2.2) are generally considered the natural starting point in the process of export-led industrialization. However, the role of MNEs in this area remains a controversial issue. In the spectacular export take-off of the East Asian NIEs in the 1960s, the key role was played by indigenous firms with the help of marketing services provided by foreign buyers – the Japanese trading houses and the large retail buying groups in developed countries (Hone 1974, Nayyar 1978, Naya 1990, Westphal et al. 1979). There are, however, strong reasons to argue that this ‘early East Asian pattern’ of local-entrepreneur dominance in exports may not be replicated in latecomer countries. First, perhaps the most important factor behind the East Asian experience was the unique entrepreneurial background of these countries. Hong Kong, Taiwan and to some extent Singapore started with a stock of entrepreneurial and commercial talents inherited from the prerevolution industrialization in China. Hong Kong and Singapore also had well-established international contacts based upon entrepot trade that involved exporting manufactured goods to begin with. Likewise, the considerable industrial experience that accumulated over the preceding five decades or so under the Japanese occupation was instrumental in the export take-off in Taiwan and Korea (Amsden and Chu 2003, Rhee et al. 1984). Therefore, there was not such a large difference between domestic firms in these countries and foreign firms with regard to knowledge of and access to market channels. The present-day newcomers to export-led industrialization (including most transitional economies) are not generally comparable to the East Asian NIEs in terms of the initial level of entrepreneurial maturation. In many of these countries, the import-substitution growth strategy pursued indiscriminately over a long period has thwarted the development of local entrepreneurship. Domestic firms are generally weakly oriented towards, and have limited knowledge of, highly competitive export markets. This observation seems even more relevant for the present-day transition economies, which have embarked on the process of integration into the global economy following a long period of central planning (Lankes and Venables 1996). Moreover, from around the mid-1980s, successful exporting firms in the East Asian NIEs have begun to play an important role as direct investors in the latecomers’ labour-intensive export industries. Two main factors accounted for this trend: the erosion of international competitiveness of labour-intensive export products from their home countries as a result of

Manufacturing for export

51

rising real wages and exchange rates; and the imposition and gradual tightening of quantitative import restrictions (QRs) under the Multi-Fibre Arrangement (MFA) by industrialized countries on certain labour-intensive exports (mostly textiles, garments and footwear) (Wells 1983, 1994). There are indications that, consistent with rapid structural transformations that were taking place in the NIEs prior to the recent economic crisis (and from which they have quickly recovered), this intermediary role of these ‘new’ investors in linking latecomers to world markets may become increasingly important in years to come. A major advantage which investors from these new countries possess is that, unlike MNEs from developed countries, they are familiar with and/or easily adaptable to the more difficult business conditions (such as poor infrastructure, bureaucratic red tape and unpredictable policy settings) in latecomers. Given that NIE firms have developed considerable specialized knowledge of small-scale and labour-intensive production procedures in the manufacture of standardized products, they have a powerful competitive advantage over both local firms and MNEs from industrialized countries (ICs) in these latecomer environments (Gereffi 1999). The location in developing countries of relatively labour-intensive component production and assembly within vertically integrated international industries (‘international production fragmentation’ or ‘outsourcing’) (Table 3.1, Category 2.3) has been an important feature of the international division of labour since about the late 1960s. The process was started by electronics MNEs based in the USA in response to increasing pressures of domestic real-wage increases and rising import competition from low-cost sources (Sharpton 1975, Helleiner 1973b, Feenstra 1998). The transfer abroad of component assembly operations now occurs in many industries where the technology of production permits the separation of labourintensive components from other stages of production. Assembly operations in the electronics industry (in particular, assembly of semiconductor devices, hard disk drives and so on) are still by far the most important. The other industries with significant assembly operations located in developing countries are electrical appliances, automobile parts, electrical machinery, optical products, musical equipment, watches and cameras. In general, industries that have the potential to break up the production process to minimize the transport cost involved are more likely to move to peripheral countries than other industries. The expansion of production fragmentation as an important facet of international production has been hastened by two mutually reinforcing developments over the past few decades. First, rapid advances in production technology have enabled the industry to slice up the value chain into finer, ‘portable’ components. Second, technological innovations in communication and transportation have shrunk the distance that once separated

52

Multinational enterprises in Asian development

the world’s nations, and improved the speed, efficiency and economy of coordinating geographically dispersed production processes (Krugman 1995, Jones 2000, Jones and Kierzkowski 2001a). There is evidence that global trade in parts and components (middle products) is growing much faster than total manufactured exports (Athukorala 2006b, Feenstra 1998, Hummels et al. 2001, Yeats 2001).5 At the formative stage of worldwide assembly operations in the late 1960s, some observers were sceptical about prospects for developing countries to rely on this form of international specialization for export expansion. They predicted that the process would be reversed because of rapid automation of production processes in developed countries (for example, Frobel et al. 1980, Cantwell 1994). However, in many high-tech industries (notably electronics and electrical products) rapid innovation and continuous technical change, which bring about a constant cycle of change and obsolescence, have proved to be formidable constraints on rapid automation as an alternative to offshore assembly. Therefore, the indications are that this form of internationalization of production will continue to expand, providing countries with the opportunity to find new niches for labour-intensive, export-oriented production (depending of course on their ability to provide an enabling domestic economic environment). The bulk of outsourcing takes the form of locating small fragments of the production process in a low-cost country and re-importing the assembled components to be incorporated in the final product (Category 2.3a in Table 3.1). However, recent years have seen a noteworthy expansion of the coverage of global assembly operations from component assembly to assembly of final products (such as computers, cameras, TV sets and motor cars) (Category 2.3b). In final assembly, labour costs, while significant, are of secondary importance compared with the availability of world-class operator, technical and managerial skills; a good domestic basis of supplies and services; relatively free access to world-priced inputs including capital; and excellent infrastructure. In other words, the location decisions of MNEs in this sphere depend on the availability of a wider array of complementary inputs that enable their facilities to be efficient by world standards. Also, given the heavy initial fixed costs, MNEs are hesitant to establish overseas plants in final assembly without considerable first-hand commercial experience in the host country. For these reasons, overseas production units of MNEs involved in such final stage assembly are normally located in countries which are at a relatively advanced stage of export-led industrialization.6 MNEs from industrialized countries are the key actors in worldwide offshore assembly operations. While MNEs from the USA dominated the scene at the formative stage of global spread of assembly activities in the late 1960s, the involvement of Japanese and Western European MNEs also

Manufacturing for export

53

has been gaining importance since the late 1970s. More recently MNEs from more advanced developing countries, notably those from the East Asian NIEs, have also joined this process of internationalization of production. In response to rapid domestic wage increases, the growing reluctance of domestic labour to engage in low-paid blue-collar employment, and stringent restrictions on the importation of labour, firms in the electronics industry and other durable consumer goods industries in NIEs in East Asia have begun to produce components and sub-assemblies in neighbouring countries where labour costs are still low. In recent years, outsourcing has begun to spread beyond the domain of MNEs. Many companies which are not parts of MNE networks now procure components globally through arm’s-length trade. Technological innovations in communication (in particular the Internet) have reduced the cost of outsourcing, particularly through reduced research costs. The process has also been facilitated by the ‘standardization’ of some components.7 However, the bulk of fragmentation trade still takes place under the aegis of MNEs (Rangan and Lawrence 1999, Hanson et al. 2001). An important new development in MNE involvement in Asia in assembly activities is the emergence of cross-border multi-plant operations within the region (Borrus 1997, Dobson and Chia 1997, McKendrick et al. 2000, Naughton 1999a and 1999b). MNE affiliates in high-tech industries located in countries like Hong Kong, Singapore and Taiwan, faced with intensifying competition emanating from deregulation and falling trade barriers, have begun to relocate some segments of the production process to neighbouring countries. This process has created a new regional division of labour, based on skill differences, differential factor prices (especially wages) and superior communication facilities. Production is thus dispersed within the region, leading to intensified intra-regional trade in the context of a global industrial network. One important implication of these crossborder production activities, which are primarily focused on the complementarity of national resource endowments, is that promoting FDI is a ‘positive sum game’ in which all players gain from trade in differentiated products (Dobson and Chia 1997). In sum, the discussion in this section suggests that, in the context of emerging patterns of international division of labour, MNE involvement through FDI is bound to be more important for latecomer countries to export-led industrialization compared with the early experience of presentday NIEs. These developments seem to have reduced the efficacy of relying predominantly on non-FDI forms of MNE as a means of acquiring export competence. With this background, we now turn to the available direct evidence concerning MNE participation in international production in developing countries for further scrutiny of these postulates.

54

Multinational enterprises in Asian development

EVIDENCE Table 3.2 brings together a data set to examine the contribution of MNEs to manufactured exports from the four East Asian NIEs (Hong Kong, South Korea, Taiwan and Singapore) and eight other developing Asian countries (China, Indonesia, Malaysia, the Philippines, Thailand, Vietnam, India and Sri Lanka). MNE involvement in export expansion is measured in terms of the percentage share accounted for by MNE affiliates in total manufactured exports (column 3). Export performance of each country is measured in terms of three indicators – export value (column 4), the share in total world manufactured exports (world market share) (column 5) and annual export growth (column 6). The final column contains summary observations on the nature of the product composition of MNE-related exports in terms of the typology developed in the previous section. MNE share in total manufactured exports (MNEXS) and world market share (WMSH) are plotted in Figure 3.1 for the countries for which continuous annual data series on the two variables are available for at least ten years. It is important to emphasize that data pieced together from diverse sources on MNE share in exports are not strictly comparable (see Appendix 3.1 for details). In particular, there is no uniform treatment of the ownership share used in identifying the ‘multinationality’ of host-country firms across these sources. Estimation errors in individual country figures are also unlikely to be consistent across countries, as data quality naturally varies. Nevertheless, the estimates assembled here are the best available and, taken together, they yield a number of important inferences. The 12 Asian countries covered in this study accounted for over twothirds of total manufacturing exports from the developing countries over the past two decades. Thus the data presented in Table 3.2 give some idea of the changing role of MNEs in manufactured exports from these countries. A rough calculation obtained by combining the data on MNE share in exports and total manufacturing exports from each country suggests that the share of MNEs in combined exports from the 12 countries increased from about 30 per cent to 50 per cent between 1980–84 and 1995–99. When Hong Kong, Taiwan and Korea are excluded from the calculations, the increase is from 33 per cent to over 60 per cent. These trends are in sharp contrast to what Nayyar (1978) found by surveying the experiences of 12 developing countries (in both Asia and Latin America) for the 1960s and the 1970s.8 According to his estimates, share of MNEs in total manufacturing exports from developing countries amounted to around 15 per cent throughout the period under study without any discernible upward trend.

55

Manufacturing for export

Table 3.2 MNE affiliates and manufactured exports from selected developing Asian countries: MNE share in total manufactured exports and selected export performance indicators1 Country

Period

(1)

(2)

Hong Kong 1970–74* 1980–84* 1985–89 1990–94 1995–99* South Korea 1970–74* 1975–79* 1980–84* 1985–89* Taiwan 1975–79 1980–84 1985–89 1990–94 1995–99 2000 Singapore 1970–74 1975–79 1980–84 1985–89 1990–94 1995–99 China 1985–89 1990–94 1995–99 2000 Indonesia 1990–94 1995–99 2000 Malaysia 1975–79 1980–84 1985–89 1990–94 1995–99 2000 Philippines 1985–89* 1990–94*

MNE Export value share in ($ million)2 exports (%) (3)

(4)

10.0 13.8 16.0 20.8 26.5 19.3 25.0 25.8 26.1 36.7 27.9 23.1 19.7 14.4 10.1 70.0 84.5 74.9 81.5 85.2 87.2 5.3 24.3 43.4 50.5 28.5 38.5 45.3 65.2 72.4 75.6 78.1 82.2 85.5 49.9 47.6

2635 14 460 23 100 28 509 25 644 6237 9227 21 703 45 482 9782 23 242 48 612 79 165 119 492 122 152 5125 7219 17 722 28 372 66 265 113 699 29 488 75 232 160 982 18 900 29 369 29 857 3384 6669 11 585 34 213 69 444 73 510 3173 6115

World Export Nature of export market growth composition of MNE share (%)2 affiliates by the late 1990s (as per the (%)2 typology in Table 3.1) (5) (6) (7) 0.52 3.53 1.10 4.02 1.19 4.20 0.95 0.04 0.59 2.21 0.93 11.32 1.07 10.50 1.65 6.15 2.30 6.61 1.13 9.11 1.76 5.74 2.46 6.81 2.61 3.02 2.72 2.87 2.76 0.78 9.34 0.84 7.62 1.35 5.09 1.43 6.40 2.16 6.92 2.59 1.55 1.49 7.95 2.44 8.25 3.65 4.54 4.74 0.62 7.39 0.67 2.11 0.68 0.40 7.62 0.51 4.32 0.59 5.97 1.11 9.46 1.58 3.45 1.67 0.16 3.73 0.21 4.59

Mostly 2.3a and 2.3b, with the latter increasing rapidly in recent years 2.3a and 2.3b, with the latter increasing rapidly in recent years 2.3a and 2.3b, with the latter increasing rapidly in recent years

2.3a and 2.3b. 2.3a still dominates, but there as been a continuing shift from 2.3a to 2.3b since about the mid-1980s Predominantly 2.1, with some increase in 2.3a recently Predominantly 2.1, with some increase in 2.3a recently Predominantly 2.3a, with some increase in 2.3b recently

Predominantly 2.3a, with a small and

56

Multinational enterprises in Asian development

Table 3.2

(continued)

Country

Period

(1)

(2)

Thailand

Vietnam

India

Sri Lanka

MNE Export value share in ($ million)2 exports (%)

World Export Nature of export market growth composition of MNE share (%)2 affiliates by the late (%)2 1990s (as per the typology in Table 3.1) (5) (6) (7)

(3)

(4)

1995–99 2000 1970–74* 1975–79* 1980–84* 1985–89* 1990–94* 1995–99* 1990–94 1995–99 2000

76.8 85.7 11.4 16.7 13.5 15.0 50.4 62.6 12.0 39.2 56.8

23 880 25 212 1686 2038 4316 9639 27 976 49 029 1554 5375 7170

0.54 0.57 0.15 0.24 0.33 0.47 0.91 1.11 0.05 0.12 0.16

1970–74 1975–79 1980–84 1985–89 1990–94 1995–99 2000 1975–79* 1980–84 1985–89 1990–94 1995–99 2000

5.0 7.9 8.7 5.7 4.6 3.5 4.4 25.7 42.8 53.6 63.5 49.0 47.2

3724 3867 5240 8830 16 031 26 223

0.50 0.47 0.40 0.46 0.53 0.61 0.63 0.01 0.03 0.04 0.05 0.07 0.08

120 421 672 1647 2983 3410

13.01

diminishing share of 2.2 6.235 2.1, 2.2, 2.3a and 2.3b, 7.23 with the latter two 3.98 increasing rapidly in 10.41 recent years 7.71 2.49 14.57 Predominantly 2.1 9.47 (mostly seasonal food) and 2.2, with a small but increasing share of 2.3a 3.65 A wide range of 1, 4.96 with some increase 1.46 in 2.2 and 2.3a recently 6.95 4.33 2.38 8.66 5.93 3.49 9.36 3.31

Predominantly 2.2, and some 1 (mostly ceramics and rubber goods) and 2.3a

Notes: 1. In all cases manufactured exports have been measured using the ISIC-based definition (i.e. all goods belonging to Division 3 of the International Standard Industry Classification) or an approximation to it. Figures reported are five-year averages unless otherwise indicated. 2. Annual averages. 3. Figures marked with an asterisk are for a single year or some years falling within the given five-year period. For details see Appendix 3.1. Sources: Sources of data on MNE export share (column 3) and data limitations are discussed in Appendix 3.1. Export data (columns 4–6) were compiled from the International Economic Data Base (based on UN Series D data tapes) of the National Centre for Development Studies, Australian National University.

57

Manufacturing for export 30

1.4

25

1.2 1.0

20

0.8

15

0.6

10

0.4

MNEXS WMSH

5

0.2 1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

1984

0.0 1983

0

(a) Hong Kong

40

3.0

35

2.5

30 2.0

25 20

1.5

15

1.0

10

MNEXS WMSH

5

0.0 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

0

0.5

(b) Taiwan

Figure 3.1 Share of MNE affiliates (MNEXS, left scale) and world market share (WMSH, right scale) in manufactured exports in selected Asian countries (%)

Commodity Composition The available data do not permit precise disaggregation of exports by MNE affiliates according to the typology developed in the previous section. However, information coming from various country case studies on the nature of the product composition of MNE-related exports (summarized in column 7, Table 3.2) does provide empirical support for our postulates

58

Multinational enterprises in Asian development 3.0

100 90 80 70 60 50 40 30 20 10 0

2.5 2.0 1.5 MNEXS WMSH

1.0 0.5

1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

0.0

(c) Singapore 60

5.0 4.5

50

4.0 3.5

40

3.0 30

2.5 2.0

20 MNEXS WMSH

10 0

1.5 1.0 0.5

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

0.0

(d) China

Figure 3.1 (continued) concerning changing export patterns and the potential role of MNEs in manufactured export expansion. It is evident that light manufactured goods and assembly activities within vertically integrated high-tech industries have been the main areas of MNE export activities. In Singapore, Malaysia and the Philippines, MNE involvement is predominantly in assembly activities. In the other second-tier exporting countries, the standard labour-intensive products still account for the bulk of exports, but the relative importance of assembly activities seems to have increased over the years in all cases. There is also evidence of a notable concentration of assembly processes in final goods assembly in China. Interestingly, there is no evidence of a shift in

59

Manufacturing for export 50

0.8

45 40

0.7 0.6

35 30 25

0.5

20

0.3

0.4

15 MNEXS WMSH

10 5

0.1

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

0.0 1990

0

0.2

(e) Indonesia Philippines 90

1.8

85

1.6

80

1.4

75

1.2

70

1.0

65

0.8 0.6

60 MNEXS WMSH

55

0.4 0.2

1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

50

(f) Malaysia

Figure 3.1 (continued)

MNE activities from component specialization to final goods assembly in Singapore. It seems that, given the highly favourable investment climate and deep-rooted operational links coupled with relatively high domestic wages, MNEs use Singapore as the regional centre for high-tech activities in component production, while undertaking more labour-intensive component assembly and final goods assembly in neighbouring countries (mostly in Malaysia but also in Thailand and the Philippines) and China.

60

0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

MNEXS WMSH 1990

100 90 80 70 60 50 40 30 20 10 0

Multinational enterprises in Asian development

(g)) Indonesia Philippines

0.18

60

0.16

50

0.14 0.12

40

0.10 30

0.08 0.06

20

2000

1998

1997

1996

1995

1994

1993

1992

1991

1990

0

1999

MNEXS WMSH

10

0.04 0.02 0.00

(h) Vietnam

Figure 3.1 (continued) The prolonged heavy concentration of MNE activities in standard labour-intensive product lines (mostly garments and toys) in Sri Lanka can be explained in terms of unfortunate developments in the investment climate (Athukorala and Rajapatirana 2000a, Chapter 6; Snodgrass 1999). Despite the government’s continued commitment to outward-oriented policy since the late 1970s, with further strengthening of general incentives for EOFDI over the years, and the availability of cheap and trainable labour, political and policy instability has been a major deterrent to MNE involvement in assembly activities. Foreign firms involved in vertically inte-

61

Manufacturing for export 12

0.7

10

0.6 0.5

8

0.4 6 0.3 4

0.2

MNEXS WMSH

2

0.1 0.0

1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

0

(i) India

0.09

70

0.08

60

0.07 50

0.06

40

0.05

30

0.04 0.03

20 MNEXS WMSH

10

0.02 0.01 0.00

1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

0

(j) Sri Lanka Notes: The correlation coefficient between MNEXS and WMSH: Hong Kong 0.91; Taiwan 0.82; Singapore
0.40; China
0.98; Indonesia
0.70; Malaysia
0.86; Philippines
91; Vietnam
0.96; India 0.32; Sri Lanka
0.90 Sources: See Appendix 3.1.

Figure 3.1 (continued) grated assembly industries, unlike those involved in light consumer goods industries such as garments, usually view country risk and the other elements in the investment climate from a long-term perspective. Two major electronics multinationals from the USA (Motorola and the Harris

62

Multinational enterprises in Asian development

Corporation) had in fact finalized plans to establish large assembly plants in the Katunayake Export Processing Zone in Sri Lanka in the early 1980s. These plans were abandoned as the political climate began to deteriorate then. In the site selection process of electronics MNEs, there is something akin to ‘herd psychology’, particularly if the first-comer is a major player in the industry. Considering this, one can surmise that, if the two projects of Motorola and Harris had been successful, many other MNEs would have followed suit, giving a major boost to the expansion of assembly exports from Sri Lanka. There is some evidence of MNE involvement in resource-based processing activities in Indonesia, Vietnam, Thailand and Sri Lanka. But the share of MNE-related exports of these product lines in total manufactured exports seems to have declined over time in the face of rapid expansion of the standard labour-intensive products and/or component assembly. The Role of MNEs in Export Expansion The observation that MNE involvement in export expansion from the NIEs (other than Singapore) is low by international standards generally remains valid in our data set. However, it is important to note that, in both Korea and Taiwan, the MNE share in exports did increase significantly from about the mid-1970s to the mid-1980s, as compared with the figures reported by Nayyar (1978) for the late 1960s. Detailed case studies of the export performance of these countries suggest that this increase reflected the important role played by MNEs in these countries, as they shifted from the early reliance on labour-intensive standard consumer goods sectors to assembly activities in vertically integrated hightech industries, and subsequently to sophisticated consumer durables production.9 The available evidence on product composition of exports by MNE affiliates in Taiwan and Korea clearly attest to this important role played by these firms in the structural transformation of exports from these countries (Ranis and Schive 1985, Schive and Tu 1991, Koo 1985, Hu and Chan 2001). Given the rapid expansion of traditional labour-intensive exports at the initial stage of export-led growth in these countries, any analysis based on MNE shares of total exports obviously fails to capture this important point. It is interesting to note that the MNE export shares in Korea and Taiwan have tended to decline from about the mid-1980s. This is most probably due to the combined effects of exports by domestic firms growing more rapidly in recent years and an increase in domestic sales by MNE affiliates in consumer durable industries in response to the strong growth in domestic demand fuelled by rapid economic growth.

Manufacturing for export

63

The relatively small role of MNEs in export expansion from Korea and Taiwan compared with Singapore, and more importantly with the secondtier exporting countries in the region, is generally interpreted as resulting significantly from the ‘guided’ industrial development policies pursued by these countries. These countries (Korea in particular), so the argument goes, followed the Japanese pattern by relying on non-equity arrangements rather than FDI to access technology and other MNE-controlled assets. However, following Goh Keng Swee (1993), the first Finance Minister and one of the key architects of modern Singapore’s spectacular economic development, one can argue that this difference, at least to some extent, emanated from the nature of the investment environment at the time when technical advances in the US electronics industry began to create (from the late 1960s) rapid growth of demand for semiconductors, whose production and assembly required the massive use of low-cost labour. At that time, China’s Cultural Revolution was reaching its height, and political stability was a key factor governing the location decisions for assembly operations by electronics MNEs. To quote Goh (1993, p. 253): It is a matter for speculation whether in the absence of the upheaval caused by the Cultural Revolution in the mid- and late 1960s, the large American multinationals – among them National Semiconductors and Texas Instruments – would have sited their offshore factories in countries more familiar to them, such as South Korea, Taiwan and Hong Kong. These had resources and skills superior to Singapore. My own judgment remains that these three areas were too close to the scene of trouble, the nature of which could not but cause alarm to multinational investors.

This argument receives further support from the fact that not only Korea and Taiwan (which, according to the revisionists, followed ‘strategic’ FDI policy) but also Hong Kong, a country that followed almost laissez-faire economic policy throughout, were largely shunned by the electronics multinationals. By the time the political risk waned and export-led growth policies became firmly rooted in these countries, wages had increased to levels which made them less attractive as labour-intensive assembly locations. The electronics revolution in Singapore, which began in the mid1960s, absorbed all unemployed labour in that country within a period of five to seven years and electronics MNEs shifted unskilled and semi-skilled simple assembly activities to neighbouring low-wage countries – Malaysia, Thailand and Indonesia, and more recently the Philippines. In the process, Singapore assumed a major regional headquarters function for the electronics industry in Southeast Asia (Hill and Pang 1992). In the following 20 years, the MNEs diversified their operations in the region, first from simple assembly to component production operations (mainly hard disk drives),

64

Multinational enterprises in Asian development

and more recently to consumer electronics, such as TV sets, radios and sound systems. The inference that MNE participation is crucial for the export success of latecomers gains further support from a comparison between China and India, the two giant economies in the region. In China, the share of exports from enterprises with foreign investment rose from 0.4 per cent in 1984 to over 46 per cent in 1996 (Table 3.2). This was accompanied by a more than ten-fold increase in manufactured exports over this period. By contrast, in India, where MNE subsidiaries are still predominantly of the old-fashioned ‘tariff-jumping’ variety, both the share of MNEs in total manufactured exports and the rate of export growth have remained low.10 Interestingly there has been a mild, yet persistent, decline in MNE share in manufactured exports from India from about the mid-1980s and the decline became sharper following the liberalization reforms initiated in 1991 (Figure 3.1). A detailed analysis of the underlying factors is beyond the scope of this study, but the explanation seems to be in the nature of the post-reform trade and foreign investment regimes. From the early 1980s India gradually relaxed restrictions on intermediate and investment goods imports, and the removal of these restrictions was intensified as part of the liberalization reforms initiated in 1991. Consequently the pressure on MNE affiliates (which are predominantly domestic-market-oriented) to export in order to become eligible for importing gradually waned and then virtually disappeared after 1991. At the same time, given the half-hearted nature of the policy regime relating to FDI and still binding bureaucratic restraints on FDI approval procedure, so far India has not been successful in attracting export-oriented foreign investors.11 Overall, there is a clear difference between the three Northeast Asian NIEs – South Korea, Taiwan and Hong Kong – and the other countries in terms of the relationship between the share of exports accounted for by MNE affiliates (MNEXS) and the share in total world manufacturing exports (WMSH) (Table 3.2, columns 3 and 5, and Figure 3.1). For the former three countries, the data do not point to any systematic relationship. By contrast, for all the other countries there is a close positive relationship suggesting that the entry of MNEs has been export creating. As a formal test of this relationship, we estimated the following regression: WMSHit 
1 MNEXSit
2WYit
3RERit
dDit


it

(3.1)

As we have already defined, the dependent variable and the explanatory variable of interest are respectively the world market share in manufactured exports (WMSH) and the share of MNE affiliates in total manufactured exports (MNEXS). WY is world income, RER is the real exchange rate, and

Manufacturing for export

65

D is a matrix of the country dummy variables included to capture unobserved country-specific fixed effects,  is the constant term,  is a stochastic error term, and i and t are country and time subscripts respectively. We are interested in testing whether there is a positive relationship between MNEXS and WMSH, when controlled for WY and RER and D, implying that MNE entry into export industries facilitates market penetration without crowding out the performance of purely local firms. WY is measured as the weighted average of the indices of real GDP (1995 100) of the ten major importing countries of each country under study calculated using market shares in 1995 as weights.12 The hypothesis relating to this variable is that market penetration becomes easier under buoyant demand conditions. The RER is measured in a similar fashion using indices of bilateral nominal exchange rate (expressed as domestic currency price of foreign currency), relative price indices (measured as foreign producer price relative to domestic consumer price) in relation to the same ten countries.13 It is used here to control for the implications of change in international competitiveness of the given countries for their export performance. An increase (decrease) in RER reflects an improvement (deterioration) in international competitiveness. Thus the coefficient on RER is expected to be positive. The country dummy variables are specified using Singapore as the base dummy. Equation 3.1 is estimated using five-year average data reported in Table 3.2 on MNEXS and WMSH, and comparable RER and WY series constructed using data obtained from the World Development Indicators database of the World Bank (data on real GDP, bilateral exchange rates and price indices) and the UN Comtrade database. Thus the estimation is based on an unbalanced data panel consisting of 51 observations. The ordinary least squares (OLS) is used as the estimation technique. All variables (other than country dummies) are used in log form so that the coefficients can be directly interpreted as elasticities. The results are reported (without country dummies) in Table 3.3. Equation 1 covers data for all 12 countries. Equation 2 is the estimate obtained after deleting observations relating to the three Northeast Asian NIEs (Hong Kong, South Korea and Taiwan). Both equations comfortably pass the F test for overall statistical significance and the Ramsey RESET test for the functional form. There is no evidence of violation of normality and heteroscadesticity assumptions relating to the error term in both cases. In spite of the desirable statistical properties, the results need to be interpreted cautiously because of the poor quality of the data, in particular obvious intercountry differences in the measurement of the key explanatory variable, MNEXS and the small sample size. In both equations, coefficients on export share accounted for by foreign firms (MNEXS) and world income (WY) carry the expected (positive) sign

66

Multinational enterprises in Asian development

Table 3.3 Determinants of export market penetration: regression results (dependent variable: country share in world manufacturing trade, WMSH) Variable Constant MNE share in exports (MNEXS) World income (WY) Real exchange rate (RER) N R2 F RESET – 2 (1) JBN, 2 – (2) BPHET, 2 – (3)

Equation 1 3.31 (2.67)***
0.96 (4.88)***
0.59 (4.53)***
0.19 (0.66) 51 0.86 25.11 1.66 0.55 2.85

Equation 2 2.45 (2.19)**
1.13 (6.91)***
0.47 (4.25)***
0.05 (0.29) 38 0.90 34.46 2.78 2.77 0.57

Notes: 1. The equations have been estimated using ordinary least squares (OLS). The figure in parentheses underneath each coefficient is the t-ratio of the coefficient computed using White-corrected standard errors. The level of statistical significance is denoted as: *10%, **5% and ***1%. 2. Country intercept dummies are not reported. Test statistics RESET Ramsey test for functional form mis-specification JBN Jarque–Bera test for the normality of residuals HET Breusch–Pagan test for heteroscedasticity.

and are significant at the 1 per cent level. The coefficient on RER is also positive as hypothesized, but is not statistically significant. The coefficient on MNEXS suggests that, on average, 1 per cent increase in the share of foreign firms in total manufacturing exports is associated with a 0.96 per cent increase in the degree of penetration of these countries in world manufacturing markets. The estimated elasticity increases marginally from 0.96 to 1.13 when the three Northeast Asian NIEs are excluded from the sample coverage. In sum, the results support the hypothesis that, when controlled for world demand conditions and unobservable countryspecific effects, the involvement of MNEs has a strong salutary effect on manufactured export expansion. The result for RER runs counter to the widely held view that international competitiveness as measured by the real exchange rate is important in export success. It could well be that the increased involvement of MNEs in export-oriented manufacturing, and in particular the growing importance of the MNE-dominated trade in parts and components, has contributed to a weakening of the link between

Manufacturing for export

67

the real exchange rate and export performance. There is evidence that exchange rate changes are generally of lesser relevance to pricing decisions governing intra-firm trade because of various other considerations impinging on global operations of MNEs (Rangan and Lawrence 1993, 1999; Helleiner 1981).

CONCLUDING REMARKS The evidence assembled in this chapter suggests that the share of MNEs in manufactured exports from developing countries has recorded a significant increase from about the mid-1970s, with the rate of MNE participation in export expansion accelerating over time. MNEs have been responsible for a larger share of exports from latecomers to export-led industrialization in Asia compared with the historical experiences of the East Asian NIEs. Contrary to the historically specific experience of Korea and Taiwan (and also Japan), for latecomer DCs the entry of MNEs is virtually essential for export success. Export-oriented FDI is not a blunt and homogeneous but a complicated and finely differentiated instrument of international economic interaction, whose role changes in line with changes in patterns of international production. In this chapter, we have focused on the direct link between international production and exports operating through FDI. The importance of MNE participation as a key ingredient for export success is actually a good deal higher than that suggested by our estimates, for two main reasons. First, as a number of recent studies have convincingly demonstrated, the presence of MNEs in a given host country has significant positive spillovers on the performance of local firms.14 Second, there is evidence to suggest that a direct MNE presence also provides a conducive setting for the exploitation of gains from other non-equity forms of MNE linkages. For instance, the entry of foreign investors from the NIEs into garments and other labourintensive product sectors in second-tier exporting countries propelled the entry of the latter countries into the global purchasing networks of international buying groups; eventually even non-FDI (local) firms benefit from such market linkages, which develop from what Wells (1994) has aptly termed ‘country reputation’. A key policy inference from our analysis is therefore that, in designing policies of outward-oriented development, investment and trade policies must be considered together as co-determinants of the location of production and patterns of trade. Given the fact that an increasing number of developing countries compete in attracting export-oriented FDI, countries that attempt to implement a selective FDI promotion policy are likely to

68

Multinational enterprises in Asian development

lose important opportunities for export expansion. Of course, enhancing national gains from export-oriented industrialization by encouraging greater participation of local companies is a legitimate objective for any country. But, under the current competitive conditions governing international production, this objective can be achieved only by providing a conducive setting for domestic entrepreneurial development as part of the overall development strategy, not through direct restrictions on the entry and operation of MNEs.

NOTES *

I am indebted to my colleague Hal Hill for his valuable contribution at the formative stage of research leading to this chapter. 1. A prime example is the treatment of foreign investment in India following the liberalization reforms initiated in 1991 (World Bank 2003b, Bajpai and Sachs 2000). Further liberalization of the FDI regime also remains a thorny issue in the ongoing debate on further economic opening in China (Naughton 1996, Lardy 2002). 2. Surprisingly even some strong proponents of open trade and investment policies seem to accept such inferences as of general relevance for all developing countries at all times. For instance, on the issue of whether China’s superior export performance relative to India can be explained in terms of a superior record in attracting FDI, Anne Krueger recently observed that ‘this was not the case in Japan, Hong Kong and Korea’ (Srinivasan 1998, p. 233). 3. The term ‘latecomer exporting countries’ is used here to refer to the developing countries which are gradually shifting from primary commodity specialization into manufactured exports following the example of the East Asian newly industrialized countries (NICs). Two alternative terms used in the literature are ‘new exporting countries’ and ‘second-tier exporting countries’. 4. As Caves (1996, p. 253) aptly put it, ‘[G]iven scale economies and the very small domestic markets of most developing countries, a foreign subsidiary will locate there either to serve the domestic market or to export exclusively, but it will not serve the domestic market and export a little . . . Accordingly, generalizations that span the export and domestic markets are somewhat suspect.’ 5. Through a disaggregation of OECD import data, Yeats (2001) found that the share of fragmentation-based trade (parts and components) accounted for 30 per cent of total manufacturing imports of OECD countries in 1996, compared with around 15 per cent in the mid-1980s. According to estimates reported in Athukorala (2006b), between 1992 and 2000 the share of these products in total world manufacturing exports and manufacturing exports from developing countries increased from 20.7 per cent to 25.4 per cent and from 19.2 per cent to 32.0 per cent respectively. 6. However, in recent years China has emerged as an important location for final assembly in many product lines largely because of the vast domestic market for these products, which naturally reduces the risk of covering the initial establishment costs (Borrus 1997). 7. Some fragments of the production process in certain industries have become ‘standard fragments’ which can be effectively used in a number of products. Examples include long-lasting cellular batteries, originally developed by computer producers and now widely used in cellular phones and electronic organizers; transmitters, which are used not only in radios (as originally intended) but also in PCs and missiles; and electronic chips, which have spread beyond the computer industry into consumer electronics, motor vehicle production and many other product sectors.

Manufacturing for export 8.

9. 10. 11. 12. 13. 14.

69

There are no recent estimates of MNE shares in exports from Latin American countries, but recent evidence on foreign direct investment in the region suggests that MNE involvement in export performance might have increased considerably over the past two decades (Blomström 1990, Fritsch and Franco 1992, UNCTAD 1995). Numerous studies have drawn attention to this phenomenon. See for example Hobday (1995), Koo (1985), Lee (1992), Naya (1990), Ranis and Schive (1985), Schive (1990) and Amsden and Chu (2003). For a fuller discussion on India’s failure to attract MNEs as a major cause of her lacklustre export performance, see Srinivasan 1998. Note that the increase in export share in the late 1980s is consistent with the tightening of import and exchange controls in response to the balance of payments crisis preceding the 1991 liberalization. That is, WYit  兺 in itYit, where  is the export market share in 1995, Y is an index (1995 100) of real GDP, and i and t are country and time subscripts. For details on the construction of this index and the reason behind the choice of the particular price indices in its construction, see Appendix to Chapter 6. For instance, in China MNE affiliates are usually responsible for product design, the supply of needed parts and components, and the sale of goods on the world market of products produced by local firms. In some cases the foreign firm even supplies specialized equipment that is required to assemble the products sold on the world market (Lardy 2002, pp. 6–7). This is also an experience widely observed across many countries with MNE presence in export-oriented industries (Blomström and Kokko 1998; Lipsey 2004).

70

Multinational enterprises in Asian development

APPENDIX 3.1 Table 3.A1

Data sources on MNE share in manufactured exports

Country

Time coverage

Source

Ownership criterion used in identifying MNE affiliates

Hong Kong South Korea

1981–96

Ramstetter (1999)

All firms with FDI

1971 1974, 1975 1977, 1986

Nayyar (1978) Koo (1985) Ramstetter (1993)

10% FDI in share capital

1973–89

Wang (1999), based on Survey of Foreign Direct Investment and Analysis of its Contribution to Taiwan’s Economic Development, various years (in Chinese) Extended using the same data source

All firms with FDI

Taiwan

1990–2000 Singapore

1973–99

Singapore Development Board, Report on the Census of Industrial Production (various years since 1973)

50% FDI in share capital

China

1985–2000

Compiled using data on exports of foreign-invested firms from China Statistical Publishing House, Statistical Yearbook (various years), and export data from International Economic Data Base (IEDB), Australian National University

All firms with FDI

Indonesia

1990–2000

Estimated from unpublished data from the Annual Survey of Manufacturing, Indonesian Bureau of Statistics

All firms with FDI

Malaysia

1975–2000

Estimated using export data from Bank Negara Malaysia, Monthly Bulletin of Statistics, and data on foreign share in total sales from the Survey of Manufacturing (Malaysian Department of Statistics), assuming that MNE export share in a given product category is proportionate to output share1

50% FDI in share capital

ITC (1987)

12.5% FDI in share capital

Philippines 1980 and 1983

Manufacturing for export

Table 3.A1 Country

71

(continued) Time coverage

Source

1985, 1990–2000

Estimated using data reported in Hill (2003)

Ownership criterion used in identifying MNE affiliates

Thailand2

1974, 1975

Tambunlertchai and Ramstetter (1991) 1979, 1986, Ramstetter (1997) 1986, 1990 1994, 1995, Dollar and Hallward- Driemeier 1996 (2000), based on a survey of 1200 randomly chosen plants in five industries – auto parts, electronics, food products, textiles and clothing.

All firms with FDI (firms approved by the Board of Investment)

Vietnam

1990–2000

Athukorala (2002b) (based on official records of Ministry of Investment and Planning, Vietnam)

All firms with FDI

India

1975/6– 1999/20003

Compiled using exports by MNE affiliates from Athreye and Kapur (2001) (1975/6–1994/5) and the data base of Research Information Systems, Delhi (1994/5–1999/2000) (based on the Survey of Public Limited Companies conducted by the Reserve Bank of India) and export data from International Economic Data Base (IEDB), Australian National University.

Public limited liability companies with FDI

Sri Lanka

1975, 1977, Athukorala and Rajapatirana 1978–93 (2000a and 2000b) 1994–2000 Athukorala (2004)

All firms with FDI

Notes: 1. This procedure is likely to result in an underestimation of MNE share in exports because MNE affiliates generally tend to be more export-oriented than local firms. However, the bias is likely to be rather small because over 80% of manufacturing exports originate in fully foreign-owned industries. 2. The estimates for Thailand are based on returns to a periodic survey conducted by the Thai Board of Investment (BOI) of the BOI-approved firms only. Since there is no legal requirement in Thailand for foreign investors to obtain BOI approval, these estimates are likely to understate the contribution of MNE affiliates to export expansion (Ramstetter 1997). 3. The original data on exports by MNEs are based on the Indian fiscal year (from 1 April in the stated (reported) calendar year to 31 March in the next year).

4. Product fragmentation and trade patterns in East Asia* Product fragmentation – the cross-border dispersion of component production/assembly within vertically integrated production processes1 – has been a key factor in the rapid expansion of MNE involvement in the global economy over the past three decades. For example, Japanese car producers export engine parts to their affiliates in Thailand, where they are assembled into engines using some other components procured from other countries in the region and exported back to Japan and to other thirdcountry markets. Computer producers in the USA increasingly rely on worldwide procurement networks spanning the globe for the assembly/ production of most of the components embodied in the final product, and more recently some producers have begun to shift even final stage assembly to some countries in Asia. German camera producers are now engaged only in design and marketing tasks, after shifting most segments of the production process to Japan and other Asian countries. The other products with significant international product sharing include television and radio receivers, optical products, musical equipment, watches, sewing machines, typewriters and other office equipment, aircraft parts, chemicals, pharmaceuticals, synthetic fibres, wearing apparel and travel goods. There is a sizeable theoretical literature examining causes and modalities of international product fragmentation and its implications for trade flow analysis and trade policy.2 Applied trade economists have, however, been rather slow in responding to this new form of international specialization. Trade flow analysis is still based on the traditional notion of a horizontal specialization scenario in which countries trade goods that are produced from start to finish in just one country. This chapter aims to examine the extent, trends and patterns of this new form of international exchange, and its implications for analysing regional trade patterns, with special emphasis on countries in East Asia. The analysis is based on a systematic separation of trade in parts and components from total trade flows, using a new data set extracted from the UN trade database. The East Asian experience is examined in the wider global context, focusing specifically on the comparative experience of that region and the 72

Product fragmentation and trade patterns in East Asia

73

North American Free Trade Agreement (NAFTA) and the European Union (EU). Our work relates to, and builds on, Ng and Yeats (2001 and 2003). Compared with these papers, this chapter offers both more current and more detailed information on the nature, trends and patterns of fragmentation trade. However, its major novelty is in the analysis of the process of product fragmentation for the analysis of intra- versus interregional trade patterns with emphasis on East Asia. There is a vast literature based on the standard trade data analysis that unequivocally points to a persistent increase in intra-regional trade in East Asia (including as well as excluding Japan) from about the early 1980s (for example, Kwan 2001, Drysdale and Garnaut 1997, Frankel and Wei 1997, Petri 1993). This evidence figures prominently in the current debate on forming regional trading arrangements covering some or all countries in East Asia. In this chapter we argue that, in a context where components trade is growing rapidly, the standard trade flow analysis can lead to misleading inferences as to the nature and extent of trade integration among countries, for two reasons. First, in the presence of production fragmentation, trade data are double-counted because goods in process cross multiple international borders before getting embodied in the final product. Thus the total amount of recorded trade could be a multiple of the value of final goods. Second, and perhaps more importantly, trade share calculated using reported data can lead to wrong inferences as to the relative importance of the ‘region’ and the rest of the world for the growth dynamism of a given country/region, even controlling for double counting in trade. This is because intra/extra-regional patterns of trade in parts and components (henceforth referred to as ‘fragmentation trade’) and trade in related final goods (‘final trade’) are unlikely to follow the same patterns. There is indeed ample evidence coming from the case study literature on multinational enterprises operating in the East Asian region that the demand for the final products comes predominantly from the rest of the world, particularly from North America and countries in the EU.3 The chapter begins with a discussion on the procedures followed in extracting data from the UN trade data tapes, and the nature and quality of the data. The next section examines the nature and extent of global trade in components and East Asia’s role in this form of trade specialization. The following section examines the implications of the rapid expansion of product fragmentation for analysing intra- and extra-regional patterns of economic integration in East Asia. The final section presents the key policy inferences.

74

Multinational enterprises in Asian development

DATA This study makes use of data extracted from the UN trade database based on Revision 3 of the Standard International Trade Classification (SITC, Rev 3). In its original form (SITC, Rev 1), the UN trade data reporting system did not provide for separating fragmentation trade (parts and components) from final manufactured goods. The SITC Revision 2, introduced in the late 1970s (and implemented by most countries only in the early 1980s), adopted a more detailed commodity classification, which provided for separation of parts and components within the machinery and transport sector (SITC 7). There were, however, considerable overlaps between some advanced-stage component production/assembly and assembly of final goods in Revision 2 (Ng and Yeats 2001).4 Revision 3, introduced in the mid-1980s, marked a significant improvement over Revision 2. In addition to redressing overlaps within SITC 7, this new version of SITC provided for separation of parts and components trade in the ‘miscellaneous goods’ sector (SITC 8). These two sectors together accounted for around 70 per cent of total world trade during the period under study. SITC Revision 3, despite its significant improvement over the previous version, does not provide for the construction of data series covering the entire range of activities involving production fragmentation. Data reported under SITC 7 do provide a comprehensive coverage of fragmentation trade. But data for SITC 8 do not seem to fully capture fragmentation trade within that commodity category. For instance, for some products such as clothing, furniture and leather products in which outsourcing is prevalent (and perhaps has been increasing), the related components are recorded under other SITC categorizing (such as pieces of textile, parts of furniture, parts of leather soles). The SITC data system does not provide adequate information to separate these components and relate them accurately to the related final product. Moreover, there is evidence that international production fragmentation has been spreading beyond SITC 7 and 8 to other product categories, in particular to pharmaceutical and chemical products (falling under SITC 5) and machine tools and various metal products (SITC 6). Assembly activities in software trade too have recorded impressive expansion in recent years. These are lumped together with ‘special transactions’ under SITC 9. So the measurement of trade in parts and components used in this chapter is presumably downward biased. These factors cause our estimates to be downward biased, and the degree of bias may have increased over the years with the gradual spread of production fragmentation to other types of products beyond SITC 7 and 8. It is important to note that parts and components trade measured using the reported trade data, regardless of the downward bias involved, provides

Product fragmentation and trade patterns in East Asia

75

only a proxy measure of fragmentation trade. On the import side, the data capture both intermediate goods used to make goods for exports and those that are used for domestic consumption. On the export side, fragmentationbased exchange encompasses both parts and components and final goods assembled using imported parts and components. Precise measurement of fragmentation trade requires combining published trade data with data on the input–output structure of trading nations. However, the latter data are not available for a sufficient number of countries to enable us to undertake a more precise analysis of global trade patterns.5 We tabulated data from the UN trade database for the period from 1992 to 2003, the most recent year for which trade data are available for all reporting countries. The year 1992 is used as starting point because by this time countries accounting for over 95 per cent of total world manufacturing trade had adopted the new system. Given the prohibitive cost of tabulation data covering the entire period, 1992, 1996 and 2003 are chosen as the most appropriate interim years for the inter-temporal comparison of trade patterns. The list of parts and components identified at the five-digit level for these two sectors provides the basis of our empirical analysis. 6 It contains a total of 225 five-digit products – 168 products belonging to SITC 7 and 57 belonging to SITC 8.7 For the purpose of our analysis, East Asia is defined to include both Japan and developing countries in the region. The latter include the newly industrialized economies (NIEs) in North Asia (South Korea, Taiwan and Hong Kong), China and members of the Association of South East Asian Nations (ASEAN) Free Trade Area (AFTA). Among the AFTA member countries, only the six largest economies – Indonesia, Malaysia, the Philippines, Thailand, Singapore and Vietnam – are covered in the statistical analysis; Brunei, Cambodia, Laos and Myanmar are ignored because of lack of data. The UN data system does not cover Taiwan (because it is not a UN member). Vietnam has not yet started reporting data under the SITC classification. Singapore has not been reporting data on its bilateral trade with Indonesia from 1964 onwards for political reasons. In these cases import and export data for the given country are based on partner country export and import records respectively. The trade data used here for Hong Kong relate to total reported trade net of entrepot trade. For other countries the official trade data do not permit separating entrepot trade.8

TRENDS AND PATTERNS OF PRODUCT FRAGMENTATION World trade in parts and components9 increased from about $440 billion in 1992 to nearly $1000 billion in 2003 (Table 4.1, Figure 4.1). The share of

76

Table 4.1

Multinational enterprises in Asian development

World trade in parts and components, 1992–2003 (%) Exports 1992

1996

Trade balance2

Imports 2003

1992

1996

2003

East Asia 29.3 38.2 39.2 23.8 30.8 37.9 Japan 15.2 15.5 11.9 3.5 4.8 4.7 Developing East Asia 14.1 22.7 27.3 20.4 26.0 33.2 China 0.8 1.7 6.1 2.7 3.0 10.7 Hong Kong SAR 1.3 0.8 0.2 1.8 1.8 0.5 Rep. of Korea 2.5 3.8 4.4 3.1 3.4 3.6 Taiwan 3.0 4.5 3.4 3.1 2.9 5.6 AFTA 6.5 11.8 13.3 9.8 14.9 12.7 Indonesia 0.1 0.3 0.4 0.9 0.9 0.3 Malaysia 2.2 3.4 3.4 2.8 3.9 3.7 Philippines 0.2 1.2 2.1 0.5 1.5 1.9 Singapore 3.0 5.6 5.7 4.0 6.2 5.0 Thailand 0.9 1.2 1.6 1.7 2.4 1.7 Vietnam – – 0.1 – 0.1 0.1 South Asia 0.1 0.2 0.2 0.7 0.4 0.6 Oceania3 0.1 0.4 0.4 0.3 1.3 1.1 NAFTA 25.9 24.0 22.8 26.0 26.6 23.7 USA 20.8 18.7 17.3 18.0 18.2 15.3 Canada 3.6 3.3 2.8 6.0 5.4 4.2 Mexico 1.5 1.9 2.9 2.0 3.0 4.1 MERCOSUR 0.6 0.6 0.5 1.1 1.6 1.1 Andean Pact – – – 0.5 0.3 0.2 Europe 46.1 53.9 38.4 47.6 51.2 36.9 EU 43.8 38.0 34.2 44.3 34.8 32.1 Eastern Europe 0.3 0.9 2.5 0.5 1.4 3.0 Rest of Europe 2.0 1.7 1.7 2.8 2.2 1.9 World 100.0 100.0 100.0 100.0 100.0 100.0 $ billion1 438.9 728.6 986.9 447.0 735.1 983.4 Developed countries Developing countries

85.7 14.3

75.7 24.3

68.7 31.3

74.8 25.2

67.4 32.6

1992 25.9 79.2 31.2 196.6 25.3 9.5 5.6 36.9 721.5 12.0 128.2 22.7 72.4 – 538.9 173.8 8.4 21.3 53.9 21.7 67.3 – 6.1 7.9 52.1 23.2

1996

2003

18.6 3.6 68.5 60.9 15.7 21.4 77.1 76.5 112.2 184.5 9.8 17.4 35.4 65.0 27.6 4.4 211.5 25.3 16.0 9.0 29.8 10.3 11.2 13.3 99.0 6.3 – 0.4 107.7 198.9 237.4 149.0 11.6 3.3 1.7 12.0 63.6 53.3 58.5 43.9 176.9 99.2 – – 4.3 4.2 7.6 6.6 61.5 16.2 28.3 12.1

59.2 40.8

Notes: 1. By definition value of exports and imports for a given year should be identical. The minor differences seem to reflect recording errors and differences in measurement arising from the use of CIF (cost, insurance and freight) price for reporting imports and FOB (free on board) price for reporting most exports. 2. Trade balance as a percentage of exports. 3. Australian and New Zealand. – Zero or negligible. Source: Compiled from the UN Comtrade database.

Share (%)

55

1000

50

900

45

800

40

700

35

600

30

500

25

400

20

300

15

200

10

P&C ($ billion)

77

Product fragmentation and trade patterns in East Asia

100 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Share of P&C in total manufacturing (left scale) Share of P&C in machinery and transport equipment (left scale) P&C, $ billion (right scale)

Source: Based on data compiled from the Comtrade database (exporter records).

Figure 4.1 World trade in parts and components (P&C), 1992–2003 these products in total world manufacturing exports increased from 18.5 to 22.0 per cent between 1992 and 2003. Components accounted for nearly a third of the total increment in world manufacturing exports between these two years. Developed countries account for the bulk of component trade (Table 4.1; Figure 4.2). However, the share of developing countries has increased sharply over the years (from 14.3 to 31.3 per cent on the export side and from 25.2 to 40.8 per cent on the import side between 1992 and 2003). The share of East Asia (including Japan) in total world exports of components increased persistently from 29.3 per cent in 1992 to 39.2 per cent in 2003. This is despite a notable decline in the share accounted for by Japan, the dominant economy in the region, in recent years. The share of developing East Asia (East Asia excluding Japan) increased from 14.1 to 27.3 per cent between these two years. Within the group, all reported countries have recorded increases in world market shares. The growing importance of China in component trade is particularly noteworthy. The share of China in total world component exports increased from less than 1 per cent to 6.1 per cent and in total imports from 2.7 to 10.7 per cent between 1992 and 2003. Contrary to the popular perception of ‘crowding out “the rest” by China’, this increase has been within an overall increase in exports from other newcomers in the region. For instance, the combined export share of the six main member countries of the ASEAN Free Trade Area (AFTA) more than doubled (from 6.5 to 13.3 per cent) between these two years. The

78

Multinational enterprises in Asian development 1000 900

Exports (US$ billion)

800 700 600 500 400 300 200 100

East Asia

Developing East Asia

Developing countries

World

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

0

Developed countries

Source: Based on data compiled from Comtrade database (exporter records).

Figure 4.2 Parts and components exports by major country groups, 1992–2003 (US$ billion) share of Hong Kong in world component trade has eroded persistently as a result of the dramatic relocation of manufacturing ventures to mainland China during this period. The component trade of Vietnam is also growing, but it still accounts for only a small fraction of regional trade. Japan’s share in world component trade on the import side is much smaller than that on the export side. The former has remained largely unchanged around 5 per cent. When Japan in excluded, market share changes of individual East Asian countries on the import side closely mirror those on the export side. Japan has persistently recorded a large surplus in component trade, reflecting the heavy involvement of Japanese companies in overseas assembly activities (mostly in East Asia) to maintain their competitiveness in final trade in third-country markets (Table 4.1, last three columns). The EU and the USA too have recorded modest surpluses. By contrast, China has been a major net importer.10 This comparison is generally consistent with the hypothesis that high-wage countries are the major suppliers of components (production of which is generally more capital intensive) while low-wage countries have a comparative advantage in assembly activities (which are labour intensive) (Ng and Yeats 2003). However, among the individual East Asian countries, the trade positions of the Philippines and Singapore have turned from deficits to surpluses in recent years. As we will see later, this

Product fragmentation and trade patterns in East Asia

79

reflects growing cross-border trade in components within the region. Regional production-sharing networks interact and supply one another, leading to an expansion of two-way trade in components. Has the formation of NAFTA and the integration of some of the new countries emerging from the former Soviet Union with the rest of Europe adversely affected developing East Asia’s relative position in world assembling activities? Indeed, proximity to industrial countries and relatively low wages by regional standards (though not compared with some of the East Asian countries) can be considered as added advantages of these countries compared with East Asian countries in production-fragmentation-based international specialization (Egger and Egger 2005, Ng and Yeats 2003, USITC 1999, Kierzkowski 2001). The data do not, however, point to any dampening effect of exports from these countries on the relative world market position of East Asia. The world market shares of Mexico and the rest of Europe (EUT less EU) have increased, but at a much slower rate than that of developing East Asia. It seems that, in spite of geographical proximity and tariff concessions under FTAs, US producers still find East Asia a more attractive location for outsourcing. A new dimension of regional production sharing in Europe has been added by the economic integration of Europe. One possible explanation is that, despite rapid industrialization for over a decade, the average manufacturing wage in East Asia still remains low compared with Europe and North America (Table 4.2). Moreover, significant differences in wages among the countries in the region have provided the basis for rapid expansion of an intra-regional product-sharing system, giving rise to increased cross-border trade in components. Substantial wage differentials among countries in the region naturally make intra-regional specialization especially profitable. Also, as firstcomers in this area of international specialization, countries in East Asia (particularly Singapore, Korea, Taiwan and Malaysia) offer considerable agglomeration advantages for companies that are already located there. Location decisions of MNEs operating in assembly activities are strongly influenced by the presence of other key market players in a given country (Barry and Bradley 1997, Ruane and Gorg 2001). Also, there is a general tendency for MNE affiliates to become increasingly embedded in host countries the longer they are present there and the more favourable the overall investment climate of the host country becomes over time. They may respond sluggishly to relative cost changes once they have invested substantial resources in domestic production facilities and in establishing information links (Rangan and Lawrence 1999). Table 4.3 presents comparative statistics on the share of components in total manufacturing exports and imports in 1992, 1996 and 2003 and their contribution to the growth of manufacturing trade between these years. It

80

Multinational enterprises in Asian development

Table 4.2 Annual compensation1 per worker in manufacturing, 1990–98 ($ per month)

All countries Canada Europe Latin America Mexico East Asia China Hong Kong India Indonesia Japan Korea Malaysia Philippines Singapore2 Taiwan Thailand Oceania Australia New Zealand

1990

1991

1992

1993

1994

1995

1996

1997

1998

2257 3156 3074 903 579 1788 178 774 348 388 3808 1134 387 314 810 1162 343 1982 2331 1633

2408 3343 3283 920 666 1950 217 866 315 409 4108 1449 420 370 911 1337 428 2142 2446 1838

2529 3350 3600 1009 765 1963 213 855 323 436 4373 1346 488 467 1067 1639 502 – 2412 –

2518 3432 3388 1146 843 2141 317 n.a. 335 414 4902 1635 570 499 1264 1638 569 2304 2377 2230

2560 3589 3503 1175 854 2129 293 850 318 483 5723 1743 573 565 1477 1695 508 – 2517 –

2652 3553 3724 1147 749 2115 395 844 427 466 5934 2059 569 580 1424 1914 523 – 2694 –

2693 3570 3728 1253 746 2138 428 1219 463 475 5849 2313 665 647 1623 n.a. 568 3082 2520 3645

2578 3506 3616 1343 831 1883 513 1077 461 459 5343 2351 657 623 1469 2032 508 3024 2400 3647

2539 3413 3618 1284 – 1787 533 1178 474 395 5054 2007 652 598 1720 1970 472 2907 2334 3480

Notes: 1. Salary/wage plus other remuneration (including superannuation contribution) paid by foreign affiliates of US MNEs. 2. Covers only workers in industrial machinery sector. – Data suppressed in the original source for confidentiality. Source: Compiled from Survey of U.S. Direct Foreign Investment Abroad, Bureau of Economic Analysis, Department of Commerce, Washington DC (various issues).

is evident that the degree of concentration of East Asian countries as a group on component trade is much higher than that of all other regions in the world. In 2003, components accounted for 27.5 per cent of total manufacturing exports from developing East Asia, compared with the world average of 22.0 per cent, 16.7 per cent for the EU and 25.6 per cent for NAFTA. Of the total increment in manufactured exports from East Asia between 1992 and 2003, over a third came from components exports. The comparable figures for the EU and NAFTA were 17.9 and 25.9 per cent respectively. Within East Asia, countries belonging to AFTA, in particular Malaysia, the Philippines, Singapore and Thailand, stand out for their heavy dependence on production fragmentation for export dynamism. In

81

Exports

East Asia Japan Developing East Asia China Hong Kong SAR Rep. of Korea Taiwan AFTA Indonesia Malaysia Philippines Singapore Thailand South Asia Oceania NAFTA USA Mexico MERCOSUR Andean Pact

128.6 66.5 62.0 3.6 5.7 11.4 12.8 28.5 0.6 10.0 0.8 13.0 4.1 0.6 0.3 113.5 91.0 6.7 2.5 0.2

1992 278.0 106.1 165.0 9.8 6.1 26.7 22.4 82.9 1.7 23.5 8.8 39.4 9.5 1.0 4.0 182.4 146.2 13.1 3.8 0.2

1996

Value of P&C ($ billion)

386.4 117.4 269.0 59.5 1.8 43.4 33.6 131.2 4.3 33.9 20.7 56.5 15.8 2.6 3.8 225.5 170.6 28.0 4.9 0.4

2003 20.3 21.2 19.3 5.5 20.2 17.1 28.3 24.7 3.7 38.7 19.8 27.0 19.1 2.6 13.2 25.3 26.8 20.7 13.1 5.0

1992 31.0 30.2 30.2 9.8 23.8 25.2 28.8 35.0 7.4 42.6 52.5 39.7 23.4 4.9 18.9 27.2 30.5 19.4 13.5 4.1

1996 27.5 27.9 27.3 15.2 12.3 25.5 39.5 40.6 13.9 42.7 63.8 46.7 26.7 4.0 17.3 25.6 29.2 21.1 12.2 5.6

2003

Share of P&C in mfg. exports (%)

Parts and components (P&C) in manufacturing trade

Country/region

(a)

Table 4.3

3.19 1.17 4.52 7.32 2.57 3.77 2.53 4.15 2.58 4.53 8.57 3.70 4.08 4.17 9.37 2.70 2.17 5.73 2.98 2.32

1992–2003

Growth of mfg. exports (%)

4.44 2.27 5.97 11.71 4.45 5.42 3.88 6.21 8.09 4.94 13.71 5.97 5.47 5.96 10.54 2.75 2.51 5.81 2.69 2.77

1992–2003

Growth of P&C exports (%)

33.5 47.5 31.2 17.1 – 30.9 52.2 49.4 25.1 44.6 70.1 59.7 31.0 4.8 17.8 25.9 32.5 21.2 11.4 6.4

1992–2003

Contribution of P&C to growth of mfg. exports (%)

82

(continued)

East Asia Japan Developing East Asia China Hong Kong SAR Rep. of Korea Taiwan AFTA Indonesia Malaysia Philippines

Imports

95.3 14.1 81.4 10.6 7.2 12.2 12.4 39.2 3.6 11.0 1.9

375.9 63.0

Developed countries Developing countries

(b)

202.4 192.1 1.5 8.8 438.9

1992

Europe EU-15 Eastern Europe Rest of Europe World1

Country/region

Table 4.3

226.5 31.3 190.9 20.8 13.0 22.6 27.5 102.4 6.7 27.1 10.9

552.4 176.2

277.5 258.5 6.9 12.2 728.6

1996

Value of P&C ($ billion)

372.6 46.2 326.7 105.3 5.1 35.5 55.5 125.5 3.1 36.5 19.1

677.5 309.4

379.5 337.5 25.4 16.7 986.9

2003

21.4 14.2 23.5 17.6 28.1 25.2 16.9 28.2 18.5 35.2 24.8

18.6 18.0

15.5 15.9 9.1 11.1 18.5

1992

31.3 19.3 34.0 21.1 31.6 27.4 35.0 39.3 23.8 47.5 43.6

20.4 22.0

16.2 17.7 13.5 12.3 20.8

1996

35.3 21.5 38.9 34.3 44.2 33.6 37.3 47.1 18.5 55.7 63.1

20.2 27.5

16.6 16.7 20.2 11.7 22.0

2003

Share of P&C in mfg. exports (%)

3.46 3.09 3.55 6.64 3.10 3.13 2.83 2.60 0.59 2.97 5.57

2.02 4.72

2.24 2.05 8.35 2.35 2.54

1992–2003

Growth of mfg. exports (%)

5.53 4.80 5.64 9.49 1.35 4.31 6.10 4.70 0.59 4.85 9.54

2.35 6.49

2.51 2.25 11.82 2.56 3.25

1992–2003

Growth of P&C exports (%)

45.6 27.8 49.8 38.4 – 40.7 57.1 67.7 18.5 74.4 76.1

22.6 31.8

18.1 17.9 21.9 12.4 26.0

1992–2003

Contribution of P&C to growth of mfg. exports (%)

83

1.9 11.1 447.0

299.0 148.0

Eastern Europe Rest of Europe World1

Developed countries Developing countries

462.4 272.7

10.1 14.9 735.1

50.3 20.9 3.8 9.0 183.9 126.0 20.7 11.0 2.0 267.0 242.0

582.5 400.9

28.8 18.7 983.4

49.6 17.2 6.8 10.0 232.7 150.3 41.1 10.5 2.1 363.4 315.9

16.1 30.1

9.2 12.6 19.0

30.0 24.7 14.1 14.6 18.9 17.5 18.7 17.9 12.6 15.0 15.3

18.4 25.1

14.3 12.5 20.4

42.8 32.9 14.6 15.2 23.6 21.7 30.6 17.2 9.7 16.6 18.9

17.0 28.2

20.3 12.3 20.3

49.2 32.5 12.6 12.1 17.7 15.4 28.7 21.5 9.1 17.4 17.6

2.45 4.28

7.91 2.18 2.90

2.55 2.61 3.74 9.92 3.49 3.47 4.89 2.48 1.11 1.99 1.74

2.67 4.01

11.33 2.08 3.16

4.57 3.73 3.28 9.11 3.22 2.95 6.67 3.22 0.18 2.59 2.31

18.1 27.2

22.2 11.9 21.5

70.8 41.0 11.6 11.8 16.8 13.9 33.0 25.7 1.8 21.1 21.8

Source: Compiled from the UN Comtrade database.

Notes: 1. By definition value of exports and imports for a given year should be identical. The minor differences seem to reflect recording errors and differences in measurement arising from the use of CIF price for reporting imports and FOB price for reporting exports. – Zero or negligible.

16.0 6.8 3.0 1.1 104.2 71.9 8.0 4.7 2.2 190.3 177.3

Singapore Thailand South Asia Oceania NAFTA USA Mexico MERCOSUR Andean Pact Europe EU-15

84

Multinational enterprises in Asian development

2003, components accounted for 40.6 per cent of total manufacturing exports in AFTA, up from 24.7 per cent in 1992. Between these two years, the share of components in total manufacturing exports almost tripled in China (from 5.5 to 15.2 per cent). Interestingly, even for Taiwan and Korea, the relative importance of components in total manufacturing exports (and imports) has increased over the years, contradicting the popular belief that these countries have shifted palpably from component production to final goods production. In all countries/regions, component trade is heavily concentrated in the machinery and transport equipment sector (SITC 7) (Table 4.4). This sector accounts for over 90 per cent of the combined component trade of SITC 7 and SITC 8 (miscellaneous manufacturing). Within SITC 7, both component exports and imports of East Asia are heavily concentrated in the electronics and electrical industries. Semiconductors and other electronics components (components within SITC 776) alone accounted for 40 per cent of components exports from East Asia in 2003. Adding to these items components of telecommunication equipment (SITC 764) and office and automated data processing machines (SITC 759) increases the concentration ratio to almost 90 per cent of total exports of components. The balance consists largely of electrical machinery (SITC 778) and auto parts (SITC 784). The degree of concentration of component trade on electronics is much larger in AFTA (over 60 per cent) compared with the regional average. These electronics and electrical products are also the major areas of activity in other countries/regions. But the trade patterns of these countries/regions are characterized by a greater presence of other items such as engines and motors (SITC 714), specialized industrial machinery (SITC 728) and internal combustion machines (SITC 713), for which transportation cost is presumably an important consideration for production location. Overall, these differences are consistent with East Asia’s competitive edge in component specialization in the electrical and electronics industries.

PRODUCT FRAGMENTATION AND TRADE PATTERNS We observed in the previous section that product fragmentation has been of growing importance for trade expansion in East Asia relative to the overall global experience as well as experiences of countries in other major regions. The purpose of this section is to examine intra-regional trade patterns in East Asia by explicitly taking into account this important development in the economic landscape.

85

7 713 714 723 724 728 735 737 741 742 743 744 745 749 759 764 771 772 776 778 784 785

SITC

(a) Exports Machinery and Transport Equipment Internal combust piston engines Non-electrical engines and motors Civil engineering equipment Textile and leather machinery Specialized industrial machinery Metal working machine tools Metal working machinery Heating & cooling equipment Pumps for liquid elevator Gas compressors and fans Mechanical handling equipment Non-electrical machine tools and appliances Non-electrical machine parts and accessories Office & audio data processing machines Telecommunication equipment Electrical machinery Apparatus for switching/protecting circuits Semiconductors & other electronic components Electrical machinery and apparatus, n.e.c. Motor vehicle parts/accessories Motor cycles/cycle parts/accessories 96.1 1.0 0.1 0.8 0.4 0.8 0.3 0.2 0.6 0.2 0.2 0.3 0.3 0.2 18.3 12.4 0.6 5.8 40.0 3.2 7.4 0.9

East Asia

94.4 2.2 0.3 0.8 0.5 1.4 0.6 0.4 0.7 0.4 0.4 0.5 0.3 0.3 12.3 8.3 0.3 8.2 31.0 4.0 17.5 1.3

Japan

96.7 0.5 0.1 0.8 0.3 0.6 0.2 0.1 0.6 0.2 0.1 0.2 0.3 0.2 20.6 14.0 0.7 4.9 43.6 2.9 3.3 0.8

Dev. East Asia 94.8 0.3 0.0 0.4 0.4 0.4 0.2 0.2 0.6 0.2 0.1 0.3 0.5 0.1 28.0 21.8 1.2 5.8 24.8 4.3 2.1 1.1

98.8 0.7 0.0 1.0 0.5 1.2 0.2 0.1 0.7 0.1 0.1 0.2 0.1 0.1 19.1 14.5 0.6 2.1 44.0 3.1 8.5 0.1

98.1 0.5 0.1 1.1 0.1 0.6 0.1 0.1 0.4 0.1 0.1 0.1 0.1 0.2 17.1 5.3 0.2 4.5 61.1 1.3 2.4 0.6

93.8 3.2 1.4 2.9 0.2 1.8 0.5 0.3 1.3 0.6 0.6 0.9 0.7 0.3 10.0 6.0 0.3 6.9 22.8 2.4 20.8 0.2

97.0 9.5 1.8 2.4 0.3 0.4 0.1 0.1 4.8 0.3 0.3 0.9 0.6 0.6 0.2 12.7 0.5 29.2 13.2 10.8 5.9 0.2

94.5 4.3 1.4 1.7 0.9 3.0 0.7 0.7 1.3 1.0 0.8 1.4 1.6 0.9 9.3 5.4 0.5 9.5 13.2 2.4 22.2 0.7

94.6 2.7 1.0 1.6 0.6 1.9 0.6 0.4 1.0 0.6 0.5 0.9 0.9 0.5 13.1 8.2 0.5 7.5 25.4 2.7 16.1 0.7

China Korea AFTA NAFTA2 Mexico EU3 World
HK

Table 4.4 Percentage composition of parts and components exports and imports by three-digit SITC categories,1 2003

86

0.0 1.1 1.0 3.9 0.2 1.1 2.6 100 399 96.0 1.3 0.3 0.9 0.4 0.1 1.1 0.3 0.5 0.2 0.3 0.2 0.3

Railway vehicles & associated equipment Aircraft and associated equipment Other Miscellaneous Manufacturing Furniture, bedding and mattresses Measuring/checking/analysing equipment Other 4

Total $ billion

(b) Imports Machinery and Transport Equipment Internal combust piston engines Non-electrical engines and motors Civil engineering equipment Textile and leather machinery Printing/book binding machinery Specialized industrial machinery Metal working machine tools Heating and cooling equipment Pumps for liquid elevator Gas compressors and fans Mechanical handling equipment Non-electrical machine tools and appliances

791 792

7 713 714 723 724 726 728 735 741 742 743 744 745

8 821 874

East Asia

(continued)

SITC

Table 4.4

91.5 1.4 0.6 0.7 0.5 0.3 1.4 0.5 0.8 0.3 0.6 0.4 0.6

100 114

0.1 1.3 1.1 5.6 0.4 2.6 2.6

Japan

96.7 1.2 0.2 0.9 0.4 0.1 1.0 0.3 0.4 0.2 0.2 0.2 0.3

100 286

0.0 1.0 0.9 3.3 0.2 0.6 2.5

Dev. East Asia

96.9 0.9 0.1 0.6 0.5 0.1 0.6 0.2 0.4 0.1 0.2 0.2 0.3

100 122

0.1 0.5 1.4 5.1 0.2 0.5 4.4

95.7 2.3 0.2 0.1 0.3 0.1 1.6 0.5 0.3 0.3 0.4 0.2 0.3

100 434

0.0 0.5 1.1 1.2 0.2 0.3 0.8

97.3 1.2 0.5 1.4 0.2 0.1 1.3 0.3 0.4 0.2 0.2 0.2 0.2

100 132

0.0 1.5 0.4 1.9 0.1 0.7 1.1

93.5 4.3 1.0 1.2 0.3 0.2 1.4 0.5 1.2 0.7 0.5 0.9 0.9

100 227

0.5 7.1 2.1 6.2 2.3 2.3 1.6

95.4 3.5 1.6 0.4 0.3 0.1 0.8 0.3 0.9 0.3 0.3 0.2 0.7

100 17

0.3 0.7 1.2 2.9 0.0 1.4 1.5

93.8 4.1 1.2 1.4 0.5 0.4 1.8 0.7 1.2 0.8 0.6 1.4 1.3

100 340

0.8 6.5 4.5 5.5 1.1 2.1 2.3

94.6 3.3 0.9 1.2 0.5 0.3 1.5 0.5 1.0 0.6 0.5 0.8 0.8

100 1047

0.5 4.3 2.7 5.4 1.3 1.8 2.3

China Korea AFTA NAFTA2 Mexico EU3 World
HK

87

Total $ billion

Office and audio data processing machines Telecommunication equipment Apparatus for switching/protecting circuits Semiconductors & other electronic components Electrical machinery and apparatus, n.e.c. Motor vehicle parts/accessories Motor cycles/cycle parts/accessories Railway vehicles & associated equipment Aircraft and associated equipment Other Miscellaneous Manufacturing Furniture, bedding and mattresses Measuring/checking/analysing equipment Other4 100 366.6

13.6 9.7 5.9 50.3 2.4 4.4 0.3 0.1 1.3 2.1 4.0 0.3 1.3 2.3 100 46.3

16.1 10.4 5.6 37.6 1.8 6.6 0.5 0.1 2.6 2.0 8.5 1.2 3.6 3.7 100 320.3

13.2 9.6 5.9 52.2 2.5 4.1 0.3 0.1 1.1 2.2 3.3 0.1 1.0 2.1

Source: Compiled from the UN Comtrade database.

Notes: 1. 5-digit SITC parts and components items aggregated at the 3-digit level. 2. Including Mexico. 3. Excluding new Eastern European member countries. 4. Mostly parts of clothing, optical equipment, watches and clocks and sport goods. n.e.c. Not elsewhere classified.

8 821 874

759 764 772 776 778 784 785 791 792

100 170.9

15.7 13.5 6.5 46.5 2.8 3.8 0.2 0.2 0.6 2.7 3.3 0.1 0.6 2.5

15.9 4.1 4.6 57.3 1.5 3.8 0.5 0.1 1.8 1.4 2.7 0.1 1.1 1.5

100 100 35.5 125.6

5.6 6.8 5.7 59.5 3.5 5.0 0.1 0.1 0.8 1.9 4.3 0.3 2.0 2.0 100 232.9

13.8 6.6 8.1 16.4 2.7 26.0 0.6 0.5 2.7 3.0 6.4 2.9 1.7 1.8

13.4 5.6 7.3 14.7 2.9 22.1 1.1 0.7 5.8 4.8 6.2 1.7 2.0 2.6

13.5 7.4 7.4 27.9 2.7 15.9 0.7 0.5 3.3 3.5 5.4 1.4 1.6 2.4 100 100 100 41.2 317.4 1046.7

11.5 8.4 13.4 23.8 3.7 22.1 0.3 0.4 0.1 2.2 4.6 1.4 1.4 1.8

88

Multinational enterprises in Asian development

The discussion is based on data on regional bilateral trade flows reported in Tables 4.5–4.7. Table 4.5 depicts regional patterns of trade in total manufacturing and components. Table 4.6 reports percentage shares of components in bilateral flows of total manufacturing exports. Data on intra-regional shares of trade in total manufacturing, components and final goods for East Asia, the EU and NAFTA are reported in Table 4.7. A number of interesting inferences emerge from a careful comparative treatment of data presented in these tables. In terms of the conventionally used trade data, intra-regional manufacturing trade (export + imports) in East Asia is significant and growing rapidly. The share of total intra-regional trade in East Asia increased from 44.1 per cent in 1992 to 53.2 per cent in 2003. Intra-regional trade in developing East Asia increased from 35.1 to 40.1 per cent between these two years. For AFTA the magnitude of these figures is much smaller, but they point to an impressive, persistent increase over the years from 16.6 to 21.3 per cent. Interestingly, in all three years (1992, 1996 and 2003) for which data are reported in Table 4.5, the degree of intra-regional trade in East Asia (including Japan) is significantly larger than the comparable estimates for NAFTA (East Asia: 47.1 per cent, 51.5 per cent and 52.3 per cent; NAFTA: 39.2 per cent, 43.1 per cent and 46.3 per cent). The estimates for the EU are larger in magnitude (70.8 per cent, 61.0 per cent, 58.2 per cent), but they have been falling over the years, unlike in East Asia. Overall, in terms of conventionally used trade data the ongoing process of regional integration in East Asia looks remarkable. However, these figures need to be interpreted cautiously for two reasons. First, unlike in the EU and NAFTA, the East Asian intra-regional trade ratio camouflages a significant asymmetry in regional trade patterns on the import and export sides. In 2003, intra-regional import flows amounted to 65.6 per cent of total manufacturing imports of East Asia, up from 55.2 per cent in 1992. Intra-regional share in total regional exports was significantly lower: 36.6 per cent in 1992 and 45.6 per cent in 2003. In other words, the region is much more heavily dependent on extra-regional trade for its growth dynamism than is (misleadingly) suggested by the total regional trade share, and this dependence has remained virtually unchanged for the last decade. This imbalance in intra-regional trade is largely a reflection of the unique nature of the involvement of Japanese MNEs in fragmentation trade in East Asia. As already noted, Japan’s trade relations with the rest of East Asia are predominantly in the form of using the region as an assembly base for meeting demand in the region and, more importantly, for exporting to the rest of the world. The data on bilateral trade balances with individual countries (measured as a percentage of exports) show that Japan has persistently maintained a trade surplus with

89

8.6 11.5 10.1 2.7 8.8 8.7 8.8 11.1 10.0

Developing 1992 44.0 East Asia (DEA)2 1996 46.8 2003 47.3 Greater 1992 56.4 China (GCH) 1996 46.2 2003 39.1 AFTA 1992 36.8 1996 45.0 2003 48.0

1992 1996 2003

EU

5.8 8.0 6.7

1992 16.3 1996 18.6 2003 14.9

NAFTA

Japan

1.9 2.4 1.7

6.5 6.8 4.4

4.7 7.4 7.4 0.0 0.0 0.0

1992 1996 2003 1992 1996 2003

36.6 43.8 45.6 25.1 34.4 35.9

East Asia

Total manufacturing exports (X)

B1

Parts and components exports (X)

3.9 5.7 5.0

9.8 11.8 10.5

35.5 35.3 37.2 53.8 37.4 30.4 28.0 33.9 38.0

31.9 36.5 38.2 25.1 34.4 35.9

1.5 2.1 2.6

3.0 3.2 3.8

23.2 19.0 23.2 45.3 26.5 19.0 7.1 8.2 13.5

17.1 16.4 22.2 9.0 10.7 17.8

1.8 2.6 1.6

4.6 5.6 4.5

11.0 14.4 11.2 6.5 7.5 6.9 19.3 23.6 21.3

11.5 15.9 11.6 11.2 17.0 11.5

8.2 8.4 11.4

44.6 48.0 55.2

25.9 24.1 23.7 19.1 25.9 27.7 27.2 23.5 20.7

30.3 27.6 25.8 32.7 30.8 28.7 100 100 100 100 100 100 100 100 100

100 100 100 100 100 100 55.7 57.3 68.2 80.5 83.0 95.6 47.3 51.8 60.0

44.4 52.3 64.9 30.2 39.8 49.6

64.1 100 59.4 100 50.8 100

6.3 9.6 9.9

20.2 100 18.6 16.0 100 23.9 14.4 100 23.2

17.1 16.0 15.4 14.7 18.8 20.9 19.7 16.0 14.2

19.6 16.6 15.7 20.8 16.2 14.9

1.2 1.6 1.4

5.9 7.3 4.5

8.8 10.6 9.6 28.1 31.7 26.5 9.0 9.5 9.3

4.2 6.3 7.0 0.0 0.0 0.0

5.1 7.9 8.5

12.7 16.6 18.8

46.9 46.6 58.7 52.4 51.3 69.1 38.3 42.3 50.6

40.3 45.9 57.9 30.2 39.8 49.6

1.8 2.3 4.1

2.6 3.2 5.2

19.5 16.4 34.0 25.0 15.1 16.0 7.1 7.7 17.8

13.3 13.9 31.8 6.7 9.2 23.6

2.7 4.4 3.5

7.4 9.8 10.0

25.3 27.8 20.5 10.6 14.7 23.6 29.3 32.6 28.3

22.0 27.5 20.7 16.7 23.8 18.3

9.2 10.1 10.1

46.2 43.5 48.6

26.2 23.7 14.4 14.6 14.7 8.6 28.3 22.6 14.5

34.1 29.1 17.7 36.6 32.9 24.0

100 100 100 100 100 100 100 100 100

100 100 100 100 100 100

62.1 100 55.9 100 49.1 100

18.8 100 17.0 100 13.8 100

12.3 13.7 11.4 14.8 13.7 11.0 13.7 14.3 12.4

15.7 14.3 12.1 16.6 13.5 12.5

EA Japan DEA GCH AFTA NAFTA EU Total EA Japan DEA GCH AFTA NAFTA EU Total

A1

Direction of parts and components trade (%)

(EA)1

Table 4.5

90

Developing 1992 East Asia (DEA) 1996 2003 Greater 1992 China 1996 (GCH) 2003 AFTA 1992 1996 2003

Japan

1992 1996 2003 1992 1996 2003

A1

Total manufacturing exports (X)

B1

Parts and components exports (X)

5.8 6.2 8.5

4.7 6.6 4.9

20.6 21.2 23.5

42.9 100 18.9 37.6 100 22.9 32.9 100 32.3

Total manufacturing imports (M)

12.0 14.9 15.3 B2

3.1 4.4 4.0

4.9 1.5 14.4

8.7 9.9 10.9

25.0 48.3 21.0

37.3 100 13.8 100 26.8 100

Parts and components imports (M)

15.8 18.5 28.3

61.9 59.9 66.9 78.9 71.7 74.2 52.7 53.8 58.7

27.1 25.5 22.8 34.1 25.2 26.1 29.0 26.3 20.0

55.2 21.2 55.7 19.7 65.6 18.1 29.0 0.0 36.8 0.0 49.5 0.0 34.8 34.4 44.1 44.7 46.5 48.0 23.7 27.5 38.7

34.1 35.9 47.0 29.0 36.8 49.5 21.3 16.7 20.1 28.0 25.1 22.8 4.8 5.8 12.0

19.5 17.3 22.7 11.9 17.4 29.8 9.5 12.3 16.2 4.3 6.4 5.9 14.4 17.1 21.4

9.7 13.1 16.4 9.5 13.8 14.0 16.6 17.6 13.6 8.4 10.8 8.4 18.5 18.1 16.6

20.4 20.7 15.3 30.9 29.0 20.2 16.3 17.5 14.3 18.1 25.0 22.7 17.6 17.4 12.5

19.2 19.1 15.5 27.3 22.5 18.2 100 100 100 100 100 100 100 100 100

100 100 100 100 100 100 64.5 62.9 70.7 80.5 83.0 95.6 60.4 59.4 59.6

33.1 31.0 24.9 28.1 31.7 26.5 31.2 28.0 19.7

59.3 28.3 59.7 26.3 69.7 22.0 26.9 0.0 37.0 0.0 52.8 0.0 31.4 31.9 45.8 52.4 51.3 69.1 29.2 31.4 39.9

31.0 33.3 47.7 26.9 37.0 52.8

10.8 7.7 13.6 25.0 15.1 16.0 3.9 4.5 9.9

9.8 7.9 14.5 4.0 8.3 18.7

16.8 18.7 24.4 10.6 14.7 23.6 21.0 21.5 24.8

16.5 18.8 24.6 13.7 17.2 21.4

20.4 21.3 15.7 14.6 14.7 8.6 20.4 20.9 19.8

25.2 25.2 17.0 49.4 42.9 23.2

13.2 13.6 10.9 14.8 13.7 11.0 12.5 12.7 10.0

13.6 13.2 10.9 14.6 10.3 9.4

100 100 100 100 100 100 100 100 100

100 100 100 100 100 100

EA Japan DEA GCH AFTA NAFTA EU Total EA Japan DEA GCH AFTA NAFTA EU Total

A2

3.4 4.4 3.7

EA Japan DEA GCH AFTA NAFTA EU Total EA Japan DEA GCH AFTA NAFTA EU Total

1992 15.4 1996 19.3 2003 19.0

(continued)

East Asia (EA)

World

Table 4.5

91

1992 1996 2003 1992 1996 2003 1992 1996 2003

Developing East Asia (DEA)2 Greater China (GCH) AFTA

Japan

1992 1996 2003 1992 1996 2003

1992 24.8 12.0 1996 26.2 10.4 2003 29.4 8.5

World

East Asia (EA)1

1992 11.8 1996 12.4 2003 14.7

EU

6.5 7.3 11.5

2.6 3.3 6.2

7.0 8.0 14.3

4.2 6.0 6.2

2.0 3.0 3.0

6.1 7.8 6.1

17.4 19.5 17.3

8.6 9.5 9.0

34.4 38.5 35.3

44.9 40.3 34.3

66.8 63.5 52.8

18.0 16.7 18.0

Total manufacturing trade (M
X)

12.8 15.8 20.9

5.5 7.3 10.6

16.5 18.8 23.7

100 100 100

100 100 100

100 100 100 7.3 7.2 5.3

B3

27.1 15.4 34.5 13.2 36.8 12.1

11.2 15.2 16.7

34.7 22.5 38.4 19.4 35.1 13.4

3.0 0.5 8.2

0.7 1.4 4.2

1.9 2.9 7.9

6.5 9.4 12.0

2.5 4.8 5.5

7.2 10.4 9.6

23.6 40.7 19.5

11.9 14.3 11.2

42.7 39.5 41.5

40.3 14.8 27.8

67.4 59.7 52.8

14.1 13.1 13.8

Parts and components trade (M
X)

11.7 21.4 24.7

3.9 8.0 11.4

12.2 19.0 21.7

100 100 100

100 100 100

100 100 100

52.7 53.2 55.6 71.3 62.7 57.9 45.5 49.7 52.9

17.6 18.3 15.5 23.6 19.4 18.1 19.9 19.1 14.5

44.1 11.2 50.1 12.8 53.2 11.7 26.1 0.0 35.2 0.0 40.5 0.0 35.1 34.9 40.1 47.8 43.3 39.8 25.6 30.6 38.3

32.8 36.2 41.4 26.1 35.2 40.5 22.3 17.9 21.9 33.8 25.6 21.0 5.8 6.9 12.8

18.0 16.8 22.4 9.7 12.9 21.8 10.2 13.4 13.3 5.1 6.8 6.3 16.6 20.1 21.3

10.9 14.8 13.7 10.8 15.9 12.3 21.4 20.9 19.4 12.0 16.2 17.3 22.4 20.6 18.8

26.4 24.6 21.6 32.2 30.2 25.8 16.7 16.7 15.0 17.0 22.8 21.8 18.6 16.7 13.4

19.5 17.7 15.6 22.4 18.3 16.0 100 100 100 100 100 100 100 100 100

100 100 100 100 100 100 60.8 60.3 69.5 33.9 41.3 50.4 54.9 56.0 59.8

22.7 21.7 17.3 15.8 15.9 12.7 21.8 19.7 14.4

50.9 14.5 55.7 15.4 67.3 13.8 29.6 0.0 39.1 0.0 50.5 0.0 38.0 38.6 52.2 18.0 25.4 37.7 33.0 36.3 45.4

36.4 40.3 53.5 29.6 39.1 50.5

14.5 11.7 23.7 4.8 5.6 13.5 5.3 5.9 13.9

11.8 11.2 23.9 6.3 9.0 22.2

20.4 22.8 22.5 7.1 10.6 13.0 24.5 26.5 26.6

19.7 23.6 22.7 16.2 22.3 19.2

22.9 22.4 15.0 24.5 25.5 20.5 23.8 21.6 17.1

30.3 27.3 17.4 38.7 35.2 23.8

12.8 13.6 11.1 41.0 33.1 28.3 13.0 13.4 11.2

14.8 13.8 11.5 16.3 12.7 11.6

100 100 100 100 100 100 100 100 100

100 100 100 100 100 100

EA Japan DEA GCH AFTA NAFTA EU Total EA Japan DEA GCH AFTA NAFTA EU Total

A3

6.3 5.1 4.1

1992 35.7 19.2 1996 33.7 14.9 2003 33.8 10.1

NAFTA

92

A3

1992 8.7 1996 10.0 2003 10.5

1992 20.1 1996 22.7 2003 24.2

EU

World

B3

Parts and components trade (M
X)

12.4 15.4 18.1

4.6 6.4 7.6

13.5 15.6 18.4

6.1 6.7 10.0

2.0 2.7 4.3

5.2 5.8 10.1

4.5 6.3 5.6

1.9 2.8 2.3

5.4 6.8 5.4

19.0 20.4 20.4

8.4 8.9 10.3

39.0 42.8 43.3

43.9 100 22.8 39.0 100 28.9 33.6 100 34.5

65.4 100 8.7 61.3 100 12.3 51.7 100 13.2 9.0 8.9 8.0

4.1 4.3 3.3

19.0 100 26.4 13.9 16.4 100 31.2 13.4 16.6 100 29.3 9.0

13.9 20.0 26.5

4.6 7.9 9.9

12.5 17.8 20.3

4.0 1.0 11.3

1.3 1.8 4.1

2.3 3.0 6.6

7.7 9.6 11.4

2.6 4.6 4.5

7.3 10.1 9.8

Source: Compiled from the UN Comtrade Database using the commodity/country classification described in the text.

7.7 7.4 6.1

4.1 3.6 2.9

1992 27.0 13.5 1996 26.9 11.3 2003 26.2 7.8

Notes: 1. Including Japan. 2. Including AFTA.

Total manufacturing trade (M
X)

24.3 44.4 20.3

10.5 12.2 10.6

44.5 41.5 45.0

38.7 100 14.3 100 27.3 100

64.6 100 57.7 100 50.9 100

16.5 100 15.1 100 13.8 100

EA Japan DEA GCH AFTA NAFTA EU Total EA Japan DEA GCH AFTA NAFTA EU Total

(continued)

NAFTA

Table 4.5

Product fragmentation and trade patterns in East Asia

93

all East Asian countries in both total manufacturing trade and trade in components, of which the latter is much larger (Athukorala 2003b, Table A-1). Secondly, the measured degree of intra-regional trade integration needs to be qualified for the growing importance of cross-border trade in components. As can be seen in Table 4.6, component trade accounts for a significant and growing share of intra-regional trade in manufacturing in East Asia, on both the export and import sides. Moreover, the share of components in intra-regional trade is much larger than their share in the region’s extra-regional trade. In 2003, components accounted for 40.2 per cent of intra-East Asian exports, compared with 32 per cent in the region’s total exports. Between 1992 and 2003, trade in components accounted for 53 per cent of growth in total imports and 46 per cent of growth in total exports in East Asia. The significance of component trade looms even larger for developing East Asia and ASEAN. At the individual-country level, cross-border component trade accounts for more than a half of total imports and exports in Singapore, Malaysia and the Philippines, and more than a third in Thailand (Athukorala, 2003b, Table A-3). Korea and Taiwan are also involved in sizeable cross-border trade with the other countries in the region. For all East Asian countries, between 1992 and 2003 the share of components in both intra-regional exports and imports has increased at a much faster rate than that in exports to and imports from countries outside the region. So far, we have noted two important peculiarities of trade patterns in East Asia compared with total global trade and the trade of the EU and NAFTA. Firstly, component trade has played a much more important role in trade expansion in East Asia relative to the overall global experience and experiences of countries in other major regions. Second, trade in components accounts for a much larger share in intra-regional trade compared with the region’s trade with the rest of the world. Given these two peculiarities, trade flow analysis based on reported trade data is bound to yield a misleading picture as to the relative importance of intra-regional trade relations (as against global trade) in the growth dynamism of East Asia (and AFTA and other subregional groupings therein). Data reported in Table 4.7 on intra-regional shares of trade in total manufacturing, components and final goods for various regional economic groupings help us to understand this important point. The intra-regional share of final manufacturing trade in East Asia declined from 52.5 per cent in 1992 to 47.6 per cent in 2003, in sharp contrast to a notable increase (from 44.1 to 53.2 per cent) recorded for the conventionally used trade share (which covers both components and final goods). While the difference between intra-regional shares of final and

94

Multinational enterprises in Asian development

Table 4.6 Share of parts and components content in regional manufacturing trade flows (%) Exporter/ Importer East Asia (EA)1

1992 1996 2003 Japan 1992 (JPN) 1996 2003 East Asia 1992 excl. Japan 1996 (DEA)2 2003 ASEAN 1992 1996 2003 NAFTA3 1992 1996 2003 USA 1992 1996 2003 European 1992 Union 1996 (EU) 2003 World 1992 (WLD) 1996 2003

Exports EAS JPN DEA ASEAN NAF USA EUU WLD 25.2 32.1 40.2 28.3 34.9 38.4 23.3 30.7 40.9 33.9 40.7 52.2 28.0 33.1 40.9 28.3 33.7 41.1 18.4 20.3 27.1 28.5 29.3 41.2

20.2 24.0 31.3 – – – 20.2 24.0 31.3 28.5 29.1 38.0 23.6 27.4 31.5 24.0 28.1 31.9 11.9 13.0 14.6 23.0 21.2 30.4

25.8 33.6 42.0 28.3 34.9 38.4 23.9 32.8 44.0 35.4 44.1 56.2 30.3 35.9 44.9 30.6 36.4 45.1 21.1 23.0 31.9 29.8 31.5 44.1

34.9 42.4 51.3 33.4 40.6 46.8 36.1 43.7 53.5 39.7 49.3 58.7 39.3 45.0 57.2 40.1 45.6 57.6 21.8 27.9 38.7 39.1 39.7 53.8

23.1 29.9 28.6 25.6 32.4 28.2 20.3 27.9 28.9 26.5 32.6 40.8 27.1 25.0 24.8 31.6 31.5 31.1 21.1 21.2 19.0 26.6 25.2 25.6

22.9 29.1 28.3 25.7 31.9 27.5 20.1 26.8 28.8 27.1 32.7 40.7 22.2 19.7 20.1 – – – 20.3 20.7 18.4 22.5 23.4 22.3

16.8 24.5 28.2 19.3 26.5 27.6 13.4 23.3 28.6 17.8 29.7 38.7 25.2 29.3 28.8 25.5 29.7 29.1 14.8 16.6 17.0 17.3 18.1 20.1

21.3 28.0 32.0 22.9 30.2 30.6 19.7 26.7 32.8 26.4 35.0 44.4 26.2 27.2 28.1 27.6 30.5 32.3 15.6 17.7 18.9 22.2 21.7 25.4

Notes: 1. Including Japan. 2. Including AFTA. 3. Including USA. – Not applicable Source: Compiled from UN Comtrade Database using the commodity/country classification described in the text.

total trade is observable for both exports and imports, the magnitude of the difference is much larger on the export side. The difference in magnitude between regional trade shares estimated in gross and net terms is much larger for developing East Asia and AFTA compared with estimates for the entire region. In 2003 only 30 per cent of final goods exports from developing Asia found markets within the region, compared with 37.6 per

Product fragmentation and trade patterns in East Asia

95

Imports EAS

JPN

DEA ASEAN NAF

USA EUU

WLD

24.9 30.3 37.1 15.4 20.6 26.8 26.3 32.0 39.2 33.5 42.4 49.4 19.5 26.9 25.5 19.8 26.6 25.0 10.7 22.5 24.4 25.8 26.5 31.3

29.0 36.9 40.3 – – – 29.0 36.9 40.3 32.2 41.9 47.1 23.9 30.4 27.9 24.4 30.1 27.1 13.3 26.5 27.8 30.0 31.1 34.6

22.1 27.0 35.9 15.4 20.6 26.8 23.9 28.9 38.7 34.9 42.8 50.7 15.5 24.6 24.2 15.7 24.4 23.8 8.0 20.1 22.9 22.4 24.0 29.9

26.9 32.9 40.7 23.7 28.1 31.1 28.5 35.2 44.8 34.1 45.0 56.8 26.9 30.1 31.0 – – – 19.3 28.6 27.5 26.4 29.1 31.1

22.8 27.9 35.4 15.4 19.3 24.2 24.8 30.2 38.4 30.4 39.3 48.6 20.4 23.6 22.8 19.0 21.7 20.2 12.6 18.9 20.3 21.7 21.4 24.5

35.1 39.5 49.6 22.1 23.2 32.8 38.7 44.7 54.3 40.0 49.5 54.6 22.2 29.5 35.2 22.6 29.3 34.9 9.1 27.1 29.8 32.6 32.6 43.3

26.5 32.9 40.4 23.5 27.6 30.7 27.9 35.4 44.6 33.5 44.7 56.4 25.0 24.3 24.8 22.8 18.9 18.9 18.9 28.5 27.8 26.8 26.3 28.4

15.6 19.2 26.7 8.5 9.4 11.6 19.9 22.8 32.5 21.2 28.6 38.9 16.8 19.3 17.1 16.4 18.5 15.9 12.4 17.5 18.4 16.5 17.2 18.8

cent in total exports. For AFTA the relevant figures were 17.0 and 21.7 per cent, respectively. Moreover, as already noted, for all East Asian countries Japan is a much smaller market for final goods exports than the USA and the EU, accounting for less than 10 per cent in all cases in 2003. It is also interesting to note that, unlike in the case of East Asia (or developing East Asia and AFTA), the estimated intra-regional trade shares for NAFTA and the EU are remarkably resilient to the inclusion or exclusion of component trade. In sum, the estimates presented in this section support the hypothesis that, in a context where fragmentation-based trade is expanding rapidly, the standard trade flow analysis can lead to misleading inferences regarding the ongoing process of economic integration through trade. Product

96

Multinational enterprises in Asian development

Table 4.7 Intra-regional trade shares: total manufacturing, parts and components, and final trade (%), 1992, 1993 and 2003

A

East Asia1

Developing East Asia2

AFTA

NAFTA3

EU4

36.6 43.8 45.6 55.2 55.7 65.6 44.1 50.1 53.2

35.6 35.6 37.6 34.7 34.7 45.4 35.3 36.2 40.4

19.5 23.8 21.7 14.4 17.3 22.7 16.8 21.5 21.6

44.6 48.0 55.2 34.4 38.5 35.3 39.0 42.8 43.3

64.1 59.4 50.8 66.8 63.5 52.8 65.4 61.3 51.7

44.4 52.3 64.9 59.3 59.7 69.7 50.9 55.7 67.3

46.9 46.8 58.9 31.4 32.0 45.9 38.1 38.8 52.5

29.3 32.7 28.5 21.0 21.7 24.9 24.6 26.7 27.0

46.2 43.5 48.6 42.7 39.5 41.5 44.5 41.5 45.0

62.1 55.9 49.1 67.4 59.7 52.8 64.6 57.7 50.9

35.0 41.2 38.8 54.3 54.1 61.9 52.5 46.8 47.6

33.7 32.5 30.1 35.7 35.4 43.3 44.6 33.9 35.2

16.3 19.3 17.0 11.9 14.5 18.7 23.9 16.8 17.7

44.1 49.6 57.3 32.6 38.2 34.1 37.6 43.2 42.9

64.4 60.0 51.1 66.7 64.3 52.8 65.5 62.0 51.9

Total Manufacturing

Exports (X)

Imports (M)

Trade (X
M)

1992 1996 2003 1992 1996 2003 1992 1996 2003

B Parts and Components Exports (X) 1992 1996 2003 Imports (M) 1992 1996 2003 Trade (X
M) 1992 1996 2003 C Final goods Exports (X)

Imports (M)

Trade (X
M)

1992 1996 2003 1992 1996 2003 1992 1996 2003

Notes: 1. Including Japan. 2. Including AFTA. 3. USA, Canada, Mexico. 4. Austria, Belgium, Denmark, Finland, France, Germany, Ireland, Italy, Netherlands, Greece, Portugal, Spain, Sweden, UK. Source: Compiled from the UN Comtrade Database using the commodity/country classification described in the text.

Product fragmentation and trade patterns in East Asia

97

fragmentation leads to double counting of trade flows in published trade data because goods in process cross multiple international borders in the course of their production sequence. The total amount of trade involving the good while in process can be a multiple of the final value of that good. Moreover, trade shares calculated using reported data can lead to wrong inferences as to the relative importance of the ‘region’ and the rest of the world for the growth dynamism of a given country/region, even controlling for double counting in trade. This is because the rate of expansion of component trade depends crucially on the demand for the related final goods. When data on component trade are excluded from trade flows, our estimates suggest that extra-regional trade is much more important than intra-regional trade for the continued growth dynamism of East Asia, both including and excluding Japan. Thus, the ongoing process of product fragmentation spearheaded by MNEs seems to have strengthened the case for a global, rather than a regional, approach to trade and investment policymaking.

CONCLUSION AND INFERENCES There is clear evidence that fragmentation trade is expanding more rapidly than conventional final goods trade. The degree of dependence on this new form of international specialization is proportionately larger in East Asia than in North America and Europe. This seems to be the outcome of the more favourable policy setting for the entry of MNEs, agglomeration benefits arising from the first-comer advantage of early entry into fragmentation-based specialization, and considerable inter-country wage differentials in the region. A notable recent development in international product fragmentation in the region has been the rapid integration of China into the regional production networks. This development is an important counterpoint to the popular belief that China’s global integration would crowd out opportunities for export-led industrialization in other countries in the region. The evidence presented in this chapter on intra-regional patterns of trade flows has implications for assessing the knock-on effects of the newfound fondness in East Asian countries for free trade agreements (FTAs). Trade in components and final assembly is postulated to be more sensitive to tariff changes than final trade (or total trade as captured in published trade data) is (Yi 2003). Normally a tariff is incurred each time a good in process crosses a border. Consequently, when there is a one percentage point reduction in tariff, the cost of production of a vertically integrated good declines by a multiple of this initial reduction, in contrast to a one

98

Multinational enterprises in Asian development

per cent decline in the cost of a normal traded good. Moreover, tariff reduction may also make it more profitable for goods that were previously produced entirely in one country to now become vertically specialized. Consequently, in theory, the trade-stimulating effect of FTA would be higher for component trade than for normal trade, other things remaining unchanged. However, in reality, much would depend on the nature of ‘rules of origin’ (ROOs) built into FTAs (Baldwin 2001, Garnaut 2003). Trade distorting effects of ROOs are presumably more detrimental to fragmentation-based trade than to conventional final goods trade, because of the inherent difficulties involved in defining the ‘product’ for giving duty exception and the transaction costs associated with the bureaucratic supervision of the amount of value added in production coming from various sources. Furthermore, maintaining barriers to trade against non-members (while allowing free trade among members) can thwart ‘natural’ expansion of fragmentation trade across countries. Formation of FTAs would therefore simply result in substituting for the existing tariff concessions rather than generating new incentives for fragmentation trade. Thus, in terms of opportunities for trade expansion through international product fragmentation, the ideal policy choice appears to be multilateral liberalization through the WTO process; the ongoing process of product fragmentation seems to have strengthened the case for a global, rather than a regional, approach to trade and investment policymaking. Fragmentation trade has played a pivotal role in the continuing dynamism of the East Asian economies and increasing intra-regional economic interdependence. But this new form of international specialization could not be sustained purely as an East Asian phenomenon, because of the growing importance of extra-regional markets for final goods. Put simply, growing trade in components has made the East Asian region increasingly reliant on extraregional trade for its growth dynamism. Fragmentation-based specialization has the potential to magnify the gains from unilateral and multilateral trade and investment liberalization. In this sequential mode of production, activities at separate stages have thin margins, and therefore are highly sensitive to even small tariffs. When a component crosses multiple borders, even low tariff rates are magnified as they are repeatedly applied to the goods in process. Thus, in the presence of growing opportunities for fragmentation trade, further trade liberalization could yield substantial gains. Of course, to be effective, opening up of the borders needs to be appropriately combined with policy initiatives to dismantle/harmonize barriers behind borders, such as regulations, national standards, competition policies and government procurements.

Product fragmentation and trade patterns in East Asia

99

NOTES * 1.

2. 3. 4. 5. 6.

This chapter draws upon Athukorala (2006b) and Athukorala and Yamashita (2006). This phenomenon has gone under alternative names, such as ‘Vertical specialization’ (Hummels et al. 2001, Irwin 2002), ‘slicing the value chain’ (Krugman 1995), ‘international production sharing’ (Ng and Yeats 2001, Yeats 2001), and ‘outsourcing’ (Rangan and Lawrence 1999, Hanson et al. 2001). Following Jones (2000), Jones and Kierzkowski (2001a, 2001b, 2001c), Venables (1999), Baldwin (2001) and others we stick to the term ‘product fragmentation’ in this chapter. The key references include Baldwin (2001), Cantwell (1994), Deardorff (2001), Jones (2000), Jones and Kierzkowski (2001a, 2001b, 2001c), Jones et al. (2004), Venables (1999). See for instance Borrus (1997); Dobson and Chia (1997); McKendrick et al. (2000). For instance, ‘television tubes’ and ‘computer processors’ were lumped together with ‘television sets’ and ‘computers’ respectively. See Hummels et al. (2001) for an application of the input–output technique for estimating fragmentation trade for some individual OECD countries. Share of SITC 7 and SITC 8 in total manufacturing trade (SITC 5–8 less SITC 68)

Exports

SITC 7 SITC 8 SITC 7
8 7. 8.

9. 10.

1992 50.4 18.1 68.5

1996 50.4 17.2 67.6

Imports 2003 55.2 16.7 71.9

1992 53.8 16.9 70.7

1996 52.1 15.5 67.6

2003 57.1 15.5 72.6

The list is available in Athukorala (2003b), Appendix A-5. Historically entrepot trade has accounted for a significant share of Hong Kong’s total trade and this share has increased dramatically in recent years because of its growing importance as a transshipment centre for mainland China’s booming foreign trade. For instance, re-exports accounted for 92.3 per cent of total exports from Hong Kong in 2004, up from 74.3 per cent in 1992. Failure to net out entrepot trade for other countries obviously leads to some double counting of trade flows, but the magnitude of the resulting error is unlikely to be significant, except perhaps for Singapore. Henceforth we use the term ‘components’ in place of ‘parts and components’ for brevity. For an analysis of the component trade as an increasingly important vehicle for China’s rapid economic integration in the East Asian region, see Athukorala (2005).

5. Multinational firms in crisis and recovery: lessons from the 1997–98 Asian crisis* The string of economic crises in emerging market economies in the 1990s and the global reverberations that followed them have added new impetus to the debate on how to reconcile international capital mobility with domestic economic stability and developmental priorities in investmentreceiving developing countries. The unqualified enthusiasm for promoting capital flows to aid economic advancement in these countries has given way to a new emphasis on finding ways and means of reconciling international capital mobility with domestic economic stability and developmental priorities. At the heart of this new policy focus is a renewed emphasis on the conventional wisdom about the need to treat foreign direct investment (FDI) flows separately from other forms of capital flows (mostly hot money) in designing national policies to monitor capital flows.1 This and the next chapter aim to inform this debate. This chapter examines the behaviour of FDI flows compared with other major forms of capital flows in the context of the 1997–98 East Asian financial crisis focusing on the following questions. Have FDI flows been more stable than other forms of capital flows? What has been the contribution of FDI firms (MNE affiliates) in the adjustment process? These issues are addressed drawing upon the experiences of the five countries in the region which were directly affected by the crisis – Thailand, Indonesia, Malaysia, the Republic of Korea (henceforth referred to as Korea) and the Philippines. The next chapter examines the implications of FDI compared with other forms of capital flows in the policy debate on what Corden (1994, p. 8) calls the ‘real exchange rate problem’ – the possibility that capital inflows bring about an appreciation of the real exchange rate (the relative price of traded to nontraded goods) with adverse affects on traded-goods production in the domestic economy – through a comparative study of Asia and Latin America. The remainder of the chapter is organized as follows. The next section discusses the a priori reasoning on differences between FDI and other forms of foreign capital flows in the context of an international financial 100

Multinational firms in crisis and recovery

101

crisis. The following two sections examine the behaviour of FDI flows compared with other forms of capital flows in the wake of the crisis and in the recovery process, and the role of the affiliates of multinational enterprises (MNEs) in adjustment to the crisis. The final section contains some concluding remarks.

ANALYTICAL CONTEXT FDI originates from the decision of an MNE to enter into international production and relocate part of its activities in a selected host country. This decision is underpinned by the desire to reap benefits from its specific advantages (in the form of technology, managerial expertise, marketing know-how and so on), which cannot be effectively leased or purchased through ‘arm’s-length’ market dealings with unrelated firms. In other words, FDI is a flow of long-term capital based on long-term profit considerations involved in international production (Caves 1996). Thus it can be hypothesized that FDI tends to be less responsive to short-term aberrations in general economic conditions in host countries. Viewed from this perspective, one would expect FDI inflows to be much more resilient than other forms of private capital – portfolio investment, bank lending and other related forms of foreign capital (broadly known as ‘financial investment’) – in the wake of an international financial crisis. Financial investment essentially involves acquisition of financial assets rather than direct involvement in international production. Returns from such investment depend on variables such as the exchange rate, interest rates and share prices, which are usually subject to short-term fluctuations. Moreover these assets can be easily disposed of at short notice, although of course at a cost.2 The operation of a subsidiary firm (MNE affiliate) in a given country essentially involves accumulation over time of intangible assets such as technological know-how, long-term supplier relationships, marketing channels, goodwill, human resources, institutional links and intimate knowledge of the host-country business environment, and other firmspecific skills. Such intangible assets are not normally transferable through arm’s-length transactions. Thus the resale price of a subsidiary could fall well short of the actual accumulated investment. Given this sunk cost, in a crisis context parent firms could well invest more in order to support troubled subsidiaries rather than contemplating liquidating or selling them (Fukao 2001). A financial crisis generates both positive and negative impacts on the profitability of MNE operations in the crisis-affected economy. On the

102

Multinational enterprises in Asian development

positive side, currency collapse can have a positive impact on FDI in at least three different ways. First, large exchange rate depreciations reduce domestic production costs and asset values, making foreign investment more profitable. Since depreciation of the exchange rate of host-country currencies makes the firms wealthier in terms of their purchasing power within the country, investment can increase. Second, the cost of investment may also be significantly reduced by falling asset prices because of the contraction in domestic demand propelled by the crisis. Third, revisions to FDI laws as part of the crisis management package in crisis-affected countries can open up new opportunities for cross-border mergers and acquisitions (‘fire-sale’ investment, à la Krugman (2001)). On the negative side, domestic demand contraction caused by output collapse and lowered immediate growth prospects can have a negative effect on domestic-market-oriented foreign investment. There are reasons to believe that significant involvement of MNE affiliates – the tangible reflection of FDI inflow – in the domestic economy of a given country can act as a facilitator of the adjustment process following a financial crisis or other economic disruption caused by an external economic shock (Blomström and Lipsey 1993; Lipsey 2000, 2001). MNE affiliates have already set up international markets to supplement external markets. They also presumably have greater access than local firms to market information, distribution channels and international marketing skills. For these reasons, affiliates of MNEs should find it easier to switch markets than other firms in response to a collapse of domestic demand as well as to benefit from newly gained competitiveness from exchange rate depreciation. On the supply side, MNE affiliates are also presumably better equipped to face domestic credit contraction (the credit crunch) that usually follows an exchange rate collapse and the exodus of short-term capital in the wake of a financial crisis. When domestic bank credit dries up and/or the cost of credit increases in the crisis context, it is still possible for these firms to obtain financing in international capital markets or to receive credit from affiliated firms. Local firms do not have that advantage, although large exporting firms might still find it possible to obtain financing in the international capital markets or receive credit from upstream firms (Krueger and Tornell 1999, p. 33). Thanks to their parent firms, unlike purely domestically owned firms, MNE affiliates in crisis-affected countries do not generally suffer lowered credit ratings.

CAPITAL FLOWS DURING THE CRISIS An important development in the global economy in the 1990s was the enormous increase in private capital flows to emerging markets, that is,

Multinational firms in crisis and recovery

103

developing countries and transition economies. Net private capital flows to these emerging markets increased from an annual average of less than US$10 billion in the latter half of the 1980s to nearly US$200 billion by the mid-1990s.3 Of total annual flows to these countries during 1990–97, countries in Asia and Latin America absorbed almost three-fourths. In the early 1990s, total flows to Asia and Latin America were roughly similar in magnitude, but during 1993–96 inflows to Asia surpassed those to Latin America by a wide margin. Korea, Malaysia, Thailand and Indonesia were among the eight countries that received more than $15 billion in net longterm private capital inflows during this period. Flows to the Philippines continued to remain low by international standards, but they increased from less than $10 million to over $5 billion by the mid-1990s. The total net inflow to the five countries increased sharply from an average annual level of $2.5 billion in the second half of the 1980s to $78 billion in 1996. In that year, net capital inflows relative to GDP stood at 10 per cent in the Philippines, 9.2 per cent in Thailand, 6.9 per cent in Malaysia, 5.4 per cent Indonesia and 4.7 per cent in Korea. The rapid increase in capital flows to the region has been accompanied by a remarkable broadening of the composition of these flows, with FDI emerging as the single largest component. (By contrast the overwhelming share of flows to Latin America took the form of portfolio investment and bank credit.) In 1997, FDI flows constituted $64.3 billion, or 52.4 per cent of the estimated $122.6 billion which flowed to these countries (up from 40 per cent in 1990). Notwithstanding the greatly increased importance of FDI, other forms of private capital flows have increased faster still, although in some cases the amounts involved were very small even as recently as 1990. The extremely rapid growth of these more short-term, volatile capital flows explains in part the crisis in several East Asian economies in 1997–99. What have been the implications of the onset of the financial crisis in 1997 for the process of global integration through capital mobility in these countries? Did FDI behave differently from other forms of capital flows in the crisis context? The remainder of this section examines these issues using Tables 5.1 and 5.2 and Figure 5.1. It is important to note that the FDI series reported in Table 5.1 and Figure 5.1 provide only partial coverage of FDI in these countries. According to the standard definition, FDI has three components: equity capital, inter-company debt and reinvested earnings. As in many other countries, data series on FDI in these countries (reported as part of the balance of payments accounts on which Table 5.1 and Figure 5.1 are based) capture only equity capital and inter-company debt. The omission of the third component (retained earnings) can lead to a considerable underestimation of the actual magnitude of FDI in a given host country, depending

104

Multinational enterprises in Asian development

Table 5.1 Capital flows in Asian crisis countries, 1990–2000 (US$ million) 1990–941 Indonesia FDI inflow FDI outflow FDI (net) Portfolio flows (net) Bank credit and other (net) Total (net) Korea FDI inflow FDI outflow FDI (net) Portfolio flows (net) Bank credit and other (net) Total (net)

1995

1998

1999

2000

356 44 400 1878

2745 72 2817 1792

4550 150 4700 1909

117

2270

126

1420

9841 1123

4548

4483

8029

1776 2326 2053 3551 4671 3438 1775 2345 1385 11 711 15 102 9917

5412 4739 673 1224

9333 4197 5136 9190

9283 4998 4285 12 177

1693 193 1500 1098

4346 603 3743 4100

2561

2416

5158

10 259

819 1501 682 5110

1996

1997

6194 2702 600 151 5594 2551 5005 3558 758

1782

7459

1778 18 848

736

1215

5757

6211

17 395

14 535 10 316

185

15 541

10 704

26 874 19 269 7605 2226

12 048 7680 4121 1686 7927 5994 3498 7021

7504 2255 5249 353

8463 3572 4891 264

9178 3622 5556 2294

Malaysia FDI inflow 17 330 FDI outflow 11 986 FDI (net) 5343 Portfolio 4396 flows (net) Bank credit and 2657 other (net)

4456

2300

2698

2936

2021

992

12 397

14 287

13 724

1671

1960

3135

2271

826 135 691 63

1478 399 1079 1190

1517 182 1335 5317

826 145 681 71

2287 160 2127 928

573 59 632 4816

2029 95 2124 236

2561

3040

3291

344

309

2292

1493

Total (net)

3316

5309

9943

1096

890

3156

3381

Thailand FDI inflow FDI outflow

1948 236

2068 885

2336 932

3746 389

6940 130

5724 373

3366 23

Total (net) Philippines FDI inflow FDI outflow FDI (net) Portfolio flows (net) Bank credit and other (net)

105

Multinational firms in crisis and recovery

Table 5.1

(continued) 1990–941

FDI (net) Portfolio flows (net) Bank credit and other (net) Total (net) ‘Crisis five’ total FDI inflow FDI outflow FDI (net) Portfolio flows (net) Bank credit and other (net) Total (net)

1995

1712 1748

1183 4082

7139

1996 1404 3545

1997 3357 4353

1998

2000

5351 816

3389 706

16 644

14 538 24 585 21 223 15 531

6989

10 599

21 909

19 487 16 875 14 454

9364

4306

22 616 14 051 8565 12 416

36 542 24 707 11 835 23 309

24 421 10 506 13 915 32 467

21 348 8155 13 193 13 294

19 306 8652 10 654 7032

16 700

34 015

21 149 40 508 26 002 18 503 13 665

37 681

69 159

67 530 25 547 15 967

17 007 5809 11 198 3762

6810 41

1999

21 787 7328 14 459 4424

7985

4021

Note: 1. Annual average. Sources: Compiled from Bank Negara Malaysia, Quarterly Bulletin of Statistics (various issues), for Malaysia, and IMF, International Financial Statistics (CD-ROM) for other countries.

on the history of MNE involvement and the source-country profile of FDI.4 For this reason, we also make use of data on outflows of FDI from the United States in Table 5.2 to examine the sensitivity of inferences. This is the only available source of time series data on FDI encompassing all three components for the period under study. Table 5.1 shows that total capital inflows to the five crisis-hit countries reversed from a net inflow of $6.8 billion in 1996 to a net outflow of $25.5 billion in 1997 and $16.0 billion in 1998. Total net inflows started to recover from about the third quarter of 1998, but the annual figures remained well below the pre-crisis level in 1999 and 2000. At the individual country level, the sharpest reversal in net inflow was experienced by Thailand, with recorded negative inflows for four consecutive years from 1997. The total net outflow of capital from Thailand during the four years from 1997 to 2000 amounted to $45 billion or almost half of the net inflows during the seven boom years of 1990–96. Capital flows to Indonesia reversed from a net inflow of over $10 billion in 1996 to a net outflow of $1.1 billion in 1997 and the value of net outflows continued to widen in the ensuing years, reaching $8 billion in 2000. Net inflows to Korea contracted by $24.5

106

Multinational enterprises in Asian development 3000 2000

Net flows ($ million)

1000 0 –1000 –2000 –3000 –4000 FDI

–5000

FPI

BNK

2001q1

2000q4

2000q3

2000q2

2000q1

1999q4

1999q3

1999q2

1999q1

1998q4

1998q3

1998q2

1998q1

1997q4

1997q3

1997q2

1997q1

1996q4

1996q3

1996q2

1996q1

–6000

(a) Indonesia

12000

Net flows ($ million)

7000 2000 –3000 –8000 FDI

FPI

BNK

–13 000

2001q4

2001q3

2001q2

2001q1

2000q4

2000q3

2000q2

2000q1

1999q4

1999q3

1999q2

1999q1

1998q4

1998q3

1998q2

1998q1

1997q4

1997q3

1997q2

1997q1

1996q4

1996q3

1996q2

1996q1

–18 000

(b) Korea

Figure 5.1 Net capital flows to Indonesia, Korea, Malaysia, the Philippines and Thailand, 1996q1–2002q1 billion between 1996 and 1997 but began to recover much more quickly (from about the first quarter of 1998) than in the other countries. Malaysia and the Philippines, too, suffered massive contractions in capital inflows ($12 billion in Malaysia and $8.8 billion in the Philippines between 1996 and 1997) but total net inflows remained positive throughout.

107

Multinational firms in crisis and recovery

3100

Net flows ($ million)

2100 1100 100 –900 –1900 –2900

FDI

–3900

FPI

BNK

–4900 1996q1 1996q2 1996q3 1996q4 1997q1 1997q2 1997q3 1997q4 1998q1 1998q2 1998q3 1998q4 1999q1 1999q2 1999q3 1999q4 2000q1 2000q2 2000q3 2000q4 2001q1 2001q2 2001q3 2001q4 2002q1

–5900

(c) Malaysia 3000

Net flows ($ million)

2000 1000 0 –1000 –2000 FDI

FPI

BNK 2000q4

2000q3

2000q2

2000q1

1999q4

1999q3

1999q2

1999q1

1998q4

1998q3

1998q2

1998q1

1997q4

1997q3

1997q2

1997q1

1996q4

1996q3

1996q2

1996q1

–3000

(d) Philippines

Figure 5.1 (continued) The behaviour of FDI inflows into these countries in the context of the crisis was strikingly different from what we can observe for total net capital flows. In 1999 there was an annual contraction in total net FDI inflows to the five countries of about 15 per cent (from $13.9 billion in 1996 to $11.2 billion in 1997). But these flows recovered swiftly to pre-crisis levels by 1998. It is therefore clear that the massive contraction of total capital flows

108

Multinational enterprises in Asian development 2500

Net flows ($ million)

1500 500 –500 –1500 –2500 –3500

FDI

FPI

BNK

–4500

2001q4

2001q3

2001q2

2001q1

2000q4

2000q3

2000q2

2000q1

1999q4

1999q3

1999q2

1999q1

1998q4

1998q3

1998q2

1998q1

1997q4

1997q3

1997q2

1997q1

1996q4

1996q3

1996q2

1996q1

–5500

(e) Thailand Notes:

FDI foreign direct investment; PFI portfolio investment; BNK bank credit.

Source: As for Table 5.1.

Figure 5.1 (continued) to these countries originated in the other two components of foreign capital – portfolio flows and bank credit. Total net FDI inflow to the five countries during 1997 and 1998 ($25.7 billion) was only 0.4 per cent lower than the total during the two pre-crisis years 1995 and 1996 ($25.6 billion). A similar comparison for portfolio capital and bank credit (including unclassified flows) reveals massive contractions of 101 per cent (from $55.7 to $0.7 billion) and 220 per cent (from $55.1 to $66.4 billion) respectively.5 Interestingly, a comparison among the five countries suggests an inverse relationship between the share of FDI in total net flows in the leadup to the crisis and the degree of contraction in net inflows in the aftermath of the crisis (Table 5.2). Table 5.3 shows that, following the onset of the crisis, there was a significant increase in FDI coming in the form of cross-border mergers and acquisitions (M&As, or ‘fire-sale FDI’ à la Krugman (2001)) to all five countries. Total approved (announced) average annual M&A in the five countries recorded a 120 per cent increase between 1990–1996 and 1997–2001 (from $4.5 to $11.1 billion). The individual country increases were: Indonesia 172 per cent, Korea 834 per cent, Malaysia 87 per cent, the Philippines 142 per cent and Thailand 269 per cent.6 In Korea the crisis-driven slowdown in net FDI inflows lasted for only about two quarters, as Figure 5.1(b) shows. From then on, these flows

109

Multinational firms in crisis and recovery

Table 5.2 Percentage change in net capital flows during 1997–98 over 1995–96 Indonesia Korea FDI 77.0 Other capital flows 408.7

82.7 132.3

Malaysia Philippines Thailand Total 27.6 161.0

16.5 87.0

293.0 206.9

0.4 160.5

Source: Computed from Table 5.1.

Table 5.3 Mergers and acquisitions by foreign firms in Asian crisis countries, 1990–2001, announced value (US$ million) 1990–941

1995

1996

1997

1998

1999

2000

2001

Indonesia Korea Malaysia Philippines Thailand

747 676 1221 1446 778

809 192 98 1208 161

530 564 768 462 234

332 836 351 4157 633

683 3973 1096 1905 3209

1164 10 062 1166 1523 2011

819 6448 441 366 2569

3529 3648 1449 2063 657

‘Crisis five’ total

4868

2468

2558

6309 10 866

15 926 10 643 11 346

Note: 1. Annual average. Source: UNCTAD (2002, Annex Table B.7).

started to increase significantly as investors began to respond to the new FDI liberalization initiatives and started to take part in takeover and acquisition activities.7 Total net flows in 1999 were almost 20 per cent higher than the levels recorded in 1996. It is important to note that the post-crisis increase in net FDI flows in Korea was somewhat aided by a decline in outward investment by Korean companies owing to their domestic financial troubles. However, the overall post-crisis increases in net inflow are not much different from that of inward flows. In Thailand, the pick-up in net FDI inflows (the pattern of which is not much different from that of inward flows, given small outward flows from the country) started from about the second quarter of 1998. In sharp contrast to the experiences of Korea and Thailand, in Indonesia net FDI flows continued to contract from the last quarter of 1997 (with the exception of a mild reversal in 1998), reflecting the continued deterioration in the overall investment climate. Net FDI flows to Malaysia declined from $7.9 billion in 1996 to $6.0 billion in 1997 (a 24 per cent contraction) and have remained virtually flat

110

Multinational enterprises in Asian development

at that level from about mid-1998, compared with a significant increase in flows to Korea and Thailand. It could well be that the prolonged period of policy and political uncertainty following the onset of the crisis, and widespread market scepticism about the fate of Malaysia’s unorthodox reform package introduced in September 1998, have played a role. However, one should be cautious in deriving inferences from a comparison of Malaysia’s post-crisis FDI experience with that of Thailand and Korea for a number of reasons. First, in Thailand and Korea acquisition by foreign companies of assets or equity of domestic companies has been an important component of foreign capital inflows during this period. Despite the severity of the downturn, corporate distress was far less widespread in Malaysia than elsewhere, and there were simply fewer ‘bargain assets’ for mergers and takeovers. Moreover, unlike Korea and Thailand, Malaysia did not resort to promoting acquisitions and takeovers by foreign companies as part of the ongoing process of corporate and banking restructuring. Second, compared with Korea and Thailand (particularly the former), Malaysia’s foreign investment regime remained much more liberal for a long time, and in some sectors the presence of MNEs had already reached very high levels by the onset of the crisis. Thus the post-crisis increase in FDI in the former countries compared with Malaysia may, to a significant extent, reflect ‘catching-up’ entry by foreign firms following the new FDI liberalization initiatives. Third, in the immediate pre-crisis years, intra-regional inflows (particularly from Korea and Taiwan) accounted for over one third of total FDI inflows to Malaysia and these flows dwindled following the onset of the crisis. Table 5.4 provides data on direct investment by US MNEs in the five countries, disaggregated by its three components – equity capital, intercompany debt and reinvested earnings. The time patterns of US FDI in the five countries are by and large consistent with those revealed by the data of aggregate FDI (balance-of-payments based) reported in Table 5.1. The time patterns are much closer for Korea, Thailand and Indonesia. In Malaysia and the Philippines the decline in US FDI in the aftermath of the crisis is much sharper than is revealed by the data in Table 5.1. These differences may reflect both estimation errors (mostly in the data reported in Table 5.1) and differences in the behaviour of FDI flows from different countries. Interestingly, the disaggregated data in Table 5.4 suggest that the inclusion or exclusion of reinvested earnings as part of total FDI does not matter much in an inter-temporal comparison of FDI; all three constituent series seem to move in unison over time. However, the data do suggest that the widely used balance-of-payments-based FDI figures (which do not cover reinvested earnings) tend to significantly understate the level of FDI in a given host country. For instance, reinvestment earnings accounted on

111

Multinational firms in crisis and recovery

Table 5.4 US direct investment in Asian crisis countries, 1994–2001 (US$ million) 1994

1995

1996

1997

444 1729 112

67 435 151

77 152 1031

612 234 867

Total

2061

519

956

21

461

Korea Equity capital Inter-company debt Reinvested earnings

154 78 314

282 185 584

102 19 668

337 49 295

Total

390

1051

752

Malaysia Equity capital Inter-company debt Reinvested earnings

197 76 280

166 243 628

Total

553

Philippines Equity capital Inter-company debt Reinvested earnings

Indonesia Equity capital Inter-company debt Reinvested earnings

1998

1999

616 197 618 231 463 1012

2000

2001

100 641 666

1 137 132

584

1207

291

69 134 566

834 391 304

771 667 862

25 278 651

681

631

1531

2300

953

179 D 754

134 91 509

63 256 278

86 926 573

32 500 1003

32 713 132

1037

1298

733

470

439

1471

549

75 152 187

D D 171

3 362 373

D D 289

103 62 122

D D 58

12 328 241

2 149 105

Total

414

269

738

107

287

213

99

47

Thailand Equity capital Inter-company debt Reinvested earnings

145 352 206

97 131 458

D D 550

D D 42

407 73 56

883 203 17

288 12 773

64 259 473

Total

703

686

849

16

424

1068

1050

668

‘Crisis five’ total Equity capital Inter-company debt Reinvested earnings

1015 2231 875

* * 1992

* * 3376

* * 1918

1120 605 817

* * 94

915 1468 3545

6 164 1283

Total

4121

3562

4593

1526

1333

1363

5929

1410

Notes: D Suppressed to avoid disclosure of data of individual companies. * Total cannot be computed because of suppression of data for one or more countries. Source: Compiled from the US Bureau of Economic Analysis electronic database, http://www.bea.doc.gov/bea/di/diacap_98.htm.

112

Multinational enterprises in Asian development

average for 57 per cent of total annual US direct investment in the five countries during 1994–2001. Fukao (2001) provides a penetrating analysis of the investment behaviour of Japanese MNEs in Indonesia, Malaysia, the Philippines, Thailand and Korea in the aftermath of the financial crisis. According to his findings, there was a noticeable decline in the number of new FDI cases by Japanese MNEs in all these countries. Compared with the US and European firms, Japanese MNEs were not actively involved in cross-border mergers and acquisitions. However, they injected considerable investment into existing subsidiaries in order to ensure their uninterrupted operation in the face of deteriorating financial conditions. Thus total Japanese FDI inflows to these countries increased from Yen 492 billion in the second quarter of 1996 and the first quarter of 1997 to US$733 billion in the corresponding period in 1997/98 (Fukao 2001, Table 8.2).

ROLE OF MNE AFFILIATES IN ADJUSTMENT AND RECOVERY Did MNE affiliates behave differently from domestically owned firms in the context of the crisis? In particular, did their response to the crisis contribute to the agility of the crisis-affected economies? A definitive analysis of this issue is not possible because of paucity of data. In this section an attempt is made to piece together fragments of relevant data from various sources. Table 5.5 reports estimates of the relative contribution of FDI to total gross domestic investment (GDI) in the five countries. Interestingly, in Korea, Malaysia, the Philippines and Thailand the FDI/GDP ratio was higher during the crisis years, compared with pre-crisis levels. In other words, FDI weathered the crisis far better than domestic private Table 5.5 Asian crisis countries: foreign direct investment as a percentage of gross domestic investment, 1990–2001

Indonesia Korea Malaysia Philippines Thailand

1990–94

1995

1996

1997

1998

1999

2000

2001

3.8 0.7 15.7 6.5 4.5

7.6 1.0 26.0 9.0 3.0

9.2 1.2 27.9 7.8 3.1

7.7 1.7 29.0 6.1 7.6

1.4 5.7 38.0 16.6 29.2

9.0 8.3 48.3 3.9 24.3

12.2 7.1 40.0 15.0 12.5

11.2 7.0 38.2 13.5 13.2

Source: Compiled from UNCTAD, World Investment Report (various years) and Bank Negara Malaysia, Quarterly Statistical Bulletin (for Malaysia).

Multinational firms in crisis and recovery

113

investment. The important inference is that FDI can act as an effective cushion against a possible collapse in domestic investment during a crisis. This inference, however, needs to be qualified owing to data problems. FDI and GDI series used here come from different data systems (balance-ofpayments accounts and national accounts respectively) and are presumably subject to estimation errors of different magnitudes, which are unlikely to be consistent over time. In particular, the FDI series capture financial flows relating to cross-border mergers and takeovers in addition to green-field FDI, but conceptually only the latter is captured in the GDI series. Moreover, as already noted, FDI series cover only equity flows and intercompany debt, whereas GDI series should capture reinvested earnings as well. In addition to these differences in the actual coverage, these series are subject to measurement errors of different magnitudes. The data generated by the US Bureau of Economic Analysis from its Annual Survey of US Direct Investment Abroad, when combined with the relevant host-country data, provide important insight into the behaviour of affiliates of US MNEs in East Asian countries during 1995–98 (Tables 5.6–5.8). Export performance of the affiliates located in the five crisisaffected Asian countries (Indonesia, Korea, Malaysia, the Philippines and Thailand) followed a steady course during the crisis years, notwithstanding the generally negative growth in total manufacturing exports from these countries. Consequently there was a persistent increase in the share of US affiliates in total manufacturing exports during 1997–2001: the average annual share, which was about 5 per cent in the first half of the 1990s, increased to over 7 per cent in the second half (Table 5.6). The time patterns of the export shares of the affiliates among the five countries were strikingly similar. Interestingly, there was no significant difference in export patterns between these five countries and the other four Asian countries listed in the table, which were not directly affected by the onset of the crisis. As can be seen in Table 5.7, the increase in the share of US MNEs in total manufacturing exports from these countries was underpinned by a notable increase in export propensity or the share of exports in total sales. It seems that, as local sales declined sharply following the onset of the crisis (by 17 per cent in all East Asian countries and 30 per cent in the ‘crisis five’ between 1997 and 1998), the MNE affiliates were quick to redirect their sales away from host-country markets in an attempt to minimize the impact of the crisis on their overall performance. The ratio of exports to total sales of these affiliates jumped in every country except China. The largest changes were in the four crisis-affected countries in Southeast Asia. Export sales ratios of affiliates in Korea remained virtually unchanged, but it is not possible to read much meaning into this figure because majority-owned affiliates are not representative of the overall US MNE presence in that

114

Multinational enterprises in Asian development

Table 5.6 Exports by majority-owned affiliates of US MNEs as a percentage of total host-country exports in East Asia, 1995–98 1990–94 1995 Hong Kong Singapore Taiwan China ‘Crisis five’ Indonesia Korea Malaysia Philippines Thailand 9 countries

2.50 21.60 3.00 0.30 5.00 0.50 0.80 12.80 17.50 7.30 5.50

1996

1.90 2.00 16.80 20.60 3.00 2.70 0.60 1.70 3.20 3.90 0.50 0.50 0.50 0.60 6.60 8.30 9.10 9.40 5.20 6.10 4.30 5.30

1997

1998

1999

2000

2001

2002

2.20 3.30 3.70 3.50 2.20 21.40 23.60 26.70 26.20 27.80 2.50 3.10 2.70 3.30 3.00 2.10 2.70 3.30 3.20 3.70 4.30 5.20 7.90 7.30 9.00 0.50 0.50 1.10 0.60 0.70 0.50 0.50 0.80 0.90 1.40 10.70 14.80 20.00 18.50 22.60 9.30 9.10 9.70 11.60 12.80 6.30 8.80 13.20 10.90 13.40 5.50 6.50 7.60 7.30 7.70

2.20 22.90 3.00 3.70 8.80 1.00 1.90 21.20 12.60 13.50 6.90

Source: Compiled from computer files of US Direct Investment Abroad, Bureau of Economic Analysis, US Department of Commerce.

Table 5.7 Exports by majority-owned affiliates of US MNEs as a percentage of total sales in East Asia, 1995–98

Hong Kong Singapore Taiwan China ‘Crisis five’ Indonesia Korea Malaysia Philippines Thailand 9 countries

1990–94

1995

1996

1997

1998

1999

2000

2001

2002

55.0 81.9 39.8 23.5 55.8 11.4 24.8 74.8 36.3 62.4 64.0

42.5 75.0 41.7 24.6 45.8 17.0 15.9 59.1 40.6 59.6 55.3

41.8 77.5 42.0 43.9 50.4 19.3 16.3 66.8 44.2 61.0 5.3

39.9 72.8 40.5 45.2 53.3 17.3 15.8 68.6 47.1 60.9 57.8

58.5 76.3 42.0 44.6 67.4 32.2 19.1 85.4 54.2 72.8 64.8

68.2 66.8 40.1 36.9 60.4 24.8 16.3 80.0 56.3 59.8 58.7

67.3 65.5 50.4 36.2 58.8 18.2 15.8 79.6 66.2 58.1 57.9

54.8 71.0 42.2 36.8 60.9 14.2 22.4 80.1 61.6 61.6 58.0

46.6 56.2 44.8 34.8 57.4 18.5 29.3 69.0 65.0 58.8 50.7

Source: Compiled from computer files of US Direct Investment Abroad, Bureau of Economic Analysis, US Department of Commerce.

country. Affiliates in Singapore, Taiwan and China, the economies less affected by the crisis, did not have significant market switching. The ability of these firms to switch sales from host-country markets to export markets clearly acted as a cushion against employment losses in the firms at a time when total manufacturing employment in these countries

115

Multinational firms in crisis and recovery

Table 5.8 Share of employment in US affiliates in total manufacturing employment

China Singapore Taiwan Hong Kong ‘Crisis five’ Philippines Indonesia Korea Malaysia Thailand

1990–95

1996

1997

1998

1999

2000

2001

2002

0.0 16.2 1.5 6.8 1.2 2.2 0.1 0.4 5.1 1.2

0.1 19.1 1.4 7.6 1.2 1.8 0.2 0.4 5.2 1.6

0.1 20.1 1.4 13.4 1.3 1.6 0.2 0.4 5.8 1.9

0.2 18.1 1.2 11.3 1.3 1.6 0.2 0.4 5.6 1.8

0.3 19.1 1.2 10.7 1.4 2.1 0.2 0.6 5.2 1.8

0.2 16.3 1.2 10.7 1.3 1.9 0.2 0.7 4.7 1.8

0.3 17.1 1.2 10.3 1.2 1.7 0.2 0.8 4.7 1.5

0.3 18.7 1.0 14.9 1.3 2.1 0.2 0.8 4.4 1.5

Source: Compiled from computer files of US Direct Investment Abroad, Bureau of Economic Analysis, US Department of Commerce.

contracted sharply. As can be seen in Table 5.8, on average their share in total manufacturing employment in all five crisis-affected countries was noticeably higher following the onset of the crisis than in the preceding six years of the decade. Moreover, as we have already noted, fixed capital formation (expenditure on plant and equipment) by these firms also remained steady (or even increased) during the crisis years, despite the massive contractions in total domestic fixed capital formation. It seems that, despite the crisis, the US firms continued to take an optimistic view about the long-run economic prospects for the region. All in all, these findings support the hypothesis that foreign-owned firms have behaved differently from domestic firms in their response to the crisis, aiding the adjustment process in the crisis-affected countries. Fukao (2001) examines the performance of affiliates of Japanese MNEs in the crisis-affected Asian countries, distinguishing between exportoriented and domestic-market-oriented firms and paying particular attention to the role played by parent companies in assisting the subsidiaries.8 The results suggest that export-oriented subsidiaries recorded a 17 per cent increase in total sales, resulting in almost a doubling of their profits between 1996 and 1997. Consequently, overall these firms were able to maintain employment levels while some firms even recruited new workers. Sales and profits of domestic-market-oriented firms did decline substantially, but there was no substantial contraction of employment in these firms, thanks to financial support provided by the parent companies. In particular, subsidiaries owned by large parent firms or by a parent company

116

Multinational enterprises in Asian development

with a larger Keiretsu network in the given host country tended to keep their employment levels unchanged. The general inference from the discussion so far in this section that the presence of MNE subsidiaries facilitates the adjustment process in host countries in the context of a financial crisis receives further support from the findings of available country-level studies which compare the relative performance of MNE affiliates and local firms in the crisis context (Indonesia: Ramstetter 2000, Blalock and Gertler 2005; Thailand: Dollar and Hallward-Driemeier (2000); the Philippines: Lamberte et al. 1999). The common finding of these studies is that MNE affiliates generally perform better than independent local firms in terms of various performance criteria such as maintaining (or even increasing) employment and investment levels, capacity utilization, profitability and export performance. The superior performance of the former is generally attributed to two factors relating to their ‘multinationality’. First, they have the ability to increase exports swiftly through the global marketing network of the MNE in the face of domestic demand contraction. Second, and perhaps more importantly, they are less connected with domestic banks and domestic financial markets and have access to financial support from the parent company or banks in their home countries. Blalock and Gertler (2005) provide an interesting econometric analysis of the way these two factors jointly contributed to the agility of MNE subsidiaries in Indonesian manufacturing in the crisis context. Their results, based on a large firm-level data set for the period 1997–98, suggest that, whereas liquidity constraints denied exporters the opportunity to take advantage of the massive currency devaluation, firms with foreign ownership were able to access credit through their parent company to expand production and exports. In other words, the support of parent firms provided foreign-affiliated firms with a form of liquidity insurance. Consequently they were able to increase capital investment by 8 per cent, domestic employment by 15 per cent and value added by 30 per cent compared with purely local firms during 1997–98. In Table 5.9, we have pieced together a data set to shed some light on the relative performance of MNE affiliates in Malaysian manufacturing during the crisis. The table covers 20 three-digit industries for which the required data are available. Manufacturing performance is measured in terms of three key variables: real output, employment and real wages. For these three variables, deviation from the overall growth trend (for the entire period 1987–2000) during the crisis years (1998–2000) was measured by fitting a least-square trend line with a slope dummy for the crisis years. The MNE presence in the manufacturing sector is measured alternatively as the MNE share in total manufacturing output and employment.

117

311–12 313 314 321 322 331 341 342 351 352 355 356 362 369 371 372 381 382 383 38321 38329

ISIC Code

6.9 0.4 0.9 3.5 4.3 10.3 1.7 2.5 1.0 1.3 4.8 5.2 0.6 3.2 1.7 0.7 4.9 5.2 29.9 5.8 19.6

(2)

(1) 6.9 0.8 1.1 2.9 1.5 5.4 1.7 2.6 6.0 1.8 4.0 3.8 1.0 4.1 2.6 0.9 4.1 5.6 30.5 5.3 21.2

Composition of employment (%)

Composition of value added (%)

20.0 58.4 69.7 51.8 43.7 14.7 13.5 8.3 62.4 64.2 65.3 63.5 17.8 10.5 8.1 64.2 35.5 23.7 95.3 93.4 90.5

(3)

MNE share in value added (%)

13.6 39.1 24.8 42.2 48.2 14.4 10.4 12.0 46.2 46.1 56.1 57.6 29.8 9.6 10.5 75.1 31.3 24.2 96.7 88.7 89.5

(4)

MNE share in employment (%)

MNE presence and post-crisis performance in Malaysian manufacturing1

Food Beverages Tobacco Textiles Wearing apparel Wood and cork products Paper and paper products Printing and publishing Industrial chemicals Non-industrial chemicals Rubber goods Plastic products Glass and glass products Non-metallic minerals Iron and steel products Non-ferrous metals Fabricated metal products Non-electrical machinery Electrical machinery Consumer electronics Semiconductors and electronic components

Table 5.9

12.5 19.5 23.8 11.3 10.7 14.2 9.3 13.9 0.6 9.8 12.6 8.5 6.7 15.8 19.0 4.9 10.4 20.5 8.8 4.1 6.3

Real output2 (5) 21.4 20.7 12.6 25.0 15.4 15.0 8.0 13.6 9.4 8.9 8.9 3.9 14.4 12.7 12.7 10.8 5.3 17.8 11.8 17.5 9.0

(6)

Employment

4.7 2.6 28.3 5.9 8.4 11.0 1.2 16.4 11.9 6.4 5.7 10.7 44.8 2.2 21.3 13.2 8.2 19.0 20.7 5.0 4.6

Real wage4 (7)

Post-crisis trend deviation3

118

100.0

Total manufacturing

100.0

1.9 3.8

(2)

(1) 1.8 6.3

Composition of employment (%)

Composition of value added (%)

Cables and wires Transport equipment

(continued)

44.4

49.9 6.5

(3)

MNE share in value added (%)

45.4

75.4 11.7

(4)

MNE share in employment (%)

9.0

17.6 17.0

Real output2 (5)

11.7

15.4 14.5

(6)

Employment

0.6

20.0 1.7

Real wage4 (7)

Post-crisis trend deviation3

Sources: Wholesale and consumer price indices are from the Ministry of Finance, Malaysia, Economic Survey (various issues). The other data series are compiled from the Department of Statistics, Malaysia, Annual Survey of Manufacturing Industries 1997 (published report and unpublished data on output and employment of MNE affiliates) and Index of Industrial Production (various issues).

Notes: 1. Data in columns 1 to 4 are for 1996 and estimates reported in columns 5–7 are based on annual data for 1987–2000. 2. Growth rate of gross output deflated by wholesale price index of domestic manufacturing. 3. Estimated by fitting the following equation: Log X  
1T
2D*T, where X is the relevant variable (real output, employment, real wage); T is time trend and D is a dummy variable which takes value 1 for 1998–2000 and zero for other years. The post-crisis trend deviation is given by [(2/1]100. 4. Growth rate of nominal wages deflated by the consumer price index.

38391 384

ISIC Code

Table 5.9

Multinational firms in crisis and recovery

119

Table 5.9 provides support for the proposition that MNE presence has acted as a cushion against output and employment contraction during the crisis. The MNE-dominated electronics industry, which accounts for over one third of manufacturing value added and employment, is among the three-digit industries with the lowest measured contraction in output and employment during the crisis years. Many other industries with a higher MNE presence are also at the lower end of the ranking of industries in terms of the degree of employment and output contraction. For the 20 three-digit industries listed in the table, the rank correlation coefficients between the MNE share in sectoral output (column 1) and trend deviation in output and employments during the crisis (1998–2000) (columns 5 and 6) are 0.28 and 0.24 respectively. Both are statistically significant at the 5 per cent level. Industries with greater MNE participation are also generally characterized by lower real wage compression during the crisis. The correlation coefficient between the MNE share in employment and the trend deviation in real wages during the crisis is 0.51, which is significant at the 1 per cent level. In sum, the data presented in this section suggest that MNE affiliates were instrumental in ameliorating the severity of economic collapse and facilitating the recovery process. This finding is consistent with the available studies on the behaviour of MNE affiliates in a number of Latin American countries during the debt crisis in the early 1980s and in Mexico during the 1994–96 financial crisis (Blomström 1990, Blomström and Lipsey 1993, Lipsey 2001).

CONCLUDING REMARKS Contrary to some pessimistic predictions, the 1997–98 Asian financial crisis did not result in a major discontinuity in FDI flows to the region, apart from a modest decline in the immediate aftermath of the crisis. The mass exodus of capital from all five countries was accounted for by foreign portfolio investment and bank credit. While the net long-term national gains from FDI inflows remain a debatable subject, the evidence presented in this chapter suggests that they play a useful stabilizing role in the crisis by limiting the fall in aggregate flows and facilitating the adjustment process. MNE affiliates, both export- and domestic-oriented but particularly the former, seem to contribute to the agility of an economy in the wake of a financial crisis through their ability to maintain output and export levels with the help of their global trading networks. FDI was also found to be much more resilient to a crisis than domestic investment, presumably because of the ability of MNE affiliates to tap international and intra-company financial

120

Multinational enterprises in Asian development

resources in the face of a crisis-driven collapse of the domestic banking system. In the case of Malaysia, there is no clear evidence to suggest that controls on ‘hot money’ movements introduced as part of the crisis management policy package adversely affected that country’s image as an attractive location for FDI.

NOTES * This is an expanded version of Athukorala (2003a). 1. For useful critical assessments of the debate, with extensive listings of the related literature, see Eichengreen (2002) and Lamfalussy (2000). 2. A large number of studies investigating the determinants (or early warning indicators) of currency crises have found that a higher ratio of FDI to total capital flows normally reduces the vulnerability of a country to an exodus of capital (Radelet and Sachs 1998; Frankel and Rose 1996; Kim and Hwang 2000; Lipsey 2001). 3. Data reported in this section, unless otherwise stated, come from the standard IMF sources, in particular various issues of International Financial Statistics and World Economic Outlook. Unless otherwise stated, denomination in dollars refers to US dollars. 4. There is evidence that the component ‘retained earnings’ in FDI is positively related with the age of operation of firms in a given country, and that US MNEs have a general tendency to rely more on retained earnings for investment expansion than MNEs from other countries do (Lipsey 2000). 5. The data sources used here do not permit precise separation of ‘other flows’ from bank credit. But based on tentative estimates derived from other sources (in particular IMF, World Economic Outlook) for all emerging market economies, we believe that more than 80 per cent of the values reported under the ‘Bank credit and other’ item in Table 5.1 consist of bank credit. 6. Unfortunately, the available data do not permit estimation of the relative contribution of M&A to gross FDI inflows. Announced M&A figures (Table 5.3) are generally believed to overstate the realized (balance of payments-based) FDI reported in Table 5.1. 7. As part of policy reforms undertaken in response to the crisis all five countries liberalized their FDI regimes, resulting in a considerable policy convergence among them. For details, see Kim and Hwang (2000); Athukorala (2002a); UNCTAD (1998). 8. The analysis makes use of rare data that bring together firm-level data for 1996 and 1997 from the Quarterly Survey on Trends of Japanese Business Activities Abroad, conducted by the Japanese Ministry of International Trade and Industry (MITI), and data on parent firms from the Japanese Ministry of Finance, Financial Reports of Limited Firms (1998).

6. Capital inflows and the real exchange rate: foreign direct investment versus short-term capital* The ‘real exchange rate problem’ (à la Corden 1994) – the possibility that capital inflows bring about an appreciation of the real exchange rate with adverse implications for the performance of the domestic economy – is central to the contemporary debate on managing capital flows in emerging market economies. This issue has gained added emphasis in the past oneand-a-half decades because of the rapid explosion of capital flows to emerging market economies as part of rapid globalization of capital. The prognosis of the real exchange problem is rooted in the Australian model of a small economy.1 At the heart of this model is a fundamental analytical distinction between goods that are exchanged internationally, termed tradables, and those that are not, termed non-tradables.2 Any excess domestic demand for (or supply of) traded goods is met through foreign trade by running a trade deficit (surplus). The domestic prices of these goods are therefore determined broadly by the world market, subject only to tariffs, export subsidies, international transport costs and, of course, the nominal exchange rate (domestic currency price of foreign currency). By contrast, domestic demand always has to be met by domestic production. Prices of non-tradables are therefore solely determined by supply and demand domestically. Perhaps the most important implication of this distinction between tradables and non-tradables is that the real exchange rate (defined as the domestic relative price of traded to non-traded goods) and thus the internal structure of production in the economy tend to change when the balance between domestic production and domestic absorption is disturbed. Capital inflows are an important source of such disturbance.3 Capital inflows (or any other inward transfer) enable a country to maintain high levels of domestic absorption over and above its domestic production. The increased spending on traded goods will be accommodated through an increase in the trade deficit with no impact on the real exchange rate. By contrast the excess demand for non-traded goods will result in an increase in the price of these goods relative to that of traded goods 121

122

Multinational enterprises in Asian development

(appreciation of the real exchange rate).4 This is, of course, part of the natural economic adjustment resulting from the economy’s choice to absorb foreign capital. But the problem with this natural (equilibrium) phenomenon is that capital inflows may well be temporary, and hence in due course real depreciation is likely, which may require a painful and politically unpalatable economic adjustment. Moreover, capital inflow-induced real appreciation can have an adverse impact on economic adjustment during the boom period, hampering the country’s ability to face such an eventuality. It tends to discourage traded-goods sectors and divert resources to over-consumption or investment in low-yielding non-tradable goods sectors. Furthermore, persistent real exchange rate appreciation could well set the stage for a speculative attack on the national currency, putting an end to the very economic boom fuelled by capital inflows.5 The above discussion on the mechanism linking capital flows and the real exchange rate is based on the implicit assumption that these flows are homogeneous. There are, however, strong reasons for hypothesizing that the degree of real exchange rate appreciation associated with a given level of FDI inflows tends be smaller in magnitude than other flows, in particular portfolio flows and bank lending. First, compared with other flows, FDI tends to have high import content. Imported machinery and equipment account for the bulk of MNEs’ contribution to investment in affiliates, particularly in developing countries. Thus the addition to domestic demand associated with a given (recorded) level of FDI is likely to be much lower than with the same level of other forms of capital inflows. Thus the pressure on non-traded goods prices resulting from FDI-related activities is presumably lower than that arising from the other forms of capital inflow. Second, as we have seen in Chapter 5, FDI is not as volatile as the other short-term flows. Therefore any possible ratchet (lingering) effect on the real exchange rate resulting from upswings in inflows is likely to be less important in the case of FDI. Third, unlike other forms of foreign capital, and perhaps more importantly, FDI comes as a package of investment, new technology and marketing know-how, and thus has the potential to change the production mix under the given production possibility set or to effect an outward shift in the production possibility frontier. Such compositional shifts in production are likely to be associated with diversification into more dynamic product lines with more favourable price trends. This ‘ladder effect’ in international specialization can bring about an increase in aggregate tradable prices relative to non-tradable prices even if the country is a price taker in world markets of individual products. Fourth, productivity improvement in tradable production resulting from FDI participation could tilt the real exchange rate in favour of non-tradables (the Balassa–Samuelson effect).6 For these reasons we treat FDI and other

Capital inflows and the real exchange rate

123

capital inflows as two separate variables in our analysis and assume the effect of the former to be smaller in magnitude than that of the latter. The purpose of this chapter is to examine the impact of capital inflows on the real exchange rate in emerging market economies with emphasis on the differential impact of FDI and other forms of capital inflow through a comparative analysis of the experiences of eight Asian economies (China, India, Indonesia, Korea, Malaysia, the Philippines, Singapore and Thailand) and six Latin American countries (Argentina, Brazil, Chile, Colombia, Mexico and Peru) during the period 1985–2000. Capital inflow is a key explanatory variable used in the literature on the determinants of real exchange rate in developing countries.7 However, to our knowledge this is the first comparative analysis of the nexus of real exchange rate and capital flows paying attention to the possible differential impact of FDI. The study is also unique in its comparative Asian and Latin American focus. There is a sizeable literature on the patterns and determinants of capital flows to Asia and Latin America and problems of macroeconomic management in the context of increased capital mobility.8 However, to our knowledge there has not been any systematic comparative analysis of the nexus of real exchange rate and capital flows encompassing countries in the two regions. The choice of the sample was primarily dictated by data availability, but it contains all major countries in the two regions which were exposed to significant international capital flows in the 1990s. There are notable differences among these countries in terms of the ‘official’ (de jure) exchange rate regime adopted during the study period, ranging from a fixed rate (currency board) regime in Argentina to a ‘freely floating’ regime in the Philippines. But the available evidence on the actual exchange rate practices of these countries (Corden 2002, Edwards 2000, Williamson and Mahar 1998) amply supports the view that during most of the period under study all countries de facto maintained fixed (but adjustable) exchange regimes, with the United States dollar as the key intervention currency. Thus, the country sample provides a near-laboratory setting for studying the issues at hand. The attractiveness of a given country for FDI depends crucially on the general investment climate, which primarily reflects the cumulative outcome of economic policy reforms (Calvo et al. 1996, Feenstra 2000, Corbo and Hernandez 2000). However there is evidence that the composition of capital inflows can also be tilted in favour of FDI (and other long-term flows) through policies specifically aimed at discouraging short-term flows. The most cited evidence of such policy-engineered compositional shift comes from Chile and Colombia.9 There is also evidence from China and Malaysia10 that the government can intervene in short-term flows and still provide a hospitable environment for FDI (Corden 2002, Stiglitz 2000,

124

Multinational enterprises in Asian development

Athukorala 2002a, Garnaut 1999). The disaggregated treatment of capital flows in our analysis should help to determine whether such policy-induced shift in the composition of capital inflows can help to ameliorate the effect of total inflows on the real exchange rate. The rest of the chapter is organized as follows. The next section provides a comparative overview of trends and patterns of capital inflows to the two regions, followed by a preliminary analysis of the relationship between capital flows and the real exchange rate through an examination of timeseries plots. This is followed by an in-depth empirical analysis of the role of capital flows among other variables impacting on real exchange rate behaviour. This section proceeds in two steps. First, a single-equation model is formulated and estimated to delineate the impact of capital inflows on the real exchange rate while controlling for other relevant variables suggested by the theory. Second, the econometric estimates are combined with data on the key variables for individual countries to make inferences about the observed differences between the two regions. The key inferences are summarized in the final section.

TRENDS AND PATTERNS OF CAPITAL INFLOW The post-Second-World-War era has witnessed two major episodes of capital inflow surges to developing countries. The first was associated with the petrodollar recycling process following the oil price increases in the 1970s. The episode started in the second half of the 1970s and lasted until the onset of the debt crisis in 1982 following the Mexican debt moratorium. The second episode, which is the focus of this chapter, began in the latter half of the 1980s and gathered momentum in the first half of the 1990s, with total net inflows to developing countries and flows to many individual countries surpassing the early-1980s peaks by 1994. The flows slowed following the Mexican crisis in late 1994 and plummeted following the onset of the Asian crisis in the second half of 1997. Unlike the previous boom, which was propelled predominantly by a once-and-for-all event, the boom of the 1990s has been the outcome of pull factors – related to better opportunities in the recipient countries – and push factors – related to lower interest rates and slow-down in economic activity in industrial countries. Some of these factors (for instance, global asset diversification of pension funds in developed countries, rapid internationalization of production in high-tech industries, domestic market reforms in developing countries) are of long-term nature in their effect.11 Thus the indications are that large capital flows are here to stay as an important global economic phenomenon.

Capital inflows and the real exchange rate

125

120 100 Asia

80 60 40 20 0 –20

Latin America

–40 –60 –80 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Source: IMF (1995 and 2001).

Figure 6.1 Net private capital flows to Asia and Latin America (US$ billion) Net private capital flows to the emerging markets (developing countries and transition economies) increased from an annual average of less than US$10 billion in the latter half of the 1980s to nearly US$200 billion by the mid-1990s (Figure 6.1).12 Of total annual flows to these countries during 1990–97, almost three-fourths were absorbed by countries in Asia and Latin America. In the early 1990s, total flows to Asia and Latin America were roughly similar in magnitude, but during 1993–96 inflows to Asia surpassed those to Latin America by a wide margin. The decline in flows caused by the Mexican crisis in 1994 was by and large confined to Mexico and Argentina. Thanks to swift action under the IMF–US Treasury deal, its repercussions on Asia (and the rest of the world) were minimal. Total flows to both regions expanded well into 1997. However, the repercussions of the Asian crisis on global capital mobility were much more dramatic. Total flows to Asia plummeted from US$114 billion in 1996 to US$19 billion in 1997 and contracted by a staggering US$55 billion in 1998, and net inflows to the region remained virtually zero until 2001.13 Reflecting the global reverberation of the Asian crisis (which was subsequently amplified by the Russian and Brazilian crises), total flows to Latin America declined from US$68 billion in 1997 to US$39 billion in 2000, arguably a much less dramatic decline in the circumstances. During the 1977–82 capital flow boom, bank loans (and other related flows) accounted for the bulk (over three-fourths) of net inflows to the developing countries. By contrast, foreign direct investment (FDI) and portfolio investment have dominated net flows in the 1990s. FDI has,

126

Multinational enterprises in Asian development

however, been more important in net flows to Asia than in those to Latin America. Table 6.1 provides summary data on total net capital flows to the 14 countries under study, distinguishing between three types of flows: foreign direct investment (FDI), portfolio investment (investment in the form of transaction and debt securities) and bank loans and other types of flows. The first two categories are essentially private flows. The third category includes trade credit (both short-term and long-term) and official capital flows (bilateral and multilateral loans and foreign aid).14 It is evident that, relative to the size of the economy, some of the Asian countries experienced much larger capital inflows than their Latin American counterparts. For instance, Malaysia absorbed inflows amounting to 10 per cent of GDP in 1991, 15 per cent in 1992 and more than 20 per cent in 1993, averaging 12 per cent for the entire boom period of 1989–96. The average annual inflows to Thailand and the Philippines also exceeded 10 per cent of GDP. None of the six Latin America countries experienced capital inflows exceeding 10 per cent of GDP. The Asian countries as a group received a higher share of inflows in the form of FDI. However there were considerable differences in the FDI share among the countries in the region. For instance, South Korea was a net overseas foreign direct investor throughout and total net inflows to that country predominantly took the form of portfolio capital and bank borrowing. The Philippines, Thailand and Indonesia also relied on bank borrowing to a significant extent, although the FDI share in total flows to these countries remained much larger than in those to their Latin American counterparts. FDI accounted for the overwhelming share of net flows to China and Malaysia. FDI flows to Latin American countries (other than Chile) were basically related to privatization of state-owned enterprises, whereas in Asia, particularly in East Asia, such flows were in the form of green-field investment. There is also evidence that a higher share of FDI inflows to Latin America had gone to non-traded sectors (construction and commercial services) and natural resource development, compared with Asia. In Asia FDI was by and large in traded-goods sectors, mostly in export-oriented manufacturing (Ito 2000, Reisen 2000). Finally, the data suggest that across the countries FDI flows have been the least volatile of the three different types of flows (see also Chapter 5). In particular, during the financial crises in Mexico and the five Asian countries FDI flows showed remarkable stability in a context where the other flows shifted sharply into negative territory. Interestingly, net FDI flows to South Korea and Thailand increased in the aftermath of the currency collapse, aided by the competitiveness newly gained through currency depreciation and liberalization of the FDI regimes as part of the crisis management strategy.

127

Capital inflows and the real exchange rate

Table 6.1 Net capital inflows1 to selected Asian and Latin American countries: 1985–99 (percentage of GDP) 1985–892 1990–942

1995

1996

(a) Asia China FDI Portfolio investment Bank loans and other India FDI Portfolio investment Bank loans and other Indonesia FDI Portfolio investment Bank loans and other Korea FDI Portfolio investment Bank loans and other Malaysia FDI Portfolio investment Bank loans and other Philippines FDI Portfolio investment Bank loans and other Singapore FDI Portfolio investment Bank loans and other Thailand FDI Portfolio investment Bank loans and other

1.71 0.51 0.41 0.99 2.14 0.00 0.00 2.12 2.53 0.49 0.02 2.87 2.63 0.11 0.08 2.48 0.51 2.33 1.04 2.83 2.31 1.04 0.18 0.22 2.66 9.34 0.45 6.54 5.04 1.11 1.17 1.95

1.02 2.56 0.23 0.24 2.25 0.12 0.47 1.52 3.26 1.06 0.64 2.01 1.57 0.20 1.50 0.45 11.11 7.10 1.08 3.51 6.32 1.28 0.11 4.84 0.37 6.35 3.91 1.41 9.94 1.64 1.35 6.59

2.98 4.83 0.11 0.58 1.33 0.56 0.44 0.07 3.96 1.85 2.03 1.19 3.18 0.36 2.39 1.40 7.75 4.70 0.49 4.39 4.34 1.46 1.61 4.10 7.00 1.11 8.82 1.95 12.31 0.70 2.43 9.89

(b) Latin America Argentina FDI Portfolio investment Bank loans and other Brazil FDI Portfolio investment Bank loans and other

1.99 0.76 0.74 1.83 4.85 3.10 0.04 7.82

2.32 1.18 3.70 2.47 7.51 2.02 1.25 4.10

1.04 3.59 1.59 1.96 0.72 3.61 0.54 1.40 3.90 8.77 3.40 5.06 0.05 1.60 0.24 3.06

1997

1998

1999

2.98 0.12 2.62 0.71 4.63 4.61 4.26 3.73 0.21 0.77 0.39 1.13 0.02 3.06 4.53 1.83 2.58 1.99 2.33 2.08 0.57 0.83 0.61 0.46 1.03 0.61 0.14 0.51 1.48 0.87 1.54 1.04 5.35 1.51 6.22 2.70 2.46 2.09 0.32 2.00 2.20 1.22 1.50 1.27 0.11 1.14 5.88 0.94 4.69 3.11 4.61 2.16 0.45 0.34 0.21 1.26 2.90 3.02 0.39 2.26 2.03 4.74 2.48 0.49 6.92 2.06 0.67 10.03 5.04 5.13 2.98 1.97 0.27 0.25 0.39 1.02 4.63 2.69 6.89 11.40 10.01 1.53 0.41 5.55 1.61 1.32 3.25 0.57 6.42 0.72 1.42 6.28 5.58 5.87 1.09 8.08 7.12 9.48 21.82 20.08 2.24 0.82 8.48 3.58 12.07 13.65 9.48 8.34 3.98 0.29 25.02 15.91 9.24 10.08 15.14 9.13 0.77 2.32 6.42 4.85 1.94 3.00 0.32 0.06 7.97 13.39 19.51 14.06 5.38 1.88 3.52 0.30 9.18 4.46 3.15 2.17

6.30 1.66 2.94 1.74 2.74 2.52 1.13 2.97

5.11 8.09 2.28 0.36 1.00 6.45 0.19 7.87

128

Table 6.1

Multinational enterprises in Asian development

(continued) 1985–892 1990–942

Chile FDI Portfolio investment Bank loans and other Colombia FDI Portfolio investment Bank loans and other Mexico FDI Portfolio investment Bank loans and other Peru FDI Portfolio investment Bank loans and other

4.85 3.10 0.04 7.82 2.21 1.47 0.13 1.06 0.35 1.20 0.45 1.41 5.34 0.08 0.00 3.74

7.51 2.02 1.25 4.10 1.87 1.37 0.32 0.06 5.66 1.53 3.42 1.50 2.06 1.79 0.33 1.34

1995

1996

1997

1998

1999

3.90 8.77 3.40 5.06 0.05 1.60 0.24 3.06 4.96 6.67 0.77 2.87 1.55 1.71 2.52 1.99 0.55 0.70 3.33 2.76 3.63 4.20 3.37 5.12 6.59 7.70 3.89 5.82 0.30 0.32 1.48 0.07

9.18 4.46 3.15 2.17 5.76 4.53 0.85 1.20 1.86 3.20 1.08 0.52 8.87 2.88 0.26 6.21

2.74 2.52 1.13 2.97 3.86 2.46 1.83 0.02 3.73 2.69 0.32 2.04 3.99 3.30 0.61 0.14

1.00 6.45 0.19 7.87 0.29 1.17 0.89 0.39 2.92 2.46 2.07 0.82 1.85 3.79 0.72 1.87

Notes: 1. Including errors and omissions. 2. Annual average. Source: Compiled from the IMF’s International Financial Statistics database.

REAL EXCHANGE RATE–CAPITAL FLOW NEXUS: A FIRST LOOK This section looks at the capital flow–real exchange rate nexus as a prelude to the econometric analysis in the next section. The real exchange rate (RER) is measured here as the ratio of the export-weighted wholesale price index of trading partner countries expressed in domestic currency relative to the GDP deflator. The rationale behind the choice of this particular measure among various proxy measures of RER is discussed in Appendix 6.2. In Figure 6.2, RER is plotted for each country together with net capital inflow measured as a percentage of GDP (denoted CFW). In each graph, the beginning and end of the capital inflow episode in the 1990s are demarcated with vertical lines. In Table 6.2, the data are summarized for the capital inflow episodes in the 1990s. The general observation that the degree of RER appreciation associated with capital inflow is uniformly much higher in Latin American countries

129

Capital inflows and the real exchange rate

(a) Asia CFW

100.0 RER 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00

87

19

19

86

50.0

00

99

20

98

19

97

19

96

19

95

19

94

19

93

19

92

19

91

19

90

50.0

19

89

19

88

19

87

19

19

86

RER

RER

Indonesia

200.0

CFW

150.0 100.0 RER 00

20

99

19

98

19

97

19

96

19

95

19

94

93

19

92

19

19

91

19

90

89

19

88

CFW (%) 6.00 4.00 2.00 0.00 –2.00 –4.00 –6.00 –8.00

19

87

19

86

50.0 19

19

RER 150.0

Korea

RER

100.0 CFW 00

20

99

19

98

19

97 19

96

95

19

94

19

93

19

92

19

19

91 19

90

89

19

88

87

19

19

19

86

50.0 19

85 19

150.0 100.0

CFW (%) 6.00 4.00 2.00 0.00 –2.00 –4.00 –6.00 –8.00

85

RER

India

CFW

19

19

85

CFW (%) 4.00 3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00

19

RER 150.0

China

19

19

85

CFW (%) 5.00 4.00 3.00 2.00 1.00 0.00 –1.00 –2.00 –3.00

Figure 6.2 Net capital inflows (percentage of GDP) and the real exchange rate (1990  100) for selected Asian and Latin American countries, 1985–2000

130

Multinational enterprises in Asian development

CFW 100.0 RER 00

99

20

98

19

97

19

96

19

95

19

94

19

93

19

92

19

91

19

90

RER 00

99

20

19

98 19

97 19

96 19

95 19

94

93

19

92

19

91

19

19

90 19

89

88

50.0 19

87

100.0

19

86

19

RER 150.0 CFW 100.0

RER

00

99

CFW

20

98

19

97

19

96

19

95

Thailand

19

94

19

93

19

92

19

91

CFW (%) 15.00 10.00 5.00 0.00

19

19

90 19

89 19

88

87 19

86

50.0 19

85

19

19

RER 150.0

Phillippines

Singapore

19

19

85

CFW (%) 15 10 5 0 –5 –10 –15 –20 –25 –30

19

89

19

88

19

87

19

19

86

50.0

CFW (%) 12.00 10.00 CFW 8.00 6.00 4.00 2.00 0.00 –2.00 –4.00 –6.00 –8.00

RER 150.0

100.0

19 95 19 96 19 97 19 98 19 99 20 00

4 19 9

3 19 9

2

19 9

1 19 9

0

9

19 9

8

19 8

7

19 8

6

19 8

19 8

5

RER

19 8

–5.00 –10.00 –15.00 –20.00

RER 150.0

Malaysia

19

19

85

CFW (%) 25.00 20.00 15.00 10.00 5.00 0.00 –5.00 –10.00 –15.00

50.0

Figure 6.2 (continued) than in their Asian counterparts is clearly borne out by this comparison. Focusing on the capital inflow episodes (Figure 6.2), the real exchange rate appreciated by almost 43.5 per cent in Argentina during its early-1990s capital inflow episode, compared with the three preceding years (Table

131

Capital inflows and the real exchange rate

(b) Latin America CFW (%) 10.00

RER 200.0

Argentina CFW

5.00 150.0

0.00 –5.00

100.0

RER

–10.00 –15.00 00

99

20

98

19

97

19

96

19

95

19

94

19

93

19

92

19

00

99

20

19

98 19

97 19

96

95

19

94

19

19

93

92 19

91 19

90

19

RER 150.0

Chile CFW

100.0 RER 00

99

20

19

98

97

19

96

19

95

94

19

19

93

92

19

91

19

90

19

89

19

RER 150.0

Columbia CFW RER

100.0

50.0

86 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00

19

19

19

88

50.0

19

85

19

89

88

19

19

19

87

50.0

CFW (%) 8.00 7.00 6.00 5.00 4.00 3.00 2.00 1.00 0.00 –1.00 –2.00 19

91

100.0

87

86 19

19

85

CFW (%) 15.00 10.00 5.00 0.00 –5.00 –10.00 –15.00

RER 150.0

Brazil

19

86 19

19

85

CFW (%) 5.00 4.00 3.00 2.00 1.00 RER 0.00 –1.00 CFW –2.00 –3.00 –4.00 –5.00

19

90

19

89

19

88

19

87

19

86

19

19

19

85

50.0

Figure 6.2 (continued) 6.2). The rates of appreciation for the other five countries ranged from 14.7 per cent (Brazil) to 33.8 per cent (Mexico). Among the Asian countries, only the Philippines experienced a rate of appreciation that comes closer to the average level for the Latin American countries. The degrees of

132

Multinational enterprises in Asian development

CFW

100.0

00

99

20

98

19

97

19

96

19

95

19

94

19

93

19

92

19

91

19

90

19

89

19

88

50.0

RER 150.0

Peru CFW

100.0 RER 00 20

99 19

98

97

19

96

19

95

19

94

19

93

19

92

19

91

19

90

19

89

19

19

88 19

87 19

86

50.0

19

85

19

87

19

19

86

RER

CFW (%) 10.00 8.00 6.00 4.00 2.00 0.00 –2.00 –4.00 –6.00 –8.00 –10.00 19

RER 150.0

Mexico

19

19

85

CFW (%) 10.00 8.00 6.00 4.00 2.00 0.00 –2.00 –4.00 –6.00

Notes: * Vertical lines denote the beginning and end of capital inflow episodes. The correlation coefficient estimated between change in RER and capital inflow (% of GDP): China0.25; India0.19; Indonesia0.72; Korea0.75; Malaysia 0.40; Philippines0.09; Singapore0.58; Thailand0.51; Argentina0.83; Brazil0.65; Chile0.33; Colombia0.55; Mexico0.67; and Peru 0.84. Source: Based on data compiled from the IMF’s International Financial Statistics database.

Figure 6.2 (continued) appreciation in Indonesia, Malaysia, Singapore, Thailand and the Philippines were quite mild, ranging from 2.3 to 11.2 per cent. In India and China, the real exchange rate continued to depreciate (rather than appreciate) during their inflow episodes. In India, capital inflows occurred against the backdrop of a significant structural adjustment reform. Discretionary devaluation as part of this reform package seems to have offset the impact of capital inflows.15 In early 1994 China decisively reformed its exchange rate mechanism, which resulted in a real depreciation of the yuan by 17 per cent over the previous year, placing the country in a position of strength to withstand massive capital inflows in the subsequent years. The average degree of depreciation reported in Table 6.2 must therefore be viewed in the context of this successful, early exchange rate adjustment (Athukorala and Warr 2002, p. 50).

133

Capital inflows and the real exchange rate

Table 6.2 Capital inflow episodes of selected Asian and Latin American countries Capital inflow episode1

Change in net capital inflow2 (% of GDP)

Maximum annual inflows during the episode

Change in RER3 (%)

Asia China, PR India Indonesia Korea, Rep. of Malaysia Philippines Singapore Thailand

1993–96 1991–94 1990–96 1990–96 1989–96 1989–96 1987–92 1987–95

5.0 0.6 1.6 5.5 5.7 4.8 7.1 9.1

4.4 3.8 5.4 4.7 21.4 10.0 5.8 12.4

6.8 22.0 8.1 7.4 6.9 11.2 2.3 5.5

Latin America Argentina Brazil Chile Colombia Mexico Peru

1990–93 1992–96 1989–97 1992–96 1989–93 1992–97

8.8 3.9 5.6 3.9 6.9 9.0

8.1 4.4 9.3 6.7 7.8 8.8

43.5 14.7 18.6 18.1 33.8 19.7

Notes: 1. The period during which the economy experienced a significant surge in net capital inflows. 2. Percentage change in average net capital inflow to GDP ratio during the episode from the average for the three preceeding years. 3. Percentage change in the average RER (1990100) during the episode relative to the average for the three preceding years. A decrease in the index denotes appreciation. Source:

As for Table 6.1.

REAL EXCHANGE RATE–CAPITAL FLOW NEXUS: EMPIRICAL ANALYSIS In the previous section we observed that, although there were considerable inter-country variations in the relationship between capital inflow and changes in the real exchange rate among the countries under study, the six Latin American countries experienced a uniformly higher degree of real exchange rate appreciation associated with capital inflows than the Asian countries. In this section we proceed to examine these similarities and differences in two steps. First a single-equation model is developed and

134

Multinational enterprises in Asian development

estimated using pooled time-series data for the 15 countries to delineate the like between the real exchange rate and capital inflows while controlling for other factors. Secondly parameter estimates of the model are combined with data on the key explanatory variables to explain inter-group and intercountry differences. Our model aims to explain the behaviour of RER in terms of capital inflows disaggregated into foreign direct investment (FDI) and other capital flows (OCFW),16 both measured relative to GDP, and a set of macroeconomic indicators chosen to represent policies implemented to compensate for the real exchange rate effect of capital flows. The postulated relationships between the dependent variables and the real exchange rate are firmly rooted in the Salter–Swan–Corden–Dornbusch model.17 The options open to policy makers to cushion the real exchange rate against pressure of appreciation arising from capital inflows under a fixed exchange rate are of three types. These are fiscal contraction, sterilizing foreign exchange market interventions and nominal exchange rate adjustment. In our model these policy stances are represented respectively by government expenditure (relative to GDP) (GEXP), excess growth in money supply (M2) measured as the difference between growth in M2 and real GDP growth (EXMG),18 and change in the nominal exchange rate (DNER). In the context of an investment boom funded by capital inflows, fiscal contraction can act as an effective stabilizer in moderating the real exchange rate effect of the boom. The absorption of capital inflows would increase demand for domestic goods and the fiscal contraction would reduce it. In addition to this general demand contraction effect, reduction in government expenditure can have a favourable switching effect because government expenditure tends to be spent more on non-tradables, unlike private consumption. The measure widely used to represent a fiscal policy stance in this type of analysis is the budgetary balance (measured as a ratio of GDP). But we believe that government expenditure is a superior indicator because in the context of an economic boom a country could well experience a ‘revenue surplus’, a reflection of a faster growth in revenue than in expenditure. Meaningful deficit comparison across countries should correct for such biases. Another problem with published data on budget deficits is that different definitions of taxation and borrowing can heavily skew the measured deficit (Sachs and Larrian 1993). When foreign capital flows into a country that has a fixed exchange rate commitment, the central bank is naturally forced to purchase excess flows (that is build up its foreign exchange reserves) in order to maintain the ‘desired’ level of stability in the nominal exchange rate. However, the mere purchase of foreign currency (non-sterilized intervention) is not going to

Capital inflows and the real exchange rate

135

solve the problem. The increase in the domestic money base (of which foreign reserves are a part) results in an increase in domestic money supply, fuelling domestic inflation and appreciating the real exchange rate. But the central bank could, at least for a time, offset this effect by combining reserve accumulation through foreign exchange market intervention with the open market sale of bonds or other monetary action to reduce domestic credit expansion. The variable EXMG is included in the model to test the effectiveness of such sterilized intervention in averting real exchange rate appreciation. Sterilized intervention is a useful policy that can provide some shortterm relief in the context of excessive capital inflow. But, to the extent that sterilization drives short-term interest rates higher, it may perpetrate excess capital inflows and real appreciation. This is going to be a particularly binding constraint if FDI and other long-term inflows account for a large share of total inflows. In such a case, one-to-one sterilization is likely to increase the short-term interest rate, leading to an increase in short-term capital (Ito 2000). Another important problem with sterilized intervention, raised particularly in the Latin American context, relates to the alleged fiscal cost; domestic market intervention of the central bank involves paying a higher interest rate than what is earned on accumulated foreign reserves, which are usually invested in low-yielding short-term securities (Calvo 1991). When such costs build up, authorities are naturally forced to backtrack from their policy commitment to support the currency. For these reasons, sterilized intervention is generally considered a far less effective policy instrument than fiscal contraction in averting the problem of real exchange rate appreciation in the face of large capital inflows. The third policy instrument considered here is nominal exchange rate adjustment. Within the boundaries set by the particular exchange rate regime chosen, countries have the ability to take steps to correct from time to time the disequilibrium in the fixed (but adjustable) nominal rate against the intervention currency (US$). Such exchange rate practice has been referred to in the literature as one of the contributory factors for the success of East Asian countries in maintaining the real exchange rate at realistic levels (Garnaut 1999, Krueger 1997, Reisin 2000). To test the effectiveness of such nominal exchange rate adjustment, we use the annual changes in the nominal exchange rate (DNER) as an explanatory variable. In addition to the above variables, we use openness to trade (OPEN) as another variable. Previous studies on real exchange rate determination in developing countries have found this to be a significant explanatory variable (Edwards and Savastano 2000, Table 13). The underlying hypothesis is that, other things remaining unchanged, greater openness to trade tends to avert undue pressure for the appreciation of the real exchange rate. There

136

Multinational enterprises in Asian development

is no unique measure of openness. Among the available alternatives, we use the Sachs–Warner binary index, which takes value 1 for an open trade regime and zero otherwise (Sachs and Warner 1995).19 On the basis of the above discussion, the real exchange rate function for the ensuing empirical analysis can be specified as RERƒ(FDI, OCFW, GEXP, EXMG, DNER, OPEN, LA*OCFW, LA*FDI, LA*EXMG, LA*GEXP, LA*DNER, LA*OPEN) (6.1) The dependent variable (RER) is the real exchange rate index (as defined in Appendix 6.2). An increase (decrease) in RER indicates real depreciation (appreciation). The independent variables are given below (with the signs expected for the regression coefficients in parentheses): FDI () OCFW () GEXP () EXMG () DNER (
) OPEN (
)

foreign direct investment capital inflow excluding FDI government expenditure excess money growth change in nominal bilateral exchange rate against the US dollar openness.

The remaining variables are the counterpart slope dummy variables for Latin America (LA), where LA takes value 1 for Latin American countries and 0 for the other (Asian) countries. These dummy interaction terms serve to test whether the magnitude of each regression coefficient for Latin American countries as a group differs significantly from the estimate for the overall country sample. The model is estimated using pooled annual data for the 14 countries over the period 1985 to 2000. In addition to the variables mentioned above, we included country-specific intercept dummies (with Thailand as the base dummy) to allow for country-specific fixed effects. The data series on OCFW and FDI were compiled from the International Financial Statistics database of the IMF. All other data series come from the World Development Indicators database of the World Bank. For the purpose of estimation, all variables are used in natural logarithms so that the estimated coefficients can be directly interpreted as elasticities.20 Pre-testing of the explanatory variables for endogeneity (using the Wu–Hausman procedure) suggested that RER and OCFW are jointly determined or more precisely, OCFW is positively correlated with the error term of the equation. We therefore estimated the equation using two-stage least squares (TSLS), instrumentalizing that variable.21

137

Capital inflows and the real exchange rate

Table 6.3 Determinants of the real exchange rate in selected Asian and Latin American countries1 (dependent variableRER)

Variable2 Constant Foreign direct investment (FDI) Capital inflow excluding FDI (OCFW) Excess money growth (EXMS) Government expenditure (GEXP) Change in nominal exchange rate (DNER) Openness (OPEN) Slope dummy variables for Latin America (LA) LA*OCFW LA*FDI LA*EXMS LA*GEXP LA*DNER LA*OPEN N R2 F SE SIV2 (11) RESET2 (1) JBN, 2 (2) ARCH 2 (1)

Equation 1

Equation 2

Parameter (t-ratio)
4.72 (41.96)***
0.32 (3.40)*** 0.55 (2.50)* 0.18 (0.84) 3.18 (3.14)***
0.61 (2.50)**
0.15 (2.67)**

Parameter (t-ratio)
4.71 (56.22)***
0.29 (3.35)*** 0.56 (2.60)**

1.16 (2.91)*** 0.54 (3.37)*** 0.19 (0.83) 0.05 (0.04) 0.56 (2.08)** 0.18 (2.42)** 224 0.50 9.91 0.14 132.35*** 0.10*** 3.26*** 0.88***

3.17 (5.53)***
0.50 (2.48)**
0.15 (2.62)**

1.14 (2.95)*** 0.52 (3.35)***

0.45 (2.18)** 0.18 (2.49)** 224 0.52 11.35 0.14 131.46*** 0.09*** 3.2*** 0.83***

Notes: 1. Country intercept dummies are not reported. 2. The figure in parentheses after each coefficient is the t-ratio of the coefficient. The level of statistical significance is denoted as: *10%, **5% and ***1%. Test statistics SE Standard error of the regression SIV Sargan’s general test of mis-specification in instrumental variable (IV) estimation RESET Ramsey test for functional form mis-specification JBN Jarque–Bera test for the normality of residuals ARCH Engle’s autoregressive conditional heteroscedasticity test.

The estimated full model is reported as Equation 1 in Table 6.3. In this equation, coefficients attached to three explanatory variables – EXMS, LA*EXMS and LA*GEXP – are not statistically significant. The model estimated after dropping these variables is reported as Equation 2.22 To facilitate the interpretation of the results, a summary of the variables used in the regressions is presented in Table 6.4.

138

Multinational enterprises in Asian development

Table 6.4 Change in explanatory variables during the capital inflow boom1 compared with the mean level for the sample period (1985–2000) FDI OCFW GEXP DNER (% of GDP) (% of GDP) (% of GDP) (annual change %) Asia2 China (1993–96) India (1991–94) Indonesia (1990–96) Korea (1990–96) Malaysia (1989–96) Philippines (1989–96) Singapore (1987–92) Thailand (1987–95)

1.1 3.8 3.4 0.8 0.5 4.5 0.1 2.1 1.0

4.3 0.4 0.2 2.2 6.3 6.3 5.8 5.0 8.0

0.5 0.6 0.4 1.0 0.2 2.3 1.9 0.9 1.6

3.4 1.5 0.7 9.6 4.2 4.7 4.0 5.8 4.8

Latin America2 Argentina (1990–93) Brazil (1992–96) Chile (1989–97) Colombia (1992–96) Mexico (1989–93) Peru (1992–97)

0.7 0.0 0.3 0.2 0.1 0.3 3.5

7.5 8.8 3.7 14.1 3.8 6.4 5.5

0.9 2.6 0.8 1.3 2.3 1.9 1.5

41.4 102.3 11.0 8.9 6.6 29.1 124.5

Notes: 1. Capital inflow boom period is given in brackets next to the country name in column 1. 2. Simple average for the listed countries. Source: Based on date compiled from the IMF’s International Financial Statistics database.

Both equations pass the standard F test for overall significance at the 1 per cent level. The overall fit (R2) is highly satisfactory for an econometric exercise based on pooled cross-country data. The equations also comfortably pass the standard diagnostic tests for functional form specification (RESET), normality (JBN) and heteroscedasticity (ARCH). The specification of the first-stage regression in instrumental variable estimation (IV) is amply supported by Sargan’s test. The following discussion is based on Equation 2, which is our preferred model. Let us first consider the results for the two capital inflow variables (OCFW and FDI). For all countries an average 1 per cent increase in OCFW brings about 0.56 per cent appreciation in the real exchange rate. By contrast, FDI inflows are associated with depreciation (rather than appreciation) of the real exchange rate. The results for the respective slope dummy variables (LA*OCFW and LA*FDI), however, suggest that the

Capital inflows and the real exchange rate

139

magnitudes of these elasticities for the Latin American country sub-group differ significantly from the overall estimates. In Latin America a 1 per cent increase in OCFW brings about, on average, 1.7 per cent appreciation in the real exchange rate. The degree of depreciation associated with a 1 per cent increase in FDI flow is much smaller for the region (0.06 per cent) than for the entire country sample (0.56 per cent). These estimates yield two important inferences. First, in both regions ‘the real exchange rate problem’ is a phenomenon specifically associated with ‘other’ capital flows (OCFW). Second, these flows have a greater dampening impact on the real exchange rate in Latin America than on that in Asia. At least part of the explanation for this difference may be rooted in differences in the macroeconomic policy histories of the two regions.23 For instance, given the proven track record of maintaining macroeconomic stability, it is likely that the pressure on non-tradable prices resulting from resource transfers to the Asian countries will be counterbalanced by an increase in domestic saving and investment. Why FDI inflows tend to depreciate (rather than appreciate) the real exchange rate is an interesting issue which requires further investigation. However, we believe that this result is consistent with our hypothesis that FDI in the tradable sector could result in an improvement in aggregate tradable prices through compositional shift in domestic output. The lower magnitude of the measured relationship for Latin America may reflect differences in the output composition of FDI-related activities in the two regions; as noted earlier, there is evidence that FDI-related activities in Latin America have a greater non-tradable bias. The coefficient attached to GEXP is uniformly applicable to both regions as the coefficient on the slope dummy variable, LA*GEXP, is not statistically different from zero. It suggests that a 1 per cent contraction in government expenditure to GDP ratio is associated with 3.17 per cent depreciation in the real exchange rate. Thus our results support the theoretical proposition that fiscal contraction is a powerful cushion for all countries against real exchange appreciation associated with capital inflows. This result for GEXP when combined with the data on this variable for individual countries (Table 6.4) provides a powerful explanation as to why the Asian countries experienced a rather mild appreciation in the real exchange rate compared with the experience of their Latin counterparts. In the former countries government expenditure remained flat or even declined during the period of the boom. In Malaysia and Thailand government expenditure in relation to GDP declined in the wake of capital influx in the early 1990s. In Latin America, in all the countries except Chile government expenditure increased, and Brazil and Argentina recorded increases of more than four percentage points. The only Asian country to record an increase in government expenditure during the boom was the Philippines.

140

Multinational enterprises in Asian development

As already noted, the excess money growth variable (EXMS) failed to yield a statistically significant coefficient and the coefficient of the other variables were remarkably resilient to its inclusion in (or exclusion from) the model. This result is consistent with the widely held views about the impotence of sterilized intervention as a policy tool in averting real exchange rate appreciation in the face of high capital inflows. Finally, the results for DNER suggest that an average 1 per cent depreciation in the nominal exchange rate change translates into 0.5 per cent depreciation in the real exchange rate. However, the coefficient estimate for the related slope dummy variable (LA*DNER) suggests that the impact is quite small (0.05 per cent) for the Latin American countries. These results are consistent with the view that the Asian countries have been more successful in averting real exchange rate appreciation through nominal exchange rate adjustment than the Latin American countries (Little et al. 1993, Corden 1996, Reisen 2000). The explanation seems to lie in wellknown wage-price rigidities, in particular wage indexation, in the latter countries which serve to quickly dissipate the impact of nominal exchange rate depreciation (Little et al. 1993, Chapter 7).

CONCLUSION Our econometric results suggest that the composition of capital flows does matter in determining their impact on the real exchange rate. ‘The real exchange rate problem’ is a phenomenon predominantly associated with ‘other’ capital flows (OCFW) and FDI seems to have a salutary effect on the real exchange rate. Moreover, owing perhaps to the legacy of a traumatic macroeconomic policy history, a given level of OCFW brings about a far greater degree of appreciation of the real exchange in Latin America than in Asia. This difference in the magnitude of the observed relationship, coupled with the greater importance of OCFW in total capital flows to Latin American countries, seems to go a long way in expanding the higher degree of real exchange rate appreciation experienced by these countries compared with their Asian counterparts. In the 1990s, the major capital-importing countries in Asia managed to cope far better with the real exchange problem associated with capital flows than their Latin American counterparts. The degree of real exchange rate appreciation associated with capital inflows was uniformly much lower in the Asian countries, despite the fact that some of these countries experienced far greater foreign capital inflows in relation to the size of their economies. According to the results of our empirical analysis, the explanation of these differences seems to lie in both country-specific factors (as

Capital inflows and the real exchange rate

141

reflected in differences in the coefficients in the real exchange rate function) and policy differences (as reflected in the levels of the relevant explanatory variables). Among the policy variables considered, fiscal contraction seems to have been used more effectively by the Asian countries to cushion the real exchange rate against the appreciation pressure of capital inflows. There is, however, no evidence to suggest that sterilized intervention can generate a lasting impact on the real exchange rate. Finally, nominal exchange rate change seems to have a significant lasting effect on the real exchange rate only in Asian countries. Owing perhaps to wage–price rigidities, the impact of a given nominal exchange rate change seems to dissipate quickly in Latin American countries.

NOTES * 1. 2.

3. 4. 5.

6.

7. 8. 9.

10.

Adapted from Athukorala and Rajapatirana (2003). So named because of its origin in the seminal works of Australian economists, Salter (1959), Swan (1960) and Corden (1960). Tradability/non-tradability of a product depends on international transport costs, and on trade taxes and other barriers to trade. In practice the line between the two categories cannot be clearly drawn because the trade-impeding effect of transport costs and trade restrictions is a matter of degree. In general most (but not all) primary commodities and manufactured goods are examples of tradables, while many (but not all) services, construction and public utilities are examples of non-tradables. The other important factors include external economic shocks such as world commodity price boom, domestic resource discovery, terms of trade shock, disruption to foreign borrowing (debt crisis) and domestic macroeconomic policy shifts. For a diagrammatic exposition, see Appendix 6.1. In the recent literature on currency crises, persistent appreciation of the real exchange rate (adjusted for fundamentals) has been identified as a major factor in setting the scene for a crisis. This is because a persistent real appreciation implies that economic fundamentals of the country may not permit the authorities to defend the currency successfully in the event of a speculative attack (Kaminsky et al. 1997, Sachs et al. 1996). Assuming purchasing parity holds for the prices of tradables, the prices of non-tradables tend to increase relative to tradables when the relative productivity of labour in the latter sector increases. For a succinct exposition of this proposition, see Sachs and Larrian (1993, pp. 672–5). Edwards and Savastano (2000, pp. 488–90 and Table 13.5) provide a comprehensive survey of various studies of real exchange rate determination. See for instance Calvo et al. (1994 and 1996), Gavin et al. (1996), Corbo and Hernadez (2000), Dornbusch and Warner (1994), Edwards 2000, Ito (2000), Larian (2000) and Fernandez-Arias and Montiel (1996). See Larian 2000, pp. 11–12 and the work cited therein. Note that here we refer to composition shift, rather than the total volume, of capital inflows. Whether the Chilean and Colombian capital controls have had any dampening effect on the volume of inflow still remains a controversial issue (Edwards 2000). Here we refer to the controls introduced by the Malaysian authorities on short-term capital inflows in 1994. The Malaysian capital controls introduced in 1998 were on outflows and are not relevant to the discussion here.

142 11. 12.

13. 14. 15. 16. 17. 18. 19.

20. 21. 22. 23.

Multinational enterprises in Asian development The relative importance of the pull and push factors has been extensively debated in the literature. See, for instance Corbo and Hernandez 2000 and the works referred to therein. Data on total capital flows (public
private) covering the entire period and with the required disaggregation are not readily available. But private capital flows depicted in Figure 6.1 provide an accurate picture of the trends in total capital flows, because net public flows dwindled from about the late-1980s, reaching less than US$10 billion (less than 5 per cent of total annual flows) by the mid-1990s. In the five crisis-affected countries in Asia (Thailand, Malaysia, Indonesia, Korea and the Philippines) the turnaround in capital flows during 1996–97 amounted to US$105 billion, more than 10 per cent of the combined GDP of these economies. Ideally, we should disaggregate the third category into private and official flows, but the required data are available only for a few countries. In any case, though, the post-war era capital inflow to India had been very small relative to the size of the economy (Athreye and Kapur 2001). OCFW covers both private and public flows. Data are not available for most of the countries under study to treat them separately. We do not spell out the dependent economy model here, for want of space. The interested reader is referred to Corden 1994, Chapter 1, which contains perhaps the best nontechnical exposition of the model. This measure assumes that money demand has a unitary elasticity with respect to real income. Sachs and Warner (1995) employ the following policy criteria to distinguish countries with closed (inward-oriented) policy regimes from those with open (outward-oriented) policy regimes: (i) non-tariff barrier coverage of intermediate and capital goods imports of 40 per cent or more, (ii) an average tariff on intermediate and capital goods imports of 40 per cent or more, (iii) a black market exchange rate that is depreciated by 20 per cent or more relative to the official exchange rate, (iv) a socialist economic system and (v) a state monopoly on major exports. The trade policy regime of a given country/year is identified as open if none of the above five conditions is applicable. Alternative regression estimates based on two other widely used alternative measures of openness – the black market exchange rate premium and the ratio of total merchandise trade (imports
exports) to GDP in traded-goods sectors (GDP net of services, construction and utilities) – produced virtually similar results. RER is directly converted into logarithms and DNER is measured as annual differences in logarithms. The remaining variables are first measured as ratios of GDP and then converted into logarithms as ln (1
X). The instruments were LIBOR, real OECD GDP, a time variable, GDP growth, gross domestic fixed capital formation (relative to GDP), lagged dependent variable, current and lagged values of other explanatory variables and country intercept dummies. This specification choice was amply supported by the Wald test for joint variable deletion. The coefficient estimates of the remaining variables are remarkably robust to the deletion of these three variables. For details on macroeconomic policies and experiences, see Corden (1996), Little et al. (1993) and Edwards (1995).

Capital inflows and the real exchange rate

143

APPENDIX 6.1 THE AUSTRALIAN MODEL: CAPITAL INFLOWS, THE REAL EXCHANGE RATE, AND ECONOMIC ADJUSTMENT IN A SMALL OPEN ECONOMY We consider a ‘small’ country which faces given world prices for its imports and exports. At these given prices, importables and exportables can be combined into a composite traded good (a tradable). The domestic price of this good is then given by the fixed foreign currency prices of the two component tradables and the nominal exchange rate, which is assumed to be fixed. Any domestic excess demand or supply of this traded good is met through foreign trade, with excess demand (supply) being met by running a trade deficit (surplus). The economy also produces a nontraded good (a non-tradable), a good which, because of either high transport cost or prohibitive trade taxes, does not enter foreign trade. When the domestic production of the economy is collapsed into these two composite goods – a tradable and a non-tradable – the crucial price in the model is the relative price of traded to non-traded good, or the real exchange rate. In Figure 6.A1 tradables are shown on the vertical axis and nontradables on the horizontal axis. The schedule NT is the constant employment transformation curve, which shows production possibilities of these goods. Social preferences for aggregate absorption of the two goods are shown by indifference curves such as U1 and U2. The curve OZ traces out the pattern of demand between tradables and non-tradables as expenditure changes. It is upward sloping because both goods are considered here as normal goods (expenditure elasticity of demand is positive in both cases). Consider point A, where U1 is tangent to NT. At this point domestic demand and supply for both tradable and non-tradables are equal, or the economy is in both internal balance (full employment) and external balance (zero current trade balance). (We ignore international trade in services here; that is, trade balancecurrent account balance.) The slope of the price line Pa which is tangent to NT at A, indicates the domestic relative price of tradables to non-tradables (the real exchange rate) appropriate to internal balance and external balance. Capital inflows are an increment to the economy’s absorption capacity. This is shown by a vertical shift in the production possibility frontier from NT to N T . The vertical distance between the two curves (AB) is equivalent to the amount of capital inflows. If the economy absorbs the entire amount of foreign capital inflows in the form of tradables, the real exchange rate and hence domestic production will

144

Multinational enterprises in Asian development

T


Pb = Pa Z Pa

T

B

Tradables

C A U2

D

U1 Pc

N
Pd = Pc

N

0 Non-tradables

Figure 6.A1 Capital inflows, the real exchange rate and economic adjustment in a small open economy remain unchanged. Such a situation is shown by point B, lying vertically above A. But point B is not an equilibrium because it is not on the line OZ. The commodity mix at this point contains more tradables and less nontradables than the desired pattern of demand between tradables and nontradables in the economy. At this higher absorption level, the consumers desire more tradables as well as non-tradables than at point A. To redress this imbalance there has to be a compositional shift in domestic production from tradables to non-tradables. This in turn requires an increase in nontradable prices relative to tradable prices, or an appreciation in the real exchange rate. The new equilibrium under the given level of capital flows is shown by point C, lying on the highest attainable indifference curve, U2. At that point, demand and supply for non-tradables are equal, and demand for tradables exceeds their domestic supply by BD ( AB, the amount of capital inflows). Note that the slope of line Pc which is tangent to point C is steeper than that of line Pb ( Pa). This implies that the move from point B to point C has involved an increase in the price of non-tradables relative to tradables, corresponding to an appreciation of the real exchange rate.

Capital inflows and the real exchange rate

145

APPENDIX 6.2 MEASUREMENT OF THE REAL EXCHANGE RATE The real exchange rate (RER) is the price of traded goods relative to the price of non-traded goods. In the absence of readily available indices of tradable and non-tradable prices, the real exchange rate has to be proxied by available domestic and world price indices and nominal exchange rates. There is no unique way of constructing a proxy measure, but all commonly used measures compute the ratio, RER 

[NER]PW PD

where NER denotes the nominal exchange rate (measured as domestic currency per foreign currency), PW is an index of foreign prices and PD is an index of domestic prices. NER and PW are weighted averages computed across trading partner countries. The country weights may be based on export shares, import shares or, most commonly, shares based on the sum of exports and imports, although the latter has no apparent theoretical foundation. The particular measure used differs according to the measures used for PW and PD. Our preferred proxy measure makes use of foreign producer (wholesale) prices for PW and the GDP deflator for PD. Country weights based on export shares are used in the construction of NER and PW series.1 By construction, the producer price index is dominated by the prices of tradables much more than by the GDP deflator. The index may thus serve as a rough proxy for the theoretical concept of the real exchange rate – the relative prices of tradable to non-tradable goods. A convenient alternative to the GDP deflator as the domestic price measure in constructing the index is the consumer price index (CPI) (Edwards 1989, Athukorala and Warr 2002). Our preferred choice is the former, for two reasons. Firstly, the GDP deflator provides a much broader, economy-wide coverage of price changes compared with the CPI. Secondly, and perhaps more importantly, in most countries the CPI, being a ‘politically visible’ economic indicator, is susceptible to manipulation. Most of the previous studies have typically used either of two other indicators, although the theoretical reasoning behind the particular measurement choice is seldom made explicit. One, which is perhaps the most widely used, particularly in publications of the IMF and the World Bank, uses a trade-weighted index of consumer prices in trading partner countries for PW and an index of consumer prices in the given country for PD. The use of this indicator as a proxy for the theoretical concept of a real exchange rate for developing countries is usually justified on the premise that, under

146

Multinational enterprises in Asian development

the low inflation conditions that prevail in developed countries (which are generally the major trading partners), producer prices and consumer prices tend to move together (Edwards 1989). The other is the J.P. Morgan index, which uses wholesale non-food manufacturing prices for both world and domestic prices. This measure may thus be viewed as an indicator of the international competitiveness of manufacturing goods produced in a given country. It is not a measure of the relative profitability of domestic production of traded goods compared to non-traded goods (internal competitiveness), which is the theoretically more appropriate concept for the present analysis. Wholesale prices of traded goods generally adjust to exchange rate changes and the dismantling of trade barriers and are thus likely to deviate from the price trends of non-traded goods.

NOTE 1. For a discussion on the conceptual basis for using export weights (rather than commonly used trade (export
import) weights) see Warr (1986).

7. Trade orientation and productivity gains from international production* The role of multinational enterprises (MNEs) in the growth process of host countries has long been a topic of intense debate. An important issue arising from the debate is the role of the domestic trade policy regime in determining the net national gains from production undertaken by foreign affiliates. Using an implication from the theory of immiserizing growth, Jagdish N. Bhagwati (1973) first noted the possibility that, under a restrictive trade regime, returns to international production may fall short of its social marginal cost. This proposition was further developed and integrated into the theory of capital mobility by Richard A. Brecher and Carlos F. Diaz-Alejandro (1977) and Richard A. Brecher and Ronald Findlay (1983). These authors demonstrated that the national income of the host country is likely to decline when foreign capital flows are attracted by artificially high rates of return induced by a restrictive trade regime. In his subsequent work, Bhagwati (1978, 1985, 1994) argued that this analytical finding may be a significant element in the explanation of the observed economic failure of import-substitution (IS) economies compared with export-promoting (EP) (or outward-oriented) economies.1 Put simply, Bhagwati’s hypothesis is that an outward-oriented trade regime has the potential to reap greater benefit from international production than an IS regime. This is because, in contrast with IS regimes, EP regimes encourage foreign direct investment (FDI) in activities where the host country has a comparative advantage. Despite the strong assertion from theory, the relevance of the nature of trade orientation for host-country gains from FDI has attracted little attention, both within policy circles and among applied economists working on the impact of FDI on growth patterns. At the policy level, many countries still resort to trade restrictions with a view to enticing MNEs to relocate production in priority sectors. And, at times, countries that embark on significant liberalization reforms tend to retain and even increase special protection to industries dominated by MNEs. In empirical studies of the determinants of growth patterns, aggregate FDI inflow (usually as a ratio 147

148

Multinational enterprises in Asian development

of GDP) is commonly used as a source of technology, without distinguishing between FDI induced by domestic protection and that undertaken predominantly on export profitability considerations. The mismatch between theory and practice may be because of the paucity of convincing empirical evidence on the real-world relevance of the theoretical postulates. The first, and to the best of our knowledge so far the only, empirical study of this subject is by V.N. Balasubramanyam et al. (1996). This study examines the impact of FDI on economic growth by estimating a growth equation that incorporates FDI as an additional factor input for 46 developing countries. The impact of trade regime is taken into account by estimating the equation separately for export-promoting (EP) and importsubstituting (IS) countries identified on the basis of a binary classification of trade regimes. The results provide empirical support for the proposition that FDI inflows have a more significant and positive impact on GDP growth in EP countries than in IS countries. The study however suffers from some limitations. First, FDI participation is measured in terms of the inflow of direct investment from abroad based on flow data provided by the standard balance-of-payments accounts. As the authors have candidly acknowledged, this is no more than a rough proxy for international production. Furthermore, FDI inflows as captured in balance-of-payments accounts generally understate capital spending of foreign affiliates,2 and the magnitude of the error tends to vary significantly across countries (Grubert and Mutti 1992). Second, as the trade orientation is measured in terms of a binary variable (1 for EP economies and zero for others), inter-country differences in the degree of trade orientation are not appropriately captured in this analysis. Thirdly, and perhaps less importantly, the use of aggregate investment data has the limitation of ignoring differences in technology among source countries. There is evidence that characteristics specific to source countries have significant implications for the growth impact of international production on the recipient country (Kokko 1994, Caves 1996). The purpose of this chapter is to undertake an empirical inquiry into the relationship between trade orientation and national gains from international production using a rich, and yet hitherto unexploited, data set on overseas operations of United States MNEs which enables us to avoid these pitfalls.3 The trade policy regime impacts on the national gains (and losses) from international production activities in a number of respects (Blomström 1991, Caves 1996). In this study, we select one such aspect for deeper scrutiny, namely productivity growth of international production itself.4 Apart from the need for narrowing the focus for deeper empirical analysis, this subject choice is justified by the current policy emphasis on international production as a main vehicle for promoting productivity growth in developing countries.

Trade orientation and productivity gains

149

The rest of the chapter is structured as follows. The next section presents the analytical framework. Data sources and methods of data compilation are discussed in the following section. This is followed by a brief discussion on inter-country differences in productivity and the key explanatory variables in order to provide the setting for interpreting the results. Then the results are presented and discussed, followed by concluding remarks. The results provide strong support for the proposition that ceteris paribus an open trading regime enhances the productivity of FDI. The results also point to a significant negative effect of a stringent domestic tax regime on efficiency gains from international production.

ANALYTICAL FRAMEWORK Consider a standard aggregate production function for the foreign affiliate: Y AF(K, L)

(7.1)

where Y is output (value added), K is capital and L labour and A is total factor productivity. Total factor productivity in turn is given by: AA(t, Z)

(7.2)

where Z represents the state of technology5 available to the affiliate and t captures exogenous changes in technology over time. It is assumed that the ability to absorb technology and adapt it to suit individual host-country situations depends on the nature of the trade policy regime of the host country (j): Zj Z(j)

(7.3)

where  is an index of outward orientation of the policy regime. Substituting (3) in (2) and differentiating with respect to time gives: A  0
 ␶

(7.4)

where a tilde above a variable denotes the rate of growth and 0 captures exogenous technological progress. According to Equation (7.4), the relevant measure of the dependent variable of our analysis is total factor productivity growth (TFPG), that part of output growth which is left unexplained by the growth of inputs.6 The index used here to measure TFPG is the Tornquist–Theil formulation (as

150

Multinational enterprises in Asian development

modified by Caves et al. (1982) for fixed cross-sectional samples) which satisfies several desirable characteristics of index numbers. This is also perhaps the most intuitive formulation of total factor productivity growth. The measurement procedure and its desirable features and limitations are discussed in Appendix 7.1. We also use the growth of labour productivity (value added per worker) (LPG) as an alternative indicator of productivity performance. LPG is obviously a less precise measure than TFPG. In reality, workers may produce more not only because of an increase in efficiency but also because they have more inputs (capital, in the two-factor case) to work with. Thus LPG spuriously captures changes in capital per worker as part of measured productivity.7 However, it is important to see the sensitivity of the results to the use of LPG in place of TFPG, because the former is the most widely used (and oldest) indicator of factor productivity. Measuring the restrictiveness of a trade policy regime (or the degree of outward orientation) is a controversial subject.8 The data set used in this study permits us to measure this variable directly in terms of the export propensity – exports as a percentage of total output (XOR) – of foreign affiliates. The underlying assumption here is that foreign affiliates located in more outward-oriented countries generally tend to be more export oriented than those located in countries with restrictive trade regimes. This assumption is largely consistent with the available evidence on the relationship between international production and trade patterns across countries. There is ample evidence that, given scale economies and the very small domestic markets of most developing countries, a foreign affiliate will locate there to serve the international market by exporting extensively, and that the choice of location depends crucially on the nature of the trade regime (Caves 1996, p. 215; Hufbauer et al. 1994).9 An obvious limitation of XOR is that, under certain circumstances, the degree of export orientation of foreign affiliates may be influenced by various factors which are not directly related to the trade regime. For instance, foreign affiliates located in countries with large domestic markets tend to be less export oriented than those located in countries with smaller domestic markets (Blomström and Lipsey 1993). Also, foreign affiliates are better placed than local firms to redirect their sales in the event of domestic demand contraction (as is evident from the performance of foreign affiliates in Latin American countries during the debt crisis in the early 1980s) (Blomström and Lipsey 1993). Specific performance criteria adopted by host-country governments as part of the regulatory mechanism relating to foreign affiliates (for example, linking profit remittance permits to export performance, as in the case of India) are another possible extraneous influence.

Trade orientation and productivity gains

151

For these reasons, we use two alternative measures of outward orientation of the economy as a check on the results based on our preferred measure (export–sales ratio), which is specific to international production.10 These are a binary index of openness (1 for open economies and zero otherwise), based on the policy criteria suggested by Jeffrey D. Sachs and Andrew Warner (1995), and the black market premium on the official exchange rate. Sachs and Warner employ the following trade-policy-related indicators to distinguish between closed and open economies: (i) non-tariff barrier coverage of intermediate and capital goods imports of 40 per cent or more; (ii) average tariff on intermediate and capital goods imports of 40 per cent or more; (iii) a black market exchange rate that is depreciated by 20 per cent or more relative to the official exchange rate; (iv) a socialist economic system; and (v) a state monopoly on major exports. Using the information provided in Sachs and Warner (1995), we designate an economy as closed if it satisfies all five criteria for the duration of the entire period covered in this study (1983–1992). As noted, the black market premium is one of the five criteria on which the Sachs–Warner index is based. In fact, in most cases, it was the ‘decisive variable’ in the Sachs–Warner categorization of economies as closed or open (Sachs and Warner 1995, p. 106). However, we use this as an alternative indicator for the following reasons. First, despite some conceptual limitations as an indicator of policy-induced openness,11 there is consistent evidence that restrictive trade regimes are generally characterized by a high black market premium on the official exchange rate (Krueger 1978, Michaely et al. 1991). It is therefore a good proxy for the overall extent of distortions in the external trade and payments regime. Secondly, it has been one of the few continuous measures of trade restrictiveness that has been found consistently to be statistically significant in the recent empirical literature on the relationship between openness and growth rate differentials across countries (Edwards 1998, Harrison 1996). Thirdly, this measure, like the export–sale ratio (or other indicators of openness derived from observed data), allows for ‘intermediate situations’ of trade orientation, whereby countries are neither totally open nor totally closed. In order to delineate appropriately the impact of trade orientation on productivity, it is important to control for other possible influences on the latter. Guided by the theory and previous empirical work on the determinants of inter-country changes in productivity, we use six additional explanatory variables. These are (1) scale of operation (SCL) proxied by total sales volume, (2) the effective tax rate measured as the share of total taxes in total gross income (TAX), (3) the share of electrical and electronic equipment in total output (ELSH), (4) initial (base-year) level of

152

Multinational enterprises in Asian development

productivity (PRBY) and (5 and 6) two intercept dummy variables to distinguish between industrial countries and high-performing Asian economies from the other countries in the sample. A potential cause of productivity growth is scale economies arising from the market size. Total sales volume of United States foreign affiliates is used as a proxy variable to control for this influence. The openness of the trade regime does not necessarily go with a domestic tax regime that is conducive to R&D activities of foreign affiliates. Countries with relatively open trade regimes could well have stringent tax regimes12 that act as a potential deterrent to productivity improvement. The choice of the effective rate of taxation (TAX) as an explanatory variable is based on this consideration.13 More specifically, the underlying hypothesis here is that higher corporate taxes, much like weak property rights, ceteris paribus discourage long-term investment in productivity improvement by foreign affiliates (independently of the impact of such taxes on the entry decision and the initial level of investment). A well-known feature of overseas operations of United States manufacturing MNEs is their heavy involvement in assembly activities (or ‘slicing product chain activities’ according to Krugman (1995)) in the electrical and electronics industry. These activities, particularly in developing countries, are relatively labour intensive. Moreover, assembly processes generally involve limited technological adaptation/diffusion compared with normal production process (Grunwald and Flamm 1985). Given these features, one may hypothesize inter-country differences in productivity growth in aggregate production of foreign affiliates to be negatively related to the differences in the degree of reliance on assembly activities. On these grounds, the share of electrical and electronic products in total manufacturing output is included as an explanatory variable. Given the tendency for convergence in growth rates over time generally found in previous studies,14 the initial (base-year) level of productivity is chosen as an additional explanatory variable. Should there be convergence in TFP growth rates after allowing for the other explanatory variables, the coefficient on the initial level of productivity will be negative and statistically significant. Finally, two dummy intercept variables – industrial country dummy (DIC) and high-performing Asian economy dummy (DHPAE) – are included to allow for the possible impact of the level of development of the host country on the productivity of international production.15 These variables are expected to capture productivity implications of differences in the quality of public infrastructure, property rights regime and transparency/ stability of related policies, all of which are positively related to the level of economic development. Ideally, one would like to capture these effects sep-

Trade orientation and productivity gains

153

arately by using continuous measures, but this is not feasible because of the paucity of the required data. In the empirical analysis we focus on inter-country variations in productivity growth. The estimating equation is: Ai 0
1OPENi
2TAXi
3SCLi
4ELSHi
5PRBYi
6DICi
7DHPAE


(7.5)

The variables (with the sign expected for the regression coefficient in brackets) are: A

productivity growth of foreign affiliates measured alternatively in terms of: LPG labour productivity growth and TFPG total factor productivity growth; openness (outward orientation) measured in terms of:

OPEN XOR (
) SWI (
) BMP ()

export–sales ratio, Sachs–Warner binary index, which takes a value of 1 for open economies and 0 otherwise, and black market premium on the official exchange rate;

SCL (
) TAX

scale of operation proxied by the sales volume in logs; tax rate measures as the share of total taxes in total gross income PRBY initial (base-year) level of productivity represented alternatively by: LPB () base-year level of labour productivity (in logs) and TFPB () base-year level of total factor productivity (in logs); DIC (
) DHPAE (
) 0  i

a dummy variable, which takes a value of 1 for mature industrial countries and 0 otherwise; a dummy variable, which takes a value of 1 for highperforming Asian economies and 0 otherwise; constant term; a stochastic error term; a country subscript.

DATA The data series for BMP and SWI come from Sachs and Warner (1995). The construction of DIC and DHPAE is based on the country classification in

154

Multinational enterprises in Asian development

Table 7.1. All other data series were compiled from reports published in the Survey of U.S. Direct Investment Abroad conducted by the Bureau of Economic Analysis (BEA) of the United States Department of Commerce.16 This is the most comprehensive and consistent source of data on international production by MNEs available for any home country. The use of data relating to MNEs from a single source (home) country has the advantage that the results are automatically controlled for differences in the nature of technology and access to technology, as well as other sourcecountry-specific factors impinging on international operations of MNEs (Pratten 1976). In other words, this data set provides a near-laboratory situation for examining the effects of host-country policy regimes on international production. The data set covers production by majority-owned United States foreign affiliates in 44 countries (listed in Table 7.1) over the period 1983–92. During that period, the 44 countries on average accounted for over 90 per cent of the total overseas production (measured in value added terms) of majority-owned foreign affiliates and for over three-quarters of that of all affiliates of United States MNEs. Output is measured in terms of value added. The data on the capital stock relate to net property, plant and equipment. Labour input is measured in terms of the number of employees.17 The data on output and capital stock by the BEA are in current dollars (estimates in host-country currency converted by the average market exchange rate (MER)). We derived real dollar-denominated output and capital stock series by deflating the original series, respectively, with the implicit price deflator for United States gross domestic product originating in non-petroleum manufacturing industries and the subindex for investment goods in the United States producer price index.18 The data are for total manufacturing, defined to cover all product sectors classified under Division 3 of the International Standard Industry Classification (ISIC 3). It would have been preferable to have some industry breakdown on international production so that differences in productivity emanating from changes in the product mix could be captured. However, apart from allowing for possible productivity differentials arising from the dependence on electronics, a detailed commodity-level analysis is not possible because data for many countries, in particular for developing ones, are too heavily suppressed at the individual industry level. In any case, the aggregation bias arising from the use of data for total manufacturing is likely to be less important when one works with data for investment flows from a single source country. Furthermore, the use of productivity growth (rather than the level of productivity) as the dependent variable will minimize the potential error arising from industry-specific fixed

Trade orientation and productivity gains

155

effects; differencing over time ‘washes out’ time-invariant components of productivity. The data set is cross sectional, with each country representing a single data point. All variables (except the initial level of productivity) are measured as period averages over 1982–92. A well-known limitation of the use of inter-country cross-sectional data in econometric estimation is that they make it difficult to control for unobserved country-specific differences. Long-term averages also tend to ignore changes that have occurred over time in the same country. These limitations can be avoided by using a panel data set compiled by pooling cross-country and time-series data. Unfortunately, this preferred data choice is not possible in this case, given the nature of data availability. Complete data are available for the series needed for productivity measurement, namely the capital stock (net property, plant and equipment), output (net of inputs) and employment (number of workers) for all the 44 countries. For other variables, data for some years have been suppressed for some countries for confidentiality reasons, and the use of averages of available data is the only available option. (There are 16 countries for which at least one data point is missing on any one of these remaining variables.) Fortunately, the data suppression does not pose a severe problem for using period averages, as the incidence of suppression has occurred fairly randomly, and for all variables data for at least four intermittent years are available to compute a meaningful period average. The other alternative would have been to construct a pooled time-series and cross-country data set for the 28 countries for which data are available on all variables covering the entire period. This is, however, an unsatisfactory compromise because the truncated sample does not give adequate coverage to developing countries and within that country group to countries representing different policy regimes. The hypothesis at hand relates to economic behaviour in the medium to long run. Thus, using a data set based on ten-year period averages is a reasonable compromise. As noted, the data pertain to overseas production of majority-owned affiliates only. This does not, however, hinder the representativeness of our sample, given the general preference of United States MNEs to hold majority or full ownership in their affiliates abroad. Over 85 per cent of total sales of United States MNEs worldwide (over 80 per cent when the focus is limited to developing countries) originate in majority-owned affiliates. At the individual country level, majority-owned affiliates account for less than 50 per cent of total sales only in two countries of our sample – the Republic of Korea (40 per cent) and India (35 per cent) – which had implemented relatively stringent ownership limits on foreign companies during the time period examined.19

156

Multinational enterprises in Asian development

PRODUCTIVITY PATTERNS As a prelude to the presentation and discussion of econometric results, estimates of productivity are reported in Table 7.1. Industrial countries are generally characterized by higher productivity growth (with lower intercountry dispersion) than that of developing countries. Among developing countries, the high-performing Asian economies (HPAEs) on average show productivity growth comparable to that of industrialized countries. However, there are vast differences among countries within the HPAE group. Note in particular the very low rate for Malaysia and Singapore despite their high export orientation. A possible reason is the heavy concentration of international production in assembly activities in the electronics industry. The average share of electronics in total sales of United States MNEs in Malaysia is as high as 80 per cent, the highest rate of electronics dependence for any single country. Measured total factor productivity growth across the sample is generally lower (mean4.14 per cent) and has a greater dispersion (CV1.34) than labour productivity growth (mean 5.79 per cent, CV1.14). However, the two measures are highly correlated (r0.90). Now we turn to regression analysis, which deals with that relationship in greater detail.

DETERMINANTS OF PRODUCTIVITY The regression results for total factor productivity growth and labour productivity growth, with trade orientation measured in terms of the export–output ratio, are reported in Table 7.2. Both equations pass the F test for overall statistical significance at the 1 per cent level, and perform well by the standard diagnostic tests relevant for cross-sectional regression analysis of this nature. In particular, they pass the tests for functional form specification, heteroscedasticity and normality at the 5 per cent level or better. The hypothesis that the contemporaneous independent variables are exogenous is not rejected in terms of the Wu–Hausman test for endogeneity. Moreover, pre-testing of the data set using the procedure suggested by Cook (1977) for detecting influential observations indicated no evidence of the presence of outlier observations, a result which is consistent with the results from the normality and functional form tests. Regression estimates are remarkably resilient to the choice between total productivity growth and labour productivity growth as the measure of the independent variable. It seems that factor proportions in international operations of foreign affiliates tend to remain largely unchanged both

157

Trade orientation and productivity gains

Table 7.1 Overseas operations of US manufacturing MNEs: estimates of productivity and related indicators Initial (1983) productivity levels (Canada 100)

Industrial countries 1 Australia 2 Austria 3 Belgium 4 Canada 5 Denmark 6 France 7 Germany 8 Greece 9 Ireland 10 Israel 11 Italy 12 Japan 13 Luxembourg 14 Netherlands 15 New Zealand 16 Norway 17 Portugal 18 Spain 19 Sweden 20 Switzerland 21 United Kingdom High-performing Asian economies 22 Hong Kong (China) 23 Indonesia 24 Republic of Korea 25 Malaysia 26 Singapore 27 Taiwan 28 Thailand Other developing countries 29 Argentina 30 Brazil 31 Chile 32 Colombia

Productivity growth (per cent) TFPG

Export orientation (per cent)

Tax rate (per cent)

Electronics in total sales (per cent)

XOR

TAX

ELSH

43.39 17.40 41.50 68.80 39.36 50.24 36.27 42.93 20.53 87.75 56.08 27.77 17.06 94.04 67.39 5.29 37.68 44.35 32.31 38.38 54.14 32.03

33.25 37.28 29.92 27.12 37.76 41.32 37.99 41.32 47.06 5.72 14.18 34.77 54.09 22.05 24.29 76.92 26.09 29.63 28.23 16.22 37.33 28.97

10.30 4.80 8.39 6.69 12.66 17.12 5.71 6.84 5.90 7.93 54.83 7.76 11.81 6.32 3.75 5.22 4.61 22.17 5.52 2.62 10.07 5.51

TFP83

LP83

LPG

81.00 77.84 58.27 86.02 100.00 89.87 87.58 98.01 66.20 122.66 61.22 88.75 94.52 69.41 65.33 65.83 102.72 35.69 43.76 98.41 118.94 69.89

76.21 77.67 60.38 76.67 100.00 75.70 77.38 90.82 38.44 150.94 63.74 74.11 103.33 70.78 59.70 51.60 95.03 26.05 40.24 91.45 113.64 62.79

8.47 5.63 5.69 3.60 2.01 0.77 9.99 6.21 2.72 1.29 8.55 5.02 8.42 5.80 8.57 6.38 9.59 4.62 8.44 6.36 12.09 10.17 10.87 6.96 7.80 6.05 17.37 11.99 11.18 6.94 1.59 1.94 5.81 4.46 11.80 7.40 16.34 13.11 6.28 5.40 8.05 2.06 7.69 5.50

44.11

28.57

7.40

4.50

57.24

21.15

44.11

54.80

26.09

11.56

6.33

67.29

13.30

39.54

41.12 38.81

25.43 25.82

1.52 9.27

1.32 3.48

9.86 43.11

38.80 48.24

15.81 49.11

34.88 72.82 37.66 28.65

20.35 59.52 23.81 19.01

4.40 3.11 17.62 4.30

3.01 2.34 11.48 3.53

79.30 87.10 49.40 64.62

10.01 5.10 16.38 16.24

79.23 44.64 41.44 38.93

44.89

38.25

1.59

2.04

18.06

36.41

4.79

45.72 35.60 37.39 60.91

42.36 33.55 34.45 52.51

2.66 7.93 11.79 1.92

2.59 7.19 4.00 3.30

20.91 16.96 33.99 5.15

54.52 50.10 12.98 33.05

2.75 8.77 3.02 1.85

158

Table 7.1

Multinational enterprises in Asian development

(continued) Initial (1983) productivity levels (Canada 100) TFP83

33 34 35 36 37 38 39 40 41 42 43

Ecuador Egypt India Jamaica Mexico Nigeria Panama Peru Philippines South Africa Trinidad & Tobago 44 Venezuela

Summary statistics Mean CV**

LP83

Productivity growth (per cent) LPG

TFPG

Export orientation (per cent)

Tax rate (per cent)

Electronics in total sales (per cent)

XOR

TAX

ELSH

12.28 6.20 4.73 42.34 29.88 2.32 39.70 3.70 35.68 5.08 27.27

88.00 55.56 46.71 19.66 36.99 20.93 11.91 21.81 31.91 45.37 19.66

10.41 0.00 0.19 0.00 8.68 3.45 0.00 3.31 27.13 1.77 0.00

46.94 26.98 20.43 31.34 13.36 55.23 98.24 43.19 12.84 50.41 87.22

34.01 12.04 14.26 25.54 12.24 12.59 12.77 1.22 2.59 31.65 5.27 11.77 12.24 10.89 12.93 49.79 5.51 0.10 71.92 1.35 2.42 33.92 2.36 5.18 9.21 11.94 13.15 39.37 8.42 8.11 81.14 5.71 4.87

52.40

47.59

0.07

1.63

2.71

33.33

5.33

54.83 56.78

62.00 45.69

4.14 1.34

5.79 1.14

36.38 68.57

32.47 53.83

13.67 1.28

Notes: * Except for TFP83 and LP83, measures are period averages (1983–92). ** Coefficient of variation. Source: Compiled from computer files of US Direct Investment Abroad, Bureau of Economic Analysis, US Department of Commerce.

across countries and over time. The choice between these two measures is not important in determining the nature of the results.20 The coefficient of XOR is significantly different from zero in both equations with the theoretically expected (positive) sign. Thus there is strong statistical support for the hypothesis that countries with trade regimes characterized by greater outward orientation tend to experience higher productivity growth in national operations of MNEs. Through various experimental runs, we found that the result for XOR is robust to the presence (or deletion) of the three ‘conditional’ variables (TAX, ICD and ELSH)21 in the regression specification. A priori the direction of causation between productivity growth and export orientation can run either way. However, as already noted, in this particular exercise the standard Wu–Hausman test failed to find any statistical evidence that the regression estimates are affected by such endogeneity.

159

Trade orientation and productivity gains

Table 7.2 Determinants of productivity growth: regression results with trade orientation measured by export–sales ratio (XOR) (N 44) Equation 1 TFPG CON XOR TAX TFP83

0.006 (0.159) 0.094 (2.317)** 0.124 (2.572)*** 0.093 (3.1090)***

LP83 SCL ELSH DC*ELSH Adj. R2 F-statistic W–Ha (F) b 2 NORM (X ) HET c (F) RESET d (F )

0.015 (3.948)*** 0.001 (0.898) 0.002 (2.155)** 0.427 6.362*** 0.092## 1.640## 0.462## 5.325#

Equation 2 LPG 0.046 (1.250) 0.163 (3.608)*** 0.077 (1.465)*

0.101 (3.145)*** 0.018 (4.458)*** 0.001 (1.367) 0.002 (2.644)*** 0.493 7.971*** 0.263## 0.214## 0.125## 2.252#

Notes: t-ratios of regression coefficients are given in brackets with statistical significance denoted as: *** 1 per cent, ** 5 per cent and * 10 per cent. # Null-hypothesis is not rejected at the 5 per cent level. ## Null-hypothesis is not rejected at the 1 per cent level. a. Wu–Hausman test for endogeneity of regressors; based on the F-distribution. b. Jarque–Bera test for normality of the error term; distributed as 2. c. White test for heteroscedasticity; based on the F-distribution. d. Ramsey test for functional form mis-specification; based on the F-distribution.

There is evidence of a significant negative effect of the degree of tax incidence on productivity gains. This result points to the importance of taking into account the nature of the domestic tax regime, in addition to the nature of the trade regime, in examining the effects of host-country policies on the operation on foreign affiliates. Relating to this point, it is important to note that the correlation between TAX and XOR (and the other two policy

160

Multinational enterprises in Asian development

Table 7.3

Correlation matrix

TFPG LPG TFP83 LP83 TAX

SCL

XOR SW1 BMP ELSH

TFPG 1.00 LPG 0.90 1.00 TFP83 0.04 0.04 1.00 LP83 0.00 0.04 0.95 1.00 TAX 0.41 0.35 0.13 0.14 1.00 SCL 0.33 0.37 0.23 0.25 0.17 1.00 XOR 0.40 0.49 0.27 0.28 0.60 0.05 1.00 SWI 0.35 0.56 0.42 0.36 0.28 0.32 0.58 1.00 BMP 0.31 0.50 0.41 0.33 0.24 0.40 0.42 0.74 1.00 ELSH 0.17 0.25 0.24 0.29 0.31 0.07 0.51 0.32 0.20

1.00

regime variables) is far from perfect (Table 7.3), suggesting that higher trade orientation does not necessarily imply the presence of a tax regime that is conducive for international production. The coefficient of the initial level of productivity is negative and statistically significant, suggesting the presence of productivity convergence among MNE operations. The estimated coefficient of this variable suggests that a country whose TFP level was one percentage point higher than that of the sample, on average, experiences 0.06 percentage point slower TFP growth in subsequent years. The estimated convergence coefficient here is fairly large in comparison with the economy-wide estimates from various cross-country growth regressions.22 This result is plausible given the fact that foreign affiliates from a given home country operating in different countries have easy assess to a common pool of technology, managerial and marketing practices, which accelerate the convergence process. The coefficient of ELSH has the expected (negative) sign, but is not statistically significant. However, when ELSH is used interactively with a dummy variable for developing countries (that is DC*ELSH, where DC takes value of 1 for developing countries and 0 otherwise), the regression coefficient attains statistical significance at the 5 per cent level. This result is consistent with the fact that simple (labour-intensive) assembly operations in the electrical and electronic equipment industry are predominantly located in developing countries. For industrial countries this broad industry category encompasses mostly capital- and technology-intensive product lines, which are generally subject to rapid productivity improvement.23 The result for DC*ELSH also helps explain the relatively low rates of productivity growth recorded by electronics assembly centres in Asia, in particular Malaysia and Singapore, despite their outward trade orientation and conducive tax regimes.

Trade orientation and productivity gains

161

Note that the two country dummy variables (DIC and DPHAE) do not appear in the regression estimates reported in Table 7.3. Despite their strong positive bivariate correlation with productivity growth (both LPG and TFPG), these variables turned out to be statistically insignificant over and above the other variables in explaining inter-country differences in productivity performance. The strong bivariate correlation seems to reflect spuriously the differences between three groups of countries (industrial countries, HPAES and other developing countries) in terms of the other relevant influences that are appropriately captured in the regression specification. Alternative regression estimates based on the Sachs–Warner binary index of openness (SWI) and the black market premium (BMP) as alternative indicators of trade orientation are reported in Table 7.4. The results based on BMP are generally statistically more satisfactory (both in terms of the overall fit of the equation and the statistical significance of individual coefficients) than those based on SWI. This is understandable because a continuous variable naturally yields greater variability in the measurement than a binary variable. This difference notwithstanding, the results based on both openness indicators are largely consistent with those based on XOR. Without further discussion on individual regression coefficients, we can therefore safely conclude that our findings on the implications of trade orientation for host-country productivity gains from participation in international production (as well as the other related results) are remarkably robust to the particular measure of trade orientation used.

CONCLUSION AND POLICY INFERENCES In this chapter we have examined the impact of the nature of trade orientation of host countries on the productivity gains from international production through a study of overseas operations of manufacturing affiliates of United States MNEs. The analysis is based on a cross section of data for 44 host countries (21 industrial and 23 developing countries) averaged over the 1983–92 period. Total factor productivity and labour productivity were used as alternative measures of productivity growth. Particular emphasis was placed on the robustness of results to alternative measures of trade orientation. Even allowing for margins of error that are inherent in any empirical work of this nature, our results lend support for the hypothesis that outward orientation of the trade regime, ceteris paribus, enhances productivity gains from international production. Furthermore, there is evidence that countries with higher tax rates tend to exhibit lower productivity growth. The results are remarkably robust to the use of labour productivity or total

162

Multinational enterprises in Asian development

Table 7.4 Determinants of productivity growth: regression results with trade orientation measured by Sachs–Warner index (SWI) and black market premium (BMP) (N  44) Equation 1 TFPG 0 SWI

0.039 (1.145) 0.029 (1.435)*

BMP TAX TFP83

0.167 (3.747)*** 0.083 (2.665)***

Equation 2 TFPG 0.063 (1.478)*

0.062 (1.773)** 0.169 (3.767)*** 0.079 (2.534)**

LPD83 SCL ELSH DC*ELSH Adj. R2 F-statistic W–Ha (F) NORMb (X 2) HETc (F) RESET d (F)

(2.476)** 0.013 (3.036)*** 0.001 (0.991) 0.002 (1.773)* 0.379 5.384*** 0.328## 0.425## 1.236## 4.649#

0.012 (2.908)*** 0.001 (1.531)* 0.002 (2.106)** 0.368 5.181 0.173## 1.723## 0.730## 3.289#

Equation 3 LPG 0.084 (0.235) 0.063 (2.983)***

Equation 4 LPG 0.062 (1.422)

0.137 (2.784)***

0.138 (2.465)** 0.141 (2.776)***

0.078 (2.476)** 0.013 (2.855)*** 0.001 (1.215) 0.002 (1.782)* 0.447 6.805*** 0.263## 1.505## 2.094## 4.011#

0.071 (2.211)** 0.013 (2.628)*** 0.002 (2.166)** 0.002 (2.370)** 0.411 6.008 0.607## 1.345## 1.324## 2.054##

Notes: t-ratios of regression coefficients are given in brackets with statistical significance denoted as: *** 1 per cent, ** 5 per cent and * 10 per cent. # Null-hypothesis is not rejected at the 5 per cent level. ## Null-hypothesis is not rejected at the 1 per cent level. a. Wu–Hausman test for endogeneity of regressors; based on F-distribution. b. Jarque–Bera test for normality of the error term; distributed as 2. c. White test for heteroscedasticity; based on the F-distribution. d. Ramsey test for functional form mis-specification; based on the F-distribution.

factor productivity as a measure of productivity growth and to the use of alternative indicators of trade orientation. We also find that the relative importance of assembly activities in total production of foreign affiliates tends to correlate negatively with inter-country

Trade orientation and productivity gains

163

differences in productivity growth in their international operations. Thus the heavy reliance of domestic manufacturing on assembly activities in the electronics industry seems to provide some explanation for the much publicized low productivity growth syndrome of the East Asian miracle economies, in particular Singapore and Malaysia. This inference by no means implies that there are no national gains from this form of international production. It brings about other well-known positive gains, particularly in the forms of employment generation and foreign exchange earning. There are two aspects of the influence of MNEs on manufacturing productivity in a given host country; the direct effects from their production activities, which are essentially a part of total production in the country, and the indirect (or spillover) effects on the operation of local firms. In this study we have focused on the former aspect only. Indirect effects are not explicitly considered in this study, but it is reasonable to assume that higher direct productivity effects usually lead to the enhancement of positive spillover effects (through labour mobility and demonstration effects on management practices). Should this be the case, the implications of the domestic policy regime for productivity gains from MNE participation in domestic manufacturing would be much greater than is suggested by the econometric evidence reported here. The key policy implication arising from the findings is that, to maximize productivity gains from international production, trade and FDI liberalization should go hand in hand. Many developing countries still resort to trade restrictions with a view to enticing MNEs to relocate production in ‘priority’ sectors. Also, many countries that embark on significant liberalization reforms tend to retain and even increase special protection to industries dominated by MNEs. The findings suggest that such inconsistencies and contradictions in liberalization reform packages can thwart anticipated gains from reforms, and perhaps even generate immiserizing growth. They also lend support to the view that, when economy-wide restrictive trade practices are difficult to dismantle (because of domestic political and/or ideological resistance), allowing MNEs to operate in export processing zones is a rational second-best policy choice. As regards the implications for the economic analysis of international production, the findings cast doubts on the prevalent practice of using the volume of FDI inflow as a general measure of the transmission of disembodied technology. Our findings make a strong case for distinguishing between FDI inflows induced by domestic protection and those undertaken predominantly on export profitability considerations in analysing the economic impact of international production on the host country.

164

Multinational enterprises in Asian development

NOTES * 1.

2.

3.

4.

5. 6.

7.

8. 9.

10. 11. 12. 13.

This chapter is based on Athukorala and Chand (2000). Bhagwati (1978) also postulates that the volume of FDI inflows to a given country should, other things being equal, be larger under an export-promoting strategy because the size of the domestic market is not a limiting factor. Here we are not concerned with this volume effect of the trade regime on FDI. For empirical analyses of this aspect, see Balasubramanyam and Salisu (1991) and Hufbauer et al. (1994). According to estimates done as part of the Harvard Multinational Enterprise Project in the early 1970s, for every dollar of capital transferred by the parent companies to their foreign affiliates in developing countries, about four dollars more of capital were collected by the affiliates from other sources, including sources internal to the developing countries (Vernon 1971). Given the rapid globalization of company financing over the past two decades, perhaps this ‘non-FDI flow’ component has also increased significantly since then. Despite rapid growth of FDI outflows from other countries in recent years, the United States is still the principal source country for FDI, accounting for over a quarter of the world stock of over $2.4 trillion. The size of the United States stock is more than the combined total of the second and third largest investors (the United Kingdom and Japan respectively) (UNCTAD 1995). Such productivity gains are distributed between the home (investor) country via profit remittances and the host country via increased production, tax receipts and perhaps future productivity-related wage increases. Productivity growth of international production may also raise the quality of technology spillover from foreign affiliates to domestic firms in the host country (Blomström and Kokko 1998). The term ‘technology’ as used here encompasses production technique, as well as marketing and managerial know-how embodied in FDI. The vast empirical literature on the determinants of factor productivity spawned by the endogenous growth theory has made a clear case for distinguishing between the level effect and growth effect on the productivity of the explanatory variables under investigation (Barro and Sala-I-Martin 1995). In this article, we focus solely on the growth effect because it is the most (if not the only) relevant aspect of productivity performance in relation to the issue at hand. The objective here is to examine the impact of trade orientation on inter-country differences in productivity performance of foreign affiliates. Assuming the production function in Equation (7.1) is linearly homogeneous in K and L and letting small letters denote per-worker quantities, labour productivity growth (lpg) can be written as total factor productivity growth (tfpg) plus growth in capital intensity (kg) weighted by the share of capital in output (sk), that is lpgtfpg
sk kg. See Edwards (1998) and the works cited therein. As it has been correctly pointed out, foreign affiliates operating in export processing zones (EPZs) can be highly export oriented even under an otherwise very restrictive trade regime. However, we do not consider that this well-known feature of export-oriented FDI in developing countries limits the usefulness of XOR as a measure of trade orientation of international production. Setting up EPZs is an effective (though obviously a ‘second best’) means of trade liberalization when it comes to opening up domestic manufacturing to international production. Note that these two measures are economy-wide indicators of openness, whereas XOR is specific to international production taking place in the country in question. The black market premium may be endogenous to the policy regime and simply pick up the tightness of the black market. Moreover, it can be affected by interest rates and penalties for dealing in the black market (Srinivasan, 1995). Australia, New Zealand and Sweden are examples. It has been suggested that a more general index of the climate of business activity should be used as a superior alternative to TAX. Unfortunately, a significant number of countries in our sample are not covered by the available general indices. For instance, the

Trade orientation and productivity gains

14. 15.

16. 17.

18.

19. 20.

21.

22. 23.

165

Index of Economic Freedom (IEF) developed by the Heritage Foundation (which is perhaps the most comprehensive in terms of country coverage) covers only 25 of the 37 countries included in our analysis. Interestingly, for these 25 countries IEF (for 1997) and TAX are highly positively correlated (with a rank correlation coefficient of
0.91). See Dollar and Wolf (1993); Barro and Sala-I-Martin (1995) and the works cited therein. In this dummy-variable treatment of the level of development of sample countries, all developing countries other than those belonging to the HPAE category are the ‘base’ dummy. A further disaggregation of countries is not possible given the nature of the country sample. Benchmark Survey for 1989 and annual surveys for other years. It would be preferable to use gross output, together with material as a third input, in the TFP measure. Also, the ideal labour input measure is hours worked, preferably disaggregated into skill categories. However, the data required for these preferable variable choices are not available. Admittedly, this deflation procedure is somewhat crude – it is based on the assumption that MERs tend to maintain purchasing power parity between the currencies of the host countries and the dollar. Ideally, one would translate current-dollar data for each country back into local currency, and then deflate the resulting local currency estimates by appropriate country-specific price indices (Mataloni 1997). Unfortunately, the data needed for this preferred procedure were not available for most developing countries in our sample. To see the sensitivity of our results to the choice between the two deflation procedures, we estimated total factor productivity growth (TFPG) for 18 of our countries (industrial countries listed in Table 7.1 except Israel, Portugal and Switzerland) using the real output series derived by Mataloni (1997, Table 2) through the preferred procedure. A comparison of these estimates with those used in this study failed to show significant rank reversal among these 18 countries (rank r 0.92). It is therefore unlikely that the use of the crude deflation procedure would have influenced significantly the results of our empirical analysis. Figures reported in this paragraph are based on United States Bureau of Economic Analysis (1992). Assuming the production function in Equation (7.1) is linearly homogeneous in K and L and letting small letters denote per-worker quantities, labour productivity growth (lpg) can be written as total factor productivity growth (tfpg) plus growth in capital intensity (kg) weighted by the share of capital in output (sk), that is lpgtfpg
sk kg. These are the variables that we have identified as ‘potentially’ important in our specification, hence the terms ‘conditional’ variables. The other two variables – SCL and TFPg (or LPRODg) – are variables that are usually included in this type of analysis. The discussion here is based on Levine and Renelt (1992). See Edwards (1998) and Felipe (1999) for comprehensive surveys of this literature. Ideally ELSH should have covered only assembly activities in the electrical and electronics industries, but disaggregated data are not available. Given the nature of data availability, the use of a dummy interaction variable (DC*ELSH) is obviously a ‘second-best’ approach to the issue at hand.

166

Multinational enterprises in Asian development

APPENDIX 7.1 MEASUREMENT OF TOTAL FACTOR PRODUCTIVITY GROWTH The method used here for the measurement of total factor productivity growth is the Tornquist–Theil formulation. Under this formulation, which perhaps is the most intuitive formulation of total factor productivity, the level of total factor productivity is defined as: TFP 

Y

(7.A1)

,

兺aiXi

where Y is output measured by value added, X is the vector of inputs and ai is the vector of input weights. The choice of input weights (a in Equation (7.A1)) requires assumptions about the underlying technology. These parameters can either be estimated econometrically or deduced from the data using index number theory. The econometric technique requires sufficient degrees of freedom, a luxury in most cases. Here, the index number approach is used, owing to the limited number of observations. The input index is defined as: ln xkj  21

n

兺 (sij
sik)(ln xij  ln xik)

(7.A2)

i1

where sij and xij are the expenditure share and quantity, respectively, of input i at observation j while n is the number of inputs. The subscript k denotes the point of reference, this being the binary comparison of point i with that of point k. The choice of this reference point and the subsequent normalization is non-trivial; here the reference point chosen is the hypothetical firm whose input expenditure shares are the arithmetic mean of all the cross-sectional units and whose input quantities are the geometric mean of the respective quantities for the entire set of observations. Hence the input share of this hypothetical firm is given as: lnx*  21

n

兺 (sfi
s) (lnxfi  i1

lnxi n ).

(7.A3)

A major advantage of using the Tornquist–Theil index in this exercise over other measures is that it does not require any estimation of the parameters in the production technology. These parameters are treated as subsumed in the expenditure and revenue data, this assumption being legitimate if firms pursue profit maximization and/or cost minimization.1 Another advantage of the index in its application in cross-sectional work is

Trade orientation and productivity gains

167

its transitivity; that is, it does not suffer from the problem of being sample dependent (Caves et al. 1982). Given the nature of data availability, we consider two inputs, capital and labour. Capital is measured as the end-of-period book real value of fixed capital. Labour is measured by the number of production workers employed. The more appropriate measure of labour input is, of course, hours worked, but data on this variable are not available. Data are also not available for adjusting the capital stock for capacity utilization or to introduce human capital as an additional variable. Given these limitations on data, our measure of total factor productivity growth is admittedly crude (Nelson and Pack 1999, Felipe 1999). However, we believe that biases resulting from these problems are likely to be less severe in the measurement of factor productivity growth of international production by MNEs from a single source country (the United States) than in a general application of the same procedure in economywide measurement. For instance, failure to identify human capital as a separate factor of production may be less problematic for international production, given the fact that MNEs generally draw upon a common pool of technology and expertise. Moreover, significant inter-country variation in the degree of capacity utilization is unlikely to be a significant factor in international production.

NOTE 1. Good et al. (1996) provide an excellent survey of this literature.

8. Multinational enterprises and the globalization of R&D: a study of US-based firms* Multinational enterprises (MNEs) play a pivotal role in the generation of technology and its transmission across countries. The potential contribution of MNE affiliates to the indigenous innovatory capability of the countries in which they operate (the host countries) is therefore central to the contemporary policy debate on the developmental impact of foreign direct investment (FDI). There are two methods by which an MNE affiliate provides technology to host countries – importing technology produced elsewhere within the global operational network of the MNE (technology transmission) and developing new technology locally through R&D (technology generation). The host-country governments generally attach much greater importance to technology generation than to technology transmission, in the hope that R&D activities undertaken within the national boundaries will have important externalities that lay the foundations for national scientific and technology activity. This expectation is reflected in strong competition among countries to attract R&D-intensive FDI through investment promotion campaigns and by offering generous R&Drelated tax concessions and high-quality infrastructure at subsidized prices. In spite of this policy emphasis, there are no systematic, up-to-date empirical analyses of the determinants of international location of R&D activity by MNEs and the role of government policy in influencing the process to their national advantage. The few available empirical studies on this subject are not only much dated, but also, based as they were on data for a single or a few intermittent years, have failed to account for the inherent dynamics of the phenomenon under study.1 This chapter aims to fill this gap by examining patterns and determinants of the international location of R&D activity by foreign affiliates of US-based MNEs using a rich new panel data set for the period 1990–2001. To the best of our knowledge, ours is the first analysis of the patterns and determinants of R&D activity of US-based MNEs using data spanning the entire decade of the 1990s, a period characterized by significant changes in international production. In comparison with previous studies, we examine inter-country variation 168

Globalization of R&D

169

in R&D intensity of MNEs by taking into account a larger number of explanatory variables suggested by the theory of MNE behaviour. This chapter begins with a succinct review of the theory of overseas R&D activities of MNEs in order to set the stage for the ensuing empirical analysis. The next section examines trends and patterns of overseas dispersion of R&D expenditure of US MNEs. The following sections deal with model specification and data for the regression analysis of the determinants of inter-country differences in R&D propensity. The results are then presented and interpreted in the context of the existing literature and this is followed by a summary of the key inferences.

THEORETICAL FRAMEWORK The R&D location decision of the MNE is governed both by considerations which compel it to keep R&D as a headquarter function (centripetal factors) and by those which tend to pull it away from the centre and into peripheral locations (centrifugal factors) (Caves 1996, p. 117). The centripetal factors are of two major forms. First, technology – the assets created by the innovatory process – is an important part of the ‘knowledge capital’ of the MNE which determines its market power or ‘ownership advantage’ in international operation. There is always the possibility that geographical decentralization of R&D will lead to leakage of proprietary technology to foreign competitors, attenuating the MNE’s market power. Such leakage can happen either through defection of R&D personnel to competitors or to start up their own ventures or simply through the ‘demonstration’ effect. Thus the desire to maintain strategic knowledge within the firm is a compelling reason for keeping R&D as a headquarter function. Second, production of technology is an activity subject to firmlevel (rather than plant-level) scale economies. The innovatory process essentially involves communication and cooperation with product design, marketing and other related key functions. There is also a need for better motivation of R&D efforts towards objectives set by the top management. For these reasons, dispersion of resources for executing parallel projects at plant level could be wasteful and reduce productivity of the overall R&D effort of the MNE (Barba Navaretti and Venables 2004, pp. 25–6). The above factors are generally expected to have a dominant impact on the MNE’s decision to keep R&D fundamentally a headquarter function. However, two ‘centrifugal’ forces necessitate some dispersion of R&D activities among various production locations. Firstly, there may be a need to adapt production processes and characteristics of products to local conditions and regulations. This consideration is particularly relevant when

170

Multinational enterprises in Asian development

demand and/or production conditions in the host country differ significantly from the conditions in the home country, or when the geographical proximity of research facilities to manufacturing facilities in the host country reduce the time lag in adjusting production techniques or product characteristics to host-country conditions. While improved communications mitigate some of the difficulties created by distance, it is presumably an imperfect substitute for the physical proximity needed for effective communication between R&D and other functional areas, notably marketing and production. Secondly, MNEs may have to undertake R&D in overseas locations in order to source technology and to benefit from localized technology spillovers in these locations, with a view to maintaining their competitive edge. Locating R&D facilities in prominent centres of excellence in specific technologies across the world would enable MNEs to enrich their own R&D. There is indeed evidence that independent R&D is the most effective way of ‘learning’ about other firms’ products and processes near the sources of the spillover, as compared with licensing, patent disclosures, the hiring of competitors’ R&D employees and reverse engineering (Levine et al. 1987). This is because knowledge spillover is positively related with proximity. R&D units set up in global innovatory centres could also serve as stations for recruiting local scientists and technicians and as points of contact with the scientific community in the host country (Serapio and Dalton 1999, Cohen and Levin 1989, OECD 1998). The early literature on R&D activities of MNEs generally considered product adaptation, which normally involves cross-border transfer of mature technologies, as the dominant motive for decentralization of R&D geographically (Vernon 1974; Caves 1996, Chapter 6; Dunning 1994; Lall 1979). Recent survey-based evidence, however, suggests that over the years the technology-seeking motive has become a significant contributing factor in decentralization of R&D by MNEs in R&D-intensive industries such as pharmaceuticals, consumer chemicals, professional and scientific equipment and office equipment (Ronstadt 1977, Pearce 1999, Fors and Svensson 1994, Birkinshaw and Morrison 1995, Vernon 2000). There are also numerous cases of acquisition of companies by MNEs outside their home base in the hope of unlocking some prized technological secrets for worldwide use. In sharp contrast to the role of a conventional R&D department that was primarily engaged in adapting established group products for the local market, the mission of the modern knowledge-seeking R&D labs is to draw upon geographically differentiated frontier technology in an attempt to preserve the technological lead of the MNE. These labs are engaged in original product development or providing inputs into programs of basic or applied research to support the longer-term evolution of

Globalization of R&D

171

the core technology of the MNE group at the world technology frontier. Thus the compositional difference between headquarter R&D operations and overseas R&D operations of MNEs is likely to have narrowed over time with the new emphasis on knowledge-seeking overseas R&D. Even if there are compelling reasons to decentralize R&D globally, the MNE’s decision to undertake R&D in a given host country also depends on the domestic business environment. The availability of and cost of hiring technical personnel, the nature of property rights legislation, tax concessions and other incentives for R&D activities, skilled labour, and the general business climate for foreign direct investment (including political stability and policy certainty, and the foreign trade regime) are among the relevant factors in making the R&D location decision. Assuming these prerequisites are met, the entry of MNEs into a given host country and the expansion of its R&D activities are likely to take place in a sequential manner. The process will begin with the establishment of production activities based entirely on technology provided by the parent company. Setting up of local R&D research support activities will take place only after the subsidiary gains experience in that particular location and if the future growth prospects are promising. The activities of the research departments may then grow, in terms of both the staff employed and the complexity of tasks, hand in hand with the expansion of the subsidiary’s business. This sequence suggests that, after some time, the R&D departments of some overseas affiliates may establish themselves as centres of technology ‘sourcing’ for other affiliates in the MNE’s global network (Lall 1979).

TRENDS AND PATTERNS OF R&D INTERNATIONALIZATION Data on R&D expenditure of US majority-owned multinational enterprises are set out in Table 8.1. The dollar value of overseas R&D activities of US MNEs increased rapidly from less than US$ 600 million in 1966 to around US$ 10 billion in 1990 and to over US$ 20 billion in 2001. Over the past decade, the share of overseas R&D expenditure in total corporate R&D expenditure (domestic
overseas) has varied in the narrow range of 11.4 per cent to 13.6 per cent. Overall, apart from some minor variations in either direction, overseas R&D expenditure has kept pace with domestic R&D expenditure. Thus, contrary to the inferences of some survey-based studies (for example, Pearce 1999, Cantwell and Piscitello 2002), there is no evidence of dramatic globalization of R&D activities in the 1990s, as far as the US-based MNEs are concerned. In spite of rapid globalization of

172

Multinational enterprises in Asian development

Table 8.1

R&D internationalization of US MNEs during 1966–2001 All sectors

Total

1966 1977 1982 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001

Manufacturing

Foreign affiliates

Total

Manufacturing share (%)

Foreign affiliates

Total Foreign affiliates

$ bn

$ bn

%

$ bn

$ bn

%

9.0 21.0 60.2 89.3 74.8 76.8 83.2 84.2 103.2 110.2 114.6 121.4 128.4 144.4 151.3 162.7

0.6 2.1 3.9 7.0 10.2 9.4 11.1 11.0 12.1 12.6 14.0 14.6 14.7 18.1 19.8 19.7

6.6 9.9 6.4 7.9 13.6 12.2 13.3 13 11.7 11.4 12.3 12 11.4 12.6 13.1 12.1

8.1 – – 78.9 64.4 67.0 73.4 74.2 90.6 97.2 102.2 107.3 113.6 121.2 125.0 132.5

0.5 – – 5.7 8.5 8.1 9.3 9.0 10.1 10.8 12.2 12.5 12.8 16.4 17.8 17.4

6.5 – – 7.2 13.1 12.1 12.7 12.2 11.2 11.1 11.9 11.7 11.3 13.5 14.3 13.1

90.5 – – 88.4 86.1 87.3 88.2 88.2 87.8 88.2 89.2 88.4 88.4 83.9 82.6 81.4

89.2 – – 81.1 83.1 86.1 84.3 82.4 83.9 85.8 86.9 85.7 87.4 90.3 90.2 88.2

Note: – data not available. Sources: Compiled from US Department of Commerce (1975, 1981, 1985, 1992) and computer files of U.S. Direct Investment Abroad, Bureau of Economic Analysis, US Department of Commerce.

MNE operations in the 1990s, the conventional wisdom about the dominant role played by centripetal factors in the MNE R&D decision (see above) still seems to hold. How does the degree of internationalization of R&D by US MNEs compare with that of MNEs from other countries? There are no data for a systematic comparison, but the available fragmentary data suggest that overseas R&D activities of MNEs based in other countries may have grown faster. For instance, the share of overseas R&D in total R&D expenditure of Swedish manufacturing MNEs increased from 9 per cent in 1970 to 13 per cent in 1978, and further to 24.7 per cent in 1994 (Fors 1998, p. 117). There are no complete records of overseas R&D activities of German MNEs, but there is survey-based evidence that the percentage of overseas employees in total R&D staff of German MNEs increased from 15 per cent in the late 1970s to over 18 per cent by the early 1990s (Golberman 1997,

Globalization of R&D

173

p. 141). Bloom and Griffith (2001, p. 350) report that in the 1990s British MNEs increased their R&D spending in their overseas research labs at a much faster pace than in labs in the UK; the overseas share of R&D expenditure of the British pharmaceutical industry increased from 48 per cent in 1994 to over 55 per cent in 1999. Internationalization of R&D by the Japanese MNE is a more recent phenomenon. However, the overseas share of total R&D of Japanese MNEs increased persistently from less than 1 per cent during 1989–90 to 2.3 per cent in 1996–97 (Kumar 2001, p. 161). Manufacturing accounts for the lion’s share (over four-fifths) of both total and overseas R&D expenditure of US MNEs (Table 8.1). However, over the past decade, the manufacturing share in overseas R&D has shown a mild but persistent increase (from 81 per cent in 1990 to over 90 per cent in 2001), in contrast to a persistent decline in this share in total overseas R&D expenditure (from 88 per cent to 83 per cent) during this period. Within manufacturing, chemical, electrical and electronic goods and motor vehicles account for over two-thirds of total overseas R&D expenditure (Table 8.2). There has been a noteworthy increase in the R&D expenditure share of electronics. Table 8.3 summarizes data on the inter-country distribution of overseas R&D expenditure in manufacturing. In order to place inter-country differences in R&D activities in the wider context of MNE operations, data on country shares of R&D expenditure and R&D intensity (R&D expenditure relative to total sales turnover) are brought together with data on the percentage distribution of the total capital stock and sales. The developed countries have remained by far the dominant location of R&D activities of US MNEs, accounting for nearly 90 per cent of total overseas R&D expenditure. However, there has been a mild, but persistent, decline in this share over time, from 94 per cent in the early 1990s to 87 per cent by the dawn of the new millennium. This decline has largely mirrored an increase in R&D shares of some high-performing East Asian economies, in particular Singapore, Korea, Malaysia and China. All Asian countries listed in the table, with the exception of Hong Kong and Indonesia, have recorded some increase in the share. In Latin America, all countries except the special case of Mexico have recorded a decline in their relative importance as locations of R&D activities for US MNEs. In sum, the decline in the developed-country share of overseas R&D expenditure is predominately a reflection of the growing importance of East Asian countries in global operations of US MNEs. Among developed countries, there has been a notable increase in the relative importance of the UK, Japan and Sweden. In the first half of the 1990s, Germany was by far the dominant location of R&D activities of US MNEs, accounting for over a quarter of the global total. However, by the

174 10 222

Total expenditure ($m)

19 201

100.0 1.7 89.6 1.7 22.7 0.8 3.9 26.4 28.0 6.0 3.1 0.0 5.2 0.3

1999–2001

9528

100.0 0.3 83.1 1.8 24.1 1.0 15.2 4.8 19.4 6.2 5.4 0.1 6.1 0.1

1990–92

14 971

100.0 5.8 87.6 1.5 23.2 1.0 4.5 20.7 29.0 7.7 2.5 0.0 4.0 0.1

1999–2001

Developed countries1

245

100.0 0.5 89.6 0.0 3.3 0.0 11.0 11.0 0.0 1.8 3.3 0.0 6.9 0.0

1990–92

1122

100.0 25.9 72.8 0.1 2.5 0.1 1.2 64.5 0.3 4.1 0.6 0.0 0.5 0.1

1999–2001

NICs2

295

100.0 1.5 95.5 6.5 29.0 1.1 5.4 5.5 4.0 13.8 4.2 0.3 0.7 1.2

1990–92

1096

100.0 7.9 88.4 4.3 15.5 0.4 3.8 8.7 15.0 40.6 1.9 0.0 1.4 0.4

1999–2001

Other developing countries3

Source:

Compiled from computer files of US Direct Investment Abroad, Bureau of Economic Analysis, US Department of Commerce.

Notes: 1. OECD Europe, North America, Japan, Australia and New Zealand. 2. Hong Kong, Korea Republic, Singapore and Taiwan. 3. Twenty-four countries for which data are available (Turkey, Argentina, Brazil, Chile, Colombia, Mexico, Panama, Ecuador, Venezuela, China, the Philippines, Indonesia, Malaysia, Thailand, India, Egypt, Nigeria, South Africa, Costa Rica, Honduras, Peru, Dominican Republic, Saudi Arabia and United Arab Emirates.

100.0 2.7 84.5 2.2 23.4 1.0 18.4 7.3 22.7 9.4 6.3 0.1 6.2 0.3

1990–92

All countries

Industry distribution of R&D expenditure of selected countries, 1990–2001 (%)

All industries Petroleum Manufacturing Food products Chemical products Primary and fabricated metals Industrial machinery and equipment Electronics Automotives Other manufacturing Wholesale trade Finance, insurance and real estate Services Other industries

Table 8.2

175

Developed countries Europe Austria Belgium Denmark Finland France Germany Greece Ireland Italy Netherlands Norway Portugal Spain Sweden Switzerland United Kingdom Canada Japan Australia New Zealand Israel

84.09 57.74 0.49 3.36 0.27 0.09 7.55 15.47 0.09 1.51 4.20 4.50 0.06 0.24 3.48 0.51 0.94 14.73 17.79 4.65 3.78 0.13 0.24

1990–92 73.19 48.14 0.66 2.72 0.35 0.22 5.23 9.20 0.09 3.03 3.10 3.45 0.23 0.24 2.48 1.22 0.60 14.63 17.11 4.32 3.08 0.54 0.69

1999–2001

FDI stock

84.95 61.43 0.46 3.26 0.26 0.08 7.88 16.31 0.11 1.94 4.96 5.05 0.08 0.30 3.82 0.70 0.81 15.26 16.04 4.40 2.93 0.16 0.15

1990–92 76.16 53.49 0.55 2.58 0.28 0.23 6.41 10.94 0.14 4.21 3.85 4.06 0.48 0.34 2.86 1.77 0.96 13.44 15.50 4.78 2.13 0.26 0.37

1999–2001

Sales

94.22 76.86 0.10 3.83 0.16 0.01 8.48 28.74 0.02 6.92 3.85 3.71 0.04 0.06 1.88 1.08 0.70 17.03 10.31 5.27 1.75 0.03 0.23

1990–92 87.23 66.63 0.49 1.79 0.35 0.42 8.22 18.91 0.04 1.90 2.90 2.13 0.09 0.09 1.03 6.10 0.87 19.36 10.85 8.03 1.66 0.05 1.92

1999–2001

R&D expenditure

1.61 1.82 0.30 1.71 0.91 0.21 1.56 2.56 0.30 5.18 1.13 1.07 0.73 0.30 0.72 2.25 1.26 1.62 0.93 1.74 0.87 0.32 2.26

1990–92

1.70 1.81 1.32 1.03 1.87 2.75 1.91 2.57 0.45 0.67 1.12 0.78 0.29 0.41 0.54 5.13 1.34 2.14 1.04 2.50 1.16 0.29 7.69

1999–2001

R&D–sales ratio

Table 8.3 Overseas affiliates of US manufacturing MNEs: FDI stock, sales, R&D expenditure and R&D–sales ratio by country/region (%)

176

Developing countries Asian NICs Hong Kong Korea, Republic of Singapore Taiwan Other Asia China Indonesia Malaysia Philippines Thailand India Latin America Argentina Brazil Chile Colombia Ecuador Mexico Panama

Table 8.3 (continued)

15.91 2.76 0.28 0.61 1.06 0.82 1.38 0.10 0.06 0.56 0.29 0.34 0.04 11.45 0.64 6.32 0.62 0.21 0.05 3.25 0.02

1990–92 26.81 5.37 0.33 1.16 3.12 0.75 6.74 2.61 0.20 1.35 0.67 1.34 0.57 14.07 1.53 4.85 0.19 0.27 0.05 6.42 0.05

1999–2001

FDI stock

15.05 3.88 0.76 0.35 1.89 0.87 1.68 0.06 0.08 0.67 0.36 0.46 0.05 9.08 0.62 3.86 0.15 0.30 0.04 3.58 0.05

1990–92 23.84 6.01 0.79 0.69 3.79 0.75 5.22 1.75 0.12 1.65 0.53 0.89 0.29 11.88 1.13 3.20 0.18 0.26 0.05 6.42 0.07

1999–2001

Sales

5.78 2.54 0.26 0.09 1.76 0.44 0.32 0.02 0.04 0.09 0.09 0.05 0.03 2.72 0.15 1.66 0.02 0.05 0.01 0.67 0.01

1990–92 12.77 5.63 0.22 0.73 3.90 0.78 3.97 2.40 0.01 1.13 0.24 0.07 0.12 3.01 0.19 1.44 0.01 0.04 0.00 1.19 0.00

1999–2001

R&D expenditure

0.56 0.95 0.49 0.35 1.35 0.74 0.27 0.41 0.80 0.20 0.34 0.15 1.09 0.43 0.36 0.62 0.18 0.22 0.21 0.27 0.18

1990–92

0.80 1.39 0.41 1.59 1.53 1.54 1.13 2.04 0.15 1.02 0.66 0.12 0.63 0.38 0.25 0.67 0.08 0.24 0.08 0.28 0.06

1999–2001

R&D–sales ratio

177

100 (46.8)

All countries (US$ billion)

100 (83.0)

0.06 0.65 0.31 0.32 100 (198.0)

0.04 0.44 0.25 0.16 100 (372.0)

0.06 0.50 0.45 0.28 100 (2.9)

0.01 0.15 0.14 0.06 100 (5.5)

0.00 0.12 0.11 0.05 1.45

0.35 0.48 0.79 0.53 1.49

0.08 0.36 0.37 0.26

Source: Compiled from US Department of Commerce (1975, 1981, 1992) and Bureau of Economic Analysis, US Department of Commerce, computer files of U.S. Direct Investment Abroad.

0.04 0.30 0.14 0.19

Peru Venezuela South Africa Turkey

178

Multinational enterprises in Asian development

end of the decade, the UK was at par with Germany, each accounting for about a fifth of the global total. In the early 1990s, Ireland (the ‘Celtic Tiger’) accounted for a sizeable share (7 per cent), reflecting perhaps the increased participation of US MNEs in the export-oriented FDI boom in the country at the time. However, the relative importance of Ireland as an R&D location declined in the ensuing years, bringing its share down to 2 per cent by the end of the decade. The R&D share of Canada has remained virtually unchanged at 10 per cent, reflecting perhaps the enduring importance of its proximity-related advantages. There is a clear mismatch between developed and developing countries in terms of the size of the R&D share compared with FDI stock and total global sales turnover. For instance, in 1999–2001, developed countries accounted for 87 per cent of total overseas R&D expenditure, compared with a share of 73 per cent in total FDI stock and 76 per cent in total sales turnover. By contrast, developing countries accounted for 26 per cent of FDI stock and 24 per cent of total sales turnover, but their share in total R&D expenditure stood at 13 per cent. Interestingly, in this comparison, the East Asian NICs occupy a middle position between developed countries and the other developing countries, with R&D shares comparable to FDI and sales shares. The average R&D–sales ratio for developed countries (1.70 per cent in 1999–2001) is more than double that of developing countries (0.8 per cent). Among developing countries, both NICs and other Asian countries show much greater R&D intensity (R&D–sales ratios of 1.4 per cent and 1.1 per cent respectively) than countries in Latin America (0.4 per cent). Among developed countries, MNE affiliates operating in Israel, Sweden, Finland, Japan and Germany (in that order) exhibit above-average R&D intensity compared with other countries. The exceptionally high figures for the small economies of Israel, Sweden and Finland seem to suggest the importance of these countries as innovatory centres, with a greater attraction for knowledge-seeking investment. Among the developing Asian countries, the R&D–sales ratio of MNE affiliates in China increased from a mere 0.4 per cent in the early 1990s to over 2.0 per cent in 1999–2001, a figure comparable to that of many developed countries. R&D intensity of MNE affiliates in Korea, Singapore and Taiwan has also increased over the years, approaching the average developed-country level. Malaysia and the Philippines have also recorded some notable increases, but they still lag behind the four NICs. Among the other developing countries, R&D–sales ratios of China, India and Brazil are notably high (notwithstanding some decline in the Indian ratio between 1990–92 and 1999–2001), perhaps because of the importance of productadaptation-type R&D activities in these large economies.

179

Globalization of R&D

Table 8.4 Percentage share of R&D expenditure of US MNE affiliates in total R&D expenditure in host countries (1990–2000 annual average) Country/country group

US affiliates’ share in total domestic R&D expenditure

All countries

3.3

Developed countries

3.4

Europe Austria Belgium Denmark France Germany Greece Ireland Italy Netherlands Norway Portugal Spain Sweden Switzerland United Kingdom Israel Canada Japan Australia New Zealand

4.9 1.2 7.5 1.1 3.2 5.5 0.6 43.8 3.0 4.6 0.3 3.0 4.5 3.6 1.5 8.7 2.4 12.1 0.6 4.1 1.4

Country/country group

Developing countries Asian NICs Hong Kong Korea, Republic of Singapore Taiwan Other Asia China Indonesia Malaysia Thailand India Philippines Latin America Argentina Brazil Chile Colombia Ecuador Egypt Mexico Panama Peru Venezuela South Africa

US affiliates’ share in total domestic R&D expenditure 1.7 0.3 0.1 0.3 11.4 1.4 2.8 16.2 3.9 21.9 2.5 0.3 11.9 4.1 2.3 3.9 0.8 2.4 3.1 0.7 10.6 2.0 34 6.0 1.9

Sources: Computed using data for Research and Development Expenditure from World Development Indicator(CD-ROM), World Bank, except for Taiwan. Data for Taiwan are from Taiwan Statistical Data Book 2001, Council for Economic Planning and Development, Taipei.

Table 8.4 depicts the relative importance of R&D expenditure of US MNE affiliates in total national R&D expenditure in host countries over the period 1990–2000. It is important to note that data on national R&D expenditure in these countries are fragmentary and not directly comparable to those of US MNEs, which are presumably collected and compiled with

180

Multinational enterprises in Asian development

greater care. Nevertheless, the general picture emerging from the table is clear; although the share of the total R&D expenditure of US MNEs is small, US MNE affiliates account for a significant share of total R&D activities in a number of host countries, among both developed and developing countries. The average share of US MNEs in total host developingcountry R&D expenditure for the period 1990–99 is 1.7 per cent, but this masks more than 10 per cent figures for Singapore, China, Malaysia, the Philippines and Mexico. Among the developed countries, individualcountry figures are relatively uniform, with the exception of high figures for Ireland, Canada and the UK. The developed-country average (3.4 per cent) is double that for developing countries.

DETERMINANTS OF R&D INTENSITY We have seen in the previous section that, while the degree of R&D intensity of MNE affiliates operating in developed countries is on average much higher than those operating in NICs and other developing countries, there are notable inter-country differences within each group. Interestingly, there is a considerable overlap between developed countries and NICs, with many developed countries recording R&D intensities comparable to or lower than those in NICs. We now turn to a more formal examination of what forces shape inter-country differences in R&D intensity. The analysis is based on a panel data set for 42 countries constructed at three-year frequency over the period 1990–2001. In this section, we first focus on model formulation and then give a brief discussion on the data before presenting the results. The Model The dependent variable of our analysis is R&D intensity defined as the ratio of R&D expenditure to total sales (RDS). The explanatory variables are specified in the context of the conceptual framework developed earlier in this chapter. They are discussed below under four main categories. Product adaptation We include three variables to capture the importance of adapting products and production processes to suit domestic market conditions in determining inter-country variation in R&D intensity. They are domestic market size measured by real gross domestic product (GDP), geographic distance measured by great-circle distance between Washington DC and the capital city of the given host country (DIST), and domestic market orientation of

Globalization of R&D

181

MNE affiliates (measured by the percentage of domestic sales in total sales turnover of affiliates) (DMS). A positive relationship is hypothesized between GDP and RDS simply because a large domestic market should provide incentives to perform R&D for adapting products and production processes to suit local demand patterns. DIST is a proxy for the ‘search problem’ that seems to induce MNEs to undertake product-adaptation-type R&D closer to its consumer base (Rangan and Lawrence 1999, p. 94). Here ‘search’ refers to acts performed in identifying potential exchange patterns and these acts become more important as economic opportunities become spatially dispersed. DIST may also capture the impact of market segregation associated with transport costs. Technological advances during the post-war era have certainly contributed to a ‘death of distance’ (à la Cairncross 1997) when it comes to international communication costs. However, there is evidence that the geographical ‘distance’ is still a key factor in determining differences in international transport costs, in particular shipping costs (Hummels 1999). For these reasons, we assume a positive relationship between DIST and RDS. At first blush, R&D activities of affiliates should depend positively on the extent to which the home market is served by their local production (Lall 1979, Hirschey and Caves 1981). However, in practice, when controlled for the market size, the impact of domestic market orientation on local R&D effort can go either way, depending on the differences in demand conditions between the host country and regional markets and the degree of market segmentation resulting from tariff and non-tariff barriers. If MNE affiliates located in a given country produce for wider regional or global markets in addition to serving the domestic market, a high degree of export orientation can in fact be positively associated with R&D intensity. In particular, this would be the case if the differences in technological levels between the subsidiary and its export market were less than the technological gap between the latter and the parent company. Domestic technological competency Domestic technological competency of the host country (henceforth referred to as the national ‘technology intensity’) is an important consideration for MNEs’ R&D location decisions. As already discussed, this is a particularly important consideration if technology seeking is a driving force behind overseas R&D activities. However, even in the case of domestic market adaptation type R&D, the domestic technology base is an important facilitating factor. We use a ‘technology effort index’ (henceforth denoted as TECH) developed by Lall (2002) to measure domestic technology intensity of host

182

Multinational enterprises in Asian development

countries. This is a composite index of two well-known R&D indicators, namely national ‘productive enterprise’ R&D expenditure and the number of patents registered by the country in the USA (both normalized by midyear population). ‘Productive enterprise’ R&D expenditure is total R&D expenditure net of R&D expenditure in agriculture, defence and various tertiary-sector activities. The latter components are deducted because they are not directly related to innovatory activities of private agents. The number of patents taken out in the USA is a good proxy for innovative activities of a country because practically all innovators who seek to exploit their technology internationally take out patents in the USA, given its market size and technology strength. The values for each variable is first standardized so that the highest country scores 1 and the lowest scores 0 and then the composite index is obtained as the average of the two (Lall 2002, pp. 8–9). Investment environment Three variables are used to capture various aspects of the economic environment of the host country, namely R&D personnel per million population (RDPN), the cost of hiring technical personnel (TPWG), tax intensives for firm-level R&D activities (TINS) and intellectual property right protection (IPR). RDPN is used to capture the ability of host countries to meet human capital requirements for undertaking R&D activities, which obviously contributes to the attractiveness of a given country as a location for R&D activities. Holding other relevant influences constant, TPWG is presumably a key determinant of the profitability of undertaking R&D locally compared with importing technological know-how from the parent company or other overseas affiliates. Tax incentives for R&D activities clearly have the potential to affect the propensity to undertake R&D, since higher tax rates depress after-tax returns, thereby reducing incentives to commit investment funds. Higher domestic corporate tax rates make importing technology a more attractive option than domestic technology generation because taxes on royalty payments for imported technology are tax deductible in the host country (Hines 1995). Intellectual property right protection (IPR) is widely considered an important policy tool for promoting innovative activities in countries with appropriate complementary endowments and policies. Private innovators will not fully exploit their capabilities, even when the other preconditions are met, unless they can appropriate returns to their innovations (Maskus 1998 and 2000). Other variables As discussed above, R&D intensity in a given country is potentially influenced by the nature of the industry mix, because the production pro-

Globalization of R&D

183

cesses of some industries are more R&D intensive than those of the others. Moreover, the need for adaptation of products to suit local market conditions varies from industry to industry. For instance, most product lines in the chemical, electrical and electronics, and automobile industries have more complex configurations than other goods, necessitating more R&D effort to modify or adapt them to markets abroad. Ideally, one should therefore work with country-level data disaggregated by industry. Unfortunately, it is not possible because industry-level R&D data of the US Bureau of Economic Analysis (BEA) are plagued by missing values (see below). As the second-best alternative, we use an index measuring the R&D potential of the industry composition (which we dub the ‘R&D potential index’, RPI) as an additional control variable.2 The proposed index is defined by the formula

RPIjt 

冤

N

兺itXijt 1 N 兺 Xijt i

N

冥

*100

i

where i denotes the share of industry i in total R&D expenditure incurred by the overseas affiliates of US manufacturing MNEs, Xij is gross output of industry i in total manufacturing output of US MNE affiliates in host country j, N is the number of industries and t is the time subscript. For a given country, j, RPI is simply the global R&D-share-weighted average of manufacturing output of US MNE affiliates normalized at the mean (unweighted) output. Since the index is intended to compare only patterns of output across countries, it should not be influenced by the relative size or scale of MNE operations in individual countries. If the industry composition of MNE output in country j is identical to the industry composition of R&D expenditure in global operations of US MNEs, the index will take on the value of 100. A higher numerical value of the index implies greater R&D potential of the output composition of MNE affiliates’ operations in the given country, the other factors influencing R&D propensity remaining constant. The capital stock of US MNE affiliates in host countries (KUSF) is used as a control variable for two reasons. First, the relative importance of the given country as an investment location can presumably be an important consideration in R&D location decision of MNEs. Second, once controlled for the market size, the FDI stock is a reasonable proxy for the duration of MNE operations in a given country (Lipsey 2000). It should capture the evolving pattern over time of R&D activities in a given country. For these reasons we expect a positive relationship between R&D intensity and KUSF.

184

Multinational enterprises in Asian development

We consider three country-group dummies – developed countries (mature industrial countries, DIC) defined to cover OECD Europe, North America, Japan, Australia and New Zealand; the newly industrialized countries in East Asia (Hong Kong, Korea, Taiwan and Singapore, DNIC); and other developing countries (DODC) to capture possible differences in the degree of R&D intensity associated with the stage of development (with DICs as the controlled group).3 DNIC and DODC will also be interacted with the other explanatory variables in alternative regression runs to test whether the hypothesized relationship between R&D intensity and each of these variables is sensitive to the stage of development of countries. Time dummy variables (TIME) are included to capture time-specific fixed effects, with the first sub-period (1990–92) as the base dummy. Finally, a ‘crisis dummy’ (CRIS) is included to allow for the possible impact of the recent financial crisis on R&D activities of MNE affiliates in Indonesia, Malaysia, Singapore, Thailand and the Philippines. This variable takes value 1 for the sub-period 1996–98 and 0 otherwise for these five countries. Based on the above discussion, the estimating equation is specified as follows: RDSit 
1GDPit
2DMSit
3DISTi
4TECHit
5RDPNit
6TPWGit
7TINSit
8IPRit
13KUSFit
9RPIit
1DNICi
2DODCi
3CRISi
TIMEt
 where RDS is research and development intensity (research and development expenditure as a percentage of sales turnover), and subscripts i and t denote countries and time respectively. The explanatory variables are listed below (with the expected sign of the regression coefficient of each variable given in brackets): GDP (
) DIST (
) DMS ( or
) TECH (
) RDPN (
) TPWG () TINS (
) IPR (
) KUSF (
) RPI (
)

real gross domestic product distance percentage of domestic sales in total affiliate sales turnover technology intensity index R&D personnel per million population wages of technical personnel tax incentives for firm-level R&D activities intellectual property right index capital stock of US firms (at the beginning of the each subperiod) An index of R&D potential of output mix

Globalization of R&D

DODC(?) DNIC(?) CRIS(?) TIME

 

185

dummy variable for developing countries other than NICs dummy variable for newly industrialized countries in East Asia financial crisis dummy (for Thailand, Indonesia, Malaysia and the Philippines) a matrix of time dummy variables (each of which takes value 1 for a given year and zero otherwise) to capture time-specific ‘fixed’ effects a constant term a stochastic error term, representing other omitted influences.

Data The data on the dependent variable and three explanatory variables (DMS, RPI, KUSF) are compiled from the electronic data files of the Annual Survey of US Direct Investment Abroad conducted by the Bureau of Economic Analysis (BEA), the US Department of Commerce. The data relate to majority-owned, non-bank affiliates of US-headquartered corporations, as tracked by the BEA. The BEA started reporting data on R&D on an annual basis with effect from 1990 and the latest year for which data was available was 2001.4 Therefore, our data set covers the 12-year period from 1990 to 2001. For confidentiality reasons, the BEA does not divulge the response of individual firms and reports only country-level data (disaggregated at the two-digit level of the standard industry classification) for those countries in which there are a sufficient number of US firms with sizeable activities. It is not possible, however, to construct continuous data series at the industry level for a sufficient number of countries because the incidence of data suppression resulting from the application of the single-firm disclosure rules is much more severe at that level. Even for total manufacturing, there are considerable gaps in data for a sizeable number of countries. Thus, with a view to achieving a reasonable time series dimension and a reasonable country coverage, we limited the sample coverage to those countries for which there are no missing values for more than two consecutive years within the period 1990–2001.5 By doing so, we were able to construct a panel data set arranged at three-year intervals6 for 42 countries (see Appendix Table 8.A1). The use of threeyear averages rather than annual data is not a serious limitation because we are focusing here on long-term relations. Information on sources and time coverage of the other data series and the list of countries are reported in Appendix Table 8.A2.

186

Multinational enterprises in Asian development

Regression Results We used the random-effect estimator as our preferred estimation technique. The alternative fixed-effect estimator is not appropriate because our model contains a number of time-invariant variables (DIST, IPR, TINS, DODC, DNIC) which are central to our analysis. A major limitation of the randomeffect estimator, compared with its fixed-effect counterpart, is that it can yield inconsistent and biased estimates if the unobserved fixed effects are correlated with the remaining component of the error term. However, this is unlikely to be a serious problem in our case because N (the number of explanatory variables) is larger than T (the number of ‘within’ observations) (Wooldridge 2002, Chapter 10). The random-effect estimator also has the advantage of taking care of the serial correlation problem. The results are reported in Table 8.5. Summary statistics for the data used in the estimation are presented in Table 8.6 to facilitate interpretation of the results. All variables, other than the two ordered qualitative variables (IPR and TINS) and the dummy variables, have been used in natural logarithms in estimation. In experimental runs, we interacted the two country-group dummies, DNIC and DODC, with other explanatory variables to test whether the hypothesized relationship between R&D intensity and each of these variables is sensitive to the stage of development of countries. Only two interaction terms – DODC*DMS and (DODC
DNIC)*IPR – turned out to be statistically significant.7 The estimate of the full model is reported as Equation 1. In this equation, the coefficients on two variables (KUSF and TINS) are statistically insignificant (with t-ratios of less than 1) with the unexpected (negative) sign. The final equation estimated after deleting these variables (our ‘preferred model’) is reported as Equation 2.8 Since there was some evidence of heteroscedasticity, t-ratios of regression coefficients were computed from standard errors estimated using White’s heteroscedasticity consistent covariance matrix estimator (White 1980). An examination of the squared multiple correction coefficient of each explanatory variable on the other explanatory variables (last column, Table 8.6) suggests that multicollinearity does not cause problems in interpreting the individual regression coefficients. For all explanatory variables except TECH, the squared multiple correction coefficient is smaller in magnitude than the R2 of the parent regressions. The relatively high intercorrelation of TECH does not seem to cause problems, because that variable has ample variability (a coefficient variation of over 2.86, the highest among all explanatory variables) (Goldberger 1991, Chapter 23). For the purpose of comparison, ordinary least squares (OLS) estimates of the two equations are reported in Appendix Table 8.A3. Also reported in the table are the random-effect and fixed-effect estimates obtained after

187

Globalization of R&D

Table 8.5

Determinants of R&D intensity: random-effect GLS estimates1



Constant term

GDP

DIST

Real gross domestic product Domestic market share of total sales Distance

TECH

Technology index

DMS

R&D personnel per million population TPWG Wages of technical personnel TINS Tax incentives for firm-level R&D IPR Intellectual property protection KUSF Stock of fixed capital of US MNEs RPI R&D potential of output composition Dummy variable DODC Developing country dummy DNIC Newly industrialized country dummy DODC* Interaction term of DMS DC and DMS (DODC
Joint interaction term of DNIC)*IPR DODC and NIC, and IPR CRIS Financial crisis dummy

RDPN

R-sq: Overall Within Between Wald test, X2 Observations

Equation 1

Equation 3

Equation 2

5.34 (2.39)**
0.26 (3.07)*** 0.73 (3.76)***
0.07 (0.46)
0.13 (1.65)**
0.21 (2.01)** 0.50 (2.36)** 0.12 (0.97)
0.18 (2.04)** 0.05 (0.73)
0.87 (4.08)***

1.90 (1.10)
0.28 (4.36)*** 0.78 (4.51)***
0.06 (0.48)
0.10 (1.25)*
0.19 (1.97)** 0.43 (2.28)**

6.25 (3.05)***
0.21 (3.01)*** 0.64 (3.66)***
0.12 (0.88)
0.12 (1.52)*
0.19 (1.96)** 0.45 (2.33)***


0.13 (1.48)*


0.18 (2.03)**

3.24 (2.09)**
0.63 (0.79)
0.96 (3.01)*** 0.19 (1.56)* 0.51 (2.17)** 0.69 0.30 0.81 179.06*** 168

3.89 (2.54)**
0.44 (0.56)
1.08 (3.40)*** 0.15 (1.32)* 1.90 (1.10) 0.68 0.19 0.81 169.18*** 168


0.85 (4.01)** 3.30 (2.18)***
0.65 (0.83)
0.99 (3.16)*** 0.20 (1.72)** 0.52 (2.24)** 0.69 0.29 0.80 183.56*** 168

Notes: 1. All variables (except DODC, NIC, TINS and IPR) are in logarithms. The t-ratios based on White’s heteroscedasticity adjusted standard errors are given in brackets, with statistical significance (one-tailed test) denoted as: *** 1 per cent, ** 5 per cent and * 10 per cent.

188

Table 8.6

RDS GDP DMS TECH RDPN DIST TPWG KUSF RPI IPR TINS

Multinational enterprises in Asian development

Summary data on variables used in the regression analysis Maximum

Minimum

Mean

2.04 15.54 4.57 4.46 8.65 9.70 4.44 12.35 4.07 8.68 5.63

2.75 8.91 2.23 4.61 3.61 6.61 1.70 5.70 1.38 1.61 1.68

0.41 12.27 4.07 0.92 6.73 8.92 3.36 8.76 2.99 5.76 3.69

Standard deviation 0.96 1.23 0.46 2.66 1.35 0.58 0.70 1.45 0.61 1.76 0.96

Coef. of variation 2.35 0.10 0.11 2.89 0.20 0.06 0.21 0.16 0.20 0.31 0.26

R2 – 0.32 0.35 0.80 0.69 0.05 0.67 0.31 0.14 0.56 0.47

Notes: All variables other than IPR and TINS are in natural logarithms. R2 Squared multiple correlation coefficient of each explanatory variable with respect to all other explanatory variables. – Not applicable.

deleting the time invariant explanatory variables. The random-effects estimates (Table 8.5) and OLS estimates (Table 8.A3) are remarkably similar, suggesting that unobserved effects are relatively unimportant in our model. Fixed-effects and random-effects estimates of the model after deleting the time invariant parameters are also closely comparable. However, the hypothesis that the fixed-effects estimator is better than the random-effects estimator is rejected by the Hausman test (Hausman 1978). The following discussion focuses solely on the results reported in Table 8.5. The coefficient on GDP is significant at the 1 per cent level, supporting the hypothesis that, other things remaining unchanged, domestic market size is a key determinant of R&D intensity of MNE affiliates. A 1 per cent change in market size is associated with a 0.28 per cent change in R&D across countries. As we anticipated a priori, the result for DMS is mixed. For the entire country sample, its coefficient is statistically significant with the negative sign, suggesting that greater domestic market orientation is negatively related with R&D intensity. However, the coefficient of the interaction dummy DODC*DMS is positive and statistically significant, suggesting that a 1 per cent increase in domestic market orientation is associated with a 0.48 per cent increase in RDS among other developing countries (that is, developing countries excluding NICs). As already noted, the interaction

Globalization of R&D

189

dummy for NICs (DNIC*DMS) was found to be statistically insignificant. These contrasting results confirm the view that, given the similarities of demand patterns between the host country and the major (mostly developed-country) markets and the virtual absence of trade barriers to trade, greater export orientation provides impetus for increase in R&D effort for MNE affiliates located in developed countries. The NICs, given their heavy export orientation, seem to exhibit a similar relationship between these two variables. By contrast, given some peculiarities in domestic demand patterns and presumably also because of remaining barriers to integrate in the global economy, there seems to be some need for undertaking product-adaptation-type R&D in ODCs. In sum, the link between the nature of market orientation and R&D intensity varies across countries, depending on the stage of development and global market integration of the countries under study. There is strong statistical support for the hypothesis that R&D intensity of domestic manufacturing in the host country is a strong attraction for MNEs to undertake R&D activities in that country. The coefficient on TECH is significant at the 1 per cent level. It suggests that a 1 per cent increase in the national technology effort is associated with a 0.15 per cent increase in R&D intensity of MNE affiliates. Among the variables included to capture the domestic investment climate, the coefficient on RDPN is statistically significant with the expected (positive) sign, providing support for the hypothesis that the availability of R&D personnel is a significant influence on the R&D location decision of MNEs. The results for TPWG corroborate this inference; the wage rate of technical personnel has a strong negative relationship with R&D intensity of MNE operations. This result, however, needs to be qualified for the poor quality of the data series (the wages of nonproduction workers) used to represent the cost of hiring technical personnel. Perhaps the estimated coefficient provides a possible lower bound, because normally the wages of R&D personnel are higher and increase at a faster rate than wages of non-production workers in general. IPR has a statistically significant (at the 10 per cent level) coefficient with the expected positive sign, but its interaction term with developing countries yields a negative coefficient of virtually the same magnitude. This finding is consistent with the view that international property protection is a positive tool for promoting R&D activities only in countries with appropriate complementary endowments and policies (Maskus 1998). The results for TINS cast doubt on the effectiveness of financial incentives as a policy tool for promoting R&D activities by MNE affiliates in host countries.9 A plausible explanation seems to be that, as the MNEs have access to intra-firm trade and other means to minimize the actual tax

190

Multinational enterprises in Asian development

burden, tax incentives are not an important consideration for MNEs in their R&D location decisions when controlled for the other relevant variables (Clausing 2001, Mansfield 1986). The coefficient on DIST has the expected positive sign, suggesting that geographical distance still matters for the overseas R&D location decision of MNEs, but this relationship is not statistically significant. There is no evidence to suggest that the relative importance of a given country in global operations of US MNEs as measured by the size of the stock of capital (KUSF) is important in explaining R&D intensity of affiliates operating in that country. Contrary to the popular belief that underpins investment promotion campaigns in many host countries, total foreign direct investment and R&D activities do not seem to go hand in hand. The coefficient on RPI is statistically significant with the expected (positive) sign, supporting the hypothesis that the industry composition does matter in explaining inter-country differences in the degree of R&D intensity of MNE affiliates. We also re-estimated Equation 2 (our preferred equation) after deleting RPI and found that individual regression coefficients attached to the other variables are remarkably resilient to its inclusion/exclusion.10 At the same time, the deletion of RPI from Equation 2 was not supported by the standard variable deletion F-test.11 The upshot is that industry composition is an important determinant of the overall R&D intensity of MNE operations in a given country over and above the other variables considered here. Finally, how do our findings compare with those of the previous studies? Our results confirm the findings of Kumar (1996 and 2001) that MNEs prefer to locate their R&D activities in countries that are able to offer, among other things, large markets and technical resources. However, we find that there is no unique relationship between the nature of market orientation of MNE affiliates and R&D intensity. There is a positive relationship between these two variables only for developing countries. For developed countries and NICs in Asia, the relationship is negative, implying that greater export orientation is associated with more, rather than less, R&D intensity. The relationship depends very much on the stage of development of a given country. Thus there is no case for supporting domesticmarket-oriented policies on the grounds that they promote local R&D activities by MNEs in developing countries. Unlike Kumar (1996 and 2001) we find some statistical support for the view that intellectual property protection can play a positive role in promoting innovatory activities, depending of course on the presence of appropriate complementary endowments and policies. Our results on the impact of tax incentives on R&D activities run counter to those of Hines (1995) and Hines and Jaffe (2001) relating to

191

Globalization of R&D

US-based MNEs and Bloom and Griffith (2001) relating to UK-based MNEs. These studies have uncovered a statistically significant positive effect of tax incentives on the distribution of inventive activity between the home country and overseas locations of MNEs. We suspect that the failure to appropriately control for relevant explanatory variables may have biased the results of these studies against the null hypothesis of their experiments. Each study has controlled for only one relevant variable (Hines and Hines and Jaffe: R&D intensity of the host country; Bloom and Griffith: domestic real output) in testing the link between internationalization of R&D and tax incentives. Interestingly, our data set permits us to replicate their results through similar (arbitrary) variable choice. For instance, truncating our model to retain TECH (our measure of the technology intensity of the host country) as the only control variable yields:12 R&D  1.27
0.20 TECH
0.18 TINS (4.84)**

(7.56)***

R2  0.47

F  75.00

(2.49)**

When GDP (our measure of real output) is used in place of TECH, R&D  5.32
0.28 GDP
0.39 TINS (9.19)*** R2  0.39

(5.44)***

(5.90)**

F  55.85

Both equations provide strong statistical support for the hypothesis that tax incentives are a significant determinant of inter-country differences in R&D intensity of US MNE affiliates. However, the (arbitrary) truncation of the model in each case is not supported by the standard variable deletion (F) test conducted against our full model (Equation 1 in Table 8.5).13

CONCLUSION We have examined patterns and determinants of overseas R&D activity by MNEs using a new panel data set relating to US-based MNEs over the period 1990–2001. It is found that domestic market size, geographic distance, overall R&D capability of the country and cost of R&D personnel are key determinants of the R&D intensity of operations of US MNE affiliates. There is also evidence that, contrary to the conventional wisdom, the impact of domestic market orientation of affiliates on R&D propensity

192

Multinational enterprises in Asian development

varies among countries depending on their stage of development. The degree of domestic market orientation has a positive impact on R&D intensity only in developing countries other than the East Asian NICs. For the latter countries and developed countries the two variables are negatively related, suggesting that greater export orientation is associated with greater (not less) R&D intensity. There is also evidence that, when controlled for the other relevant variables, the industry composition does matter in explaining inter-country variations in R&D intensity. R&D-related tax incentives do not seem important in explaining inter-country differences in R&D intensity when appropriately controlled for other relevant variables. Intellectual property protection seems to matter for mature economies with complementary endowments. Overall, our findings serve as a caution against governments paying too much attention to turning MNE affiliates into technology creators as part of their foreign direct investment policy. An MNE’s decision to undertake R&D activities in a given country seems largely endogenous to its overall growth and development process. Excessive concern as to where R&D is performed may tend to downplay the more important role of MNEs as a conduit of technology transfer. Even if MNE affiliates generate little or no technology locally, they can still play an important role in improving local innovative capabilities through technology transfer.

NOTES *

1.

2. 3.

Based on a paper (written jointly with Archanun Kohpaibboon) presented at the conference, Technology and Long-run Economic Growth in Asia, organized by the N.W. Posthumus Institute and Hitotsubashi University, 8–9 September 2005, Tokyo. We are grateful to Kyoji Fukao, Hiroyuki Odagiri, F. Odaka and other conference participants and Hal Hill, Xin Meng and Russell Thomson for useful comments and constructive criticism. Most of the existing econometric analyses of overseas R&D of MNEs have specifically focused on change in propensity to locate R&D overseas at the industry or firm level, ignoring the geographic dimension. (For surveys of this literature see Caves 1996, Golberman 1997 and Kumar 2001). So far, three studies have examined inter-country distribution of overseas R&D activities of US MNEs using data compiled from the Benchmark Survey of US Direct Investment Abroad conducted by the US Bureau of Economic Analysis. These are Kumar 1996 (based on pooled country-level data for 1977, 1982 and 1989), Kumar 2001 (pooled country-level data for 1982, 1989 and 1994), and Hines 1995 (country-level data for 1989). Similar studies for other countries are Kumar 2001 (Japanese MNEs, pooled country- and industry-level data for 1982, 1989 and 1994), Odagiri and Yasuda 1996 (Japanese MNEs, firm-level data for 1990), Zejan 1990 (Swedish MNEs, country- and firm-level data for 1978) and Fors 1998 (Swedish MNEs, pooled industry- and firm-level data for 1978 and 1990). We are grateful to Kyoji Fukao for suggesting this index. In experimental runs we also tested further desegregation of DODCs into East Asian developing countries (other than NICs) and other developing countries. These two

Globalization of R&D

4. 5. 6.

7. 8. 9.

10. 11. 12. 13.

193

groups were finally combined (to form DODCs) because we were not able to detect a statistically significant difference between the two sub-groups in relation to the hypothesized impact of the explanatory variables on R&D intensity. For details on this database see Hanson et al. (2001, Appendix). In cases where the reported amount is greater than zero but less than $500 000, we set the level of investment at $250 000. That is, an observation is a country’s performance average over a three-year period, yielding four observations (averages for 1990–92, 1993–95, 1996–98 and 1999–2001) for each country. If a data point is missing within any three-year period, a two-year average is used, and when two data points are missing, the available data point is used as the three-year average. Of the total 168 observations on R&D, only 17 observations have been ‘approximated’ in this way. As we will discuss below, this is much in line with our expectations (as discussed under the model specification). This specification choice is amply supported by the standard variable deletion (F) test; the joint test for zero restriction on the coefficients of the four variables yielded F (4, 152)  1.097. The data series on TINS captures the state of tax incentives for R&D circa 1999/2000 (see Appendix Table 8.A2). However, this does not seem to be a serious problem because most changes in effective tax incentives occurred in the 1980s. For instance, see United Nations (1996), Bloom et al. (2002, Figures 1 and 2). This alternative estimate is available from the authors on request. The test for zero restriction on the coefficients of RPI performed on Equation 2 is F (1, 151) 8.37, which is significant at the 1 per cent level. The following two equations are OLS estimates. The results of the variable deletion (F) test for the two equations are F (16 147) 7.02 and F (16 147) 12.71 respectively.

194

Multinational enterprises in Asian development

APPENDIX 8.1: COUNTRY COVERAGE, VARIABLE DEFINITION AND DATA SOURCES AND SUPPLEMENTARY REGRESSION RESULTS Table 8.A1

Country coverage

Industrial countries

Developing countries

Europe Austria Belgium Denmark Finland France Germany Greece Ireland Italy Netherlands Norway Portugal Spain Sweden Switzerland United Kingdom Canada Japan Australia New Zealand Israel

Asian NICs Hong Kong Korea, Republic of Singapore Taiwan Other Asia China Indonesia Malaysia Philippines Thailand India Latin America Argentina Brazil Chile Colombia Ecuador Mexico Panama Peru Venezuela South Africa Turkey

195

Globalization of R&D

Table 8.A2

Variable definition and data sources

Variable

Source

Time coverage

RDS

Research and development expenditure as a percentage of total sales turnover

Compiled from the electronic data files of the Annual Survey of US Direct Investment Abroad, the Bureau of Economic Analysis, http//www.bea.doc. gov./bea/uguide.htm#_1_23

1990–2001

DMS

Domestic market share of total sales

As above

1990–2001

CHEM

Percentage of chemical products in total affiliate output

As above

1990–2001

RPI

Index of R&D potential – a composite index of R&D potential of output composition

As above

1990–2001

KUSF

Stock of fixed capital of US MNEs (at the beginning of the 3-year period)

As above

1990–2001

GDP

Real gross domestic product

Word Development Indicator Database, World Bank, http://www.worldbank.org

1990–2001

DIST

Great-circle distance between The Western Cotton Research the capital city of the given Laboratory database, US country and Washington DC Department of Agriculture, www.wcrl.ars.usda.gov/ cec/java/lat-long.htm

Not applicable

TECH

Technology effort index – a composite index of productive enterprise R&D expenditure and the number of patents registered in the USA, both normalized by mid-year population

Lall (2002)

Circa 1999

RDPN

R&D personnel per million population

UNESCO Statistical Yearbook, Geneva: United Nations

1990–2001

TPWG

Wages of technical personnel

US Department of Commerce, Bureau of Economic Analysis, Benchmark Survey of US Investment Abroad 1994

1996 (?)

TINS

Index of tax incentives for firm-level R&D (ranges from 1 (no

World Economic Forum, Global Competitiveness Report

2000 and 2001 (average)

196

Multinational enterprises in Asian development

Table 8.A2 (continued) Variable

Source

Time coverage

incentives) to 7 (incentives most prevalent)) IPR

Index of intellectual property protection (ranges from 1 (least binding) to 10 (most stringent))

World Economic Forum, Global Competitiveness Report

1990–2001

197

Globalization of R&D

Table 8.A3

Determinants of R&D intensity: alternative regression results1 OLS



Constant term

GDP

Real gross domestic product Domestic market share of total sales Distance

DMS DIST TECH RDPN TPWG TINS IPR KUSF RPI

Technology index R&D personnel per million population Wages of technical personnel Tax incentives for firm-level R&D Intellectual property protection Stock of fixed capital of US MNEs R&D potential of output composition

Dummy variables DODC Developing-country dummy NIC Newly industrialized country dummy DODC* Interaction term of DMS DC and DMS (DODC
Joint interaction term of NIC)*IPR ODC and NIC, and IPR CRIS Financial crisis dummy R-sq: Overall Within Between F-statistic Wald, X2 Observation

OLS

Fixed effects2

Equation 1 Equation 2

Equation 3

Random effects2 Equation 4

3.98 (2.30)**
0.26 (3.71)*** 0.74 (3.86)***
0.08 (1.08)
0.16 (2.15)**
0.19 (1.90)** 0.50 (3.04)*** 0.08 (0.93)
0.14 (1.58)* 0.04 (0.84)
0.64 (2.62)***

4.65 (2.90)***
0.22 (4.67)*** 0.66 (3.69)***
0.12 (1.85)**
0.15 (1.96)**
0.17 (2.19)** 0.47 (3.17)***

10.95 (7.26)*** 0.33 (2.56)** 0.69 (2.33)**

7.80 (5.61)***
0.21 (2.60)*** 0.48 (2.84)***


0.23 (1.34)


0.24 (2.40)**


0.15 (1.77)**


0.30 (2.51)** 0.02 (0.14)
1.17 (5.02)***


0.15 (2.54)** 0.06 (0.99) 1.01 (4.81)***

4.12 (2.78)***
0.33 (0.50)
1.12 (3.70)*** 0.11 (1.37)* 0.53 (2.01)** 0.66

4.17 (2.92)***
0.33 (0.52)
1.14 (4.70)*** 0.10 (1.40)* 0.55 (2.06)** 0.67


0.70 (1.27) 0.23 (1.41)*


0.06 (0.88) 0.05 (1.48)*

19.33*** – 168

21.75*** – 168

0.10 0.29 0.08 4.29*** – 168

0.60 0.27 0.70 21.57*** 128.01*** 168


0.62 (2.56)**

Notes: 1. All variables (except DODC, NIC, TINS and IPR) are in logarithms. The t-ratios based on White’s heteroscedasticity adjusted standard errors are given in brackets, with statistical significance (one-tailed test) denoted as: *** 1 per cent, ** 5 per cent and * 10 per cent. – Not applicable. 2. The hypothesis that the fixed-effects estimator is better than the random-effects estimator is rejected by the Hausman test (Hausman 1978) (X2 (2,11) 6.56).

9. Multinational firms and factor proportions in manufacturing: does parentage matter?* An important development in the scenario of internationalization of economic activity since about the late 1970s is the emergence of multinational enterprises from developing countries. As an outcome of rapid industrial development in the post-war era, the more industrialized of the developing countries now provide home to many indigenous firms which have grown in strength to expand production beyond national frontiers, mostly (but not exclusively) to neighbouring developing countries. These new multinationals (or third-world multinational enterprises (TWMNEs) as they are popularly known), still account for only a minor share of the total stock of direct foreign investment in developing countries,1 but in a significant number of them the incidence of such investment is already quite substantial. There is evidence that the number of TWMNEs, the sophistication of their activities and their international spread have greatly increased over the years (Yeung 1999, Chudnovsky and Lopez 2000). Beginning with the pioneering works of Diaz-Alejandro (1977), Wells (1977) and Lecraw (1977), a sizeable body of literature, mostly explanatory in nature, has developed on this subject.2 A basic assertion of this literature is that TWMNEs possess a number of unique characteristics which distinguish them from the ‘traditional’ (developed-country) multinationals (DCMNEs). One such hypothesized difference relates to the relative appropriateness of their technology to host developing countries. A well-known argument in the long-running discussion on labour absorption in the manufacturing sector of developing countries is that the technology introduced by DCMNEs (which virtually monopolize world trade in technology) is highly capital intensive and, therefore, tends to reduce the employment potential of industrialization (Sen 1980, Singer 1973). This argument is based on the presumption that a given investor’s technology reflects peculiarities of the factor market conditions in their home country. On the basis of the same presumption, the conventional wisdom on TWMNEs holds that the technology of these firms is more appropriate to the relative labour and capital cost in the host (capital-receiving) developing country. This 198

Factor proportions in manufacturing

199

belief has nurtured the hope that TWMNEs might be able to ‘make a special kind of contribution to the development of poor countries’ (Wells 1983, p. 1). The hypothesis that DCMNE affiliates in developing countries are more capital intensive than indigenous firms has been subjected to extensive empirical testing in recent years.3 However, empirical evidence on the relative capital intensity of TWMNE affiliates is sparse; to our knowledge there have been only three previous reported studies. The hypothesis that TWMNEs are more labour intensive has apparently been supported by the three major studies which have investigated the issue (Wells and Warren 1979, Busjeet 1980, Lecraw 1977).4 These studies are based on rather sketchy data covering a small number of countries and employing questionable analytical procedures. Wells and Warren’s (1979) study of Indonesian manufacturing firms, as well as Busjeet’s (1980) study of manufacturing firms operating in the export processing zones of Mauritius and the Philippines, compared average factor intensity without controlling for other firm attributes. Lecraw’s (1977) more carefully designed study of manufacturing firms in Thailand attempted to control for the influence of these factors. However, this study was based on data from the ‘old breed’ of TWMNEs (set up in the early 1970s) which tended to locate almost exclusively in those sectors employing relatively unsophisticated technologies. There is evidence that in recent years TWMNEs have been rapidly moving into more sophisticated product lines requiring technologies, skills and production scales previously considered to be the exclusive preserve of DCMNEs (Lall 1983a). As the scope for technological adaptation is rather limited in such product lines (Forsyth et al. 1980), it could be argued that differences in labour intensities found in these studies may have reflected conditions prevailing at an early stage of TWMNEs which is now past. Thus much more systematic empirical investigations covering a variety of host-country situations are needed before any valid generalization on this issue can be made. This chapter presents results from an analysis of the differences in capital intensity between TWMNEs, DCMNEs and indigenous private firms in Sri Lankan manufacturing industry. Access to hitherto unexploited data has allowed us to avoid some of the major drawbacks of previous studies and to test more rigorously important hypotheses regarding the choice of technique, as well as to explore the dynamic adjustments of MNEs over time to host-country market environments. The chapter began with a survey of the limited empirical evidence available on the subject. The next section provides an overview of the trends and patterns of direct foreign investment in Sri Lankan manufacturing industry. The following section sets out the methodology and describes the data base. Then the results are

200

Multinational enterprises in Asian development

presented and analysed. The principal conclusions are contained in the concluding section.

FOREIGN DIRECT INVESTMENT IN SRI LANKAN MANUFACTURING As in many other developing countries, import restriction was the major factor which triggered the entry of MNEs into Sri Lankan manufacturing. In response to a deteriorating balance-of-payments situation, Sri Lanka moved rapidly to a regime of stringent import and exchange restrictions in the 1960s. When their market shares were threatened by import controls, many foreign firms set up affiliates within Sri Lanka to undertake the domestic production and/or assembly of product lines which had previously been supplied from overseas production centres.5 By the end of the decade, there were 51 MNE affiliates in operation, and they accounted for about 28 per cent (38 per cent if the public sector is excluded) of total manufacturing output (Fernando 1971, p. 84). The period from 1970 to 1977 witnessed a marked reduction in the tempo of foreign capital participation in the economy. The left-of-centre government in power during this period placed a heavy emphasis on expanding the role of the public sector in the economy and foreign direct investment (FDI) did not receive much encouragement (Athukorala and Rajapatirana 2000a). There was a marked shift in development policy in favour of FDI as part of the liberalization reforms initiated in 1977. The most important aspect of the new FDI policy was the setting up of the Greater Colombo Economic Commission (GCEC) in 1978 with wide-ranging powers to establish and operate Export Processing Zones (EPZs).6 As an important part of the FDI policy, steps were also taken to enter into Investment Protection Agreements and Double Taxation Relief Agreements with the major investing countries. A guarantee against nationalization of foreign assets without compensation was provided under Article 157 of the new Constitution of Sri Lanka, adopted in 1978. Until 1990, there was no major change in the policy towards non-FTZ (free trade zone) foreign ventures. Majority local ownership continued to be the general rule for approving such projects. Even though more liberal ownership criteria (including up to 100 per cent foreign ownership) were applied in approving export-oriented firms, these firms were not eligible for the lucrative incentives offered to EPZ firms. As part of the second-wave liberalization, a new ‘Investment Policy Statement’ was announced in 1990 with several important changes to the foreign investment policy framework in line with the increased outward orientation of the economy. Activities of the

201

Factor proportions in manufacturing

Foreign Investment Advisory Committee (FIAC) and GCEC were brought together under a new Board of Investment (BOI) in order to facilitate and speed up investment approval within a unified policy framework applicable to both import-substituting and export-oriented investors. Restrictions on the ownership structures of joint-venture projects outside EPZs were abolished. EPZs’ privileges were extended to local investors who establish new export-oriented projects in all parts of the country; these were given freetrade status (in addition to the areas demarcated as EPZs). This provision, which was initially applicable only to investors who were prepared to implement their projects prior to 30 September 1991, was extended in February 1993 to local investors starting new export ventures as well as existing companies which set up production facilities outside the Western Province. Since then, this has become a permanent feature of the BOI approval procedure. The new policy orientation certainly proved to be very attractive to foreign investors. Having averaged less than half a million dollars a year during 1970–76, net FDI inflows increased rapidly in the post-reform period, reaching US$ 64 million in 1982 (Figure 9.1). The outbreak of war in 1983 severely disrupted this impressive trend and annual flows were subsequently only in the US$17–58 million range. The second-wave reforms and the temporary cessation of hostilities during the first half of the 1990s 16

250

14 12 10

150

8 100

6

FDI/GDFCF, %

FDI, US$ million

200

FDI, US$ million FDI as a percentage of GDFCF

4

50

2 0 1970–75* 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

0

Note: * Annual average. Source: Based on data compiled from Central Bank of Sri Lanka, Annual Report (various issues).

Figure 9.1 Foreign direct investment in Sri Lanka: value in US$ million (left scale) and as a percentage of gross domestic fixed capital formation (GDFCF) (right scale)

202

Multinational enterprises in Asian development

witnessed a notable surge in FDI, which increased to an all-time high of US$184 million in 1993. The relative contribution of net FDI inflows to private sector fixed capital formation increased from 0.1 per cent during 1970–77 to 5 per cent during 1978–89 and to over 12 per cent during 1990–94. However, FDI inflows, while remaining remarkably high compared with the pre-reform era, have declined since 1995 both in US$ value and in relation to fixed capital formation.7 This reflects continuing uncertainty in the investment climate, due mostly to the long-running ethnic conflict in the country (World Bank 2004). Data relating to individual projects set up with FDI participation (henceforth referred to as ‘foreign-invested enterprises’ (FIEs) point to a clear shift in FDI from domestic-market-oriented to export-oriented activities (Table 9.1). Of 2042 FIEs set up during 1978–2002, 1341 received approval under the BOI special incentive scheme for export-oriented firms. These firms accounted for over 90 per cent of the envisaged investment and employment of all projects. This is in sharp contrast to the heavy importsubstitution bias in FIEs during the post-reform era: during 1967–77, a total of 82 foreign manufacturing firms were set up in Sri Lanka. Of these, only 13 were export-oriented ventures (garments 9; gem cutting 2; ceramicware 1; wall-tiles 1) (Athukorala 1995). As regards the source-country composition of FIEs, the majority of firms are of developing-country parentage (Table 9.2). South Korea is by far the largest investor, with capital participation in 112 firms, accounting for 16 per cent of total investment and 14 per cent of total exports of all export-oriented firms set up during 1978–2002. Standard labour-intensive manufacturing has been the main attraction to export-oriented foreign investors (Table 9.3), with a heavy concentration in the garment industry. However, since the later 1980s, there has been a noticeable increase in the number of FIEs in other labour-intensive activities, in particular footwear, travel goods, plastic products and diamond cutting and jewellery. There has also been an increase in processing of primary products which were previously exported in raw form, notably rubber-based products (heavy duty tyres, rubber bands and surgical gloves) and ceramics. Initially, the surge of FDI in the garment industry was a response to the imposition by developed countries of quota restrictions on garment imports from traditional East Asian developing countries under the Multi-Fibre Arrangement (MFA). However, by 1983, garment exports from Sri Lanka had also come under quota restrictions, and the BOI stopped approval of new investors in quota-restricted product categories. Since then the newly approved FIEs in the garment industry have been predominantly involved in the production of non-quota garments. According to BOI firm-level data, in 1990 there were 23 non-quota

203

1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998

4 19 33 61 90 109 133 144 165 181 204 230 263 304 382 545 707 855 982 1162 1372

2 125 186 300 338 328 367 396 415 495 497 488 522 566 686 1014 1225 1459 1781 2160 2435

2 84 114 193 222 209 231 239 255 314 318 304 349 386 477 674 756 887 1061 1282 1437

181 10 433 14 765 29 834 35 196 40 964 44 111 47 665 52 249 56 041 61 942 76 144 87 126 103 811 121 127 149 451 169 652 189 255 205 465 238 868 272 230

0 10 17 26 29 33 37 41 48 54 61 72 82 99 138 245 351 444 513 622 778

Foreign Employment Number share in investment (US$ m)

All firms1

0 112 169 222 210 198 185 188 202 239 241 243 294 340 459 744 911 1127 1414 1759 2039

Investment (US$ m)

0.0 67.9 62.0 73.2 72.6 73.0 73.3 73.7 75.3 72.3 73.5 72.3 78.4 79.2 78.8 70.1 64.5 62.7 60.5 60.3 59.9

Foreign share in investment (%)

Export-oriented firms 2

0 9588 13 424 26 946 27 653 31 897 32 835 35 699 38 892 41 538 46 182 57 277 65 159 79 809 95 533 120 171 138 612 156 645 171 665 202 689 235 311

Employment

0.0 52.6 51.5 42.6 32.2 30.3 27.8 28.5 29.1 29.8 29.9 31.3 31.2 32.6 36.1 45.0 49.6 51.9 52.2 53.5 56.7

0.0 89.6 90.9 74.0 62.1 60.4 50.4 47.5 48.7 48.3 48.5 49.8 56.3 60.1 66.9 73.4 74.4 77.2 79.4 81.4 83.7

0.0 91.9 90.9 90.3 78.6 77.9 74.4 74.9 74.4 74.1 74.6 75.2 74.8 76.9 78.9 80.4 81.7 82.8 83.5 84.9 86.4

Total Total Total number investment employment of firms

Share of export-oriented firms (%) in:

Envisaged investment and employment in firms in commercial operation, 1978–2002 (cumulative annual data)

Number Investment (US$ m)

Table 9.1

204

1582 1743 1882 2042

2811 2930 2707 2815

1599 1661 1483 1508

317 842 342 355 366 400 397 393

938 1065 1189 1341

2418 2564 2389 2516

Investment (US$ m)

57.3 57.0 54.8 53.4

Foreign share in investment (%)

Export-oriented firms 2

278 827 301 526 325 291 356 124

Employment

59.3 61.1 63.2 65.7

86.0 87.5 88.3 89.4

87.7 88.1 88.8 89.6

Total Total Total number investment employment of firms

Share of export-oriented firms (%) in:

Source:

Compiled from official records of the Sri Lanka Board of Investment.

Notes: 1. Firms set up under Section 17 (domestic-market-oriented firms) and Section 16 of foreign investment approval legislation (the BOI Act). 2. Projects approved under Section 16 of the BOI Law.

1999 2000 2001 2002

Foreign Employment Number share in investment (US$ m)

All firms1

(continued)

Number Investment (US$ m)

Table 9.1

205

Factor proportions in manufacturing

Table 9.2 Source country composition of export-oriented firms in commercial operation as at end of 20021 No. of Total Foreign Foreign share Export Employment firms investment investment in total (%) (%) (%) (%) investment (%) Developed countries Australia Austria Belgium– Luxemburg Canada Denmark France Germany Israel Italy Japan Netherlands Norway Sweden UK USA Other developed countries

260

34.6

37.7

73.1

42.0

32.3

21 2 22

4.8 0.0 3.1

6.4 0.0 3.2

89.4 34.4 70.0

3.7 0.0 8.0

1.9 0.0 4.6

4 3 6 44 2 4 54 10 6 7 45 27 3

0.1 1.0 0.2 5.9 0.0 0.2 4.8 2.2 0.2 1.4 6.8 3.5 0.6

0.1 0.4 0.1 5.3 0.0 0.2 5.6 2.0 0.2 1.8 8.2 3.6 0.5

71.9 25.4 35.5 60.5 100.0 100.0 77.4 62.9 59.3 89.2 80.3 68.9 85.2

0.1 0.2 0.3 3.9 0.0 0.2 4.9 1.7 0.1 1.4 7.0 9.8 1.1

0.1 0.2 0.3 4.5 0.1 0.2 7.0 2.3 0.2 0.5 3.8 6.1 0.6

Developing 286 countries China 11 Hong Kong 55 India 46 Indonesia 3 Korea, South 112 Malaysia 5 Maldives 2 Malta 2 Mauritius 2 Pakistan 8 Singapore 16 Saudi Arab 2 United Arab 2 Emirates Taiwan 12 Thailand 2 Other developing 6 countries

38.4

47.4

82.8

38.5

37.0

0.3 9.7 4.4 0.5 15.7 2.7 0.0 0.1 0.4 0.4 2.9 0.0 0.3

0.3 10.2 4.2 0.8 21.6 4.0 0.0 0.2 0.2 0.5 3.9 0.0 0.4

68.9 70.6 64.2 97.8 92.1 98.6 35.2 96.1 23.8 96.4 89.7 100.0 100.0

0.6 19.7 2.9 0.0 10.6 0.5 0.0 0.8 0.0 0.6 1.3 0.0 0.0

0.3 15.8 2.4 0.0 13.5 0.5 0.1 0.4 0.1 1.0 1.1 0.0 0.1

0.7 0.0 0.1

0.9 0.0 0.1

85.5 100.0 84.3

1.4 0.0 0.1

1.6 0.1 0.0

1.5 74.4

1.2 86.3

52.6 77.7

1.0 81.5

1.4 70.7

Unclassified2 All FIEs

14 560

206

Multinational enterprises in Asian development

Table 9.2 (continued) No. of Total Foreign Foreign share Export Employment firms investment investment in total (%) (%) (%) (%) investment (%) Local firms All firms

373 933

25.6 100.0

13.7 100.0

36.0 67.0

18.5 100.0

29.3 100.0

Notes: 1. Projects approved during 1978–2002 under the BOI special provisions (Section 16 of the BOI Law). 2. Projects set up with investment from tax haven countries (Bahamas, Bermuda, Channel Islands, Gibraltar, Virgin Islands). FIE Foreign-invested enterprise. Source: Compiled from official records of the Board of Investment.

garment-producing FIEs, which accounted for 10 per cent of total garment exports by BOI-approved firms. By 2002 these figures had increased to 36 and 42 per cent respectively. Another important recent development is the setting up of a number of yarn- and textile-producing FIEs to supply inputs to the garment industry. The investment promotion campaign of the GCEC in the early stage aimed at attracting FDI into assembly activities in electronics and other high-tech industries. However, it seems that the increase in political risk, following the eruption of the ethnic conflict in 1983, has prevented a positive response. Foreign firms involved in vertically integrated assembly activities, unlike those involved in light consumer goods industries, view investment risk from a long-term perspective because output disruption in a given location can disturb production plans for the entire production chain. In fact, two electronics multinationals, Motorola and the Harris Corporation, abandoned plans to set up assembly plants in Sri Lanka in the early 1980s as the political climate began to deteriorate. There is evidence that there is something of a herd mentality in the site selection process of multinational electronics firms, particularly if the first-comer is a major player in the industry. If the two projects of Motorola and the Harris Corporation had been successful, other multinationals would have followed suit (Athukorala and Rajapatirana, 2000a, pp. 106–8; Snodgrass 1999). Instead, today only a handful of small-scale firms from Germany, Japan and Korea are involved in electronics assembly activities in Sri Lanka. Foreign direct investment has been instrumental in dramatically transforming Sri Lanka’s export trade. The export structure of the country evolved during the colonial era relied heavily on a limited range of primary

207

Factor proportions in manufacturing

Table 9.3 Industry composition of export-oriented manufacturing firms with FDI,1 1982, 1995 and 2002 1982 No. of firms 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 2 2.1 2.1a 2.1b 2.1c 2.3 2.4 2.5 2.6 2.7 2.7 3

3.1 3.2

Resource-based manufacturing Food, beverages and tobacco Natural rubber-based products Ceramics/granite products Coir products Wood products Gem cutting Other Standardized consumer goods Textiles and garments Handloom and textile products Textile fabric, knitted and woven Garments Leather goods Plastic goods Footwear Sport goods Diamond cutting and jewellery Other Component production/ assembly in fabricated metal products and machinery and transport equipment Electronics and electrical goods Other Total foreign

1995

Export (%)

No. of firms

2002

Export (%)

No. of firms

Export (%)

4

3.3

58

11.5

151

16.4

1 1

– 2.8

17 16

5.2 3.2

63 28

4.3 9.3

1 1 – – –

0.2 0.3 – – –

10 6 2 7 –

2.3 0.5 0.1 0.2 –

23 20 7 6 4

2.0 0.5 0.1 0.1 0.0

27

94.9

155

79.6

321

71.1

21 0

86.9 0

81 8

59.4 8.1

152 15

51.6 6.0

3

1.1

10

9.5

23

6.3

18 1 1 2 – 2

85.8 – 0.4 1.0 – 6.6

63 7 16 6 31 17

41.8 3.0 2.8 2.5 2.8 8.9

114 6 30 5 21 21

39.2 0.3 2.1 1.3 4.9 3.4

–

–

7

1.2

86

7.4

3

1.8

32

8.9

97

12.5

1

1.6

19

4.2

19

5.6

2 34

0.2 100

1.3 303

4.7 100

78 569

Note: 1. Firms approved under Section 17 (special export provision) of BOI Law. – Zero or negligible. Source: Compiled from official records of the Board of Investment.

6.9 100

208

Multinational enterprises in Asian development

commodities (Snodgrass 1966) and this structure hardly changed during the first three decades of the post-independence period. By 1977, the share of manufacturing (excluding petroleum products) in total exports was only 4 per cent, while the balance came from the traditional ‘primary trio’ (tea, rubber and coconut) and other primary products. Following the 1977 policy reforms, manufacturing exports emerged as the most dynamic element in the export structure. Exports of manufactured goods grew (in current US$ terms) at an annual compound rate of over 30 per cent during 1978–2002, lifting their share in total exports to over 70 per cent. The share of FIEs in total manufacturing exports increased from 24 per cent in 1977 to over 80 per cent in the mid-1990s. The contribution of foreign firms to total increment in manufacturing exports increased from 46 per cent during 1978–85 to over 80 per cent by the turn of the century. FIEs accounted for nearly 80 per cent of the total increment in manufactured exports between 1985 and 2002 accounted for by FIEs, compared with 46 per cent between 1978 and 1985. In 2002, TWMNEs accounted for 46 per cent of total manufacturing exports by FIEs, with firms from Hong Kong and Korea accounting for 24 per cent and 13 per cent respectively.

METHODOLOGY AND DATA We employ multiple regression analysis to examine the relative capital intensity of TWMNEs in relation to DCMNEs and indigenous (local private) firms while controlling for the effect of other firm attributes expected to affect capital intensity. Depending on the availability of data, three industry characteristics (firm size, export orientation, the age of firms) are taken into account. The impact of product composition is taken into account by performing the regression analysis at individual industry level. The complete model is as follows: KL = KL(DCMNE, LPF, SZ, WG, XOUT, AG, AG*DCMNE), where KL SZ DCMNE

= capital intensity (defined as the beginning-of-the-year book value of machinery and equipment per production worker) = firm size (represented by gross output) = a dummy variable set equal to 1 if the firm is a DCMNE affiliate and 0 otherwise

Factor proportions in manufacturing

LPF AG XOUT

209

= a dummy variable set equal to 1 if the firm is a local private firm and 0 otherwise = age of firm = export orientation of firm (measured by the export–output ratio).

In specifying the nationality-of-ownership (parentage) dummies, we treat TWMNE ownership as the base dummy. Therefore, the estimated coefficients of DCMNE and LPF variables are expected to indicate the difference in capital intensity of affiliates of developed-country multinationals and indigenous private firms respectively in relation to TWMNEs. According to the conventional wisdom on TWMNEs (see above), the expected sign of the coefficient of DCMNE is positive. In the literature there is no explicit postulate as to the capital intensity difference between indigenous firms and TWMNE affiliates (that is, as to the sign of the coefficient of the LPF variable in our model). In the only study to test this difference so far (Lecraw 1977), it was found to be positive (and statistically significant). Unfortunately, Lecraw did not discuss the implication of his result. However, drawing upon the existing literature, such a result can be explained as follows. Local firms usually have to choose their equipment from the currently available range of alternatives in a market dominated by DCMNEs. In contrast, TWMNEs which establish affiliates can draw on the technological expertise developed over a long period in their home country. They can either design machinery to suit the relatively low-wage situation in the host country, drawing upon their own historical experience (Wells 1983), or utilize second-hand machinery which is now outdated by the factor market situation in the home country but is well suited to the new host country (Wells 1978). We would therefore, expect the sign of the coefficient of TWMNE to be negative. A well-known counter-argument to the ‘inappropriate technology’ allegation against DCMNEs is that, within a given industry, firm size is more likely than parentage to explain differences in capital intensity between DCMNE affiliates and local firms (Biersteker 1978). The rationale behind this argument, derived from the theory of the firm, is that optimum factor proportions and the scale of operation are intimately linked in such a way that the relative importance of labour declines as the scale of production increases.8 If this were so, the sign of the coefficient of SZ should be positive. Export-oriented firms are usually set up with the aim of exploiting the comparative advantage of the host country. In contrast, inward-looking (import-substituting) investments are induced mostly by various product and factor market distortions which usually impart a bias towards higher capital intensity. Therefore, we can expect production techniques used by

210

Multinational enterprises in Asian development

firms engaged in manufacturing for export to be more labour intensive than those used by firms which are domestic-market-oriented (Helleiner 1988, Chen 1983, p. 199). Accordingly, the coefficient of the XOUT variable is expected to be negative. There is evidence that firms, in particular affiliates of MNEs, do modify and adapt their technology to local factor market conditions (Pack 1976, Morley and Smith 1977, Kirkpatrick et al. 1984). Indeed, it is quite possible that affiliates of DCMNEs could have undertaken technological adaptation that is essentially time dependent (Watanabe 1981). We take into account the possibility of such adjustment over time by including AGE as an additional explanatory variable, both in its own right and interactively with DCMNE. The sign of this variable is expected to be negative. Data The econometric analysis is undertaken using two separate data sets relating to 1981 and 2002. As already discussed, the year 1981 represents the early stage of MNE involvement in Sri Lankan manufacturing, whereas by 2002 there was significant MNE presence across a wider spectrum of industries. Thus comparison of results for the two time points would provide a firm empirical basis for testing the hypothesis under consideration. The first data set comes from the annual Survey of Manufacturing, 1982, conducted by the Sri Lanka Department of Census and Statistics (reference year 1981). The identification of MNE affiliates for the purpose of data extraction from unpublished returns was done using the list of firms prepared by Lakshman and Athukorala (1985). Initially, data were extracted from unpublished returns to the survey for all firms employing more than ten workers for all four-digit industries which each contained at least one foreign firm. Whenever returns were incomplete (for at least one variable relevant for the study) or irremediably inconsistent, they were not included in the list. The list at this stage contained 132 firms belonging to 16 four-digit industries. However, the number of firms in the three ownership categories was inadequate for statistical analysis in seven four-digit industries. Hence, 31 firms had to be excluded from the analysis. The remaining firms belonged to four industry groups – textiles, wearing apparel, chemicals and metal products. The first two groups were at the four-digit level while the latter two were combinations of four-digit industries.9 The aggregation in the latter two groups was unavoidable given the need for an adequate number of observations in each industry classification. The survey data contained information on the value of total fixed capital as well as that of plant and machinery in fixed capital. In compiling the

Factor proportions in manufacturing

211

data series for the capital intensity variable (KL), we use the latter for the following reasons. Firstly, employment is likely to be more closely related with outlay on plant and machinery than with outlay on land and buildings. Secondly, the literature suggests that DCMNEs tend to spend relatively more on buildings compared with TWMNEs (Wells 1983, p. 27) and indigenous firms (Agarwal 1977). Hence, the use of total fixed assets as the measure of capital intensity may lead to misleading results.10 The second data set is extracted from the administrative records of the Board of Investment relating to export-oriented firms (investment projects approved after 1978 under Section 16 of the BOI Law). There were 993 such firms in commercial operation by the end of 2002. Of these firms, we selected all firms which had been in commercial operation for more than one year and for which complete information was available. There were 597 firms which satisfied both criteria. The BOI does not maintain records on the wage bill of individual firms. Data on the wage rate (WG) were therefore compiled from the Annual Survey of Industries 2001 (SLDCS, 2003) by assigning average annual wage per production worker computed for each industry identified at the three-digit level of the International Standard Industry Classification (ISIC) to firms belonging to that industry. Given that all firms in the sample are export oriented (exports accounting for more than 90 per cent of total output), the export–output ratio (XOUT) does not enter the regression estimated on the basis of this data set.

RESULTS The model specified in the previous section was estimated using ordinary least squares (OLS). Both linear and log-linear functional forms were tested. While broadly similar results were obtained with both functional forms, the linear form generally provided estimates with superior statistical properties for the 1981 data set. The results are reported in Table 9.4. To facilitate interpretation of results, the mean values of the major variables differentiated by parentage are reported in Table 9.5. Note that the equation for the metal product industry does not contain the XOUT variable because none of the firms in this industry had recorded exports. In the equation estimated for the total sample of firms (Equation 5), the coefficient of DCMNE is significant only at the 10 per cent level, providing mild statistical support for the hypothesis that DCMNEs are more capital intensive than TWMNEs. The coefficient of LPF is, however, not statistically different from zero, suggesting that purely local firms are similar to TWMNEs in terms of factor proportions. At the disaggregated level, the

212

Table 9.4 Variables

Multinational enterprises in Asian development

Regression results, 1982 (dependent variable: KL) Textile (1)

Constant

7.95 (0.61) DCMNE
56.63 (2.27)** LPF 0.58 (0.06) XOUT
5.66 (0.81) SZ 0.43 (0.20) WG 2.99 (3.04)*** AG*DCMNE 3.33 (1.94)** R2 F

0.85 21.61***

N

22

Wearing apparel (2)
7.81 (1.78)**
7.81 (1.85)** 1.96 (1.10)
0.06 (0.02) 0.44 (1.05)
0.04 (0.12) 0.45 (0.47) 0.50 5.15*** 26

Chemical (3)

Metal products (4)


26.68
4.91 (2.13)** (0.19) 1.61 28.85 (0.09) (0.81) 16.04
3.45 (1.17) (0.15)
9.26 – (0.52) 0.18
1.93 (1.09) (3.26)***
0.11 1.15 (0.12) (0.52) 0.08
0.58 (0.18) (0.26) 0.02 1.07 37

0.53 4.35*** 16

Total sample (5)
1.91 (0.19) 13.12 (1.37)*
0.39 (0.39) 9.11 (1.40)*
0.02 (1.32)* 1.65 (2.80)*** 0.50 (1.45)* 0.24 4.72*** 100

Notes: 1. Estimated by OLS. t-ratios are given in parentheses under each coefficient. Level of significance (one-tailed test) denoted: * significance at 10 per cent, ** significance at 5 per cent, *** significance at 1 per cent. – Not applicable.

statistical support for higher capital intensity of DCMNEs is much stronger for the textiles and wearing apparel industries (Equations 1 and 2), with the coefficient of DCMNE attaining significance at the 5 per cent and 1 per cent levels respectively. The results for LPF are, however, broadly similar to those of the aggregate equation. For wearing apparel, if anything, there is weak support for the alternative hypothesis that TWMNEs are more capital intensive than indigenous firms. The equation for the chemical industry does not pass the F-test for overall significance. This suggests that the variables included in the model are not capable of explaining a significant part of the variation in capital intensity in this industry. This may be because all firms in this group use similar (developed-country) technology without adaptation. A study on technology licensing and patent rights in Sri Lankan manufacturing

213

Factor proportions in manufacturing

Table 9.5

Firm characteristics by parentage (mean values), 1982

Variable

Parentage Number of firms

1. Textiles

DCMNE TWMNE LPRV DCMNE TWMNE LPRV DCMNE TWMNE LPRV DCMNE TWMNE LPRV DCMNE TWMNE LPRV

2. Wearing apparel 3. Chemical products 4. Metal products 5. Total sample

5 6 11 8 8 10 15 6 16 6 5 5 34 25 42

Average output (Rs’000) (SZ)

Export Wage per Capital per output employee worker (Rs) ratio (%) (Rs) (KL) (XOUT) (WG)

24 375 29 253 3 467 24 122 28 324 22 500 26 861 7706 4892 39 776 14 228 1583 28 130 20 750 8317

5.7 35.5 8.5 8.0 91.0 85.5 4.0 2.6 9.7 – – – 28.2 37.4 8.7

8883 5285 6329 8500 7461 6118 18 421 8422 7466 14 882 9876 6325 14 059 7638 6711

56 059 3653 3435 11 775 6488 3850 25 264 24 068 16 198 40 236 13 843 9456 29 260 11 498 9113

Notes: DCMNE Developed–country multinational enterprise TWMNE Third-world multinational enterprise LPRV Local private firm. Source: Compiled using firm-level data (unpublished) from The Survey of Manufacturing Industry – 1981, Department of Census and Statistics, Sri Lanka.

(ESCAP/UNCTC1984) showed that in this sector there was a widespread developed-country management presence through licensing agreements which extended to all firms irrespective of ownership. Our division of firms on the basis of FDI does not capture this element of foreign technology participation. The equations for the metal product group are significant at the 5 per cent level in terms of the standard F-test. However, the only statistically significant explanatory variable is firm size, which is significant at the 1 per cent level in both equations. It is interesting to note that the large differences in the mean levels of capital intensity between DCMNEs and the other two groups observed in Table 9.5 cannot be attributed to parentage; rather, they appear to be related to firm size. Among other explanatory variables, the coefficient of XOUT is insignificant and the sign is not consistent when the analysis is conducted at the

214

Multinational enterprises in Asian development

disaggregated level of industry groups. However, in the equation for the total sample it has the expected sign and one of its coefficients is significant at the 5 per cent level. This suggests that the degree of export orientation may be relevant for explaining inter-industry variation of capital intensity of production but not intra-variation. The wage rate variable is significant (with the expected sign) only in one industry (textiles). Note that the same is true for the interaction variable of age with DCMNE ownership. In sum, there is no consistency among the four industry groups with regard to the relevance of the four firm characteristics considered in explaining interfirm differences in capital intensity. Regression estimates based on the 2002 data are reported in Table 9.6, with the summary data on the variables given in Table 9.7. The result for the DCMNE variable is broadly similar to that based on the 1981 data. The coefficient on DCMNE is positive but not statistically significant in the equation for all firms (Equation 8). As in 1981, it attains significance with the positive sign in the equation for textiles and garments (Equation 2).11 Unlike in the estimates for 1981, it is positive and significant in the equation for chemical products (Equation 4). This contrasting result is consistent with compositional shifts in this product group between the two years. In the early 1980s this product group was dominated by more capitalintensive chemical products, mostly destined for the domestic market. By contrast, in 2002 it was dominated by plastic goods (mostly toys) and rubber products, in which developing-country MNEs presumably have greater competitive advantage in international production (Athukorala and Rajapatirana 2000a). Unlike in the results for 1981, here we have some statistical support for the hypothesis that production processes of purely local firms are more labour intensive than those of TWMNEs. The coefficient of LPF is statistically significant with the negative sign in the overall equation. It also carries the negative sign with t-ratios greater than 1 in all but one of the equations for individual products. The difference in findings is presumably consistent with the nature of market orientation of the sample of firms and the prevailing labour market rigidities in Sri Lanka. The 2002 data relate only to export-oriented firms, whose profitability depends crucially on the ability to exploit the relative labour cost advantage of being located in a low-wage country. However, given the stringent labour legislation and other rigidities in formal labour market practices in Sri Lanka, presumably purely local firms are better placed to do so than their foreign counterparts. Among the other variables, AG has turned out to be much more important in explaining inter-firm variations of factor proportions than in 1982. The coefficient of this variable is statistically significant at the 5 per cent

215


5.95 (1.49)* 0.45 (0.97) 0.48 (1.23)*
0.10 (1.40)*
0.10 (0.29) 0.30 (2.45)**
0.01 (1.58)* 0.16 2.51** 1.23 0.12 1.01 49

Constant


7.81 (1.46)*
0.29 (1.71)** 0.19 (1.04)
1.92 (4.48) 1.31 (1.04) 0.34 (4.82)**
0.01 (3.55)*** 0.20 11.24*** 1.8 2.78 11.43 251

Textiles and wearing apparel (2) 3.24 (0.22)
0.25 (0.31) 0.84 (1.01) 0.24 (1.57)**
0.77 (0.56)
0.43 (0.71) 0.07 (1.03) 0.12 1.60 0.15 0.17 0.47 25

Wood and paper products (3) 6.45 (0.77)
0.65 (1.79)** 0.01 (0.02)
0.07 (0.87)
1.08 (1.53)* 0.24 (2.12)**
0.01 (1.32)* 0.12 2.93*** 2.45 0.67 0.18 85

Chemical and plastic products (4) 2.03 (0.11) 0.01 (0.15) 0.59 (0.46) 0.11 (0.62)
0.75 (0.50) 0.15 (0.71)
0.003 (0.34) 0.03 1.22 0.05 0.01 0.26 39

Non-metallic mineral products (5)
10.00 (1.03)
0.22 (0.48) 0.57 (1.01)
0.14 (1.35)* 0.37 (0.43)
0.04 (0.12) 0.002 (0.06) 0.02 0.85 0.03 9.45 2.01 41

Machinery and transport equipment (6)


4.58 (1.39)*
0.07 (0.90) 0.22 (2.76)***
0.07 (2.55)**
0.19 (0.67) 0.25 (6.27)***
0.01 (3.72)*** 0.28 18.79*** 0.28 8.69 1.97 597

(8)

(7)
6.79 (18.76)***
0.16 (0.59) 0.12 (1.31)* 0.13 (2.32)** 0.08 (0.17) 0.21 (2.19)** 0.01 (1.01) 0.21 6.68*** 0.50 0.16 0.46 107

All firms

Other manufacturing

Test statistics: RESET Ramsey test for functional form mis-specification, JBN Jarque–Bera test for the normality of residuals, HET Breusch–Pagan test for heteroscedasticity.

Notes: 1. Estimated by OLS, with KL, SZ and WG expressed in natural logarithms. T-ratios are given in parentheses under each coefficient. Level of significance (one-tailed test) denoted: * significance at 10 per cent; ** significance at 5 per cent, *** significance at 1 per cent. For Equation 2, t-ratios are based on White-corrected standard errors because of the presence of heteroscedasticity.

R2 F RESET, 2 JBN, 2 BPHET, 2 N

AG2

AG

WG

SZ

LPF

DCMNE

Food, beverages and tobacco (1)

Regression results, 2002 (dependent variable: KL)

Variables

Table 9.6

216

Table 9.7

Multinational enterprises in Asian development

Firm characteristics by parentage (mean values), 20021

Variable

1. Food, beverages and tobacco

2. Textiles and wearing apparel

3. Wood and paper products

4. Chemical products

5. Non-metallic mineral products

6. Fabricated metal products, machinery and transport equipment 7. Other manufacturing

8. All industries

Parentage

DCMNE TWMNE LPRV TOTAL DCMNE TWMNE LPRV TOTAL DCMNE TWMNE LPRV TOTAL DCMNE TWMNE LPRV TOTAL DCMNE TWMNE LPRV TOTAL DCMNE TWMNE LPRV TOTAL DCMNE TWMNE LPRV TOTAL DCMNE TWMNE LPRV TOTAL2

Number of firms 13 8 27 48 53 81 117 251 9 10 7 26 29 32 24 85 17 8 4 29 16 18 8 42 51 30 26 107 195 189 213 597

Average output (Rs m) (SZ) 359 32 98 161 618 622 197 435 617 113 17 262 455 320 161 269 276 90 31 191 259 286 61 232 262 106 56 168 872 469 291 388

Wage per employee2 (Rs)

Capital per worker (Rs ’000) (KL)

– – – 88 030 – – – 59 310 – – – 59335 – – – 116 936 – – – 117 847 – – – 79 048 – – – 68 704 – – – 76 734

519 1527 698 786 880 309 170 369 1945 1600 434 1370 1928 1001 429 1051 530 1326 5328 1399 1050 1256 684 1068 781 873 331 703 1680 706 413 774

Notes: 1. The data relate to export-oriented firms (firms which export more than 90 per cent of total output). 2. Average annual wage computed as the total wage bill for production workers divided by the number of workers. – Date not available.

level or better, with the negative sign in the overall equation as well as four of the other six product-level equations. In each case the quadratic term of AG also enters the equation with a positive and statistically significant coefficient. These results provide strong support for the hypothesis that the

Factor proportions in manufacturing

217

degree of labour intensity of production increases over time, though at a diminishing rate.12 The results for SZ and WG are rather mixed.

CONCLUSION DCMNEs generally exhibit a higher degree of capital intensity than TWMNEs and indigenous private firms in terms of average capital per worker. However, such a simple comparison can be misleading, as firm attributes other than parentage can affect capital intensity. Our analysis of data from the Sri Lankan manufacturing sector which controlled for the effect of such other factors highlights the fact that the link between parentage and capital intensity is industry specific. There were marked differences in capital intensity between DCMNEs and TWMNEs in the textile and wearing apparel industries. However, no such differences could be observed in the chemical and metal product industries in 1981 or in the food and beverages, wood and paper products, nonmetallic mineral and transport equipment industries in 2002. Some significant differences between TWMNEs and indigenous private firms were found only in 2002. All in all, our results support the view that developing-country MNEs may adopt more appropriate technology only in those industries where the range of technological possibilities is wide enough to enable significantly different techniques of production to be utilized (Lall 1983a). The textiles and wearing apparel industries are classic examples of such industries. It is significant that in these industries even the firms with developedcountry parentage tended to adopt more labour-intensive techniques over time. Our results also show that the degree of export orientation is more relevant in explaining inter-industry rather than intra-industry differentials in capital intensity. As expected, labour intensity is higher in those exportoriented industries in which labour-rich developing countries tend to have greater comparative advantage. Thus our results suggest that the inter-industry variations in the degree of technological flexibility are an important determinant of the capacity of developing-country multinational affiliates to adopt more appropriate techniques than developed-country multinationals. To the extent that they locate within industries where most developing countries typically have a comparative advantage, they can make a greater contribution to employment and economic growth than developed-country multinationals. However, over time these differences tend to diminish.

218

Multinational enterprises in Asian development

NOTES * 1. 2. 3. 4. 5. 6.

7. 8. 9.

10.

11. 12.

An expanded/updated version of Athukorala and Jayasuriya (1988). By 1980 the share of developing-country investors in the stock of foreign direct investment in developing countries was between 4 and 8 per cent (Jenkins, 1986, p. 458). Lall (1983a) and Wells (1983) synthesize much of the literature spread until the early 1980s. For a comprehensive compilation of the key papers, see Yeung 1999. For comprehensive surveys of this literature, see Balasubramanyan (1983) and Kirkpatrick et al. (1984, pp. 104–16). For a useful summing up of the methodology and findings of these studies, see Wells (1983, Chapter 2). Fernando (1971) provides a detailed and interesting account of this early stage of MNE involvement in Sri Lankan manufacturing. The first EPZ, at Katunakaye near the Colombo International Airport (henceforth KEPZ), was opened in 1978. The remarkable success of the KEPZ paved the way for setting up a second EPZ in Biyagama (BEPZ) in 1982 and a third in Koggala (KGEPZ) in June 1991. The mild reversal in the declining trend since 2000, mostly reflecting increased inflows, related to the privatization program. The assumption here is that the production function of the firm is of the non-homothetic type. This is valid for industries where economies of scale are important. The four commodity groups (with relevant ISIC number in parentheses) are: (i) Textiles (3211) (ii) Wearing apparel (3220) (iii) Chemicals – paints and varnish (3521), drugs and medicines (3522), soap and cosmetics (3523) and miscellaneous chemicals (3529) (iv) Metal products – non-electrical machinery (3829), electrical machinery (3839) and hand-tools and general hardware (3811). The use of the book value of fixed assets to construct the KL series is subject to the problem that firms have different accounting practices in the valuation of assets. There is also the well-known ‘vintage’ issue (the difficulties of comparing machines of different ages). In the absence of data on fixed investment by age, it is not possible to overcome these limitations. All previous studies in this area have encountered these problems. For details on various methods of measuring capital intensity and related issues, see Bhalla (1975). It was not possible to estimate a separate equation for textiles because there were only 11 firms belonging to that industry. Both an intercept and a slope dummy for these firms were tested in estimating the equation but were found to be statistically insignificant. Note that we were not able to retain AG*DCMNEs in the final estimates because of its high correlation with AG. This variable deletion was amply supported by the standard F-test for variable deletion.

10. Foreign direct investment in economic transition: the experience of Vietnam Attracting FDI has been a key focus of market-oriented policy forms in transitional economies, countries which have embarked on a process of systemic transformation from central planning to market orientation. It is generally believed that FDI can play a ‘special’ role in economic transition in these countries as a catalyst for revitalizing the private sector which remained dormant under the command economy era. Yet there is a dearth of systematic comparative analysis of FDI flows to these countries and their developmental impact. The limited literature so far has predominantly focused on the experiences of China and a few major economies in Eastern Europe.1 This chapter is an attempt to fill this gap through a case study of the Vietnamese experience. It surveys the evolution of FDI policy in Vietnam in the context of overall policy reforms and the current state of the investment climate, examines the experience of attracting FDI from a comparative regional and global perspective, with a view to informing the debate on how to reform policy regimes for better harnessing FDI in the context of an overall strategy of economic transition. A key theme running through the analysis is that both the rate of FDI involvement in the economy and the national developmental gains from FDI depend crucially on the level of economic transition as reflected in the extent of privatization/restructuring of state-owned enterprises, market-based decision making and the creation of a legal and institutional framework for foreign and domestic private investment (Lankes and Venables 1996, McMillan and Woodruff 2002). The chapter begins with an overview of FDI policy and the investment environment in Vietnam. Trends and patterns of FDI are then examined, focusing in turn on trends in total inflows against the backdrop of regional and global trends and source-country and sectoral/industry composition. The next section examines the development impact of FDI, with emphasis on the implications of the incomplete reform agenda and policy inconsistencies cropping up in the reform process in determining the actual outcome in contrast to initial expectations. The key findings and policy recommendations are presented in the concluding section. 219

220

Multinational enterprises in Asian development

INVESTMENT CLIMATE As already noted (Chapter 2), the term ‘investment climate’ covers both the foreign investment regime (rules governing foreign investment and specific incentives for investors) and the general investment environment, which encompasses various considerations impinging on investment decisions such as political stability, macroeconomic environment and attitudes of host countries towards foreign enterprise participation. In this section we survey the evolution of FDI regimes under market-oriented reforms and then provide a comparative assessment of the current state of the overall investment climate. FDI Policy Significant opening of the economy to FDI was an important element of Vietnam’s ‘renovation’ (doi moi) reforms initiated in 1986. Since then the FDI regime in Vietnam has gone through several changes in the post-war period as an integral part of the ongoing policy transition towards a market-oriented economy. In general, policies have become more liberal and the sectors open to foreign investment have expanded over the past two decades. However, much remains to be done to deepen the reforms and improve the business environment in order to make the economies more attractive to foreign investors and to enhance national gains from their participation in the national economy. The first law on FDI was passed by the Vietnamese National Assembly on 29 December 1987. The law specified three modes of foreign investor participation, namely (i) business cooperation contracts (BCCs), (ii) joint ventures and (iii) fully foreign-owned ventures. Foreign participation in the fields of oil exploration and communications was strictly limited to BCCs. In some sectors, such as transportation, port construction, airport terminals, forestry plantation, tourism, cultural activities, and production of explosives, joint ventures with domestic state-owned enterprises (SOEs) were specified as the mode of foreign entry. Fully foreign-owned ventures were to be allowed only under especial considerations governing policy priorities of domestic industrial development. The duration of foreign ownership of approved projects was limited to a maximum of 20 years, apart from in exceptional circumstances. The government provided a constitutional guarantee against nationalization of foreign affiliates and revoking ownership rights of enterprises. The incentives offered to foreign investors included exemption from corporate tax for a period of two years, commencing from the first profit-making year, followed by a preferential corporate tax rate between 15 and 25 per cent in priority sectors (as against

Foreign direct investment in economic transition

221

the standard rate of 32 per cent). Foreign investors were permitted to repatriate after-tax earnings subject to a 10 per cent withholding tax. Overseas remittance of payments for the provision of technology services and repayment of principal and interests on loans were freely allowed. The specific emphasis on joint ventures with SOEs as the prime mode of foreign entry reflected the government’s choice to rely on FDI as a vehicle for industrial transition while ensuring state dominance in the economy. However, in 1990 the foreign investment law was amended to permit existing joint ventures to form new joint ventures with other foreign partners. In 1991 legislation was passed to permit setting up export processing zones (EPZs) which offered especial incentives to firms involved in the production of goods for export and the provision of services for the production of export goods. In 1992 the duration of foreign participation was extended from 20 years to 50 years, and 70 years in special cases. A new Law on Foreign Investment, enacted in 1996, permitted private enterprises to enter into joint ventures with foreign investors and procedures for the approval of investment projects were streamlined. Under this law, authority to issue licences for projects, up to a specified size, was delegated to local governments. The tax holiday for investment in priority sectors was extended to 8 years, with a beyond-tax holiday tax of 10 per cent. A three-tier withholding tax of 5, 7 and 10 per cent, based on the ‘priority status’ of investment, was introduced in place of the original flat rate of 10 per cent. These revisions to the foreign investment law, however, coincided with a growing resentment within Communist Party circles. This resentment, which was fuelled by the massive influx of FDI following the initial phase of reforms, resulted in a number of restrictive policy measures which raised serious concerns in the international investment community about Vietnam’s attempts to project its image as a new investment centre. These included a proposal to establish liaison offices of the Communist Party in all foreign ventures, doubling of commercial and residential rents for foreign enterprises and expatriate staff, imposition of a maximum time limit of three years on work permits issued to foreigners employed in FDI projects, restrictions on foreign capital participation in labour-intensive industries, and imposition of domestic content requirements and export performance requirements on foreign-invested enterprises (FIEs) in a number of key industries. There is also evidence that the foreign investment approval procedure turned out to be more selective, with greater emphasis placed on promoting investment in key high-tech industries such as metallurgy, basic chemicals, machinery, pharmaceuticals, fertilizers, electronics and motor vehicles. Notwithstanding the new (1996) legislation permitting domestic private enterprises to enter into joint ventures with foreign firms, joint

222

Multinational enterprises in Asian development

ventures with SOEs continued to receive powerful support in senior policy circles as the prime mode of FDI entry (Truong and Gates 1996). Policy reforms following the economic downturn during 1997–99 have placed greater emphasis on FDI promotion. Under an amendment to the FDI law on 9 June 2000, FIEs and parties to business cooperation contracts (BCCs) were given freedom to change the mode of investment and to split, merge and consolidate enterprises. (Recently there have been several cases of joint ventures being converted into 100 per cent owned FIEs.) The three-tier withholding tax on profit transfers was reduced to 3, 5 and 7 per cent. The approval procedure of new investment proposals was streamlined, with automatic registration of export-oriented FIEs. Foreign investors were allowed to implement ‘less sensitive’ projects (that is, those deemed to have no implications for national defence, cultural and historical heritage or the natural environment) without licensing scrutiny of the Ministry of Planning, provided they are export oriented. Vietnamese permanently living overseas who invested in Vietnam were liable to pay a withholding tax of 3 per cent. FIEs were allowed (under special permission from the State Bank of Vietnam (SBV)) to open accounts with overseas banks and to mortgage assets attached to land and land-use rights as security for borrowing from credit institutions permitted to operate in Vietnam. Local authorities, notably the Ho Chi Minh City council, were given greater autonomy to improve the investment climate and ease administrative hurdles for foreigninvested projects. The implementation of a new Enterprise Law in 2000, which permits greater participation of domestic private enterprises in the economy, significantly contributed to improving investor confidence in the reform process. In April 2003, 100 per cent foreign-owned companies were allowed to become shareholding companies (that is, they were permitted to establish joint ventures). The withholding tax on profit remittance was abolished with effect from April 2004. In November 2005 a new unified Investment Law was promulgated to replace the Law on Foreign Investment and the Law on Domestic Investment. Key features of this landmark legislation (which took effect in July 2006) include treating foreign and domestic investors equally with regard to institutional and legal procedures for approval and monitoring and the incentives offered, providing investors with complete freedom in the choice of the particular mode of business entry (that is joint venture or full ownership), abolishing local context and export performance requirements, and introducing a decentralized three-tier system of investment approval. Under the new investment approval procedures, projects under US$1 million require only business registration (that is, no requirement for investment approval), projects between US$1 and 20 million are approved at the provincial level, and only projects beyond this

Foreign direct investment in economic transition

223

investment level require the approval of the Ministry of Planning and Investment (MPI). The FDI regime in Vietnam has certainly become more liberal and investor friendly since about 2000. In spite of these reforms, considerable restrictions still remain (even after the new investment law became operational in July 2006) compared with more open policy regimes in neighbouring Southeast Asian countries. For instance, the foreign partner is required to contribute 30 per cent of registered equity capital of a joint venture (which may be lowered to 20 per cent, depending on the field of operation, technology, market, business efficiency, and socio-economic benefit of the particular project). BCC remains the only permitted mode of foreign entry into the oil exploration and telecommunication sectors. Only joint ventures or BCCs are allowed in air transportation and airport construction, industrial explosive production, forestry, culture and tourism. Vietnam’s current business legislation is not conducive to cross-border merger and acquisition (M&A) activity. Foreign investors are permitted to acquire only up to 49 per cent of shares in local companies (both listed and unlisted) in some designated business sectors. This requires the approval of the Prime Minister. This restrictive approach to M&A is presumably a major constraint on the expansion of FDI inflows to Vietnam, because cross-border M&A has been an increasingly important mode of FDI inflows in recent years. Business Environment After one-and-a-half decades of policy reforms, how do international investors rate Vietnam among other countries as an alternative host? Have recent attempts to reform the FDI regimes and streamline the investment approval procedures brought about anticipated results? What are the investors’ key concerns about the investment environment? There is no unique way to answer these and related questions, but information summarized in Tables 10.1–10.3 provides some useful insights. The World Economic Forum in its Global Competitiveness Report ranks countries in terms of two composite indices measuring long-term growth prospects and business competitiveness (Table 10.1). In 2005, among 117 countries covered Vietnam ranked well below all other high-performing countries in the region. The Economic Freedom Index measures overall quality of the institutional and policy setting for private sector growth. On this index, Vietnam has continued to be at the bottom of the ranking (Table 10.2). The Doingbusiness database of the World Bank Group ranks countries in terms of ten indicators of business environment of almost all countries

224

Multinational enterprises in Asian development

Table 10.1 World Economic Forum growth competitiveness and macroeconomic environment indices, 2005 (ranking among 117 countries) Growth competitiveness1 Singapore Hong Kong, SAR Taiwan Malaysia Korea, Rep Thailand China India Vietnam Indonesia Pakistan Philippines Bangladesh Sri Lanka Highest Lowest

6 28 5 24 17 36 49 50 81 74 83 77 110 98 (1) Finland (117) Chad

Business competitiveness2 1 8 17 19 25 26 33 50 60 64 69 71 83 94 (1) Singapore (117) Zimbabwe

Notes: 1. Based on an index of long-term growth prospects which encompasses the quality of the macroeconomic environment, public institutions and domestic technology. 2. Business competitiveness ranking based mostly on micro aspects, with emphasis on the quality of a country’s business environment. Source: World Economic Forum, Global Competitiveness Report 2005–06, September 2005, Geneva: World Economic Forum.

in the world and provides detailed information on the characteristics of individual countries relating to each indicator. Information from this database is summarized in Table 10.3 for Vietnam, together with the average figures for all countries in the East Asian and Pacific region. Overall, Vietnam ranks poorly among the countries in the region (and globally). In particular, it performs very poorly in terms of labour market flexibility (hiring and firing), credit availability, registering property, investor protection, ease of trading across borders and enforcing contracts. There is clear evidence that recent attempts to streamline and expedite investment approval procedures have not yet delivered expected results; the complex bureaucratic requirements and procedures for getting a project approved and implemented, which often take several years to complete, remain a major obstacle to investors. Finally, a number of recent firm-level surveys

225

Foreign direct investment in economic transition

Table 10.2 Index of Economic Freedom1 (ranking among all countries covered) 1995

1977

2000

2005

1 2 12 17 19 20 42 55 54 84 90

1 2 12 37 50 49 56 65 105 104 125 120 109 133 152 161

1 2 27 38 68 71 92 98 110 111 121 125 134 141 142 155

161

157

Hong Kong Singapore Taiwan Korea, Rep Malaysia Thailand Sri Lanka Philippines Pakistan China, PR India Nepal Indonesia Bangladesh Vietnam Korea, D. Rep

68 85 98 101

1 2 15 25 37 35 38 67 81 111 122 123 67 114 141 150

Country coverage

101

150

–

Notes: 1. An index of factors that most influence the institutional setting of economic growth. Based on 50 variables grouped into categories: trade policy, fiscal burden of government, monetary policy, capital flows and foreign investment, banking and finance, wages and prices, property rights, regulations, and informal market activity. – Not covered in the given year. Source: The Heritage Foundation, Index of Economic Freedom database (www.heritage.org/research/features/index).

have identified many constraints on doing business in Vietnam (Le et al. 2002, World Bank 2002). Corruption is rated as by far the biggest hurdle. Poor infrastructure and the high cost of electricity due to lack of generating capacity also increase the cost of doing business.

TRENDS AND PATTERNS OF FDI There are two alternative data series on gross FDI inflows to Vietnam, one compiled by the Ministry of Planning and Investment (MPI) from administrative records relating to realized investment projects and the other coming from the balance-of-payments accounts of the State Bank of Vietnam

226

Multinational enterprises in Asian development

Table 10.3 Indicators of ease of doing business: Vietnam in global and regional context, 2005 (rank among 155 countries) Criteria of ease of doing business 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Starting a business Dealing with licences Hiring and firing Registering property Getting credit Protecting investors Paying taxes Trading across borders Enforcing contracts Closing business Overall

Details on each criterion 1. Starting a business Procedures (number) Time (days) Cost (% of income per capita) 2. Dealing with licences Procedures (number) Time (days) Cost (% of income per capita) 3. Hiring and firing workers Difficulty of hiring index Rigidity of hours index Rigidity of firing index Rigidity of employment index Hiring cost (% of salary) Firing cost (weeks of wages) 4. Registering property Procedures (number) Time (days) Cost (% of property value) 5. Getting credit Legal rights index1 Credit information index2 Public registry coverage (% adults) Private bureau coverage (% adults) 6. Protecting investors3 Disclosure index

Vietnam

Regional average (East Asia and Pacific)

82 18 122 39 106 143 107 83 102 95 99

11 50 50.6

8.2 52.6 42.9

14 143 64.1

18.0 157.7 137.4

44 40 70 51 17.0 98.0

26.0 29.6 23.0 26.2 8.8 44.2

5 67 1.2

4.6 62.2 5.0

3 3 1.1 0.0

5.3 1.8 1.7 9.6

4

5.6

227

Foreign direct investment in economic transition

Table 10.3

(continued)

Criteria of ease of doing business

7.

8.

9.

10.

Director liability index Shareholder suits index Investor protection index Paying taxes Payments (number) Time (hours) Total tax payable (% of gross profit) Trading across borders Documents for export (number) Signatures for export (number) Time for export (days) Documents for import (number) Signatures for import (number) Time for import (days) Enforcing contracts Procedures (number) Time (days) Cost (% of debt) Closing business Time (years) Cost (% of estate) Recovery rate (cents in the dollar)

Vietnam

Regional average (East Asia and Pacific)

1 2 2.3

4.2 6.2 5.3

44 1050 31.5

28.2 249.9 31.2

6 12 35 9 15 36

7.1 7.2 25.8 10.3 9.0 28.6

37 343 30.1

30.0 406.8 61.7

5.0 14 19.3

3.4 28.8 24.0

Notes: 1. The legal rights index ranges from 0 to 10 with higher scores indicating that those laws are better designed to expand access to credit. 2. The credit information index ranges from 0 to 6 with higher scores indicating better scope, access and quality of credit available through public registries or private bureaus. 3. The indicators below describes three dimensions of investor protection: transparency of transactions (disclosure index), liability for self-dealing (director liability index), and shareholders’ ability to sue directors for misconduct (shareholder suits index), and strength of investor protection index. Each index varies between 0 and 10, with higher values indicating better performance. The investor protection index is the average of the three sub-indices. Source: Doingbusiness database, The World Bank Group, http://www.doingbusiness.org/ExploreEconomies.

(SBV), the central bank, (Figure 10.1). The former series covers capital contributions by both foreign and local partners, while the latter by definition covers only the actual foreign capital inflows resulting from the establishment of foreign-invested enterprises in the country. Apart from this fundamental

228

Multinational enterprises in Asian development 3500

US$ million

3000

FDI1

2500 2000 1500 FDI2

1000 500

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

0

FDI1: total realized investment in licensed foreign-invested project FDI2: gross FDI inflow based on balance of payments records

Note: * FDI1 includes capital contribution by local firms in joint ventures. Sources: FDI1: data provided by Ministry of Planning and Industry, Hanoi. FDI2: UNCTAD, World Investment Report, Geneva (various issues).

Figure 10.1

Foreign direct investment in Vietnam,* 1990–2005

difference relating to the coverage, the two series suffer from their own measurement problems. For instance, the MPI series simply captures approved investment in implemented projects and no adjustment is made for a possible difference between investment commitment and actual investment. The balance-of-payments data on FDI are compiled by the SBV from quarterly and semi-annual survey reports received from foreign-invested enterprises, supplemented by reports from the SBV’s provincial branches. However, problems persist with the survey response rate: not all FIEs provide the requested information and the response rate varies from year to year. Moreover, so far no effort has been made by the SBV to distinguish between intra-firm borrowing and other non-resident liabilities in estimating FDI (IMF 2006). The inclusion of the latter introduces an upward bias into the estimates. Given these differences relating both to coverage and to the compilation procedures, the two series are not strictly comparable. But they are broadly similar in terms of the overall trends. Data reported in Table 2.1 (in Chapter 2) help us to examine Vietnam’s relative performance as a host to FDI in a global context. According to the standard FDI data based on balance-of-payments records of individual countries reported in this table, foreign investors’ response to economic opening in Vietnam was swift and notable. Annual gross FDI inflows to Vietnam surged from almost zero in the first half of the 1980s to an annual average of US$780 million in 1990–95 and to US$2587 million in 1997. FDI

Foreign direct investment in economic transition

229

amounted to over a third of total domestic capital formation and nearly 10 per cent of GDP during 1995–97. From 1997 there was a precipitous fall, bottoming at US$ 1200 million in 2002. Since then there has been a notable recovery, reaching US$1610 million in 2004. An inspection of official investment approval records suggests that this trend would continue well into 2005 (and beyond); total registered investment in realized FDI projects increased persistently from US$19 billion in 2001 to US$29 billion in 2005. The surge of FDI in the aftermath of the policy shift from ‘plan to market’ was a common pattern observed across almost all other transition economies (Huang 2003, Pomfret 1991, Lankes and Venables 1996, Lankes and Stern 1997). Significant initial reforms and the general media-propelled euphoria about opening of a ‘new investment frontier’ naturally heightened investor interest in becoming the first to exploit new investment opportunities. Moreover, in the immediate aftermath of economic opening, there were many quick-return as well as low-risk long-term investment opportunities to be grabbed in infrastructure development and provision of utilities (power, telecommunication and so on) and resource extraction (oil exploration, for example). Massive injection of funds by international developmental agencies, such as the Asian Development Bank (ADB), into infrastructure and energy projects provided an added impetus for investment in related areas. Once these initial stimuli dissipated, the sustainability of the investment surge depended very much on the ability of the government to deliver the promised reforms and the ‘natural’ attractiveness of the country as an investment location. The onset of the East Asian financial crisis in mid-1997 acted as an additional factor in the cessation of the post-reform surge in FDI in Vietnam. Investors from East Asian ‘miracle economies’ – in particular Malaysia, South Korea and Singapore – played a key role in the investment surge on the back of the boom in their economies in the lead-up to the crisis. These substantial intra-Southeast Asian FDI flows were severely disrupted by the onset of the financial crisis in mid-1997 (see below). In addition to this direct effect, the financial crisis presumably had a damaging impact (at least in the short to medium term) on the investor bullishness about East Asia as a favoured investment location in general. However, one should not overstate the role of the East Asian crisis in the cessation of the post-reform FDI boom. A close look at investment approval data in Vietnam suggests that investor interest in that country began to decline from about mid1996, following the failure of the sixth Communist Party Congress to deliver anticipated further reforms and the onset of a political backlash against foreign firms on the basis of their perceived adverse socio-economic implications (World Bank 2002, Kokko 1997). There was also a notable increase in the failure rate of licensed FDI projects (that is, the percentage

230

Multinational enterprises in Asian development

of withdrawn projects out of total licensed projects) in the second half of the 1990s compared with the early post-reform years (Kokko et al. 2003). In an overall international comparison, Vietnam has continued to be a small player in the global investment scene. Even during the investment boom of 1992–95, FDI in Vietnam amounted to a mere 1.2 per cent of total FDI flows to developing countries. During 2000–05, this figure declined to 0.6 per cent. In 2004, for the first time FDI inflows to Vietnam (US$1610 million) exceeded those to Thailand (US$1064 million). However, it is premature to read too much meaning into this rank reversal because foreign investors in Thailand are still in a process of adjusting to massive excessive investment made in that country during the pre-crisis economic boom. Ownership Structure A well-known feature of foreign investment in Vietnam highlighted in the early discussion on FDI is the dominance of joint ventures. During 1988–94 joint ventures accounted for over 70 per cent of total approved projects and 75 per cent of total registered capital. The bulk of these joint ventures (over 90 per cent) had state-owned enterprises as the local partners (Kokko 1997, p. 3). Two major concerns were raised regarding SOE dominance. First, to the extent that such investment was in activities made financially attractive by protection and other forms of government patronage, the net national gain would have been much less than the reported figures suggest – and could even have been negative. Second, the process of channelling FDI to SOEs could create strong constituencies against future liberalization reforms. These enterprises were not only large and powerful but also intimately connected with various levels of political decision making and were in a position to use their political influence to oppose trade liberalization or other reforms that might reduce their privileges (Riedel and Comer 1997, Truong and Gates 1996). From the mid-1990s there has been a significant increase in the share of fully foreign-owned firms in total approved investment (both in terms of the number of projects and value of committed capital) at the expense of the relative position of joint ventures. By 2001 the former form of FIEs accounted for over 80 per cent of total approved projects and 65 per cent of total registered investment. The main underlying factor in this shift in the ownership structure appears to be the more flexible ownership criterion adopted by the Vietnamese authorities in approving export-oriented FDI. According to our analysis of detailed investment approval data provided by the MPI, the relative importance of export-oriented firms and that of fully foreignowned firms among the total FIEs seems to have increased hand in hand from about the mid-1990s.

Foreign direct investment in economic transition

231

Source Country Composition The geographic origin of FDI in Vietnam is characterized by a clear regional bias. During 2000–05, investors from ASEAN, Northeast Asia and China together accounted for over a half of FDI to Vietnam (Table 10.4). This is in sharp contrast to the other Southeast Asian countries (Indonesia, Malaysia, the Philippines, Singapore and Thailand) where the bulk of FDI originates in OECD countries. However, over the years, the relative position of ASEAN countries in the source-country composition has declined as a result of the growing importance of investors from the other East Asian countries and OECD countries. During 2000–05, Northeast Asia and China accounted for 44 per cent of total approved investment in realized projects, with OECD countries and ASEAN countries accounting for 36 and 20 per cent respectively. At the individual country level, the relative position of Singapore, which was the largest host country until the late 1990s, has declined (from 16 per cent during 1988–99 to 12.5 per cent during 2000–05) and that of South Korea and Taiwan has increased (from 9 to 16 per cent, and from 12 to 23 per cent respectively). Investment from China also has increased rapidly, but from a low base, reaching 4 per cent of total investment during 2005. During the early years of market-oriented reforms in Vietnam, analysts often referred to the US economic embargo as a major constraint on the country’s ability to rely on FDI in the process of economic transition. However, interestingly the lifting of the embargo in 1994 and the signing of the Vietnam–USA Free Trade Agreement in 2001 has not ushered in a significant change in the source-country composition of FDI in Vietnam. The share of US investors in total approved investment in realized projects amounted to a mere 1.5 per cent during 2000–05. US FDI in developing countries in the Asian-Pacific region is heavily concentrated in assembly activities in vertically integrated high-tech industries, mostly in electronics (Lipsey 1998). Investors in these product lines place a much greater weight on the stability and transparency of the domestic investment climate than do investors involved in standard export-oriented labour-intensive production (like clothing or footwear) or domestic-marketoriented industries. Reflecting this cautious approach to site selection, the first investment project by a US electronics MNE in Vietnam materialized only in 2006, by which time Vietnam’s commitment to market-oriented reforms and promoting FDI had become firmly rooted.2 Industry Composition In the early stage, extraction of crude petroleum and gas and the construction and services sectors were the major areas of attraction to foreign

232

Multinational enterprises in Asian development

Table 10.4 Vietnam: source-country composition of FDI inflows, 1988–2005 Source country/ region

OECD countries1 Australia Belgium Canada Denmark France Germany Italy Japan Luxembourg Netherlands Norway Sweden Switzerland UK USA Other European transition economies Russia Czech Rep ASEAN Laos Thailand Brunei Indonesia Malaysia Philippines Singapore Northeast Asia and China China Taiwan South Korea Hong Kong Other countries2 Total US$ million

1988–99

2000–05

Number of realized projects

Approved investment (%)

Number of realized projects

Approved investment (%)

848 92 12 34 6 149 35 – 270 – 39 7 9 30 37 108 20 79

30.6 3.0 0.1 0.6 0.1 5.8 0.6 0.0 9.1 0.0 1.6 0.1 1.0 1.7 3.2 3.5 0.3 4.4

741 23 13 20 27 15 36 21 330 15 23 7 2 3 31 157 18 –

36.1 4.8 0.2 0.4 0.9 0.3 0.9 0.4 19.6 5.8 10.2 0.1 0.2 0.4 0.5 1.0 1.5 –

62 5 495 4 126 – 18 80 27 238 1118

4.1 0.1 23.2 0.0 2.9 – 0.9 3.0 0.6 15.8 31.0

– – 309 3 4 15 13 104 3 165 2085

– – 19.8 0.1 2.8 0.2 0.9 3.2 0.0 12.5 43.7

88 458 266 306 260 2800

0.4 12.4 8.5 9.7 10.7 100.0 37 088

269 964 798 54 95 3230

4.4 22.8 15.7 0.8 0.4 100.0 13 930

Notes: 1. OECD Europe, Australia, New Zealand and Japan. 2. Predominantly tax-haven islands. – Zero or negligible. Source: Compiled from data provided by the Ministry of Planning and Investment, Hanoi.

233

Foreign direct investment in economic transition

Table 10.5 Vietnam: sectoral distribution of cumulative approved investment,1 1991, 1995, 2000, 2005 (%) 1991

1995

2000

2005

Primary production Crude oil Agriculture and forestry Manufacturing industry Foodstuff Sea food Textiles, clothing and footwear Other Construction New resident parks New cities Office-building EPZ & IZ infrastructure construction

50.64 45.21 5.43 15.66 3.41 1.77 2.18 8.30

27.93 24.10 3.82 33.66 18.17 10.21 0.52 4.77 3.26

16.36 10.51 5.85 49.01 23.85 14.90 0.74 9.53 4.69

25.74 19.80 5.93 41.93 6.77 0.56 11.23 23.38 16.74 8.31 0.18 6.36 1.88

Service Transportation & telecommunications Hotel–tourism Finance–banking Cultural–health–education Others

20.99 10.12 7.60 2.77 0.03 0.46

26.44 7.10 12.18 4.93 0.03 2.21

18.86 4.67 9.69 2.48 0.15 1.87

15.59 2.65 8.37 2.30 1.02 1.27

100 361

100 6269

100 14 954

100 27 986

Total US$ million

– – – – – –

– – – – –

– – – – –

–

Notes: 1. Figures for a given year show the cumulative approved investment since 1988. The data cover realized projects only. – Data not available. Source: Compiled from data provided by the Ministry of Planning and Investment, Hanoi.

investors, with the manufacturing sector accounting for less than a fifth of registered investment in total approved projects (Table 10.5). The relative importance of manufacturing has, however, increased over the years. By 2005, manufacturing accounted for 42 per cent of cumulative approved investment in realized projects. During the early years, much of FDI investment in manufacturing was in production for the domestic market. During 1988–90, less than 20 per cent of total approved projects had export–output ratios of over 50 per cent. From the late 1990s there has been a notable compositional shift in manufacturing FDI from domesticmarket-oriented to export-oriented production. By 2000, over 70 per cent

234

Multinational enterprises in Asian development

of approved FIEs in manufacturing had export–output ratios of 50 per cent or more, the majority clustering within the 80–100 per cent category (Athukorala 2002b, Table 10, updated to 2004 using the same data sources). Until about the late 1990s, most of the export-oriented FDI projects were in the garment, footwear, and furniture and other wood product industries. Over the past five years foreign investors have begun to enter into assembly activities in the electrical and electronics industries. The decline in FDI during 1998–2002 was largely confined to non-traded sectors (in particular the construction industry) and import-competing (domestic-market-oriented) manufacturing. FDI coming to export-oriented industries (in particular garments, food processing, and assembly activities in the electronics and electrical industries) has continued to increase, though at a slower rate than in the early 1990s. The share of export-oriented projects in total committed legal investment has persistently increased from about 1997. The explanation seems to lie in Vietnam’s strong comparative advantage in labour-intensive production and assembly activities and the fact that export-oriented FIEs are more resilient to adverse developments in the domestic policy scene, provided the trade policy regime assures easy/uninterrupted access to imported inputs. Spatial Distribution Table 10.6 presents data on the spatial distribution of approved investment in operational projects in Vietnam. There has been a heavy concentration of projects in the South East (mainly the Ho Chi Minh City area) and in the Red River Delta (around Hanoi). These two regions accounted for 61 and 28 per cent respectively of the total approved investment during 1988–2005. Ho Chi Minh City alone accounted for over a fifth of this investment. There has not been any notable change in the pattern of spatial distribution over the period 1998 to 2005. Thus there is little evidence that the government’s incentive schemes to encourage foreign investors to move to remote regions have yielded the anticipated outcome. These spatial patterns of FDI location clearly point to the importance of transportation and other infrastructure facilities and access to administrative services in determining investment location.

ECONOMIC IMPACT A systematic comparative analysis of the economic implications of FDI in Vietnam is not possible given the paucity of data. The investment monitoring organization in Vietnam, the Ministry of Planning and Investment,

Foreign direct investment in economic transition

Table 10.6

235

Vietnam: spatial distribution of FDI,1 1988–2005

Mountain North Red River Delta Ha-noi Hai-phong Hai-duong Central coast Thanh-hoa Da-nang Central High Land Dak Nong South East Ho Chi Minh City Dong-nai Baria-Vungtau2 Mekong Delta Long-an Can-tho Total US$ million

1988

1990

1995

2000

2005

0.27 0.49 0.41 0.00 0.00 27.36 0.00 27.28 0.00 0.00 69.54 19.36 0.00 50.18 2.34 0.00 1.39 100.00 367

0.16 16.91 13.79 0.91 0.84 8.28 0.16 6.58 0.20 0.00 73.13 40.44 1.04 31.57 1.32 0.13 0.51 100.00 1583

0.92 27.08 17.66 4.64 0.87 7.38 2.37 3.11 0.66 0.00 60.55 33.85 10.97 12.44 3.40 0.69 0.32 100.00 17 827

0.71 29.27 19.30 3.54 1.28 9.20 1.11 2.04 2.40 0.00 56.03 27.80 8.54 13.58 2.39 0.73 0.39 100.00 38 594

1.57 28.39 18.27 3.99 1.41 5.35 1.40 0.95 0.52 0.02 61.16 23.99 16.65 9.38 3.00 1.50 0.22 100.00 51 018

Notes: 1. Figures for a given year show the cumulative approved investment since 1988. The data cover realized projects only. 2. Includes investment in the petroleum industry. Source: Compiled from data provided by the Ministry of Planning and Investment, Hanoi.

does collect information annually on the operation of approved projects, but there is no system in place to systematically process and publish the data. The Central Statistical Organization has been conducting an annual census of industry based on a well-designed questionnaire. Some of the basic data tabulations are available from CSO publications, but the large bulk of the valuable information gathered in the census remains underutilized. This section is based on whatever data are available from MPI and CSO publications and some fresh tabulations from unpublished data obtained from these two organizations. FDI has undoubtedly made a notable contribution to the process of economic transition in Vietnam (Table 10.7). The share of foreign-invested enterprises in GDP increased persistently from 6.3 per cent in 1995 (the earliest year for which this information is available) to 15 per cent in 2003, and they accounted for over 20 per cent of the total increment in real GDP between

236

Multinational enterprises in Asian development

Table 10.7 Foreign-invested enterprises (FIEs) in the Vietnamese economy: key indicators, 1995–2003 FIEs share in . . .

1995 1996 1997 1998 1999 2000 2001 2002 2003

GDP Gross domestic investment Gross industrial output Fixed capital in industrial enterprises Employment in industrial enterprises (total) Employment in industrial enterprises (female)

6.3 32.3

7.4 28.6

9.1 31.3

10.0 25.0

12.2 18.2

13.3 18.7

13.8 18.4

13.8 18.0

14.5 17.5

25.1

26.7

28.9

32.0

34.7

35.9

35.3

35.4

36.0

–

–

–

–

–

–

35.9

34.1

30.9

–

–

–

–

–

–

11.5

12.4

14.8

–

–

–

–

–

–

16.3

18.6

22.9

Note: – Data not available. Source: Compiled from Central Statistical Office, Statistical Data of Vietnam Socio-Economy, Hanoi: Statistical Publishing House (various issues).

these two years. The share of FIEs in gross industrial production increased from 25 per cent in 1995 to 36 per cent in 2003, accounting for over 30 per cent of total increment in gross industrial output between these two years. This notable contribution of FIEs to expansion in GDP and industrial output seems to have occurred against the backdrop of a persistent decline in the share of FIEs in gross domestic capital formation in the economy, from 32 per cent in 1995 to 18 per cent in 2003. There are two possible explanations. First, it may be that efficiency of factor usage (productivity) in FIEs has improved over time (more on this later in this section). Second, the decline may be a reflection of the notable increase (discussed above) in FIE involvement in export-oriented production. It is not possible to separate the impact of these two factors because of the data constraint. In 2003, FIEs accounted for 15 per cent of total industrial employment (up from 12 per cent in 2001) and 23 per cent of total female employment (up from 16.3 per cent) in the country. The evidence of increase over time in the female share in total FIE employment is consistent with the increased export orientation of FIE production. Further information on the patterns of employment in FIEs in the industrial sector is given in Table 10.8. Total industrial employment in FIEs increased at an annual rate of 23 per cent during the four-year period from 2000 to 2003, compared with 8 per cent growth in employment in non-FIE (purely local) firms. FIEs contributed

237

24

20 21 22 23

16 17 18 19

15

13 14

10 11

VSIC code

Total industry (a) Mining and quarrying Mining of coal and lignite Extraction of crude petroleum and gas Mining of metal ores Other mining and quarrying (b) Manufacturing Food products and beverages Tobacco products Textiles Wearing apparel Footwear and leather products Wood and wood products Paper and paper products Publishing and printing Coke and refined petroleum products Chemicals and chemical products 72

69 44 28 1

13 144 317 373

2024 317

7 62

2253 149 74 7

Employment (’000)1

14.3

14.9 10.7 1.7 48.7

2.8 19.6 19.2 38.5

22.3 12.9

1.4 0.5

19.9 4.4 0.4 88.0

2.3

1.6 1.1 0.1 0.1

0.0 5.4 17.2 30.8

98.6 7.6

0.0 0.1

100.0 1.2 0.0 1.1

14.6

14.2 21.1 40.6 7.0

2.3 11.5 43.3 22.4

23.0 10.9

34.7 18.4

22.7 3.8 19.1 3.8

4.4

13.4 7.5 9.0 9.5

4.3 6.5 9.9 5.7

9.2 8.1

15.8 0.9

8.3 2.1 2.9 20.8

38.1

11.0 29.0 9.1 1.5

2.7 31.3 58.6 75.2

45.6 17.1

2.5 11.2

44.6 11.2 3.8 100.0

35.8

12.9 17.4 20.6 92.6

19.5 14.2 11.7 11.8

15.9 24.2

32.5 16.9

17.0 101.5 19.3 112.0

17.2

7.5 12.5 20.1 15.2

29.6 9.9 10.4 8.3

11.5 10.3

11.5 8.8

12.5 13.9 18.2 38.9

FIE Composition Growth Growth of FIE share in Average Average share in of FIE of FIE employment employment wage of wage of employment1 employment1 employment1 in local increment FIEs (Dong local firms (%) (%) (%) firms1 (%) from 2000 mn)1,2 (Dong mn)1,2 to 2003 (%) (%) (%)

Table 10.8 Vietnam: contribution of foreign-invested enterprises to industrial employment, 2000–03

238

35 36

34

33

32

31

30

27 28 29

26

25

VSIC code

(continued)

Rubber and plastic products Non-metallic mineral products Basic metal products Fabricated metal products Machinery and equipment n.e.c Office, accounting and computing machines Electrical machinery and apparatus n.e.c Radio, television and communication equipment Medical and optical instruments, watches and clocks Motor vehicles, trailers and semi-trailers Other transport equipment Furniture, manufacturing n.e.c

Table 10.8

56 117

22

9

20

50

3

33 65 43

161

68

Employment (’000)1

22.4 34.6

32.5

45.8

49.2

56.8

98.3

8.8 21.0 10.0

8.1

27.0

3.2 8.6

1.5

0.8

2.0

5.3

0.6

0.6 2.7 0.8

2.6

3.6

32.7 31.4

24.2

17.5

15.7

15.9

11.6

15.1 21.4 16.1

16.3

22.3

9.2 19.0

14.5

9.5

1.4

9.5

66.1

5.7 17.3 11.1

11.3

15.4

57.1 50.0

47.3

62.8

98.3

70.2

79.9

22.1 26.2 15.4

11.9

37.0

19.4 12.8

23.9

21.3

22.9

17.7

21.7

37.1 20.5 22.2

28.7

16.5

13.6 10.1

12.9

12.2

17.8

20.8

15.6

18.1 11.4 12.7

12.6

13.9

FIE Composition Growth Growth of FIE share in Average Average share in of FIE of FIE employment employment wage of wage of employment1 employment1 employment1 in local increment FIEs (Dong local firms (%) (%) (%) firms1 (%) from 2000 mn)1,2 (Dong mn)1,2 to 2003 (%) (%) (%)

239

17

63

80

0.5

1.2

1.0

0.0

0.1

0.1

0.8

0.9

0.8

6.6

4.3

4.8

0.0

0.2

0.2

55.7

47.5

48.3

17.9

33.0

29.8

Source: Compiled from unpublished returns to the Annual Manufacturing Census (2000, 2002, 2003, 2004) provided by the Central Statistical Office, Hanoi.

Notes: 1. Period average. 2. At constant (2000) price. n.e.c not elsewhere classified. VSIC Vietnam Standard Industry Classification (based on the International Standard Industry Classification, ISIC).

41

40

(c) Electricity, gas and water supply Electricity, gas, steam and hot water supply Collection, purification and distribution of water

240

Multinational enterprises in Asian development

over 44 per cent of the increase in total industrial employment between 2000 and 2004. The average wage of FIEs has been consistently higher than that of non-FIEs across most of the industries, a pattern consistent with the findings of a large literature on the wage behaviour of foreign affiliates of MNEs in various countries (Lipsey 2004). While employment in FIEs has increased notably, their share in industrial employment has persistently lagged behind the comparable share in gross industrial output (Table 10.8). This seems to reflect the capital intensity bias inserted into FIE production, the heavy-industry emphasis of the investment approval policy in the 1990s (as discussed above), and the continuing domestic market bias in the trade policy regime (Athukorala 2006a). Despite recent expansion in labour-intensive export-oriented product sectors, the output composition of FIEs is still dominated by highly capital-intensive sectors promoted by the protectionist trade regime. For instance, chemical, metallic and non-ferrous mineral, fabricated metal product, consumer electronics and motor vehicle production accounted for over 70 per cent of total output, compared with a combined employment share of less than 20 per cent. There are signs that, with the continuing increase in the relative importance of export-oriented ventures among FIEs, the employment potential of FIEs has already begun to improve. Of particular significance in this connection is the growing importance of assembly activities in the global electronics industry and other high-tech industries as an area of involvement for foreign investors in Vietnam. However, Vietnam has a long way to go in replicating the East Asian success story in this sphere. Perhaps the most visible contribution of FIEs to the Vietnamese economy is in export expansion. The share of FIEs in total non-oil merchandise exports from the country increased persistently from 2.5 per cent in 1991 to 30.2 per cent in 2000 and 43.5 per cent in 2005 (Table 10.9). The role of FIEs is especially important in some key export industries, such as footwear, where they accounted for over three-fourths of total exports, garments and textiles (35 per cent), and electronics and electrical goods (mostly components) (95 per cent). The two export processing zones in Ho Chi Minh City have begun to make an impressive contribution to export expansion (see Appendix 10.1). Table 10.10 provides data on the commodity composition of exports by FIEs. The patterns emerging from the data are basically consistent with the typology of MNE involvement in export-oriented production in developing countries presented in Chapter 3. Contrary to the policy makers’ expectations, FIEs in so-called heavy industries such as chemicals and chemical products, basic metal products, fabricated metal products and motor vehicles have not begun to make a significant contribution to export expansion,

241

Foreign direct investment in economic transition

Table 10.9

Export performance of foreign-invested enterprises

Year

Exports of FIEs (US$ million) Total

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

52 112 269 352 1473 2132 3203 3215 4682 6811 6796 7877 10 161 14 487 17 300

Crude oil 0 0 0 0 1033 1346 1413 1233 2092 3491 3123 3275 3821 5671 7000

Others 52 112 269 352 440 786 1790 1982 2590 3320 3673 4602 6340 8816 10 300

FIE share in . . . Total exports (%)

Non-oil exports (%)

2.5 4.3 9.0 8.7 27.0 29.4 34.9 34.3 40.6 47.0 45.2 47.2 50.4 54.7 56.4

2.5 4.3 9.0 8.7 10.0 13.3 23.0 24.4 27.4 30.2 30.9 34.3 38.8 42.3 43.5

Source: Compiled from data provided by the Central Statistical Organization, Hanoi.

despite special incentives linked to export performance requirements. The standard labour-intensive goods (in particular garments, footwear and wood products) dominated the product composition to begin with. From about the late 1990s, exports of parts and components of office, accounting and computing machines, electrical machinery and apparatus and other machinery have begun to gain importance. This important structural shift in export composition is an indication of the importance of FDI participation in linking Vietnam to the ongoing process of product fragmentation in global manufacturing (Athukorala 2005 and 2006b). However, so far this product line in Vietnam has been dominated by small- and medium-scale assembly plants, the only significant large-scale player being Hitachi, which runs an assembly plant in the South with a total employment of over 4000. So far Vietnam has not been successful in attracting any of the other major electronics multinationals such as Siemens, National Semiconductors, Motorola, Seagate and Nixdorf which play a significant role in the electronics revolution in Singapore and Malaysia and more recently in the Philippines. One possible explanation for the reluctance of high-tech industries to set up assembly plants in Vietnam seems to lie in the unsettled investment climate. Foreign firms involved in vertically integrated assembly

242

Multinational enterprises in Asian development

Table 10.10 Commodity composition of exports by foreign-invested enterprises, 1996–2005 (%) VSIC 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

Food products and beverages Tobacco products Textiles Wearing apparel Footwear and leather products Wood and wood products Paper and paper products Publishing and printing Coke and refined petroleum products Chemicals and chemical products Rubber and plastic products Other non-metallic mineral products Basic metal products Fabricated metal products Machinery and equipment n.e.c Office, accounting and computing machines Electrical machinery and apparatus n.e.c Radio, television and communication equipment Medical and optical instruments, watches and clocks Motor vehicles, trailers and semi-trailers Other transport equipment Furniture, manufacturing n.e.c

Total US$ million

1996–98

2002–04

5.9 0.0 10.7 9.5 30.9 0.0 0.3 0.0 0.0 2.1 1.2 1.2 0.2 1.3 1.0 0.0 29.2 2.2 0.5 0.2 0.4 3.2

6.5 0.0 4.0 11.9 21.2 0.4 0.6 0.1 0.0 2.9 2.1 0.9 0.9 2.2 8.6 0.2 24.8 1.4 0.6 1.9 4.1 4.7

100

100

1109

3114

Notes: VSIC Vietnam Standard Industry Classification (based on the International Standard Industry Classification, ISIC). n.e.c not elsewhere classified. Source: Compiled from data provided by the Ministry of Planning and Investment, Hanoi.

activities, unlike those involved in light consumer goods industries, view investment risk from a long-term perspective. A consideration central to any assessment of national gains to host countries from FDI is the contribution of FIEs to productivity growth in the national economy. FIEs are expected to contribute to productivity growth both directly, through their role as part of the domestic economy, and indirectly, through spillover effects on the performance of domestic firms. In the remainder of this section, we undertake a preliminary analysis

Foreign direct investment in economic transition

243

of the direct productivity implications of FIEs in Vietnamese manufacturing using data (at the two-digit industry level) for the four years from 2000 to 2004 tabulated from unpublished returns to the annual Industrial Census conducted by the Central Statistical Organization. The most widely used indicator of factor productivity is labour productivity (LP), measured as valued added per unit of labour input. Growth of labour productivity refers to an increase in the value of goods produced by the average worker (or the increased efficiency of the average worker). In reality, workers may produce more not only because of an increase in efficiency but also because they have more inputs (capital, in the two-factor case) to work with. Thus LPG (labour productivity growth) spuriously captures change in capital per worker as part of measured productivity. Total factor productivity (TFP) – the residual output after accounting for all factor inputs (in this case, capital and labour) – avoids this problem and this is our preferred productivity measure. However, it is important to see the sensitivity of the results to the use of LPG in place of TFPG (total factor productivity growth), because the former is the most widely used (and oldest) indicator of factor productivity. Estimates of labour productivity growth and total factor productivity growth are reported in Table 10.11, together with some supplementary data to facilitate the analysis. The methodology and data source are briefly described in notes to the table. Interestingly there is a sharp contrast in productivity performance of FIEs during 2000–03 in terms of the two alternative indicators. LPG of FIEs in total manufacturing contracted at a compound rate of 2.4 per cent during this period in a context where LPG of all firms remained virtually unchanged. By contrast, TFPG of FIE production increased at a compound rate of 2.2 per cent compared with a mere 0.6 increase recorded by pure local firms and 1.2 per cent by all firms. At the disaggregated level, FIEs’ contribution to productivity improvement is particularly impressive in office, accounting and computing machines (11.1 per cent), electrical machinery (9.8 per cent), and other transport equipment (17.9 per cent) – industries which have become increasingly export oriented over time (see above). By contrast, in most of the domesticmarket-oriented heavy industries, where FDI participation was encouraged by the government at the initial stage of reform, FIEs’ productivity growth is negative or near zero. Moreover, in these industries there is no notable difference in productivity performance between FIEs and local firms. The interesting contrast between the measured LPG and TFPG for FIEs points to an important ongoing structural change in their performance, namely a decline in the degree of capital intensity of production. In total manufacturing, the degree of capital intensity of FIEs declined at a compound rate of 5.1 per cent during this period (whereas capital intensity of

244

32

31

25 26 27 28 29 30

15 16 17 19 20 21 22 24

VSIC code

68.9

63.4 1.7

2.4

2.8 8.6 2.2 2.1 1.3 0.4

70.4 20.1 4.4 3.0 3.7 1.1 1.4 1.4 5.2

38.9 34.5 0.5 33.8 60.1 21.2 15.0 2.0 44.3 33.4 27.2 31.8 40.3 33.8 99.0

(2) Total

2.5

3.2

1.8 4.2 1.7 1.7 0.8 0.8

54.1 13.5 0.0 2.2 5.0 0.4 0.5 0.0 4.4

1.0

1.7

3.7 12.7 2.7 2.5 1.8 0.0

85.5 26.2 8.5 3.7 2.5 1.8 2.3 2.6 5.9

(3) (4) FIEs Local

Composition output (%)

(1)

FIE share in output (%) (6) FIEs

1.78

0.06

0.47 2.56 7.55 1.26 2.37 7.45

2.86 1.65 2.92 1.53 2.12 1.64 2.70 2.40 3.63

(7) Local

4.23

2.46

0.46

4.43

4.43 2.01 9.35 1.59 3.06 14.38 5.78 2.40 4.41 5.08 6.82 1.10

0.01 5.12 0.06 3.24 1.98 22.80 0.55 1.87 0.09 3.36 1.43 1.96 0.78 7.46 2.43 4.48 2.57 1.84

(5) Total

Growth of capital intensity (%) (9) FIEs

(10) Local

3.1

23.2

1.4

18.9

9.5

31.3

5.0 7.5 3.9 0.6 1.5 0.9 1.1 6.6 3.6 18.3 17.2 20.1 1.2 2.4 2.1 3.4 2.7 2.3

0.3 2.4 1.0 8.5 4.0 10.5 6.5 2.2 6.4 3.6 12.4 0.6 5.1 6.6 5.7 16.5 14.5 16.9 11.2 3.7 12.5 2.2 6.4 2.5 4.4 4.6 6.6

(8) Total

Labour productivity growth (LPG) (%)

Vietnam manufacturing: estimates of productivity growth and related data (2000–03)

Total manufacturing Food products and beverages Tobacco products Textiles Footwear and leather products Wood and wood products Paper and paper products Publishing and printing Chemicals and chemical products Rubber and plastic products Non-metallic mineral products Basic metal products Fabricated metal products Machinery and equipment Office, accounting and computing machines Electrical machinery and apparatus Radio, television and communication equipment

Product

Table 10.11

0.3

9.8

0.5 0.6 1.9 14.3 2.3 10.8

1.2 2.3 2.1 5.3 0.6 5.8 0.8 2.2 3.4

(11) Total

0.6 2.6 2.1 4.1 0.1 5.2 0.9 2.3 4.5

(13) Local

0.3

9.1

0.4

10.8

1.2 1.6 1.2 0.1 1.8 4.2 10.2 17.6 2.2 2.2 11.1 3.0

2.2 1.7 3.4 6.9 1.2 8.1 1.0 0.5 2.2

(12) FIEs

Total factor productivity growth (TFPG) (%)

245

Motor vehicles, trailers and semi-trailers Other transport equipment Furniture and miscellaneous manufacturing 61.5 44.9

80.6 3.4 2.0

2.6 4.3 1.9

4.3 2.5 2.1

1.1 2.11 2.90

4.42 5.73 1.04

1.05 6.47 5.84

8.53 17.0 5.6

2.6 7.4

1.6 4.6

17.8 18.8 21.7 17.9 1.1

5.1 18.6 0.4

6.4 16.5 2.1

0.7

Data source and method of data compilation Estimates are based on data compiled from unpublished returns to the annual Industrial Census conducted by the General Statistical Office during the four years 2000 to 2003. Firm-level data were aggregated to three-digit VSIC using concordance provided by the GSO. The real output (value added) was derived by deducting real intermediate inputs from real gross output. Nominal gross output is deflated by the two-digit level producer price indices. Intermediate input price indices were derived by applying input weights (derived from the 2000 Input–Output Table) to two-digit producer price indices. Capital stock data were deflated by the implicit deflator for fixed capital formation derived from national accounts. Labour input share is estimated as the share of nominal wages in nominal value added. Clothing (VSIC 38), coke and refined petroleum products (VSIC 23) and medical and optical products (VSIC 33) are not covered in the estimates because of serious data gaps which made it difficult to construct reliable data series on the capital stock.

Notes: LPG  Go  GL, where Go and GL denote annual compound growth of output (value added) and labour (number of production workers). TFPG  GO  SL GL  SK GK, where GO, GL and GK denote annual compound growth of output (value added) labour and the stock of capital; and SL and SK denote the average value shares of labour and capital and material in output. VSIC  Vietnam Standard Industry Classification (based on the International Standard Industry Classification, ISIC).

35 36

34

246

Multinational enterprises in Asian development

local firms increased by 2.9 per cent). As already noted, this in turn is a reflection of the increased export orientation of FIE production. By contrast, TFPG is negatively related with the change in capital intensity (whereas, as already noted, labour productivity growth is positively related with capital intensity).3

CONCLUSIONS AND POLICY INFERENCES There has been significant improvement over the past decade in Vietnam’s legal and institutional framework for the approval and monitoring of foreign investment, culminating in the promulgation of the 2006 Investment Law. Investment incentives and the tax law have also been revised and streamlined. However, complex bureaucratic requirements and procedures for getting a project approved and implemented, which often take several years to complete, remain perhaps the biggest obstacle to investors. According to interview-based surveys of investors, this process typically requires the involvement of at least eight central government agencies and numerous local government entities. Side by side with the restrictive FDI regime are various obstacles to import trade. Vietnam’s investment screening criteria are not consistent with national development priorities. As the experiences of latecomers to export-led industrialization in Asia have vividly demonstrated, export expansion through greater FDI participation is a surer way to generate employment and reduce poverty. With its vast human capital base and strategic location in the world’s most dynamic economic region, Vietnam is well placed to follow that path. Under this growth strategy, ‘labour intensity’ should be treated as a virtue, rather than a drawback, in approving foreign investment projects. The trends in FDI flows to Vietnam over the past one-and-a-half decades largely reflect changes/shifts in the domestic investment climate rather than global trends. When the data are appropriately adjusted for some large urban development projects, the approval of which was expedited from time to time to ‘arrest’ the decline in total FDI inflows, it is clearly seen that the FDI boom in the first half of the 1990s came to an end by 1996, well before the onset of the East Asian crisis. Further reform implemented in response to this decline and reconfirmation of government commitment to promote FDI seem to have contributed to reversing the downturn from about the early 2000s. At first look, the contribution of FIEs to growth in manufacturing output and GDP has been very impressive. But the contribution of FIEs to employment expansion has lagged behind their rapid output growth. The explanation seems to lie in the dominance of capital-intensive product

Foreign direct investment in economic transition

247

sectors in the output mix of the FIE sector, an outcome guided by industrial promotion initiatives under a protectionist trade regime. Fortunately, there are signs that with the continuing increase in the relative importance of export-oriented ventures among FIEs the employment potential of FIEs has already begun to improve. Of particular significance in this connection is the growing importance of assembly activities in the global electronics industry and other high-tech industries as an area of involvement for foreign investors in Vietnam. However, Vietnam has a long way to go in replicating the East Asian success story in this sphere. There is no evidence from the Vietnamese experience to suggest that imposition of domestic-content requirements on foreign-affiliated firms can act as a useful infant industry tactic in forging backward linkages with domestic downstream suppliers. This policy instrument has only served to provide a protected/assured market for SOEs involved in intermediate goods production industries. Nor is there evidence to suggest that export performance requirements have achieved the anticipated objective of integrating domestic manufacturing into global networks of MNEs on a costeffective and sustainable basis. Given the handsome profits assured for their domestic sales (which naturally account for the bulk of total output), MNEs’ subsidiaries have so far met export requirements imposed upon them without much fuss. Most of these forced exports have gone to marginal markets (mostly in Africa and Eastern Europe) which are not covered by the global sales networks of the parent companies and their subsidiaries located in export-oriented economies in East Asia and where low price, rather than quality standards, is the prime factor in export ‘success’.

NOTES 1. Key references, which also provide useful listings of related works, include Huang (2003), Lardy (2002), Naughton (1996), Lankes and Stern (1997) and MacBean (ed.) (2000). 2. On 28 February 2006, Intel Corporation, the world’s largest semiconductor producer, announced that it will invest $300 million to build a semiconductor testing and assembly plant in Ho Chi Minh City as part of its worldwide expansion of production capacity. When completed this will be the seventh assembly site of Intel’s global network and is projected to eventually employ about 1200 workers. www.intel/com/pressroom/archive/ releases/20060228corp.htm. 3. Labour productivity growth (lpg) is the sum of total factor productivity growth (tfpg) and growth in capital intensity (kg) weighted by the share of capital in output (sk), that is, lpg  tfpg
sk kg.

248

Multinational enterprises in Asian development

APPENDIX 10.1: EXPORT PROCESSING ZONES IN VIETNAM A widely debated aspect of the trade–FDI nexus in developing countries concerns the role of export processing zones (EPZs). The early studies, both those focused on selected aspects of EPZs, such as employment creation, linkages to the rest of the economy and net foreign exchange earnings, and the few available systematic cost–benefit analyses of selected EPZs, found their developmental contribution to be at best marginal (Warr 1990, pp. 35–6). A number of more recent studies, however, have identified them as a useful transitional tool in the process of integrating the national economy within the world economy (ADB 1997, Johansson and Nilsson 1997, Radelet and Sachs 1997). According to the ‘new view’, the early studies, given their narrow focus on the direct economic impacts (or national profitability), have ignored important catalyst effects of EPZs on potential domestic exporters, operating through exposure to marketing know-how and technology, direct demonstration effects and linking, and bringing international buyers to the country. In many countries EPZs failed to generate these externalities and to provide a conducive setting for an export take-off, not because of any intrinsic limitations of the zones but because they were used as an appendage to a highly regulated domestic economy. Moreover, the ‘footloose industry argument’ against EPZs (that is, that they possessed weak linkages with the rest of the economy and would quickly migrate in response to rising domestic costs) ignored the inevitable time lags involved in the process of linkage formation with the domestic economy by new entrants to EPZs. There is convincing evidence that EPZ firms tend to increase their local purchases and shift over to more sophisticated production processes as their operations in the host country mature, provided the local business environment is conducive to such behaviour. Of the six EPZs set up in Vietnam since the mid-1990s, only three are currently in operation: Linh Trung and Tan Thuan (both in Ho Chi Minh City) and Nomura in the North (Hai Pong). The other three have been converted into industrial zones, given the poor investor response from solely exportproducing foreign investors. This mixed performance record, however, does not seem to warrant the inference that EPZs have no ‘special’ role to play in Vietnam’s drive to promote export-oriented FDI. On the contrary, the performance record of the two Southern zones suggests that exportoriented investors have a distinct preference for locating in these zones and the export performance of firms located there has been superior to that of the country as a whole.

Foreign direct investment in economic transition

249

Investment in the two zones, in terms of both the number of new firms entering and the value of committed investment, has increased continuously over the past seven years, despite the decline in total FDI inflows to the country. During this period, exports from the two EPZs have increased much faster than total manufactured exports by FIEs. Their share in total non-oil manufactured exports increased from 11 per cent in 1995 to over 35 per cent in 2003. By 2003 total employment stood at 75 000, slightly more than a quarter of total FIE employment in the country. The other EPZs in Vietnam have failed not because of the irrelevance of the EPZs as a ‘special’ (of course, second-best) tool of export promotion through FDI but because of various implementation problems. For instance, Nomura Zone has developed high-quality infrastructure, including port facilities, but it has suffered from problems of labour availability since its inception because of its remote location. In the case of the other failed zones too, the location decisions were influenced by concern with the importance of spatial development and particularly the progress of Vietnam’s poor and remote regions, rather than the preconditions for the successful operation of the zones. Both Southern zones are well located with easy access to port facilities and urban infrastructure. Moreover, the EPZ administration in Ho Chi Minh City seems to have made use of the flexibility provided under the decentralized FDI administration mechanism of the country to create a more favourable business environment (including a speedy import clearance) for firms located in the zones. Further initiatives for improving the performance of EPZs and enhancing national gains from them include (a) removing the current rigid restrictions on the area of operation of approved firms, (b) expanding the product coverage beyond manufacturing into services, trading and other related activities, and (c ) extending the current duty rebate and turnover tax rebate schemes for exporters to cover local producers who supply inputs to firms in EPZs.

References ADB (Asian Development Bank) (1997), Emerging Asia: Changes and Challenges, Manila: ADB. ADB (Asian Development Bank) (2006), ‘Foreign direct investment in the transition economics of the Mekong: trends, policy, and economic impacts’, draft report, Manila: ADB, Mekong Department. Agarwal, Jamuna (1977), ‘Factor proportions in foreign and domestic firms in Indian manufacturing’, Economic Journal, 86, 589–94. Ali, M. Yunus (1999), ‘Economic liberalisation and growth in Bangladesh: the role of foreign private investment’, in I. Mohammed Alaudin and Samiul Hasan (eds), Development, Governance and Environment in South Asia: A Focus on Bangladesh, London: Macmillan, pp. 87–102. Amsden, Alice H. and Wa-wen Chu (2003), Beyond Late Development: Taiwan’s Upgrading Policies, Cambridge, MA: MIT Press. Athreye, Suma and Sandeep Kapur (2001), ‘Private foreign investment in India: pain or panacea?’, World Economy, 24 (3), 399–424. Athukorala, Prema-chandra (1995), ‘Foreign direct investment and manufacturing for export in a new exporting country: the case of Sri Lanka’, World Economy, 14 (2), 543–64. Athukorala, Prema-chandra (2002a), Crisis and Recovery in Malaysia: The Role of Capital Controls, 2nd edn, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Athukorala, Prema-chandra (2002b), ‘Foreign direct investment and manufacturing exports: opportunities and strategies’, background paper to the World Bank study ‘Vietnam’s exports: policies and prospects’, Hanoi: World Bank Vietnam. Athukorala, Prema-chandra (2003a), ‘FDI in crisis and recovery: lessons from the 1997–98 Asian crisis’, Australian Economic History Review, 43 (2), 197–213. Athukorala, Prema-chandra (2003b), ‘Product fragmentation and trade patterns in East Asia’, Trade and Development Working Paper 2003/21, Division of Economics, Research School of Pacific and Asian Studies, Australian National University. Athukorala, Prema-chandra (2004), ‘External sector policies for accelerated growth’, background paper for Sri Lanka Development Policy Review, Washington, DC: World Bank (June) (processed). 250

References

251

Athukorala, Prema-chandra (2005), ‘Component trade and China’s regional economic integration’, in R. Garnaut and L. Song (eds), The China Boom and its Discontents, Canberra: Asia Pacific Press, pp. 215–39. Athukorala, Prema-chandra (2006a), ‘Trade policy reforms and the structure of protection in Vietnam’, World Economy, 29 (2), 161–87. Athukorala, Prema-chandra (2006b), ‘Product fragmentation and trade patterns in East Asia’, Asian Economic Papers, 4 (3), 1–27. Athukorala, Prema-chandra and Satish Chand (2000), ‘Trade orientation and productivity gains from international production: a study of overseas operations of United States transnational corporations’, Transnational Corporations, 9 (2), 1–30. Athukorala, Prema-chandra and Hal Hill (2003), ‘FDI and host country development: the East Asian experience’, in B. Bora (ed.), Foreign Direct Investment: Research Issues, London: Routledge, pp. 168–94. Athukorala, Prema-chandra and Sisira Jayasuriya (1988), ‘Parentage and factor proportions: a comparative study of third-world multinationals in Sri Lankan manufacturing’, Oxford Bulletin of Economics and Statistics, 50 (4), 409–23. Athukorala, Prema-chandra and Sisira Jayasuriya (1994), Macroeconomic Policies, Crises, and Growth in Sri Lanka 1969–90, Washington, DC: World Bank. Athukorala, Prema-chandra and Sisira Jayasuriya (2003), ‘Food safety issues, trade and WTO rules: a developing country perspective’, World Economy, 26 (9), 613–37. Athukorala, Prema-chandra and Jayant Menon (1996), ‘Foreign direct investment and industrialisation in Malaysia: exports, employment and spillovers’, Asian Economic Journal, 10 (1), 29–44. Athukorala, Prema-chandra and Sarath Rajapatirana (2000a), Liberalization and Industrial Transformation: Sri Lanka in International Perspective, Delhi and Oxford: Oxford University Press. Athukorala, Prema-chandra and Sarath Rajapatirana (2000b), ‘Liberalization and industrial transformation’, Economic Development and Cultural Change, 48 (3), 543–72. Athukorala, Prema-chandra and Sarath Rajapatirana (2003), ‘Capital inflows and the real exchange rate: a comparative study of Asia and Latin America’, World Economy, 26 (4), 613–37. Athukorala, Prema-chandra and Kishor Sharma (2005), ‘Foreign investment in Nepal’, in Raghbendra Jha (ed.), Economic Growth, Economic Performance and Welfare in South Asia, Houndmills, Hampshire: Palgrave Macmillan, pp. 323–39.

252

Multinational enterprises in Asian development

Athukorala, Prema-chandra and Nobuaky Yamashita (2006 forthcoming), ‘Production fragmentation and trade integration: East Asia in a global context’, North American Journal of Economics and Finance. Athukorala, Prema-chandra and Peter G. Warr (2002), ‘Vulnerability to a currency crisis: lessons from the Asian experience’, World Economy, 25 (1), 33–57. Athukorala, Prema-chandra, Chris Manning and Piyasiri Wickramasekara (2000), Growth, Employment and Migration in Southeast Asia: Structural Change in the Greater Meuong Region, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Bajpai, Nirupam and Jeffrey Sachs (2000), ‘Foreign direct investment in India: isues and problems’, Development Discussion Paper No. 759, Harvard Institution for International Development. Balasubramanyam, V.N. (1983), ‘Transnational corporations: choice of techniques and employment in developing countries’, in B. Weisbrod and H. Hughes (eds), Human Resources Employment and Development, London: Macmillan. Balasubramanyam, V.N. and Vidya Mhambare (2003), ‘FDI in India’, Transnational Corporations, 12 (2), 45–72. Balasubramanyam, V.N. and M.A. Salisu (1991), ‘Export promotion, import substitution and direct foreign investment in developing countries’, in K.A. Koekkoek and L.B.M. Mennes (eds), International Trade and Global Development, London: Routledge, pp. 191–207. Balasubramanyam, V.N., M.A. Salisu and David Sapsford (1996), ‘Foreign direct investment and growth in EP and IS countries’, Economic Journal, 106 (434), 92–105. Baldone, S., F. Sdoagati and L. Tajoli (2001), ‘Patterns and determinants of international fragmentation of production: evidence from outward processing trade between the EU and Central Eastern European countries’, Weltwirtschaftliches Archiv, 137 (1), 80–104. Baldwin, Robert E. (2001), ‘The implications of increasing fragmentation and globalization for the World Trade Organization’, in Leonard K. Cheng and Henryk Kierzkowski (eds), Global Production and Trade in East Asia, Boston: Kluwer Academic Publishers, pp. 249–71. Barba Navaretti, Giorgio and Anthony J. Venables (2004), Multinational Firms in the World Economy, Princeton, NJ: Princeton University Press. Barro, R. and X. Sala-I-Martin (1995), Economic Growth, New York: McGraw Hill. Barry, Frank and John Bradley (1997), ‘FDI and trade: the Irish host country experience’, Economic Journal, 107, 798–811.

References

253

Bennett, Douglas and Kenneth E. Sharp (1979), ‘Transnational corporations and the political economy of export promotion: the case of the Mexican automobile industry’, International Organization, 32 (2), 177–201. Bhagwati, Jagdish (1973), ‘The theory of immiserizing growth: further applications’, in Michael Connolly and Alexander Swoboda (eds), International Trade and Money, Toronto: University of Toronto Press, pp. 45–54. Bhagwati, Jagdish (1978), Anatomy and Consequences of Exchange Control Regimes, New York: Balinger Publishing. Bhagwati, Jagdish (1985), ‘Investing abroad’, Esmée Fairbairn Lecture, Lancaster: University of Lancaster Press. Bhagwati, Jagdish (1994), ‘Free trade: old and new challenges’, Economic Journal, 104 (423), 231–46. Bhagwati, Jagdish (1996) [1985], ‘Investing abroad’, in Douglas D. Irwin (ed.), Political Economy and International Economics: Collected Essays of Jagdish Bhagwati, Cambridge, MA: MIT Press, pp. 309–39. Bhalla, A.S. (1975), ‘The concept and measurement of labor intensity’, in A.S. Bhalla (eds), Technology and Employment in Industry: A Case Study Approach, Geneva: ILO, pp. 11–33. Biersteker, J.J. (1978), Distortion or Development? Contending Perspectives on the Multinational Corporation, Cambridge, MA: MIT Press. Birkinshaw, Julian M. and Allen J. Morrison (1995), ‘Configuration of strategy and structure in subsidiaries of multinational corporations’, Journal of International Business Studies, 26 (4), 729–50. Blalock, Garrick and Paul J. Gertler (2005), ‘Foreign direct investment and externalities: the case for public intervention’, in Theodore H. Morgan, Edward M. Graham and Magnus Blomström (eds), Does Foreign Direct Investment Promote Development?, Washington, DC: Institute for International Economics, pp. 73–106. Blomström, Magnus (1990), Transnational Corporations and Manufacturing Exports from Developing Countries, New York: United Nations Centre for Transnational Corporations. Blomström, Magnus (1991), ‘Host country benefits of foreign direct investment’, in D.G. McFetridge (ed.), Foreign Investment, Technology and Economic Growth, Calgary: University of Calgary Press, pp. 93–108. Blomström, Magnus and Ari Kokko (1998), ‘Multinational corporations and spillovers’, Journal of Economic Surveys, 12 (3), 247–77. Blomström, Magnus and Robert E. Lipsey (1993), ‘Foreign firms and structural adjustment in Latin America: lessons from the debt crisis’, in Gote Hansson (ed.), Trade Growth and Development, London: Routledge, pp. 109–32.

254

Multinational enterprises in Asian development

Bloom, Nicholas and Rachel Griffith (2001), ‘The internationalisation of UK R&D’, Fiscal Studies, 22 (3), 337–55. Bloom, Nick, Rachel Griffith and John Van Reenen (2002), ‘Do R&D tax credits work? Evidence from a panel of countries 1979–1997’, Journal of Public Economics, 85 (1), 1–31. Borrus, Michael (1997), ‘Left for dead: Asian production networks and the revival of US electronics’, in Barry Naughton (ed.), The China Circle: Economics and Technology in the PRC, Taiwan and Hong Kong, Washington, DC: Brookings Institution Press. Bose, Swadesh R. (1983), ‘The Pakistan economy since Independence (1947–70)’, in Dharma Kumar and Meghnad Desai (eds), The Cambridge Economic History of India, Vol. 2, Cambridge, UK: Cambridge University Press, pp. 995–1026. Brecher, Richard A. and Carlos F. Diaz-Alejandro (1977), ‘Tariff, foreign capital and immiserizing growth’, Journal of International Economics, 7 (3), 317–22. Brecher, Richard A. and Ronald Findlay (1983), ‘Tariff, foreign capital and national welfare’, Journal of International Economics, 14 (2), 277–88. Brooks, Douglas H. and Hal Hill (2004), Managing FDI in a Global Economy: Asian Experiences, Houndmills, UK: Palgrave Macmillan. Brown, Drusilla K., Alan V. Deardorff and Robert M. Stern (2003), ‘The effect of multinational production on wages and working conditions in developing countries’, National Bureau of Economic Research (NBER) Working Paper, Cambridge, MA: NBER. Bui, Tuan (2004), ‘Viet Nam’, in Douglas H. Brooks and Hal Hill (eds), Managing FDI in a Global Economy: Asian Experiences, Houndmills: Palgrave Macmillan, pp. 281–308. Busjeet, V. (1980), ‘Foreign investors from less-developed countries’, unpublished doctoral dissertation, Harvard Business School. Cairncross, Frances (1997), The Death of Distance: How the Communication Revolution will Change Our Lives, London: Orion Business Books. Calvo, Guillermo A. (1991), ‘The peril of sterilization’, IMF Staff Papers, 31 (4), 921–6. Calvo, Guillermo A., Leonardo Leiderman and Carmen M. Reinhart (1994), ‘Capital inflows to Latin America: the 1970s and 1990s’, in Edmar L. Bacha (ed.), Economics in a Changing World, London: Macmillan, pp. 123–48. Calvo, Guillermo A., Leonardo Leiderman and Carmen M. Reinhart (1996), ‘Inflows of capital to developing countries in the 1990s’, Journal of Economic Perspectives, 10 (2), 123–39.

References

255

Cantwell, John (1994), ‘The relationship between international trade and international production’, in David Greenaway and L. Alan Winters (eds), Surveys in International Trade, Oxford: Basil Blackwell, pp. 303–28. Cantwell, John and Lucia Piscitello (2002), ‘The location of technological activities of MNCs in European regions: the role of spillovers and local competencies’, Research Policy, 8 (1), 69–96. Caves, D., L. Christensen and W. Diewert (1982), ‘Output, input and productivity using superlative index numbers’, Economic Journal, 92, 73–96. Caves, R.E. (1996), Multinational Enterprise and Economic Analysis, 2nd edn, Cambridge, UK: Cambridge University Press. Chen, Edward K.Y. (1983), Multinational Corporations, Technology and Employment, London: Macmillan. Chia Siow Yue (1985), ‘The role of foreign trade and investment in the development of Singapore’, in Walter Galenson (ed.), Foreign Trade and Investment: Economic Development in the Newly Industrializing Asian Countries, Madison, WI: University of Wisconsin Press, pp. 259–97. Chudnovsky, Daniel and A. Lopez (2000), ‘A third wave of FDI from developing countries: Latin American TNCs in the 1990s’, Transnational Corporations, 9 (2), 31–73. Clausing, Kimberly A. (2001), ‘The impact of transfer pricing on intrafirm trade’, in James R. Hines (ed.), International Taxation and Multinational Activity, Chicago: University of Chicago Press, pp. 173–94. Cohen, Benjamin (1975), Multinational Firms and Asian Exports, New Haven, CT: Yale University Press. Cohen, W.M. and D.A. Levin (1989), ‘Empirical studies of innovation and market structure’, in R. Schmalensee and R.D. Willig (eds), Handbook of Industrial Organization, Volume 2, Amsterdam: Elsevier Publishing B.V., pp. 1060–107. Cook, Robert D. (1977), ‘Detection of influential observations in linear regressions’, Technometrica, 9 (1), 15–18. Corbo, Vittorio and L. Hernandez (2000), ‘Private capital inflows and the role of economic fundamentals’, in Felipe Larian B. (ed.), Capital Flows, Capital Controls, and Currency Crises in Latin America in the 1990s, Ann Arbor: University of Michigan Press, pp. 83–100. Corden, W. Max (1960), ‘The geometric representation of policies to attain internal and external balance’, Review of Economic Studies, 18 (1), 1–22. Corden, W. Max (1994), Economic Policy, Exchange Rates and the International System, Oxford: Oxford University Press. Corden, W. Max (1996), ‘Pragmatic orthodoxy: macroeconomic policies in seven East Asian economies’, Occasional Paper 61, San Francisco, CA: International Center for Economic Growth.

256

Multinational enterprises in Asian development

Corden, W. Max (2002), Too Sensational: On the Choice of Exchange Rate Regimes, Cambridge, MA: MIT Press. Deardorff, Alan V. (2001), ‘Financial crisis, trade, and fragmentation’, in Leonard K. Cheng and Henryk Kierzkowski (eds), Global Production and Trade in East Asia, Boston: Kluwer Academic Publishers, pp. 9–31. Desai, Ashok V. (2002), ‘Comment’ [on Saxenian (2002)] in Anne O. Krueger (ed.), Economic Policy Reforms and the Indian Economy, Delhi and Oxford: Oxford Universty Press, pp. 203–10. Diaz-Alejandro, C.F. (1977), ‘Foreign direct investment by Latin Americans’, in T. Agmor and C.P. Kindleberger (eds), Multinationals from Small Countries, Cambridge, MA: MIT Press, pp. 167–96. Dobson, W. and Chia S.Y. (1997), Multinationals and East Asian Integration, Singapore: Institute of Southeast Asian Studies. Dollar, David and Mary Hallward-Driemeier (2000), ‘Crisis, adjustment and reform in Thailand’s industrial firms’, World Bank Research Observer, 15 (1), 1–22. Dollar, David and Edward N. Wolf (1993), Competitiveness, Convergence and International Specialization, Cambridge, MA: MIT Press. Dornbusch, R. and A. Warner (1994), ‘Mexico: stabilization, reform, and no growth’, Brooking Papers in Economic Activity, 1, pp. 252–315. Drysdale, Peter and Ross Garnaut (1997), ‘The Pacific: an application of a general theory of economic integration’, in C. Fred Bergsten and Marcus Noland (eds), Pacific Dynamism and the International Economic System, Washington, DC: Institute for International Economics, pp. 183–224. Dunning, John H. (1992), Multinational Enterprises and the Global Economy, Reading, MA: Addison Wesley. Dunning, John H. (1994), ‘Multinational enterprises and the global innovatory capacity’, Research Policy, 25 (1), 67–88. Edwards, Sebastian (1989), Real Exchange Rates, Devaluation and Adjustment: Exchange Rate Policies in Developing Countries, Cambridge, MA: MIT Press. Edwards, Sebastian (1995), Crisis and Reform in Latin America: From Despair to Hope, New York: Oxford University Press. Edwards, Sebastian (1998), ‘Openness, productivity and growth: what do we really know?’, Economic Journal, 108, 383–98. Edwards, Sebastian (2000), ‘Capital flows, real exchange rates, and capital controls: some Latin American experiences’, in Sebastian Edwards (ed.), Capital Flows and the Emerging Economies: Theory, Evidence and Controversies, Chicago: University of Chicago Press, pp. 197–253. Edwards, Sebastian and M.A. Savastano (2000), ‘Exchange rate in emerging economies: What do we know? What do we need to know?’, in Anne

References

257

O. Krueger (ed.), Economic Policy Reforms: The Second Stage, Chicago: University of Chicago Press, pp. 453–510. Egger, Harmut and Pete Egger (2003), ‘Outsourcing and skill-specific employment in a small economy: Austria after the fall of the iron curtain’, Oxford Economic Papers, 55 (4), 625–43. Egger, Harmut and Pete Egger (2005), ‘The determinants of EU processing trade’, World Economy, 28 (2), 147–68. Eichengreen, Barry (2002), Financial Crises and What to Do About Them, New York: Oxford University Press. ESCAP/UNCTC (1984), ‘Technology transfer from transnational corporations in Sri Lanka’, in ESCAP/UNCTC, Cost and Conditions of Technology Transfer Through Transnational Corporations, Bangkok, pp. 151–83. Feenstra, Robert (1998), ‘Integration of trade and disintegration of production in the global economy’, Journal of Economic Perspectives, 14 (4), 31–50. Feenstra, Robert C. (2000), ‘Facts and fallacies about foreign direct investment’, in Martin Feldstein (ed.), International Capital Flows, Chicago: University of Chicago Press, pp. 331–50. Felipe, Jesus (1999), ‘Total factor productivity growth in East Asia: critical survey’, Journal of Development Studies, 35 (4), 1–41. Fernandez-Arias E. and P. Montiel (1996), ‘The surge in capital inflows to developing countries: an analytical overview’, World Bank Economic Review, 10 (1), 51–77. Fernando, Lionel E.N. (1971), ‘Some aspects of private foreign enterprises in Ceylon’, PhD thesis, University of Hull. Finger, J. Michael (1975), ‘Tariff provisions for offshore assembly and exports of developing countries’, Economic Journal, 85 (338), 365–71. Fors, Gunnar (1998), ‘Locating R&D abroad: the role of adaptation and knowledge-seeking’, in Pontus Brounerhjelm and Karolina Ekholm (eds), The Geography of Multinational Firms, Boston: Kluwer Academic Publishers, pp. 117–34. Fors, Gunnar and R. Sevensson (1994), R&D in Swedish Multinational Corporations, Stockholm: Industrial Institute for Economic and Social Research. Forsyth, D.J.C. (1980), Technology Policy for Small Developing Countries, London: Macmillan. Forsyth, D.J.C. and R.F. Solomon (1977), ‘Choice of technology and nationality of ownership in manufacturing in a developing country’, Oxford Economic Papers, 29 (2), 259–83. Frankel, Jeffrey A. and Andrew Rose (1996), ‘Currency crashes in emerging markets: an empirical treatment’, Journal of International Economics, 41, 351–66.

258

Multinational enterprises in Asian development

Frankel, Jefrey A. and Shang-Jin Wei (1997), ‘The new regionalism and Asia: impact and policy options’, in Aravind Panagariya, M.G. Quibria and Narhari Rao (eds), The Global Trading System and Developing Asia, Oxford: Oxford University Press, pp. 83–130. Freeman, Nick J. (2004), ‘The prospects for foreign direct investment in the transitional economies of South East Asia’, in Nick J. Freeman and Frank L. Bartels (eds), The Future of Foreign Investment in Southeast Asia, London: Routledge, pp. 188–200. Freeman, Nick J. and Frank L. Bartels (eds) (2004), The Future of Foreign Investment in Southeast Asia, London: Routledge. Fritsch, Winston and Gustavo H.B. Franco (1992), ‘Foreign direct investment and patterns of industrialisation and trade in developing countries’, in Gerald K. Helleiner (ed.), Trade Policy, Industrialisation and Development: New Perspectives, Oxford: Clarendon Press, pp. 197–224. Frobel, Folker, Jurgen Heinrichs and Otto Kreye (1980), The New International Division of Labour: Structural Unemployment in Industrialised Countries and Industrialisation in Developing Countries, Cambridge, UK: Cambridge University Press. Fukao, Kyoji (2001), ‘How Japanese subsidiaries in Asia responded to the regional crisis: an empirical analysis based on the MITI Survey’, in Takatoshi Ito and Anne O. Krueger (eds), Regional and Global Capital Flows: Macroeconomic Causes and Consequences, Chicago: University of Chicago Press, pp. 267–310. Garnaut, Ross (1999), ‘Exchange rates in the East Asian crisis’, in H.W. Arndt and Hal Hill (eds), Southeast Asia’s Economic Crisis: Origins, Lessons and the Way Forward, Singapore: Institute of South East Asian Studies, pp. 93–106. Garnaut, Ross (2003), ‘Australia and Japan: time to be important to each other again’, address to the Australia–Japan Business Committee conference, Kyoto, 5 October. Gavin, M., R. Hausmann and L. Leiderman (1996), ‘The macroeconomics of capital flows to Latin America’, in Ricardo Hausmann and Liliana Rojas-Suarez (eds), Volatile Capital Flows: Taming their Impact on Latin America, Washington, DC: Inter-American Bank, pp. 1–60. Gereffi, Gary (1999), ‘International trade and industrial upgrading in the apparel commodity chain’, Journal of International Economics, 48 (1), 37–70. Goh, Keng Swee (1993), ‘What causes fast economic growth?’, 4th K.T. Li Lecture, 13 October, Harvard University, reproduced in Linda Low (ed.), Wealth of East Asian Nations: Speeches and Writings by Goh Keng Swee, Singapore: Federation Press, pp. 243–58.

References

259

Golberman, Steven (1997), ‘Decentralization of research and development by multinational companies: determinants and future prospects’, in Jan Fagerberg, Pär Hansson, Lars Lundberg and Arne Melchior (eds), Technology and International Trade, Cheltenham, UK and Lyme, USA: Edward Elgar, pp. 140–57. Goldberger, Arthur S. (1991), A Course in Econometrics, Cambridge, MA: Harvard University Press. Good, D.H., M.I. Nadiri and R.C. Sickles (1996), ‘Index number and factor demand approaches to the estimation of productivity’, Working Paper 5790, New York: National Bureau of Economic Research. Gorg, Holger (2000), ‘Fragmentation and trade: US inward processing trade in the EU’, Weltwirtschaftliches Archiv, 136 (3), 403–22. Government of Sri Lanka (1972), The Five-Year Plan 1972–1977, Colombo: Government Publication Bureau. Grubert, Harry and John Mutti (1992), ‘Financial flows versus capital spending’, in Peter Hooper and J. David Richardson (eds), International Economic Transactions, Chicago: Chicago University Press, pp. 293–320. Grunwald, Joseph and Kenneth Flamm (1985), The Global Factory: Foreign Assembly and International Trade, Washington, DC: Brookings Institution. Guisinger, Stephen (1985), Investment Incentives and Performance, New York: Praeger. Guisinger, Stephen and Gerald Scully (1991), ‘Pakistan’, in Demetris Papageorgiou, Michael Michaely and Armene M. Choksi (eds), Liberalizing Foreign Trade, Volume 5, Oxford: Basil Blackwell, pp. 197–282. Hanson, Gordon H., Raymond J. Mataloni and Mathew J. Slaughter (2001), ‘Expansion strategies of U.S. multinational firms’, in Susan M. Collins and Dani Rodrik (eds), Brookings Trade Forum 2001, Washington, DC: Brookings Institution Press, pp. 245–82. Harrison, Ann (1996), ‘Openness and growth: a time-series, cross-country analysis for developing countries’, Journal of Development Economics, 48 (3), 419–47. Hausman, J.A. (1978), ‘Specification tests in econometrics’, Econometrica, 46, 1251–71. Helleiner, Gerald K. (1973a), ‘Manufactured exports from less-developed countries and multinational firms’, Economic Journal, 83 (329), 21–47. Helleiner, Gerald K. (1973b), ‘Manufacturing for exports, multinational firms and economic development’, World Development, 1 (7), 13–22. Helleiner, Gerald K. (1981), Intra-firm Trade and Developing Countries, London: Macmillan.

260

Multinational enterprises in Asian development

Helleiner, Gerald K. (1988), ‘Direct foreign investment and manufacturing for export in developing countries: a review of the issues’, in Sidney Dell (ed.), Policies for Development, London: Macmillan, pp. 123–53. Helleiner, Gerald K. (1989), ‘Transnational corporations and direct foreign investment’, in Hollis Chenery and T.N. Srinivasan (eds), Handbook of Development Economics, Vol. 2, Amsterdam: North Holland, 1142–80. Hill, Hal (2003), ‘Industry’, in Arsenio M. Balisancan and Hal Hill (eds), The Philippines Economy: Development, Policies and Challenges, New York: Oxford University Press, pp. 219–53. Hill, Hal (2004), ‘Foreign direct investment and Southeast Asian economic development: issues and challenges’, in Nick J. Freeman and Frank L. Bartels (eds), The Future of Foreign Investment in Southeast Asia, London: Routledge, pp. 256–67. Hill, Hal and Pang Eng-Fong (1992), ‘Technology exports from a small very open NIC: the case of Singapore’, World Development, 19 (5), 553–68. Hines, James R. (1995), ‘Taxes, technology transfer, and the R&D activities of multinational firms’, in Martin Feldstein, James R. Hines and R. Glenn Hubbard (eds), The Effects of Taxation on Multinational Corporations, Chicago: University of Chicago Press, pp. 225–52. Hines, James R. and Adam B. Jaffe (2001), ‘International taxation and the location of R&D activity’, in James R. Hines (ed.), International Taxation and Multinational Activity, Chicago: University of Chicago Press, pp. 201–30. Hirschey, Robert C. and Richard E. Caves (1981), ‘Research and transfer of technology by multinational enterprises’, Oxford Bulletin of Economics and Statistics, 43 (2), 115–30. Hirschman, Albert O. (1969), ‘How to divest in Latin America, and why’, Essays in International Finance No 76, Princeton, NJ: International Finance Division, Princeton University. Hobday, Michael (1995), Innovation in East Asia: The Challenge to Japan, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Hone, Angus (1974), ‘Multinational corporations and multinational buying groups: their impact on the growth of Asia’s manufactured exports’, World Development, 2 (2), 145–9. Hu, Sheng-Chenh and Vei-Lin Chan (2001), ‘Impact on production and trade of electronics in Taiwan’, paper presented at the 2nd Pacific Trade and Development Conference, The Australian National University, 20–22 August, Canberra. Huang, Yanshen (2003), Selling China: Foreign Direct Investment during the Reform Era, Cambridge, UK: Cambridge University Press.

References

261

Hufbauer, Gary, Daritus Lakdawala and Anup Malani (1994), ‘Determinants of direct foreign investment and its connection to trade’, UNCTAD Review, 39–58. Huff, W.G. (1994), The Economic Growth of Singapore: Trade and Development in the Twentieth Century, Cambridge, UK: Cambridge University Press. Hughes, Helen and You Poh Seng (eds) (1969), Foreign Investment and Industrialisation in Singapore, Canberra: Australian National University Press. Hummels, David (1999), ‘Have international transport costs declined?’, paper in progress, Graduate School of Business, University of Chicago. Hummels, David, Jun Ishii and Kei-Mu Yi (2001), ‘The nature and growth of vertical specialization in world trade’, Journal of International Economics, 54 (1), 75–96. IMF (International Monetary Fund) (1993), Balance of Payments Manual, 5th edn, Washington, DC: IMF. IMF (International Monetary Fund) (1995 and 2001), International Capital Markets: Developments, Prospects and Key Policy Issues, Washington, DC: IMF. IMF (International Monetary Fund) (2006), Vietnam: 2005 Article IV Consultations: Staff Report, Washington, DC: IMF (www.imf.org). Irwin, Douglas A. (2002), Free Trade Under Fire, Princeton, NJ: Princeton University Press. ITC (International Trade Centre) (1987), Analysis of Trading Channels for the Foreign Trade of the Philippines, Geneva: ITC. Ito, T. (2000), ‘Capital flows in Asia’, in Sebastian Edwards (ed.), Capital Flows and the Emerging Economies: Theory, Evidence and Controversies, Chicago: University of Chicago Press, pp. 255–97. Jenkins, R. (1986), ‘Third world multinationals: rhetoric or reality?’, Journal of Development Studies, 22 (2), 458–63. Johansson, H. and L. Nilsson (1997), ‘Export processing zones as catalysts’, World Development, 25 (12), 2115–28. Jones, Ronald W. (2000), Globalization and the Theory of Input Trade, Cambridge, MA: MIT Press. Jones, Ronald W. and Henryk Kierzkowski (1990), ‘The role of services in production and international trade: a theoretical framework’, in Ronald W. Jones and Anne O. Krueger (eds), The Political Economy of International Trade: Essays in Honour of Robert Baldwin, Oxford: Basil Blackwell, pp. 31–48. Jones, Ronald W. and Henryk Kierzkowski (2001a), ‘A framework for fragmentation’, in Seven W. Arndt and Henryk Kierzkowski (eds), Fragmentation: New Production Patterns in the World Economy, New York: Oxford University Press, pp. 17–34.

262

Multinational enterprises in Asian development

Jones, Ronald W. and Henryk Kierzkowski (2001b), ‘Horizontal aspects of vertical fragmentation’, in Leonard K. Cheng and Henryk Kierzkowski (eds), Global Production and Trade in East Asia, Boston: Kluwer Academic Publishers, pp. 33–51. Jones, Ronald W. and Henryk Kierzkowski (2001c), ‘Globalization and the consequences of international fragmentation’, in Rudigar Dornbusch, G. Calvo and M. Obstfeld (eds), Money, Factor Mobility and Trade: The Festschrift in Honor of Robert A. Mundell, Cambridge, MA: MIT Press, pp. 365–81. Jones, Ronald W., H. Kierzkowski and C. Lurong (2004), ‘What does the evidence tell us about fragmentation and outsourcing?’, International Review of Economics and Finance, 14 (3), 305–16. Joshi, Vijay and I.M.D. Little (1996), India’s Economic Reforms, 1991–2001, Oxford: Clarendon Press. Kaminsky, G., S. Lizondo and C.M. Reinhart (1997), ‘Leading indicators of currency crises’, IMF Working Paper WP/97/79, Washington, DC: IMF. Kidron, M. (1965), Foreign Investment in India, Oxford: Oxford University Press. Kierzkowski, Henryk (2001), ‘Joining the global economy: experience and prospects of the transition economies’, in Seven W. Arndt and Henryk Kierzkowski (eds), Fragmentation: New Production Patterns in the World Economy, New York: Oxford University Press, pp. 231–53. Kim, June-Dong and Sang-In Hwang (2000), ‘The role of foreign direct investment in Korea’s development: productivity effects and implications for the currency crisis’, in Takatoshi Ito and Anne O. Krueger (eds), The Role of Foreign Direct Investment in East Asian Economic Development, Chicago: University of Chicago Press, pp. 267–94. Kirkpatrick, C.H., N. Lee and F.I. Nixon (1984), Industrial Structure and Policy in Less Developed Countries, London: Allen and Unwin. Kohpaiboon, Archanun (2006), Multinational Enterprises and Industrial Transformation: Evidence from Thailand, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Kokko, Ari (1994), ‘Technology, market characteristics and spillovers’, Journal of Development Economics, 43 (2), 279–93. Kokko, Ari (1997), ‘Managing the transition to free trade: Vietnamese trade policy for the 21st century’, mimeo, Stockholm School of Economics, Stockholm. Kokko, Ari, Katarina Kotoglou and Anna Krohwinkel-Karisson (2003), ‘The implementation of FDI in Vietnam: an analysis of the characteristics of failed projects’, Transnational Corporations, 12 (3), 41–78. Koo, Bohn Young (1985), ‘The role of direct foreign investment in Korea’s recent economic growth’, in Walter Galenson (ed.), Foreign

References

263

Trade and Investment: Economic Development in the Newly Industrializing Asian Countries, Madison, WI: University of Wisconsin Press, pp. 176–216. Krueger, Anne O. (1978), Foreign Trade Regimes and Economic Development, Cambridge, MA: Ballinger. Krueger, Anne O. (1997), ‘Trade policy and economic development: how we learn’, American Economic Review, 87 (1), 1–22. Krueger, A.O. and A. Tornell (1999), ‘The role of bank restructuring in recovering from crisis’, NBER Working Paper, No. 7042, Cambridge, MA: National Bureau of Economic Research. Krugman, Paul (1995), ‘Growing world trade: causes and consequences’, Brookings Papers on Economic Activity, 25th anniversary issue, pp. 327–77. Krugman, Paul (2001), ‘Fire-sale FDI’, in S. Edwards (ed.), Capital Flows and the Emerging Economies, Chicago: Chicago University Press, pp. 43–60. Kumar, Nagesh (1994), Multinational Enterprises and Industrial Organisation: The Case of India, Delhi: Sage. Kumar, Nagesh (1996), ‘Intellectual property protection, market orientation and location of overseas R&D activities by multinational enterprises’, World Development, 24 (4), 673–88. Kumar, Nagesh (2001), ‘Determinants of location of overseas R&D activity of multinational enterprises: the case of US and Japanese corporations’, Research Policy, 30 (2), 159–74. Kumar, Rajiv (1989), India’s Export Processing Zones, Delhi: Oxford University Press. Kwan, C.H. (2001), Yen Bloc: Toward Economic Integration in Asia, Washington, DC: Brookings Institution Press. Lakshman, W.D. and Prema-chandra Athukorala (1985), ‘Transnational trading corporations and the structure of trading channels in Sri Lankan foreign trade’, Research Paper, ESCAP/UNCTC Joint Unit on Transnational Corporations, Bangkok, mimeo. Lall, Sanjaya (1979), ‘The international allocation of research activity by US multinationals’, Oxford Bulletin of Economics and Statistics, 41 (4), 313–31. Lall, Sanjaya (1983a), ‘The rise of multinationals from the third world’, Third World Quarterly, 5 (3), 618–26. Lall, Sanjaya (1983b), ‘Transnationals and the third world: changing perceptions’, National Westminster Bank Quarterly Review, 2–16 May. Lall, Sanjaya (2002), ‘Indicators of the relative importance of IPRs in developing countries’, Development Studies Working Paper 85, Queen Elizabeth House, University of Oxford.

264

Multinational enterprises in Asian development

Lall, Sanjaya (2003), ‘Foreign direct investment, technology development and competitiveness: issues and evidence’, in Sanjaya Lal and Shujiro Urata (eds), Competitiveness, FDI and Technological Activity in East Asia, Cheltenham, UK and Northampton, MA, USA: Edward Elgar, pp. 12–56. Lall, Sanjaya and Paul Streeten (1977), Foreign Investment, Transnationals and Developing Countries, London: Macmillan. Lamberte, Mario B., Caecer B. Cororaton, Magarita F. Guerrero and Aniceto C. Orbeta (1999), ‘Impacts of the Southeast Asian financial crisis on the Philippine manufacturing sector’, Philippine Institute of Development Studies (PIDS) Discussion Paper 99-09, Manila: PIDS. Lamfalussy, A. (2000), Financial Crises in Emerging Markets: An Essay on Financial Globalization and Fragility, New Haven: Yale University Press. Lankes, Hans-Peter and Nicholas Stern (1997), ‘Capital flows to Eastern Europe’, in Martin Felstein (ed.), International Capital Flows, Chicago: Chicago University Press, pp. 57–110. Lankes, Hans-Peter and Anthony J. Venables (1996), ‘Foreign direct investment in economic transition: the changing pattern of investment’, Economics of Transition, 4 (2), 331–47. Lardy, Nicholas R. (1996), ‘The role of foreign trade and investment in China’s economic transformation’, in A.G. Walder (ed.), China’s Transitional Economy, Oxford: Clarendon Press, pp. 102–220. Lardy, Nicholas R. (2002), Integrating China in the Global Economy, Washington, DC: Brookings Institution Press. Larian B.F. (2000), ‘Capital flows, capital controls, and currency crises in Latin America’, in Felipe Larian B. (ed.), Capital Flows, Capital Controls, and Currency Crises in Latin America in the 1990s, Ann Arbor: University of Michigan Press, pp. 1–16. Le, B.L., T. Bui and D.V. Hung (2002), ‘FDI and development of manufacturing industries in Vietnam’, unpublished project report, Hanoi: Institute of World Economy. Lecraw, Donald (1977), ‘Direct investment by firms from less developed countries’, Oxford Economic Papers, 29 (3), 442–57. Lee, Joseph S. (1992), ‘Capital and labour mobility in Taiwan’, in Gustav Ranis (ed.), Taiwan: From Developing to Mature Economy, Boulder: Westview, pp. 305–55. Levine, Ross, A. Klevorick, R. Nelson and S. Winters (1987), ‘Appropriating the returns from industrial research and development’, Brooking Papers on Economic Activity, 3: 783–820. Levine, Ross and David Renelt (1992), ‘A sensitivity analysis of crosscountry growth regressions’, American Economic Review, 82 (4), 942–63. Lim, Linda and Pang Eng Fong (1991), Foreign Investment and Industrialisation in Malaysia, Singapore, Taiwan and Thailand, Paris: OECD.

References

265

Lin, Tzong-biau and Victor Mok (1985), ‘Trade, foreign investment, and development in Hong Kong’, in Walter Galenson (ed.), Foreign Trade and Investment: Economic Development in the Newly Industrializing Asian Countries, Madison, WI: University of Wisconsin Press, pp. 219–56. Lindblad, J. Thomas (1998), Foreign Investment in Southeast Asia in the Twentieth Century, London: Macmillan. Lipsey, Robert E. (1998), ‘Trade and production networks of US MNEs and exports by their Asian affiliates’, in John Dunning (ed.), Globalization, Trade and Foreign Direct Investment, Amsterdam: Elsevier Science, pp. 204–16. Lipsey, Robert E. (2000), ‘The role of foreign direct investment in international capital flows’, in Martin Feldstein (ed.), International Capital Flows, Chicago: University of Chicago Press, 307–30. Lipsey, Robert E. (2001), ‘Foreign investment in three financial crises’, NBER Working papers 8084, Cambridge, MA: National Bureau of Economic Research. Lipsey, Robert E. (2004), ‘Home and host country effects of FDI’, in Robert E. Baldwin and L. Alan Winters (eds), Challenges to Globalization: Analyzing the Economics, Chicago: Chicago University Press, pp. 333–82. Little, Ian M.D. (1982), Economic Development: Theory, Policy and International Relations, New York: Basic Books. Little, Ian M.D., Richard N. Cooper, W.M. Corden and S. Rajapatirana (1993), Boom, Crisis and Adjustment: The Macroeconomic Experiences of Developing Countries, Oxford: Oxford University Press. MacBean, Alasdair (ed.) (2000), Trade and Transition: Trade Promotion in Transitional Economies, London: Frank Cass. Mansfield, Edwin (1986), ‘The R&D tax credit and other technology policy issues’, American Economic Review, 76 (2), 190–91. Mansfield, Edwin, D. Teece and A. Romeo (1979), ‘Overseas research and development by US-based firms’, Economica, 46 (182), 187–96. Maskus, Keith (1998), ‘The role of intellectual property rights in encouraging foreign direct investment and technology transfers’, Duke Journal of Contemporary International Law, 9 (1), 109–61. Maskus, Keith (2000), Intellectual Property Rights in the Global Economy, Washington, DC: Institute for International Economics. Mataloni, Raymond J. (1997), ‘Real products of U.S. companies’ majorityowned foreign affiliates in manufacturing’, Survey of Current Business, 77 (4), 8–17. McKendrick, David G., Richard F. Doner and Stephen Haggard (2000), From Silicon Valley to Singapore: Location and Competitive Advantage in the Hard Disk Drive Industry, Stanford, CA: Stanford University Press.

266

Multinational enterprises in Asian development

McMillan, John and Christopher Woodruff (2002), ‘The central role of entrepreneurs in transition economics’, Journal of Economic Perspectives, 16 (9), 153–93. Michaely, Michael, Demetris Papageorgiou and Armeane M. Choksi (1991), Liberalizing Foreign Trade: Lessons of Experience in the Developing World, Oxford: Basil Blackwell. Moran, Thedore (1998), Foreign Direct Investment and Development, Washington, DC: Institute for International Economics. Morley, S.A. and Smith G.W. (1977), ‘The choice of technology: multinational firms in Brazil’, Economic Development and Cultural Change, 24 (2), 239–64. Naughton, Barry (1996), ‘China’s emergence and prospects as a trading nation’, Brookings Papers on Economic Activity, 2, 293–344. Naughton, Barry (ed.) (1999a), The China Circle: Economics and Technology in the PRC, Taiwan and Hong Kong, Washington, DC: Brookings Institution Press. Naughton, Barry (1999b), ‘The global electronics revolution and China’s technology policy’, NBR Analysis, 10 (2), 5–28. Naya, S. (1990), ‘Direct foreign investment and trade in East and Southeast Asia’, in Ronald W. Jones and Anne O. Krueger (eds), The Political Economy of International Trade, Oxford: Basil Blackwell, pp. 288–312. Nayyar, D. (1978), ‘Transnational corporations and manufactured exports from poor countries’, Economic Journal, 88, 59–84. Nelson, Richard R. and Howard Pack (1999), ‘The Asian miracle and modern growth theory’, Economic Journal, 109, 416–36. Ng, Francis and Alexander Yeats (2001), ‘Production sharing in East Asia: who does what for whom, and why?’, in Leonard K. Cheng and Henryk Kierzkowski (eds), Global Production and Trade in East Asia, Boston: Kluwer Academic Publishers, pp. 63–109. Ng, Francis and Alexander Yeats (2003), ‘Major trade trends in East Asia: what are their implications for regional cooperation and growth?’ Policy Research working paper 3084, Washington, DC: World Bank. Odagiri, Hiroyuki and Hedeto Yasuda (1996), ‘The determinants of overseas R&D by Japanese firms: an empirical study at the industry and company levels’, Research Policy, 25 (4), 1059–79. OECD (Organization for Economic Cooperation and Development) (1998), The Internationalization of Industrial R&D: Patterns and Trends, Paris: OECD. Pack, H. (1976), ‘The substitution of labor for capital in Kenyan manufacturing’, Economic Journal, 86, 45–58.

References

267

Pearce, Robert D. (1999), ‘Decentralised R&D and strategic competitiveness: globalised approaches to generation and use of technology in multinational enterprises’, Research Policy, 29 (2), 157–78. Petri, Peter (1993), ‘The East Asian trading bloc: an analytical history’, in Jeffrey A. Frankel and Miles Kahler (eds), Regionalism and Rivalry: Japan and the United States in Pacific Asia, Chicago: Chicago University Press, pp. 21–48. Pomfret, Richard (1991), Investing in China: Ten Years of the Open Door Policy, New York: Harvester Wheatsheaf. Pratten, C.F. (1976), Labour Productivity Differentials with International Companies, Cambridge: Cambridge University Press. Radelet, S. and J. Sachs (1997), ‘Asia’s reemergence’, Foreign Affairs, 76 (9), 44–59. Radelet, S. and J.D. Sachs (1998), ‘The East Asian financial crisis: diagnosis, remedies, prospects’, Brookings Papers on Economic Activity, 1, 1–74. Rahman, Mashur and Zaid Bakht (1997), ‘Constraints to industrial development: recent reforms and future directions’, in M.G. Quibria (ed.), The Bangladesh Economy in Transition, Delhi: Oxford University Press, pp. 77–114. Ramstetter, Eric D. (1993), ‘Prospects for foreign firms in developing economies in the Asian and Pacific region’, Asian Development Review, 11 (1), 151–85. Ramstetter, Eric D. (1997), ‘Thailand: international trade, multinational firms and regional integration’, in W. Dobson and Chia S.Y. (eds), Multinationals and East Asian Integration, Singapore: Institute of Southeast Asian Studies, pp. 107–30. Ramstetter, Eric D. (1999), ‘Comparison of foreign multinationals and local firms in Asian manufacturing over time’, Asian Economic Journal, 13 (2), 163–203. Ramstetter, Eric D. (2000), ‘Survey of recent developments’, Bulletin of Indonesian Economic Studies, 36 (3), 3–45. Rangan, Subramanian and Robert Z. Lawrence (1993), ‘The response of U.S. firms to exchange rate fluctuations: piercing the corporate veil’, Brookings Papers on Economic Activity, 1, 341–79. Rangan, Subramanian and Robert Z. Lawrence (1999), A Prism on Globalization, Washington, DC: Brookings Institution Press. Ranis, Gustav and Chi Schive (1985), ‘Direct foreign investment in Taiwan’s development’, in Walter Galenson (ed.), Foreign Trade and Investment: Economic Development in the Newly Industrializing Asian Countries, Madison, WI: University of Wisconsin Press, pp. 85–137.

268

Multinational enterprises in Asian development

Reisen, H. (2000), Pensions, Savings and Capital Flows: From Ageing to Emerging Markets, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Reserve Bank of India (2004), Monthly Bulletin, January, Table 46. Rhee, Yung W., Bruce Ross-Larrson and Gary Pursell (1984), Korea’s Competitive Edge: Managing the Entry into World Markets, Baltimore, MD: John Hopkins University Press. Riedel, James (1993), ‘Vietnam: on the trails of the tigers’, World Economy, 16 (4), 401–22. Riedel, James and Bruce Comer (1997), ‘Transition to a market economy in Vietnam’, in Wing Thye Woo, Stephen Parker and Jeffrey D. Sachs (eds), Economies in Transition: Comparing Asia and Eastern Europe, Cambridge, MA: MIT Press. Rodrik, Dani (1999), The New Global Economy and Developing Countries: Making Openness Work, Washington, DC: Overseas Development Council (distributed by Johns Hopkins University Press). Ronstadt, Robert (1977), Research and Development Abroad by U.S. Multinationals, New York: Praeger. Ruane, Frances and Holger Gorg (2001), ‘Globalization and fragmentation: evidence for the electronics industry in Ireland’, in Seven W. Arndt and Henryk Kierzkowski (eds), Fragmentation: New Production Patterns in the World Economy, New York: Oxford University Press, pp. 144–64. Sachs, Jeffrey and Felipe Larrian (1993), Macroeconomics in the Global Economy, New York: Harvester Wheatsheaf. Sachs, Jeffrey and Andrew Warner (1995), ‘Economic reforms and the process of global integration’, Brookings Papers on Economic Activity, 25th anniversary issue, 1–95. Sachs, Jeffrey, Aaron Tornell and A. Velasco (1996), ‘Financial crises in emerging markets: the lessons from 1995’, Brookings Papers on Economic Activity, 1, 147–215. Salter, W.E. (1959), ‘Internal and external balance: the role of prices and expenditure effects’, Economic Record, 35, 226–38. Saxenian, AnnaLee (2002), ‘Bangalore: the Silicon Valley of Asia?’, in Anne O. Krueger (ed.), Economic Policy Reforms and the Indian Economy, Delhi and Oxford: Oxford Universty Press, pp. 169–93. Schive, Chi (1990), The Foreign Factor: The Multinational Corporation’s Contribution to the Economic Modernization of the Republic of China, Standford, CA: Hoover Institution Press. Schive, Chi and Jenn-Hwa Tu (1991), ‘Foreign firms and structural change in Taiwan’, in Eric D. Ramstetter (ed.), Direct Foreign Investment in Asia’s Developing Economies and Structural Changes in the Asia-Pacific Region, Boulder, CO: Westview Press, pp. 142–71.

References

269

Sen, Amartya (1980), ‘Labor and technology’, in J. Cody, H. Hughes and D. Wall (eds), Policies for Industrial Progress in Developing Countries, New York: Oxford University Press, pp. 121–58. Serapio, M. and D. Dalton (1999), ‘Globalization of industrial R&D: an examination of foreign direct investment in R&D in the United States’, Research Policy, 28 (3), 303–16. Sharpton, Michael (1975), ‘International subcontracting’, Oxford Economic Papers, 27 (1), 94–135. Singer, Hans W. (1973), ‘International policy and its effect on employment’, in R. Jolly et al. (eds), Third World Employment, Harmondsworth: Penguin. SLDCS (Sri Lanka Department of Census and Statistics) (2003), Annual Survey of Industries 2001, Colombo: DCS. Snodgrass, Donald R. (1966), Ceylon: An Export Economy in Transition, Homewood, IL: Richard D. Irwin. Snodgrass, Donald R. (1999), ‘The economic development of Sri Lanka: a tale of missed opportunities’, in Robert I. Rotberg (ed.), Creating Peace in Sri Lanka: Civil War & Reconciliation, Washington, DC: Brookings Institution Press, pp. 89–107. Srinivasan, T.N. (1995), ‘Comment’, in Jeffrey Sachs and Andrew Warner (eds), Brookings Papers on Economic Activity, 25th Anniversary Issue, 103. Srinivasan, T.N. (1998), ‘India’s export performance: a comparative analysis’, in Isher J. Ahluwalia and I.M.D. Little (eds), India’s Economic Reforms and Development, Delhi: Oxford University Press, pp. 197–228. Stiglitz, J.E. (2000), ‘Capital market liberalization, economic growth, and instability’, World Development, 28 (6), 1075–86. Swan, Trevor (1960), ‘Economic control in a dependent economy’, Economic Record, 36 (1), 56–66. Tambunlertchai, Somsak and Eric D. Ramstetter (1991), ‘Foreign firms in promoted industries and structural change in Thailand’, in Eric D. Ramstetter (ed.), Direct Foreign Investment in Asia’s Developing Economies and Structural Change in the Asia-Pacific Region, Boulder, CO: Westview, pp. 65–102. Truong, David H.D. and Carolyn L. Gates (1996), ‘Vietnam in ASEAN – economic reform, openness and transformation: an overview’, ASEAN Economic Bulletin, 13 (2), 159–68. UNCTAD (United Nations Conference on Trade and Development) (1994), World Investment Report 1994: Transnational Corporations, Employment and Work Place, Geneva: United Nations. UNCTAD (United Nations Conference on Trade and Development) (1995), World Investment Report 1995: Transnational Corporations and Competitiveness, Geneva: United Nations.

270

Multinational enterprises in Asian development

UNCTAD (United Nations Conference on Trade and Development) (1996), World Investment Report 1996: Investment, Trade and International Policy Arrangement, Geneva: United Nations. UNCTAD (United Nations Conference on Trade and Development) (1998), The Financial Crisis in Asia and Foreign Direct Investment: An Assessment, Geneva: United Nations. UNCTAD (United Nations Conference on Trade and Development) (2001), Measurement of the Transnationalization of Economic Activity, Geneva: UNCTAD. UNCTAD (United Nations Conference on Trade and Development) (2002), World Investment Report 2002: Transnational Corporations and Export Competitiveness, Geneva: United Nations. UNCTAD (United Nations Conference on Trade and Development) (2005), World Investment Report: Transnational Corporations and the Internationalization of R&D, Geneva: UNCTAD. UNCTC (United Nations Centre on Transnational Corporations) (1983), ‘Salient features and trends in foreign direct investment’, New York and Geneva: United Nations. United Nations (1996), Incentives and Foreign Direct Investment, New York: United Nations. United States Bureau of Economic Analysis (1992), U.S. Direct Investment Abroad 1989 Benchmark Survey, Final Results, Washington, DC: United States Government Printing Office. Urata, Shujiro (2001), ‘Emergence of an FDI–trade nexus and economic growth in East Asia’, in Joseph E. Stiglitz and Hsaid Yususf (eds), Rethinking the East Asian Miracle, New York: Oxford University Press, pp. 409–59. US Department of Commerce (1975), U.S. Direct Investment Abroad, 1966: Final Data, Washington, DC: Government Printing Office. US Department of Commerce (1981), U.S. Direct Investment Abroad, 1977, Washington, DC: Government Printing Office. US Department of Commerce (1985), U.S. Direct Investment Abroad: 1982 Benchmark Survey Data, Washington, DC: Government Printing Office. US Department of Commerce (1992), U.S. Direct Investment Abroad: 1989 Benchmark Survey, Final Results, Washington, DC: Government Printing Office. USITC (United States International Trade Commission) (1999), Production Sharing: Use of U.S. Components and Material in Foreign Assembly Operations, 1995–1998, USITC Publication 3265, Washington, DC: USITC. Venables, Anthony J. (1999), ‘Fragmentation and multinational production’, European Economic Review, 43 (3), 935–45.

References

271

Vernon, Raymond (1962), ‘International investment and international trade in the product cycle’, Quarterly Journal of Economics, 80 (2), 190–207. Vernon, Raymond (1971), Sovereignty at Bay: The Multinational Spread of U.S. Enterprises, New York: Basic Books. Vernon, Raymond (1974), ‘The location of economic activity’, in John H. Dunning (ed.), Economic Analysis and the Multinational Enterprise, London: George Allen and Unwin, pp. 89–114. Vernon, Raymond (1977), Storm Over the Multinationals: The Real Issues, London and Basingstoke: Macmillan. Vernon, Raymond (2000), In the Hurricane’s Eye: The Troubled Prospects of Multinational Enterprises, Cambridge, MA: Harvard University Press. Wang, Wen-Theu (1999), ‘Foreign direct investment, spillovers and catching up: the case of Taiwan’, unpublished PhD dissertation, Australian National University. Warr, Peter G. (1986), ‘Indonesia’s other Dutch disease: economic effects of the petroleum boom’, in J.P. Neary and S. van Wijnbergen (eds), Natural Resources and the Macroeconomy, Oxford: Basil Blackwell, pp. 288–320. Warr, Peter G. (1990), ‘Free trade zones’, in Chris Milner (ed.), Export Promotion Strategies: Theory and Evidence from Developing Countries, London: Harvester. Warr, Peter G. (2006), ‘The Gregory thesis visits the tropics’, Economic Record, 82 (257), 177–94. Watanabe, S. (1981), ‘Multinational enterprises, employment and technology adaptations’, International Labor Review, 120, 693–704. Wei, Shang-Jin (2000), ‘Why does China attract so little foreign direct investment?, in Takatoshi Ito and Anne O. Krueger (eds), The Role of Foreign Direct Investment in East Asian Economic Development, Chicago: University of Chicago Press, pp. 239–65. Wells, L.T. (1977), ‘The internationalization of firms from developing countries’, in T. Agmor and C.P. Kindleberger (eds), Multinationals from Small Countries, Cambridge, MA: MIT Press, pp. 69–84. Wells, L.T. (1978), ‘Foreign investment from the third world: the experience of Chinese firms from Hong Kong’, Columbia Journal of World Business, 13 (1), 39–48. Wells, Luis T. (1983), The Third World Multinationals: The Rise of Foreign Investment from Developing Countries, Cambridge, MA: MIT Press. Wells, Luis T. (1994), ‘Mobile exporters: new foreign investors in East Asia’, in K.A. Foot (ed.), Foreign Direct Investment, Chicago: University of Chicago Press, pp. 173–91. Wells, L.T. and V. Warren (1979), ‘Developing country investors in Indonesia’, Bulletin of Indonesian Economic Studies, 15 (1), 69–84.

272

Multinational enterprises in Asian development

Westphal, Larry E., Y.W. Rhee and G. Pursell (1979), ‘Foreign influence on Korean industrial development’, Oxford Bulletin of Economics and Statistics, 41 (4), 359–88. White, H. (1980), ‘A heteroscedasticity consistent covariance matrix estimator and a direct test of heteroscedasticity’, Econometrica, 48, 817–38. Williamson, John and M. Mahar (1998), ‘A survey of financial liberalization’, Essays in International Finance, no. 211, Princeton, NJ: Department of Economics, Princeton University. Wooldridge, Jeffrey M. (2002), Econometric Analysis of Cross Section and Panel Data, Cambridge, MA: MIT Press. World Bank (2002), Vietnam’s Exports: Policies and Prospects, Hanoi: World Bank Vietnam (draft report). World Bank (2003a), ‘Trade policies in South Asia: an overview’, Poverty Reduction and Economic Management, South Asia Region, Washington, DC: World Bank. World Bank (2003b), India: Sustaining Reform, Reducing Poverty, Washington, DC: World Bank. World Bank (2004), Sri Lanka Development Policy Review, Washington, DC: World Bank. Yeats, Alexander (2001), ‘Just how big is global production sharing?’, in Seven W. Arndt and Henryk Kierzkowski (eds), Fragmentation: New Production Patterns in the World Economy, New York: Oxford University Press, pp. 108–43. Yeung, Henry Wai-chung (1999), ‘Competition in the global economy: the globalization of business firms from emerging economies’, in Henry Wai-chung (ed.), The Globalization of Business Firms from Emerging Economies, Cheltenham, UK and Northampton, MA, USA: Edward Elgar, vol. 1, pp. ix–xiii. Yi, Kei-Mu (2003), ‘Can vertical specialization explain the growth of world trade?’, Journal of Political Economy, 111 (1), 52–102. Yoffie, David B. (1993),‘Foreign direct investment in semiconductors’, in Kenneth A. Froot (ed.), Foreign Direct Investment, Chicago: University of Chicago Press, pp. 197–230. Yusuf, Shahid (2003), Innovative East Asia: The Future of Growth, New York: Oxford University Press. Zejan, M.C. (1990), ‘R&D activities in affiliates of Swedish multinational enterprises’, Scandinavian Journal of Economics, 92 (3), 487–500.

Index affiliates of MNEs 47, 64, 79, 102 in China 69 East Asian crisis, role in recovery FDI as percentage of GDI 112, 113 Japanese affiliates 115–16 Malaysian manufacturing 116, 117–18, 119 US affiliates, employment 115 US affiliates, exports 114 in high-tech industries 53 in India 64 in Indonesian manufacturing 116 manufacturing exports share 55–6, 57–61 R&D activities 192 see also multinational enterprises AFTA (ASEAN Free Trade Area) 88 components trade 77, 80–84 agro-based processed food 50 annual export growth 54, 55–6 Argentina 125, 130 see also Latin America ASEAN countries components trade 93 exports to china 37 FDI industries 39 FDI inflows 33–4 foreign investment regime 16–20 ASEAN Free Trade Area (AFTA) 88 components trade 77, 80–84 Asian financial crisis 14, 17, 19, 229 capital flows 102–12 Asian crisis countries 104–5 mergers and acquisitions 109 net capital flows in Asian countries 106–8 net capital flows percentage change 109 US direct investment 110, 111 crisis management policy 14, 17 FDI, effect on 33–4

recovery, role of affiliates FDI as percentage of GDI 112, 113 Japanese affiliates 115–16 Malaysian manufacturing 116, 117–18, 119 US affiliates, employment 115 US affiliates, exports 114 assembly activities 51, 52–3, 58–60, 78, 163 Association of South East Asian Nations see ASEAN countries Australian model of a small economy 121, 143–5 automation of production processes 52 Balasubramanyam, V.N. 148 Bangladesh FDI industries 41 foreign investment regime 24 bank lending 103, 108, 119, 122, 126, 127–8 banking 39 Bhagwati, Jagdish 147, 164 Blalock, Garrick 116 Brecher, Richard A. 147 business licensing 13 Cambodia Chinese investment 39 FDI industries 40 FDI inflows 33 foreign investment regime 19 Canada 178 capital controls, Malaysia 18 capital inflow–real exchange rate, comparative analysis 123–41 capital flow–real exchange nexus 128–33 capital inflows, trends and patterns 124–8 conclusion 140–41 empirical analysis 133–40 273

274

Multinational enterprises in Asian development

capital inflows 121–3, 143–4 capital intensity in Sri Lankan manufacturing, analysis 198–217 data 210–11 foreign direct investment 200–208 methodology 208–11 results 211–17 cascading tariff structures 49 Caves, R.E. 68 centrifugal factors of R&D location decision 169–70 centripetal factors of R&D location decision 169, 172 chemical industries 49 China 59 components trade 77, 78, 84, 97 Cultural Revolution 63 FDI industries 40 FDI inflows 32–3, 35–7 final assembly 68 foreign investment regime 20–21 and India, comparison 64 manufactured exports 64 MNE affiliates 58, 69 as overseas investor 37, 39, 42 pre-revolution industrialization 50 trade restrictions by EU and USA 39 ‘China fear’ 35–9 clothing industry see garments industrycommunication innovations 51–2, 53 components trade (parts and components) 77–9, 84, 93 assembly 40, 51, 52, 58–60 cross-border 79, 93 developed countries 77 developing East Asia and ASEAN 77–9, 93 exports and imports, percentage 85–7 exports by major country groups 78 high-tech industries 37 intra-regional trade 93 in manufacturing trade 81–3 production 51, 59, 63–4 regional manufacturing trade flows 94, 95 regional patterns 89–92 two-way trade 79

UN trade database 74–5 world trade 52, 75–7, 78–9 consumer durables production 62 Corden, W.M. 121 country reputation 67 country risk 61–2 crisis management packages 102 cross-border components trade 79, 93 cross-border mergers and acquisitions (M&As) 14, 102, 108, 112, 223 cross-border production activities 41, 53 Cultural Revolution, China 63 currency collapse 102 currency crises 120, 141 DCMNEs (developed countries multinational enterprises) 198–9, 217 developed countries components trade 77 multinational enterprises (DCMNEs) 198–9, 217 R&D locations of US MNEs 173, 178 Diaz-Alejandro, Carlos F. 147 direct performance indicators 2, 12 Doingbusiness database 223–4 domestic credit contraction 102 domestic demand 62 domestic-market-oriented production 17, 39 domestic technological competency 181–2 domestic wages 53, 59 double counting of trade flows 97 East Asia average manufacturing wage 79, 80 components trade 77, 78, 80, 84 East Asia, newly industrialized countries (NIEs) 53, 54, 67 East Asian financial crisis 14, 17, 19, 229 capital flows 102–12 Asian crisis countries 104–5 mergers and acquisitions 109 net capital flows in Asian countries 106–8

Index net capital flows percentage change 109 US direct investment 110, 111 crisis management policy 14, 17 FDI, effect on 33–4 recovery, role of affiliates FDI as percentage of GDI 112, 113 Japanese affiliates 115–16 Malaysian manufacturing 116, 117–18, 119 US affiliates, employment 115 US affiliates, exports 114 Eastern Europe, liberalization reforms 36 economic adjustment in small open economy 143–4 Economic Freedom Index 223, 225 economic integration of Europe 79 efficiency-seeking investment 42, 47, 49 efficiency-seeking production 39–40 electronics industry 37, 39, 40, 51, 53 assembly activities 163 components trade 84 location decisions 62, 63, 206 revolution in Singapore 63 entrepot trade, Hong Kong 13, 50, 99 entrepreneurial talents 50 EOFDI see export-oriented foreign direct investment EPZs see export processing zones equity capital 26, 103 ethnic conflict, Sri Lanka 201, 202, 206 European multinational enterprises 52–3 European Union (EU) components trade 78 final goods exports 95 intra-regional trade shares 95 trade restrictions on China 39 exchange rate depreciation 102 see also real exchange rate export expansion 49–50, 52, 54, 55–6 role of MNEs 62–7 export incentives 48–9 export-led industrialization 49–50, 67 export of manufactured goods, analysis commodity composition 57–62

275

conclusion 67–8 data sources 70–71 evidence of participation 54–62 export expansion, role in 62–7 typology 47–53 export-oriented affiliates 115–16 export-oriented economies 17, 21, 147 see also import-substitution export-oriented foreign direct investment (EOFDI) 18, 45, 49, 60, 67, 202 China 21 India 22 Sri Lanka 24–5 export-oriented investors 47, 64 export-oriented production 39–40, 41, 49 export performance 54, 55–6 export processing zones (EPZs) 19, 20, 24, 221 Sri Lanka 200, 201, 218 Taiwan 15 Vietnam 248–9 export-promotion (EP) regime, China 21 export success of latecomers 64 exports by US MNE affiliates 114 external economic shocks 141 extra-regional trade 88, 93, 97 FDI see foreign direct investment FIEs see foreign-invested enterprises (FIEs) final assembly 52, 58–60, 68, 97 final goods exports 94–5 final trade 73, 78, 97 financial crisis 101–2 financial crisis, East Asia 14, 17, 19, 229 capital flows 102–12 Asian crisis countries 104–5 mergers and acquisitions 109 net capital flows in Asian countries 106–8 net capital flows percentage change 109 US direct investment 110, 111 crisis management policy 14, 17 FDI, effect on 33–4

276

Multinational enterprises in Asian development

recovery, role of affiliates FDI as percentage of GDI 112, 113 Japanese affiliates 115–16 Malaysian manufacturing 116, 117–18, 119 US affiliates, employment 115 US affiliates, exports 114 financial services 39 Findlay, Ronald 147 Finland MNE affiliates 178 fiscal contraction 134, 139, 141 for export dynamism 80 foreign direct investment (FDI) 11, 13, 103, 122–3, 126 data limitations 26, 36 definition 26 domestic-market-oriented activities 202 in economic transition, Vietnam case study business environment 223–5 conclusions and policy inferences 246–7 FDI impact 234–46 FDI policy 220–23 FDI trends and patterns 225–34 indicators of ease of doing business 226–7 industry composition 231, 233–4 ownership structure 230 productivity growth 243–6 source country composition 231, 232 spatial distribution 234, 235 global contraction 34 global investment patterns 35–6 inflows 101, 107–10, 122 inflows, comparative data (19802004) 26–34 developing Asian countries in global context 27–9 developing countries and developing Asia 32 as percentage of gross domestic fixed capital formation 30–31 liberalization 39, 45 foreign-invested enterprises (FIEs) 40, 202 Vietnam 235–46

foreign investment regimes 13–26 ASEAN countries 16–20 Bangladesh 24 Cambodia 19 China 20–21 Hong Kong 13 India 21–3, 64 Indonesia 16, 17–18 Korea 13–15 Lao PDR 19–20 Malaysia 16, 17, 18, 110 Myanmar 20 Nepal 25 Pakistan 23 Philippines 16, 17, 18 South Asia 21–6 Sri Lanka 21, 24–5 Taiwan 15–16 Thailand 16 Vietnam 18–19, 220–23 foreign ownership 17 India 22 in Korea 14, 15 of land 14, 15 of land and real estate properties 15 Malaysia 18 fragmentation-based specialization 97, 98 fragmentation trade 53, 73, 98 see also product fragmentation free trade agreements (FTAs) 97, 98 free trade zones (FTZs) 22 Fukao, Kyoji 115 garments industry 39, 40, 67, 206, 217 Germany R&D activities of US MNEs 178 R&D expenditure 172 Gertler, Paul J. 116 global assembly operations 52 global components trade 52, 75–7 global foreign direct investment flows 34 global industrial networks 41, 119 global innovatory centres 170 global investment patterns 35–6 Goh, Keng Swee 63 government expenditure 134, 139 growth dynamism 88, 97, 98

Index Harris Corporation 62, 206 Harvard Multinational Enterprise Project 164 headquarter R&D operations 169, 171 herd mentality 62, 206 high-tech industries 37, 41, 52, 53, 59 high-wage countries 78 Ho Chi Minh City 40 Hong Kong 50, 53, 63, 64 components trade 78 entrepot trade 13, 50, 99 foreign investment regime 13 investment in China 35 IMF (International Monetary Fund) 17 data sources 120, 136 import-substitution 13, 17, 40, 46 Bangladesh 24 economic failure of 147 growth strategy 50 industrialization strategy 21 Nepal 25 Sri Lanka 24 see also export-oriented economies independent R&D 170 Index of Economic Freedom 223, 225 India and China comparison 64 FDI industries 40–41 FDI inflows 34 FDI, potential for 42–3 foreign investment regime 21–3, 64 liberalization reforms 64 MNE affiliates 64 Indonesia 62, 63 capital flows 103, 105, 106 FDI, efficiency-seeking 40 FDI inflows 33, 34, 109 foreign investment regime 16, 17–18 mergers and acquisitions 108 MNEs affiliates 116 industry distribution of R&D expenditure 173, 174 industry profile in developing Asia 39–41 innovatory centres 178 innovatory process 169 Inspector Raj 23 intangible assets 101

277

Intel Corporation 37, 247 intellectual property rights 22 inter-company debt 103 inter-country wage differentials 79, 80, 97 international division of labour 51 International Monetary Fund (IMF) 17 data sources 120, 136 international product fragmentation 51–2, 72, 79, 80, 95, 98 trade patterns 84–97 direction of components trade 89–92 share of components in manufacturing trade flows 94, 95 total manufacturing, components and final trade 96 trends and patterns 75–84 components export by categories 85–7 components exports by country groups 78 components in manufacturing trade 79–80, 81–3 developed world components trade 77 East Asia 77, 78–9 manufacturing wage, East Asia 79, 80 NAFTA components trade 79 world components trade 75–7 international production, productivity growth analytical framework 149–53 conclusion and policy inferences 161–3 correlation matrix 160 data 153–5 productivity determinants 156–61 productivity growth determinants, regression results 159, 162 productivity patterns 156 productivity and related indicators 157–8 total factor productivity growth measurement 166–7 international specialization 122 internationalization of R&D 172–3

278

Multinational enterprises in Asian development

Internet 53 intra-company loans 26 intra-regional FDI inflows 110 intra-regional import flows 88 intra-regional trade 41, 88 components 93, 94, 95 final manufacturing 93–5 in global network 53 in manufacturing in East Asia 93 patterns 84–97 product-sharing system 79 specialization 79 total manufacturing, components, and final trade 96 investment inflows, China 32–3 Ireland, as R&D location 178 Israel, as R&D location 178 Japan 88 affiliates in crisis-hit countries 115–16 car producers 72 components trade 77, 78 FDI inflows to crisis-hit countries 112 final good exports 95 fragmentation trade 88 multinational enterprises 52–3 overseas assembly activities 78 US affiliates, R&D intensity 178 Japanese occupation 50 Joint Venture Law (1979), China 20 joint ventures 14, 15, 21, 23, 221, 230 Kaohsiung, Taiwan 15 Katunayake Export Processing Zone (KEPZ) 62, 218 Khmer Rouge 19 knowledge-seeking investment 178 knowledge-seeking (overseas) R&D 170–71 knowledge spillover 170 Korea 50, 84, 93, 110 capital flows 103, 105–6 export expansion 62–3 FDI flows 34, 108–9 FDI industries 39 Five Year Plan (1962) 14 Foreign Investment Promotion act 14, 15

foreign investment regime 13–15 mergers and acquisitions 14, 15, 108 labour intensity 117 labour-intensive production 40, 52, 67 assembly locations 63 exports 41, 50–51 final consumables 49, 50 manufacturing 202 labour productivity growth 247 ladder effect in international specialization 122 Lao PDR FDI industries 40 FDI inflows 33 foreign investment regime 19–20 latecomers to manufacturing for export 50, 53 share of exports 64, 67, 68 Latin America capital inflow–real exchange rate, comparative analysis 123–41 capital flow–real exchange rate nexus 128–33 capital inflow, trends and patterns 124–8 conclusion 140–41 empirical analysis 133–40 debt crisis 119 foreign direct investment 69 R&D activities of US MNEs 173 liberalization reforms 163 Eastern Europe 36 India 64 Sri Lanka 200–202 Taiwan 15 Licence Raj 23 life-cycle investors 47 light manufacturing export sector 13 liquidity insurance 116 location decisions 47 assembly activities 79 electronics firms 62, 63, 206 final assembly 52 R&D 169–71, 181 low-wage countries 78 M&As see mergers and acquisitions Malaysia 58, 59, 63, 80, 93 capital flows 103, 106, 107

Index FDI flows 33, 34, 109–10 FDI industries 40 foreign investment regime 16, 17, 18, 110 high-tech industries 37, 40, 41 mergers and acquisitions 108 MNE manufacturing, post-crisis 116, 117–18, 119 manufacturing for export, analysis commodity composition 57–62 conclusion 67–8 data sources 70–71 evidence of participation 54–62 export expansion, role in 62–7 typology 47–53 manufacturing wages 79, 80 market-seeking investment 42, 47 mergers and acquisitions by foreign firms in Asian crisis 109 mergers and acquisitions (M&As) of domestic financial institutions, Korea 15 Mexico 79 financial crisis 119, 125 see also Latin America MFA (Multi-Fibre Arrangement) 51, 202 middle products see components trade MNEs see multinational enterprises Motorola 62, 206 Multi-Fibre Arrangement (MFA) 51, 202 multinational enterprise affiliates 47, 64, 79, 102 in China 69 East Asian crisis, role in recovery FDI as percentage of GDI 112, 113 Japanese affiliates 115–16 Malaysian manufacturing 116, 117–18, 119 US affiliates, employment 115 US affiliates, exports 114 in high-tech industries 53 in India 64 in Indonesian manufacturing 116 manufacturing exports share 55–6, 57–61 R&D activities 192 see also multinational enterprises

279

multinational enterprise subsidiaries see multinational enterprise affiliates multinational enterprises 1–3, 6–10 economic development, role in 3–6 see also multinational enterprise affiliates Myanmar 20 National Semiconductors 63 nationalization 23, 24 Nayyar, D. 54 negative list of investment approval 17 Nepal, foreign investment regime 25–6 NIEs (newly industrialized economies) 53, 54, 67 nominal exchange rate 121 nominal exchange rate adjustment 134, 135, 140, 141 non-sterilized intervention 134 non-tradability/tradability 141 non-tradables/tradables 121, 143–4 North American Free Trade Agreement (NAFTA) 36, 80, 88 components trade 79 intra-regional trade shares 95 obsolescence 52 offshore assembly 52 oil price increases 124 openness to trade 135–6 ‘other’ capital flows 139 outsourcing 51, 52–3 see also product fragmentation overseas R&D operations 171 expenditure 172–3 ownership restrictions 17 Pakistan, foreign investment regime 23 parts and components see components trade performance indicators 2 Philippines 58, 63, 80, 93 capital flows 103, 106, 107 components trade 78–9 FDI industries 40 FDI inflows 33–4 foreign investment regime 16, 17, 18 mergers and acquisitions 108

280

Multinational enterprises in Asian development

pioneer status 16, 17 policy reforms in response to crisis 120 political instability 61 portfolio investment 103, 108, 119, 122, 126, 127–8 present-day transition economies 50 private capital flows 103 private foreign investment (PFI) see foreign direct investment process patents 22 product adaptation 170, 180–81 product composition (of MNE-related exports) 57–62 product fragmentation 51–2, 72, 79, 80, 95, 98 trade patterns 84–97 direction of components trade 89–92 share of components in manufacturing trade flows 94, 95 total manufacturing, components and final trade 96 trends and patterns 75–84 components export by categories 85–7 components exports by country groups 78 components in manufacturing trade 79–80, 81–3 developed world components trade 77 East Asia 77, 78–9 manufacturing wage, East Asia 79, 80 NAFTA components trade 79 world components trade 75–7 product patents 22 production technology, advances in 51 productivity gains from international production analytical framework 149–53 conclusion and policy inferences 161–3 correlation matrix 160 data 153–5 productivity determinants 156–61 productivity growth determinants, regression results 159, 162 productivity patterns 156

productivity and related indicators 157–8 total factor productivity growth measurement 166–7 property rights 49 quantitative import restrictions (QRs) 51 real exchange rate 66–7, 121–2, 143–4 measurement of 145–6 real exchange rate–capital inflow, comparative analysis 123–41 capital flow–real exchange nexus 128–33 capital inflows, trends and patterns 124–8 conclusion 140–41 empirical analysis 133–40 real exchange rate problem 121, 139, 140 regional division of labour 53 regional integration in East Asia 88 regional production-sharing networks 79 regional trading agreements (RTAs) 49 reinvested earnings 26, 36–7, 103, 110, 120 renovation (doi moi) reforms, Vietnam 18–19 Republic of Korea see Korea RER see real exchange rate research and development (R&D) as headquarter function 169, 171 location decisions 169–71 research and development (R&D) globalization, US firms data 185, 195–6 expenditure 173, 174–7, 179–80 model 180–85 R&D intensity determinants 180–91 domestic technological competency 181–2 investment environment 182 other variables 182–5 product adaptation 180–81 regression results 186–91, 197 theoretical framework 169–71 trends and patterns 171–80 resource-based manufacturing 49–50

Index resource-based processing activities 62 resource-rich developing countries 49 retained earnings 26, 36–7, 103, 110, 120 round-tripping capital 35 RTAs see regional trading agreements rules of origin (ROOs) 98 Sachs, Jeffrey 136, 142 seafood 50 second-tier exporting countries 58, 63 semiconductors 63, 84 services sector 39, 41 SEZ see special economic zones short-term capital flows 123–4 Singapore 63, 80, 93 assembly activities 58 components trade 78–9 efficiency-seeking production 39 entrepreneurial background 50 FDI industries 39–40 FDI inflows 33 foreign investment regime 16 high-tech industries 39, 41, 53, 59, 63 site selection 47 assembly activities 79 electronics firms 62, 63, 206 final assembly 52 R&D 169–71, 181 slicing the value chain see product fragmentation small open economy, economic adjustment 143–4 software industry 41, 42–3, 44, 74 South Korea 63, 64, 126, 202 see also Korea special economic zones (SEZs) 19, 20–21 spillovers 2–3 Sri Lanka 21, 60–62 capital intensity in manufacturing, analysis 198–217 data 210–11 FDI 200–208 methodology 208–11 results 211–17 ethnic conflict 201, 202, 206 export processing zones (EPZs) 200, 201, 218

281

export trade 205–8 FDI industries 41 FDI inflows 201–2 foreign investment enterprises (FIEs) 202 foreign investment regime 21, 24–5 liberalization reforms 200–202 standard fragments 68 standardization of components 53, 68 standardized products 51 sterilized intervention 134–5, 140, 141 subsidiary firms see multinational enterprise affiliates Sweden multinational enterprise affiliates 178 R&D expenditure 172 Taiwan 50, 53, 63, 64, 93 components in total manufacturing exports 84 export expansion 62 export processing zones (EPZs) 15 FDI industries 39 foreign investment regime 15–16 investment in China 35 tariff changes 97–100 tariff-jumping 47, 64 tariff protection 23 tax holidays 16, 25, 221 tax incentives 14, 20, 220–21 technological innovations 51–2, 53 technology effort index 181–2 technology generation 168 technology licensing 21, 23 technology sourcing centres 171 technology spillovers 170 technology transmission 168 Texas Instruments 63 Thailand 59, 62, 63, 80, 93, 110 capital flows 103, 108 FDI flows 33, 34, 109 FDI industries 40 foreign investment regime 16, 17 mergers and acquisitions 108 negative inflows 105 third world multinational enterprises (TWMNEs) 198–9, 217 Tiananmen Square 32 tradability/non-tradability 141 tradables/non-tradables 121, 143–4

282

Multinational enterprises in Asian development

trade flow analysis 72, 97 trade policy regimes 142 trade restrictions 147, 163 on China 39 transitional economies 219 transportation, technical innovations 51–2 TWMNEs (third world multinational enterprises) 198–9, 217 UK R&D activities of US MNEs 178 R&D overseas expenditure 173 UN trade database 74–5 US Annual Survey of US Direct Investment Abroad 113 components trade 78 direct investment in Asian countries 38, 110, 111 electronics industry 51, 62 exports in crisis-hit countries 113 FDI inflows 33 final goods exports 95 manufacturing, employment share 115 manufacturing, productivity and indicators 157–8 R&D globalization study data 185, 195–6 domestic technological competency 181–2 expenditure 173, 174–7, 179–80 investment environment 182 model 180–85 other variables 182–5 product adaptation 180–81 R&D intensity determinants 180–91 regression results 186–91, 197 theoretical framework 169–71 trends and patterns 171–80 trade restrictions on China 39 vertical specialization see product fragmentation vertically integrated assembly industries 61, 206 vertically integrated high-tech industries 37, 42

vertically integrated international industries 51 Vietnam components trade 37, 78 ease of doing business indicators 226–7 export expansion 240–41, 242 export processing zones (EPZs) 248–9 FDI in economic transition, case study business environment 223–5 conclusions and policy inferences 246–7 FDI impact 234–46 FDI policy 220–23 FDI trends and patterns 225–34 indicators of ease of doing business 226–7 industry composition 231, 233–4 ownership structure 230 productivity growth 243–6 source country composition 231, 232 spatial distribution 234, 235 FDI industries 40 FDI inflows 33 foreign-invested enterprises (FIEs) 235–46 foreign investment regime 18–19, 220–23 investment data 225–8, 234–5 Investment Law (2006) 222–3 joint ventures 230 wages 42, 51, 53, 59, 79 differentials in East Asia 79, 80, 97 Warner, Andrew 136, 142 Western European multinational enterprises 52–3 World Bank database 136 world components trade 52, 75–7 World Economic Forum 223, 224 world manufactured exports 55–6 world manufacturing exports 54 World Trade Organization (WTO) 14, 16, 49 worldwide procurement networks 72 Yeats, Alexander 68