East Asia's Economic Integration: Progress and Benefit

East Asia’s Economic Integration Progress and Benefit

Edited by

Daisuke Hiratsuka and Fukunari Kimura

East Asia’s Economic Integration

Other titles from IDE-JETRO: Mariko Watanabe (editor) Recovering Financial Systems China and Asian Transition Economies Daisuke Hiratsuka (editor) East Asia’s De Facto Economic Integration Hisayuki Mitsuo (editor) New Developments of the Exchange Rate Regimes in Developing Countries Tadayoshi Terao and Kenji Otsuka (editors) Development of Environmental Policy in Japan and Asian Countries Masahisa Fujita (editor) Regional Integration in East Asia From the Viewpoint of Spatial Economics Akifumi Kuchiki and Masatsugu Tsuji (editors) Industrial Clusters in Asia Analyses of Their Competition and Cooperation Mayumi Murayama (editor) Gender and Development The Japanese Experience in Comparative Perspective Nobuhiro Okamoto and Takeo Ihara (editors) Spatial Structure and Regional Development in China An Interregional Input-Output Approach Akifumi Kuchiki and Masatsugu Tsuji (editors) The Flowchart Approach to Industrial Cluster Policy Masahisa Fujita (editor) Economic Integration in Asia and India

East Asia’s Economic Integration Progress and Benefit Edited by Daisuke Hiratsuka and Fukunari Kimura

ª Institute of Developing Economies (IDE), JETRO 2008 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No portion of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, Saffron House, 6-10 Kirby Street, London EC1N 8TS. Any person who does any unauthorized act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The authors have asserted their rights to be identified as the authors of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2008 by PALGRAVE MACMILLAN Palgrave Macmillan in the UK is an imprint of Macmillan Publishers Limited, registered in England, company number 785998, of Houndmills, Basingstoke, Hampshire RG21 6XS. Palgrave Macmillan in the US is a division of St Martin’s Press LLC, 175 Fifth Avenue, New York, NY 10010. Palgrave Macmillan is the global academic imprint of the above companies and has companies and representatives throughout the world. Palgrave and Macmillan are registered trademarks in the United States, the United Kingdom, Europe and other countries. ISBN-13: 978–0–230–55362–0 ISBN-10: 0–230–55362–1

hardback hardback

This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. Logging, pulping and manufacturing processes are expected to conform to the environmental regulations of the country of origin. A catalogue record for this book is available from the British Library. A catalog record for this book is available from the Library of Congress. 10 9 8 7 6 5 4 3 2 1 17 16 15 14 13 12 11 10 09 08 Printed and bound in Great Britain by CPI Antony Rowe, Chippenham and Eastbourne

Contents List of Figures List of Tables Notes on the Contributors Preface

1

ix xii xv xvi

From De Facto to De Jure Economic Integration in East Asia: Past, Present and Future Daisuke Hiratsuka and Fukunari Kimura 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8

Introduction De facto economic integration The problems of de jure integration in East Asia Evolution of a core–periphery structure Growing indigenous enterprises with MNEs Benefit from economic integration Policy implications of individual chapters Conclusion

1 1 3 6 11 13 14 16 22

Part I: The Making of Economic Integration in East Asia

27

2

The Modality of East Asia’s Economic Integration Fukunari Kimura

29

2.1 2.2 2.3 2.4

Forming a basic architecture of economic integration De facto and de jure economic integration in East Asia De jure economic integration in terms of trade in goods De jure economic integration in terms of policy measures other than tariffs 2.5 Geographical boundary of de jure plurilateral integration 2.6 Concluding remarks

29 30 32

The East Asian Noodle Bowl Syndrome Richard E. Baldwin

45

3.1 Introduction 3.2 Current state of play

45 47

3

v

36 38 40

vi Contents

3.3 Fragility and emerging tensions 3.4 New East Asian Regional Management Effort (NEARME) 3.5 Concluding remarks 4

Rules of Origin, Local Content and Cumulative Local Content in East Asia: Application of an International Input–Output Analysis Ikuo Kuroiwa 4.1 4.2 4.3 4.4 4.5 4.6

5

Introduction Rules of origin Calculation of local content Empirical results Sector analysis Conclusion

Making Sense of the Timeliness of Transportation in Economic Integration Nobuaki Hamaguchi 5.1 Introduction 5.2 Development of timely transportation and revealed geographical advantage 5.3 The timeliness issue in Japanese trade in East Asia 5.4 Concluding remarks

64 72 78

82 82 83 85 89 101 109

118 118 122 124 137

Part II: The Evolution of the Core–Periphery Structure

141

6

The Evolution of Core–Periphery Structure in East Asia Nobuaki Hamaguchi

143

6.1 Introduction 6.2 The perspective of new economic geography analysis for regional integration 6.3 Regional integration in East Asia 6.4 Perspective of the economic integration in East Asia

143

7

144 148 158

Industrialization through Vertical Production Networks: Can Laos Participate? Daisuke Hiratsuka, Souknilanh Keola and Motoyoshi Suzuki

162

7.1 Introduction 7.2 Industrialization through vertical production networks

162 164

Contents vii

7.3 Transport costs 7.4 Potential of industrialization through vertical production networks in Laos 7.5 High fixed cost and border-related barriers in Laos 7.6 Case study of high border-related costs in Laos and their effect on its participation in vertical production networks 7.7 Summary and conclusion 8

Structural Change in Intermediate, Consumption and Capital Goods Trade During Economic Integration: the EU Experience Bart Los and Jan Oosterhaven 8.1 8.2 8.3 8.4

166 168 170

172 179

184

Introduction EU integration and trade and location theory EU inter-country input–output tables and model Earlier inter-country input–output research on EU integration 8.5 Trends in European trade, 1975–95 8.6 Summary and conclusion

192 203 215

Part III: The Current Status of Indigenous Enterprises

223

9

Spillovers and Linkages between Local and Foreign Plants Kazuhiko Yokota

225

9.1 Introduction 9.2 Industrial structure and local–multinational linkages in Thailand 9.3 Spillover and linkage effects 9.4 Further discussion on spillovers and demand linkage effects, and trade liberalization 9.5 FTA and its impact on local plants 9.6 Conclusion

225

10 Do Local Suppliers Benefit from Economic Integration? Evidence from Outward FDI from ASEAN Daisuke Hiratsuka 10.1 Introduction 10.2 What determines outward FDI? 10.3 Pattern of outward FDI from ASEAN

184 186 189

228 231 242 249 251

255 255 257 262

viii Contents

10.4 Intra-regional FDI in ASEAN 10.5 Conclusion and policy implications Part IV: Quantitative Implications of Economic Integration 11 Comparing Bilateral and Multilateral ASEAN10þ4 Free Trade Agreements: Possible Impacts on Member and Non-member Countries Michel Fouquin 11.1 11.2 11.3 11.4

Introduction The open regionalism concept and AFTA Quantitative model and possible scenario ASEAN10 bilateral FTAs with Japan, China, Korea and India 11.5 An Asian single market? 12 How Will ASEANþ3 Integration Accelerate Investment? A CGE Analysis Ken Itakura 12.1 12.2 12.3 12.4 12.5 Index

Introduction Investment theory in the Dynamic GTAP model Data aggregation and simulation design Simulation results Concluding remarks

268 277 281

283 283 284 290 293 305

321 321 322 324 328 334 339

List of Figures 2.1 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4.1(a) 4.1(b) 4.1(c) 4.1(d) 4.2(a) 4.2(b) 4.2(c) 4.2(d) 4.2(e) 4.3(a) 4.3(b) 4.3(c)

Fragmentation: an illustration The East Asian ‘noodle bowl’ syndrome Placement of Japanese plants in East Asia, 1975–2004; emergence of China Unilateral tariff cutting in Phase II Formal unilateralism; reductions in applied MFN tariffs AFTA utilization rates Intra-ASEAN trade by HS Chapter, 2003 Total manufacturing growth, 1990–2002 An example of interdependence in ‘Factory Asia’ East Asian bilateral trade flows rounded to nearest 1% of regional flows Changes in local content (1990–2000): manufacturing sector Changes in import content (1990–2000): manufacturing sector Impact of trade structural changes (1990–2000): manufacturing sector Impact of technological structural changes (1990–2000): manufacturing sector Changes in local content (1990–2000): manufacturing sector Changes in import content (1990–2000): manufacturing sector Impact of value added coefficient changes (1990–2000): manufacturing sector Impact of trade structural changes (1990–2000): manufacturing sector Impact of technological structural changes (1990–2000): manufacturing sector Changes in local content (1990–2000): electronics sector Changes in import content (1990–2000): electronics sector Impact of trade structural changes (1990–2000): electronics sector

ix

36 48 51 56 56 58 59 62 66 72 95 95 96 96 99 99 99 100 100 102 103 103

x List of Figures

4.3(d) 4.4(a) 4.4(b) 4.4(c) 4.4(d) 5.1 5.2 5.3 6.1 6.2 6.3 6.4 6.5 7.1 7.2 7.3 7.4 7.5 8.1 8.2 8.3 9.1 9.2

Impact of technological structural changes (1990–2000): electronics sector Changes in local content (1990–2000): automotive sector Changes in import content (1990–2000): automotive sector Impact of trade structural changes (1990–2000): automotive sector Impact of technological structural changes (1990–2000): automotive sector World trade share of intra-regional trade Advantage of better logistics Japanese manufacturing FDI in East Asia: accumulated number of cases by industry Share of East Asia’s intra and extra trade in world trade Main routes and feeder services of East Asia container lines Thailand’s regional per capita value added of manufacturing sector China’s provincial per capita GDP (1990 and 2002) Thailand’s regional per capita GDP Relation between trade volume (1981–2004) with and distances to Laos Relation between number of approved inward FDI (1988–2004) and distances to Laos Cooperation of Tokyo Coil Laos with Tokyo Coil Thailand Cooperation of Asahi Maxima Laos with Asahi Denki Thailand Number of times to pass through trade-related barriers The ‘ideal’ full-information inter-country input–output table Development of specialization ratios for internal EU exports, 1965–85 Weighted averages of IO cell-level Grubel-Lloyd indexes by industry of origin Output difference and labour productivity (all local plants except for machinery) Output difference and labour productivity (local machinery plants)

103 105 106 106 106 119 122 126 146 154 155 157 160 168 169 175 176 178 190 200 213 246 247

List of Figures xi

10.1

Parts procurement of a hard disc drive assembler located in Thailand 10.2 FDI inward and outward stock by region 10.3 Inward and outward FDI stocks of the ASEAN countries 10.4 Outward FDI flows of the ASEAN countries 10.5 Share of FDI stock abroad from Malaysia by geographical destination 10.6 Share of FDI stock abroad from Thailand by geographical destination 10.7 Share of FDI stock abroad from Malaysia by geographical destination 11.1 Asia-Pacific regional and bilateral trade agreements (June 2005) 12.1(a) FTA effect on GDP 12.1(b) FTA effect on capital accumulation 12.2 FTA effect on investment 12.3 Investment in Vietnam for baseline and policy scenario 12.4 Investment in Malaysia for baseline and policy scenario 12.5 FTA effect on foreign ownership of physical capital 12.6 FTA effect on export volume 12.7 FTA effect on import volume

261 263 264 265 265 266 267 287 329 330 331 332 332 333 334 335

List of Tables 2.1

Matrix of FTAs involving countries in the Asia-Pacific region (as of November 2007) 3.1 Widening and deepening of the Asian manufacturing matrix, 1985, 1990, 2000 3.2 Intra-East Asian preference margins vis-a`-vis EU and North America 3.3 Bound and applied tariffs in East Asia 4.1 Rules of origin in FTAs 4.2(a) Local content (1990) 4.2(b) Local content (2000) 4.3 Changes in local content 1990–2000 4.4 Local and import content (2000): manufacturing sector 4.5 ASEAN and ASEAN-China cumulative local content: manufacturing sector 4.6 Local and import content (2000): manufacturing sector 4.7 ASEAN and ASEAN-China cumulative local content: manufacturing sector 4.8 Local and import content (2000): electronics sector 4.9 ASEAN and ASEAN-China cumulative local content: electronics sector 4.10 Local and import content (2000): automotive sector 4.11 ASEAN and ASEAN-China cumulative local content: automotive sector 5.1 Japan–China trade 5.2 Share of air transportation and containerization in ocean transportation in Japan–China trade 5.3 Japan–China trade in machinery and electrical equipment 5.4 Japan–China trade in textile industry by transportation mode 5.5 Japan–ASEAN trade 5.6 Share of air transportation and containerization in ocean transportation in Japan–ASEAN trade 5.7 Japan–ASEAN trade of machinery and electrical equipment

xii

30 53 60 68 84 90 92 93 94 97 98 101 102 104 105 107 127 129 131 132 135 136 137

List of Tables xiii

6.1 6.2 6.3 7.1 7.2 8.1 8.2 8.3 8.4 8.5 8.6 8.7 8.8 8.9 9.1 9.2 9.3 9.4 9.5 9.6 9.7 9.8 9.9 9.10 10.1 10.2 10.3 10.4

10.5

11.1

Volume of sea container transportation in East Asia in 2003 Production share of Chinese main agglomerations Malaysia: number of approved FDI and per capita GDP by state Courier service cost from and to selected countries Documents needed for an out-plan import of machinery by land Pearson correlations between levels and changes in specialization, EU5, 1965–85 ROW export specialization as a percentage of EU export specialization, 1965–85 Origin of total intermediate inputs (%) Average self-sufficiency ratios (%), 1975–95, by sector Growth of self-sufficiency ratios (%), by sector Origins of tradable intermediate inputs (%), 1975–95 Origins of tradable consumption goods (%), 1975–95 Origins of tradable capital goods (%), 1975–95 Averages of IO cell-level Grubel-Lloyd indexes by country of origin (%) Industrialization in ASEAN Categorization of industries Directions of resource reallocation Summary statistics of variables Spillover effect and plant productivity Demand linkage effect and plant productivity Averages of variables by ownership Output difference and plant productivity Trade liberalization and plant productivity ISIC industry codes Intra-ASEAN FDI flows by source and host country Share of intra-ASEAN FDI flows by source country Share of intra-ASEAN FDI flows by host country FDI inflow from extra- and intra-ASEAN in manufacturing industry (approved base) by industrial sector (1999–2003) Share of intra-ASEAN FDI flows in manufacturing industry (approved base) by industrial sector (1999–2003) Applied tariffs in East Asia, the EU and NAFTA, by sector, 2002

149 152 159 171 173 199 201 204 205 206 207 209 211 214 227 229 232 236 238 240 244 248 250 252 269 270 271

273

275 289

xiv List of Tables

11.2 11.3 11.4 11.5 11.6 11.7 11.8 11.9 11.10 11.11 11.12 11.13 11.14 11.15 11.16 12.1 12.2 12.3 12.4 12.5

Macroeconomic impacts (in %, SC1) Impact on welfare (US$ million, SC1) Impact on agricultural trade (in %, SC1) Impact on production and prices (in %, SC1) Trade in industrial goods (in %, SC1) Terms of trade (in %, SC1) Impact on world prices (in %, SC1) Major changes in bilateral trade (US$ million, SC1) Macroeconomic results (% change, SC2) Trade impacts on agricultural goods (%, SC2) Impact on production and prices (in %, SC2) Impact on industrial trade (in %, SC2) Terms of trade (% change, SC2) World prices for developing countries (% change, SC2) Major changes in trade (US$ million, SC2) Aggregation of regions Aggregation of sectors Average tariff rates (2010, %) Effect on export volume by sector in 2020 Effect on import volume by sector in 2020

294 295 297 300 301 302 303 304 306 308 309 311 312 313 314 325 326 328 336 337

Notes on the Contributors Richard E. Baldwin is Professor of International Economics at the Graduate Institute of International Studies and Policy Director of the Centre for Economic Policy Research. Michel Fouquin is Deputy Director of the Centre d’Etudes Prospectives et d’Informations Internationales (CEPII). Nobuaki Hamaguchi is Professor at the Research Institute for Economics and Business Administration, Kobe University. Daisuke Hiratsuka is Director General of the Development Studies Center, Institute of Developing Economies (IDE). Ken Itakura is Associate Professor at the Graduate School of Economics, Nagoya City University. Souknilanh Keola is Research Fellow of the Development Strategies Studies Group, Development Studies Center, Institute of Developing Economies (IDE). Fukunari Kimura is Professor in the Faculty of Economics, Keio University. Ikuo Kuroiwa is a Senior Research Fellow Sent Abroad (Singapore), Development Studies Center, Institute of Developing Economies (IDE). Bart Los is Associate Professor of International Economics and Business, University of Groningen. Jan Oosterhaven is Professor of Spatial Economics at the Faculty of Economics, University of Groningen. Motoyoshi Suzuki is JICA’s Policy Adviser to the Committee for Planning and Investment, LAO P.D.R. Kazuhiko Yokota is a Research Associate Professor at the International Centre for the Study of East Asian Development.

xv

Preface This book is intended for policy-makers, academics and students interested in ongoing economic integration in East Asia. Our goal is to provide readers with both theoretical and empirical frameworks needed to understand the current state and direction of economic integration in East Asia, while at the same time clarifying issues that must be addressed in the move towards fully-fledged economic integration. We have provided academic resources and analysis that graduate students can use for conducting further research on the topic. The main points under discussion are accessible to general readers by way of tables, figures, explanations and concluding remarks in each chapter. This volume is the second in our series on economic integration in East Asia. The first volume, East Asia’s De Facto Economic Integration, focused on a deeper understanding of the ongoing de facto economic integration in East Asia. This volume, East Asia’s Economic Integration: Progress and Benefit, focuses on de jure economic integration and the concerns arising from economic integration. The twelve chapters in this volume address the following questions: What institutional arrangements are required to support the current de facto economic integration? What will happen in the region if economies are more deeply integrated? Will each small ASEAN country be able to attract industries? In particular, will Cambodia, Laos, Myanmar and Vietnam benefit from economic integration? Will the rural areas in each country benefit from economic integration?

xvi

1 From De Facto to De Jure Economic Integration in East Asia: Past, Present and Future Daisuke Hiratsuka and Fukunari Kimura

1.1

Introduction

Economic integration in East Asia, particularly in trade, has made remarkable progress, along with high economic growth, enhancement of regional competitiveness and agglomeration of industry in the region. The rapid progress of de facto (informal) economic integration through trade and foreign investment has prompted interest in de jure (formal) integration to institutionalize regional cooperation. This would not only provide an efficient and stable regional trade and investment system but also narrow socio-economic disparities in East Asia and complement international trading systems governed by the World Trade Organization (WTO). In fact, in East Asia, regional trade arrangements have prevailed in the form of free trade agreements (FTAs) centring on ASEAN, such as ASEAN-China, ASEAN-Korea and ASEAN-Japan, to name a few. FTAs will certainly improve predictability as compared with the current trade system which carries the risk of most Asian countries raising their tariff until the bound tariff in the WTO. The suspension of the Doha round of the WTO trade negotiations in July 2006 might catalyze the regional trade arrangements, so that, at the bilateral, plurilateral and wider levels, East Asia will enter into a new era of regional trading arrangements.1 Such movement is complementary to WTO processes and may induce WTO members to resume the stalled Doha trade talks. However, the evolution and proliferation of FTAs at many different levels have raised a number of queries about where East Asia is headed. Firstly, what kinds of problems do the current FTAs face and what are their weaknesses? Which particular institutional arrangements are required to support the current de facto economic integration? Secondly, and more importantly, what will occur if there is greater economic 1

2 East Asia’s Economic Integration

integration in East Asia? More specifically, will regional production concentrate on China due to the home market effect or will each small ASEAN country be able to continue agglomeration of industries? Will the so-called CLMV countries (Cambodia, Laos, Myanmar and Vietnam) benefit from economic integration? Will the benefits of economic integration trickle down to rural areas in each country? These concerns stem from the fact that East Asian nations vary considerably in their levels of economic development, and there are significant peripheries within each country: for example, income disparity, in terms of per capita GDP, in 2006, varied more than one hundred times between the lowest and the highest figures: for example, US$237 for Laos compared to more than US$30,000 for Australia, Japan, Singapore, Brunei and New Zealand. In addition, there are major disparities among the domestic regions in each country. In East Asia, there are the so-called core–periphery structures, within each country as well as across them (Krugman, 1991; Krugman and Venables, 1995). Thirdly, of equal importance, there is concern as to whether indigenous enterprises can develop together with multinational ones. This stems from the view that since economies of scale work strongly in economic integration, multinational enterprises (MNEs) are expected to gain major benefits in East Asia where they have large operations. Fourthly, what are the policies for solving the problems in East Asia: to provide an efficient and stable regional trade and investment system, to narrow the development gaps, or to foster cooperation between indigenous firms and foreign ones? This book aims to answer these questions from a regional perspective. In this introductory chapter, we provide an overview of the main issues and set out what the subsequent chapters address. We begin by discussing the status quo of de facto economic integration. Second, we highlight which features of the current FTAs have been designed and practised in East Asia and the type of management of FTAs most appropriate for the region. Third, we outline our views on what has been happening in East Asia, at country/city level as well as firm level, in the wake of the development of de facto economic integration. This discussion will contribute to the debate on what will happen in East Asia if the region is integrated further by de jure integration. Since the start of this research project in 2004, entitled East Asia’s Challenges, we have conducted several field studies in East Asia. The results of these field studies will form part of our thesis on what has been happening in East Asia in the wake of the de facto economic integration. Fourth, we will discuss the impact of regional trade arrangements in East Asia, and the concerns

From De Facto to De Jure Economic Integration 3

arising from regional integration. Fifth, we shall briefly introduce each chapter and focus on policy implications. In the concluding section, we shall outline the policy package for economic integration.

1.2 1.2.1

De facto economic integration The world’s factory: East Asia

East Asia, narrowly defined as ASEANþ3, is the most attractive area in terms of market size, economic growth rate, and population revealing potential markets. The economic size of ASEANþ6, or member countries of the East Asia Summit (EAS), which includes India, Australia and New Zealand (as proposed by the Japanese government), in 2006, was US$11 trillion, accounting for about one-quarter of the world’s GDP. East Asia’s GDP grew more than 5.3 times in the period between 1980 and 2006. In 1980, the economic size of East Asia was just 62 per cent of the NAFTA economy. However, in 2006, it was 70 per cent of the NAFTA economy, and was almost the same economic size as that of the European Union (EU) that year. The total population of ASEANþ6 is 3 billion, which is almost half of the world’s population. Needless to say, East Asia is the region with the greatest potential. These facts suggest that East Asia, i.e. ASEANþ3, or ASEANþ6, has emerged as the world’s factory. The majority of the world’s production of hard disc drives (HDDs), takes place in East Asia: in Singapore, Thailand, China and, most recently, in Korea. The American companies Seagate, Maxtor and Western Digital have research and development (R&D) functions and highly capital-intensive processes of silicon bar production in the United States, but their production facilities are located in East Asia. The same applies for the Japanese companies Hitachi, Fujitsu and Toshiba, all of which base their major production processes in Singapore, Malaysia, Thailand and China. All major Japanese suppliers locate most of their production functions in these ASEAN countries and China. In addition, in 2005 East Asia produced 97 per cent of the world’s personal computers and 79 per cent of the world’s mobile phones, and in 2004, it produced 67 per cent of the world’s semiconductors. East Asia is quite literally the world’s factory that produces and exports a variety of goods. Two types of production fragmentation have advanced in East Asia. The first is intra-firm in which MNEs split production processes at several stages, and base those in different countries depending on which is the best location. This type of fragmentation is sometimes known as vertical specialization. MNEs achieve economies of scale for each production

4 East Asia’s Economic Integration

process, and at the same time, in particular in East Asia, a Ricardian type of division of labour occurs within firms such that capital- and labourintensive products and processes are allocated in different countries. We saw several such cases: (i) industrial machinery assembled in Thailand with the casting process in Vietnam; (ii) condensers for which the machinery process took place in Thailand and the manual process in Laos; (iii) footwear for which the machinery process took place in Thailand and the manual process in Myanmar; and so on. The second type is inter-firm (arm’s length) production fragmentation in which MNEs purchase parts not only from the host country’s suppliers but also from third countries’ suppliers in multiple countries. These suppliers are the so-called small and medium-sized enterprises (SMEs) that wish to expand their business beyond their host countries and their industry sector. Sometimes, home electric appliance parts suppliers produce goods for automobile-related suppliers, for example. The two types of intra-firm and inter-firm production fragmentation are mixed, forming production networks. For example, with HDDs, Seagate locates the assembly of the advanced model, and the most capitalintensive processes such as disc and R&D functions in Singapore, while the assembly of the low-price model and labour-intensive products and parts takes place in China and Thailand (Hiratsuka, 2006). For this division of labour, the Seagate Singapore factory purchases some parts from the Seagate Thailand factory. Nidec and Minebea are engaged in HDD spindle motor assembly in Thailand, and Showadenko in Singapore, and Fujidenki and Komag in Malaysia produce discs. Therefore the Seagate Singapore factory purchases parts of HDD spindle motors from Thailand and discs from Malaysia and Singapore, and the Seagate Thailand factory purchases spindle motors from Thailand, and discs from Malaysia and Singapore. Eventually, the assembling of HDDs will be concentrated across a few countries and parts and components will be procured from various countries, thus forming multilayered production networks in East Asia. This is the story for HDD, whose parts and components are very light and compact. A different picture emerges for goods with high transport costs. For example, the automobile industry is located in many countries and East Asia is becoming one of the world’s major automobile production areas due to the interaction of growing markets, low wage rates and agglomeration of industry. Thailand, for instance, has been both the production and the export base for pickup trucks. Ford/Mazda, GM/ Isuzu, Mitsubishi and Toyota have large operations in specific industrial districts such as Samutprakan, Rayon and Chonburi. Automobile parts are purchased mainly from domestic suppliers in order to reduce

From De Facto to De Jure Economic Integration 5

transport costs. In particular, the Asian currency crisis in 1997 increased the purchasing from domestic suppliers because of the rise in import prices. As a consequence, Japanese automobile supporting industries have agglomerated there. Those production networks, either across countries or within a country, are supported by a ‘just in time’ production operation that links assemblers, suppliers and markets with forwarders/logistic companies. Normally, each factory employs two shifts and has a 24-hour operation. Products are shipped very frequently, e.g. every two to four hours for domestic customers, and two or three times a week for foreign ones. The ‘just in time’ production operation, which is simply a system that ensures necessary parts are delivered to assemblers ‘just in time’, enables efficient production since warehouse space is limited and necessary parts can be conveyed to the workers speedily. Local staff in charge of purchasing examine parts and material for quality, enquire about prices and delivery schedules, and place orders to suppliers. They are able to maintain accurate information about prices and other details as there is regular exchange of information. Assemblers normally outsource to three suppliers for one part in order to encourage competition and promote R&D activity. Once the sales-purchase agreement is made, parts suppliers ask logistic companies to arrange for sea, air or land cargo transportation, to take care of the export and import procedures, and to deliver parts ‘just in time’, from one factory to another and from one country to another. Sometimes logistic companies follow a milk-run style, picking up parts from each supplier and dropping off at assemblers, and sometimes they supply parts to assemblers from ‘just in time warehouses’ or ‘replenishment hub centres’, where companies assigned by suppliers carry parts. Local logistic companies can participate in the production and distribution networks in East Asia. The shipment of assembled or finished goods is, however, quite a different story. Head offices in Japan and the United States receive orders from clients, allocate production to their factories across East Asia, and ask international logistic companies from East Asia factories to their sales representative offices. In sum, marketing and delivery are managed by the head offices, and factories in East Asia are simply efficient production bases controlled by the head offices. 1.2.2

Drive forces of regional integration

Through the production networks, intermediate goods are delivered from factory to factory, from factory to wholesalers/retailers/consumers, within countries and/or across countries. When we examine this process from a trade perspective, it is evident that regionalization in trade has

6 East Asia’s Economic Integration

progressed in East Asia. Indeed, in 2004, the intra-regional trade ratio of ASEANþ6 amounted to 43 per cent. This figure is higher than that of NAFTA (42 per cent), but lower than that of the EU (58 per cent). Most traded goods are intermediate goods rather than finished goods. Why has there been such a rapid increase in the intra-regional trade ratio? It can be interpreted in the following way. Firstly, the economies of other countries have achieved dramatic growth. When the Japanese economy used to be a hegemony, many of the manufactured products were exported to other East Asian economies. In contrast, other East Asian economies have now succeeded in industrialization, transforming from the periphery (unindustrialized) to the core (industrialized), resulting in expansion of manufactured goods among the countries, as well as to Japan. Secondly, East Asia has attained higher economic growth than other major regions such as NAFTA and the EU. In 1980, the economy of East Asia was just 25 per cent of the US economy, but it has now risen to 60 per cent. Thirdly, transport costs are decreasing for various reasons. The most favoured nation (MFN) tariff has been cut in the region, driven by competition for attracting investments (see Chapter 3). Exemptions of tariffs on intermediate goods have been introduced. Infrastructure projects are carried out by public and private sectors, and logistic networks have developed. Of course, there has also been a fall in air and sea cargo fares (see Chapter 5). Thanks to low transport costs in East Asia, manufacturing production bases are linked together, like a systematized factory with distribution networks. It is characteristic of East Asia that many countries have participated in the production and distribution networks, compared to which Europe and North America do not cover as many countries (Ando and Kimura, 2003). The process of production fragmentation has increased trade of intermediate goods within the same industries, raised East Asia’s intra-regional trade ratio, and contributed to the region’s rapid economic growth. In particular, in the machinery industry, production fragmentation has developed, resulting in high import value of parts and components in machinery, compared to the EU15 and NAFTA.

1.3 1.3.1

The problems of de jure integration in East Asia Progress of trade liberalization

It is often claimed that de facto economic integration has proceeded in East Asia, whereas de jure integration has been absent. However, de jure economic integration, particularly in the elimination of tariffs, has progressed far more in East Asia than is generally recognized. ASEAN

From De Facto to De Jure Economic Integration 7

has formed a high level preferential trade arrangement (PTA) in a number of exclusion lists as well as in the rules of origin (RoO) with a low local content ratio of 40 per cent. Since 1992, ASEAN has gradually reduced the tariff rates for goods within the member countries. By the beginning of 2003, the original six ASEAN members (ASEANþ6) of Brunei, Indonesia, Malaysia, the Philippines, Singapore and Thailand had succeeded in reducing preferential tariff rates to less than 5 per cent with plans to reduce to 0 per cent 99.4 per cent of tariff lines by 2010. Also, the newer ASEAN members of Cambodia, Laos, Myanmar and Vietnam will reduce the tariff to 0 per cent for 98.2 per cent of tariff lines by 2015. This means that ASEAN will realize the ASEAN free trade area (AFTA) by 2015, with few exclusions from tariff elimination. Considering that each country has put into place a programme to exempt tariffs on intermediate goods and that the Information and Technology Agreement provides for participants to completely eliminate duties on IT products, the actual utilization ratio of the AFTA is not so low and is in fact increasing. This is an indication that AFTA has been effective. Besides AFTA, the ASEAN-China FTA was implemented in July 2005, reducing the tariff of normal track to 0 per cent by 2010 for ASEANþ6 and by 2015 for new members. Each country has offered sensitive lists, according to the rules specified in the agreement, which are to be reduced to not more than 50 per cent by 2015 for ASEANþ6 and China and by 2018 for the newer ASEAN members. In December 2005, the ASEAN-Korea FTA was concluded with a provision that the agreement with Thailand would be effective once it had been signed. Since 2006, ASEANþ6 and Korea have applied the preferential tariff for normal track that covers more than 90 per cent of all the tariff lines and the total value of imports based on the 2004 trade statistics, with the aim of reducing them to 0 per cent by 1 January 2012 and 1 January 2010 respectively. Among the remaining 10 per cent, the two sides will eliminate tariffs to 0–5 per cent for 7 per cent of all the tariff lines by 2016. The remaining 3 per cent are regarded as highly sensitive goods which are classified into five groups from group A to group E according to sensitivity. The group E goods of 40 tariff lines at HS-6 level are allowed to be excluded from the scheme. Japan enforced bilateral FTAs with Singapore in November 2002, and with Malaysia in July 2006, and signed with the Philippines in September 2006, Thailand in April 2007, and Brunei in June 2007, and is currently negotiating with Indonesia and Vietnam. Besides these bilateral FTAs, Japan has negotiated plurilateral FTAs with all of ASEAN which will permit the utilization of the preferential tariffs based on accumulated RoO between

8 East Asia’s Economic Integration

Japan and ASEAN. That is, Japan can utilize the ASEAN-Japan FTA when goods are exported to another ASEAN country through an ASEAN country. The FTAs in East Asia tend to employ the AFTA’s RoO as standard. In the case of FTAs, RoO determine the ‘nationality’ of a product, and only products that are considered to have originated in FTA member countries are eligible for preferential tariff concessions. In contrast to other FTAs, AFTA has a simple and uniform format for RoO as shown in the following formula: 2 3 Value of Value of undetermined 4 imported non-ASEAN þ origin intermediate 5 intermediate inputs inputs  100%  60% FOB price ðof the finished productÞ This formula means that at least 40 per cent of the product’s content must originate from a member country. The AFTA’s 40 per cent of local content ratio is very low compared to that of other FTAs: e.g. NAFTA and EU have a 60 per cent local content ratio. In addition, its calculations are very easy. 1.3.2

The rules of origins and the overlapping FTA issue

The proliferation of FTAs will lower trade costs, and improve the predictability of trade. But the RoO requirements may impose managerial and trade costs on firms instead of tariff elimination when goods are eligible for preferential treatment in the importing country. The overlapping FTA means that a product will face different tariffs and application forms, which may increase administration costs and discourage exporters from using the preferential scheme. The JETRO’s survey on FTA utilization (2007b) found that among the 729 exporters who responded, only 5.1 per cent (37 firms) are already utilizing the preferential tariff schemes in the Asia Pacific region (ASEAN, Australia, China, Japan, India, New Zealand and Republic of Korea), and 8.5 per cent (62 firms) plan to do so (13 firms now utilizing schemes plan to continue doing so). Among the 97 firms currently utilizing or planning to utilize preferential tariff schemes, most (33 per cent) see ‘no problem at this point’, but 27.8 per cent think ‘[the existence of] different rules [in the region] complicates procedures to prove the country of origin, and leads to increased costs’. Firms were also asked for their views on the future direction of RoO under FTAs in the region. Of these, 63.9 per cent of respondents think RoO should be harmonized by some means. The evidence suggests that preferential tariff schemes are not widely used and the existence of multiple RoO places a heavy burden on firms.

From De Facto to De Jure Economic Integration 9

1.3.3

‘Shallow’ integration

Another problem is that most of the bilateral FTAs in East Asia that liberalize trade in goods still remain ‘shallow’, based on the preferential trade arrangements between developing countries. The so-called Enabling Clause allows less-developed contracting parties to eliminate tariffs and non-tariff measures exempt from the most-favoured-nation principle. They are a long way from ‘deep’ integration which involves establishing or expanding the institutional environments in order to facilitate trade and location of production. ‘Deep’ integration contains policies designed to encourage trade and facilitate segmentation of production processes and distribution networks. In this regard, business environment measures that support business activities and attract FDI are recommended for each country. Business environments vary from country to country, according to trade facilitation and production conditions. For instance, customs clearance in Indochina takes a lot of time: 70 hours at the Vietnam borders (JETRO, 2007a). Hence, Japan has proposed the ASEAN Common Investment Climate Initiative to enhance logistic efficiency and to strengthen the competitiveness of small and medium enterprises of ASEAN (AEM-METI, 2006). Given that East Asia is at diverse stages of economic development, there is no guarantee that developing economies can participate in East Asia’s production and distribution networks. The trade facilitation measures to promote the participation of developing countries in the production and destruction networks should be the key element for FTAs in East Asia. Equally importantly, regulations on service industry and investments are still in place in East Asia. East Asian nations are very conservative in liberalizing service industries and investment. Obviously, liberalization of service is a significant part of ‘deep’ integration. 1.3.4

Lack of political will in Northeast Asia

While regional trade arrangements have progressed among the ASEAN nations, Northeast Asia has yet to see such arrangements between China, Japan and Korea. Japan has proposed FTA to Korea, and in December 2003 the two countries began FTA negotiations, but no trade talks have been held since the sixth round of talks in November 2004. China proposed a China–Japan FTA in 2004 and 2006 at the ASEAN meetings. However, Japan’s stance relates to whether China fulfilled its WTO liberalization obligations up to 2007. Instead, as a precursor to trade elimination, Japan has proposed a tripartite cooperation between China, Japan and Korea to create an attractive environment for trade and investment, in areas such as

10 East Asia’s Economic Integration

customs, transportation and quality supervision, protection of intellectual property rights, minimizing trade frictions, and liberalizing service and investment (MOFA, 2004). Why has ASEAN progressed so far with de jure integration but Northeast Asia has not? Firstly, it is easier for China, Japan and Korea to conclude FTAs with the ASEAN countries, due to their smaller economies, than with bigger countries. Secondly, the domino effect has worked in that China has proposed FTAs with all of ASEAN, with Japan and then Korea following suit in order to maintain political influences on ASEAN. Thirdly, each side of ASEAN and its partners, China, Japan and Korea, feels the benefit from FTAs. ASEAN and China both seek markets with tariff elimination. In contrast, the Japan and ASEAN FTA aims at the improvement of the business environment and economic and technical cooperation. Japan wishes to enhance the competitiveness of ASEAN, which will benefit the large number of Japanese firms which operate there and thus benefit the Japanese economy. Meanwhile, ASEAN can get economic assistance to strengthen indigenous firms and to improve institutions in global competition. Korea was concerned that it might be excluded from the ASEAN market. However, common FTA strategies are not shared among China, Japan and Korea. Japan and Korea hope that China will liberalize service industries and investment as well as significant trade goods such as steel and automobile production, but China is unlikely to agree with this. China, on the other hand, hopes that Japan and Korea will liberalize protected industries such as agricultural products. 1.3.5

Architecture of East Asian integration

Several regional cooperation frameworks are beginning to intersect in East Asia. Since 1997, the year of the Asian currency crisis, ASEANþ3 has fostered regionalism in the form of functional cooperation in a total of seventeen fields including trade, finance and currency, labour, environment, tourism, energy and health. Japan has proposed the ASEANþ6 framework, sometimes referred to as ASEANþ3þ3 or EAS, which includes India, Australia and New Zealand. In December 2005, the first East Asia Summit (EAS) was held in Malaysia, following the ASEAN Summit meeting. At that time, the ASEANþ3 was commonly understood to be the framework that supported de facto economic integration characterized by East Asia’s production networks, while the EAS was simply the place to foster dialogue on global issues rather than on regional cooperation. In fact, unlike the ASEANþ3, the EAS does not hold the senior official meeting (SOM). However, in April 2006, Japan announced its global strategy to promote economic

From De Facto to De Jure Economic Integration 11

partnership among the EAS members, and proposed a comprehensive economic partnership for ASEANþ6 as a whole. This seems to have generated a conflict between ASEAN and China on the one hand, who supported the ASEANþ3 and Japan, India, Australia and New Zealand on the other, who supported the ASEANþ6 or EAS. Apart from this issue, the 2006 APEC Summit in Hanoi agreed to examine the FTAAP (the Free Trade Area for the Asia Pacific), and to submit the results at the 2007 APEC meeting in Australia. This has generated another conflict. The United States fears being excluded from East Asia’s regional trade arrangements, is opposed to East Asia’s regionalism, and is determined to launch the FTAAP. On the other hand, ASEAN is opposed to the American idea of FTAAP since the least developed countries of Cambodia, Laos and Myanmar are excluded from APEC. Which type of regional cooperation is the most appropriate for East Asia? The WTO integration process, the ASEAN þ3, the ASEAN þ6 and the APEC process can be seen as a concentric circle rather than as contradictory forces. There is a valuable lesson to be learned from the EU experience which has shown that regionalism brings not only economic prosperity but also peace. The ASEANþ3 and the ASEANþ6 will foster a sense of community in East Asia, and lead East Asia to economic prosperity and peace.

1.4 1.4.1

Evolution of a core–periphery structure Core–periphery structure among countries

Business sectors, academics and policy-makers share concerns about the effect economic integration will have on where industry is located. In particular, the ASEAN countries fear that industry will be concentrated in China where there are large markets and producing powers. However, the multipolar geographical structure will continue in East Asia. It is certain that agglomeration forces in an integrated market will be larger than ever in China, due to the economy of scale effect. They will, however, increase wage rates and land prices, and eventually dispersion forces will come to operate since manufacturing will move from higher wage countries to lower wage ones, as Krugman and Venables (1995) argue.2 Indeed, the term ‘China plus one’ has been coined. ‘China plus one’ means that MNEs are exploring new frontiers beyond China. China is losing its attractiveness for foreign direct investment for export purposes because of increasing wage rates and land prices. Instead, Vietnam is now attracting foreign direct investment due to high quality and cheap labour as well as a greatly improved business climate thanks to government efforts.

12 East Asia’s Economic Integration

The success in industrialization of less developed countries with lower wages such as Cambodia, Laos, Myanmar and Vietnam (CLMV) is dependent on how far they can reduce border barriers. The CLMV countries, however, lack infrastructure and have poor trade institutions and little start-up business assistance from government. If they overcome these handicaps, in particular the poor institutions, and are able to reduce transport and set-up costs, MNEs will move manufacturing to the countries with cheaper labour. As mentioned earlier, MNEs have been seeking new frontiers. In this context, the ASEAN countries whose wages are now relatively high compared to China are rapidly losing their earlier advantage of cheap labour. In the Europe of the 1980s, MNEs went to Portugal and Spain where wages were low compared to major industrialized countries, but in the twenty-first century MNEs are extending their production facilities to new frontiers in Hungary and Poland where wages are lower than in Portugal and Spain. Economic integration means that economic benefits trickle down from high wage to low wage countries, and eventually to lower wage countries as MNEs seek new frontiers. Unless these higher wage countries can alter their business climate (by improving accessibility to knowledge and information, enhancing human resource capability in high skill activities, and increasing research and development capacity), MNEs will not be able to upgrade existing production facilities, and relocate them. In sum, in order to reap the benefits of economic integration, high wage countries will have to improve their business climate and enhance human resource capability, and low wage countries will have to lower border barriers and improve their trade and investment climate.

1.4.2

Core–periphery structure among cities within country

In this section, we examine whether economic integration will narrow or widen socio-economic disparities among cities and/or domestic regions. In other words, will economic benefits trickle down to rural areas and/ or cities? There are some arguments on the trickle-down economic benefit among domestic regions. One argument stemming from the core–periphery structure theory is that in a world of complete free factor mobility, consumers and producers will concentrate on a few cities where there are a variety of goods and a large market (Krugman, 1991). Venables (2004) finds that in Europe income disparities between countries are decreasing while income disparities among cities are increasing. Baldwin and Wyplosz (2004) further observe that European integration has been

From De Facto to De Jure Economic Integration 13

accompanied by only modest relocation of industry among nations but industry has become more concentrated spatially at sub-national level. Another argument is that, in economic integration, as economy and population sizes grow gradually, small cities grow to be medium-sized cities, medium-sized cities grow to be large cities, and at the same time, the total number of small, medium and large cities increases (Fujita et al., 1999a, 1999b). This implies that the trickle-down economic benefits diffuse to small cities in rural areas as the economy and population grow. Urban areas in East Asia have grown in the past few years in terms of population and industry as compared with the 1980s when there was little growth. The coastal areas of China, in particular, have industrialized since the 1980s. Southeast Asia shows a similar pattern. For example, in Lamphun, 50 km south of Chiang Mai, there is a concentration of Japanese manufacturing affiliates. In Phisanulok, about 380 km north of Bangkok, Thai Yazaki operates a wire harness plant with 4500 workers. Seagate, a leading producer of HDDs, has a large plant in Nakhon Rachasima, 300 km north of Bangkok. In Malaysia, many parts suppliers are located in Johor, close to Singapore, and also in Malacca where the super highway has extended from Singapore and Kuala Lumpur.

1.5

Growing indigenous enterprises with MNEs

In economic integration where economies of scale are significant, MNEs will expand production and distribution networks. This raises the question of how indigenous enterprises will fare in economic integration, whether they will be able to participate in production and distribution networks, and whether they will disappear from the market as a result of competition with foreign ones. In particular, have ASEAN indigenous enterprises benefited from economic integration? In Penang, Malaysia, several indigenous enterprises have developed through linkages with American electronics companies. These are ENG Teknologi, Globetronics, Kobay Technology, and Pentamaster Corporations. They all share a common history as multinational electronics companies in Penang have outsourced system equipments and parts to them. They have expanded their business in various areas, and have moved overseas in the wake of their clients, thus becoming global players. MMS Ventur, engaged in equipment systems, used to be a small factory shop in 2004, but now has a large office space and a factory on an industrial estate where there are many MNEs. The Thai Yazaki Phisanulok factory has outsourced wire harness production to indigenous

14 East Asia’s Economic Integration

enterprises in Khon Kaen in Thailand and Vientiane in Laos. Khon Kaen and Vientiane firms were established when they were outsourced. These cases suggest that the growth of indigenous enterprises depends on how far they are able to strengthen business linkages with MNEs. In this regard, American MNEs have outsourced to indigenous enterprises, far more than the Japanese MNEs. Most indigenous enterprises in Penang have strong linkages with American MNEs. Dell Computer Penang, for instance, has outsourced most of its parts (barring core parts that affect performance), systems equipment and packagingrelated goods to indigenous enterprises. Japanese MNEs, however, have not had the right to choose their outsourcers. Their headquarters hold purchaser meetings in Japan, inviting Japanese suppliers operating overseas in order to outsource. But, in recent years, Japanese MNEs have tended to outsource to indigenous enterprises in order to reduce costs in a situation of increasing global competition. Canon Hitech Thailand, for example, is increasing purchases from indigenous parts suppliers, in order to reduce transport costs which are 2–5 per cent of parts values. Economic integration has provided huge business opportunities not only through expanding production and distribution networks but also through expanding markets. Economic integration has provided huge business opportunities to final assemblers. Pensonic, a Malaysian electrical appliance firm, is increasing sales by the so-called original equipment manufacturer (OEM)3 which outsources production to more efficient producers in China, Indonesia, Thailand and Vietnam and resells to the ASEAN integrated market. ASEAN enterprises have a major presence in agricultural products, including food and beverages. San Miguel, a major brewery in the Philippines, has expanded its business overseas. San Miguel operates breweries in Vietnam, Australia and Indonesia and maintains four breweries in China, including in Hong Kong. The Charoen Pokphand (CP) Group, a leading Thai agro-based company, has expanded into Bangladesh, Cambodia, China, Hong Kong, India, Indonesia, Malaysia, Myanmar, Singapore, Taiwan, Turkey and Vietnam. Thus there are many indigenous enterprises, as in the cases cited above, that have successfully made linkages with MNEs, and are expanding sales in the integrated ASEAN market.

1.6

Benefit from economic integration

What benefits are generated by economic integration? The traditional approach of economic integration (Viner, 1950) focuses on trade creation and diversion effects. The extent of trade creation and diversion

From De Facto to De Jure Economic Integration 15

effects depends mainly on the degree of trade barriers as well as trade dependency of each member country and within member countries. In East Asia, there are zero tariffs on intermediate goods generally for export purposes, but tariffs on finished goods still remain very high. This fact has been confirmed by a study compiled in our previous book, East Asia’s De Facto Economic Integration. Freudenberg and Paulmier (2006) found that East Asian countries apply high tariffs for protected sectors, namely agriculture, food and beverages, and light industry including textiles and apparel, and surprisingly, that in every sector East Asian countries apply higher effective tariffs on products manufactured within the region than on imports from the EU and NAFTA. This effectively means that East Asia is discriminating against itself. It must be acknowledged that national borders have impeded international trade in East Asia. National borders impose export and import procedures requiring extensive paperwork from related agencies. This raises the transport costs across borders, and consequently obstructs international trade. Even developed countries’ transport costs associated with national borders are very high. Anderson and Wincoop (2004) estimated transport costs by using a gravity model, broadly defined in developed countries. A rough estimate of the tariff equivalent of trade costs is 170 per cent: 21 per cent of transportation of goods, 44 per cent of border-related trade barriers, and 55 per cent of retail and wholesale distribution costs (1.7 5 1.21  1.44  1.55  1). The Japanese manufacturing affiliates that were interviewed said that they were keen to reduce transport costs. These are equivalent to about 1.5–2.5 per cent of the cost of purchasing parts and materials, and this estimate would be higher if packaging costs are included. In electronics industries, transport costs are almost the same as labour costs. If transport costs are reduced by elimination of tariffs as well as trade facilitation measures, it would provide greater opportunities for dynamic economic development. There are many studies which examine the relationship between border barriers and international trade. McCallum (1995), for instance, finds that domestic trade in Canada and the United States is twenty times that of international trade between the two countries, even after controlling for geographical distance and economic size. This figure may seem rather large. However, it may be true that the national border between Canada and the United States has a big impact on international trade. These studies suggest that non-tariff border barriers are more than tariff barriers, and thus national borders have impeded trade across countries. Elimination of tariffs is not sufficient to realize deep economic

16 East Asia’s Economic Integration

integration. If East Asia can integrate fragmented markets into a seamless single market, the positive effect would be enormous for the countries concerned. However, two things must be kept in mind. Firstly, it will damage the countries that are excluded from regional integration such as Taiwan, South Asia and some other economies to a greater or lesser degree. Secondly, regional integration means the achievement of comparative advantage among regional member countries, but not efficient international division of labour at the global level. Obviously, the liberalization at the WTO is better than the FTA integration in efficient resource allocation.

1.7

Policy implications of individual chapters

The chapters in this book fall into four main parts. Part I, ‘The Making of Economic Integration in East Asia’, discusses the architecture of East Asian economic integration. Part II, ‘The Evolution of the Core–Periphery Structure’, explores the evolution of core–periphery structures in East Asia at inter-country as well as inter-city level. Part III, ‘The Current Status of Indigenous Enterprises’, considers whether indigenous enterprises can thrive with economic integration. Finally, the chapters in Part IV discuss various aspects of the ‘Quantitative Implications of Economic Integration’ for the East Asian economy. 1.7.1

Part I: The Making of Economic Integration in East Asia

Chapters 2 and 3 present contrasting assessments of the proliferation of FTAs in East Asia. In Chapter 2, ‘The Modality of East Asia’s Economic Integration’, Fukunari Kumura discusses the current status and the future prospects of de jure economic integration in East Asia and argues that the reduction of tariffs would induce the restructuring of industries which for a long time have been protected by high tariffs. Although until quite recently East Asia was lagging behind the worldwide boom of regionalism, bilateral FTAs in East Asia and other neighbouring regions are now actively being formulated. In East Asia, de facto economic integration has started earlier than de jure economic integration, and thus the characteristics of the former have strongly influenced the process and contents of the latter, particularly with regards to promoting FDI and international production/distribution networks. Tariff removals led by AFTA and followed by other FTAs have started to work effectively in restructuring import-substitution-type industries while pragmatism in utilizing policy measures other than tariffs promotes the improvement of business environments for network-forming industries. Despite

From De Facto to De Jure Economic Integration 17

its economic size, ASEAN is currently playing a pivotal role in de jure economic integration in East Asia, which seems to support open architecture in the future form of de jure integration rather than constructing a ‘fortress East Asia’, while at the same time it nurtures a strong consciousness of development issues in the region. In Chapter 3, ‘The East Asian Noodle Bowl Syndrome’, Richard Baldwin presents the contrasting view that the current web of FTAs has imposed additional administration costs instead of tariffs. There have been many ex ante studies based on Computable General Equilibrium (CGE) models but the real costs and benefits of the current web of FTAs can be better assessed using firm level studies. Baldwin argues that the features of the current East Asia de jure economic integration are fragile for three reasons. Firstly, in East Asia, the manufacturing sectors of all the nations are tightly interlinked via regional trade: firms in factory economies have sourced parts from other factory economies. However, ‘Factory Asia’ operates in the fragile trade system of ‘tariff binding overhang’ where a tariff binding committed to the WTO is higher than an applied tariff. Hence, there is a possibility that a tariff-raising war will break out in East Asia since the tariff binding overhang countries can raise the MFN tariffs without violating the WTO rule. Therefore, Factory Asia needs regional public goods to manage conflicts in the region. Secondly, the current state of East Asian de jure economic integration or regionalism is marked by the ‘noodle bowl syndrome’ where the FTAs are different in exclusion lists, dispute settlement mechanisms, and so on. The author discusses the fact that real regionalism means discriminatory (preferential) tariff liberalization against other regions. In this sense, there has not yet been any real regionalism in East Asia. Thirdly, in East Asia FTA management is necessary. Otherwise, the three largest bilateral trade flows of China–Japan, China– Korea and Japan–Korea will not covered by FTAs because domestic produce in those countries will be opposed to FTAs with big trade partners. This increases the need for the East Asian FTA management to harmonize lists and to include the three largest countries. Policy-makers should pay attention to these suggestions. There are several free trade schemes in existence in East Asia. Above all, AFTA is the most advanced one in coverage of tariff elimination. Does the business sector use the preferential tariff scheme? In order to avoid tariff fraud, whereby goods from third nations are transported through a member country to gain preferential access, FTA requires the importer to prove that the goods actually originated in the member countries. This requirement raises the question of whether FTA is effective in East Asia.

18 East Asia’s Economic Integration

In Chapter 4, ‘Rules of Origin, Local Content, and Cumulative Local Content in East Asia: Application of an International Input–Output Analysis’, Ikuo Kuroiwa conducts an empirical study, based on the Asian international input–output tables developed by the Institute of Developing Economies, to address this very question. Kuroiwa’s study shows that the local content ratios of East Asia are higher than the AFTA threshold of 40 per cent (with the exception of electronics in Malaysia, the Philippines and Thailand, and refined petroleum in Singapore and Thailand). The electronics industry has fragmented with the production process split and is now located in different countries. The electronics industry of Malaysia, the Philippines and Thailand has satisfied the ASEAN cumulative local content of 40 per cent. Thus, the study finds that AFTA and ASEAN-China FTA (AFTA-standard RoO applied) will be effective, and may promote international trade among countries within the region. Chapter 5, ‘Making Sense of the Timeliness of Transportation in Economic Integration’, by Nobuaki Hamaguchi, emphasizes the timely delivery and short lead time of goods in globalization where multinational firms have split production lines and locate them optimally such that they can minimize total direct production costs by taking advantage of international factor price differences. There are roughly four ways to realize timely and short delivery of goods across countries. First, good transport infrastructure plays a crucial role for delivery time. Second, a substantial saving of shipping time is possible by improving the efficiency of customs procedures and shipment handling. This can be achieved by institutional reform without requiring heavy investment in infrastructure. Third, the exchange of information with extensive use of information and communication technology can contribute to timely and short delivery. Fourth, the capability of implementing timely outsourcing requires a reasonably good quality of human resources. 1.7.2

Part II: The Evolution of the Core–Periphery Structure

In East Asia, production in the electronics industry is becoming increasingly fragmented, while, in some industries such as automobiles, production has agglomerated in certain countries. What is happening in East Asia as a result of the combined effects of fragmentation and agglomeration? Where will industries locate in East Asia? Will the distribution of economic activities in East Asia become more uneven? In Chapter 6, ‘The Evolution of the Core–Periphery Structure in East Asia’, Nobuaki Hamaguchi discusses the trickle-down of industrialization, from the point of view of spatial economics, focusing mainly on issues within each country. The study points out that there is an

From De Facto to De Jure Economic Integration 19

increasing development gap within each country, as with China’s coastal provinces and inland provinces and the growing concentration in Bangkok and its surrounding areas in Thailand. However, looking closer at the core regions, the study finds a geographical spread of agglomeration. In China, the income gap among coastal provinces has narrowed. In Thailand too, there is a spread of industrialization to the east and north of Bangkok. Thus, as the agglomeration in core regions grows, more people enjoy higher incomes. The resulting stark core–periphery structure, however, implies a wider inter-regional developmental gap. The study suggests that, as with the EU where an agenda for the integration process includes regional policies to narrow development gaps as a common goal, the regional integration in East Asia needs a formal framework to narrow the development gap across countries as well as within each country. In East Asia, the income disparities among countries are relatively high compared to other regions of the world, but there have been no formal arrangements to economically help unindustrialized countries such as Cambodia, Laos, Myanmar and Vietnam. Will these countries successfully overcome their handicaps and be integrated with other East Asian countries? In Chapter 7, ‘Industrialization Through Vertical Production Networks: Can Laos Participate?’, the authors examine the case of Laos from the perspective of periphery economies. Apparel, electrical appliances, electronics, motorcycle and automobile industries have all agglomerated in Thailand, dispersing to rural areas. Whether Laos can transform from the periphery to semi-core in the future depends on how many plants operating in Thailand will move to Laos. This study finds there has been a remarkable improvement in trade procedures among other institutions as a result of an attempt by the government to lower border-related barriers. The improvement in the institutions has made it possible for some MNEs in Thailand to relocate parts of their production processes to Laos. The study gives an indication that Laos will be able to overcome the handicaps of a small market and an unindustrialized economy if the government is successful in reducing border barriers and improving the business climate with cooperation from foreign governments as well as international organizations. It is a major concern for both policy-makers and academics as to whether small countries benefit more or less in an integrated market. In Chapter 8, ‘Structural Change in Intermediate, Consumption and Capital Goods Trade During Economic Integration: the EU Experience’, Los and Oosterhaven describe the European experience in the context of this particular issue by utilizing inter-country input–output tables of

20 East Asia’s Economic Integration

seven major European countries. An initial study, from earlier intercountry input–output studies on EU integration, showed that small countries are more specialized than large countries. The authors then investigated whether domestic transactions inside EU countries grew faster or slower than trade with the rest of the EU, and in particular, whether there were differences between small and large countries. The results obtained from their original studies are that, in intermediate inputs, there is no difference between small and large countries, and that surprisingly, in most countries, domestically produced inputs are becoming more important in the long run. By contrast, for small and large countries, in consumption goods and capital goods, there is a tendency over time to decrease domestically produced inputs in all seven EU countries, and increase intra-EU inputs and extra-EU inputs. Consumption goods are generally bought from outside the country. Next, they calculated the Grubel-Lloyd index for the EU countries and found that intra-industry trade is larger than inter-industry trade but the latter is increasing for all the EU countries except Denmark. Some industries are becoming more dependent on intra-industry, while others are more dependent on inter-industry. What are the implications for East Asia from these European experiences? Firstly, there is no evidence to indicate that large countries benefit more from economic integration than small countries. Rather, small countries can benefit if they specialize in specific industries. Secondly, for the seven major EU countries, increasing trade with the rest of the EU means that countries joining the integration process later can benefit from integration. This finding indicates that the CLMV countries can benefit from integration if they join the league later on. Of course, the countries involved in economic integration in East Asia and Europe are quite different, in particular with regard to disparities in income, and in terms of border characteristics, in that most European countries have land borders, whereas East Asian ones have sea borders. However, it is possible for East Asia to share the European experience and for small ASEAN countries to benefit from economic integration even if they join the league later. 1.7.3

Part III: The Current Status of Indigenous Enterprises

Do indigenous enterprises join the league of economic integration? In Chapter 9, ‘Spillovers and Linkages between Local and Foreign Plants’, Kazuhiko Yokota addresses this issue by examining the effects of MNEs on local firms’ productivity using plant-level data from Thailand. The study has many significant findings. Firstly, positive spillovers and linkage effects exist in the machinery industry, but not in other industries.

From De Facto to De Jure Economic Integration 21

Secondly, in machinery industry and labour-intensive industries, trade liberalization is strongly and positively correlated with higher plant productivity. That is, the impact of FTA/EPA differs from industry to industry according to particular characteristics. Therefore, thirdly, local firms in the machinery industry may receive the largest benefits from FTA/ EPA. And fourthly, large foreign plants and small local plants coexist in the machinery industry. All these findings suggest that MNEs and local firms in the machinery industry are vertically linked in production. In other words, MNEs demand intermediate goods from upstream local firms and their relationship is complementary to the machinery industry. This implies that competitive suppliers in the upstream in an industry such as the machinery industry are crucial to the success of FTA/EPA. Have indigenous enterprises really benefited from globalization or economic integration? The answers to this question are found in Chapter 10, ‘Do Local Suppliers Benefit from Economic Integration? Evidence from Outward FDI from ASEAN’, by Daisuke Hiratsuka. Outward foreign direct investment (FDI) from developing countries is a proxy indicator to measure how great a role enterprises of developing countries have played in the world market and how they benefit from globalization where border barriers are reduced. This study finds that ASEAN enterprises have extended their business activities to East Asia, becoming regional and global players. Firstly, Singapore has taken a lead, followed by Malaysia and Indonesia, although most of their FDI is directed towards Singapore. Secondly, ASEAN country transnational corporations (TNCs) went initially to neighbouring countries, resulting in a high share of outward FDI within ASEAN and East Asia, and then to large markets such as the United States as they grew. They seem to progress from being local players, then ASEAN regional players, and finally global players. Thirdly, ASEAN enterprises have established strong linkages with MNEs. They followed MNE customers when they expanded their facilities to lower wage countries. Finally, thanks to the home market effect, some ASEAN enterprises have become global players in agro-based industries and textile industries. 1.7.4

Part IV: Quantitative Implications of Economic Integration

Chapter 11, ‘Comparing Bilateral and Multilateral ASEAN10þ4 Free Trade Agreements: Possible Impacts on Member and Non-member Countries’, by Michel Fouquin, recalls that East Asia has maintained higher barriers to trade within Asia than with other partners, and conducts simulation analyses of comparisons with the four bilateral FTAs (ASEAN-China, ASEAN-Japan, ASEAN-Korea and ASEAN-India) and the full FTA among ASEAN10þ4 (China, Japan, Korea and India). Firstly,

22 East Asia’s Economic Integration

this study finds that out of the four bilateral ASEAN hub FTAs, ASEAN10 gets the most benefit. Secondly, the full liberalization among ASEAN10þ4 imposes severe competition on ASEAN10 with China in the Japanese and Korean markets. Therefore, ASEAN10 will lose compared to the four bilateral ASEAN FTAs. Thirdly, ASEAN will benefit from liberalization on agricultural products. Fourthly, the highly protected economy may lose out from the liberalization process since it causes a decrease in prices, employment and wages. Lastly, excluded economies from the FTAs, in particular Hong Kong and Taiwan, whose main trade partners are ASEAN, China, Japan and Korea, would lose out due to the trade diversion effect from the four bilateral FTAs. Theirs would be a major loss from the full liberalization among ASEAN10þ4. Although the trade liberalization of the ASEANþ3 may bring more trade, production and investment to the region, there are concerns raised by the ASEAN countries that potential economic benefits from the integration could be mostly reaped by the large economies, especially fast-growing China. They are also concerned about the investment coming from abroad, since the economic growth of ASEAN has been fuelled by attracting foreign investment to manufacturing industries. One of the prominent issues for the ASEAN countries to consider is whether the investment from abroad would be diverted outside ASEAN, once the ASEANþ3 FTA was established. Chapter 12, ‘How Will ASEANþ3 Integration Accelerate Investment? A CGE Analysis’, examines this particular issue. Ken Itakura’s simulation results of the ASEANþ3 FTA by a CGE model tell us that establishing the ASEANþ3 FTA can bring economic benefits, in terms of GDP increase and investment increase, to the participating countries. Additionally, all of the ASEANþ3 countries can attract more foreign investments. These findings imply that the ASEANþ3 FTA has a potential for accelerating the investment in the region and that all the countries in the FTA can gain from the integration by committing to free trade.

1.8

Conclusion

This book complements the previous volume, East Asia’s De Facto Economic Integration (2006, Palgrave Macmillan). The two books provide some perspectives on what has been happening in East Asia where economies are being integrated by de facto as well as by de jure integration. We can conclude from the two books that fragmentation of MNEs slices production processes and locates them in different countries/cities according to the most advantageous location. East Asia, i.e. ASEANþ3, is diverse

From De Facto to De Jure Economic Integration 23

in terms of location and government support programmes, such as tax exemption systems, labour costs, labour skills, development of supporting industries, and so on. Hence, the differences form a sort of historical path dependency in that specific industries have agglomerated in different countries/cities: for example, the advanced electronics industry in Singapore and Penang, the assembly of electrical appliances in Guangzhou, spindle motors for HDDs in Greater Bangkok, and so on. In East Asia, this kind of diversity in agglomeration of industry has resulted in the fragmentation of production processes and vertical production networks. This is related to the development of a logistics network as well as policy measures to exempt tariffs on intermediate goods. These measures contributed to a high regional trade ratio of 52 per cent in 2004. The production fragmentation has extended, covering gradually almost all countries in the region. The benefits of economic integration have trickled down to the rural areas, and then even to the so-called peripheries (unindustrialized small countries) such as Cambodia and Laos. Such evolution of the core–periphery structure is the result not only of de jure integration and FTAs but also of improvements in the business climates in individual countries. Efforts to improve business climates are very important. Laos is attempting to do this by reforming institutions. Unfortunately, Myanmar has not joined the league of economic integration. If Myanmar improves its business climate, the trickledown effects of economic benefit will spread throughout the country. Along with regional economic integration, MNEs have been extending regional production networks. There is empirical evidence that indigenous enterprises can benefit from economic integration through gradual participation in production and distribution networks. As global competition advances, MNEs have to outsource more to indigenous enterprises in order to reduce transport costs. The procurement of materials and intermediate goods from domestic producers is cheaper in terms of transport costs than procuring from foreign producers. Therefore, indigenous enterprises will enjoy greater business opportunities from regional integration. In East Asia, regional trade arrangements have prevailed in the form of a free trade area (FTA). We have noted that widescale FTA such as East Asia FTA is better than that of ASEANþ1. We have addressed the issue of regional trade arrangements generating positive effects to member countries but negative ones to non-members. ASEAN countries are concerned that there will be a concentration of investments in China due to the home market effect. However, our estimated results confirm that small ASEAN countries can benefit in terms of investment.

24 East Asia’s Economic Integration

We suggest that regional trade arrangements have to contain elements not only of elimination of tariffs but also other efforts such as trade and investment facilitation measures and issues of economic cooperation. Given that East Asia is at diverse stages of industrial development, and that, in particular, Cambodia, Laos, Myanmar and Vietnam are lagging behind in economic development, economic cooperation is very important. Japan will be able to play a significant role in facilitating cooperation with the governments of ASEAN countries and China, since major parts of the MNEs operating in ASEAN and China are Japanese affiliates. Lastly, we would like to reiterate that regionalism or regional trade arrangements on a wider scale such as ASEANþ3, ASEANþ6 (EAS) and APEC will foster a sense of community, and bring peace as well as economic prosperity.

Notes 1. The regional level could include ASEANþ3 (ASEAN10 plus China, Japan and Korea), or ASEANþ6 (ASEANþ3, India, Australia and New Zealand). 2. When transport costs fall, at first, manufacturing will move to a country with a variety of intermediate goods, thereby organizing a core–periphery structure. As transport costs fall further, when the advantage of low wage rates offsets the disadvantage of being far from markets and suppliers, manufacturing will move from a high wage country to a low wage country. This hypothesis implies that whether less developed countries such as Cambodia, Laos and Myanmar can succeed in industrialization depends on how much they can reduce border barriers. 3. When a company purchases products or components from another company and resells the products or components with the purchasing company’s name or logo on them.

References AEM-METI (2006). The thirteenth consultations between the ASEAN Economic Ministers and the Minster of Economy, Trade and Industry of Japan, 23 August, Kuala Lumpur, Malaysia. Anderson, James E. and Eric van Wincoop (2004). ‘Trade Costs’, Journal of Economic Literature, 42: 691–751. Ando, Mitsuyo and Fukunari Kimura (2003). ‘The Formation of Production and Distribution Networks in East Asia’, National Bureau of Economic Research, Working Paper, 10167. Baldwin, Richard and Charles Wyplosz (2004). The Economics of European Integration, London: McGraw-Hill.

From De Facto to De Jure Economic Integration 25 Freudenberg, Michael and Thierry Paulmier (2006). ‘A Comparison of De Jure Economic Integration in East Asia: Is East Asia Discriminating against Itself?’, in D. Hiratsuka (ed.), East Asia’s De Facto Economic Integration, Basingstoke: Palgrave Macmillan. Fujita, Masahisa, Paul Krugman and Tomoya Mori (1999a). ‘On the Evolution of Hierarchical Urban Systems’, European Economic Review, 43: 209–51. Fujita, Masahisa, Paul Krugman and Anthony J. Venables (1999b). The Spatial Economy: Cities, Regions and International Trade, Cambridge, MA: MIT Press. Hiratsuka, Daisuke (2006). ‘Vertical Intra-Regional Production Networks in East Asia: a Case Study of the Hard Disc Drive Industry’, in D. Hiratsuka (ed.), East Asia’s De Facto Economic Integration, Basingstoke: Palgrave Macmillan. JETRO (2007a). ASEAN Logistics Networks 2006, CD Rom, JETRO. JETRO (2007b). FY2006 Survey of Japanese Firms’ International Operations (http:// www.jetro.go.jp/en/news/releases/20070228845-news/Survey1.pdf, 7 March 2008). Krugman, Paul (1991). ‘Increasing Returns and Economic Geography’, Journal of Political Economy, 99, 3: 483–99. Krugman, Paul and Anthony J. Venables (1995). ‘Globalization and the Inequality of Nations’, Quarterly Journal of Economics, 110, 4: 857–80. McCallum, John (1995). ‘National Borders Matter: Canada–US Regional Trade Patterns’, American Economic Review, 85, 3: 615–23. Ministry of Foreign Affairs (MOFA) (2004). ‘The Progress Report of the Trilateral Cooperation among the People’s Republic of China, Japan and the Republic of Korea’ (adopted on 27 November 2004, by the Three-Party Committee in Vientiane, the Lao PDR). Venables, Anthony J. (2004). ‘European Integration: a View from Spatial Economics’, paper prepared for the international symposium on ‘Globalization and Regional Integration from the Viewpoint of Spatial Economics’, Institute of Developing Economies, Japan External Trade Organisation, Tokyo, 2 December. Viner, J. (1950). The Customs Union Issue, New York: Carnegie Endowment for International Peace.

This page intentionally left blank

Part I The Making of Economic Integration in East Asia

This page intentionally left blank

2 The Modality of East Asia’s Economic Integration Fukunari Kimura

2.1

Forming a basic architecture of economic integration

Until recently, East Asia was lagging behind the worldwide boom of regionalism, but this is no longer the case. Table 2.1 shows the networks of bilateral free trade agreements (FTAs) in East Asia and other neighbouring regions as of November 2007. A double circle represents an FTA signed and/or entered into force, a single circle denotes an FTA that has been agreed or already under negotiation, and a triangle shows an FTA under consideration or for which a feasibility study has been initiated. ASEANþ3 (China, Japan, Korea) is almost filled with concluded FTAs except for Northeast Asia. FTAs in ASEANþ3þ3 (India, Australia, New Zealand) also seem to be almost within reach in the short run. The ASEAN free trade area (AFTA) has become a hub of FTAs in East Asia. A growing number of trans-Pacific FTAs are also observed, stimulating discussion on a variety of competing ideas for future plurilateral FTA arrangements. However, the argument about a basic architecture of East Asian economic integration is far from settled. As the wave of integration has influenced various sectors and covered a number of countries in the region, the dimension of discussion has necessarily become multifaceted and entangled. It has become increasingly difficult even to share the understanding of ‘facts’ on the interface of the economy and economic policies. This chapter tries to fill the wide gap between a simplistic tariff-only argument on FTAs and an extremely ambitious argument on constructing an East Asian ‘community’ by reorganizing our views on the current status of de facto economic integration in East Asia. Section 2.2 discusses the complicated relationship between de facto and de jure economic integration in East Asia, which is different from an unwarranted simplistic view that de jure integration determines the 29

30 The Making of Economic Integration in East Asia Table 2.1 Matrix of FTAs involving countries in the Asia-Pacific region (as of November 2007) Japan Japan Korea China ASEAN India Australia NZ

˚o ˜ A ˚¢ A ˚o ˚ 6A ˚ u` ˜A A ˚o ˜ A ˚o ˜ A

Korea

China

ASEAN

India

AUS

NZ

˚o ˜ A

˚¢ A ˚¢ A

˚o ˚ 6A ˚ u` ˜A A ˚ u`* A ˚ u` A ˚ u` A ˚o ˜ A ˚o ˚ 6A ˚ u` ˜A A ˚o ˚ 6A ˚ u` ˜A A

˚o ˜ A ˚o ˜ A ˚¢ A ˚o ˜ A

˚o ˜ A ˚¢ A ˚o ˜ A ˚o ˚ 6A ˚ u` ˜A A

˚¢ A ˚o ˜ A ˚o ˚ 6A ˚ u` ˜A A

˚¢ A ˚ u`* A ˚o ˜ A ˚¢ A ˚¢ A

˚ u` A ˚¢ A ˚o ˜ A ˚o ˜ A

˚ u` A ˚ u` A

˚ u`: Entered into force or signed (*Korea-Thailand has not signed yet) A ˚o ˜ : Under negotiation or agreed to negotiate A ˚o ˚ 6A ˚ u`: Some bilateral FTAs are A ˚ u` while plurilateral FTA is A ˚o ˜A ˜ A ˚ ¢: Under consideration (G-G base) or feasible study initiated A

characteristics of de facto integration. In fact, de facto integration has progressed in East Asia, and its initial characteristics have influenced industrialization through vertical production networks. Thereafter, feedback between the two has developed. Section 2.3 reviews the current status of de jure economic integration in terms of trade in goods in East Asia by reviewing FTAs already in place and under negotiation. A possible ‘spaghetti bowl’ phenomenon is briefly discussed while considering the current attainment of de facto economic integration. Section 2.4 evaluates the current status of de jure economic integration in various fields other than tariff removals. It argues that policy coordination is crucial in further promoting international production/distribution networks or de facto economic integration in East Asia. Section 2.5 provides an overview of the current debates on the geographic architecture of East Asian economic integration and section 2.6 presents concluding remarks.

2.2 De facto and de jure economic integration in East Asia Usually when economic integration is discussed the central issue is typically how and to what extent de jure economic integration, e.g. the conclusion of FTAs, will promote de facto economic integration. In such a context, the causal arrow is primarily regarded as heading from de jure to de facto. To understand the nature and characteristics of economic integration in East Asia, however, it is necessary to critically reorganize our thinking on the relationship between de jure and de facto integration. In East Asia, de facto economic integration has deepened to a

The Modality of East Asia’s Economic Integration 31

substantial level while de jure economic integration has followed with a time lag. Then multifaceted feedbacks are observed between de facto and de jure in both directions. It is therefore necessary first to understand properly the characteristics of de facto economic integration in East Asia. In this regard, the following three points are important. First, de facto economic integration in East Asia has advanced rapidly, but the East Asian economy is still far from being perfectly integrated. International trade theory has a concept of ‘integrated world economy equilibrium’, which defines an extreme case of economic integration.1 By using this concept, the degree of economic integration can conceptually be measured as a diversion from the perfectly integrated economy equilibrium in terms of the mobility and price equalization of various economic elements including goods, services, technology and others. Such a test concludes that East Asia’s de facto economic integration still has a long way to go, compared with, for example, that of the European Union. In East Asia, intra-regional trade ratios have drastically increased, and intraindustry trade instead of inter-industry/one-way trade has rapidly expanded in proportion. However, factor prices, particularly wages, differ widely across countries, and countries are actually at different stages of development. In fragmentation and agglomeration, production processes are extensively fragmented beyond national borders in order to take advantage of differences in wages and development stages. Thus, development issues will certainly continue to be central in the coming decades, together with economic integration in East Asia.2 Second, de facto economic integration has developed primarily with market forces, but not in the absence of policy support. On the contrary, the policy environment and the setting of economic and politico-economic conditions have been extremely important for the development of de facto economic integration in East Asia. In general, developing countries have actors both in opposition and in support of trade liberalization and the introduction of foreign capital. In ASEAN countries, however, the capability of extending infant industry protection policy has been relatively weak on the side of both entrepreneurs and governments, and thus from the early stages of development the introduction of foreign capital has taken the form of a dual-track approach that has simultaneously promoted both import-substitution-type and export-oriented-type foreign direct investment (FDI). As FDI has mostly been introduced in the manufacturing sector, rather than in the natural-resource-exploiting sector, the fear of foreign capital has been relatively light. Since the recession in the latter half of the 1980s and the drastic policy transition in China in the early 1990s, ASEAN countries together with China have shifted their

32 The Making of Economic Integration in East Asia

policy support to export-oriented or network-forming industries, and the improvement of the business environment has been accelerated. Selective acceptance policy for FDI with complicated regulations and incentives has gradually receded, and aggressive introduction of FDI in a setting open to the international market has become a main feature. Although such policy changes may not be deliberately designed from the beginning, they are gradually formulated with the accumulation of small adjustments and trouble-shooting, in response to requests from foreign companies.3 Third, the fact that de facto economic integration starts before de jure economic integration certainly affects the path of the following political effort. Although the promotion of FDI is emphasized in the current FTA formation worldwide, the case of East Asia is special in that multinational enterprises (MNEs) themselves become influential actors who directly and indirectly influence the process and contents of de jure economic integration. Developing countries in the region face policy tasks on how to invite FDI in order to connect themselves with international production networks, form industrial agglomerations so as to stabilize the industrial structure, and help local firms to penetrate production networks. In the discussion on the ingredients of de jure economic integration, the importance of international production/distribution networks is inevitably emphasized. In addition, the boundary of production networks affects the formation of the concept of ‘East Asia’.

2.3 De jure economic integration in terms of trade in goods De jure economic integration in East Asia started from the formation of the ASEAN free trade area (AFTA) in the early 1990s. The initial motivation for AFTA was obviously that of promoting inward FDI and a fear of China as a great attractor of FDI. The direct effect of AFTA, however, was rather limited in the 1990s, as was the effect of a satellite arrangement called the ASEAN Industrial Cooperation (AICO) scheme.4 The slow start of AFTA is usually regarded as a product of its fragile implementation scheme based on loose peer pressure among member countries, but there are also some economic reasons. Although an active formation of production networks was observed in East Asia in the early 1990s, intra-ASEAN trade was not particularly large. At that time, export-oriented/network-forming FDI was mainly supported by country-specific incentive schemes including export-processing zones, bonded factories, and a duty-drawback system5 while import-substituting type industries were basically kept intact under trade protection. The situation, however, was drastically altered

The Modality of East Asia’s Economic Integration 33

by the eruption of the Asian currency and financial crisis. Deep recession in the domestic economy forced ASEAN member countries to restructure import-substituting-type industries. On the other hand, the intra-ASEAN networking of export-oriented/network-forming industries was considerably developed by then. The changes in the economic environment accelerated the process of AFTA in the late 1990s and thereafter. In terms of the coverage of trade liberalization and the rules of origin (RoO), AFTA ended up with a fairly clean FTA. The format of liberalization is to prepare four commodity categories, i.e. the inclusion list (IL) commodities, the tentative exclusion list (TEL) commodities, the general exclusion list (GEL) commodities, and the sensitive/highly sensitive list (SL/HSL) commodities; to specify the liberalization schedule of each category; and to classify commodities into the categories under each country’s discretion. The application of preferential tariffs is determined by the rule of commodity-wise reciprocity. In six ASEAN original countries (Brunei, Indonesia, Malaysia, the Philippines, Singapore and Thailand), about 98 per cent of commodities are already incorporated in the IL category where virtually all commodities have 0–5 per cent tariffs. Most of these tariffs will be removed by the beginning of 2008, and the complete tariff removal is supposed to be implemented by the beginning of 2010. ASEAN latecomers (Cambodia, Laos, Myanmar and Vietnam (CLMV)) will follow the same path with some lag; the due dates of cutting tariffs to 0–5 per cent are 2006 for Vietnam, 2008 for Laos and Myanmar, and 2010 for Cambodia, and the complete tariff removal will be obtained by 2018 in these countries. The RoO is basically a simple cumulative 40 per cent value-added rule. The liberalization coverage meets the requirement of GATT Article XXIV, namely ‘substantially all the trade’, though it takes time to complete tariff removals over a period of more than ten years, which is claimed to be allowed under the Enabling Clause. Although the actual usage of preferential tariffs started slowly, the utilization ratios have steadily increased over the past few years. According to information from the ASEAN Secretariat, the ratios of exports under preferential schemes to total intra-ASEAN exports by Thailand and Malaysia are 27.5 per cent and 15.5 per cent in 2004, respectively.6 Taking into consideration the existing duty-free status of semiconductor-related products and other existing tariff evasion schemes such as AICO and the duty-drawback system, we must regard these ratios as fairly high. Even under the control of loose peer pressure, AFTA seems to be working effectively. Reshuffling of industry location choices among ASEAN countries has been accelerated, particularly in former import-substitution-type industries such as automobiles, domestic electric appliances, iron and steel and others.

34 The Making of Economic Integration in East Asia

The trade-in-goods portion of ASEAN-China FTA (ACFTA) and that of ASEAN-Korea FTA (AKFTA) were quickly negotiated and concluded because they basically applied the same method of tariff removals and the same system of RoO as AFTA had used.7 A novel feature of ACFTA is the introduction of the ‘early harvest’ scheme, which allows some of the agricultural and fishery products trade (HS01-08) to be free of charge before the whole negotiation is concluded.8 Otherwise, the tariff removal scheme is very similar to the AFTA method. ACFTA sets up commodity categories as normal track 1 (most of the commodities), normal track 2 (within 150 commodities), sensitive list (within 400 commodities and less than 10 per cent in trade values), and highly sensitive list (40 per cent of sensitive list commodities and within 100 commodities). Each category is accompanied by its own liberalization schedule. Commodities in normal track 1 and normal track 2 will be free of tariffs by 2010 and 2012, respectively. Commodities in the sensitive list require tariff reduction to less than 20 per cent by 2012 and to 0–5 per cent by 2018. Commodities in the highly sensitive list will be less than 50 per cent tariffs by 2015. The application of preferential tariffs is again under the rule of commodity-wise reciprocity. The liberalization will be designed so as to follow the liberalization under AFTA with a few years’ time lag, and a large proportion of important manufactured goods are classified in the sensitive or highly sensitive list so that the effects on trade will be small in the short run. The RoO is basically the same as AFTA; i.e. the cumulative 40 per cent value-added rule is applied. Bilateral FTAs between Japan and ASEAN countries do not apply the AFTA method.9 Tariff removal negotiations follow the commodity-bycommodity negotiation method rather than the category-setting method. Such a method sometimes helps to achieve a high-level commitment to liberalization in key sectors, one example of which is the automobile sector in Malaysia.10 On the other hand, it may allow a low coverage of liberalization once a country has other ways to provide counter-offers, an example of which is the agricultural sector in the negotiation between Japan and Thailand.11 The RoO is also set on the commodity-by-commodity basis, which stems from the belief that the change in commodity classification method or the product-stage method would be more convenient than the value-added method for some commodities. The uncoordinated formation of FTAs in East Asia has become a serious concern for some policy-makers and academics. Networking of bilateral FTAs does not automatically mean regional integration. If some commodities are traded from country A through B to country C,

The Modality of East Asia’s Economic Integration 35

bilateral FTAs between A and B and between B and C may not help; a trilateral FTA is required among A, B and C in order to trade without tariff. A plurilateral FTA is certainly needed for the complete economic integration of the region.12 The semi-governmental study group of ASEANþ3 actually discusses the possibilities for strapping some of the existing bilateral FTAs so as to straighten out the trade liberalization scheme and RoO for the whole region. It should, however, be noted that the current status of de facto economic integration is notably different from the situation in other parts of the world. International production/distribution networks in East Asia have already generated massive international trade, particularly of machinery parts and components, and such trade has largely been free of tariffs.13 East Asian countries conducted unilateral tariff removals for semi-conductor related parts and components under the initiative of the Information Technology Agreement (ITA) in the latter half of the 1990s. In addition, there was extensive use of the duty-drawback system and other trade facilitating measures. Thus, international transactions by network-forming industries were conducted almost under free trade. This, of course, does not mean that East Asian countries are almost under free trade; they still keep fairly high tariffs for import-substitutiontype industries. However, we can at least say that such sectors with heavy trade protection are no longer crucially important in the realm of international trade in East Asia. The so-called ‘spaghetti bowl’ phenomenon, first described by Bhagwati (1995), has recently been highlighted in the media. Although the spaghetti bowl phenomenon has never been formally defined to my knowledge, the usual claim is that unorganized formation of bilateral FTAs may deter international trade through the complication of RoO and discretionary exemptions from liberalization lists.14 In the context of East Asia, such criticism is sometimes directed towards Japanese bilateral FTAs. Such claims, however, must be considered with caution. First, the complication of RoO and others would reduce additional trade creation due to the conclusion of an FTA but would not be likely to deter the original trade flows. The conclusion of an FTA certainly requires the negotiation cost, and the cost of thicker tariff books may be another expense. However, other than those, it cannot be a big deal for private firms. If firms find a new FTA to be cumbersome, they simply give up using the preferential tariff and keep using the original most-favoured-nation (MFN) tariff. A new FTA may be useless but cannot be too harmful. Second, the convergence of the RoO system may not necessarily facilitate international trade. The AFTA method based on the cumulative

36 The Making of Economic Integration in East Asia

40 per cent value-added rule seems to be popular among negotiators but may not necessarily be very convenient for private firms to trade commodities with complicated production processes. The facilitation of international transactions is important, not the unified system of RoO.

2.4 De jure economic integration in terms of policy measures other than tariffs The restructuring of import-substitution-type industries can be taken care of by tariff removals. Further activation of network-forming industries, on the other hand, requires policy reform in addition to tariff removals. The ‘comprehensiveness’ of FTAs is often claimed in terms of the coverage of policy modes but has a special implication in East Asia because of its motivation for de jure economic integration generated by the nature of de facto economic integration. The mechanics of international production/distribution networks require a multifaceted policy support. In order to make the fragmentation of production processes at the firm level, two elements are essential: (i) the saving of production cost per se in production blocks, and (ii) low service link cost for connecting production blocks (see Figure 2.1 for an illustration). As for (i), firms take advantage of differences in various aspects of location advantages. Differences in wage levels in different locations are still very important, but the flexibility in cutting out each production block provides a larger degree of freedom in designing production/distribution networks. In other words, elements of location advantages can be more explicitly multidimensional. As for (ii), tariff

Before fragmentation Old big factory After fragmentation PB

SL

PB

SL PB

SL PB

SL

PB

SL PB: Production Block SL: Service Link

Figure 2.1 Fragmentation: an illustration

The Modality of East Asia’s Economic Integration 37

removal is certainly important, but other types of trade/FDI facilitation also become crucial factors. Furthermore, fragmentation at the firm level goes together with agglomeration at the industry level in East Asia. Arm’s length transactions, i.e. transactions between unrelated firms, require geographical proximity, and such forces generate industrial clusters. Policy support is again extremely important in the formation of efficiencyenhancing agglomeration.15 In this respect, bilateral FTAs between Japan and ASEAN countries present the most advanced features at this moment in time. AFTA, ACFTA and AKFTA are still limited to tariff removals; they are accompanied by additional efforts in trade in services, investment and others, though such effort does not really seem to generate huge impacts.16 In reviewing non-tariff features of bilateral FTAs between Japan and ASEAN countries, we have to follow up not only the contents of concluded FTAs but also the attached agreements to FTAs and even the process of negotiations. As a matter of fact, FTAs concluded by Japan already significantly affect the business environment in ASEAN countries. Bilateral FTAs between Japan and ASEAN countries include four particular elements. The first is trade and FDI facilitation. Trade costs have recently attracted academic interest in the international trade literature.17 In the context of East Asia, they are particularly crucial to international production/distribution networks in the form of service link costs in fragmentation. Japanese FTAs and various parallel talks emphasize the improvement of customs clearance, logistics, transport infrastructure, one-stop service for FDI, services of industrial estates, and others; these ideas are extracted mainly from the private sector including both Japanese and local companies.18 Second, practical institutional building is incorporated, which includes investment rule, intellectual property rights protection, public procurement, and others. The Japanese are not good at the legal, rulemaking approach. Their approach is more pragmatic and emphasizes the improvement of the business environment in the context of production/distribution networks. There is also an emphasis on capacity building with economic/technical cooperation and the enhancement of stability and accountability of the policy environment. Third, the establishment of business–government dialogue is emphasized. In business activities, firms face various small issues and problems, which do not necessarily fit a costly government-to-government-based dispute settlement. It would be better to prepare a system of business– government dialogue for quick trouble-shooting. In this context, the development of local chambers of commerce is extremely important.

38 The Making of Economic Integration in East Asia

And the governmental sector should include both the central and local governments. Japanese FTAs take care of these issues either within or outside the text of FTAs. Fourth, links with other policy channels are emphasized. To further activate international production/distribution networks, FTAs must work with economic/technical cooperation policy, international finance policy, energy and environment-related policy, and others. Japan has various policy channels to pursue its economic objectives. Bargaining in FTA negotiations between Japan and ASEAN countries is not quite like the usual ‘tariff removals debate’ but is actually a question of ‘the improvement of business environment vs. economic/technical cooperation’. It should be noted that these four aspects are likely to be applied and become effective in a non-discriminatory way, rather than simply helping the Japanese and their related firms in the region. International production/distribution networks are not only for Japanese firms but also for others, including non-Japanese MNEs and local firms. The discriminatory features of FTAs are often criticized, but such criticism does not seem applicable in this context. One of the recent challenges for policy-makers is how to adopt these policy elements in countries at different developmental stages. Immediate policy requirements vary across countries. At an initial development stage, countries must attract FDI in order to connect with international production/distribution networks. Here, the focused improvement of location advantages and service links is required. When countries progress further up the ladder, the formation of agglomeration becomes a crucial task in stabilizing the industrial structure. Then, at the stage of development similar to ASEAN forerunners, the issue of particular significance is how to foster indigenous entrepreneurship and local firms and to make them penetrate international production/distribution networks. Given the current status of East Asia, which includes countries at different stages of development, the proper designing of FTA schemes and other supporting policies becomes a challenge.

2.5 Geographical boundary of de jure plurilateral integration The proliferation of bilateral FTAs in East Asia has recently stimulated discussion on the architecture of plurilateral FTAs and deeper economic integration. Based on the current boundary of de facto economic integration and the accumulation of meetings and studies, plurilateral

The Modality of East Asia’s Economic Integration 39

FTAs and an East Asian ‘community’ in the geographical boundary of ASEANþ3 would certainly be a plausible idea. Is there likely to be such a development? Could ASEANþ3 become an individual unit for deeper integration, even heading for political integration in the future as with the European Union? Countries such as China and Malaysia as well as some opinion leaders in the region seem to support a move in this direction. The author has some doubts concerning such a deepening of East Asian integration in a ‘closed architecture’ at least in the short and medium term. Rather, ‘open architecture’ and parallel developments of competing plurilateral FTAs seem to be more likely. The reasons are threefold. First, the formation of FTAs among Japan, Korea and China is likely to take several years or more. The three countries will probably enter into negotiation for a trilateral investment treaty as a first step towards a trilateral FTA, which would take at least another two years. Japan does not have any intention of negotiating a narrowly scoped FTA with China, i.e. only for trade in goods, and thus FTA negotiations will be started only after the conclusion of an investment treaty. Whatever the forum, it is not possible to have an East Asian-wide plurilateral FTA without connecting these three countries.19 Second, the core ingredients of de facto economic integration, i.e. international production/distribution networks, have a dynamic open setting. Key participants in the networks are certainly East Asian countries, but the networks also extend to the North American and European markets. India has started importing massive amounts of parts and components from East Asia and will eventually enter the networks. Transactions between Australia and East Asia are also increasing. Production/ distribution networks are always evolving and open to outsiders. In this sense, it is unlikely that ‘fortress East Asia’ will be realized. Third, a number of East Asian countries are concluding and/or negotiating FTAs with countries outside East Asia, in parallel with continuing the effort of East Asian integration; such countries include Singapore, Korea, Japan, China, Thailand and Malaysia. Therefore, in the discussion on plurilateral frameworks, it is very likely that trans-Pacific integration and others would be on the agenda running parallel with East Asian integration.20 ASEAN has long been criticized for being a lenient organization directed by weak peer pressure but in fact it has proved to be a very successful model for developing countries. Despite its economic size, ASEAN has become a hub of FTAs, which is partly due to windfalls from the complication of Northeast Asian countries and also in part due to their own effort in strengthening the member countries’ integrity. ASEAN is

40 The Making of Economic Integration in East Asia

actually concluding FTAs with almost all major trading partners and is thus virtually demonstrating ‘open regionalism’. Their own efforts to deepen integration even beyond economic matters are accelerating. The effort to enhance integrity among ASEAN member countries greatly helps ASEAN latecomers attract attention to their development. At least in the short and medium term, ASEAN will continue to play a pivotal role in East Asian de jure economic integration.

2.6

Concluding remarks

This chapter has reviewed the current status of economic integration in East Asia, particularly in the context of interactions between de jure and de facto economic integration. We find that de facto economic integration, though it is still far from a perfectly integrated economy, affects the design of de jure economic integration in East Asia. This distinguishes East Asian economic integration from existing examples of economic integration in other parts of the world such as the EU and NAFTA. It would certainly be useful to draw lessons from other parts of the world, but the peculiar aspects of East Asian economic integration must be taken into account. The current attainment of de facto economic integration in East Asia is still accompanied with wide differences in development stages across countries and regions. Development issues will continue to be at the centre of political/economic debates in East Asia. It will be particularly important to design the ingredients of FTAs in order to reflect diversity in development stages. In other parts of the world, some negative perceptions of globalization still exist. Although its applicability may depend on initial conditions of each country/region, the East Asian model of economic development must surely be a useful reference for countries reconstructing development strategies. Regionalism is booming all over the world. However, simply concluding FTAs and liberalizing trade are not enough to take advantage of globalizing forces for economic development. Overcoming the fear of FDI is far more important. It is hoped that the rest of the world will pay due attention to the case of East Asia. Notes 1. 2.

See a typical textbook of international trade theory such as Helpman and Krugman (1985). As for the relationship between fragmentation and the development of East Asian countries, see Kimura (2006a). The nature and characteristics of international production networks in East Asia, particularly in the context of

The Modality of East Asia’s Economic Integration 41

3. 4.

5.

6.

7. 8. 9.

10. 11.

12.

fragmentation theory, is concisely summarized in Kimura (2006b). The seminal work of fragmentation theory is Jones and Kierzkowski (1990). Policy reform in East Asian developing countries is discussed in detail in Kimura (2004). The AICO scheme was initiated with AFTA, together with the ASEAN Investment Area (AIA), so as to accelerate the cross-border division of labour among ASEAN countries. It has been utilized to some extent by automobile and other sectors, though the virtual requirement for bilateral trade balancing has prevented it from reaching the expected level of usage. A duty-drawback system is a policy measure for providing a duty-free status for imported parts and components used in producing exported products. It may be interpreted as a more flexible version of export-processing zones or bonded factories. It is, however, quite often accompanied by a complicated combination of regulations and incentives as well as troublesome and time-consuming bureaucratic procedures. For example, firms must typically submit their production plan well in advance and have to handle raw materials under the scheme strictly separate from other materials. We can easily imagine that such a system would distort the economy and generate room for rent-seeking activities in exchange for duty-free status. East Asian countries have extensively used various kinds of duty-drawback system in order to attract FDI. Obtained from Sukegawa (2005). The use of preferential schemes in other ASEAN countries seems to follow with some time lag. Bureaucratic procedures for obtaining certificates of origin (Form D) under the AFTA scheme have often been criticized. However, some recent improvements have been observed in the procedure as well as ‘learning-by-doing’ on the user side. Firms seem to judge the preferential tariff scheme as being less troublesome than other measures for evading tariff imposition such as the duty-drawback system. The main portion of ACFTA started cutting tariffs in July 2005. The trade-ingoods portion of AKFTA was signed in August 2006 except by Thailand. The early harvest scheme started with Thailand in October 2003 while other ASEAN countries had some delay due to their domestic adjustments. Japan concluded FTAs with Singapore (effective in November 2002), Malaysia (effective in July 2006), Thailand (effective in November 2007), the Philippines (signed in September 2006), Brunei (signed in June 2007), and Indonesia (signed in August 2007), and started negotiation with Vietnam in January 2007. Japan has also negotiated over an FTA with ASEAN as a whole, the major negotiation of which was completed in November 2007. In the Japan-Malaysia FTA, Malaysia committed to liberalize trade in automobiles within ten years, a target which has not been reached even by AFTA. In the Japan-Thailand FTA, Japan is going to exclude a number of agricultural commodities from the liberalization list, such as rice, sugar, molasses and others. Ando and Kimura (2006) evaluate how far the agricultural sector in Japan was liberalized through concluding FTAs and check commodities that will be at issue in negotiating FTAs with other countries. Note that a plurilateral FTA still requires certificates of origin for trade among member countries as far as the tariff structure for non-member countries is

42 The Making of Economic Integration in East Asia

13.

14.

15. 16.

17. 18.

19.

20.

different across countries. In this sense, FTAs are different from customs unions, which can ‘theoretically’ abolish certificates of origin for intraregional trade. Kimura et al. (2007) present over-time changes in tariff revenue import ratios in East Asian countries to show that international trade by network-forming industries is virtually close to free trade. The growing literature on the theoretical aspects of RoO includes Krueger (1993, 1997), Krishna and Krueger (1995), Falvey and Reed (1998), Rosellon (2000), Rodriguez (2001) and Ju and Krishna (2002, 2005). Empirical studies on RoO include Estevadeordal (2000), Devlin and Estevadeordal (2001), Augier et al. (2004, 2005), Estevadeordal and Suominen (2005) and Carreere and de Melo (2006). To the author’s knowledge, though, any serious empirical analysis on the case of East Asia has yet to be conducted. For a more detailed explanation of the mechanics of international production/distribution networks, see Kimura (2006b). Korea does not seem to be very enthusiastic towards the improvement of the region-wide investment climate despite the fact that a number of Korean firms invest in developing East Asia and take advantage of international production/distribution networks. One reason would be the well-planned network construction with major trunk lines of vertical links by Korean firms as latecomers vis-a`-vis Japanese firms, so that investment climate issues are largely solved through local negotiations (see Chang and Delios, 2006). Compared with Japanese firms, location choices of Korean firms are much more concentrated in a limited number of places. The relative weight of ASEAN as the FDI destination of Korean firms is much lighter than that of Japanese firms; cumulative FDI ratios between FDI in ASEAN and FDI in China by Korean firms are about 1 to 2 while those by Japanese firms are roughly 2 to 1. FDI destination of Korean firms inside China is also strongly biased towards the Bohai Sea Rim Region, which amounts to 70 per cent in terms of the number of affiliates (see Kimura et al., 2006). See, for example, Anderson and Wincoop (2004). Various forums for Japanese firms provide ideas and suggestions for trade and FDI facilitation. See, for example, the website of the Japan Business Council for Trade and Investment Facilitation (http://www.jmcti.org/mondai/top.html). The bilateral negotiation between Korea and China may be initiated though; the conclusion of Korea-US FTA is likely to accelerate the process, without waiting for Japan. Japan, Thailand and others strongly claimed the inclusion of India, Australia and New Zealand when the East Asian Summit was organized in 2005, intending to somewhat dilute the political implication of having a ‘closed architecture’. The idea of ASEANþ6, i.e. ASEANþ3 plus India, Australia and New Zealand, advocated by Mr Nikai, the then Minister of Economy, Trade and Industry, Government of Japan, first confused everybody including Japanese and people in East Asia but ended up confirming the recent development of ‘open architecture’. The idea of APEC FTA (FTAAP) proposed by the United States in the APEC Ministerial Meeting in November 2006 was readily accepted by participating countries as one of the competing parallel ideas of plurilateral FTAs.

The Modality of East Asia’s Economic Integration 43

References Anderson, James E. and Eric van Wincoop (2004). ‘Trade Costs’, Journal of Economic Literature, 42: 691–751. Ando, Mitsuyo and Fukunari Kimura (2006). ‘Japanese FTA/EPA Strategies and Agricultural Protection’, mimeo available at http://www.coe-econbus.keio.ac.jp/ data/DP2006-024.pdf Augier, P., M. Gasiorek and C. Lai-Tong (2004). ‘Rules of Origin and the EU-Med Partnership: the Case of Textiles’, The World Economy, 27, 9: 1449–73. Augier, P., M. Gasiorek and C. Lai-Tong (2005). ‘The Impact of Rules of Origin on Trade Flows’, Economic Policy, 20, 43: 567–624. Bhagwati, J. (1995). ‘US Trade Policy: the Infatuation with Free Trade Areas’, in J. Bhagwati and A. O. Krueger (eds), The Dangerous Drift to Preferential Trade Agreements, Washington, D.C.: AEI Press: 1–18. Carreere, C. and J. de Melo (2006). ‘Are Different Rules of Origin Equally Costly? Estimates from NAFTA’, in O. Cadot, A. Estevadeordal, A. Suwa-Eisenmann and T. Verdier (eds), The Origin of Goods: Rules of Origin in Regional Trade Agreements, Oxford and New York: Oxford University Press: 191–212. Chang, Sea-Jin and Andrew Delios (2006). ‘Competitive Interactions between Global Competitors: the Entry Behavior of Korea and Japanese Multinational Firms’, JCER Discussion Paper No. 97 (August), available at http://www.jcer.or.jp/report/ discussion/index1.html Devlin, Robert and Antoni Estevadeordal (2001). ‘What’s New in the New Regionalism in the Americas?’ INTAL-ITD-STA Working Paper 6, Inter-American Development Bank (May). Estevadeordal, A. (2000). ‘Negotiating Market Access in the Americas: the Case of NAFTA’, Journal of World Trade, 34, 1: 141–66. Estevadeordal, A. and K. Suominen (2005). ‘Rules of Origin in Preferential Trading Arrangements: Is All Well with the Spaghetti Bowl in the Americas?’ Economia: Journal of the Latin American and Caribbean Economic Association, 5, 2: 63–92. Falvey, R. and G. Reed (1998). ‘Economic Effects of Rules of Origin’, Weltwirtshaftliches Archiv, 134, 2: 209–29. Helpman, Elhanan and Paul R. Krugman (1985). Market Structure and Foreign Trade: Increasing Returns, Imperfect Competition, and the International Economy, Cambridge, MA: MIT Press. Jones, R. W. and H. Kierzkowski (1990). ‘The Role of Services in Production and International Trade: a Theoretical Framework’, in R. W. Jones and A. O. Krueger (eds), The Political Economy of International Trade: Essays in Honor of R. E. Baldwin, Oxford: Oxford University Press. Ju, J. and K. Krishna (2002). ‘Regulations, Regime Switches and Non-Monotonicity when Non-Compliance is an Option: an Application to Content Protection and Preference’, Economic Letters, 77: 315–21. Ju, J. and K. Krishna (2005). ‘Firm Behaviour and Market Access in a Free Trade Area with Rules of Origin’, The World Economy, 38, 1: 290–308. Kimura, Fukunari (2004). ‘New Development Strategies under Globalization: Foreign Direct Investment and International Commercial Policy in Southeast Asia’, in Akira Kohsaka (ed.), New Development Strategies: Beyond the Washington Consensus, Basingstoke: Palgrave Macmillan: 115–33.

44 The Making of Economic Integration in East Asia Kimura, Fukunari (2006a). ‘The Development of Fragmentation in East Asia and Its Implication for FTAs’, in Daisuke Hiratsuka (ed.), East Asia’s De Facto Economic Integration, Basingstoke: Palgrave Macmillan. Kimura, Fukunari (2006b). ‘International Production and Distribution Networks in East Asia: Eighteen Facts, Mechanics, and Policy Implication’, Asian Economic Policy Review, 1, 2 (December): 326–344, available at http://www.blackwellsynergy.com/toc/aepr/1/2 Kimura, Fukunari, Kazunobu Hayakawa and Zheng Ji (2006) ‘Are Korean Firms Doing Well? Evidence from Shandong Province in China’, mimeo available at http:// faculty.washington.edu/karyiu/confer/seoul06/papers/kimura_hayakawa_ji.pdf Kimura, Fukunari, Yuya Takahashi and Kazunomu Hayakawa (2007). ‘Fragmentation and Parts and Components Trade: Comparison between East Asia and Europe’, North American Journal of Economic and Finance, 18, 1 (February): 23–40. Krishna, K. and A. O. Krueger (1995). ‘Implementing Free Trade Areas: Rules of Origin and Hidden Protection’, NBER Working Paper, No. 4983. Krueger, A. O. (1993). ‘Free Trade Agreements as Protectionist Devices: Rules of Origin’, NBER Working Paper, No. 4352. Krueger, A. O. (1997). ‘Problems with Overlapping Free Trade Areas’, in Takatoshi Ito and Anne O. Krueger (eds), Regionalism versus Multilateral Trade Agreements, Chicago: University of Chicago Press: 9–23. Rodriguez, P. L. (2001). ‘Rules of Origin with Multistage Production’, The World Economy, 24, 2: 201–20. Rosellon, J. (2000). ‘The Economics of Rules of Origin’, Journal of International Trade & Economic Development, 9, 4: 397–425. Sukegawa, Seiya (2005). ‘ASEAN no taigai keizai senryaku to sono inpakuto’ (Foreign Economic Strategies of ASEAN and Their Impacts), in Keiichi Umada and Hiromi Ooki (eds), Shinko-koku no FTA to Nihon Kigyo (FTAs by BRICs-ASEAN and Japanese Firms), Tokyo: JETRO: 98–121.

3 The East Asian Noodle Bowl Syndrome Richard E. Baldwin

3.1

Introduction

The manufacturing network in East Asia is a marvel of globalization, allowing the region to produce a vast range of world class goods. The key is the outsourcing of billions of different parts and components to factories spread across the region. In short, we can think of this as ‘Factory Asia’. This was set up by East Asian corporations and the corporations are Factory Asia’s ‘mid-level managers’ and middle management is doing a marvellous job. A potential problem lies in the fact that Factory Asia has no ‘top-level’ management. This chapter argues that the current state of East Asian regionalism is fragile due to three facts: • East Asians’ industrial competitiveness is heavily interdependent and hangs critically on the free flow of intra-regional trade. • The tariff-cutting that made Factory Asia possible was done unilaterally by ASEAN and China; they are not ‘bound’ in the WTO. What this so-called ‘bindings overhang’ means is that East Asian tariffs could go back up overnight without violating any WTO rules (indeed, we saw some of this during the 1997 crisis). Since the smooth functioning of Factory Asia is in everyone’s interest, the lack of WTO discipline would not be a serious problem if there were some mechanism for managing conflicts – making sure that small problems do not tumble out of control due to a lack of communication and coordination. • The third source of fragility is the fact that there is no ‘top-level management’ in the region to substitute for WTO discipline. European regionalism has both top-level management (the EU) and WTO 45

46 The Making of Economic Integration in East Asia

discipline (Europeans have bound their tariffs at very low levels). The same is true of North American regionalism. The implications of these three facts are straightforward. While it has long been recognized that East Asia is short of ‘international public goods’, progress on setting up such goods has been hindered by a lack of vision.1 Questions such as ‘who should take the lead?’ and ‘what should be the long-run goals?’ have no easy answer and so have been left unanswered. The time for East Asian vision, however, is gone. What East Asia needs now is management, not vision. Moreover, the need is pressing. Given its vast political, cultural and economic diversity, East Asia has always been a region prone to conflict among small nations (of course, it is also prone to conflicts among big nations but that is true of all regions of the world). Given the extreme supply-side integration of the region, such small-country conflicts could pose a threat to the competitiveness of Japanese, Korean, Chinese and ASEAN firms in the US and European markets. For example, a serious commercial dispute between, say, Malaysia and Thailand, or Indonesia and Malaysia could hinder the production of Japanese firms. If a simple bilateral conflict spun out of control and began to spread – as happened in Europe during the 1930s – Japanese firms might find it very difficult to provide low-cost, high-quality products to European and American markets in a timely fashion. The same is true for Korean, Chinese and ASEAN firms. It should also be noted that a new source of tensions is about to appear in the region. The commercially important elements of East Asian regionalism, the China-ASEAN FTA and the Japan-ASEAN bilaterals, are just starting to cut tariffs on a discriminatory (i.e. preferential) basis. Such discrimination has led to trade tensions in other regions. Although East Asia may be lucky enough to avoid such problems, the extreme interdependence of East Asia’s competitiveness suggests that such things should not be left to luck. The region should establish a ‘New East Asian Regional Management Effort’ with a reinforced ASEANþ3 being the most likely candidate for the job. The first priority should be to bind the region’s unilateral tariff cuts in the WTO. 3.1.1

Chapter plan

The rest of the chapter is organized into four sections. The next section, Section 3.2, presents the state-of-play in East Asian regionalism, arguing that East Asian trade arrangements are marked by a ‘noodle bowl syndrome’ (an unorganized tangle of bilateral trade deals). This

The East Asian Noodle Bowl Syndrome 47

section also explains how the noodle bowl arose by distinguishing three phases: ‘rampant unilateralism’, 1980–90; ‘regionalism delayed, unilateralism accelerated’, 1990–2000; and ‘real regionalism’, 2000 to the present. It argues that the development of ‘Factory Asia’ was the driving force in the first two phases, but China’s FTA demarche towards ASEAN triggered a domino effect in the region that led to the current ‘noodle bowl’ of FTAs. Section 3.3 fleshes out the argument that East Asian trade is fragile. It also contends that East Asia is about to experience a series of shocks as real tariff discrimination appears in the region for the first time. Section 3.4 presents some ideas on how the region might manage this fragility in the short run and remove it in the long run through the development of an East Asian free trade association modelled on Europe’s ‘other’ regional arrangement, EFTA. Section 3.5 presents concluding remarks.

3.2

Current state of play

When it comes to East Asian regionalism, the state of play is easily summarized – it is a mess. Dozens if not hundreds of trade deals are under discussion, under negotiation, or already signed. Even limiting the universe to the deals that have been signed or are near signing, it is clear that East Asian regionalism is marked by the ‘noodle bowl syndrome’. 3.2.1

The East Asian noodle bowl2

Figure 3.1 makes the point graphically. The figure shows each FTA in the region that has been or is near to being signed. Specifically, by the end of 2005 the region had signed what amounts to 57 FTAs. The ASEAN-China FTA (ACFTA) is counted as ten separate deals due to ASEAN’s rather unique method of preferential liberalization. The reality is extremely complex but roughly speaking China and each ASEAN chooses its own ‘sensitive list’ and bilateral market access depends upon the interaction of the two lists. Nations do not get preferences for items on their own list and are not granted preferences for items on the counterparty’s list. All bilateral links inside the ASEAN FTA (AFTA) are listed separately for the same reason. The fractured nature of ASEAN’s basic approach to preferential trade implies that the degree of market access faced by an AFTA exporter of any particular product varies according to the ASEAN destination market concerned. For example, we cannot view the MalaysiaIndonesia preferential tariff structure as identical, or even close to the Singapore-Philippines preferential tariff structure. Thus AFTA acts as if

48 The Making of Economic Integration in East Asia

China Japan Korea

HK

Myanmar Philippines

Cambodia

Thailand Indonesia

Laos

Vietnam Singapore

Brunei

Malaysia

Figure 3.1 The East Asian ‘noodle bowl’ syndrome Note: The map shows FTAs signed or under negotiation in January 2006. East Asia is defined here as the ten ASEANs, China, Japan and Korea. Source: Author’s compilation.

it were 45 bilateral trade relationships (10 times 9 divided by 2). Fortunately, the existence of ASEAN implies that 55 bilaterals are not completely dissimilar – ASEAN has imposed some discipline on rules of origin (RoO), product exclusion practices and phase-in modalities. Finally, Japan has signed FTAs with Malaysia and Singapore with these being de jure as well as de facto separate agreements (see Whalley and Banda, 2005 for a detailed assessment of ASEAN’s trade deals). But by the end of 2006, it seems likely that the region will have at least thirteen more deals – ten de facto bilaterals from the impending ASEANKorea FTA, and three more bilaterals between Japan and the advanced ASEANs, Thailand, the Philippines and Indonesia. That brings the total to 70. Japan may also conclude FTAs with Vietnam and Brunei and ASEAN as a whole, bringing the total to at least 75. Of course, someone could count these differently and come up with a different number, but that just serves to strengthen the basic point. A classic aspect of the noodle bowl syndrome is that it is not entirely clear how many deals there actually are in the region. How did East Asian trade relations get so tangled up?

The East Asian Noodle Bowl Syndrome 49

3.2.2

No real regionalism yet

Strange as it may seem given Figure 3.1, there is as yet no real regionalism in East Asia. Regionalism means preferential trade liberalization, i.e. discriminatory trade liberalization, since preference is just another word for discrimination. By that definition, there is not yet any real regionalism in East Asia. The two most commercially important sets of arrangements that have been signed – the ASEAN-China FTA and the ASEAN-Japanese bilaterals – are supposed to be phased in over the next five years, but they have not yet undertaken any serious discriminatory tariff-cutting. The only scheme that has been substantially implemented – AFTA – is not actually used (more on this below). How did Factory Asia get established without ‘real’ regionalism? The answer to this question reveals many critical facts about the region and many aspects of East Asian trade relations that are highly unusual by global standards. In answering the question, it is useful to distinguish three phases of East Asian regionalism. • Phase I, rampant unilateralism: From the mid-1980s to 1990, tariffs on intra-regional trade came down but this was due to unilateral tariff-cutting in the region driven by competition for investments and jobs related to the development of what can be called Factory Asia. • Phase II, regionalism delayed, unilateralism accelerated: From roughly 1990 to 2000, East Asia witnessed an acceleration of unilateral tariff cuts as China’s emergence heightened the competition among East Asians for jobs and investment linked to the ever expanding Factory Asia. Additionally, two formal arrangements began – the ASEAN FTA (AFTA) and Asian-Pacific Economic Cooperation (APEC) – but neither created much discrimination nor had much effect on trade flows, as we shall see. • Phase III, rampant regionalism: In November 2000, Chinese Premier Zhu Rongji triggered a domino effect by suggesting that China might be interested in an FTA with the ASEANs. This idiosyncratic initiative faced excluded nations with a new situation which in turn strengthened proFTA political forces in the excluded nations, especially Japan and Korea. The result was a sequence of FTAs, each of which served to further strengthen the pro-FTA forces in all nations in the region. The ASEAN-Japan FTAs, the Korea-ASEAN FTA and the Japan-Korea FTA are direct reactions to the China-ASEAN FTA according to my reading of history. Consider each phase in turn.

50 The Making of Economic Integration in East Asia

3.2.2.1

Phase I: unilateralism and competition for Factory Asia jobs

East Asian trade was suppressed before 1985 by three factors: by the very unequal distribution of economic size within the region (Japan was the only large economy), by the great disparities in development levels in the region (which implied the natural trade pattern would follow the North/South template of inter-industry trade and this entails low trade volumes), and finally by the ‘dual-track’ development strategies widely adopted in the region that blocked the imports of manufactured goods for final consumption while simultaneously fostering manufactured exports to nations outside the region (Ando and Kimura, 2005). Starting in the mid-1980s, three interlinked factors launched the expansion of intra-Asian trade and what might be called ‘rampant unilateralism’. The factors are: (1) Erosion of Japan’s comparative advantage in manufacturing. The phenomenal growth of Japanese incomes and wages in the 1980s and 1990s eroded the nation’s comparative advantage in manufacturing. Specifically, Japanese productivity growth in labour-intensive processes was outstripped by overall Japanese productivity growth. Wage growth, of course, is linked to overall productivity, so rising Japanese wages were not fully compensated by labour productivity growth in labourintensive activity and unit labour costs in labour-intensive processes started to rise. Japanese businesses reacted by seeking lower cost manufacturing sites for labour-intensive stages of production and the obvious solution was to offshore these stages of manufacturing to nearby East Asian nations. This tendency, which has been called the ‘hollowing out’ of the Japanese economy, started the development of what can be called ‘Factory Asia’ or the ‘Asian Manufacturing Matrix’.3 Instead of Japanese goods being made in Japan and sold in the US or Europe, a new pattern of so-called ‘triangle trade’ emerged. Firms that were headquartered in Japan would produce certain hi-tech parts in Japan, ship them to factories in ASEAN nations for labour-intensive stages of production (including assembly) and then ship the final products to Western markets or back to Japan. This division of East Asia into headquarter (HQ) economies and factory economies strengthened as Taiwan, Korea, Singapore and Hong Kong experienced their own ‘hollowing out’ and followed the lead of Japanese manufacturing companies in offshoring the most labour-intensive production stages to East Asian nations who had a comparative advantage in such tasks (i.e. nations whose low wages more than compensated for their low labour productivity).

The East Asian Noodle Bowl Syndrome 51

(2) Reduced cost of moving goods and ideas. Information technology advances and the falling costs of transportation, especially air freight, facilitated and accelerated the development of the Asian Manufacturing Matrix by making complex production structures easier and cheaper to manage while at the same time making them more flexible and more reliable. (3) China. China’s opening up and domestic pro-market reforms brought something like a half a billion low-wage/low-productivity workers to the gates of ‘Factory Asia’. This accelerated the erosion of the HQ nations’ comparative advantage in labour-intensive production processes while simultaneously expanding the attractiveness of the offshoring solution. In short, China added a pull-factor to push-factors and this quickened the hollowing out of the HQ economies ( Japan, Korea, Taiwan, Singapore and Hong Kong). Development of the Asian Manufacturing Matrix Figure 3.2 shows the number of plants from Japanese electrical machinery and automobile industries that set up in the listed East Asian nations (plainly, the number of plants is only a rough indicator of the actual degree of offshoring). The offshoring process started gradually but picked up

900 800

China

700 600 500 400 300

Thailand

200

Malaysia Indonesia

100 0 1975 Figure 3.2 China

Vietnam 1980

1985

1990

1995

2000

2004

Placement of Japanese plants in East Asia, 1975–2004; emergence of

Note: The figure shows the sum of plants in the auto and electrical machinery industries. Source: ‘The Coming Age of China-plus-One’, Fujita and Hamaguchi (2006) presentation at IDE-JETRO workshop Janauary 2006 (original based on data from ‘Kaigai shinshutsu kigyo soran’, various years).

52 The Making of Economic Integration in East Asia

speed in the late 1980s. Between 1975 and 1990, the total number rose three and a half times with almost half of this increase coming between 1985 and 1990. The figure also shows that China was not a major player in the competition for plant locations prior to 1990. The big locations in Phase I were Thailand and Malaysia. These plants generated new trade, almost all of it in machinery and much of it in parts and components. The typical ‘triangle trade’ involved a Japanese firm sending high-end parts to an affiliated plant located in, say, Thailand. In some cases the goods would be shipped back to Japan for either final sale or further processing, but often the goods were shipped for final consumption to the US or European market. More direct evidence on the development of the Asian Manufacturing Matrix comes from the IDE-JETRO international input–output table. This shows the country of origin of imported manufactured goods purchased by the manufacturing sector of each East Asian economy. Table 3.1 has three panels corresponding to the Matrix in 1985, 1990 and 2000. The top panel shows the situation in 1985 when the Matrix was very simple. With the exception of Singapore, East Asian nations sourced their imported manufactured inputs from Japan and the rest of the world – all the rows are dominated by zeros except those of Japan and the rest of the world (mainly the US and Europe). By 1990 (second panel), the Matrix was more complex: ‘Triangle trade’ still dominated the picture with the low-wage nations (first five columns) buying inputs from Japan and the rest of the world but providing no inputs in return. Now, however, Japan is not the only HQ economy. Taiwan, Korea and Hong Kong experienced their own hollowing-out phases and new triangle trade appears. This new triangle trade involves the shipment of parts from the new HQ economies to the ‘factory economies’ (China and the advanced ASEANs, Indonesia, Malaysia, the Philippines and Thailand). This can be seen from the emergence of new non-zero entries in the rows for Taiwan, Korea and Singapore. By 2000, the Matrix was really complex. Firms based in the ‘factory economies’ began to source parts from other factory economies rather than from the HQ economy alone. In particular, Thailand, Malaysia and China became important suppliers of parts to other ‘factory economies’ including to each other. In short, the input–output matrix went from simple triangle trade to a much more complex situation where the ‘factory economies’ were both makers and buyers of parts and components.

Table 3.1 Widening and deepening of the Asian manufacturing matrix, 1985, 1990, 2000 (share of manufactured inputs bought by column nation’s manufacturing sector from the row nation; numbers less than 2% are zeroed out; own-nation purchases are also zeroed out) 1985 Indonesia Malaysia Philippines Thailand China Taiwan Korea Singapore Japan RoW 1990 Indonesia Malaysia Philippines Thailand China Taiwan Korea Singapore Japan RoW

China

Indonesia

Malaysia

Philippines

Thailand

China

Taiwan

Korea

Japan

8% 16% 0% 0% 14% 3%

2%

3%

Singapore

3% 12% 15%

7% 14% 19%

4% 19%

9% 14%

12% 11%

7% 10%

8% 16%

8%

Indonesia

Malaysia

Philippines

Thailand

Singapore

Taiwan

Korea

Japan

10% 17%

8% 13%

5%

5%

3%

8%

8% 23%

7% 10% 20%

4% 2% 2% 8% 21%

3% 3% 14% 22%

3% 3% 2% 18% 44%

53

(Continued)

54

Table 3.1 (Continued) 2000 Indonesia Malaysia Philippines Thailand China Taiwan Korea Singapore Japan RoW

China

Indonesia

2% 2%

3%

2% 4%

7% 16%

Malaysia

4% 3% 5% 4% 13% 15% 20%

Philippines

Thailand

Singapore

Taiwan

3%

4%

2% 12%

2%

3% 5% 8% 6% 20% 20%

4% 3% 3% 4% 16% 17%

3% 5% 3% 4%

4%

19% 38%

14% 15%

Korea

Japan

7% 11%

4%

2%

Notes: The columns would sum to 100% if we had included each nation’s supply of inputs to its own manufactured sector (a number that is often greater than 50%) and if we had not zeroed out the low numbers (less than 2%). RoW equals Rest of World. Source: IDE-JETRO, Asian input–output matrix (7 sectors) for 1985, 1990 and 2000; see, for example, www.ide.go.jp/English/Publish/Books/Sds/ 082.html

The East Asian Noodle Bowl Syndrome 55

This rise of China’s position in the Matrix is especially noteworthy between 1990 and 2000. At the beginning of the decade, it neither bought nor sold much manufactured input in East Asia. By the end of the decade, we see many entries for the Chinese column (which shows its purchase pattern) and the Chinese row (which shows which nations depend a lot on inputs from China). The flourishing of intra-ASEAN trade is also clear from the comparison of 1990 and 2000. The message of Table 3.1 is clear. By 2000, the competitiveness of manufacturing firms in East Asia depended in a serious way on the smooth functioning of regional trade. A disruption of trade between, say, Malaysia and China, could cause serious problems for Japanese and Korean firms trying to sell in the US. Impact on East Asian tariffs: rampant unilateralism Japanese firms’ offshoring strategies fit in nicely with the export-track of the dualtrack development strategies being pursued by the ‘factory economies’. To attract such investment, the ASEANs unilaterally reduced their tariffs on triangle trade in what may be viewed as a ‘race to the bottom’. Often this came in the form of ‘duty-drawbacks’ and dutyfree treatment for plants located in Export Processing Zones (EPZs).4 While it is hard to find direct evidence that the unilateral tariffcutting in this phase was caused by competition for Factory Asia investment, anecdotal evidence is abundant (see, for example, Kuchiki, 2003). Regardless of what caused it, the unilateral tariff-cutting is obvious from Figures 3.3 and Figure 3.4. Figure 3.3 shows the average tariff calculated as total tariff revenue divided by total imports. This is the broadest definition of the average applied tariff since it encompasses all tariff schemes, including duty-drawback arrangements. Figure 3.4 shows a much narrower measure of average tariff, but one that more directly measures the extent to which tariffs were unilaterally reduced. Specifically, Figure 3.4 shows the applied MFN tariff rates on machinery trade (the category that accounts for the bulk of these nations’ imports). Both figures clearly show why the first two phases are called ‘unilateralism’. It is useful to think of this sort of tariff-cutting as ‘quasiregionalism’ because its effect is to reduce tariffs only on intra-regional trade, but it was not formally discriminatory. It was, in other words, de facto preferential tariffs liberalization that involved no de jure preferences. It has also been called Asian-style regionalism (run by businessmen rather than by lawyers and diplomats).

56 The Making of Economic Integration in East Asia

Philippines 20

Thailand

18

Korea

16

Malaysia

14

China

12

Indonesia

10

Singapore

8 6 4 2 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001

0

Figure 3.3 Unilateral tariff-cutting in Phase II Note: Tariff revenue divided by total imports. Source: Ando and Kimura (2005).

45 China

%

40 35

China, Taiwan

30

Province of Indonesia

25

Korea, Republic of

20 15

Malaysia

10

Philippines

5 2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

Thailand

1988

0

Figure 3.4 Formal unilateralism; reductions in applied MFN tariffs Source: Kimura et al. (2005: Figure 6).

3.2.2.2

Phase II: regionalism delayed, unilateralism accelerated

The second phase began in 1991 with the call by Malaysian Premier Mahathir for an East Asian Economic Community (EAEC). This led to the ASEAN Free Trade Agreement (AFTA) in 1992, but Mahathir’s vision

The East Asian Noodle Bowl Syndrome 57

for the region was much broader. His vision, however, was not appreciated in Washington. The US feared that an Asian-only economic bloc might come to involve or even be dominated by communist China, a nation whose economic resurgence caused concern (the US was still quite uncertain about Chinese motives in the early 1990s). According to my reading of history, the US countered Mahathir’s vision by pushing the Asian Pacific Economic Cooperation (APEC) in 1993. The US was adopting a well-known strategy in the regionalism game – undermining a preferential trade arrangement by proposing a larger one.5 The US’s diversionary tactic worked. East Asian nations cared and still care far more about the US market, and the US’s security role in the region was still seen as reassuring in the rapidly shifting post-Cold War world. The ‘exclusively Asian’ aspects of Mahathir’s vision struck many in the region as too risky. In 1993, the oxymoron ‘Open Regionalism’ was embraced by nations throughout the region and the APEC’s ‘Bogor Goals’ were adopted. This is why Phase II is called ‘regionalism delayed’. The source of the tag-line ‘unilateralism accelerated’ should be clear from Figures 3.3 and 3.4. APEC was explicitly designed to rule out preferential trade liberalization which is and always has been the defining element of regionalism. AFTA did little more to foster regionalism. By contrast, the unilateral tariff-cutting that had begun in the 1980s accelerated in the new decade. Extremely low utilization rates under AFTA How can we be sure that the average tariff shown in Figure 3.3 does not reflect AFTA’s preferential tariff cuts? We have two pieces of direct evidence. First, Figure 3.4 shows that the average applied most-favoured-nation (MFN) rate for the sectors responsible for most trade in the region (machinery) came down more or less in tandem with the average rate shown in Figure 3.3 so that at least much of the tariff-cutting came from unilateral reductions in non-discriminatory tariffs. The second piece of evidence comes from the fact that almost no trade took advantage of AFTA’s preferential (i.e. discriminatory) tariff rates. This is not widely recognized since the ASEAN Secretariat promotes statistics that measure progress by counting the number of tariff lines that have been partially or fully liberalized, and not the share of intraASEAN trade that is conducted over the preferential tariff. As we shall see, almost no one uses AFTA preferences. Information from the late 1990s shows that the AFTA preferences were not being used (see Figure 3.5). Overall, less than 3 per cent of the intra-ASEAN trade benefited from AFTA’s preferences. Getting the preferential tariff rate

58 The Making of Economic Integration in East Asia

0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0 Indonesia

Malaysia

Philippines

AFTA utilisation rate, 1998

Thailand

Total

AFTA utilisation rate, 1999

Figure 3.5 AFTA utilization rates Source: PriceCooperWaterhouse, Presentation to the 10th Meeting of the ASEAN DirectorsGeneral of Customs, 24 July 2002.

requires the importer to prove that the good actually originated in ASEAN (to avoid tariff fraud whereby goods from third nations are trans-shipped through an ASEAN country to gain preferential access). In AFTA this requires Form D. As it turns out, the vast majority of traders found it more advantageous either to pay the MFN applied rate (and thus avoid the administrative cost and delay of Form D) or to take advantage of other schemes such as duty-drawback programmes or duty-free treatment in export processing zones. Figures for 2002 are not much better according to JETRO (2003). That paper reports that only 11.2 per cent of Thailand’s imports from AFTA took advantage of AFTA’s Common Effective Preferential Tariff (CEPT). Malaysia’s data suggest that just 4.1 per cent of its exports to AFTA enjoyed the CEPT preference. By comparison, utilization rates below 50 per cent are considered very low for European FTAs (Augier et al., 2005). Why are AFTA preferences not used? The reasons for the low utilization rates are simple – AFTA’s margins of preference on the high tradevolume goods are too small to compensate for the administrative cost and there is delay in applying for preferential tariff treatment. Evidence for this can be seen in Figure 3.6 and Table 3.2. Figure 3.6 plots the intra-Asian trade for each of the 99 HS chapters for the year

The East Asian Noodle Bowl Syndrome 59

99 98 97 96 95 94 93 92 91 89 8890 87 86 83 82 81 80 79 78 76 74 75 73 72 71 70 69 68 67 66 65 64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 4140 39 3738 36 35 34 33 32 31 30 29 28 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 09 08 07 06 05 04 03 02 01 0

85

84

27

10,000,000

20,000,000

30,000,000

40,000,000

Figure 3.6 Intra-ASEAN trade by HS Chapter, 2003 Source: ASEAN Secretariat website, External Trade Statistics page.

2003. Each bar represents the value of trade in a single chapter. Of course, most of the labels are illegible but that is because most of the bars are equally small. Intra-ASEAN is completely dominated by Computer/machinery and Electrical equipment (HS chapter 84 and 85 respectively), although there is also non-negligible trade in Lubricants, Fuels and Oil (HS 27). Table 3.2 shows that the MFN tariffs on these high-volume goods are very low – less than 2 per cent, so AFTA’s preferences are irrelevant. Experience from Europe and NAFTA shows that even modest costs and/or delays from complying with rules of origin will induce exporters to pay the MFN rate, if the MFN rate is low. Given that the margins of preference are razor thin – between 0 and 1.5 per cent – it is no surprise that AFTA was almost never used. A second salient point from Table 3.2 sheds some light on APEC’s non-role. The low East Asian tariffs in the table are for the goods involved in Factory Asia trade, namely General machinery, Electrical machinery (including electronics) and speciality petroleum derivatives. This unilateralism was not extended to other manufactured goods such as Transportation machinery (cars and trucks), or Light industry. This outcome is perfectly logical when thinking of the unilateralism as being driven by competition for Factory Asia jobs and investment, but it makes little sense when one thinks about APEC’s volunteerism.

60

Table 3.2 Intra-East Asian preference margins vis-a`-vis EU and North America Sector

Mining products (HS 25-27) General machinery (HS 84) Electrical machinery (HS 85) Others Wood and paper Precision apparatus Agriculture Light industry Food and beverages Textiles and clothing Transportation machinery Pottery products Chemicals Basic metals All products

Exporter to East Asia

Preference margins

East Asia

North America

1.7 1.5 1.4 1.4 1.4 1.2 41 26.8 21.8 7.3 4.6 2.9 2.4 1.8 7.4

2.6 1.9 1.5 1.7 1.3 1.3 29.7 8.3 26.4 7.6 2.8 3.6 3 2.6 5.5

EU

Over N. American. exporters

Over EU exporters

0.9 0.4 0.1 0.3 0.1 0.1 11.3 18.5 4.6 0.3 1.8 0.7 0.6 0.8 1.9

0.3 1.4 0.3 0.6 0.5 0.7 10.7 4.1 9.8 3.1 5.3 1.3 1.4 0.7 1.5

1.7 2.5 2.2 2.6 1.5 2 30.9 12.8 25.8 7.8 8.6 4.4 2.7 2.3 7.2

Note: Tariff data for 2002; see the chapter for aggregation schemes. Source: Author’s reorganization of data drawn from Freudenberg and Paulmier (2005: Table 3).

The East Asian Noodle Bowl Syndrome 61

Tariff liberalization in most nations is determined by the balance of serious political forces within each nation. If one wants to understand why a nation’s government finds it politically optimal to remove a tariff that it previously found politically optimal to impose, one has to focus on domestic political factors in each nation. Group photos of APEC leaders in colourful shirts cannot be considered a driving force. If a trade initiative is to have an effect on tariffs, it has to alter the array of pro- and anti-trade forces within the nations concerned. For example, rising competition for Asian Manufacturing Matrix investment created new, pro-liberalization forces within the East Asian factory economies. By contrast, nothing in the APEC process had an independent impact on the domestic political forces inside APEC nations. Of course, APEC leaders were happy to assign credit to the Bogor process for liberalization that was done for totally unrelated reasons. From a public relations perspective, it is better to present liberalization as part of a broad, inclusive process rather than the result of a race-to-the-bottom competition among East Asian developing economies. 3.2.2.3

Setting the stage for Phase III

Two events at the end of Phase II set the stage for the ‘real’ regionalism that started in Phase III. Asian financial crises The spectacular growth of East Asian economies in the mid-1990s came to a screeching halt with the 1997 Asian financial crisis. The big exception was China; it escaped almost unscathed (see Figure 3.7) and this firmed up the nation’s role as primus inter pares in the gallery of East Asian developing nations. The Asian crisis had an important influence on East Asian regionalism, although it was not, in my reading of history, a pivotal event. In particular: • It deeply shook East Asians’ faith in their development strategies and their heavy emphasis on US and European markets. • It created a sense of commonality among East Asian nations and networking among policy elites. • Many East Asian firms that had been protected from global competition by high tariffs suddenly found themselves faced with the full fury of globalization; the resulting ‘industrial restructuring’ was far from systematic and it certainly was not optimal, but it did act as a brutal form of ‘natural selection’. The average East Asian firm that survived the 1997 crisis was systematically more capable of standing up to international competition than the average East Asian firm before the crisis. This was important since it exogenously reduced the

62 The Making of Economic Integration in East Asia

400 350 Indonesia Malaysia Philippines Singapore China Japan Korea HK

300 250 200 150 100 50 2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

0

Figure 3.7 Total manufacturing growth, 1990–2002 Source: Hiratsuka (2006); UNIDO Statistics Database, 3-digit ISIC Code (Rev. 2); total manufacturing output.

anti-liberalization pressure from import-competing firms. On the margin, this shift in political economy forces within each nation made it politically optimal to remove some tariffs that East Asian governments previously found it politically optimal to retain. • It revealed the unsuitability of APEC as a vehicle for regional cooperation. • China’s economic stability during the ‘storm’ gave the country a big boost in its competition for new jobs and factories from the evergrowing Asian Manufacturing Matrix (see Figures 3.2 and 3.7). China’s looming WTO membership Although the Asian crisis is usually viewed as the key turning point, China’s WTO membership was far more important in rearranging the alignment of pro-trade and antitrade forces in the region. As the 1990s drew to a close, China’s longrunning attempt to join the WTO began to look like it would succeed. Chinese membership, many believed, would provide an external lock-in of the reforms the nation had unilaterally adopted. This, in turn, greatly heightened the attractiveness of China as a location for FDI from the HQ economies ( Japan, Taiwan, Korea, Hong Kong and Singapore). This became an increasingly serious worry for the other factory economies in the region, especially Indonesia, Malaysia, Thailand, Vietnam and the Philippines. And this set the stage for the next phase of Asian regionalism.

The East Asian Noodle Bowl Syndrome 63

3.2.2.4

Phase III of Asian regionalism: China triggers a domino effect

At the November 2000 China-ASEAN Summit, Chinese premier Zhu Rongji broached the idea of an FTA between China and ASEAN. This Chinese initiative came as something of a surprise to ASEANs but it was in line with China’s behaviour in the 1990s. One key element of the Chinese economic development strategy of the 1980s and 1990s was the desire to avoid antagonizing non-Chinese states. Since China’s success in attracting industrial jobs and investment was increasingly viewed as a threat by some ASEAN nations, Zhu Rongji decided to make a big-hearted gesture that he hoped would assuage ASEANs’ concerns over Chinese competition for Factory Asia jobs and investments. The Chinese offer was generally welcomed by the ASEANs and led, three years later, to the ASEAN-China Free Trade Agreement (ACFTA), which promises to eliminate tariffs on almost all bilateral trade between China on the one hand, and the ten ASEANs on the other by the year 2010 (2015 for the poorer ASEANs). The first domino effect triggered by China’s move was among the ASEANs themselves. During the ACFTA talks, the ASEANs agreed to adopt the same deadlines and zerotariff goals for intra-AFTA trade.6 China’s proposal to the ASEANs set off alarm bells all around the region, but especially in the main HQ economies, Japan and Korea. Preferential liberalization between two of Japan’s major markets – China and ASEAN – would create discrimination against goods shipped from Japan. ACFTA markets absorbed more than a third of Japanese exports in 2003, with the figure set to rise much higher if China’s boom continues as expected. Importantly, the level of discrimination might be substantial since China and the ASEANs have relatively high MFN tariffs on many industrial goods, especially the sort of finished products at which Japan excels – consumer electronics, autos and the like. If Europe’s experience is a guide, Japanese firms might react to such discrimination by moving more production facilities to locations inside ACFTA. Korea faced an isomorphic problem. In the early years of the new century, no one knew what ACFTA might become, but Japan and Korea simply had to have a plan in case major tariff discrimination did arise. In short, China’s demarche in November 2000 triggered a domino effect that has produced the noodle bowl that East Asian regionalism is today. Japan’s and Korea’s reaction According to the domino theory and the historical precedent in Europe and North America, ‘Plan A’ for Japan should have been to redress the discrimination by joining ACFTA. This path, however, was blocked. ACFTA was not an organization that

64 The Making of Economic Integration in East Asia

outsiders could join.7 With Plan A out of the running, a Plan B was needed. This was always going to be convoluted since its goal was to redress possible discrimination from a trade bloc without formally joining the trade bloc. Other parts of the world’s experience with the regionalism game suggested two Plan-B responses: (i) form a separate trade bloc with other excluded nations (Japan-Korea FTA), and (ii) seek bilaterals with the smaller/more susceptible members of the new trade bloc ( Japan-ASEAN bilaterals). The Japanese government opted to follow both paths. The Korean government faced a similar situation and adopted similar solutions. If the Japan-Korea FTA is signed, it is likely to trigger another domino effect since China will face the possibility of discrimination in two of the three largest economies in the region. A Korea-US FTA would also trigger a domino effect, but unless the US proves willing to radically alter its FTA ‘template’ – i.e. the basic FTA deal offered to the many other nations that have signed or are negotiating FTAs with the US – the Korea-US FTA is likely to be politically unpalatable to Korea.8 Moreover, the US side is likely to have political difficulties with a Korea-US deal, even if it follows the template since the Korean manufacturing sector is far more competitive than that of other US FTA partners. Of course, high-minded geo-strategic reasons may offset the opposition of US and Korean special interest groups, so anything can happen, but such reasons do not yet appear to be pressing. 3.2.3

Summary

In early 2008, East Asian regionalism looks like it will be built on four pillars: the China-ASEAN FTA, the Japan-ASEAN bilaterals, the Korea-ASEAN FTA, and AFTA. The Japan-Korea FTA looks unlikely as of February 2008, but if it is agreed, it will form a fifth pillar and certainly trigger another round of domino effects. In short, trade relations in East Asia are a mess, or more precisely, it looks like they may soon become a mess since virtually no discriminatory liberalization has yet occurred in the region. AFTA is not used and the other FTAs have only just started cutting tariffs. If commitments are followed, however, things will soon change. All the signed FTAs promise to eliminate bilateral tariffs by the end of 2010 (2015 for the poorer ASEANs). If that actually happens, discrimination will emerge in a big way.

3.3

Fragility and emerging tensions

East Asian regionalism is based on a fragile system that will soon be subject to new tensions. This section fleshes out the facts and logic behind these points.

The East Asian Noodle Bowl Syndrome 65

3.3.1

Sources of fragility

The fragility of the East Asian system stems from the extreme interdependence of nations’ competitiveness, the lack of WTO discipline, and the lack of top-level management in the face of the noodle bowl complexity. 3.3.1.1

The noodle bowl syndrome

The level of complexity shown in Figure 3.1 implies that trade relations in the region are heavily interlinked. Since the degree of protection among any two nations can alter political pressures for protection in other nations – that is the basic mechanism of the domino theory – the overlapping and intersecting nature of the noodle bowl is a source of fragility. The outcome of a trade dispute between, for example, Japan and Malaysia in the auto industry, has the potential to cause problems for Malaysian–Indonesian trade relations, yet the Japan–Malaysia problems would be handled by a dispute settlement body that is completely unrelated to the one that would handle Malaysian–Indonesian problems. While various agreements and institutional arrangements may spontaneously coordinate when common problems arise, the complexity of the system makes this outcome far from certain. Who, for instance, would be in charge of convening a meeting of East Asian leaders in the event of a local trade dispute threatening to spin out of control? 3.3.1.2

Interdependence in Factory Asia

Factory Asia started simple, but now all the nations’ manufacturing sectors are tightly interlinked via regional trade. The facts and figures can be found in the bottom panel of Table 3.1 for 2000. Figure 3.8 makes the point with an example. The diagram shows the sources of the various parts of a diskdrive that is assembled in Thailand by the affiliate of a Japanese company. The disk itself is sourced from the US, Japan and Malaysia, the filter cap comes from Hong Kong, etc. Of course, the disk itself consists of several parts, some of which are imported from the US, Japan and/or Malaysia and the same can be said of most of the parts listed. If we traced out the ultimate source of every bit of a disk-drive, the map would be impossibly complex. And one should note that the disk-drive itself is an intermediate good that will be shipped somewhere and assembled – together with an equally complex web of sourced parts – into some electronic device. The key point here is simple: • The competitiveness of firms from each East Asian nation now depends upon good trade relations between all East Asian nations.

66 The Making of Economic Integration in East Asia USA Disk, Head, Suspension

China PCBA, Carriage HGA, Base, Head Suspension Hong Kong Filter Cap

Thailand Spindle Motor, Base, Carriage, Flex Cable, Pivot, Seal, VCM, Top Cover, PCBA, HGA, HAS

Japan Cover, Disk, Screw, Seal, Ramp, Top Clamp, Latch, Plate Case, Label, Filter PCBA, Suspension

Philippines Damping Plate, Coil Support, PCBA

Malaysia Base, Pivot, Spacer, VCM, Base Card, Top Clamp, Disk Singapore Cover, Screw, Pivot, PC ADP, Disk

Taiwan Top Clamp

Mexico Head

Indonesia Suspension, VCM, PCBA

Figure 3.8 An example of interdependence in ‘Factory Asia’ Note: This shows the nations where parts are sourced for a hard-disk drive assembled in Thailand; the disk-drives are then shipped on to various markets to be used in various electronic devices. Source: Adapted from Hiratsuka (2005: Figure 2).

The Japanese firm’s disk-drive business can be hurt by trade trouble between China and Malaysia, between Thailand and Indonesia, or indeed between almost any pair of nations in the region. The Japanese government may be able to look after Japan’s trade links with other East Asian nations, but it would be virtually powerless to mediate a dispute among other East Asian nations. It is not useful to highlight scary, low-probability events, but it is crucial to point out that regional trade flows are the key to each East Asian nation’s competitiveness in the world economy, if something happened to disrupt intra-regional trade. 3.3.1.3

Lack of WTO discipline: the bound versus applied tariff ‘overhang’

The interdependence of the region’s manufacturing sectors is quite common worldwide, even if the phenomenon is taken to extremes in East

The East Asian Noodle Bowl Syndrome 67

Asia. How do North America and Europe deal with the system-fragility created by trade in intermediate inputs? There are two answers: • When it comes to the trade where the interdependence is most marked – intra-regional trade – North America and Europe have well-functioning ‘top-level management’. The US makes sure NAFTA trade runs smoothly and the EU does the same for trade in the European and Mediterranean region. • There is also a trans-Atlantic exchange of investment and intermediate goods in the manufacturing sector, but just as China and Japan find it impossible to agree who should fill the CEO post when it comes to East Asian trade, the US and EU cannot decide who should be in charge of trans-Atlantic trade. Instead, they let WTO discipline provide the top-level management. In performing this role, the most important of the WTO’s rules is Article 2 – tariff bindings. Article 2 commits members to ‘bind’ their MFN tariffs. In plain language, a tariff binding is a commitment to never raise a tariff that is ‘bound’ in the WTO. As the result of fifty years of GATT rounds, European and North American MFN tariffs are bound at very low rates since they came down in tandem with intra-regional preferential tariffcutting. Article 2 discipline is one reason why North Atlantic regionalism had a happy ending. The discrimination caused by regionalism on either side of the Atlantic was tamed by multilateral tariff-cutting and Article 2 discipline. There are two key aspects to this historical precedent: • Progressive MFN liberalization made sure that margins of preference never got too big. Since discriminatory market access is a major source of inter-bloc trade tension, the GATT rounds tended to prevent deeper and wider preferences from fuelling inter-bloc trade tensions. • WTO bindings of EU tariffs on US exports and US tariffs on EU exports meant that a broad trans-Atlantic tariff war would be illegal under GATT. Of course, the WTO permits all manner of exceptions – anti-dumping duties, countervailing duties, safeguards and the like – but these are both limited to specific products and require specific and well-disciplined procedures. Moreover, the US and EU could, as sovereign states, decide to ignore WTO rules, but doing so could not be taken lightly.

68 The Making of Economic Integration in East Asia

To use an analogy, think of trade disputes as forest fires that can sometimes get out of hand. The WTO’s rules do not make the world trade system fireproof, but tariff bindings act as an enormous ‘firebreak’. Nations must think long and hard before they violate Article 2. More to the point, a government can face down domestic calls for retaliation by claiming that such moves would be GATT-illegal and would therefore have negative repercussions far beyond the specific trade dispute. Why can this not work for East Asia? East Asia’s tariff bindings overhang The East Asian Manufacturing Matrix depends upon four HQ economies – Japan, Korea, Singapore and Taiwan – and six factory economies – China, Malaysia, Thailand, Indonesia, the Philippines and Vietnam. Table 3.3 shows that extent to which GATT Article 2 disciplines the trade of these nations. Japan, like the US and the EU, has bound almost all its tariffs at quite low rates; the simple average bound tariff is just 2.3 per cent. Japan’s average applied rate is almost identical.9 Since Japan’s tariff bindings (‘Final bound’) match the tariffs that Japan actually charges (‘Applied’), it would be impossible for Japan to raise its tariffs much without violating the GATT’s Article 2. This fact alone makes it almost unthinkable for domestic political forces in Japan to ask the Government of Japan for such a thing. In other words, Japan’s post-war tariff-cutting is WTO disciplined.

Table 3.3 Bound and applied tariffs in East Asia

Japan Korea China Malaysia Thailand Indonesia Philippines Vietnam Taiwan

Singapore

Binding coverage

Final bound1

Applied1

99.6 94.5 100.0 83.7 74.7 96.6 66.8

2.3 10.1 9.1 14.9 24.2 35.6 23.4

2.5 6.7 9.5 9.1 13.3 6.7 5.8

57.6 21.9 24.3 80.8 n.a. 51.2 45.6

n.a. 3.2 2.7 3.1 3.4 1.2 5.4

100.0 69.2

4.8 6.3

5.5 0.0

53.5 100.0

n.a. 0.3

MFN duty free imports, share in total imports

Import duties as share of total imports

Notes: 1Simple average of ad-valorem duties, non-agricultural goods. Vietnam is in the process of joining the WTO. Source: WTO country profiles, 2005.

The East Asian Noodle Bowl Syndrome 69

The situation for Korea is broadly similar. Almost all its tariffs are bound at moderate rates – the average is 10 per cent – but the MFN tariffs that Korea actually charges are below the level at which they are bound. This means that Korea could raise its MFN tariffs by 3 percentage points on average without violating Article 2. In the world of trade diplomacy, this is called a ‘binding overhang’ since much of Korea’s unilateral tariffcutting in the past two decades is not covered by commitments not to raise them. Note that once one takes account of all the special tariff treatment that Korea unilaterally extends to its imports, the ‘real’ average Korean tariff is just 3.2 per cent (‘Import duties as a share of total imports’). GATT’s Article 2 discipline provides an important firebreak should the Korean government ever find itself in a tariff-raising firestorm of domestic political pressure, but not as much of a firebreak as it provides to the Japanese government. China’s binding overhang is worse than Korea’s by some measures. For example, counting all the unilateral tariff breaks China extends, its tariff revenue is only 2.7 per cent of the value of imports, but its unweighted average binding is 10 per cent. Of course, the WTO allows China to unilaterally withdraw the unilateral tariff concessions it made in its duty-drawback and Export Processing Zone deals, so the gap between the actual and bound rates means that the WTO does not provide much of a firebreak when it comes to the Chinese government resisting domestic calls for tariff hikes in the event of an intra-East Asian trade dispute. The binding-overhang problem is far worse for the four big ASEAN economies (Malaysia, Thailand, Indonesia and the Philippines). A much narrower range of their tariffs are bound, and these are bound at very high rates, 15–35 per cent. These nations unilaterally grant duty-free MFN treatment to over half their imports (this is part of the ‘rampant unilateralism’ referred to above), and the real average tax on imports is extremely low: about 3 per cent for the biggest traders, Malaysia and Thailand, about 1 per cent for Indonesia and 5 per cent for the Philippines. What this means is that the governments of the large ASEAN nations could not tell would-be trade war-makers in their domestic polity that they cannot raise their tariffs without violating their WTO commitments. If a tariff-raising ‘forest fire’ breaks out in the domestic political environments of these nations, WTO discipline will do very little to stop its spread. Vietnam is not yet a WTO member, so the Article 2 firebreak would do nothing to help the Vietnamese government resist domestic political pressure for higher tariffs in the event of a raging trade dispute. Taiwan’s

70 The Making of Economic Integration in East Asia

situation is similar to Korea’s. Singapore is a paragon of WTO discipline; the city-state bound almost all its tariffs at zero. 3.3.1.4

Lack of top-level management at Factory Asia

No organization or government is making sure that things run smoothly in East Asia, although there are several bodies that deal with various bilateral problems.10 In a region as interdependent as East Asia, however, bilateral bodies may very well prove insufficient for dealing with complex trade disputes that involve several nations in the region – especially disputes that lead to ‘reverse domino’ effects, i.e. tariff hikes in one nation that are reactions to tariff hikes in other nations. 3.3.1.5

Summary of East Asia’s system-fragility

If we think of East Asia as a magnificent factory then duty-free trade flows are the conveyor belts that bring the parts from one stage of processing to the next. But this is not a linear factory of the type common in the twentieth century. Factory Asia is a twenty-first-century, just-in-time factory where the inputs are arriving continuously from all directions. If even a few of the ‘conveyor belts’ break down, production in many parts of the factory will be disrupted. This intense interdependence has fostered and been fostered by unilateral tariff-cutting, but this trade liberalization is not disciplined by WTO rules. Nor is there any top-level management to ensure that localized trade disputes do not spread and thus threaten the smooth functioning of the system as a whole. If some unpredictable event starts a fight among even a small subset of the ‘middle-level managers’, everyone could be harmed. Such an outcome would harm all involved, so one hopes the middle-level managers could be persuaded to settle their differences. But whose job would it be to sit everyone down in one room to talk things through? ‘Fragile’ may be too strong a word for this system; indeed it might withstand all manner of pressures, but that would be by happenstance, not by design. No one designed Factory Asia and no one is in charge of making sure it continues running smoothly. 3.3.2

Emerging pressures

Why should one be concerned about the system? After all, things have gone well for a couple of decades, so why worry? This is wishful thinking based on a false premise. East Asia has no experience with discriminatory trade liberalization. The many FTAs that are in place have not yet created discrimination.

The East Asian Noodle Bowl Syndrome 71

As argued above, none of the FTAs in the region that are likely to be effectively implemented – the Japan-ASEAN bilaterals, the ASEANChina FTA and perhaps the ASEAN-Korea FTAs – has yet cut bilateral tariffs in a serious way. The region’s other arrangement, AFTA, created almost no discrimination since no one used it. Why does this matter? Real regionalism means discriminatory tariff liberalization and discrimination sometimes creates tension. For example, China will soon grant better market access to a Malaysia-based firm than it does to a Japanbased firm – a fact that will surely bother Japanese firms. In other regions of the world, such tensions have led to policy reactions, some of which have been disruptive to trade. The most common policy reaction has been to abort the implementation of the discriminatory liberalization. This is often the case when it comes to South–South FTAs like the China-ASEAN arrangement.11 A nastier outcome, however, is also possible. The classic case was the round of tariff escalations among the European trade blocs in the 1920s and 1930s. But even without referring to the nightmare-scenario of inter-war Europe, it is clear that bilateral liberalization can increase pressure on governments to offset some pressure on domestic industry by raising protection against exports from excluded nations. For example, the liberalization of the Malaysian auto sector with respect to Japan might lead to new political pressures inside Malaysia to raise tariffs against, say, Korean firms, or to delay liberalization with other ASEANs. If the FTAs signed in early 2008 actually do what they say they will do, East Asia will witness the emergence of a complex pattern of tariff discrimination. Figure 3.9 shows a striking fact: the biggest trade flows in the region are the only ones not covered by FTAs. The diagram shows the importance of the various bilateral trade flows in the region with small flows zeroed out to improve readability. The three largest bilateral trade flows – China–Japan, China–Korea and Japan–Korea – are not covered by FTAs and seem unlikely to become so in the medium term. Thus it is exactly the biggest flows in the region that will face discrimination. Korean firms selling into China will face discrimination compared to Thai firms; Chinese firms selling into Japan will face discrimination compared to Malaysian firms, etc. If the Japan-Korea FTA gets signed, then China will face discrimination in both Japan and Korea compared to Korean firms on the one hand and Japanese firms on the other. The logical implication of this fragility is obvious. The time for East Asian vision is gone; the time for East Asian management is now.

72 The Making of Economic Integration in East Asia

East Asian bilateral trade flows rounded to nearest 1% of regional trade, 2002; Width proportional to %. China HK & Macao

Japan

Myanmar Korea Cambodia Philippines Thailand Indonesia

Laos

Vietnam Singapore Brunei

Malaysia

Figure 3.9 East Asian bilateral trade flows rounded to nearest 1% of regional flows Source: IMF DOTS database.

3.4

New East Asian Regional Management Effort (NEARME)

East Asian regionalism could have developed from a visionary beginning, but it did not. Vision takes time, a lot of time; Europeans, for example, discussed the continent’s trade arrangements for more than a decade before the EEC was formed. In East Asia, events just moved too fast. In just four years – from 2000 to 2004 – the notion of economically significant FTAs in Asia went from unlikely to obvious. All East Asian nations have a stake in ensuring that regional trade is well managed, but nothing and no one takes responsibility for this ‘collective action’ problem. This is why the pressing need is for management, not vision. This section considers some aspects of a regional management effort that could keep the status quo running smoothly while discussion of vision takes its course. 3.4.1

NEARME: institutions

Because the need for management is pressing, it seems clear that the region should rely on existing institutional arrangements; it would

The East Asian Noodle Bowl Syndrome 73

take years to set up a new organization. Given this constraint, ASEANþ3 is probably the best-suited body for the management effort. It is very easy to criticize ASEANþ3 – for example it is unwieldy to have so many leaders in one place when they differ so greatly in terms of economic and trade importance – but it has many positive sides. Chief among these is that it already exists, so all manner of delicate diplomatic issues can be avoided (just the question of who should participate could take years to resolve). Second, ASEANþ3 has a track-record of cooperation and thus has some credibility in the region. Third, it is not viewed as a creature of any single regional or non-regional power. Fourth, it has not been viewed as a threat to nations outside the region. Fifth, it encompasses all the nations whose competitiveness in European and American markets depends upon intra-regional trade flows, i.e. the set of nations who have the greatest stake in ensuring that intra-regional trade is well managed, without including any significant nation that is a bystander when it comes to this ‘collective action problem’. Sixth, it would be very easy to enlarge the group to include new members if and when the need arose. ASEANþ3’s institutional foundations, however, would need some firming up. Although it is not an institution per se, it could quickly begin to operate as one by acquiring a secretariat of its own. A structure for the political direction of the grouping is already in place, but it would need a secretariat of its own with high-quality, expert staff (legal, economic, customs and trade matters). The GATT – i.e. the WTO before the 1994 Marrakech accord – would provide a good role model for the New East Asian Regional Management Effort. The GATT operated as an international organization without actually being an international organization. Indeed, although it would be a very bad idea to call the new body an ‘Asian WTO’ (given the fate of the Asian IMF idea), this is effectively what NEARME would be. The GATT operated on a consensus basis, engaged in ‘soft law’ surveillance, enforcement and adjudication, and yet was the keystone to managing the rapid development of the world trade system in the post-war period. More to the point, many economists believe that GATT’s rules and its peer-pressure approach were responsible for the fact that FTAs in Europe and elsewhere did not threaten the multilateral system. The GATT, in short, was the ‘management effort’ that ensured waves of regionalism in the rest of the world have gone smoothly. By focusing on regional issues and tailoring itself to East Asian sensitivities, NEARME could foster a level of management, i.e. discipline, in the region that is more solid than what the WTO offers under the

74 The Making of Economic Integration in East Asia

Enabling Clause and more adapted to regional needs than the discipline it offers under Article 24. 3.4.2

Management priorities

The immediate priority of this New East Asian Regional Management Effort could be to ensure that the emerging pattern of tariff discrimination does not get in the way of the smooth functioning of Factory Asia. In other words, the overriding priority should be to protect the current very favourable status quo in East Asian trade relations. 3.4.2.1

Priority 1: bind the MFN applied rates

The first priority in the New East Asian Regional Management Effort should be to trim the ‘bindings overhang’ and thus bring WTO constraints on East Asia’s low tariffs into line with the political constraints (the fact that all would lose from a tariff war). The first step would be to convince the ASEANþ3 nations to bind their applied rates in the WTO. The invigorated ASEANþ3 body should, as a bloc, insist that nations doing this get full credit in the Doha round market access talks. Indeed, this move might encourage other developing nations to bind their applied rates and this, in turn, would prove to be a very positive contribution to unblocking progress in the ongoing WTO round. As a bonus, it would give the new organization a high and strongly positive international profile, demonstrating to the world that the new grouping was WTO-friendly. Binding the non-preferential rates would also improve the investment climate in the region since it would make trade policy more predictable. 3.4.2.2

Priority 2: transparency and confidence-building deliverables

The second priority for a New East Asian Regional Management Effort would be to improve transparency in the region. There are several obvious ‘deliverables’ here. ASEANþ3 could set up an information clearing house for preferences, RoO and the like. Indeed, as part of the new management effort, the participating nations could commit themselves to ‘notifying’ the new ASEANþ3 secretariat of any and all trade policy or legal changes that affect trade in the region. In the rosiest scenario, the new management could implement a scheme like the one that the EU pushed on the Euro-Med region, namely the Pan-European Cumulation System (PECS). In the late 1990s, Europe was marked by a ‘spaghetti bowl’ of FTAs that was as complex as East Asia’s noodle bowl is today. The EU managed the spaghetti bowl by harmonizing the tariff structures of all

The East Asian Noodle Bowl Syndrome 75

the various bilaterals, imposing a common set of RoO and insisting on cumulation throughout the region. The EU hegemonic role in the region was critical to this effort. While there is no East Asian hegemony to play a similar role, and the outcome is unlikely to be as harmonious as PECS, the new NEARME could take measures to reduce the inconsistencies among East Asian bilaterals. The New East Asian Regional Management Effort could also agree minimum ‘standards’ for East Asian FTAs – trade agreements that meet these conditions could be called ‘ASEANþ3 compliant’. Of course, the point of departure would have to be a set of conditions that was mild enough to fit all the existing FTAs (just as the GATT had to ‘legalize’ existing preferences in 1947 such as those of the British Commonwealth). But even this would be an improvement over the current lack of discipline faced by the FTAs that are notified under the Enabling Clause. The value of such a set of standards would become apparent if nations failed to fulfil the liberalization promises they made in the FTAs. Given the vast differences around the world, the WTO has found it impossible to improve the discipline of FTAs, especially those notified under the Enabling Clause. In a sense, this ‘ASEANþ3 compliant’ status would provide a level of discipline intermediate to Article 24 and the Enabling Clause. Such a thing might prove attractive to other South– South FTAs. If history is any guide, the domino effect in East Asia will spread to many more countries in the neighbourhood. In Europe, for example, the playing out of several waves of domino effects has left the EU with preferential trade deals with every WTO member except nine. It is therefore conceivable that the thirteen members of the ASEANþ3 group will end up signing a very large number of bilaterals in the coming years. It would be very helpful to the smooth functioning of Factory Asia to have agreed some discipline on the nature of these arrangements before they are signed. As it stands, there is no systematic surveillance and enforcement mechanism covering all the East Asian FTAs. Given the diverse political situation of nations in the region, such surveillance and enforcement would have to be of the ‘fair broker’ or technocratic type. There is no way that ASEANþ3 members could agree to give such a body any sort of ‘teeth’ like those that the EU’s Commission has, but the very existence of an unbiased observer and reporter of facts can often go a long way to resolving disputes or avoiding them in the first place. Just to take one example, nations have often been known to implement subtle measures that ‘nullify and impair’ the effect of tariff reductions they have

76 The Making of Economic Integration in East Asia

committed to. If the tariff-cutting was done in the context of the WTO, an aggrieved party could always ask a WTO panel to make a determination on the matter. In East Asia, there is no single body for investigating such claims. Actually, the old GATT panels are a better analogy than the current system of WTO panels. Before the Uruguay Round, the findings of GATT panels could be blocked – and occasionally they were – but the very existence of such a fair-broker body reduced trade tensions around the world. While the over-riding need in East Asia is for management, the region will need a longer-term solution eventually. 3.4.3 Longer-term goals for a New East Asian Regional Management Effort The goal of regional organizations throughout the world is generally twofold: to foster political harmony and economic efficiency, all without actively causing harm to non-members. A very broad range of organizational and institutional arrangements have been used to attain these goals. These arrangements lie on a continuum of supranationality. At the supranational end of the spectrum is the EU with its legal system, supremacy of the EU Court, and decision-making by majority voting. At the other end of the spectrum are the many purely inter-governmental arrangements such as NAFTA. In between are regional trade arrangements – such as the European Free Trade Association (EFTA) – that are inter-governmental in terms of decision-making but which have institutions and organizations that induce and lock-in cooperation. The crucial dividing line between inter-governmental and supranational lies in decision-making. If the regional organization can make decisions that bind all the members, yet need not be approved by each member, then we can say that the organization is supranational. For example, the EU decides its common external tariff – including antidumping duties – on the basis of majority voting. Consequently, EU members frequently find themselves having to adopt trade policies that they do not want.12 In inter-governmental organizations such as NAFTA and EFTA, all nations must agree to everything. 3.4.3.1

EFTA: the appropriate European model for East Asia

The EU model is a non-starter for East Asia. Imagine how ridiculous it would be to propose an East Asian organization in which China would have to adopt a trade policy it disagreed with because Korea and Japan voted for it. When the EU was born, the transfer of national sovereignty to a supranational trade body was the goal and economic integration was the means. This attitude, which is rare even in Europe today, is not part

The East Asian Noodle Bowl Syndrome 77

of any East Asian government’s plans. And this fact is completely independent of the question of existing conflicts among potential members of any such arrangement. Whether we are talking about Japan and China, Indonesia and Singapore, or Thailand and Korea, this group of countries would not accept the pooling of sovereignty that an Asian version of the EU would imply. EU-style supranationality – even supranationality limited to trade policy – is not for Asia, either now or for the foreseeable future. NAFTA provides an alternate template, but it too is wrong for East Asia. NAFTA is one of the few regional arrangements outside of Europe that really ‘works’ in the sense of liberalizing trade that would not otherwise have been liberalized and significantly shifting trade patterns. However, NAFTA’s success has nothing to do with its institutional arrangements. It works because a single nation is in charge, de facto. The US market is eighteen times the size of Mexico’s and twelve times that of Canada. Since market size equates with negotiating power in trade arrangements, it is easy to see why NAFTA works. A single nation has 90 per cent of the negotiating power and yet really cares about the organization’s success for political, geo-strategic and commercial reasons. NAFTA’s institutions are not notably more developed than those of ASEAN: indeed, in many ways ASEAN is more highly structured since at least it has a permanent secretariat. It seems likely that EFTA provides the best model for the long-run shape of East Asian economic integration since EFTA is purely intergovernmental but it has an efficient secretariat and enough institutional structure to deal with new challenges such as new members and expansion of the integration regime. In its early days, EFTA included advanced industrial nations (e.g. the UK and Switzerland) but also developing nations (Portugal) and produced ‘special and differential treatment’ (SDT) rules for Portugal. Given the great disparities in East Asia, the Asian EFTA should also include SDT-like policies and a ‘structural funds’ set-up as in the EU. This arrangement would: • mitigate the hub-and-spoke bilateralism that has emerged in the region; this would be especially beneficial to small East Asian nations that have fallen into the spoke trap; • avoid the noodle bowl problem by bringing coherence to preferences and rules of origin; • create ‘conditionally open regionalism’ in East Asia, which would result in many of the efficiency aspects of APEC’s ‘open regionalism’ while still harnessing the critical political forces that are generated by preferential trade liberalization;

78 The Making of Economic Integration in East Asia

• and, since Japan and Korea are developed nation members of the WTO, the arrangement would need to obey Article 24 disciplines. This discipline would also make it less threatening to third nations. It should be noted that since Japanese MFN tariffs are very low, and Korean MFN tariffs are relatively low, membership in this arrangement would expose Union members to something close to world market prices. This has two merits. It greatly reduces the scope for inefficient switches from low-cost non-member suppliers to high-cost member suppliers (trade diversion). It also means that joining the Union would almost certainly foster further liberalization, as in the case of Mexico.

3.5

Concluding remarks

As of early 2006, preferential tariff liberalization had yet to begin in East Asia. The ASEAN-China FTA and the Japan bilateral FTAs with ASEANs are just starting to be phased in; the only FTA in the region that is close to being implemented, AFTA, does not count since almost none of the intra-ASEAN trades uses the preferences. Since ‘discrimination’ is another word for ‘preferential’, we can say that East Asia has not yet seen what happens when tariffs are cut on a discriminatory basis. To date, the smooth growth of intra-regional trade has been driven by unilateral tariff-cutting. Consequently, the remarkably harmonious state of regional trade relations during the last two decades gives us little indication of how the system will react to future tensions including those arising from the impending discriminatory tariff liberalization. Given the extreme interdependence of almost all East Asian manufacturers on intra-regional trade and given the near total absence of WTO discipline on the tariff reductions that made all this possible, the prospect of trade tensions should be worrying to all the governments and companies in the region. To date, the middle management of Factory Asia – the private companies – have done an excellent job of keeping East Asian trade running smoothly. But middle management cannot solve all the problems. If shocks to the system led some nations to start raising tariffs, rounds of tariff retaliation could lead to substantial disruption of intra-East Asian trade. Of course, everyone would be hurt by this, but it would not be the first time in the course of human events that small-minded disputes caused huge problems due to a lack of ‘top-level’ management. Avoiding such outcomes requires a broader

The East Asian Noodle Bowl Syndrome 79

perspective than that which any one nation or trade arrangement can provide. In short, managing the noodle bowl should be a priority for all East Asian governments and companies.

Notes 1. See, for example, Fukuyama (2005), Severino et al. (2005), or the East Asian Vision Group’s report on ASEAN’s website. 2. The mess was first dubbed the ‘noodle bowl syndrome’ in various media outlets, but it should be thought of as the East Asian version of Bhagwati’s famous spaghetti bowl problem. It is difficult to track down the exact spaghetti-bowl citation since Bhagwati does not provide a reference when he uses the term. The first reference I could find was in a column by New York Times journalist Peter Passel, where he quotes Bhagwati as saying: ‘spaghetti bowl of tangled, inconsistent trade standards that just can’t be good for efficiency’ (New York Times, 4 February 1997). 3. The best papers on what I call Factory Asia are Ando (2006), Ando and Kimura (2005), Kimura and Ando (2003a, 2003b, 2005), Ng and Yeats (2003), and Fukao et al. (2003). 4. Duty-drawbacks involve the suspension of tariffs on parts and components that are imported, subjected to some processing and then re-exported. EPZ are geographically specific areas in which specified goods are imported and exported duty-free since the goods do not enter into commerce in the country’s market but rather stay physically in the EPZ. 5. Most famously, this same strategy was used by the UK to delay or prevent the formation of a French–German trade bloc in the 1940s and 1950s. As the largest importer in the region, the possibility of improved access to the UK’s market was very attractive to exporters on the continent. Thus when Britain proposed a Europe-wide FTA as an alternative to deep economic and political integration among the Six, many Europeans were interested exactly because the broader formation would also include preferential access to the UK’s market. The Britishinspired proposal did undercut some of the political and economic pressure for deep integration, but the Six eventually rejected it and went off to write the Treaty of Rome and set up what became the EU. The UK attended the initial meeting but withdrew at an early stage (see Baldwin and Wyplosz, 2006: Chapter 1 for details). 6. Before the 2003 Bali II Accord, AFTA preferences had to be between 0 and 5 per cent; afterwards, intra-AFTA should be zero. The AFTA deadlines were also rearranged to match the ACFTA deadlines. 7. See Baldwin (1993, 1997) on the domino theory and Baldwin (2003) for its application to East Asia. 8. The Bush administration has a hard time putting together a Congressional majority for FTAs and so prefers to use new agreement ‘templates’ (i.e. combinations of concessions and exemptions by the US and concessions and exemptions by the counterparty) that have won Congressional approval in the past.

80 The Making of Economic Integration in East Asia 9. The applied average is a bit higher for ‘continent protection’ duties such as anti-dumping and countervailing duties. 10. For example, the China-ASEAN FTA says that it will set up a dispute-settlement arrangement, but it does not appear to be in place yet. The Japan-Malaysia FTA has set up a bilateral dispute-settlement arrangement and it is likely that the other Japan bilaterals will have similar mechanisms. AFTA has something of a dispute-settlement procedure but given the extremely disjointed nature of ASEAN, it is not very effective and has not been chosen as the focal disputesettlement body for ASEAN’s FTAs with China, Korea and Japan. 11. If the preferences lead to the relocation of firms from one poor nation to another (usually the larger of the two since firms prefer to locate in the large market), the ‘losing’ nation often calls off the deal. Indeed, of the hundreds of regional trade deals in the world, only a couple of dozen actually influence trade and only a couple of those are South–South arrangements. 12. A recent example came with the EU–China conflict over textile quotas. Several nations, including Sweden, the Netherlands and the UK, did not want quotas against Chinese exports, but they were outvoted in the Council of Ministers and therefore were forced to restrict the trade. See Baldwin and Wyplosz (2006) for details on the EU’s trade policy-setting institutions and procedures.

References Ando, Mitsuyo (2006). ‘Fragmentation and Vertical Intra-industry Trade in East Asia’, North American Journal of Economics and Finance, 17, 3 (December): 257–81. Ando, M. and Kimura, F. (2005). ‘The Formation of International Production and Distribution Networks in East Asia’, in T. Ito and A. Rose (eds), International Trade (NBER-East Asia seminar on economics, vol. 14), Chicago: University of Chicago Press (first version, NBER Working Paper 10167). Augier, P., Michael Gasiorek and Charles Lai Tong (2005). ‘The Impact of Rules of Origin on Trade Flows’, Economic Policy, 43, July. Baldwin, Richard E. (1993). ‘A Domino Theory of Regionalism’, NBER Working Paper 4465 (Cambridge). Published as: ‘The Domino Theory of Regionalism’, in R. Baldwin, P. Haaparanta and J. Kiander (eds), Expanding Membership of the European Union, Cambridge: Cambridge University Press, 1995. Republished as Chapter 23 (pp. 479–502) in J. Bhagwati, P. Krishna and A. Panagariya (eds), Trading Blocs: Alternative Approaches to Analyzing Preferential Trade Agreements, Cambridge, MA: MIT Press, 1999. Baldwin, Richard E. (1997). ‘The Causes of Regionalism’, The World Economy, 20, 7: 865–8. Baldwin, Richard E. (2003). ‘The Spoke Trap: Hub and Spoke Bilateralism in East Asia’, KIEP Discussion Paper, 04-02, December. Baldwin, R. and C. Wyplosz (2006). The Economics of European Integration, 2nd edn, London: McGraw-Hill (1st edn, 2003). Freudenberg, Michael and Thierry Paulmier (2005). ‘A Comparison of De Jure Economic Integration in East Asia: Is East Asia Discriminating Against Itself?’ IDE Working Paper. Also in D. Hiratsuka (ed.), East Asia’s De Facto Economic Integration, Basingstoke: Palgrave Macmillan, 1996.

The East Asian Noodle Bowl Syndrome 81 Fujita, Masahisa and Nobuaki Hamaguchi (2006). ‘The Coming Age of Chinaplus-One: the Japanese Perspective on East Asian Production Networks’, World Bank Background Paper. Fukao, K., H. Ishito and K. Ito (2003). ‘Vertical Intra-industry Trade and Foreign Direct Investment in East Asia’, Journal of the Japanese and International Economies, 17: 468–506. Fukuyama, F. (2005). ‘Re-Envisioning Asia’, Foreign Affairs ( January/February): 75–87. Hiratsuka, Daisuke (2005). ‘Vertical Intra-Regional Production Networks in East Asia: a Case Study of the Hard Disc Drive Industry’, IDE Working Paper. Hiratsuka, Daisuke (ed.) (2006). East Asia’s De Facto Economic Integration, Basingstoke: Palgrave Macmillan. JETRO (2003). ‘Current Status of AFTA and Corporate Responses’, November, Japan: JETRO. Kimura, Fukunari and Mitsuyo Ando (2003a). ‘Fragmentation and Agglomeration Matters: Japanese Multinationals in Latin America and East Asia’, North American Journal of Economics and Finance, 14, 3. Kimura, Fukunari and Mitsuyo Ando (2003b). ‘Intra-regional Trade among China, Japan, and Korea: Intra-industry Trade of Major Industries’, in Yangsen Kim and Chang Jae Lee (eds), Northeast Asian Integration: Prospects for a Northeast Asian FTA. Seoul: KIEP: 245–79. Kimura, Fukunari and Mitsuyo Ando (2005). ‘Two-dimensional Fragmentation in East Asia: Conceptual Framework and Empirics’, International Review of Economics & Finance, 14, 3: 317–48. Kimura, Fukunari, Yuya Takahashi and Kazunobu Hayakawa (2005). ‘Fragmentation and Parts and Components Trade: a Comparison of East Asia and Europe’, mimeo available at http://www.apeaweb.org/confer/hito05/papers/kimura_ etal.pdf Kuchiki, A. (2003). ‘Agglomeration and Exporting Firms in Industrial Zones in Northern Vietnam: Players and Institutions’, in M. Kagami and M. Tsuji (eds), Industrial Agglomeration, IDE-JETRO, Tokyo: JETRO. Ng, F. and A. Yeats (2003). ‘Major Trade Trends in East Asia: What are their Implications for Regional Cooperation and Growth?’ World Bank Policy Research Working Paper 3084, World Bank. Severino, R. et al. (2005) ‘Framing the ASEAN Charter: an ISEAS Perspective’ (Singapore: Institute of Southeast Asian Studies), available at http://www.iseas. edu.sg/Framing_ASEAN_Charter.pdf Whalley, John and O. G. Dayaratna Banda (2005). ‘Beyond Goods and Services: Competition Policy, Investment, Mutual Recognition, Movement of Persons, and Broader Cooperation Provisions of Recent FTAs Involving ASEAN Countries’, University of Western Ontario, Center for International Governance and Innovation (CIGI), March.

4 Rules of Origin, Local Content and Cumulative Local Content in East Asia: Application of an International Input–Output Analysis1 Ikuo Kuroiwa

4.1

Introduction

Rules of origin (RoO) are an integral part of all trade rules. In the case of free trade agreements (FTAs), RoO determine the ‘nationality’ of a product, and only products that are considered to have originated in FTA member countries are eligible for preferential tariff concessions. Therefore, it is evident that RoO significantly influence the effectiveness of FTAs. In East Asia (especially for ASEAN countries and China), local content or cumulative local content have become the most important criteria for determining the origin of a product. This chapter has three purposes: (1) to provide an overview of RoO relative to FTAs in East Asia; (2) to calculate local content in eight East Asian economies, including investigation of the factors of change in local content of some sectors; and (3) to develop a projection of the impact of both AFTA and ASEAN-China FTA based RoO on cumulative local content of ASEAN countries. In order to calculate the local content of industries in East Asia, the Asian international input–output tables for 1990 and 2000 were used. These tables register all international and domestic inter-industry transactions in East Asia. They thus may be used to help explain the rapidly changing and increasingly internationalized production network in this region. 82

Rules of Origin, Local Content and Cumulative Local Content 83

The chapter is structured as follows. First, RoO in East Asia are discussed. Second, methods of calculating local content are explored, and decomposition analysis of the factors of change in local content is introduced in a technical note. Third, local content as well as the ASEAN and ASEAN-China cumulative local content of the manufacturing sector is calculated. Value added-based local content, considered to represent real local content of these sectors, is also presented. Finally, local content of the electronics and automotive sectors (demonstrating respectively the largest decline and largest rise in local content between 1990 and 2000) is examined; reasons for these changes are then presented.

4.2

Rules of origin

In this section, RoO in FTAs are reviewed. Focus is then placed on RoO in East Asia with particular emphasis on AFTA and the AFTA-China FTA. 4.2.1

Rules of origin in FTAs

There are three types of RoO in FTAs: (1) Change in Tariff Classification (CTC), (2) Value Content (VC), and (3) Technical Process (TP). In CTC, origin is granted if the exported product falls into a different part of the tariff classification from any imported inputs that are used in its production. CTC may require the product to be altered in its chapter (2 digits under the Harmonized System), heading (4 digits), sub-heading (6 digits), or item (8–10 digits) in the exporting country. VC can be defined either as the minimum percentage of the value of the product that must be added in the exporting country (local content) or the maximum percentage of imported inputs (import content). TP defines certain manufacturing or processing operations that a product must undergo in the exporting country in order to confer origin, manufacturing, or processing procedures that do not confer origin. Some elements of RoO are designed to promote trade between FTA member countries. De minimis allows a certain percentage of imported inputs to be used without affecting the origin of the final product. The De minimis rule, however, applies to CTC and TP but does not affect VC. On the other hand, cumulation allows producers to import inputs from other FTA member countries without undermining the origin of the product. There are three types of cumulation rules: (1) bilateral cumulation in which imported inputs originating in an FTA partner country can be counted as qualifying content in addition to originating materials in an exporting country when used in the country’s exports to that partner country; (2) diagonal cumulation: in this case, imported

84 The Making of Economic Integration in East Asia

inputs originating anywhere in the FTA partner countries can be counted as qualifying content when used in a country’s exports to the FTA area; and (3) full cumulation: here, any processing activities carried out in any FTA partner country can be counted as qualifying content regardless of whether the processing is sufficient to confer originating status (Breton, 2003).2 4.2.2

Rules of origin in East Asia

In Table 4.1, six FTAs are compared in terms of CTC, VC ratio, TP, Cumulation, and De minimis. Most FTAs apparently employ multiple criteria for setting RoO rather than applying a single rule. Among the four East Asian FTAs, the Japan-Singapore and Korea-Chile FTAs are more stringent and therefore more restrictive than AFTA and the ASEANChina FTA. This is because Japan and Korea have depended heavily on the frameworks of existing FTAs, especially NAFTA. Thus, RoO in the FTAs of these two countries are similar to those of Western FTAs in terms of CTC, VC ratio, cumulation, and De minimis (Cheong and Kwon, 2005). In contrast, AFTA and the ASEAN-China FTA have a simple and uniform format for RoO. In these two FTAs, CTC is not necessary, and TP is not mentioned; they require only one criterion of 40 per cent local content.3 As shown below, the rule for cumulation is also less stringent.4 In order to be eligible for Common Effective Preferential Tariffs (CEPT) under AFTA and similar arrangements under the ASEAN-China FTA, a product imported into the territory of a member country from another Table 4.1 Rules of origin in FTAs NAFTA

EUMexico FTA

JapanKoreaSingapore Chile FTA FTA

AFTA

ASEANChina FTA

CTC

Yes

Yes

Yes

Yes

VC ratio TP

60–50% Yes

50–30% Yes

60–40% Yes

45–30% Yes

Cumulation De minimis

bilateral

bilateral

bilateral

bilateral

Not necessary 40% No mention partial

Not necessary 40% No mention diagonal

7%

10%

8–10%

8%

No mention

No mention

Sources: Breton (2003), Estevadeordal and Suominen (2003), Cheong and Kwon (2005), ASEAN Secretariat (2005).

Rules of Origin, Local Content and Cumulative Local Content 85

member country must satisfy one of the following conditions: (1) the product must be wholly produced or obtained in the exporting member country or (2) at least 40 per cent of the product’s content must originate from a member country. Condition 1 is relevant to primary products such as mineral and agricultural products,5 while Condition 2 is applicable to other products which may require a substantial amount of imported inputs in production. Regarding the cumulation rule, diagonal cumulation was initially introduced in both AFTA and the ASEAN-China FTA. However, this diagonal cumulation was later replaced by partial cumulation in AFTA. In partial cumulation, ‘if the material has less than 40 per cent ASEAN cumulative local content, the qualifying ASEAN national content shall be in direct proportion to the actual domestic content provided that it is equal to or more than the agreed threshold of 20 per cent’ (ASEAN Secretariat, 2005). Therefore, unlike diagonal and full cumulation, ASEAN cumulative local content is only partially taken into account as long as the intermediate inputs have more than 20 per cent but less than 40 per cent ASEAN cumulative local content.6

4.3

Calculation of local content

In this section, methods of calculating local content are explored. First, the RoO in AFTA (a prototype of the ASEAN-China FTA) are examined in detail. These RoO are then applied to the methods of calculating local content with the use of the Asian international input–output tables.

4.3.1

Rules of origin in AFTA

According to the RoO in AFTA, the criterion for maximum import content is given by the following formula: ½ðValue of Imported Non-ASEAN Intermediate InputsÞ þ ðValue of Undetermined Origin Intermediate InputsÞ  FOB Price ðof the finished productÞ  100%  60%7

ð1Þ

where the FOB Price is decomposed as follows: FOB ¼ Ex-Factory Price þ Other Costs for Export ðtransport costs from factory to port, storage and warehousing, port handling, brokerage fees, service charges, etc:Þ ð2Þ

86 The Making of Economic Integration in East Asia

Further, the Ex-Factory Price is decomposed into the following two terms: Ex-Factory Price ¼ Values of Intermediate Inputs þ Value Added ðwages, profits, etc:Þ

ð3Þ

On the other hand, since (Values of Intermediate Inputs) ¼ (Value of Imported ASEAN Intermediate Inputs) þ (Value of Imported NonASEAN Intermediate Inputs) þ (Value of Undetermined Origin Intermediate Inputs) þ (Value of Domestic Intermediate Inputs), Formula 1 may be rewritten as ½ðValue of Imported ASEAN Intermediate InputsÞ þ ðValue of Domestic Intermediate InputsÞ þ ðValue AddedÞ þ ðOther Costs for ExportÞ  FOB Price  100%  40%

ð4Þ

In the above formula, the left hand side is considered to represent ASEAN cumulative local content; thus, the above formula requires that the ASEAN cumulative local content be at least 40 per cent of the FOB price. 4.3.2

Calculation of direct input-based local content

In this study, the Asian international input–output tables for 1990 and 2000 were used to calculate the local content of manufacturing sectors in East Asia. Although there are limitations relative to consistency with the RoO in AFTA, these tables are the most comprehensive and suitable analytical tools to examine local content. The Asian tables are Isard-type international input–output tables, and they cover five ASEAN countries including Indonesia, the Philippines, Thailand, Malaysia and Singapore, three Northeast Asian countries or regions including China, Korea and Taiwan, and the countries of Japan and the United States. 4.3.2.1

Local content

Local content of domestic industries was calculated by removing the term ‘Value of Imported ASEAN Intermediate Inputs’ from Formula 4. Using the Asian international input–output tables, it may be calculated as follows: lcjS ¼ ð

n X i¼1

S S XSS ij þ Vj Þ=Xj ¼

n X i¼1

S aSS ij þ vj

ð5Þ

Rules of Origin, Local Content and Cumulative Local Content 87

where VSj and XSj respectively represent value added and output of Sector j in Country S, while XSS ij denotes the amount of Commodity i in Country S needed by Sector j in Country S (domestic intermediate inputs from Sectors i into j). Further, n is the number of sectors in Pn S SS each economy; i ¼ 1 aij and v j represent respectively the sum of domestic intermediate input coefficients and the value added coefficient of Sector j in Country S. In Formula 5, local content is determined by intermediate input and value added coefficients which represent the input structure of industry. No roundabout (or indirect) production process is considered, so they may be considered ‘direct input-based local content’.8 In Formula 5, however, XSj is valued with ex-factory prices rather than FOB prices. Unlike Formula 4, ‘other cost for export’ is not included in the numerator of Formula 5. Considering the size of ‘other cost for export’ vis-a`-vis the FOB price, however, the accrued errors do not seem significantly large. 4.3.2.2

ASEAN cumulative local content

As shown in Formula 4, ASEAN cumulative local content is the sum of local content and import content from ASEAN member countries. Therefore, ASEAN cumulative local content may be calculated as follows: acjS ¼ ð

n XX R"AS i ¼ 1

S S XRS ij þ Vj Þ=Xj ¼

n XX

s aRS ij þ vj

ð6Þ

R"AS i ¼ 1

where AS is a set of ASEAN member countries in which country S itself is included. It should be noted that Brunei, which joined ASEAN in 1984, and CLMV countries (Cambodia, Laos, Myanmar and Vietnam) that joined the ASEAN in the 1990s, are not covered in the Asian international input–output tables. In Formula 6, all imported ASEAN intermediate inputs are counted as qualifying content regardless of whether or not the processing is sufficient to confer originating status. In other words, full cumulation, instead of partial or diagonal cumulation, is implicitly assumed in Formula 6.9 4.3.2.3

ASEAN-China cumulative local content

According to the ASEAN-China FTA, ASEAN-China cumulative local content is the sum of ASEAN cumulative local content and import content from China; the same 40 per cent local content criterion is applied

88 The Making of Economic Integration in East Asia

for preferential tax concessions. ASEAN-China cumulative local content may then be calculated as follows: accjS ¼ ð ¼

n XX

R"AS i ¼ 1 n XX

XRS ij þ

aRS ij

þ

R"AS i ¼ 1

n X

S S XCS ij þ Vj Þ=Xj

i¼1 n X

aCS ij

ð7Þ þ

vjs

i¼1

where C stands for China. 4.3.3

Calculation of value added-based local content

Although Formulas 5–7 correspond with concepts of local content in AFTA or the ASEAN-China FTA, they may not demonstrate real local content of these countries or regions. This is because production of intermediate inputs supplied by domestic industry (which is always counted as qualifying content for tariff concession) may require other intermediate inputs imported from a second country. Therefore, value added may not accrue entirely in the first country; there may be leakage of value added from the first country. For example, if an engine is supplied by domestic industry, this portion will be given originating status for the domestic automotive industry. However, production of an engine may require substantial amounts of imported intermediate inputs (metal, engine parts, etc.), and the real local content may be significantly smaller than the direct input-based local content indicates. 4.3.3.1

Local content

Matsumura and Fujikawa (1998) present a method for calculating real local content. It is calculated by computing the proportion of valued added accruing to the domestic industry from the entire (direct as well as indirect or roundabout) production processes. Consistent with the notation used in this paper, real local content can be calculated as follows: lcSj ¼

n X

gSS ij

ð8Þ

i¼1 1 ˆ ˆ where ½gRS ij  ¼ C ¼ V(1  A) . V and A represent respectively a diagonal matrix of value added coefficients and an input coefficient matrix of the Asian international input–output table. gSS ij (an element in Matrix G) indicates value added which is induced by one unit of final demand for Sector j in Country S, and that accrues to Sector i in Country S. Therefore, lc*Sj represents the proportion of value added accruing to

Rules of Origin, Local Content and Cumulative Local Content 89

all domestic industries in Country S, and local content in Equation 8 may be called ‘value added-based local content’.10 4.3.3.2

ASEAN cumulative local content

Region-wide cumulative content is calculated by extending the countries or regions covered in Equation 8. For example, the value addedbased ASEAN cumulative local content is calculated as follows: acSj ¼

n XX

gRS ij

ð9Þ

R"AS i ¼ 1

4.3.3.3

ASEAN-China cumulative local content

Analogously, value added-based ASEAN-China cumulative local content is calculated using the following: accSj ¼

n XX R"AS i ¼ 1

4.4

gRS ij þ

n X

gCS ij

ð10Þ

i¼1

Empirical results

Although the Asian international input–output tables cover ten countries or regions, the prime focus of this study was only the local content of the eight East Asian economies. Japan and the US were entered only as trade partners of these countries; all transactions in the Asian tables were converted into a common 27-sector classification, as shown in Table 4.2.11 4.4.1

Changes in local contents by sector, 1990–2000

Table 4.2 demonstrates the local content of 27 domestic industries in eight East Asian economies for 1990 and 2000. These tables clearly show that the primary industries (Sectors 1–3) and non-tradable goods and services (Sectors 24–27) had higher local content. With some exceptions (especially in Singapore), these sectors had more than 80 per cent local content. Conversely, manufacturing sectors (Sectors 4–23) had lower local content, and some were even below the AFTA threshold of 40 per cent (see, for example, Refined Petroleum in Singapore and Thailand and Electronics in the Philippines, Malaysia and Thailand in 2000). It is clear that manufacturing sectors, which are most relevant to the RoO in FTA, should be given priority in the analysis of local content. It is also notable that local content declined in many sectors in the period 1990–2000. Table 4.3 shows (in descending order) changes in average

1. Agriculture 2. Crude Petroleum 3. Other Mining 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

Foods** Spinning Wearing Apparel Timber Pulp & Paper Basic Chemicals Chemical Products Refined Petroleum Rubber Products Non-Metallic Mineral Products Basic Metal Metal Products Machinery Electronics Other Electric Machinery Motor Vehicles Motor Cycles Other Transport Equipments Precision Machines Other Manufacturing Products

90

Table 4.2(a) Local content (1990) IND

PHI

THA

MAL

SIN

CHN

KOR

TWN

AVE*

98.7 99.0 98.8

97.0 99.6 82.4

94.3 96.6 97.4

93.5 96.2 91.1

74.5

97.6

68.8

98.1 99.7 99.1

99.3

97.1 92.0 85.6

93.8 97.2 90.3

97.0 76.0 91.1 98.1 87.6 72.7 71.0 81.2 84.7 92.4

93.8 69.0 58.7 91.5 66.4 76.7 71.6 35.5 77.9 88.1

91.3 77.1 84.2 77.9 59.4 78.7 69.3 44.6 89.9 94.5

91.2 73.3 70.3 96.7 71.0 84.1 79.1 69.7 96.7 85.6

65.3 55.4 61.1 71.8 64.9 56.8 69.7 19.2 39.4 69.3

94.0 94.9 78.2 92.7 94.5 91.7 94.6 96.0 93.5 99.5

91.5 86.0 77.7 66.9 85.7 73.5 82.2 36.1 84.9 91.3

86.8 89.7 84.5 74.2 84.7 74.6 74.2 39.4 85.8 87.2

88.9 77.7 75.7 83.7 76.8 76.1 76.5 52.7 81.6 88.5

80.3 80.5 47.7 66.1 79.5 77.8 81.9 73.3

68.3 78.4 75.5 52.2 72.9 66.6 90.2 88.5

66.5 58.7 66.7 40.5 57.1 53.7 69.2 80.0

56.6 56.0 71.3 56.2 52.0 56.8 89.4 66.1

62.2 59.7 58.3 47.3 60.7 53.2 56.1 68.0

96.3 95.4 95.1 83.2 97.0 90.3 77.6 88.4

81.0 90.0 87.1 72.3 91.5 91.2 85.7 82.3

77.8 81.3 81.4 64.4 81.4 79.7 85.5 76.1

73.6 75.0 72.9 60.3 74.0 71.1 79.5 77.8

80.8 52.7

82.7 70.0

79.9 68.1

61.8 48.4

56.6 57.1

89.1 87.0

81.4 88.6

68.2 87.9

75.1 70.0

24. 25. 26. 27.

Electricity, Gas, and Water Construction Trade & Transport Service

All (Sectors 1–27)***

94.2 88.3 96.8 95.6

92.7 87.6 94.1 95.0

99.1 83.6 93.2 97.0

90.8 80.2 95.9 94.2

74.1 73.2 75.6 83.3

99.4 97.1 98.7 99.0

85.6 97.0 93.4 97.5

80.1 94.4 93.4 93.0

89.5 87.7 92.6 94.3

92.4

87.8

84.8

85.0

63.4

95.5

89.0

85.9

85.5

Notes: * In all tables and figures below, average represents an arithmetic mean of the eight East Asian countries. IND, PHI, THA, MAL, SIN, CHN, KOR, and TWN, respectively, represent Indonesia, the Philippines, Thailand, Malaysia, Singapore, China, Korea and Taiwan. ** Shaded areas indicate the manufacturing sector. *** All (Sectors 1–27) represents local content of all sectors (Sectors 1–27) combined.

91

Table 4.2(b) Local content (2000) PHI

THA

MAL

SIN

CHN

KOR

TWN

AVE

96.4 97.8 96.9

92.6 91.9 87.5

94.5 99.5 98.9

92.2 89.0 85.1

76.9

97.4

93.0 94.5 89.7

75.9

98.8 97.1 95.1

98.7

95.1 91.8 79.9

4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

Foods Spinning Wearing Apparel Timber Pulp & Paper Basic Chemicals Chemical Products Refined Petroleum Rubber Products Non-Metallic Mineral Products Basic Metal Metal Products Machinery Electronics Other Electric Machinery Motor Vehicles Motor Cycles Other Transport Equipments Precision Machines Other Manufacturing Products

94.6 80.4 83.0 92.4 74.5 76.9 77.7 74.7 79.0 91.2 82.7 77.0 51.9 80.6 75.2 83.7 92.7 62.3 83.0 71.7

94.5 55.4 65.4 82.4 68.6 65.2 65.2 40.4 52.5 78.9 66.1 60.4 64.3 34.5 63.4 66.4 59.1 69.8 58.6 56.9

89.7 85.4 85.4 76.5 69.1 72.2 71.0 34.4 86.3 85.4 69.3 62.5 57.6 39.1 50.5 59.8 73.0 67.3 71.2 69.3

85.9 70.3 56.9 85.4 69.5 78.1 67.4 73.7 78.4 81.9 54.1 60.4 64.1 36.6 56.0 68.5 75.9 65.6 52.4 68.9

62.1 56.3 61.1 68.9 77.4 69.7 74.3 14.8 73.7 70.4 64.6 73.1 60.2 50.1 62.2 58.7 51.1 79.0 58.7 60.3

96.9 92.1 90.3 92.5 83.4 89.3 92.9 82.6 90.2 95.8 91.8 93.1 91.8 75.3 90.3 91.0 94.2 94.2 88.3 87.4

91.2 83.2 86.5 81.6 87.0 76.5 81.1 44.0 83.8 93.0 81.5 91.4 88.1 63.4 87.0 92.3 87.0 80.9 82.7 90.3

86.2 86.1 78.7 71.5 79.9 63.5 68.7 42.6 77.7 79.2 77.7 85.4 74.2 58.0 74.4 80.4 90.9 69.9 61.3 83.3

87.6 76.1 75.9 81.4 76.2 73.9 74.8 50.9 77.7 84.5 73.5 75.4 69.0 54.7 69.9 75.1 78.0 73.6 69.5 73.5

24. 25. 26. 27.

Electricity, Gas, and Water Construction Trade & Transport Service

94.1 83.9 91.9 96.4

84.5 85.6 91.2 93.4

97.9 83.4 98.7 96.2

91.0 76.7 83.7 86.7

92.9 81.7 79.1 86.6

96.6 94.8 97.3 96.5

78.4 96.9 87.1 97.3

98.4 89.2 92.8 96.0

91.7 86.5 90.2 93.6

90.0

80.1

82.0

69.9

69.3

93.2

87.2

83.1

81.9

All (Sectors 1–27)

Source: Asian International Input–Output Tables (1990, 2000).

92

1. Agriculture 2. Crude Petroleum 3. Other Mining

IND

Rules of Origin, Local Content and Cumulative Local Content 93 Table 4.3 Changes in local content 1990–2000* Motor vehicles** Other Manufacturing Products

3.9

Agriculture

0.9

Trade & Transport

2.4

3.6

Construction

1.2

Crude Petroleum

2.6

2.2

Foods

1.2

Machinery

3.9

0.4 0.2

Motor Cycles Spinning

1.5 1.5

3.9 4.0

Chemical Products Refined Petroleum Basic Chemicals Timber

1.7

Rubber Products Non-Metallic Mineral Products Other Electric Machinery Other Transport Equipments Precision Machines Electronics

Electricity, Gas, and Water Metal Products Wearing Apparel Basic Metal

0.1

Other Mining

0.6

Pulp & Paper

0.6

Service

0.7

1.8 2.2 2.3

4.1 4.2 5.5 5.6

Notes: * Numbers in the table indicate the changes in average in the period 1990–2000. ** Shaded areas indicate the manufacturing sector. Source: Asian International Input–Output Tables (1990, 2000).

local content for the eight East Asian economies between 1990 and 2000. Only 5 out of 27 industries increased local content in this period. Among them, automotive (þ3.9 per cent) and electronics (5.6 per cent) sectors are at the opposite ends of the spectrum. These two industries and an exploration of the causes of these changes are discussed later. 4.4.2

The manufacturing sector

In order to investigate local content of the whole manufacturing sector, Sectors 4 to 23 in the Asian tables were aggregated into one sector, and the methods of analysis introduced in the previous section were applied. 4.4.2.1

Local content

Table 4.4 demonstrates the input structure of the manufacturing sector in 2000. Local content is decomposed into three components: inputs from its own sector (the manufacturing sector), inputs from other domestic sectors, and value added components. Considering the size of each economy, it is clear that large economies, especially in terms of population, tend to have high local content. China and Indonesia, for example, had higher local content than more advanced economies such as those of Korea and Taiwan. Further, the contribution of China’s

94 The Making of Economic Integration in East Asia

own manufacturing sector (40.8 per cent) was outstanding; this reflects a relatively strong industrial base of China’s supporting industry. On the other hand, small and open economies, such as Singapore and Malaysia, had less than 60 per cent local content, and the ASEAN economies in general had lower local content than Northeast Asian economies. The second part of Table 4.4 shows import contents of the manufacturing sector. These include inputs from ASEAN, Northeast Asia (NEA), Japan, the US and the rest of the world (ROW). Although inputs from Japan declined significantly as shown below, they were still fairly large in 2000, exceeding 5 per cent in Singapore, Malaysia, Taiwan, the Philippines and Thailand. Inputs from the US also exceeded 5 per cent in Malaysia and Singapore. On the other hand, only Malaysia and Singapore had more than 5 per cent inputs from ASEAN or Northeast Asia. Looking at the Asian table, it is seen that bilateral linkages between Malaysia and Singapore were so strong that as much as 6.5 per cent and 5.9 per cent respectively in their ASEAN cumulative local content originated in bilateral trade. Figure 4.1(a) shows factors related to changes in local content in the period 1990–2000. These include: (1) changes in value added coefficients (changes in shares of value added in total inputs), (2) changes in trade structure (changes in trade coefficients or import shares), and (3) changes in technology (changes in technological input coefficients). Details of the methodology may be found in the technical note. Figure 4.1(a) shows that local content decreased in all countries with the exception of Singapore.12 In particular, local content in the Malaysian Table 4.4 Local and import content (2000): manufacturing sector

MAN* DOM VA LC** ASEAN NEA JAP USA ROW IC

IND

PHI

THA

MAL

SIN

CHN

KOR

TWN

AVE

18.0 30.8 35.4 84.2 1.5 2.0 1.7 1.3 9.3 15.8

11.1 22.6 31.5 65.1 4.5 4.9 6.4 4.9 14.3 34.9

18.9 20.4 28.1 67.4 4.0 4.0 6.0 2.8 15.9 32.6

18.5 15.8 22.9 57.1 10.6 6.1 7.7 6.4 12.1 42.9

16.3 13.8 21.4 51.4 8.5 5.0 8.4 5.6 21.1 48.6

40.8 23.5 25.7 89.9 0.7 1.9 1.2 0.6 5.7 10.1

35.5 15.8 27.7 78.9 2.0 1.7 3.4 3.0 10.9 21.1

28.6 18.3 24.1 71.0 3.9 3.1 6.8 3.9 11.3 29.0

23.4 20.1 27.1 70.6 4.4 3.6 5.2 3.6 12.6 29.4

Notes: * MAN, DOM, and VA indicate input coefficients of respectively the manufacturing sector, other domestic sectors, and value added. ** Local content (LC) þ Import content (IC) ¼ 100%. Source: Asian International Input–Output Table (2000).

Rules of Origin, Local Content and Cumulative Local Content 95

manufacturing sector was reduced by 18 per cent; this was followed by the Philippines with a 13.2 per cent reduction. It is also notable that changes in local content (C-LC) were in parallel with changes in trade structure (TRAD). This implies that trade structural change was a leading factor for local content changes. Figure 4.1(b) reveals how import content changed by country of origin in the period 1990–2000. Together, these figures show that changes in local content summed with changes in important content equal zero. The latter is further decomposed into changes in inputs from ASEAN, NEA, Japan, the US and ROW. It is notable that ASEAN economies had

5 0 IND

PHI

THA

MAL

SIN

CHN

KOR

TWN

AVE

5 10 15 20 C-LC

VA

TRAD

TECH

*C-LC  VA  TRAD  TECH

Figure 4.1(a)

Changes in local content (1990–2000): manufacturing sector*

20 15 10 5 0 IND

PHI

THA

MAL

SIN

CHN

KOR

JAP

USA

TWN

AVE

5 C-IC

ASEAN

NEA

ROW

*C-IC  ASEAN  NEA  JAP  USA  ROW  TRAD (Figure 4.1(c))  TECH (Figure 4.1(d))  ROW For ASEAN, NEA, JAP, and USA respectively, it holds that Figure 4.1(b)  Figure 4.1(c)  Figure 4.1(d)

Figure 4.1(b)

Changes in import content (1990–2000): manufacturing sector*

96 The Making of Economic Integration in East Asia 20 15 10 5 0 IND

PHI

THA

MAL

SIN

CHN

KOR

TWN

AVE

5 TRAD

ASEAN

NEA

JAP

USA

* TRAD  ASEAN  NEA  JAP  USA

Figure 4.1(c) Impact of trade structural changes (1990–2000): manufacturing sector*

20 15 10 5 0 5

IND

PHI

THA TECH

MAL ASEAN

SIN NEA

CHN JAP

KOR

TWN

AVE

USA

* TECH  ASEAN  NEA  JAP  USA

Figure 4.1(d) Impact of technological structural changes (1990–2000): manufacturing sector* Source: Asian International Input–Output Tables (1990, 2000).

sharply increased inputs from ASEAN; Northeast Asian economies had such increases relative to inputs from NEA. This implies that geographical proximity may have affected procurement of intermediate inputs. Inputs from Japan and the US did not increase significantly or even decrease, especially in Singapore and Korea. The latter indicates a significant shift in import content away from Japan and the US towards other East Asian economies. Changes in import contents from ASEAN, NEA, Japan and the US, as seen in Figure 4.1(b), were further decomposed into the impact of trade structural changes (TRAD) and technological changes (TECH) by country of origin in Figures 4.1(c) and (d) respectively (see Equation 17). East Asian economies in Figure 4.1(c), especially Singapore and Korea,

Rules of Origin, Local Content and Cumulative Local Content 97

clearly indicate the diversion of import shares away from Japan and the US to East Asia. Since the impact of technological changes was removed from import content changes, Figure 1(c) provides strong evidence that trade structural changes were a major driver in import content changes. On the other hand, Figure 1(d) shows that impact of technological changes was minimal, except in some ASEAN economies; this may reflect the relatively stable technological structure in Northeast Asia.13 4.4.2.2

ASEAN and ASEAN-China cumulative local content

As of 1990 and 2000, AFTA and the ASEAN-China FTA were not fully implemented, but it is still meaningful, using past trends, to project the impact of the cumulation rule on local content of ASEAN countries.14 Table 4.5 shows the ASEAN and ASEAN-China cumulative local content of the manufacturing sector. In 1990, ASEAN cumulative local content of each ASEAN member country was higher than the respective local content by from 0.9 per cent in Indonesia to 7.3 per cent in Singapore. Further, ASEAN-China cumulative local content exceeded ASEAN cumulative local content by from 0.2 per cent in the Philippines to 1.8 per cent in Singapore. These facts suggest that local content may significantly increase due to the cumulation rule in open economies such as Singapore and Malaysia. Although local content (other than in Singapore) decreased in the period 1990–2000, some of this decrease was offset by an increase in inputs from neighbouring ASEAN countries and China. Therefore, ASEAN as well as ASEAN-China cumulative local content did not decrease as much as local content per se. For example, due to stronger linkages with neighbouring ASEAN countries, ASEAN content in 2000 was higher Table 4.5 ASEAN and ASEAN-China cumulative local content: manufacturing sector IND

PHI

THA

MAL

SIN

AVE

LC (90)

85.0

78.3

73.5

75.2

47.5

71.9

AS (90)* AS-C (90)** LC (00)

85.9 86.3 84.2

80.1 80.3 65.1

76.0 76.7 67.4

79.4 80.0 57.1

54.8 56.6 51.4

75.2 76.0 65.0

AS (00) AS-C (00)

85.7 86.5

69.6 70.3

71.3 72.9

67.7 69.2

59.9 62.0

70.8 72.2

Notes: *AS represents ASEAN cumulative local content. **AS-C represents ASEAN-China cumulative local content. Source: Asian International Input–Output Tables (1990, 2000).

98 The Making of Economic Integration in East Asia

than local content in values ranging from 1.5 per cent in Indonesia to 10.6 per cent in Malaysia. On average, however, cumulative local content decreased in this period due to rising dependency on the outer world.15 4.4.2.3

Value added-based local content

Table 4.6 shows value added-based local content of the manufacturing sector in 2000. Value added-based local content was considerably lower than direct input-based local content (Table 4.4), and the differences between the two varied from 6.7 per cent in Indonesia to 14.9 per cent in Korea. Taiwan and Singapore also had a difference of more than 14 per cent. This implies that the direct input-based local content in these three economies included overestimates of more than 14 per cent relative to real local content.16 When direct input-based local content is used, the real import content from ASEAN countries seems overestimated, while that from Japan and the US appears underestimated. It can thus be seen that direct inputbased local content tends to include an overestimate of real local content when countries do not have well-developed local supporting industries, and vice versa. Figures 4.2(a)–(e) correspond to Figure 4.1(a)–(d), except for Figure 4.2(c) that demonstrates the impact of changes in value added coefficients (see Equation 24). There are many similarities between the numbers for direct input and value added-based local content, but there are also notable differences. First, the impact of value-added coefficient changes on import content was very small in all economies (see Figure 4.2(c)). Second, the Table 4.6 Local and import content (2000): manufacturing sector (value addedbased)

MAN DOM LC ASEAN NEA JAP USA ROW IC

IND

PHI

THA

MAL

SIN

CHN

KOR

TWN

45.0 32.4 77.5 1.3 2.1 2.3 1.8 15.0 22.5

36.6 20.2 56.8 3.4 4.5 7.6 5.7 22.1 43.2

36.3 20.6 56.9 3.1 4.2 7.7 3.8 24.3 43.1

28.8 14.8 43.6 7.3 6.3 10.5 8.3 24.0 56.4

26.0 11.3 37.3 6.5 5.2 10.4 7.0 33.5 62.7

48.8 30.4 79.3 1.0 2.5 2.6 1.4 13.1 20.7

44.7 19.4 64.0 2.4 2.4 5.3 4.6 21.3 36.0

34.8 21.9 56.7 3.7 3.6 9.5 5.5 21.1 43.3

Source: Asian International Input–Output Table (2000).

AVE 37.6 21.4 59.0 3.6 3.8 7.0 4.8 21.8 41.0

99 5 0 5

IND

PHI

THA

MAL

SIN

CHN

KOR

TWN

AVE

10 15 20 25 C-LC

VA

TRAD

TECH

*C-LC  VA  TRAD  TECH

Figure 4.2(a)

Changes in local content (1990–2000): manufacturing sector*

25 20 15 10 5 0

IND

PHI

THA

MAL

SIN

CHN

KOR

TWN

AVE

5 C-IC

ASEAN

NEA

JAP

USA

ROW

*C-IC  ASEAN  NEA  JAP  USA  ROW  VA (Figure 2 (c))  TRAD (Figure 2(d))  TECH (Figure 2(e))  ROW For ASEAN, NEA, JAP, and USA respectively, it holds that Figure 4.2(b)  Figure 4.2(c)  Figure 4.2 (d)  Figure 4.2 (e)

Figure 4.2(b)

Changes in import content (1990–2000): manufacturing sector*

25 20 15 10 5 0 5

IND

PHI

THA VA

MAL ASEAN

SIN NEA

CHN

KOR

JAP

USA

TWN

AVE

*VA  ASEAN  NEA  JAP  USA

Figure 4.2(c) Impact of value added coefficient changes (1990–2000): manufacturing sector*

100 The Making of Economic Integration in East Asia 25 20 15 10 5 0 5

IND

PHI

THA TRAD

MAL

SIN

ASEAN

CHN NEA

KOR JAP

TWN

AVE

USA

* TRAD  ASEAN  NEA  JAP  USA

Figure 4.2(d) Impact of trade structural changes (1990–2000): manufacturing sector*

25 20 15 10 5 0 5

IND

PHI

THA

MAL

TECH

SIN ASEAN

CHN NEA

KOR JAP

TWN

AVE

USA

* TECH  ASEAN  NEA  JAP  USA

Figure 4.2(e) Impact of technological structural changes (1990–2000): manufacturing sector* Source: Asian International Input–Output Tables (1990, 2000).

diversion of import shares away from Japan and the US to East Asia, especially to ASEAN countries, became less clear when value added-based local content was used (see Figure 4.2(d)). Table 4.7 shows value added-based ASEAN and ASEAN-China cumulative local content. Value added-based content was considerably lower than direct input-based content in ASEAN countries. In particular, these differences were extremely large in Malaysia and Singapore, and they also increased over time. For example, in Malaysia, the difference between direct input and value added-based ASEAN cumulative local content increased from 12.3 per cent to 16.8 per cent.17

Rules of Origin, Local Content and Cumulative Local Content 101 Table 4.7 ASEAN and ASEAN-China cumulative local content: manufacturing sector (value added-based) IND

PHI

THA

MAL

SIN

AVE

LC (90)

80.1

70.1

63.8

64.0

35.7

62.8

AS (90) AS-C (90) LC (00)

80.9 81.4 77.5

71.9 72.2 56.8

66.1 67.1 56.9

67.1 67.9 43.6

42.7 44.7 37.3

65.8 66.7 54.4

AS (00) AS-C (00)

78.7 79.7

60.2 61.2

60.1 62.0

50.9 52.9

43.9 46.3

58.7 60.4

Source: Asian International Input–Output Tables (1990, 2000).

4.5

Sector analysis

As shown in Table 4.3, the electronics and automotive sectors demonstrated respectively the largest decline and the largest rise in local content between 1990 and 2000. This section includes focus on these two sectors as well as an exploration of the causes of these changes. However, due to limited space, only the analysis of direct input-based local content is viewed. 4.5.1 4.5.1.1

Electronics sector Local content

Table 4.8 shows that in 2000, local content of the electronics sector in the Philippines, Malaysia and Thailand was below 40 per cent. They thus had extremely high dependency on imported intermediate inputs. In addition to strong dependency on Japan and the US, the import content from both ASEAN and Northeast Asia was very high in the Philippines (8.3 per cent, 8.6 per cent), Thailand (11.6 per cent, 10.1 per cent), Malaysia (17.2 per cent, 9.3 per cent) and Singapore (10.8 per cent, 6.5 per cent). The Asian table indicates that these four countries had strong linkages with each other. For example, the electronics sectors in Singapore and Malaysia imported respectively 8.2 per cent and 10.7 per cent of their total inputs from one another. As shown below, in addition to linkages with neighbouring countries, linkages between ASEAN and Northeast Asia were strengthened. Consequently, the procurement of intermediate inputs in the electronics sector became rather diversified. Figures 4.3(a) and (b) show that all countries, except for Indonesia and Singapore, had increased import content. Import content from ASEAN and Northeast Asia increased considerably, while that from Japan and the US decreased in countries such as Thailand, Singapore and Taiwan. Similar tendencies were observed in other East Asian countries.

102 The Making of Economic Integration in East Asia Table 4.8 Local and import content (2000): electronics sector IND

PHI

THA

MAL

SIN

CHN

KOR

TWN

AVE

ELE DOM VA LC

3.1 47.0 30.5 80.6

1.6 8.7 24.2 34.5

6.3 16.4 16.4 39.1

5.6 14.8 16.1 36.6

15.4 15.0 19.8 50.1

27.6 24.5 23.2 75.3

16.0 21.0 26.4 63.4

13.6 24.6 19.8 58.0

11.1 21.5 22.0 54.7

ASEAN NEA JAP USA ROW IC

1.8 2.4 2.8 1.4 11.0 19.4

8.3 8.6 16.8 12.7 19.2 65.5

11.6 10.1 11.7 8.6 18.8 60.9

17.2 9.3 10.7 11.4 14.8 63.4

10.8 6.5 11.7 6.8 14.1 49.9

3.0 5.6 2.6 1.8 11.6 24.7

4.5 3.7 7.9 8.3 12.2 36.6

8.3 5.7 11.7 5.7 10.6 42.0

8.2 6.5 9.5 7.1 14.0 45.3

Source: Asian International Input–Output Table (2000).

20 15 10 5 0 5 10 15 20 25

SIN IND PHI

THA

C-LC

Figure 4.3(a)

CHN

MAL

VA

TRAD

KOR

TWN

AVE

TECH

Changes in local content (1990–2000): electronics sector

The impact of trade structural changes is more clearly demonstrated in Figure 4.3(c). Here, the impact of technical changes was removed from import content changes. Figure 4.3(c) shows the diversion of import shares moving away from Japan and the US to East Asia. Note that inputs from Japan declined in all countries due to changes in trade structure. Further, the electronics sector in the Philippines, Thailand and Malaysia strengthened linkages with Northeast Asia to a greater extent than with neighbouring Southeast Asian countries (vice versa in Korea and Taiwan). These facts suggest that the procurement of intermediate inputs in the electronics sector spread geographically throughout East Asia over the period 1990–2000. Figure 4.3(d) shows that in some ASEAN countries, technological changes had a strong effect on the import content of the electronics industry.

103 25 20 15 10 5 0 5 10 15 20

IND

PHI

THA

C-IC

Figure 4.3(b)

25 20 15 10 5 0 5 10 15 20

IND

MAL

ASEAN

SIN

NEA

CHN

JAP

KOR

USA

TWN

AVE

ROW

Changes in import content (1990–2000): electronics sector

PHI

MAL

SIN

CHN

KOR

TWN

AVE

THA

TRAD

ASEAN

NEA

JAP

USA

Figure 4.3(c) Impact of trade structural changes (1990–2000): electronics sector

25 20 15 10 5 0 5 10 15 20

IND

PHI

THA

TECH

MAL

SIN

ASEAN

CHN

NEA

JAP

KOR

TWN

AVE

USA

Figure 4.3(d) Impact of technological structural changes (1990–2000): electronics sector Source: Asian International Input–Output Tables (1990, 2000).

104 The Making of Economic Integration in East Asia

4.5.1.2

ASEAN and ASEAN-China cumulative local contents

Table 4.9 provides information on the ASEAN and ASEAN-China cumulative local content of the electronics sector. Although local content of the electronics sector in the Philippines, Thailand and Malaysia declined over the period 1990–2000, ASEAN and ASEAN-China cumulative local content in 2000 was considerably greater than local content in these countries due to rising import content from ASEAN neighbours and China; in Malaysia, for example, the difference between local and ASEAN cumulative local content increased from 8.4 per cent to 17.2 per cent. However, ASEAN cumulative local content in these countries, which is the sum of local content and inputs from neighbouring ASEAN countries, declined due to rising dependency on the outer world. ASEAN-China cumulative local content increased only in Thailand.18

4.5.2

Automotive sector

4.5.2.1

Local content

Table 4.10 shows that local content in the automotive sector was considerably larger than local content in the electronics sector. For example, local content of automotive industries in Korea, China, Indonesia and Taiwan exceeded 80–90 per cent in 2000. Unlike the electronics sector, import content from ASEAN and Northeast Asia was relatively small and that of Japan was predominantly large, as seen, for example, in Thailand (17.5 per cent). Figure 4.4(a) shows that with the exception of the Philippines, East Asian countries increased local content in the automotive sector. The Malaysian automotive sector, in particular, increased local content by 11.7 per cent. The value added coefficient reduced local content, while technological and trade structural changes raised it in many countries.

Table 4.9 ASEAN and ASEAN-China cumulative local content: electronics sector IND

PHI

THA

MAL

SIN

AVE

LC (90)

66.1

52.2

40.5

56.2

47.3

52.5

AS (90) AS-C (90) LC (00)

69.2 69.6 80.6

58.4 58.4 34.5

51.2 51.4 39.1

64.6 64.9 36.6

55.5 55.8 50.1

59.8 60.0 48.2

AS (00) AS-C (00)

82.4 83.2

42.7 43.3

50.8 54.5

53.8 55.7

61.0 63.3

58.1 60.0

Source: Asian International Input–Output Tables (1990, 2000).

Rules of Origin, Local Content and Cumulative Local Content 105 Table 4.10

Local and import content (2000): automotive sector IND

PHI

THA

MAL

SIN

CHN

KOR

TWN

AVE

AUTO DOM VA LC

11.3 29.0 43.4 83.7

2.4 41.3 22.6 66.4

8.4 24.9 26.5 59.8

16.7 15.9 35.9 68.5

2.9 29.7 26.0 58.7

36.4 30.8 23.8 91.0

32.1 36.8 23.4 92.3

22.9 23.5 34.0 80.4

16.6 29.0 29.5 75.1

ASEAN NEA JAP USA ROW IC

0.8 0.9 5.6 0.8 8.3 16.3

5.1 6.1 8.4 1.9 12.2 33.6

2.2 2.2 17.5 1.7 16.7 40.2

3.7 3.2 14.1 2.2 8.3 31.5

4.9 3.4 9.6 4.1 19.3 41.3

0.1 0.9 2.1 0.4 5.5 9.0

0.2 0.5 2.4 1.3 3.4 7.7

0.7 1.2 7.0 1.5 9.3 19.6

2.2 2.3 8.3 1.7 10.4 24.9

Source: Asian International Input–Output Table (2000).

20 15 10 5 0 5

SIN

IND

10

PHI

THA

MAL

AVE KOR

15 C-LC

Figure 4.4(a)

TWN CHN

VA

TRAD

TECH

Changes in local content (1990–2000): automotive sector

In Figure 4.4(b), it is notable that except for Taiwan, inputs from Japan declined. They fell more than 10 per cent in the Philippines, Malaysia and Singapore. On the other hand, inputs from the US did not demonstrate such a clear tendency. In fact, they increased slightly in several countries. Thus, although import content from Japan was still dominant, as shown in Table 4.10, there was a substantial shift in import content away from Japan (and the US, if any, to a lesser extent) to East Asia in the period 1990–2000. Further, unlike the electronics sector, the automotive sector in ASEAN increased inputs from neighbouring ASEAN countries more than from Northeast Asia (vice versa in Northeast Asia). Figure 4.4(c) shows clearly the diversion of import shares away from Japan to East Asia. Diversion from the US was more ambiguous. Figure 4.4(d) shows that the impact of technological change was rather large in some ASEAN countries.

106 15 10 5 0 CHN

5 10

IND

KOR

TWN AVE

THA SIN

15

PHI

MAL

20 C-IC

Figure 4.4(b)

ASEAN

NEA

JAP

USA

ROW

Changes in import content (1990–2000): automotive sector

15 10 5 0 5

CHN

10 15

PHI

THA

MAL

KOR

TWN

AVE

SIN

20 TRAD

ASEAN

NEA

JAP

USA

Figure 4.4(c) Impact of trade structural changes (1990–2000): automotive sector

15 10 5 0 5 10

IND

PHI

THA

MAL

SIN

CHN

KOR

TWN

AVE

15 20 TECH

Figure 4.4(d) sector

ASEAN

NEA

JAP

USA

Impact of technological structural changes (1990–2000): automotive

Source: Asian International Input–Output Tables (1990, 2000).

Rules of Origin, Local Content and Cumulative Local Content 107 Table 4.11 sector

ASEAN and ASEAN-China cumulative local content: automotive

IND

PHI

THA

MAL

SIN

AVE

LC(90)

77.8

66.6

53.7

56.8

53.2

61.6

A(90) A-C(90) LC(00)

78.3 78.4 83.7

67.2 67.2 66.4

55.0 55.7 59.8

57.5 57.6 68.5

57.1 57.4 58.7

63.0 63.3 67.4

A(00) A-C(00)

84.5 84.8

71.5 72.9

61.9 62.6

72.2 73.1

63.5 65.4

70.7 71.8

Source: Asian International Input–Output Tables (1990, 2000).

4.5.2.2

ASEAN and ASEAN-China cumulative local content

Table 4.11 reflects the input structure of the automotive sector. Inputs from neighbouring ASEAN countries and China appear relatively small. Thus, ASEAN and ASEAN-China content was not significantly greater than local content, although the difference between them increased over the period 1990–2000. Unlike the electronics sector, ASEAN and ASEAN-China cumulative local content in the automotive sector increased in all countries due to rising local and import content from neighbouring countries.19 4.5.3

Factors affecting local content

As seen above, there was a striking contrast in trends of local content in the electronics and automotive sectors. Regarding factors that affect local content, Baba (2005) pointed out that (1) characteristics of parts and components (especially transport cost and architecture), and (2) industrial policies in East Asia (whether import substitution or exportoriented) are important determinants of the development of local supporting industry. Using his analytical framework, the above phenomena may be explained as follows. 4.5.3.1

Characteristics of parts and components

Transport costs for parts and components are important factors in the determination of local content. For example, parts and components in the automotive industry (body parts, engine parts, brakes, suspension assemblies, etc.) are much more bulky and are heavier than those in the electronics industry (semiconductors, magnetic heads, PCBs, condensers, etc.). Therefore, automotive assemblers have stronger incentives to save on transport costs by procuring their parts locally.20 In fact, the

108 The Making of Economic Integration in East Asia

study shows that parts suppliers of the Japanese automotive industry, the dominant automotive producer in Southeast Asia, tend to locate near their assemblers. This is because such geographical proximity not only lowers the transport and inventory costs (or facilitates ‘just-in time deliveries’ as in case of Toyota) but also facilitates product development coordination between parts suppliers and assemblers (Dyer, 1994, 1996). In addition to relatively low transport costs, modular architecture has become common in the electronics industry (PCs and mobile phones for example). Since modular architecture has led to standardized specifications (especially interfaces between modules or components), parts and components can be readily substituted for imports. On the other hand, integral architecture is more common in the automotive industry where parts and components are not necessarily standardized (Fujimoto, 2004). Thus, as Dyer pointed out, geographical proximity is important for facilitating the development of specific parts and components for the automotive assemblers. 4.5.3.2

Industrial policy in East Asia

The automotive industry has been strategically important due to strong linkages with supporting industries. In fact, until recent years, many East Asian countries heavily protected automotive parts suppliers as well as assemblers. For example, the Thai government introduced the local content requirement policy in 1971, and until it was abolished in 2000, it made a valuable contribution to raising the local content of the Thai automotive industry (Terdudomtham, 2000). Further, because there was a lack of competition in the domestic market, the automotive industry did not need to import parts and components to increase competitiveness. Brand-to-Brand Complementation (BBC) and ASEAN Industrial Cooperation (AICO) schemes, which were initiated respectively in 1988 and 1996, reduced trade barriers of automotive parts and components within ASEAN. They thus seem to have contributed to increasing import content of the automotive industry from neighbouring ASEAN countries (see Figure 4.4(b)). Export-oriented industries in East Asia have been given privileges over other industries. These have included unlimited access to imported intermediate inputs and exemption of import duty. Thus, the electronics industry, which was one of the most successful export-oriented industries in East Asia, did not have much incentive to raise the share of local procurement. Further, due to fierce competition in the international market, they needed to import electronics parts in order to strengthen competitiveness. These factors appear to have promoted

Rules of Origin, Local Content and Cumulative Local Content 109

diversification of the procurement of parts and components in the electronics industry.

4.6

Conclusion

After reviewing the RoO in AFTA and the ASEAN-China FTA, various kinds of local content were measured using the Asian international input–output tables. Major findings in this study can be summarized as follows: (1) The manufacturing sector had relatively low local content vis-a`-vis primary industry and non-tradable goods and services. Among all industries, the electronics and automotive industries had respectively the largest decline and the largest rise in local content in the period 1990–2000. (2) In the manufacturing sector, all countries other than Singapore had reduced local content. This was primarily due to trade structural changes. Regarding import content, although Japan and the US remained key suppliers of intermediate inputs, diversion of import shares from Japan and the US to East Asian countries was observed in the manufacturing sector. Similar tendencies (especially diversion from Japan) were more clearly demonstrated in the electronics and automotive sectors. (3) Value added-based local content of the manufacturing sector was 6.7 per cent to 14.9 per cent lower than direct input-based local content. In other words, direct input-based local content, which was adopted in the RoO in FTAs, led to a (significant) overestimate of the real local content. Similarly, the real import content from ASEAN countries was also generally overestimated, while direct input-based import content led to an underestimate of the real dependency on Japan and the US. (4) Due to a decline in local content in the 1990s, local content of the electronics sectors in 2000 was below 40 per cent in the Philippines, Thailand and Malaysia, although their ASEAN and ASEAN-China cumulative local content exceeded this threshold. At the same time, the procurement of electronics parts and components was diversified. They thus had fairly large dependency on East Asian countries as well as on Japan and the US. (5) Due to a rise in local content in the 1990s, the local content of the automotive sector in Korea, China, Indonesia, and Taiwan exceeded 80–90 per cent in 2000. In addition, dependency on Japan was quite large, although this dependency had fallen sharply in the 1990s. The

110 The Making of Economic Integration in East Asia

procurement of parts and components in the automotive sector was geographically confined due to higher transport costs, etc. (6) The trend of local content over the period 1990–2000 in the electronics and automotive sectors can be explained by looking at the characteristics of parts and components in each industry as well as industrial policies in East Asia. BBC and the AICO schemes in ASEAN also seem to have helped increase procurement of automotive parts and components from neighbouring ASEAN countries. (7) Use of the cumulation rule in AFTA and the ASEAN-China FTA tends to increase local content of industry when that industry has large import content from other FTA member countries. For example, the cumulation rule increased local content of the electronics industry more significantly than local content of the automotive industry. Due to rising dependency on neighbouring ASEAN countries and China, the contribution of the cumulation rule increased in the 1990s. This trend is further expected to accelerate as trade barriers in East Asia continue to be removed due to the implementation of FTAs and other trade facilitation arrangements. Technical note Decomposition analysis of local content In decomposition analysis, using input–output tables for two points of time, factors which have affected output, value added, balance of payments, etc., can be found for each industry. Decomposition analysis was applied in this study. International input–output tables were used to investigate the factors of change in local content. 1.

Direct input-based decomposition analysis

In decomposition analysis of direct input-based local content, changes in local content are decomposed into three factors: (1) a change in value added coefficients, (2) a change in trade structure, and (3) a change in technology. The first factor is found by the following simple formula:

Dv ¼ vtþ1  vt

ð11Þ

where vt is a value added coefficient at time t. The sum of the second and third factors is found by a change in input coefficient matrices between two points in time:

DA ¼ Atþ1  At

ð12Þ

RS where A ¼ [aRS ij ], and a ij denotes the amount of Commodity i produced in Country R and needed by Country S per each unit output of Commodity j. Next, using data

Rules of Origin, Local Content and Cumulative Local Content 111 obtainable from the international input–output tables, an input coefficient aRS ij may be expressed as a product of the following two coefficients:21 RS

˙ ~S aRS ij ¼ t ij aij

ð13Þ

where a˜ijS represents a ‘technological input coefficient’ denoting the sum of domestic and imported Commodity i needed by Country S per unit output of Commodity j. A ‘trade coefficient’ t˙ RS ij reflects which fraction of intermediate demand for Commodity i exercised by Sector j in Country S is satisfied by Country R (import share of Country R). Alternatively, Equation (13) may be expressed in matrix form:

~ A ¼ T_ A

ð14Þ

˜ ˜ijS ], and stands for the Hadamard product (cell by cell where T_ ¼ [t˙ RS ij ], A ¼ [a multiplication). Substituting Equation (14) into (12) yields the following:

~ tþ1  T_ t A ~t DA ¼ T_ tþ1 A ~ tþ1  T_ t A ~ tþ1 þ T_ t A ~ tþ1  T_ t A ~t ¼ T_ tþ1 A ~ tþ1 þ T_ t D A ~ ¼ D T_ A

ð15Þ

Alternatively, Equation (12) may be written as

~ þ D T_ A ~t DA ¼ T_ tþ1 DA

ð16Þ

Averaging Equations (15) and (16) gives

1 _ 1 ~ ~ ~ _ DA ¼ D T_ ðA tþ1 þ At Þ þ ðTtþ1 þ Tt Þ DA 2 2

ð17Þ

where the first and second terms show respectively an impact of the above mentioned Factor 2, a change in trade structure (change in trade coefficients or import _ and Factor 3, a change in technology (change in technological input shares DT), ˜ ). coefficients DA 2.

Value added-based decomposition analysis

Decomposition of value added-based local content is more complicated, but a similar technique can be employed. A change in an element in Matrix G in Equation (8) can be expressed as follows:

DG ¼ Gtþ1  Gt ^ tþ1 Ltþ1  V ^ t Lt ¼V 1 ^ 1 ^ ^ ¼ D VðL tþ1 þ Lt Þ þ ðVtþ1 þ Vt ÞD L 2 2

ð18Þ

112 The Making of Economic Integration in East Asia ^ t is a diagonal value added coefficient matrix at time t, Lt ¼ (I  At)1, DV ^¼ where V ^ ^ Vtþ1  Vt and DL ¼ Ltþ1  Lt. D L can be rewritten as

DL ¼ Ltþ1  Lt ¼ Ltþ1 ðI  At ÞLt  Ltþ1 ðI  Atþ1 ÞLt

ð19Þ

Substituting Equation (14) into (19) gives

~ t ÞLt  Ltþ1 ðI  T_ tþ1 A ~ tþ1 ÞLt DL ¼ Ltþ1 ðI  T_ t A ~ t ¼ Ltþ1 ½D ðT_ AÞL

ð20Þ

Alternatively, Equation (20) may be rewritten as

~ tþ1 : DL ¼ Lt ½D ðT_ AÞL

ð21Þ

Then, from Equations (17) and (20),

DL ¼

1 ~ tþ1 þ A ~ t ÞLt þ Ltþ1 ½ðT_ tþ1 þ T_ t Þ D AL ~ t g: fLtþ1 ½D T_ ðA 2

ð22Þ

Similarly, from Equations (17) and (21),

DL ¼

1 ~ tþ1 þ A ~ t ÞLtþ1 þ Lt ½ðT_ tþ1 þ T_ t Þ D AL ~ tþ1 g: fLt ½D T_ ðA 2

ð23Þ

From Equations (18), (22), and (23), it then follows that

1 ^ DV ðLtþ1 þ Lt Þ 2 1 ^ ~ ~ ~ ~ ^ _ _ þ ðV tþ1 þ Vt ÞfLtþ1 ½D T ðAtþ1 þ At ÞLt þ Lt ½D T ðAtþ1 þ At ÞLtþ1 g 8 1 ^ ~ ~ ^ _ _ _ _ þ ðV tþ1 þ Vt ÞfLtþ1 ½ðTtþ1 þ Tt Þ D ALt þ Lt ½ðTtþ1 þ Tt Þ D ALtþ1 g 8 ð24Þ

DG ¼

In Equation (24), the first, second, and third terms show respectively the impact on local contents of the earlier mentioned factors: (1) a change in value added ^ (2) a change in trade structure (change in trade coefficients or coefficients DV, _ and (3) a change in technology (change in technological import shares DT), ˜ ). input coefficients DA

Rules of Origin, Local Content and Cumulative Local Content 113

Notes 1. An earlier version of this chapter was distributed as Institute of Developing Economies Discussion Paper No. 78. 2. In bilateral cumulation, only use of components from the (bilateral) partner country is favoured. In both diagonal and full cumulation, all beneficiary trading partners of the cumulation area are favoured. However, full cumulation is more liberal than diagonal cumulation, since the originating status in the FTA partner countries is not necessary and hence allows a greater use of third-country materials (Estevadeordal and Suominen, 2003). 3. However, the criterion of 40 per cent local content is difficult to satisfy for some products. Thus other criteria, especially CTC, were later introduced into aluminiums, wood-based, wheat flour, and other such products in AFTA. Also, in the China-ASEAN FTA, product-specific rules are scheduled for introduction into those products that may not satisfy the above criterion (ASEAN Secretariat, 2005). 4. In assessing East Asian FTAs, AFTA and the ASEAN-China FTA may have high scores in market access with simple and uniform RoO. However, some reservation is necessary relative to the quality of their market access. This is due to the fact they took recourse to the Enabling Clause; this allows developing countries to establish FTAs without satisfying the requirements set in GATT Article 24 (Cheong and Kwon, 2005). 5. In addition to primary products, Condition (1) is also applicable to waste, scraps, and used articles fit only for the recovery of raw materials (ASEAN Secretariat, 2005). 6. For example, consider the case where the material imported from other ASEAN member countries has 30 per cent ASEAN cumulative local content. In this case, 0 per cent, 30 per cent, and 100 per cent of the imported material price will be added to the calculation of ASEAN cumulative local content of finished products in diagonal, partial, and full cumulation respectively. 7. The original formula for this criterion is as follows (ASEAN Secretariat, 2005): [(Value of imported non-ASEAN material, parts, or produce) þ (Value of undetermined origin materials, parts, or produce)] 7 FOB price  100%  60%. For this chapter, original expressions were changed into more usable economic expressions. 8. Matsumura and Fujikawa (1998) called the local content in Formula 5 ‘direct technology criterion-based local content’. A different terminology is used in this study. 9. Although accrued errors are not significantly large,Pfreight P and insurance on imported intermediate inputs are not included in R"AS ni¼1 XRS ij in Formula 6, due to the format of the Asian international input–output tables. 10. It is notable that feedback effects among the ten endogenous countries are taken into account when the Asian international input–output models (rather than single-country models) are employed (see Kuroiwa, 1995). 11. The original Asian international input–output tables for 1990 and 2000 had respectively 78 and 76 sector classifications. 12. Singapore is in a unique position in this study. Singapore used to have much lower local content than other countries until 1990. Then it became the only

114 The Making of Economic Integration in East Asia country which increased local content (i.e. decreased import content) in many sectors from 1990–2000 (see Table 4.2). 13. Although there is no convincing evidence derivable from this study, it is generally accepted that the technological structure as reflected by input coefficients becomes stable, as the industry matures. 14. AFTA was initiated in 1992, and the reduction of tariff rates started in 1993. Consequently, almost all products in the Common Effective Preferential Tariff (CEPT) Scheme Inclusion List (IL) in the six ASEAN countries (Indonesia, the Philippines, Thailand, Malaysia, Singapore and Brunei) were reduced to the 0–5 per cent tariff range in 2003. On the other hand, agreement was reached on the ASEAN-China FTA in 2002, and the early harvest package was initiated in 2004. 15. As an exercise, the ASEAN-Korea, ASEAN-Japan, and ASEAN-China-KoreaJapan (ASEANþ3) content of the manufacturing sector was calculated using a similar methodology. The table below shows that the ASEAN-Korea content increased local content to almost the same extent as did ASEAN-China content. On the other hand, ASEAN-Japan content was considerably greater than that of the other two, reflecting ASEAN’s greater dependency on Japan. It is notable that the contribution of the cumulation rule increased over the period 1990–2000. For example, the difference between ASEANþ3 and local content in Malaysia increased from 10.1 per cent to 21.8 per cent. ASEAN-Korea, ASEAN-Japan, and ASEANþ3 cumulative local content: manufacturing sector

ASK (90)* ASJ (90)** ASþ3 (90)*** ASK (00) ASJ (00) ASþ3 (00)

IND

PHI

THA

MAL

SIN

AVE

86.3 88.4 89.2 86.4 87.4 89.0

80.7 82.6 83.5 71.9 76.0 79.1

76.7 81.8 83.3 72.5 77.3 80.1

79.9 84.2 85.3 69.7 75.4 78.9

56.1 66.3 69.4 61.5 68.3 72.0

76.0 80.7 82.1 72.4 76.9 79.8

Notes: * AS-K represents ASEAN-Korea cumulative local content. ** AS-J represents ASEAN-Japan cumulative local content. *** ASþ3 represents ASEAN-China-Korea-Japan cumulative local content. Source: Asian International Input–Output Tables (1990, 2000).

16. Theoretically, the difference between direct input-based local content and value added-based local content tends to be larger, as the economy becomes more dependent on imported inputs, especially in upstream industry (inclusive of raw material industry). Since it can be prohibitively costly for each exporter to estimate value added-based local content of its exported products, it is simply not feasible to work out an alternative to the current procedure. It is, however, important to note that the errors can be significantly large in some sectors or economies. 17. Regarding ASEAN-Korea, ASEAN-Japan and ASEANþ3 content, it is notable that the differences between direct input-based and value added-based

Rules of Origin, Local Content and Cumulative Local Content 115 cumulative local content became smaller when Japan was incorporated into the scheme. ASEAN-Korea, ASEAN-Japan, and ASEANþ3 cumulative local content: manufacturing sector (value added-based)

ASK (90) ASJ (90) ASþ3 (90) ASK (00) ASJ (00) ASþ3 (00)

IND

PHI

THA

MAL

SIN

AVE

81.3 84.0 84.9 79.5 81.0 82.8

72.6 75.5 76.6 62.2 67.7 70.8

66.9 73.5 75.2 61.2 67.8 70.8

67.7 74.3 75.8 52.9 61.4 65.4

43.9 55.7 58.9 45.4 54.3 58.2

66.5 72.6 74.3 60.2 66.4 69.6

Source: Asian International Input–Output Tables (1990, 2000).

18. The table below shows the cumulative local content of the electronics sector in ASEAN-Korea, ASEAN-Japan and ASEAN þ 3 arrangements. As in the manufacturing sector, inclusion of Japan in such an arrangement appears to increase the cumulative content drastically. However, its impact declined over this period; that of the inclusion of Korea and China grew by a sizeable magnitude. ASEAN-Korea, ASEAN-Japan, and ASEANþ3 cumulative local content: electronics sector

ASK (90) ASJ (90) ASþ3 (90) ASK (00) ASJ (00) ASþ3 (00)

IND

PHI

THA

MAL

SIN

AVE

70.5 77.3 79.0 83.4 85.2 87.0

59.4 75.7 76.7 48.0 59.5 65.3

52.7 66.5 68.3 54.4 62.5 69.9

65.7 73.9 75.3 56.8 64.5 69.4

57.9 74.3 77.0 63.1 72.7 77.2

61.2 73.6 75.2 61.2 68.9 73.8

Source: Asian International Input–Output Tables (1990, 2000).

19. Compared with the electronics sector, the impact of the inclusion of Japan is quite large in the automotive sector. For example, in 1990, the difference between the ASEAN-Japan and the ASEAN cumulative local content in Malaysia was as large as 26.5 per cent; that for ASEAN-China content was only 0.1 per cent. However, as seen in Figure 4.4(b), import content from Japan declined over this time, and so did the impact of the inclusion of Japan in the cumulative content. Import content from China and Korea increased only slightly in this period. 20. Transport costs also affect the procurement of parts and components within the electronics sector. Letchumanan and Kodama (2000) demonstrated that weights of parts in the electronics industry are positively correlated with

116 The Making of Economic Integration in East Asia the percentage of local procurement; the heavier the parts, the higher the percentage of local procurement. ASEAN-Korea, ASEAN-Japan, and ASEANþ3 cumulative local content: automotive sector

ASK (90) ASJ (90) ASþ3 (90) ASK (00) ASJ (00) ASþ3 (00)

IND

PHI

THA

MAL

SIN

AVE

78.7 90.1 90.8 84.8 90.1 90.7

69.1 89.3 91.2 74.2 79.8 84.1

55.5 75.7 77.0 62.6 79.4 80.8

57.6 84.0 84.2 73.3 86.3 88.3

57.3 77.6 78.2 64.5 73.1 76.0

63.6 83.3 84.3 71.9 81.7 84.0

Source: Asian International Input–Output Tables (1990, 2000).

21. See Hoen (2002: 138) for the technique of decomposition analysis, especially that of international input–output. Hoen provides an illuminating example of this technique. Although he employed the technique for a different purpose, some of the techniques used in this study were also used in his.

References ASEAN Secretariat (2005). ‘Rules of Origin for the CEPT Scheme for AFTA’, http:// www.aseansec.org Baba, T. (2005). Supporting Industry in Asia (in Japanese Ajia no Susono Sangyo), Tokyo: Hakuto-Shobo. Breton, P. (2003). ‘Notes on Rules of Origin with Implications for Regional Integration in Southeast Asia’, paper presented at the seminar ‘Regional Trade Agreements in Comparative Perspective: Latin America and the Caribbean and Asia Pacific’, PECC Trade Forum, 22–23 April, Washington, D.C. Cheong, I. and K. D. Kwon (2005). ‘Assessing the Quality of FTAs and Implications for East Asia’, paper presented at the seminar ‘Building an Asia-Pacific Economic Community’, APEC Study Center Consortium Conference 2005, 22–25 May, Jeju, Korea. Dyer, J. H. (1994). ‘Dedicated Assets: Japan’s Manufacturing Edge’, Harvard Business Review, November–December: 174–8. Dyer, J. H. (1996). ‘How Chrysler Created an American Keiretsu’, Harvard Business Review, July–August: 42–56. Estevadeordal, A. and K. Suominen (2003). ‘Rules of Origin in FTAs: a World Map’, paper presented at the seminar ‘Regional Trade Agreements in Comparative Perspective: Latin America and the Caribbean and Asia Pacific’, PECC Trade Forum, 22–23 April, Washington, D.C. Fujimoto, T. (2004). Philosophy of Manufacturing in Japan (in Japanese Nihon no Monozukuri Tetsugaku), Tokyo: Nihon Keizai Shinbunsha. Hoen, A. R. (2002). An Input–Output Analysis of European Integration, Amsterdam: North Holland.

Rules of Origin, Local Content and Cumulative Local Content 117 Kuroiwa, I. (1995). ‘The Spatial Structure of the Asia-Pacific Economy’, PhD dissertation, University of Pennsylvania. Letchumanan, R. and F. Kodama (2000). ‘Reconciling the Conflict between the ‘‘Pollution-Heaven’’ Hypothesis and an Emerging Trajectory of International Technology Transfer’, Research Policy, 29: 59–79. Matsumura, F. and K. Fujikawa (1998). Economic Analysis of Home Production (in Japanese Kokusanka no Keizai Bunseki), Tokyo: Iwanami Shoten. Terdudomtham, T. (2000). ‘Thai Policies for the Automobile Sector: Focus on Technology Transfer’, in R. Busser and T. Sodoi (eds), Production Networks in Asia and Europe: Skill Formation and Technology Transfer in the Automobile Industry, London: Routledge Curzon.

5 Making Sense of the Timeliness of Transportation in Economic Integration Nobuaki Hamaguchi

5.1

Introduction

International trade has grown dramatically since the second half of the 1980s. The trade value in 2005 was 5.5 times larger than in 1985, growing on average by 8.5 per cent per year during this period. The growth of intra-Asia trade was even more dramatic. During the same period, it grew 14.3 per cent annually and the share in the world trade has increased from 2.0 per cent to 6.5 per cent, drawing level with intra-NAFTA trade, while the intra-EU trade share has tended to decline in the long run (Figure 5.1). Why has there been so much increase in trade? Krugman (1995) offered two possible explanations, i.e. tariff reduction and transport cost.1 He suggested that both these factors may have contributed to new aspects of the growing trade such as intra-industry trade of differentiated goods, ‘slicing-up the value chain’, and the growth of the lowwage manufacturing exporters, for example, the case of the Asian newly industrialized economies (NIEs). In his argument, though, international economics gives credit to the liberalization of tariff policies as the primary cause for trade expansion, while emphasis on the transport cost reduction remains a rather journalistic point of view and is yet to be theoretically modelled and empirically tested. Although admittedly international economics did not pay much attention to the effects of transport costs, the development of spatial economics, or new economic geography,2 has focused attention on this issue. Now the transport cost is analysed by academics in more rigorous way. Lima˜o and Venables (2001) found that a 10 per cent increase in transport costs typically reduces trade volumes by approximately 20 per cent. 118

The Timeliness of Transportation in Economic Integration 119

30 25

(%)

20 15 10

Intra Asia

Intra NAFTA

2004

2002

2000

1998

1996

1994

1992

1990

1988

1986

1984

1982

0

1980

5

Intra EU

Figure 5.1 World trade share of intra-regional trade Note: Asia consists of China, Hong Kong, Indonesia, Japan, Korea, Malaysia, Philippines, Singapore and Thailand. Source: IMF, Direction of Trade Statistics, CD-ROM.

By taking the transport cost into consideration, economic integration can be readdressed in a broader sense as a process which reduces trade costs between producers and consumers across different countries. In this regard, trade costs3 are broadly defined to include tariffs and nontariff barriers, as well as costs of transportation, communication and distribution, etc. There is growing interest among international economists in the fact that trade costs and related transportation policy may affect one country’s degree of integration to the world market, thus affecting its comparative advantage structure and income distribution. From an empirical point of view, however, the broadly conceptualized trade costs are complex and hard to measure in practice. Although distance is still the most commonly used proxy for trade costs, it is not the only factor affecting trade volume. For example, natural geographic conditions certainly matter. Lima˜o and Venables (2001) point out that a landlocked country may have a trade volume 60 per cent lower than a comparable coastal economy because the inland transport cost is much higher than the ocean freight. The question of efficiency of the transportation system was studied using an econometric model by Clark et al. (2004) who found that ocean transport costs may be seriously

120 The Making of Economic Integration in East Asia

affected by port infrastructure, poor handling at customs, and the existence of organized crime. Ocean transport cost also may be affected by fuel prices and technological changes. According to Hummels (1999), contrary to what we expect from technological advances in transportation, we may be paying more for transport cost than before due to the adoption of more fuel-consuming advanced technologies such as bigger and faster vessels for containerized cargo and air cargo shipments.4 Technological advancement has shortened the required time for delivery but the associated monetary costs may be higher. Duranton and Storper (2005) questioned the puzzle that the expenditure for transportation in international trade has increased since the 1950s while the trade volume has increased. One key aspect in their analysis is that development of transport technology induced trade of higher quality goods which before had hardly ever been transported. Today we can buy a wide variety of fresh foods delivered all the way from foreign producers paying high transport costs. Such a ‘quality effect’ more than offsets the direct effect of productivity improvement in the transportation industry. Secondly, the literature of the new economic geography such as Fujita et al. (1999) suggests that lower transport costs stimulate agglomeration of industries. Goods are produced in fewer locations for a larger number of consumers as a result of the agglomeration effect, thus increasing the need for transportation. One related issue is the geographical spread of the fragmented production process. Improvement of transport technology has enabled firms to split production lines which used to be integrated in one single factory and locate them optimally so as to minimize total direct production costs by taking advantage of international factor price differences. The production cost savings should offset the transport cost of moving intermediate products between different segments of production as well as the communication cost to organize the production system efficiently. In order to develop this production network, availability of an efficient transportation system is essential. Last, but not least, there is a growing demand for timeliness in transportation which may have increased the total expenditure for transportation. Hummels (2001) argues that lengthy shipping times impose inventory-holding and depreciation costs on shippers. He found that for manufactured goods, each day’s travel is worth an average of 0.8 per cent of the value of the goods per day, equivalent to a 16 per cent tariff for the average length of ocean shipment. He also argues that each additional day in ocean transit reduces the probability that a country will export to the US by 1 per cent (all goods) to 1.5 per cent

The Timeliness of Transportation in Economic Integration 121

(manufactured goods). Such a cost may become a serious concern for retailers as well as manufacturers engaged in production fragmentation. Evans and Harrigan (2005) also argue that the demand for timeliness affects the pattern of international specialization. They found that US retail stores tend to source apparel products from adjacent countries such as the Caribbean nations and Mexico for their ‘lean retailing’ management. They prefer low inventory and frequent restocking to bulky low-cost procurement because clothing is subject to consumers’ volatile tastes and retailers are worried about being left with vast stocks of unpopular clothing. The same logic applies to ‘just-in-time’ inventory management by timely procurement, typically conducted by Japanese firms. For example, assembling factories of Japanese electronics firms could enjoy much cheaper wage locating factories in inland China but timely parts procurement, including some key components imported from Japan, is not possible in such places. Even paying higher wages, the logistic condition in the coastal areas continues to attract Japanese affiliates who are committed to delivering final products to Japanese consumers without delay. For such business practices, distance acts as a proxy for time to customers, but not necessarily shipping costs. This chapter argues that countries with good accessibility to major markets will have higher wages because they specialize in getting goods, of high value and differentiated style, delivered to markets quickly. On the other hand, countries which lack conditions for meeting the demand for timeliness can only specialize in standardized products. Under such conditions, workers can only compete with cheaper labour cost to attract industries. Nordas et al. (2006) argue that a broader range of products are becoming time-sensitive and in some developing countries the time-consuming administrative process prevents local manufacturing firms from exporting such goods. Without reforms to that effect, firms will be confined to the local market or become exporters of low-value, bulky standardized products. Hence, the strategy of the transport infrastructure and active integration to the regional and the world markets constitutes an essential part of the development strategy of the non-core periphery region. As Carruthers et al. (2004) have shown, East Asia is characterized by a significant disparity in the quality of transport infrastructure both between countries and also within each country. They recommend development of regional mega ports from what are now feeder ports, to enable them to direct service avoiding high trans-shipment costs at current hub ports, in addition to developing efficient loading/unloading services. Reduction

122 The Making of Economic Integration in East Asia

of the cost of inland transport and improving the access to the port are also necessary in order to mitigate the social cost of heavy congestion in port cities.

5.2 Development of timely transportation and revealed geographical advantage In order to illustrate the advantage of good accessibility to major markets, consider the following simple model.5 Suppose that there are three countries: major market H and potential production sites N and S. There are firms in H indexed by i 2 [0,1] considering outsourcing production in either N or S. These firms are uniformly distributed along the horizontal axis of Figure 5.2. Country N has an efficient airport which allows firms to conduct timely outsourcing by using air transportation. By locating in country S, on the other hand, firms are able to employ cheaper labour. Let iU  1 represent a marginal firm where firms i 2 [0, iU] are sensitive enough to timeliness using expensive transportation and firms i 2 [iU, 1] prefer cost saving in labour. How much firms can engage in timely outsourcing (location of iU) is constrained by the availability of an advanced logistic facility. Thus, better transportation enlarges the possibility of timely outsourcing leading to higher iU. For example, containerization has reduced inter-modal loading/unloading time. The temperature-controlled logistic system, or ‘cold chain’, prompted a wide variety of fresh food trade. Information technology enables not only efficient control of inventory and customs procedures but also the transmission of computer-aided product design (CAD) data,

∧ w LL QQ B

∧ wL

A 0

iL

Figure 5.2 Advantage of better logistics

iU

1

The Timeliness of Transportation in Economic Integration 123

enhancing the flexibility of production. These technological improvements induce more firms to relocate from the right-hand side to the left-hand side of iU leading to the rightward shift of iU. We consider that timeliness gives higher satisfaction to consumers, allowing firms to charge higher prices. Firms have an incentive to pay higher wages to workers who accept harder working conditions in committing to the timelines (including night duty and high-speed machinery^ represents the wage based works). In Figure 5.2, the vertical axis w premium paid to workers of timely outsourcing firms in country N while the QQ curve depicts a labour supply which increases as the wage premium rises. On the demand side of the labour market, if the wage in country N is too expensive, few firms are attracted to flexible outsourcing while more firms start flexible outsourcing if the wage differential is smaller. This relationship is depicted by the LL curve. While the flexible outsourcing is technically feasible up to the firm indexed iU for the above-mentioned reason, it depends on the wage level as to how many firms actually do so. The intersection of QQ and LL (point A) gives equilibrium. The size of firms engaging the flexible outsourcing in country N is given by iL and the corresponding wage premium of country N is given by ^ L. This result confirms the possibility of a higher wage if a country w is geographically well-located or has a highly developed transport infrastructure which enables the outsourcing of products whose demand in H is sensitive to timeliness. For firms in the range of iUiL, flexible outsourcing is technically possible but they prefer producing in country S obtaining labour cost savings because the wage premium in country N is too high. We now address the effect of technological advances in transportation between N and H enabling more H-firms to conduct outsourcing, while the transportation condition between S and H remains unchanged. This effect can be represented by the move of iU to the right (expressed by the bold broken line). This causes the associated shift of LL to the right because of the increase of labour demand in timely outsourcing. This changes the equilibrium point from A to B, provoking the increase of ^ L. Thus, the deepening of economic firms located in N and the rise of w integration through logistics improvement, making outsourcing feasible for a wider range of goods, will favour the countries closer to the major markets. This suggests that advanced outsourcing locations, such as Singapore and coastal China, will be able to continue enjoying higher wages provided they can successfully improve access to major markets and attract newer activities, while transferring the less timeliness-conscious

124 The Making of Economic Integration in East Asia

and cost-sensitive production to logistically disadvantageous regions like inland China, Indonesia, Vietnam and Lao PDR. This model also can discuss the impact of the change in market demand so as to become more sensitive to timeliness. Such change may occur when consumer demand becomes more diversified in the quest for seasonal or booming products or because of a short product cycle due to intense technological competition such as in the case of semiconductors and electronics products. This kind of market structure change can be depicted in Figure 5.2 as the upward shift of the LL curve reflecting the higher evaluation of the timeliness by each firm, which realized new equilibrium at point B, even without the improvement of transportation infrastructure discussed above. Thus, the geographical advantage of privileged accessibility to major markets will be revealed when the market demand becomes more intolerant to time cost in trade. In reality, a more complete explanation for the advantage of a port city requires an examination of the consumer demand-driven agglomeration effect.6 Namely, the higher income in such locations attracts more firms which, in turn, create more jobs and income. This circular causation will further raise the wage in port cities until the physical constraint eventually imposes certain diseconomies such as high land rent and other living costs. This effect could be shown in Figure 5.2 as the clockwise rotation of the LL curve with pivot point at iU which causes the reduction of firms in country N and increase in country S.

5.3

The timeliness issue in Japanese trade in East Asia

In the previous section, we argued that the demand for timeliness forms part of the complex nature of trade costs in merchandise transactions around the world and that workers can benefit from the transport advantage by engaging in flexible outsourcing for higher wages. In this section, we develop an analysis of how this argument applies to the current situation in East Asia. As seen before, the intra-regional trade in East Asia has achieved remarkable growth. Many observers, including Fujita and Hamaguchi (2007), have characterized this phenomenon as the geographical expansion of the value chain or production network. The production process is now fragmented across countries to take advantage of the factor price differences within the region. Naturally, such factor cost savings should exceed the cost of linking the geographically dispersed production process. In particular, when such a value chain is controlled by vertically integrated multinational firms, much

The Timeliness of Transportation in Economic Integration 125

care will be taken over timely delivery and short lead times in order to meet the final demand market flexibly. Japanese manufacturing firms were given the hallmark of efficient value chain management in cooperation with their suppliers located nearby for just-in-time delivery of parts and components. We can observe that such a value chain, which used to be closed within a restricted range of distance, has expanded internationally to the dimension of Northeast and Southeast Asia, thanks to the development of transportation and information and communications technology. This is due not only to the shipping cost of goods but also as a result of the wider availability of the Internet and video conferences and easier travel for the eventual need for face-to-face meetings. According to the China National Tourism Administration data (http://www.cnta.gov.cn/), the number of Japanese visitors to China increased gradually from 223,500 to 463,300 between 1981 and 1990. Then it grew abruptly to 2.2 million in 2000 and further increased to 3.4 million in 2005. Direct scheduled flights between Japan and China were established in 1974 only between Tokyo–Osaka and Shanghai–Beijing. Today, seventeen Japanese cities have direct flights to China and reciprocally seventeen Chinese cities have direct flights to Japan, amounting to 540 pairs of direct flights per week. In Japan, Tokyo/Narita has 42 per cent of departing flights to twelve Chinese cities, followed by Osaka/Kansai (26 per cent to eleven cities), Nagoya/Chubu (11 per cent to nine cities) and Fukuoka (10 per cent to seven cities), while in China, Shanghai has the largest share of departing flights to Japan (36 per cent to sixteen cities), followed by Beijing (22 per cent to six cities), Guangzhou (10 per cent to five cities) and Dalian (9 per cent to six cities). Some Japanese local airports have direct flights only to Shanghai. The destination of flights from major metropolises is influenced by business needs. Flights from Nagoya are apparently influenced by the location of Toyota’s operation in China, while from Osaka, the Kansai region-based production locations of electronics firms are widely popular, such as Dalian, Shenyang, Hangzhou, Nanjing and Xiamen. Figure 5.3 depicts the distribution of the accumulated number of cases of Japanese foreign direct investment during 1989–2004 by industrial categories. It is evident that textile and electric equipment industries account for a greater part of investment cases in both ASEAN and China. Next, Table 5.1 demonstrates the distribution of trade from/to Japan to/from China in 1988 and 2005, by industrial classification grouped into twenty-two categories. Japanese exports to China tend to concentrate in XVI (machinery and electrical equipment), XVIII (precision

126 The Making of Economic Integration in East Asia

5,000 Food

Cases (1989–2004 accumulated)

4,500 4,000

Textiles 3,500 Wood and pulp 3,000

Chemical

2,500

Basic metal

2,000

Electric equipment

1,500

Transport equipment

1,000

Machinery

500

Others

0 ASEAN

China  Hong Kong

Figure 5.3 Japanese manufacturing FDI in East Asia: accumulated number of cases by industry, 1989–2004 Source: Ministry of Finance, Outward Direct Investment (Industry & Country), http://www.mof. go.jp/english/e1c008.htm

instruments), XV (base metals and articles of base metal), VI (chemical and allied industries), XI (textiles), and VII (plastics and rubbers). They consist of capital goods, parts and components, and industrial materials. On the other hand, while the share of V (mineral products) in Japan’s imports from China has declined, there was a sharp increase of machinery and electric equipment. Also, although the relative share of textiles has declined, there was a substantial increase in absolute value in category. It is notable that there is substantial trade both ways between China and Japan in the textile sector as well as the machinery and electrical equipment sector, reflecting its intra-industry nature. This may be seen as counter-intuitive if it is judged based on the traditional comparative advantage principle which could suggest a dominance of labour-abundant China in the former and Japan’s superior position in the capital-intensive machinery industries. In this regard, it is worth recalling that the two industries were most noticeable in the figures of the foreign direct investment as shown in Figure 5.3. This leads us to the conjecture that development of the vertical production linkage between Japan and China, involving offshore production of Japanese firms in China

Table 5.1 Japan–China trade Japan’s exports

Japan’s imports

1988

I II III IV V VI VII VIII IX X XI XII XIII XIV XV XVI

2005

1988

2005

1000 yen

%

1000 yen

%

1000 yen

%

1000 yen

%

286,517 1,435,862 358,817 1,591,903 1,299,273 81,486,240 60,131,847 2,214,156 232,379 13,211,859 75,189,255 92,719 16,897,543 453,763 350,285,288 483,457,002

0.0 0.1 0.0 0.1 0.1 6.7 5.0 0.2 0.0 1.1 6.2 0.0 1.4 0.0 28.9 39.8

25,095,120 3,934,449 1,048,603 12,921,289 147,931,182 805,492,929 499,089,347 7,163,337 3,508,151 109,358,545 381,762,432 1,530,222 89,368,060 39,536,341 1,079,514,445 4,024,558,929

0.3 0.0 0.0 0.1 1.7 9.1 5.6 0.1 0.0 1.2 4.3 0.0 1.0 0.4 12.2 45.5

92,553,959 123,491,830 1,620,609 63,315,309 308,867,634 71,558,851 3,166,682 15,587,110 31,240,267 2,195,660 380,078,444 14,501,379 5,710,545 2,745,776 92,915,701 16,275,582

7.3 9.8 0.1 5.0 24.4 5.7 0.3 1.2 2.5 0.2 30.1 1.1 0.5 0.2 7.4 1.3

205,788,807 234,882,950 3,986,735 490,592,679 440,933,774 427,329,979 292,857,882 269,118,719 183,594,487 77,860,428 2,270,852,532 334,144,596 185,572,402 32,995,429 692,486,498 4,327,998,199

1.7 2.0 0.0 4.1 3.7 3.6 2.4 2.2 1.5 0.7 19.0 2.8 1.6 0.3 5.8 36.2 (Continued ) 127

128

Table 5.1 (Continued) Japan’s exports

Japan’s imports

1988

2005

1988

2005

1000 yen

%

1000 yen

%

1000 yen

%

1000 yen

%

XVII XVIII XIX XX XXI XXII

68,051,006 40,388,962 0 7,224,101 72,656 9,563,639

5.6 3.3 0.0 0.6 0.0 0.8

447,436,233 644,713,470 2,337 72,477,999 66,214 440,343,393

5.1 7.3 0.0 0.8 0.0 5.0

261,794 3,113,337 10,659 19,105,000 6,303,126 9,465,069

0.0 0.2 0.0 1.5 0.5 0.7

178,693,619 437,998,811 48,150 791,426,358 2,814,270 87,989,813

1.5 3.7 0.0 6.6 0.0 0.7

Total

1,213,924,787

100.0

8,836,853,027

100.0

1,264,084,323

100.0

11,969,967,117

100.0

(I) Live animals and animal products; (II) Vegetable products; (III) Animal or vegetable fats, oil & waxes; (IV) Prepared food, beverage, spirits, tobacco; (V) Mineral products; (VI) Chemical and allied industries; (VII) Plastics/rubbers & articles thereof; (VIII) Raw hides, skins, leather & furs; (IX) Wood/ wood charcoal/cork/straw/plaiting materials and articles thereof; (X) Wood pulp, paper, paperboard, scrap/waste paper and articles thereof; (XI) Textiles & textile articles, (XII) Footwear, headgear, umbrellas, walking sticks, riding crops & parts thereof; (XIII) Articles of stone, plaster, cement, asbestos, mica or similar materials, glass & glassware; (XIV) Pearls, precious stones/metals and articles thereof, imitation jewellery, coins; (XV) Base metals & articles of base metal; (XVI) Machinery & mechanical appliances, electrical equipment/appliances, parts & accessories; (XVII) Vehicles, aircraft, vessels & associated transportation equipment; (XVIII) Optical, photographic, measuring, checking, precision, medical or surgical instruments, clocks & watches, musical instruments, parts & accessories thereof; (XIX) Arms & ammunition, parts & accessories; (XX) Miscellaneous manufactured articles; (XXI) Works of art, collectors’ pieces, antiques; (XXII) Re-exports. Source: Japan Customs Trade Statistics (www.customs.go.jp/toukei/info/ind.exehtm).

The Timeliness of Transportation in Economic Integration 129

and related intermediate goods trade, has had a major role in the occurrence of intense intra-industry trade. If this is the case, we may expect that trade is associated with rising concern for timeliness in delivery. In this regard, Table 5.2 provides information of the intensity of the use of air transportation and container shipment. The share of air transportation in the total value of exports represents the traders’ interest in getting the products delivered as soon as possible. The merit of containerization is, in turn, an efficient inter-modal transportation from vessels

Table 5.2 Share of air transportation and containerization in ocean transportation in Japan–China trade (%) Air transportation

Containerization/Ocean transportation

Japan to China

China to Japan

Japan to China

China to Japan

1988

2005

1988

2005

1988

2005

1988

2005

I II III IV V VI VII VIII IX X XI XII XIII XIV XV XVI XVII XVIII XIX XX XXI XXII

19.0 3.6 0.2 0.7 0.2 2.1 0.8 9.0 3.4 17.1 7.6 5.8 1.0 48.8 0.2 3.3 0.1 13.5 – 8.6 80.6 31.0

0.6 11.5 0.9 1.3 0.1 9.1 11.1 24.0 1.8 11.6 9.7 26.6 33.1 91.2 4.6 40.6 1.2 61.6 0.0 30.4 29.7 69.0

13.0 1.7 0.3 0.5 0.0 5.5 4.9 25.1 0.2 13.9 11.4 1.6 1.8 57.5 0.6 21.8 8.8 56.0 0.0 6.0 88.6 15.9

12.5 4.8 1.2 0.4 0.0 8.7 4.8 20.9 0.6 4.4 12.1 6.8 8.3 84.6 4.8 42.0 2.1 50.6 5.7 13.4 94.7 51.8

70.4 4.7 89.2 66.4 41.0 59.1 57.3 99.6 66.0 36.0 89.0 70.4 90.7 32.0 4.4 63.1 14.7 62.8 – 78.9 100.0 48.8

97.2 97.5 72.1 87.0 10.3 51.9 99.6 100.0 95.3 99.0 99.9 99.8 99.8 96.1 35.3 87.3 75.1 97.4 100.0 100.0 100.0 96.1

38.4 36.8 19.0 54.1 0.6 63.8 96.4 90.6 48.0 79.8 89.6 97.3 50.7 60.6 35.6 91.2 60.6 92.9 95.1 92.1 89.5 26.4

85.0 94.7 65.0 94.9 6.4 93.0 99.7 100.0 96.2 99.8 100.0 100.0 99.1 98.2 71.6 99.6 97.2 99.9 100.0 100.0 98.8 98.1

Total

3.0

30.1

6.5

22.8

44.0

74.4

48.2

92.2

Source: Japan Customs Trade Statistics.

130 The Making of Economic Integration in East Asia

to trucks and trains, and vice versa. It emphasizes, at the same time, economies of scale because a small lot cargo cannot make a full container load. Both may contribute to shorten the delivery time but, in general, air transportation corresponds to greater necessity for speed and flexibility (small lot and frequent delivery). Table 5.2 also shows that the containerization ratio of ocean trade is approaching 100 per cent for many goods.7 This demonstrates a great improvement of ocean transportation between the two countries, owing to the construction of huge container terminals in major ports of China, taking advantage of the scale economies. The ratio of air transportation of XVI (machinery and electric equipment) is high and rapidly rising to about 40 per cent for both ‘Japan to China’ and ‘China to Japan’. Looking at this sector more closely in Table 5.3, we find that the utilization of air transportation is most intense in electrical machinery and equipment and parts (Harmonized System 2 digits code 85), especially in the trade flow from Japan to China in which a substantial shift of transport mode occurred from ocean container (80.8 per cent of trade value in 1988) to air (increasing from 3.9 per cent in 1988 to 61.4 per cent in 2005). This corresponds to the just-in-time delivery of electrical parts to the assembling process in China. Such development owes much to the diversification of direct flights between various cities in the two countries, as we have seen above. The air transportation ratio for XI (textiles) is moderate8 but the total trade value transported by air in this category is significant. For example, according to Tables 5.1 and 5.2, exports by air from China to Japan (i.e. Japan’s imports) in 2005 account for 12.1 per cent of 2,271 million yen of total exports. This is equivalent to 275 million yen, or about 10 per cent of the total exports by air from China to Japan in that year, which is far less than the group of electric machinery representing 67 per cent but is still the second largest. We can see clearly in the related Table 5.4 that the Japan–China trade in the textiles industry is characterized by the Japanese exports of ‘fibre, yarn and fabric’ and Chinese exports of ‘apparel’, and this tendency is much clearer in 2005 than in 1988. Exports by air of ‘fibre, yarn, and fabric’ from Japan (from 6.3 per cent to 9.1 per cent) and ‘apparel’ from China (from 9.2 per cent to 11.9 per cent) have become relatively bigger in 2005. It is also worth noting that the unit weight value of textiles has declined most sharply for exports by air. This means that the need for air transportation is not only for the high-end use products such as silk. Even standardized products can be exported by air in order to engage in

China to Japan

Japan to China

Table 5.3 Japan–China trade in machinery and electrical equipment Products

Transportation

1988

2005

84

Air Ocean container Ocean non-container Total

6,077,553 90,095,422 133,572,436 229,745,411

2.6% 39.2% 58.1% 100%

315,764,820 1,307,547,283 256,559,644 1,879,871,747

16.8% 69.6% 13.6% 100%

85

Air Ocean container Ocean non-container Total

9,909,750 205,024,982 38,776,859 253,711,591

3.9% 80.8% 15.3% 100%

1,316,854,271 781,433,232 46,399,679 2,144,687,182

61.4% 36.4% 2.2% 100%

84

Air Ocean container Ocean non-container Total

341,587 3,361,341 347,867 4,050,795

8.4% 83.0% 8.6% 100%

780,753,734 1,326,888,155 9,507,924 2,117,149,813

36.9% 62.7% 0.4% 100%

85

Air Ocean container Ocean non-container Total

3,212,715 8,246,668 765,404 12,224,787

26.3% 67.5% 6.3% 100%

1,038,724,967 1,170,679,723 1,443,696 2,210,848,386

47.0% 53.0% 0.1% 100%

84: Nuclear reactors, boilers, machinery & mechanical appliances, computers. 85: Electrical machinery & equipment & parts; telecommunications equipment, sound recorders, television recorders. Source: Japan Customs Trade Statistics.

131

132

Table 5.4 Japan–China trade in textile industry by transportation mode Transportation Japan to China

Ocean container Ocean non-container Total Air

China to Japan

Value ð1; 000 yenÞ

Air

Ocean container Ocean non-container Total

Products

1988

2005

Fibre, yarn, fabric Apparel Fibre, yarn, fabric Apparel Fibre, yarn, fabric Apparel

4,749,561 939,408 61,151,093 684,693 7,508,673 155,827

6.3% 1.2% 81.3% 0.9% 10.0% 0.2%

34,805,341 2,399,910 338,804,946 5,400,729 328,502 23,004

9.1% 0.6% 88.7% 1.4% 0.1% 0.0%

Fibre, yarn, fabric Apparel

73,409,327 1,779,928

97.6% 2.4%

373,938,789 7,823,643

98.0% 2.0%

Fibre, yarn, fabric Apparel Fibre, yarn, fabric Apparel Fibre, yarn, fabric Apparel

8,365,979 34,943,257 141,607,139 160,138,757 32,811,900 2,211,412

2.2% 9.2% 37.3% 42.1% 8.6% 0.6%

3,981,623 269,745,621 143,533,023 1,853,314,194 11,155 266,916

0.2% 11.9% 6.3% 81.6% 0.0% 0.0%

Fibre, yarn, fabric Apparel

182,785,018 197,293,426

48.1% 51.9%

147,525,801 2,123,326,731

6.5% 93.5%

Japan to China

Ocean container Ocean non-container Total Air

China to Japan

Unit value ð1; 000 yenÞ=KGÞ

Air

Ocean container Ocean non-container Total

Fibre, yarn, fabric Apparel Fibre, yarn, fabric Apparel Fibre, yarn, fabric Apparel

8.1 17.7 0.4 2.5 0.5 1.8

4.0 11.4 0.9 1.3 2.6 4.2

Fibre, yarn, fabric Apparel

0.5 4.2

1.0 1.8

Fibre, yarn, fabric Apparel Fibre, yarn, fabric Apparel Fibre, yarn, fabric Apparel

15.8 5.3 0.7 1.3 0.3 0.4

4.2 4.7 0.6 1.5 2.1 1.6

Fibre, yarn, fabric Apparel

0.6 1.5

0.6 1.7

Source: Japan Customs Trade Statistics.

133

134 The Making of Economic Integration in East Asia

flexible production, thus avoiding the risk of accumulating large stocks of unpopular products. There was a consensus a decade ago in Japan that offshore garment production in China is restricted to standardized merchandise, while small orders in short delivery time can be attended flexibly only by domestic producers (METI, 1997: Part II, Chapter 1, Section 1). However, more recently, thanks to information technology, computer-aided design and manufacturing techniques which allow the integration of information flow from customer orders at retail shops to the production line and the expanded availability of direct air transportation services, the timely outsourcing in China which seemed unlikely just a decade ago, has now become a reality. This largely meets the prediction of the simple model of the previous section that improvement of transportation technology stimulates more firms to engage in timely outsourcing despite rising wage rates in surrounding areas of Shanghai and Guangdong where many Japanese investments are concentrated. With respect to Japanese trade with ASEAN, a similar set of figures is presented in Tables 5.5 to 5.7. First, Table 5.5 shows that Japan’s exports to ASEAN are strongly concentrated in industry group XVI (machinery and electric equipment) in both 1988 and 2005, while the imports showed a significant shift from group V (mineral products) to group XVI. Japanese trade in group XVI with ASEAN was considerably larger than that with China in 1988 but the latter surpassed the former in 2005. The utilization of air transportation in this group grew from 24.2 per cent to 53.7 per cent for exports from Japan to ASEAN and from 47.6 per cent to 62.7 per cent for imports of Japan from ASEAN, as shown in Table 5.6. Compared to Table 5.2, these figures are even higher than those with China, although the physical distance to ASEAN is much greater. Table 5.7 gives more detailed information. In 2005 air transportation corresponds to 76.6 per cent of Japanese exports to ASEAN in electric equipment and parts, having risen from 37.8 per cent in 1988. Similarly, 60.1 per cent of imports from ASEAN to Japan of this category are made by air. One possible explanation for such high intensity of air transportation might be the stronger consciousness of time in production in ASEAN such as in the case of semiconductors and other components for computers. The switching to air cargo-based high value products competing in time cost but not in factor cost will be necessary to maintain the competitiveness of the region against cheaper labour countries like China. In this regard, the obvious implications are the need for

Table 5.5 Industry

Japan–ASEAN trade Japan’s exports

Japan’s imports

1988

2005

1988

2005

1000 yen

%

1000 yen

%

1000 yen

%

1000 yen

%

I II III IV V VI VII VIII IX X XI XII XIII XIV XV XVI XVII XVIII XIX XX XXI XXII

14,155,890 4,963,795 2,116,622 10,918,124 15,871,224 167,743,116 112,039,014 3,912,459 376,572 29,061,561 74,942,969 583,153 38,987,395 13,427,158 382,750,474 1,328,191,893 374,404,033 89,345,517 23,450 20,858,351 49,211 47,920,324

0.5 0.2 0.1 0.4 0.6 6.1 4.1 0.1 0.0 1.1 2.7 0.0 1.4 0.5 14.0 48.6 13.7 3.3 0.0 0.8 0.0 1.8

15,207,176 3,559,870 913,865 13,973,597 50,229,024 475,547,843 350,460,671 3,626,881 1,856,029 60,250,904 77,506,824 581,894 90,580,694 99,386,553 1,045,265,700 3,915,838,341 895,500,987 317,872,402 24,622 41,285,798 42,572 452,005,735

0.2 0.0 0.0 0.2 0.6 6.0 4.4 0.0 0.0 0.8 1.0 0.0 1.1 1.3 13.2 49.5 11.3 4.0 0.0 0.5 0.0 5.7

205,307,158 104,499,249 25,883,664 62,743,494 1,315,769,180 44,776,035 116,956,427 4,112,685 392,915,404 3,030,249 42,529,442 3,786,841 8,865,011 58,382,316 161,982,663 104,044,602 2,496,770 8,754,162 0 24,161,678 211,216 42,162,508

7.5 3.8 0.9 2.3 48.1 1.6 4.3 0.2 14.4 0.1 1.6 0.1 0.3 2.1 5.9 3.8 0.1 0.3 0.0 0.9 0.0 1.5

152,348,620 122,827,610 40,570,461 252,813,815 2,125,050,947 210,697,509 375,337,036 11,116,694 314,690,279 62,229,286 130,679,870 26,923,971 62,278,842 51,245,493 303,698,069 2,344,329,515 83,793,357 174,860,989 0 128,886,112 72,733 249,709,625

2.1 1.7 0.6 3.5 29.4 2.9 5.2 0.2 4.4 0.9 1.8 0.4 0.9 0.7 4.2 32.5 1.2 2.4 0.0 1.8 0.0 3.5

Total

2,732,642,305

7,911,517,982

7,224,160,833

135

Note: For industry classification see footnote of Table 5.1.

2,733,370,754

136 The Making of Economic Integration in East Asia Table 5.6 Share of air transportation and containerization in ocean transportation in Japan–ASEAN trade Air transportation (%)

Japan to ASEAN

Containerization/Ocean transportation (%) ASEAN to Japan

Japan to ASEAN

ASEAN to Japan

1988

2005

1988

2005

1988

2005

1988

2005

I II III IV V VI VII VIII IX X XI XII XIII XIV XV XVI XVII XVIII XIX XX XXI XXII

2.4 1.7 0.1 1.3 0.1 5.7 3.2 4.1 5.9 3.5 2.8 9.5 32.4 89.7 2.6 24.2 0.2 29.4 0.0 7.9 98.0 53.9

3.4 6.8 4.9 4.5 0.3 18.7 12.1 8.9 4.7 7.6 11.5 26.3 46.0 96.7 7.4 53.7 1.1 64.3 84.9 24.9 91.2 81.8

4.0 7.1 0.1 0.8 0.0 27.3 0.5 54.8 0.0 12.7 12.9 35.4 4.2 96.4 0.4 47.6 5.0 75.9 – 5.8 63.7 57.6

12.8 13.3 0.1 0.9 0.0 15.3 1.9 49.4 0.2 6.0 11.5 24.6 34.4 89.7 2.8 62.7 7.2 70.4 – 8.0 81.8 85.4

45.6 95.5 60.3 92.8 16.9 63.3 88.6 99.3 69.6 67.3 91.4 87.4 78.0 82.9 19.0 66.4 22.7 77.7 79.7 87.6 0.0 50.5

94.3 98.7 97.4 98.2 20.0 77.2 97.9 100.0 99.2 99.9 99.7 99.5 99.8 99.2 37.7 82.3 61.2 97.8 100.0 99.8 100.0 77.8

76.6 19.2 10.5 56.3 0.2 52.0 98.7 96.2 10.7 96.9 98.9 100.0 79.0 43.9 30.6 99.8 91.1 99.3 – 99.3 100.0 51.8

90.0 32.1 8.0 90.0 2.3 70.8 100.0 99.4 33.6 87.9 100.0 100.0 100.0 97.6 51.9 99.7 92.1 99.8 – 99.8 100.0 95.4

Total

15.7

38.7

6.5

27.8

53.5

70.6

22.5

49.3

Source: Japan Customs Trade Statistics.

high-quality ground transportation systems connecting factory gates and the airport and efficient transportation service providers that play key roles, as well as high institutional capacity for the custom clearance procedure. In an environment of progressively integrated wider regional markets, firms have greater choices for alternative locations. Recalling the simple model discussed in the previous section, it is important for middleincome countries/regions to develop efficient transportation in order to maintain jobs with higher wages. Without good transportation, the

The Timeliness of Transportation in Economic Integration 137

Japan to ASEAN

Table 5.7 Japan–ASEAN trade of machinery and electrical equipment Products

Transportation

84

Air Ocean container Ocean non-container Total

85

ASEAN to Japan

84

85

1988

2005

58,795,037 301,257,288

9.3 47.7

416,333,302 1,014,808,116

24.3 59.3

271,809,354

43.0

281,255,690

16.4

631,861,679

100.0

1,712,397,108

100.0

Air Ocean container Ocean non-container Total

262,891,507 367,064,577

37.8 52.7

1,688,319,364 475,013,354

76.6 21.6

66,374,130

9.5

40,108,515

1.8

696,330,214

100.0

2,203,441,233

100.0

Air Ocean container Ocean non-container Total

25,103,955 17,984,123

58.1 41.6

502,852,097 331,197,003

60.1 39.6

95,659

0.2

2,252,456

0.3

43,183,737

100.0

836,301,556

100.0

Air Ocean container Ocean non-container Total

24,459,491 36,366,472

40.2 59.8

967,810,408 540,174,932

64.2 35.8

34,902

0.1

42,619

0.0

60,860,865

100.0

1,508,027,959

100.0

Note: For product classification see Table 5.3. Source: Japan Customs Trade Statistics.

labour market tends to feel stronger downward pressure due to the competition with less developed countries.

5.4

Concluding remarks

The proposition of associating openness to world trade and stronger economic growth is widely accepted. In this chapter, we have proposed a point of view that emphasizes the important role of good transport infrastructure for a region to become competitive in the timely outsourcing to major markets which, in turn, ensures higher income based on such transport advantage rent. This policy implication should be addressed in medium-income developing countries which still lack the innovative capacity to compete in technological frontiers but at the same time face severe competitive pressure from lower wage countries.

138 The Making of Economic Integration in East Asia

We also need to take into consideration more complex issues. First, the timely logistics depend not only on the airport facility but also on high frequency of direct flights to the major markets, which, in turn, cannot be sustained without substantial volume of potential demand. Second, substantial saving of logistic time is possible through improvement in the efficiency of customs procedures and shipment handling. This can be achieved by institutional reform without requiring heavy investment in infrastructure. Third, logistics concern not only the transportation of goods but also the exchange of information with extensive use of information and communication technology. Fourth, the capability of implementing timely outsourcing requires reasonably good quality human resources. Although our analysis is so far based on monocentric (Japan-based) logistic competition, we can naturally extend our concept to multicentric or hierarchical systems of logistics networks. One simple policy implication for development strategy might be to build a good logistic connection to the nearest higher rank countries along the development ladder. To prove this conjecture will be a task for the future.

Notes 1. Baier and Bergstrand (2001) present an estimation that tariff reduction contributed three times as much as transport cost reduction to the growth of bilateral trade flows among OECD countries between the 1950s and 1980s. 2. See Fujita et al. (1999). 3. The broad concept of trade costs is discussed by Anderson and Wincoop (2004). 4. Fink et al. (2002) demonstrate the effects of anti-competitive practices of private carriers. 5. This analysis is based on Evans and Harrigan (2005) but is slightly modified. 6. See, for example, Fujita and Mori (1996). 7. In particular, containerization of goods related to transportation equipment (XVII) has advanced very much due to the substitution of exports of assembled automobiles for containerized knockdown parts which are now being assembled in China. 8. A higher air transportation ratio is observed in the case of live animals, skins, leather and furs, precious stones, precision instruments, and works of art. This owes more to the nature of these goods (high value per unit weight) than to the necessity of timely logistics to maintain an efficient value chain.

References Anderson, James E. and Eric van Wincoop (2004). ‘Trade Costs’, Journal of Economic Literature, 42: 691–751. Baier, Scott L. and Jeffrey H. Bergstrand (2001). ‘The Growth of World Trade: Tariffs, Transport Costs and Income Similarity’, Journal of International Economics, 53: 1–28.

The Timeliness of Transportation in Economic Integration 139 Carruthers, Robin, Jitendra N. Bajpai and David Hummels (2004). ‘Trade and Logistics: an East Asian Perspective’, in Kathie Krumm and Homi Kharas (eds), East Asia Integrates: a Trade Policy Agenda for Shared Growth, Washington, D.C.: World Bank: 77–93. Clark, Ximena, David Dollar and Alejandro Micco (2004). Port Efficiency, Maritime Transport Costs and Bilateral Trade, NBER Working Paper 1053. Duranton, G. and M. Storper (2005). Rising Trade Costs? Agglomeration and Trade with Endogenous Transaction Costs, CEPR Discussion Paper 4933. Evans, Carolyn and James Harrigan (2005). ‘Distance, Time, and Specialization: Lean Retailing in General Equilibrium’, American Economic Review, 95: 292–313. Fink, Carsten, Aaditya Mattoo and Ileana Cristina Neagu, ‘Trade in International Maritime Services: How Much Does Policy Matter?’, The World Bank Economic Review, 16: 81–108. Fujita, Masahisa and Nobuaki Hamaguchi (2007). ‘Brand Agriculture and Economic Geography: a General Equilibrium Analysis’, Discussion Paper Series 207, RIEB, Kobe University. Fujita, Masahisa, Paul Krugman and Anthony Venables (1999). The Spatial Economy: Cities, Regions and International Trade, Cambridge, MA: MIT Press. Fujita, Masahisa and Tomoya Mori (1996). ‘The Role of Ports in the Making of Port Cities: Self Agglomeration and Hub Effect’, Journal of Development Economics, 49: 93–120. Hummels, David (1999). Have International Transportation Costs Declined? Mimeo, University of Chicago. Hummels, David (2001). Time as a Trade Barrier. Mimeo, Perdue University. Krugman, Paul (1995). ‘Growing World Trade: Causes and Consequences’, Brookings Paper on Economic Activity, 1: 327–77. Lima˜o, Nuno and Anthony Venables (2001). ‘Infrastructure, Geographical Disadvantage, Transport Costs, and Trade’, The World Bank Economic Review, 15: 451–79. METI (1997). White Paper on Small and Medium Enterprises in Japan. Nordas, H. K., E. Pinali and M. Geloso Grosso (2006). ‘Logistics and Time as a Trade Barrier’, OECD Trade Policy Working Papers, No. 35, OECD Publishing. doi:10.1787/664220308873.

This page intentionally left blank

Part II The Evolution of the Core–Periphery Structure

This page intentionally left blank

6 The Evolution of Core–Periphery Structure in East Asia Nobuaki Hamaguchi

6.1

Introduction

In this chapter, we introduce the perspective of new economic geography (NEG) to the analysis of regional integration in East Asia. Economic development of East Asian countries has been often discussed based on the analogy of the ‘flying-geese’ pattern, related to the trickle-down effect of industrialization from advanced countries to less developed countries, leading to a spatial dispersion. From this perspective, locations of economic activities are determined by a comparison of production cost. The rapid catch-up of China in the 1990s has changed this characterization substantially. It is not only the seemingly unlimited cheap labour but also the quickly growing middle-class consumer market that attracts business as well as all kinds of activities in China. The orderly designed ‘flying-geese’ type of regional division of labour no longer applies as in the traditional form. NEG can offer some insights for understanding this new scenario. NEG models emphasize product differentiation and economies of scale, rather than pure competition in cost. The basic proposition of NEG establishes that when transport costs decrease, economic activities tend to agglomerate in a particular location due to an endogenously magnified scale economy, leading to the core–periphery spatial structure. This theoretical field emerged in the early 1990s from the concern for the future of the regional balance of Europe after the establishment of the European Union in 1992. Krugman (1991) conjectured that the integration might make European economic geography look somewhat like that of the United States causing locational specialization and concentration. Baldwin and Wyplosz (2004) observe that European integration has been accompanied by only a modest relocation of industry among 143

144 The Evolution of the Core–Periphery Structure

nations, but at the sub-national level industry has become more concentrated spatially. This fact has led to narrower income inequality across nations and to an increase in inequality within nations. Baldwin and Wyplosz conclude that the impact of integration on industrial location change across countries was governed by the comparative advantage effect, but within countries agglomeration economies became relevant due to much freer labour mobility within, rather than across, nations. In short, the current process of globalization and regional integration witnesses a freer trade of goods and a higher mobility of location of firms and people. Some activities seek lower production costs leading to dispersion from the traditional core to the periphery. Such dispersion may trigger new dynamics of concentration in particular places leading to the local agglomeration phenomenon. The newly emerging cores prosper quickly, providing higher real income in the process of attracting more people to join. Their income level tends to catch up with that of the traditional core, while the disparity is widening against the other regions which remain as peripheries, where workers are immobile because the economic activity is bound to the land. Hence, the impact of globalization and regionalization can be inferred from the NEG analysis of the type of new pattern of economic geography that is emerging.

6.2 The perspective of new economic geography analysis for regional integration In this chapter, the definition of agglomeration economies is that each individual enjoys the benefit of increasing consumers’ welfare or firms’ profit as the size of economic activities in the same region becomes larger. For example, it is often seen that stores that sell goods which are seldom bought, continue to thrive without closing down in a large shopping centre. In this case agglomeration economies mean that differentiated speciality stores can offer the consumer greater choice, thus giving them higher satisfaction (i.e. increasing their real incomes with a given amount of nominal income). Agglomeration economies increase the number of consumers visiting shopping centres which then enables a wider variety of speciality stores to thrive there. Thus, if the scale of a market is large, diverse goods and services are supplied through the differentiation, and the consumers benefiting from this are more attracted and recursively expand the scale of the market. This interactive causal relation cumulatively forms an industrial agglomeration.

The Core–Periphery Structure in East Asia 145

Based on similar logic, we can consider a model involving agglomeration economies from the production aspect. The multilayered supporting industries would be important for production since the assembling plants receive supplies from a large number of parts suppliers. These multiple layers signify the presence of diversified and specialized intermediate goods. If expanding diversity makes the assembling firm more efficient by lowering the cost of production, the firm would then be attracted by those benefits and further demand for intermediate goods would be created. This process enables firms with more specializations, supplying the intermediate goods, to exist. Diversity of intermediate goods and the expanding scale of assembling firms interact and form an industrial agglomeration as in the case discussed above where the diversity of goods attracted consumers. The concept of agglomeration economies in NEG is not an exogenous factor to be assumed beforehand, but formed endogenously as a result of market transactions. Here we can confirm that the intensity of agglomeration economies is closely related to the transportation cost. If the transportation cost of consumer goods is extremely high, firms decentralize their location to be close to consumers because they can supply the products only to the limited area. In this case, consumers are able to buy the products only from the firms located nearby and trade between different places does not take place. Thus, consumers are unable to enjoy the benefits of diversity and there are no agglomeration economies. If the transportation cost decreases sufficiently, firms tend to agglomerate in a particular location due to the cumulative market magnification described above. The benefit of agglomeration economies exceeds the attraction of local demand because firms will not lose much demand in remote markets due to the reduced transportation cost. The relationship between the transportation cost of intermediate goods and the formation of agglomeration modelled by Fujita and Hamaguchi (2001) is given as follows: (1) when the transportation cost of intermediate goods is very high and that of consumer goods is relatively low, the interaction of intermediate goods and final assembly is the strongest and both industries agglomerate in the same location; (2) if the transportation cost of final consumption goods is relatively high and that of intermediate goods is relatively low, final assembly is decentralized and only intermediate goods agglomerate in the places where maximum demand can be obtained; (3) if the transportation cost of intermediate goods is extremely low while that of consumer goods is extremely high, demand for each intermediate goods no longer depends on location, both consumer goods firms and intermediate

146 The Evolution of the Core–Periphery Structure

(%)

goods firms are decentralized and the distribution of industries in the economic space becomes more uniform. Thus, combining the consumer market-driven and intermediate input-driven agglomeration mechanisms, we can conclude that sufficient decrease in transport costs of consumption goods and remaining relatively high transport cost of intermediate goods may create industrial agglomerations. This theoretical deduction explains the reality in the following way. According to the IMF’s International Financial Statistics, East Asia (including Japan, China, Asian NIEs and ASEAN) represented around 20 per cent of the world total GDP in both 1990 and 2005, while its share in the world exports expanded from 20.4 per cent to 26.6 per cent. This means that East Asian exports have grown out of proportion to their economic size. The global trade integration which improved access to major markets stimulated exports of final consumption goods from East Asia to the world market. Thus, East Asia as a whole became an industrial agglomeration in the world economy. This process is accompanied by a rapid increase in the trade of intermediate goods within East Asia. Since transaction of intermediate goods requires timeliness in delivery, their transportation is very sensitive to distance. Still, thanks to the improvement in the transportation system, intermediate goods transactions are not restricted to local but have a regional dimension. Hence, the case of East Asia illustrates one of the propositions of NEG: agglomeration is associated with trade (which is, in turn, induced by a decrease in transport cost in a broad sense) (see Figure 6.1). Given the fact that East Asia as a whole constitutes an industrial agglomeration, a number of questions may arise. Where will industries

100 90 80 70 60 50 40 30 20 10 0

44.1 68.9 10.9 20.4

4.4 12.9 13.8 1980

Extra East Asia Intra East Asia World to East Asia East Asia to world

24.5 2004

Figure 6.1 Share of East Asia’s intra and extra trade in world trade Source: International Monetary Fund, Direction of Trade Statistics.

The Core–Periphery Structure in East Asia 147

locate in East Asia? Will the distribution of economic activities in East Asia become more uneven? In NEG propositions, a lower degree of product differentiation and a lesser importance of scale economies make such products more sensitive to transport costs and tend to disperse spatially, attracted by local demand and lower production cost. On the other hand, agglomeration forces are strong for products which are highly differentiated and there are strong increasing returns to scale. It is also worth recalling that in the European case, the progress of integration led to a clearer core–periphery structure within, rather than between, nations due to the lack of international labour mobility. In East Asia, contrary to how we consider China as an unlimited pool of unskilled labour, industrial agglomerations are restricted to some coastal areas where the wages are rising. Some industries are hollowing out of such core locations attracted by the labour cost differential. Next, NEG models, stylized by Fujita et al. (1999), show that there exists a particular range of transportation cost bringing about multiple equilibria under which both agglomeration and dispersion are possible. Under such circumstances, once an industrial agglomeration is formed, the endogenous agglomeration forces sustain the existence of core–periphery structure from temporary shocks. Thus actual spatial structure is strongly related to historical path dependence, which is another proposition of NEG with an implication for empirical analysis. Namely, although the market potential of China is strong, it does not necessarily mean that other agglomerations which already existed should disappear. It still leaves the possibility that a small temporary shock will trigger the cumulative agglomeration mechanism described above and result in drastic structural change (i.e. from dispersion to concentration, and vice versa). In the meantime, the NEG framework so far has narrowly focused on the impact of transportation cost on the spatial distribution of economic activities. It still has very little to say about the basic problems of economic development such as capital accumulation and technological innovation. However, the classic idea of agglomeration economies as described by Marshall (1890: Chapter 10) was about the knowledge spillover in localized industries which has much in common with the models of endogenous growth theory. This original idea can be incorporated into NEG theory by introducing the diversity of people communicating with each other, which may enhance innovation through the combination of a broad variety of expertise. Knowledge spillover has a geographical aspect in terms of the cost of communication because, despite the development of communication technologies, certain types of knowledge which tend to be tacit by nature can only be properly exchanged on a face-to-face basis.

148 The Evolution of the Core–Periphery Structure

6.3 6.3.1

Regional integration in East Asia Development of inter-regional transportation

For the last decade, regional integration has been intensified in various parts of the world through agreements of free trade among countries that are geographically close to each other. While the economic benefit of free trade is often emphasized, political will plays an important role in the integration process. For example, the consensus for a permanent peace in the region after the devastation of war constitutes the foundation of European integration. The NAFTA was promoted by the prominent political and economic leadership in the United States. Increasing international presence is a common aim of South–South integrations such as ASEAN and MERCOSUR. In East Asia, there is no prominent leadership or definitive consensus urging immediate regional integration. The region is quite diverse in ethnic, religious, linguistic and cultural background, and maintains a difference in political and economic structure. Each country is discrete in terms of intervention, and it is not easy to form a uniform political will in such circumstances. However, as discussed in detail in the other chapters of this volume, the intra-regional flow of trade and investment in East Asia is growing quite rapidly, despite the absence of a comprehensive legal framework of region-wide integration. One of the factors contributing to such trade growth is the development of a transport infrastructure. Coincidentally, the linear distance between Japan and Thailand is almost the same as that between the east coast and west coast of the United States. Therefore, as regions East Asia and North America are on similar scales. However, since major cities in East Asia are by the ocean, mega ports have been created in various locations to take advantage of the geographical features. Ocean container cargos are gathered into hub ports (Singapore, Hong Kong, Kaohsiung and Pusan) by the feeder services from small peripheral ports and large container ships ply between these ports. In these ports, large-scale investments are made responding to the increased size and speed of container ships to maintain their status as the key ports of the world. At the same time, Shenzhen, Shanghai and Qingdao in China and Port Kelang, Laemchabang and Tanjung Pelepas which serve as base ports in Asia, joined the top twenty of the world’s container handling ranking in 2003 (Ocean Commerce Ltd., International Transportation Handbook 2005: 822). Table 6.1 shows that 30.94 million TEU (twenty-feet equivalent unit), which corresponds to almost half of the 65.67 million TEU of the world’s

Table 6.1 Out/In

Volume of sea container transportation in East Asia in 2003 (Unit: 1,000 TEU) Japan

Japan South 174.9 Korea China 1,042.0 Hong 6.8 Kong Taiwan 157.1 Singapore 41.7 Malaysia 117.6 Philippines 121.5 Indonesia 185.0 Thailand 155.1 Vietnam 32.8 2,034.5 Intra regional total import

South Korea

224.4

354.0 2.2 32.8 13.7 35.7 33.1 69.0 31.0 30.4 826.2

China

502.9 504.1

Hong Kong

Taiwan Singapore Malaysia Philippines Indonesia Thailand Vietnam

Intra-reginal total export

To world total export

Share of intra region (%)

349.1 139.3

223.7 60.4

113.1 50.5

105.9 43.1

76.8 55.7

72.1 43.7

121.7 35.1

36.0 72.5

1,825.7 1,179.2

3,954.3 2,764.8

46.2 42.6

720.7

205.7 5.5

156.7 3.5

144.9 3.1

91.5 9.2

57.8 2.8

59.4 4.2

53.1 1.5

2,885.9 119.5

13,398.6 594.7

21.5 20.1

43.7

80.2 80.8

56.0 43.0 69.2

33.3 54.5 21.9 3.1

39.2 76.0 36.1 8.4 28.6

0 24.5 34.5 1.9 23.5 14.0

1,358.9 588.7 607.9 283.4 710.9 648.3 128.4 10,336.8

2,689.2 964.4 1,679.1 524.4 2,209.6 1,790.7 372.5 30,942.5

50.5 61.0 36.2 54.0 32.2 36.2 34.5 33.4

80.7 688.3 228.3 116.8 78.5 126.8 100.5 43.8 23.2 15.3 60.7 114.9 119.8 17.7 5.5 2,211.4 1,825.7

59.4 65.6 24.7 65.4 46.8 0 757.1

0 15.1 161.3 57.3 9.4 610.6

8.6 62.6 47.4 8.2 584.8

39.5 29.8 22.0 492.7

32.1 1.6 322.9

0.9 409.5

261.6

Source: UNCTAD Review of Maritime Transport (2004: Table 49).

149

150 The Evolution of the Core–Periphery Structure

containerized cargo transportation, is exported from East Asia, and onethird of that (10.34 million TEU) is traded intra-regionally. In the interregional trade, exports from Japan, China and Asian NIEs (Hong Kong, Singapore, Korea and Taiwan) are large and the share of the container transportation within this group is 62 per cent of the whole East Asian inter-regional container transportation. In contrast to that, the share of these countries is small compared to their GDP. It is remarkable that Southeast Asian countries (Singapore, Malaysia, Philippines, Indonesia, Thailand and Vietnam) have a relatively large share: 10 per cent with Japan, 7 per cent with China, and 10 per cent within the region. In China and Hong Kong, the share of inter-regional trade is just above 20 per cent, and is relatively small compared to the shares in the other countries. This suggests that they play a role of global production bases as a ‘factory of the world’. On the other hand, 46 per cent of the ocean container cargo exported from Japan is heading to the intraregion of East Asia because Japan exports intermediate goods and capital goods needed by the other East Asian countries to export to the world. The inter-regional air cargo transportation in East Asia is gaining importance and its share in the world air freight transportation in 2001 was 12.7 per cent, which is next to 19.6 per cent of that between East Asia and North America, and 17.1 per cent of that between East Asia and Europe. Four of the five routes with the largest traffic volumes are those between Japan and the Asian NIEs (Korea, Hong Kong, Singapore and Taiwan). These are mainly used for the transportation of electronic parts. According to Bowen and Leinbach (2003), advanced air cargo services, which track air cargo using communication technology, have been used extensively, especially in Manila and Penang where many American and European factories for multinational semiconductor enterprises are located. While ocean container shipments have provided gains from scale economies, air transportation meets the need for speed and flexibility in trade. Such aspects are more important for goods such as semiconductors whose inventory cost is high due to a short product life cycle. East Asia became an industrial hub of the world economy along with worldwide trade liberalization and the development of inter-continental transportation. This owes a lot to the individual countries’ efforts in building up infrastructures, and the progress in transportation and distribution systems with the use of information technology, all of which have contributed to a substantial reduction of the general cost of international trade (Anderson and Wincoop, 2004). As Kuroiwa (this volume) suggests, production has been fragmented, taking advantage of differences in factor prices, and close linkages by the trade of intermediate

The Core–Periphery Structure in East Asia 151

goods have been developed between different locations in East Asia. This diversity serves as a source of international competitiveness for East Asian products. In the following sections we will discuss some characteristics of the rapidly growing new industrial agglomerations in East Asia. Development of the transportation infrastructure mentioned above provided the basic conditions for the formation of such new cores which are suitably connected to the transportation network. Strong linkages with the global market provided the scale economy. Then, the agglomeration economies described in section 6.2 took effect cumulatively, resulting in a dense concentration of productive activities in a limited number of locations. 6.3.2

Industrial agglomeration in East Asia

The deepening of regional integration in East Asia has already brought about several changes. Industrial agglomerations are connected by efficient transportation forming a multi-cored industrial belt spreading from Japan and Korea in the northeastern corner, passing through the coastal provinces of China, reaching Southeast Asia. 6.3.2.1

China

In China, manufacturing industries are agglomerated mainly in the Yangtze Delta, Zhu Jiang Delta and Bohai Bay Rim. Table 6.2 shows that the bulk of the production of colour televisions and air conditioners takes place in the Zhu Jiang Delta, while the Yangtze Delta produces mainly personal computers, integrated circuits, synthetic textiles and washing machines. Bohai Bay Rim has the same amount of production of refrigerators as the Zhu Jiang Delta, and of metal cutting tools and automobiles as the Yangtze Delta. The beginnings of the agglomeration of each industry in each area could be investigated one by one, but each region is open for foreign direct investment, and equips infrastructures for export. In these regions, the diversity of the firms producing intermediate goods and the increasing productivity of the assembling factory interact to produce a synergy effect, and many firms including Japanese ones have set up affiliated plants. In these coastal regions where incomes are growing rapidly, the synergy effect takes place in such a way that diversity is expanded by the entry of multinational manufacturing and service firms attracted by the local market and it further magnifies the local market size. An empirical study by Fan and Scott (2003) found a correlation between industrial agglomeration and productivity, especially in consumer electronics, computers and garments. Although the location of the automobile industry targeting the domestic market has

152

Table 6.2 Production share of Chinese main agglomerations (by product) (Unit: %)

Chang Jiang Delta Pearl River Delta Bohai sea rim Western inland area Fujian province Others

Colour TV

Refrigerator

Air conditioner

Metal cutting machine

Personal computer

Chip

Chemical fibre

Washing machine

Car

11.1

18.2

24.2

37.2

42.0

59.3

65.1

51.2

19.8

38.6 17.1 14.7

25.2 29.5 1.5

45.5 16.7 4.8

6.6 36.7 3.0

27.4 21.7 0.1

26.5 5.7 0.0

3.9 12.1 1.5

9.1 21.8 1.9

4.2 18.9 10.2

3.9 14.6

0.0 25.6

0.0 8.8

0.4 16.1

7.5 1.2

0.9 7.6

5.1 12.3

0.0 16.0

2.0 44.9

Note: Share is derived from the number of production units in each area. Bohai sea rim refers to province of Shandong, Liaoning, Hebe, Peking city, Tianjin city; Chang Jiang Delta refers to province of Jiangsu, Zhejiang, Shanghai city; Western inland area refers to Sichuan province, Chongqing city; and Pearl Delta refers to province of Guangdong. Source: National Bureau of Statistics of China, website.

The Core–Periphery Structure in East Asia 153

been decentralized by national policy, and Fan and Scott (2003) could not identify a positive relationship between agglomeration and productivity; the concentration in a few regions and firms would progress if it grows to export industries. This draws attention to the fact that an attempt to form an industrial agglomeration based on research and development of software and biotechnology using intellectual resources and talents in the universities, has started to show results in Beijing, Shanghai and other cities that have prominent universities. Many new firms are born every year using investment through the companies shared by the universities, assistance for starting up in the incubator facilities, subsidies for entrepreneurs who studied abroad, and other systems. Soon after, multinational enterprises also start to develop research and development offices in these regions. The multinational enterprises at first aimed at adapting to the Chinese market and partial outsourcing of research and development for Chinese firms, but now these regions are ranked as one of the world’s most important bases for developing the cutting edge of technology. 6.3.2.2

Southeast Asia

As a case example in Southeast Asia, we will describe the automobile industry in Thailand and Malaysia going high-tech. In Thailand, the manufacturers of cars and the suppliers of parts, which are mainly Japanese companies, have formed an automobile industrial agglomeration which is now the biggest in East Asia outside Japan. Automobile production in Thailand was 460,000 in 2001, and it reached 930,000 three years later in 2004. In the whole production, 630,000 are commercial vehicles which are mainly one-ton-class pickup trucks. Also the market is characterized by the fact that 430,000 in 630,000 whole domestic sales are commercial vehicles. This peculiar market structure was influenced by the policy of excise tax cuts on commercial vehicles. This may not be just a matter of price; the pickup truck demand is also supported by the preferences of the Thai people. Their export is also specialized in pickup trucks. They are mainly exported to ASEAN and Oceania, and the export has doubled from 170,000 to 330,000 in these three years. This is a typical case of the home market effect, in that it tends to have export competitiveness in the goods in relatively high demand. Japanese automobile firms had invested in the Thai automobile industry earlier on. Now their share of the entire car production, sales and export is nearing 90 per cent (see Japanese Chamber of Commerce, Bangkok, www.jcc.or.th/Data/pdf/JCC_thai_japan.pdf). According to the

154 The Evolution of the Core–Periphery Structure

Singapore

To Europe

Southern China

Eastern China

Japan and Korea

Klang (Malaysia)

Hong Kong

Kaohsiung

Pusan Kwangyang

Shenzhen Yantian

Qingdao Ningbo Shanghai

Laemchabang (Thailand) Tanjung Pelepas (Malaysia)

Feeder

Tanjung Priok (Indonesia)

Feeder

Feeder

Amoy

Nagoya Kobe/Osaka Yokohama/ Tokyo

Feeder

Feeder

To North America

Singapore

Southeast Asia

Figure 6.2 Main routes and feeder services of East Asia container lines Source: Based on UNCTAD’s Review of Maritime Transport (2004: Figure 14) and International Transportation Handbook 2005, Ocean Commerce Ltd.

Toyo Keizai Shinposha’s database of Japanese foreign affiliates, automobile manufacturers like Toyota, Nissan, Hino, Isuzu and Mazda (which started in Thailand in 1962, 1962, 1965, 1966 and 1975 respectively) are located in Samutprakarn province near Bangkok, and latecomers, Mitsubishi and Honda (which started operation in 1987 and 1992 respectively) located their factories in Bangkok or Ayutthaya province and Pathum Thani province in the north. In the 1994 edition of the data, forty-one companies producing auto bodies, engines and parts are recorded. Fifteen of them are located in Samutprakarn province and seventeen in the Bangkok metropolitan area, showing that they are clearly agglomerated around the capital. After that, they deployed to the eastern areas. For example, Mazda established a joint venture company named Auto Alliance International Thailand in Rayong province. In Chonburi province, Isuzu set up a factory of metal casting parts in 1994 and Mitsubishi opened an engine factory in 2003. In the 2004 edition of the data, 147 automobile parts companies were established. In the eastern region, thirty-three companies are located in Chonburi province and thirty-one in Rayong province, forming an agglomeration larger than twenty in Samutprakarn province and sixteen in the Bangkok metropolitan area. As a result, the eastern coastal regions from Bangkok to Laemchabang port are connected as an industrial area by the development of highways, and supply from the eastern region to the market in the metropolitan area and drafting of parts from the existing industrial agglomeration in the metropolitan area can be made without hindrance. In addition, by locating close to an export port, exports have been made much easier. Thailand is a base for supplying automobile parts to ASEAN countries, and is also the fourth biggest import counterpart of Japan.

The Core–Periphery Structure in East Asia 155

60 40 20

1999

1996

1993

1990

1987

1984

1981

0

Bangkok and nearby area Central area Eastern area Figure 6.3 Thailand’s regional per capita value added of manufacturing sector (Unit: 1000 baht, at 1988 price) Source: National Statistical Office.

Figure 6.3 shows that the manufacturing industries in Thailand were concentrated in Bangkok and its suburbs in the early 1980s and expanded to the neighbouring central and eastern areas after the export boom in the late 1980s. It is clear that the production in the eastern area had increased along with the growth of the automobile industry mentioned above. However, by contrast with these three regions serving as cores, the production in the other regions had not changed for twenty years, and it is suggested that a typical core–periphery structure has been maintained. In Malaysia, reducing the disparity in income levels has been set as a goal of the national development plans, and the industrial location policy has been conducted aiming at decentralization. In the northern island state of Penang, the better quality workers with higher levels of education and fluency in English as well as a good airport service have attracted the global production hubs of the multinational semiconductor firms. Some local linkages have also emerged there. When many multinational enterprises were established in Penang and faced a shortage of skilled workers, the state government established the Penang Skill Development Centre (PSDC) in cooperation with the multinational firms who provided finance and staff for the technical training of local employees and prospective students in areas such as quality control, AutoCAD, software programming, and factory automation. Some local capital spin-offs

156 The Evolution of the Core–Periphery Structure

from the multinational enterprises have also emerged to constitute the local supporting industry that is engaged in sub-contracting. This environment enables the multinational firms to decrease costs, and strengthens the location advantage of Penang. The local linkage developed in Penang brings agglomeration economies and strengthens the ‘stickiness’ of the place against the mobility of footloose industry. By contrast, the linkage with the local parts industry was not promoted in the assembly of household electrical goods which was spread out in industrial complexes around the capital, and the main source of profit was the cheap wage rate. Thus, there is a risk that factories will leave these areas when opportunities develop in other countries with lower wages. This is also a problem in the automobile industry which is protected by the national car project. Although the competition has intensified with Thailand where the agglomeration of automobile parts industry has rapidly been strengthened by open policy, Thailand has an irresistible ascendancy in agglomerating automobile parts industries due to the scale economy expanded by export. If the automobile industry in Thailand, which now specializes in commercial vehicles, diversifies to include passenger vehicles (which Malaysia produces as well), the automobile industry in ASEAN would be concentrated in one place. Thailand would then become the ‘Detroit of Asia’, while Malaysia would be a satellite base producing only a small part of the domestic demand just as in the other ASEAN countries, since the progress of inter-regional trade liberalization in ASEAN strengthens agglomeration economies. 6.3.3

Core–periphery structure and the gap of development

From the viewpoint of general equilibrium analysis, the location pattern of one sector of the economy depends on the others. Namely, the existence of agglomeration (or core) specialized in modern industry can be sustained by the food supply from the agricultural sector of the periphery where production is bound to land and labour is immobile. Since the former enjoys higher productivity and more real income than the latter, the core–periphery structure leads to the problem of a gap between them. Moreover, this structure is enhanced through the self-multiplication mechanism mentioned above. Competition between agglomerations leads to greater polarization of economic power in a particular region and the less developed countries and regions are left behind in terms of development. In contrast to the EU, where an agenda for the integration process includes regional policies to narrow development gaps as a common goal, the regional

The Core–Periphery Structure in East Asia 157

integration in East Asia without a formal framework addresses this issue only as a country-based problem and there is no region-wide strategy. In China, as we have seen, industrialization has been concentrated in three regions. As many studies have shown, on the one hand the gap between the provinces has narrowed in both coastal and inland regions, but on the other hand the gap between the coastal and inland regions has widened continuously. Figure 6.4 illustrates the per capita GDP in each province in 1990 and 2002 expressed as a proportion of the national average. In the coastal regions, the income levels in Shanghai and Beijing which were salient in 1990 declined relatively, while those in the provinces ranging from the Yangtze Delta to the Zhu Jiang Delta rose. This confirms that the latter caught up with the former. By contrast, in many provinces in the inland regions, these figures are below 1 in each year, and furthermore deviations from the average were widened in 2002. As we saw in Table 6.2, the core–periphery structures continue to be strengthened, which involves the widening disparity in income levels between coastal and inland regions through the agglomeration of industrialization.

Xinjiang Ningxia

Peking 4

Tianjin Hebe Liaoning

Qinghai 3

Gansu Shanxi

Shanghai Jiangsu

2

Zhejiang

Tibet 1

Coastal area Fujian

Yunnan 0

Guizhou

Shandong Guangdong

Sichuan

Guangxi

Chongjin

Kainan

Konan

Shanxi

Hubei Henan Jiangxi Anhui 1990

Jilin Heilongjiang

Inner Mongolia

2002

Figure 6.4 China’s provincial per capita GDP (1990 and 2002) (yuan at 1995 price, whole country average ¼ 1) Source: China National Statistical Bureau, 2003 China Statistical Yearbook.

158 The Evolution of the Core–Periphery Structure

In Southeast Asia, it is evident that although the regional decentralization of investment was encouraged as a means of reducing disparity in income levels in Malaysia, industrial investment and population growth are still concentrated in Selangor state in the metropolitan area surrounding Kuala Lumpur, and the core–periphery structure is locked in. On the other hand, the southernmost state of Johor is linked to the agglomeration economies in neighbouring Singapore as its suburban production area, and Penang is established as a world base for semiconductor production. Each state achieved growth independently from the agglomeration economies of the Kuala Lumpur–Selangor area. As a result, three cores have been formed in the Malaysian regional economic structure. Johor has caught up with the national average in per capita GDP, and Penang has caught up with the average of the metropolitan area in which Kuala Lumpur and Selangor state are combined (see Table 6.3). However, there has been a widening in the income gap between Kuala Lumpur, where the highly waged headquarters and service industries are concentrated, and Selangor, an industrial area with a constant flow of cheap labour. Johor and Singapore seem to have the same relationship. In contrast, in the industrial agglomeration in Penang, the incoming population is less than the national average in spite of active investment, which suggests that the labour productivity has been increasing. Incomes by regions in Thailand, illustrated in Figure 6.5, confirm the growth of the manufacturing industry in the eastern region (see Figure 6.3). Bangkok, where the share of service industry is high, had achieved a higher growth of GDP than the central regions, at least until the currency crisis in 1997. This figure also suggests that growth of the other regions has stagnated. In contrast to Malaysia where a multicore structure is emerging, the over-concentrated structure has not changed and the regional gap is more serious in Thailand.

6.4

Perspective of the economic integration in East Asia

East Asia is characterized by the diversity of its developmental conditions. This is different from the cases of North America and Europe where income levels of the cores of regional integration (the USA and Canada in the former and the EU15 in the latter) are roughly the same. Taking advantage of diversity, East Asia has implemented the international division of labour and experienced a pattern of regional development known as the ‘flying-geese pattern of industrialization’. In this analogy, only the ‘lead goose’ flies with its own force, while the rest

Table 6.3 Malaysia: number of approved FDI and per capita GDP by state 1980–2000 Average annual population’s growth rate (%)

Johor Malacca Negeri Sembilan Selangor þ Kuala Lumpur Selangor Kuala Lumpur Perak Kedah Penang Perlis Kelantan Pahang Terengganu Sabah Sarawak Malaysia

2.5 1.5 2.0 0.8 1.6 5.0 1.7 1.9 1.9 2.4 1.6 4.6 2.1 2.5 2.6 2.5

Per capita GDP (whole country ¼ 1.00)

Number of approved FDB

1985

1990

1995

2000

2003

86–90

91–95

96–00

01–03

0.88 0.77 0.86 1.73 1.44 1.88 0.77 0.58 1.05 0.69 0.40 0.74 0.64 1.07 0.96 1.00

0.85 0.96 0.78 1.60 1.41 1.65 0.79 0.56 1.17 0.62 0.43 0.76 1.24 0.83 0.93 1.00

0.93 1.05 0.84 1.49 1.32 2.12 0.86 0.59 1.40 0.71 0.42 0.70 1.54 0.67 0.86 1.00

0.96 1.05 0.93 1.54 1.25 2.05 0.81 0.60 1.43 0.67 0.42 0.68 1.54 0.66 0.91 1.00

0.97 1.03 0.92 1.44 1.27 1.99 0.82 0.58 1.45 0.64 0.40 0.68 1.57 0.64 0.94 1.00

760 112 101 1,063 946 107 192 165 432 10 25 70 36 129 120 3,205

986 153 202 1,271 1,156 115 290 279 514 39 26 75 86 223 153 4,297

857 164 165 1,148 1,051 97 259 233 519 13 44 116 79 130 181 3,908

601 97 114 911 846 65 159 166 372 10 23 52 18 66 96 2,542

Source: Office of the Prime Minister, Seventh Malaysia Plan, Eighth Malaysia Plan, Mid-term Review of the Eighth Malaysia Plan.

159

160 The Evolution of the Core–Periphery Structure

160 120 80 40

Bangkok and nearby area Eastern area

2001

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

0

Central area Others

Figure 6.5 Thailand’s regional per capita GDP (Unit: 1000 baht, at 1988 price) Source: National Statistical Office.

depend on shedding activities transferred from more advanced countries. Therefore the regional structure can be seen to have only a single core. The great advance of transportation and communication technology has made East Asia one of the major industrial agglomerations of the world, as a result of the agglomeration process based on market magnification and input–output linkage. Within East Asia, multiple cores, which have their own agglomeration forces, are being formed and industries in each core tend to stay there for long periods. This is in contrast to the flying geese analogy’s continuous transition to technological sophistication. As Baldwin and Wyplosz (2004) point out in the case of European integration, we are witnessing an enlargement of the development gap within the countries of East Asia. Examples of this are China’s coastal provinces and inland provinces and, in Thailand, the intensification of concentration to Bangkok and its surrounding areas. Looking closer at the core regions, however, we can observe the geographical spread of agglomeration. In China, the income gap among coastal provinces has narrowed. The same argument applies to Thailand where industrialization sprawls out to the east and the north of Bangkok. Thus, as agglomeration in the core regions grows, more people enjoy higher incomes. However, the resulting stark core–periphery structure implies a wider inter-regional developmental gap. As far as international disparities are concerned, it might be possible that high wages in the core will surrender cost-sensitive activities to the periphery. The regions currently regarded as the peripheries may join globalization and become new cores. From the perspective of

The Core–Periphery Structure in East Asia 161

NEG, it is very helpful to build up a good transport infrastructure (in a broad sense, including technical assistance in customs control), which enables a country to access global and regional markets more efficiently. We should learn from European experiences which have always paid attention to regional income disparities along with market integration.

References Anderson, James E. and Eric van Wincoop (2004). ‘Trade Costs’, Journal of Economic Literature, 42: 691–751. Baldwin, Richard and Charles Wyplosz (2004). The Economics of European Integration, London: McGraw-Hill. Bowen Jr., John and Thomas R. Leinbach (2003). ‘Air Cargo Services in Asian Industrializing Economies: Electronics Manufactures and the Strategic Use of Advanced Producer Services’, Papers in Regional Science, 82: 309–32. Fan, C. Cindy and Allen J. Scott (2003). ‘Industrial Agglomeration and Development: a Survey of Spatial Economic Issues in East Asia and a Statistical Analysis of Chinese Regions’, Economic Geography, 79: 295–319. Fujita, Masahisa and Nobuaki Hamaguchi (2001). ‘Intermediate Goods and the Spatial Structure of an Economy’, Regional Science and Urban Economics, 31: 78–109. Fujita, Masahisa, Paul Krugman and Anthony Venables (1999). The Spatial Economy: Cities, Regions and International Trade, Cambridge, MA: MIT Press. Krugman, Paul (1991). Geography and Trade, Cambridge, MA: MIT Press. Kuroiwa, Ikuo (2008). ‘Rules of Origin, Local Content, and Cumulative Local Content in East Asia: Application of an International Input-Output Analysis’, in this volume. Marshall, Alfred (1890). Principles of Economics, vol. 4, London: Macmillan.

7 Industrialization through Vertical Production Networks: Can Laos Participate? Daisuke Hiratsuka, Souknilanh Keola and Motoyoshi Suzuki

7.1

Introduction

It is obvious that in East Asia de facto economic integration, whereby the manufacturing production process is split and located across many countries by multinational corporations (MNCs) virtually without any formal integration such as free trade arrangements, is far more advanced than de jure economic integration. Even though the process of de jure, or formal, economic integration in East Asia is still in its early stages, the region had already signed several free trade agreements by the end of 2005, while several others are currently under negotiation. However, many members of these free trade agreements have different sets of exclusion lists, making these trade deals complicated and difficult to manage. Moreover, there is virtually no mechanism that allows members who are lagging behind in economic terms to catch up with their developed counterparts in the economic integration process. Currently, the new members of ASEAN, namely Cambodia, Lao PDR (Laos), Myanmar and Vietnam, generally referred to as CLMV, are being granted a grace period before having to fulfil their due obligation of lowering tariff barriers in the tariff reduction programme under the Common Effective Preferential Tariff (CEPT) scheme, but this is the only mechanism in place in the ASEAN framework to economically rescue its new members. The logic behind this is that despite the existence of both positive and negative effects in economic integration, lowering tariff barriers is, in general, considered to have a negative impact on the industrialization of domestic economies. The economic gap between states in East Asia as a whole, and in Southeast Asia in particular, is remarkable. As shown by the joint study report at the ASEAN-Japan Research Institute Meeting (2003),1 major differences 162

Industrialization through Vertical Production Networks 163

remain in terms of living standards and in the stages of industrial development between the six original member states and the CLMV countries. The report emphasized that the specific and differential treatment of simply allowing a longer period of implementation was not sufficient to make up the difference and that ASEAN needed a regional programme and an institutional process so as to enable the late-starters to overcome their handicaps. The CLMV countries have two kinds of handicaps. First, they are economic peripheries, in other words, they are unindustrialized.2 Second, their home markets are quite small, particularly for Cambodia and Laos, both in terms of income and population size. The question arises as to whether these economic peripheries will be able to overcome the handicaps and to develop together with other industrialized East Asian economies in the process of economic integration which will be promoted by de jure integration. In Vietnam, with a population of over 70 million, and a gradual but firm progress in industrialization through the increased presence of foreign direct investment (FDI), there is little need for concern over the problems stated above. Indeed, MNCs have rapidly expanded their production networks to Vietnam in recent years, partly because wages in big cities in the coastal areas of China have increased, and partly because the production risk is higher due to the concentration of production in one country. However, for small home market peripheries, meaning an unindustrialized economy with small domestic market residing next to or near an industrialized economy with a bigger domestic market, particularly Laos, it is still very unlikely that MNCs will extend their production facilities there, through which industrialization could be realized, except for enterprises of neighbouring economies such as China, Thailand and Vietnam which have extended their market-seeking FDI in new frontiers of the country. This chapter examines how economic peripheries can overcome the handicaps of having small markets and being unindustrialized, and how they might move ahead with industrialization and successfully integrate with other East Asian economies. Traditionally, East Asian nations, with the exception of Singapore and Taiwan, which had small home markets, have adopted the dual-track policy that mixes exportoriented policies with import-substitution or protection policies in manufacturing as well as the service sector (Fouquin et al., 2006). However, we doubt that this sort of policy is viable for economic peripheries with small home markets. Therefore, an alternative solution is needed,

164 The Evolution of the Core–Periphery Structure

rather than the traditional dual-track policy. We attempt to systematize the conceptual framework of a sub-regionally complementary industrialization strategy as advocated by Suzuki and Keola (2005), incorporating the concepts of core and periphery structure and transportation costs (broadly defined) which are properties of spatial economics, and we examine the possibility of sub-regionally complementary industrialization at firm level, by investigating broadly defined transport costs, focusing on Laos as an economic periphery with a small home market.

7.2 Industrialization through vertical production networks Suzuki and Keola (2005) have advocated an alternative development strategy for Laos, a small economic periphery. They use the term, ‘sub-regionally complementary industrialization’.3 However, there is no clear definition and a lack of theoretical framework in the concept of ‘sub-regionally complementary industrialization’. For the purposes of clarification, we first use the term ‘industrialization through vertical production networks’ instead. We define this as a small periphery that is to be gradually industrialized by complementing its neighbour core (industrialized country) in a particular part or manufacturing process where the neighbour has lost its advantage, thereby participating in vertical production networks across countries in the wider region. Consequently a small periphery (unindustrialized country) will move forward with industrialization and be integrated into a subregion by participation in vertical production networks. It is assumed that the agglomeration force consequently induces the dispersion of manufacturing production locations, and an isolated town is replaced by a business district (Fujita et al., 1999). The idea of industrialization through vertical production networks is not something new. It has actually been studied and labelled extensively by trade economists. Hanson et al. (2003) defined vertical production networks as a form of ‘vertical FDI’ through which multinationals spread, across different locations, the various activities they perform, such as R&D, input production and input processing. Hummels et al. (2001) discussed vertical specialization as production processes that involve a sequential, vertical trading chain stretching across many countries, with each country specializing in particular stages of a good’s production process. Deardorff (1998) explained fragmentation as the splitting of a production process into two or more steps that can be undertaken in different locations but that lead to the same final product.

Industrialization through Vertical Production Networks 165

The key idea here is that countries increasingly link sequentially to produce goods. Vertical production networks, vertical specialization, and fragmentation are discussed in the context of trade and location theory, but not in terms of industrialization, at least not explicitly. Our concern, however, is with the industrialization of the least developed countries. Consequently, we use the term ‘industrialization through vertical production networks’ here instead. We believe that an economic periphery with a small home market can overcome its disadvantages of being remote from both the market and the supplier, by complementing in manufacturing with its large and industrialized neighbours where wages and land prices are relatively higher due to advancement of industrial agglomeration in those countries. Industrialization through vertical production networks might be a practical development strategy for a small economic periphery. Generally, due to the home market effect, a large economy tends to be able to trade with and to host FDI from distant economies, while a small economy tends to rely on the trade with and investment from economies with geographical distance proximity (see Chapter 10). This means that industrialization of a small economy tends to depend on trade with and oneway FDI from its industrialized neighbours, or other main trade and FDI players in the world. In other words, a country with a relatively huge population may try to industrialize by trading or cooperating with production in more distant cores due to the home market effect, while, for a country with a small home market, trading and cooperating with cores with geographical proximity, or industrialization through vertical production networks, seem to be the more realistic option. Of course, this is not to suggest that small countries should avoid trade or cooperation in production with industrialized nations further afield but simply that this is more likely to happen on a limited scale. Industrialization through vertical production networks is a lesson in experiences from existing ideas and practices. Even Singapore and Taiwan were just small peripheries in the 1960s and transformed into industrialized economies. Since the 1980s, due to increased wages, MNCs operating in Singapore have extended their production facilities to the hinterlands of Malaysia and Thailand (Hiratsuka, 2006). These were also peripheries at that time. Initially, MNCs moved the low value-added manufacturing processes, and then the other higher valueadded ones. In some instances in the electronics industry, the low value-added, tiny parts of the manufacturing processes were moved first, and then the other parts gradually followed. The key parts and assembly processes were the last to be moved.4 Seagate, a leading producer

166 The Evolution of the Core–Periphery Structure

of the hard disc drive, is a good example of this. One part of its hard disc drive production facility was moved from Singapore to Thailand in the early 1980s, followed by the assembly process. This event transformed Thailand into the world’s hard disc drive supplier (McKendrick et al., 2000; Hiratsuka, 2006). The automobile and textile industries, on the other hand, moved their low-valued knock-down assembly process and apparel process first, followed by other processes.

7.3

Transport costs

Our concern is about the particular conditions under which manufacturing in the core will move to its neighbouring periphery with a small home market. Many studies provide theoretical arguments as to why manufacturing in the core moves to the periphery. Krugman and Venables (1995) discuss it with the incorporation of the concept of transportation costs between the core and periphery. They claim that when transportation costs (broadly defined) fall until such levels that the location advantage of low wages in the periphery offsets its disadvantages of being remote from both market and suppliers, manufacturing in the core will move to the periphery. This argument suggests that reduction of transportation costs (broadly defined) and the wage gap would be the key elements in moving manufacturing from the core to the periphery, and industrializing the periphery. What are broadly defined transportation costs? They are sometimes referred to as transportation costs as well as trade costs, and include all costs needed to transport goods to users. Anderson and Wincoop (2004) define trade costs as that which includes all costs incurred in getting goods to a final user other than the marginal costs of producing the goods themselves. They break down roughly into international trade costs and local distribution costs (wholesale and retail). In the current context, however, broadly defined transportation costs mean international trade costs. International trade costs, according to Anderson and Wincoop (2004), can be broken down into (1) transportation costs (narrowly defined, both freight costs and shipping time costs), (2) policy barriers (tariff and non-tariff barriers), (3) language barriers, (4) currency barriers, (5) information barriers related to remoteness, and (6) contract enforcement costs (insecurity costs associated with contractual enforcement and with corruption).5 Shipping time costs impose costs for holding larger buffer-stock inventories at final destinations to accommodate variations in arrival time, as well as depreciation costs (Hummels, 2001). In principle, it is the customer who has to bear all these components of

Industrialization through Vertical Production Networks 167

international trade costs. Since international trade procedures are a heavy burden for customers, forwarder businesses which provide customs clearance services as well as transport services have played a significant role in increasing international trade, in particular, vertical production networks. In general, therefore, border-related barriers, with the exception of tariff costs, are included in transportation service charges offered by forwarders. International trade costs are large. Based on the gravity model, Anderson and Wincoop (2004) estimate trade costs for industrialized countries to be 21 per cent transportation costs including 10.7 per cent of direct transportation costs and 9 per cent for time costs (1.21 ¼ 1.107  1.09),6 and 44 per cent of border-related trade barriers. A breakdown of the 44 per cent border-related trade barriers is an 8 per cent policy barrier, a 7 per cent language barrier, a 14 per cent currency barrier, a 6 per cent information barrier, and a 3 per cent contract enforcement and securities barrier for industrialized countries (1.44 ¼ 1.08  1.07  1.14  1.06  1.03). These trade costs are seemingly large for developing countries, particularly for least developed countries (LDCs). Direct transportation costs (freight costs) would be high for a small LDC since the costs depend on trade volume.7 Lima˜o and Venables (2001) found that the freight costs depend on infrastructures such as road and rail networks, and the costs in the landlocked countries are higher than those in the coastal countries. As for indirect costs, Hummels (2001) estimates time cost or willingness-to-pay for shipping time savings using variation across exporters and commodities in the relative price per speed trade-off for air and ocean shipping. Each day saved in shipping time is worth 0.8 per cent ad valorem for manufactured goods. These studies suggest that transportation costs (narrowly defined including freight costs and shipping time) would be high for LDCs due to lack of infrastructure and as well as low trade volume. Information barriers and border costs associated with information barriers would be high for a small LDC since the costs depend on remoteness. Freight costs and information costs are difficult to manage for an individual country. Manageable border costs are contract enforcement and insecurities. The costs of writing contracts and enforcing them are greater across borders. Anderson and Marcouiller (2002) show that insecurity lowers trade substantially, when associated with contractual enforcement problems and corruption, due to the low quality of institutions. This is evidence that, in developing countries, the low quality of institutions is a high border-related trade barrier. This section suggests that improving institutions as well as infrastructure will reduce transportation costs (broadly defined), and promote

168 The Evolution of the Core–Periphery Structure 7,000 Thailand Trade volume (million US $)

6,000 5,000 4,000 3,000 2,000

Vietnam

1,000

China Singapore Japan Cambodia Germany Malaysia Korea 0 Myanmar Philippines 0 2,000 4,000 6,000 8,000

France UK New Zealand 10,000 12,000

USA Canada 14,000

16,000

Distance between capital cities (km)

Figure 7.1 Relation between trade volume (1981–2004) with and distances to Laos Source: Trade volume from Direction of Trade Statistics (IMF) and Distance from CEPII, http:// www.cepii.fr

industrialization through vertical production networks by the movement of manufacturing from the core to the periphery.

7.4 Potential of industrialization through vertical production networks in Laos Industrialization through vertical production networks is a combination of existing ideas and practices. From the perspective of spatial economics, Laos is an almost perfect case of an economic periphery with a small home market. Laos is an unindustrialized country of about 6 million people, twothirds of whom live in mountainous areas. Laos is now importing most industrial products mainly from the border-sharing countries of Thailand or China in greater or lesser degrees respectively. These countries have experienced agglomerations of industry to some extent, due to economies of scale coming from export production and their huge home markets. As mentioned before, a small economy tends to trade with and to host investment from economies with geographical proximity. Figure 7.1 shows Laos’ trading partners and Figure 7.2 shows its sources of inward foreign direct investment. Laos’ largest trade partner is Thailand, followed by Vietnam and China. The role of Vietnam, where the degree of industrial agglomeration is still far behind that of Thailand and China, is significant in trade and inward foreign direct investment in Laos. This is perhaps partly explained by the ‘special relationship’ between the two countries after the

Industrialization through Vertical Production Networks 169

Approved inward FDI cases (case)

400 350

Thailand

300 250 200 China 150 Korea

100

France

Vietnam

Malaysia Singapore Taiwan Japan Germany Cambodia Philippines 0 Myanmar 0 2,000 4,000 6,000 8,000

50

USA England New Zealand 10,000

12,000

Canada 14,000

16,000

Distance between capital cities (km)

Figure 7.2 Relation between number of approved inward FDI (1988–2004) and distances to Laos Source: Approved inward FDI from Department of Domestic and Foregin Investment (Laos) and Distances from CEPII, http://www.cepii.fr

establishment of their current regimes in the mid-1970s. As for inward foreign direct investment, Thailand has also been the largest source, followed by China. Vietnam too has been an important investor in Laos. The above evidence indicates that industrialization in Laos will depend on trade and investment from neighbouring economies, in particular, Thailand. In the context of the Krugman and Venables (1995) study, broadly defined transportation costs and differences in wages play significant roles in moving manufacturing from the core to the periphery, leading to industrialization through vertical production networks. There is a big difference in wages between Laos and Thailand. The daily minimum legal wage in Thailand, as of January 2006, is 184 baht (US$4.56) for Bangkok and its neighbouring four provinces, 166 baht (US$4.12) for Chonburi, and 145 baht (US$3.65) for Lumphun.8 The official minimum wage in Laos, on the other hand, is 98000 kips per month, equivalent to US$9.68 per month.9 However, in general, the actual average wage in foreign factories including overtime seems to be about 500,000 kips or about 2000 baht/month (or 91 baht per day assuming one works twentytwo days a month) in Vientiane, the capital city of Laos. It is evident that the average wage, including over time, in Laos is still only about 49.4 per cent, 54.8 per cent and 62.7 per cent of those minimum legal wages in Bangkok, Chonburi and Lumphun provinces respectively. As companies are generally hiring workers for wages higher than the minimum rate

170 The Evolution of the Core–Periphery Structure

regulated by law, one can safely expect the real difference between wages in Laos and Thailand to be higher than the figures given above. Apart from low wages, what Laos can utilize as its location advantages to complement Thailand are, (1) its geographical proximity or the sharing of a relatively long common border, (2) a low language barrier, and (3) a low currency barrier. While geographical proximity is somewhat easy to understand, the term ‘low language barrier’ may need some additional explanation. Laos and Thailand share a common ethnic background and their languages are very similar. Thus, it is possible for almost any Lao or Thai to understand each other’s spoken languages without any formal training. This extremely low language barrier lowers communication costs and makes it possible for foreign and local companies operating in Thailand to set up and operate factories in Laos with minimal costs for overcoming language barriers. Similarly, currency barriers should also be addressed. Trade between Laos and Thailand can now be settled in one another’s currencies. However, since trade between them is dominated by the export of goods from Thailand to Laos, the currency used in settlements is mainly the Thai baht. Generally the Lao people and firms also have Thai baht bank accounts at the banks in Laos. This results in lower trade costs as there is no need to change currency to and from the US dollar.

7.5

High fixed cost and border-related barriers in Laos

Considering the geographical proximity, low language and currency barriers, the broadly defined transportation costs between Laos and Thailand or China or Vietnam should be low. As has been emphasized already, however, due to the small volume of transportation, freight costs in less or least developed countries are greater than in developed countries. This is particularly true in the case of landlocked Laos. As can be seen in Table 7.1, the freight costs from Vientiane to Bangkok are higher than those from Bangkok to New York. For instance, the minimum charge for a courier service from Vientiane to Bangkok is US$34, as compared with US$18 from Bangkok to New York. The cost of sending 5 kg by courier is US$70 which is much higher than US$48 for the same service from Bangkok to New York. However, the courier service charge for 100 kg from Vientiane to Bangkok is US$357, lower than US$684 from Bangkok to New York. This reversal price could have resulted from the high fixed cost in Laos. Freight costs can be broken down into fixed cost and distance charge. Fixed cost per distance, which is composed of

Industrialization through Vertical Production Networks 171 Table 7.1 Courier service cost from and to selected countries (unit: US$) To\From

Bangkok min 5 kg 100 kg Singapore min 5 kg 100 kg Vientiane min 5 kg 100 kg

Bangkok

Beijing

Hanoi

Kuala Lumpur

Vientiane

New York

Singapore

18 32 364

18 30 344

13 23 260

18 36 463

18 48 693

10 18 208

13 67 298

18 88 444

18 88 474

10 48 193

18 88 474

22 111 684

30 57 357

39 84 674

34 70 560

30 57 357

45 99 789

30 57 357

Source: Compiled by authors based on Facts and Figures, 2000 Edition, ASEAN Secretariat, Jakarta.

office rent, labour cost and other maintenance costs, goes down as volume and distance of transport increase. The offered prices in Table 7.1 include not only freight costs but also border-related barriers, as discussed by Anderson and Wincoop (2004). It is not easy to estimate the cost for all the components of international trade, such as freight cost, time cost, policy barriers (tariffs and non-tariff barriers), information costs, contract enforcement costs, currencies costs, legal and regulatory costs, and local distribution costs in the same way as in Anderson and Wincoop (2004), due to lack of sufficient and reliable trade statistics within and between Laos and foreign countries. We can, however, see from the above that fixed cost per distance is high for courier services to and from Laos. Below are some possible explanations. The average per weight freight cost would generally be higher due to the small volume of goods flows between Laos and Thailand. The fact that freight cost from Thailand to Laos is relatively lower than from Laos to Thailand may partly be explained by the fact that imports from Thailand to Laos exceed exports from Laos to Thailand. Seemingly equally important are the export and import procedures in Laos, where exporters and importers are required to submit many documents to different government offices. This process can take anything from a few days to a whole month in some cases, as will be discussed later on. In general, a shorter time is taken for documents that are in-plan or submitted and approved by the authorities in advance, and

172 The Evolution of the Core–Periphery Structure

a longer time is required for documents that are not planned (out-plan). The courier service process seems to be quite smooth for documents (one to two working days from Vientiane to Tokyo) but this may not be the case if larger items are imported or exported, requiring payment of higher tariffs through the courier service. The lack of flexible and comprehensive procedures is obvious, and it is expected that service providers will have to set their price relatively higher to cover these difficult to predict costs for their operation in Laos. The different procedures for planned and unplanned import–export means that businesses such as factories need to import materials needed for production and also to export their products (or semi-products). Table 7.2 shows the documents needed for out-planned import by a foreign direct investor. It should be noted that the processing time for foreign direct investment-related documentation is what took the longest in this case. While the problem may not be due entirely to the FDIMC (the Foreign Direct Investment Committee which has now changed to the DDFI or the Department of Domestic and Foreign Investment), because coordination with the related authority is also required, the lengthy process means that Laos can only participate in a limited number of vertical production networks. It should, however, be noted that the requirement for this sort of advanced approval is not unique to Laos but is also practised in Thailand. However, efforts have been made to ensure a quicker processing time for in-plan import–export, without inspection or almost instantly in some cases, while the time needed for out-plan import–export is now much shorter.

7.6 Case study of high border-related costs in Laos and their effect on its participation in vertical production networks What direct evidence can we obtain for broadly defined transportation costs and how do these affect the ability to participate in vertical production networks? We will examine two cases where one has successfully taken advantage of reduced border-related costs and expanded production capacity, while the other is still affected by high border-related costs. 7.6.1

Case A

Tokyo Coil Engineering, a Japanese company manufacturing camera strobe transformers and trigger coils, provides an example of industrialization through vertical production networks as well as direct evidence of

Table 7.2 Documents needed for an out-plan import of machinery by land No. 1

2

3

4

Name of documentation 1 2 3

Invoice and packing list Inform of company Inform of FDIMC

4

Inform of FDIMC

5 6 7 8 9 10 11 12

Invoice and packing list Industry import form Import planing Import summary every 3 months Business Licence Tax Licence Investment Licence Factory Construction Approval document

13

Inform of FDIMC

14 15 16

Invoice and packing list Industry import form Form of commerce decrease 1%

17 18 19 20 21 22 23 24 25

Form of commerce decrease 1% Inform of FDIMC Invoice and packing list Industry import form Customs declaration form Detail of material list form Bank Guarantee Visa form Outward carrier report

Office

FDIMC

Import export FDIMC

Industry department

Period

Remark

5~15 days

By Truck out plan

Import–export industry division

1 day

Commerce

Commerce department

To pay 0.3% for inform plan 1 day

Customs department at Thanaleng for pay 1%

1 day

173

Source: Based on interview by authors.

Ministry

174 The Evolution of the Core–Periphery Structure

transportation costs at firm level. The company set up an overseas plant in Thailand in 1990, and expanded its overseas operation to Laos10 in 1999. It also has factories in China but our interest focuses on factories in Thailand and Laos which have cooperated in the production process. The Lao factory provides support to the main factory located in Lam Phun (26 km southeast of Chiang Mai) in Thailand. Manufacturing processes in Thailand have gradually been moved to Laos, and at present, all trigger coils and old models of transformers are produced in Laos.11 At the factory in Vientiane, Laos participates in producing trigger coils and old models of transformers which require lower level of skills, compared to the new model of transformers, all of which were manufactured in Thailand. On the other hand, the Thailand factory has produced the new model of transformers, and carried out R&D including kaizen (innovation), repair and instalment of machinery both in Thailand and Laos. The high technology factory in Thailand has taken two shifts (regular four-hour overtime) and a 24-hour operation system to pay off the high cost of machinery, and to attain economy of scale. The low technology factory of Laos, however, has taken one shift and an eight-hour plus overtime operation system to exploit low wages. The Ricardian type technology differences between Thailand and Laos are shown in the above operating system. Materials of plastic parts and copper wire are shipped from the Thai factory to Laos, and manufactured trigger coils and transformers are shipped back to the Thai factory for final inspection before being shipped to customers in Japan, Hong Kong and Thailand. The cooperation and logistic route between the Tokyo Coil factory in Laos and Thailand is illustrated in Figure 7.3. The factory in Laos regularly imports in-plan raw materials from the Thai factory in Lam Phun by truck twice a month. Occasionally, it also imports the out-plan ones by plane. The out-plan imports increased by seven to eight times a month in 2006 with the production expansion of the Tokyo Coil Engineering group.12 In the case of out-plan, approval by the Department of Domestic and Foreign Investment under the Planning and Investment Committee would be needed in addition to due procedures in in-plan cases. The import procedure, which used to take one week in the past, has been shortened to a few days by improving the institutional quality. All the transportation costs of importing materials from the Thai factory to the Lao factory are borne by the Thai factory. As for exports, processed products of the Lao factory are packed in small boxes measuring 30  40  21 cm and transported (four to six times a month) from Vientiane to Chiang Mai by air via Bangkok and

Industrialization through Vertical Production Networks 175

Export

Import Airport

Airport Export

Import

Lao factory Export

Thai factory

Flow of raw material Flow of products Import (Thanaleng bonded warehouse)

Export Import Border

Figure 7.3 Cooperation of Tokyo Coil Laos with Tokyo Coil Thailand

then by truck to the Thai factory in Lam Phun. It now takes a maximum of three days including two days of export procedures, which can be submitted to the authority in advance. The Lao factory transports parts to the airport in Vientiane in its own trucks. One shipment in July 2006 was 529 kg and its transportation cost was US$816, equivalent to 2.1 per cent of shipment value, US$40,190.13 This transportation cost is quite high compared with the labour cost of Tokyo Coil Laos, which was 12.5 per cent of shipment value,14 and was borne by Tokyo Coil Thailand. The possibility of an exporting route from the Lao factory to the Thai factory by truck is being examined, in order to reduce the transportation costs as border barriers are lowering due to institutional improvements in their case.15 Tokyo Coil Engineering has a long operational experience, using transport service by an MNC freight forwarder, and active interaction with authorities in Laos, which may be the reason why it has improved its import and export procedure time. The company actually made a lot of complaints when they started operating in Laos as it took weeks and sometimes months to import materials. The situation has improved, and though it is still not as good as in an industrialized country, it is difficult to deny the connection between the reduced time for import–export and

176 The Evolution of the Core–Periphery Structure

the expansion trend of production capacity for the factory both in the number of employees and the instalment of a new production line during the past few years. 7.6.2

Case B

Another example is Asahi Maxima (Laos) Co., Ltd. which suffers from lengthy procedure times (Figure 7.4). Asahi Maxima Laos has a mother factory called Thai Asahi Denki in Samut Prakarn, a suburb of South Bangkok. Both are affiliates of Asahi Denki in Osaka. Besides Thailand, it also has overseas plants in Indonesia and China. The factory in Laos16 began operation in 2003. Wages are very low: 500,000 kips, about 2,000 baht (less than one third of the factory in Thailand), or about US$50 per month including a daily overtime of two and half hours. Asahi Maxima Laos imports raw materials by land (truck, or truck and train) from Thailand to Laos once a month and sends back processed products to the Thai factory, on the same route, twice a month. Transportation of imports has been handled by a family-based forwarder in Thailand, and costs are borne by the Thai factory, except for the very short distance from the Lao factory to the friendship-bridge or border check point. A major problem is the unstable import time.

Thai factory Lao factory

Import

Export

Export

Import (Thanaleng bonded warehouse)

Flow of raw material Flow of products Export Import Border

Figure 7.4 Cooperation of Asahi Maxima Laos with Asahi Denki Thailand

Industrialization through Vertical Production Networks 177

With a short field survey, we found a good example of still unstable import time for this company even after a few years of operation in Laos. In this example, out-plan materials arrived at the Thanaleng international trade terminal in Laos on 18 July 2006, and although the import procedure had been started on 5 July 2006, the materials did not reach the factory until 27 July 2006. In this case this was caused by a very slow approval process at the Department for Promotion and Management of Domestic and Foreign Investment (DDFI). Generally, importing requires permission from the Ministry of Commerce, and the Customs Department (Ministry of Finance). In addition, permission is required from DDFI in order for the foreign direct investment company to benefit from the tariff exemptions or reductions that it is entitled to by law. Asahi Maxima Laos took three days to obtain permission from the Food and Drug Department, two weeks from DDFI, and two days from the Ministry of Commerce and Customs Department. If 0.8 per cent or more for every single day in transportation is assumed, as computed by Hummels (2001), the full shipping time would be 8 per cent (0.8 per cent  10 days) for waiting time. Even though the import tariff was exempted, a warehouse charge for 10 days at Thanaleng terminal was imposed. More importantly, very slow import procedures are said to have made it impossible for this Lao factory to produce short, or even middle lead time products. This case indicates that contract enforcement costs, in some cases, can be very high in Laos.17 As far as exports are concerned, there are not many problems at the moment.18 However, the contraction of production capacity of the Lao factory in this case seems to be correlated to the border-related costs it faces, in addition to other problems such as a high turnover rate, and the lack of discipline of the Lao worker which the company claims is prevalent. We have discussed only two cases, but these two companies have been pioneers through their participation in vertical production networks in the electronics industry in Laos. Transportation costs, including time, have also proved important in larger vertical specialization (in term of numbers) in the clothing industry which was not included in our analysis. We have not discussed this industry because most of its companies were initially located in Laos in order to bypass import quotas imposed on clothing products, which is an external factor in terms of Laos’ participation in vertical production networks. The evidence from Tokyo Coil and Asahi Maxima shows that borderrelated costs may play an important role in the potential of Laos to engage in short lead time production, in particular, to participate in the ‘just-in-time’ production system in the vertical production networks.

178 The Evolution of the Core–Periphery Structure

If these costs decrease, there can be a greater degree of participation in the vertical production network as can be observed in the case of Tokyo Coil’s production capacity which has increased both in terms of the number of employees and investment in a new production line. If the border-related barrier remains high, only long lead time production19 would be possible in Laos. This would mean infrequent procurement of materials and delivery of products, i.e. once or twice a month. For countries interested in taking part in the ‘just-in-time’ production system in vertical production networks, attention must be paid to the fact that delivering semi-finished components from a local factory to others on foreign soil ‘on a daily basis’ is nothing special among relatively developed ASEAN countries. These activities would certainly be impossible without cooperation from the relevant authorities. Yi (2003) developed a two-country Ricardian trade model in which each country specialized only in a particular stage. In that, he argued for tariffs (border-related barriers) to be imposed twice: once when the first stage enters a foreign country, and once when the second-stage goods are imported back into the home country. He went on to say that if tariffs are high enough, all vertical specialization is eliminated, but when tariffs fall low enough so that vertical specialization kicks in, the effects of further tariff reduction become much higher. In reality, this customs procedure is carried out four times in total on imports and exports between Laos and Thailand for production networks as in the case stated above; the number of times required to pass through border-related barriers is four (see Figure 7.5). Border-related barriers would be multiplied by four (or two for a country þ two in another Thai Export

Laos Import

Import

Export

Figure 7.5 Number of times to pass through trade-related barriers

Industrialization through Vertical Production Networks 179

country) making its effect much bigger when it increases and much smaller when it decreases. Tokyo Coil’s case confirms that improvement in customs procedures seems to contribute to further movement of manufacturing processes to Laos. As the number of border-related barriers to be overcome is four, there exists the possibility for improvement. Consequently, when a single stop service system is implemented, border-related barriers coming from trade procedures will be eliminated to a certain extent, leading to the creation of environmentally friendly industrialization through vertical production networks as well.

7.7

Summary and conclusion

The CLMV countries have two handicaps in being economic peripheries (unindustrialized), and having small home markets. Will these economic peripheries be able to overcome these handicaps, and to develop together with other industrialized East Asian economies in the process of economic integration, which will be promoted by de jure integration? With geographical proximity, and generally low language and currency barriers, we argue that even economic peripheries with small home markets can be industrialized by systematically participating in certain manufacturing processes, in particular those that can be shifted from neighbouring cores to peripheries in the same sub-region. We defined and conceptualized this as industrialization through vertical production networks. As Krugman and Venables (1995) have argued, when transportation costs (broadly defined) fall low enough, manufacturing in the core will move to the periphery. This means that industrialization through vertical production networks depends on broadly defined transportation costs. Broadly defined transportation costs are roughly composed of international trade costs (freight costs and time costs, and border-related barriers). We addressed the fact that customs clearance is needed four times for each cross-border manufacturing cooperation between two countries participating in a vertical production network, i.e. (1) export procedure on the Thai side, (2) importing procedure on the Lao side, (3) exporting procedure on the Lao side and (4) exporting procedure on the Thai side) in the vertical production network. Therefore, the effect of reduced border-related barriers such as a smoother export– import procedure in each border would be multiplied. We investigated the freight costs, time costs and the border-related barriers at firm level. We provided evidence that not only freight costs but also the border-related barriers were high in Laos when compared

180 The Evolution of the Core–Periphery Structure

with neighbouring Thailand. We found one case where the degree of participation in the production network was enhanced after the company successfully took advantage of reduced border-related barriers as a result of institutional improvement. However, there was another case where the company was still suffering from high border-related barriers, where the degree of participation in the production network was marginalized, e.g. unstable import time meant that only tiny and long lead time products could be manufactured. We concluded that border-related barriers play an important role in Laos’ potential to participate in vertical production networks and that they have been reduced (even though more can still be done) in general by extensive cooperation between the Lao government and external bilateral and multilateral donors. However, for various reasons, some companies are still unable to enjoy the benefits of reduced border barriers, making their cooperation in vertical production networks limited to a certain degree. Whether Laos can participate in vertical production networks will depend greatly on the border-related costs that it will be able to remove as well as improvement in the following areas. It is quite clear that MNCs will play an important role in the viability of industrialization through vertical production networks, as these are the ones to decide what processes should be split, and where. Many other things, besides making an effort to reduce border-related barriers or trade facilitation, are also necessary. These must go hand in hand with investment facilitation as MNCs would play a major role in industrialization through vertical production networks. This has been limited in Laos, partly due to the budgetary constraints of the government. There are still no industrial estates in Laos, although plans to create some include an economic special zone in the southern province of Savannakhet. Firms have to pay a relatively large amount of money to acquire land and construct infrastructure that is needed to begin production. Lastly, since the driving force behind shifting parts of the production network to unindustrialized countries by MNCs is to reduce cost, which usually involves taking advantage of lower wages, improvement in the area of quantity and quality of the labour force is also important. Although labour cost in Laos is much lower than in Thailand, seemingly with enough numbers for a certain agglomeration of industrial factories, a major problem is that most of this labour force still lacks the discipline and mindset needed for industrialization, as can been seen in their high absence and turnover rates among other things. Time may be needed for the agricultural-based

Industrialization through Vertical Production Networks 181

Lao labour force to adapt to working in industrial factories, but recognizing the problem and making an effort to improve the situation would definitely be a good step to start with.

Notes 1. The meeting was initiated by the Institute of Developing Economies. 2. The term ‘periphery’ is used to refer to the core–periphery structure in which manufacturing concentrates in one country while leaving others unindustrialized (see Krugman, 1991). 3. Here we use the term ‘sub-regionally complementary industrialization’, following Suzuki and Keola (2005). They state that as many as 5,000 Japaneseaffiliated companies are based in Thailand today. Thanks to this concentration of plants, industries in Thailand requiring large supplies of precision parts and sophisticated technology, such as the automotive and home appliance industries, have come to boast global scale quality and cost competitiveness. However, due to the rise in the costs of labour or production, Japanese-affiliated firms in Thailand are now struggling with the issue of cost reduction. 4. Order can definitely vary among industries, as low and high value-added processes or parts can differ between industries. 5. Anderson and Wincoop (2004) broke down trade costs into transportation costs (both freight costs and time costs), policy barriers (tariffs and non-tariff barriers), information costs, contract enforcement costs, costs associated with the use of different currencies, legal and regulatory costs, and local distribution costs. 6. Indirect costs include holding costs for goods in transit, inventory costs, and preparation costs associated with shipment size. 7. A logistics company calculates freight cost, considering average (monthly) trade volume and average fixed investment and operation costs. Freight costs, therefore, tend to be low for developed countries and high for developing countries. Interview with Logitem Group, Hanoi, Vietnam, on 28 April 2006. 8. The exchange rate of the baht is 40.25 baht per US dollar, and the figures for the minimum legal wages of Thailand and the exchange rate of the baht against the US dollar are from the Key Economic Indicators of the Bank of Thailand. 9. The exchange rate of kip is 10,106 kip to the US dollar. 10. Though the executives of the Lao factory repeatedly stated that expansion of production networks to reduce location (or country) risk was the reason for having the factory in Vientiane, there is no doubt that the wage gap between Thailand and Laos is one of the main factors making the operation in Laos commercially viable. The Lao factory has been managed by an experienced Thai female manager, who was sent to take charge of setting up the factory in Laos, and is operated by women workers and indirect staff. 11. The number of workers increased from 300 to 500 in 2003. 12. Interview at Vientiane factory on 27 July 2006. 13. Interview with Tokyo Coil Laos, on 26 July 2006. 14. Telephone interview with a staff member at Tokyo Coil Laos by the authors on 8 June 2006.

182 The Evolution of the Core–Periphery Structure 15. Telephone and e-mail interview of Thai manager in Tokyo Coil factory in Laos by authors in June 2006. 16. At the beginning, it engaged in micro-motor production, but the rejection rate of products was too high to reduce the total cost of the Thai factory. Currently, Asahi Maxima Laos no longer engages in producing micro-motors, but has shifted to a very labour-intensive resistor production in which almost all of the operations are carried out by hand. The number of employees was reduced from over 500 workers for micro-motor production to 260 workers for resistor production. 17. Interview with Asahi Maxima Laos, 27 July 2006. 18. On 27 July 2006, 195 boxes were shipped. The transportation cost was 16,300 baht, broken down as follows: 7,500 baht for three truck charges between the Lao factory to Nongkhai, a border city on the Thai side; 5,000 baht for the train between Nongkhai and Bangkok, and 3,800 baht for documents. Each component charge includes a service commission charge of the forwarder. The transportation cost between Bangkok railway station and the Thai factory is to be borne by the Thai factory. The transportation cost of 16,300 baht or US$407.50 was about 2.7 per cent of the shipment value, US$14,878.65. 19. According to a survey funded by the Swiss government conducted in mid2005, the lead time of most garment factories in Laos is about one to three weeks and orders are usually placed three months before delivery, and reducing lead time would be necessary in order to raise Laos’ competitiveness as a production location.

References Anderson, James E. and Douglas Marcouiller (2002). ‘Insecurity and the Pattern of Trade: an Empirical Investigation’, Review of Economic Statistics, 84, 2: 345–52. Anderson, James E. and Eric van Wincoop (2004). ‘Trade Costs’, Journal of Economic Literature, 42: 691–751. ASEAN-Japan Research Institute Meeting (2003). Joint Study Report ASEAN-JAPAN Comprehensive Economic Partnership: Vision and Tasks Ahead. The Institute of Developing Economies (www.ide.go.jp/Japanese/Lecture/Sympo/pdf/e_report_all.pdf). Deardorff, Alan V. (1998). ‘Fragmentation in Simple Trade Models’, paper presented to the session on ‘Globalization and Regionalism: Conflict or Complements?’ North American Economics and Finance Association, Chicago. Fouquin, Michel, Daisuke Hiratsuka and Fukunari Kimura (2006). ‘Introduction: East Asia’s De Facto Economic Integration in East Asia’, in D. Hiratsuka (ed.), East Asia’s De Facto Economic Integration, Basingstoke: Palgrave Macmillan. Fujita, Masahisa, Paul Krugman and Tomoya Mori (1999). ‘On the Evolution of Hierarchical Urban Systems’, European Economic Review, 32: 209–51. Hanson, Gordon H., Raymond J. Mataloni Jr. and Matthew J. Slaughter (2003). ‘Vertical Production Networks in Multinational Firms’, NBER Working Paper, 9723 (http://www.nber.org/papers/w9723). Hiratsuka, Daisuke (2006). ‘Vertical Intra-Regional Production Networks in East Asia: a Case Study of the Hard Disc Drive Industry’, in D. Hiratsuka (ed.), East Asia’s De Facto Economic Integration, Basingstoke: Palgrave Macmillan.

Industrialization through Vertical Production Networks 183 Hummels, David (2001). ‘Time as a Trade Barrier’, Working Paper, Purdue University. Hummels, David, Jun Ishii and Kei-Mu Yi (2001). ‘The Nature and Growth of Vertical Specialization in World Trade’, Journal of International Economics, 54: 75–96. Krugman, Paul (1991). ‘Increasing Returns and Economic Geography’, Journal of Political Economy, 99, 3 (June): 483–99. Krugman, Paul and Anthony J. Venables (1995). ‘Globalization and the Inequality of Nations’, Quarterly Journal of Economics, 110, 4: 857–80. Lima˜o, Nuno and Anthony J. Venables (2001). ‘Infrastructure, Geographical Disadvantage, Transportation Costs and Trade’, World Bank Economic Review, 15, 3: 451–79. McKendrick, David G.., Richard F. Doner and Stephan Haggard (2000). From Silicon Valley to Singapore: Location and Competitive Advantage in the Hard Disc Drive Industry, Stanford, CA: Stanford University Press. Suzuki, Motoyoshi and Souknilanh Keola (2005). ‘Sub-regionally Complementary Industrialization Strategy for Laos under Economic Unity’, Macroeconomic Policy Support for Socio-Economic Development in the Lao PDR, Phase 2, Main Report, Volume 1: 25–42. Yi, Kei-Mu (2003). ‘Can Vertical Specialization Explain the Growth of World Trade?’ Journal of Political Economy, 111, 1: 52–102.

8 Structural Change in Intermediate, Consumption and Capital Goods Trade During Economic Integration: the EU Experience Bart Los and Jan Oosterhaven

8.1

Introduction

The East Asian integration process was preceded by another integration process. From the early 1950s onwards, an increasing number of European countries started to tear down mutual tariff walls and reduce non-tariff barriers to trade. This chapter aims to analyse the consequences of earlier integration in Europe, and to give insights into potential future effects of similar tendencies in East Asia. Production is becoming increasingly fragmented. This is leading to an expansion of intra-industry trade and in East Asia also to an expansion of intra-regional trade (see Kuroiwa, Chapter 4, this volume). In East Asia this process is further stimulated by increasing de jure economic integration. This raises concerns such as: What will be the consequences for industrial location? Will large economies benefit most? Will small economies develop together with the larger ones? What national policies are needed to profit from the integrating market? In this chapter we will contribute to understanding these phenomena by sharing the European economic integration experience. However, countries involved in the economic integration in East Asia and those in Europe are quite different, especially as regards the differences in economic development that have been much smaller in Europe than in East Asia.1 Also most of the borders in Europe are land borders, which enable small firms to trade with neighbouring regions just across the border, whereas most trade in East Asia has to cross sea borders requiring larger shipments. Consequently, transportation costs are likely to be considerably higher in East Asia than they are in Europe. Thus, extrapolating from the EU experience to the East Asian situation entails a number of important caveats. It will not be impossible, however, especially as 184

Trade During Economic Integration 185

regards the size of the countries involved, as both regions consist of economically small countries along with economically large countries. We will analyse whether domestic transactions inside EU countries (intra-country trade) grew faster or slower than trade with the rest of the EU (intra-EU trade) and trade with the rest of the world (ROW trade). In particular, we will investigate whether there is a difference in this between small and large countries. We will also investigate whether trade in intermediate goods grew faster or slower than trade in consumption goods and trade in capital goods, and whether this differs between intra-country, intra-EU and ROW trade. Thus, we will be able to give an indication as to whether outsourcing by multinationals has become more important at the aggregate sectoral level. Finally, we will analyse whether intra-industry trade grew faster or slower than inter-industry trade, and in particular whether there is a difference in this between intermediate, consumption and capital goods trade, and between intra-country, intra-EU and ROW trade. Thus, this chapter will investigate the nature of the changes in trade and in specialization for the EU. Before we do this we will give an overview of the theories that relate the spatial concentration of industry (location theory) to the specialization of domestic demand (home market effect), the specialization of exports (inter-industry trade) and the growth of intra-industry trade (new trade theory) (section 8.2). From this it follows that inter-country input–output tables represent the best data to investigate these changes, as these are the only data that integrate both sectoral supply (i.e. domestic production 1 inter-country imports) and sectoral demand (i.e. domestic use 1 inter-country exports) in one accounting framework. Section 8.3 will briefly explain the nature of the data used and section 8.4 will summarize earlier research results based on the inter-country input–output (IO) tables of the University of Groningen for the period 1959–85. The paradox we find – increasing specialization of exports along with increasing intra-industry trade – may only be solved if import specialization grows in the same direction. The virtual absence of inter-industry trade in intermediate goods and the dominance of intra-industry trade in intermediates seem to confirm this. Section 8.5 will present an extension of part of this research for the period 1975–95. The main difference with the earlier research results will be the systematic distinction between intermediate, consumption and capital goods. The reason for this distinction is the overview of the literature that indicates that partly different forces may be at work (see also Kuroiwa, Chapter 4, this volume). We find that most ‘early’ tendencies in economic openness and in intra-industry trade continue in

186 The Evolution of the Core–Periphery Structure

later periods, but some trends reverse. Import shares for intermediate inputs started to decline, mainly after 1990, whereas those for consumption and capital goods continued to increase. Intra-industry trade for the larger trading (goods) sectors also changed, namely from increasing to decreasing, whereas the increase continues when small trading (services) sectors are given an equal weight. In our concluding section 8.6, we discuss some potential implications of our findings for the current process of Asian integration, fuelled by multilateral trade agreements.

8.2

EU integration and trade and location theory

Scientifically the economic rationale for starting economic integration in Europe in the 1950s is somewhat of a puzzle. The classical trade theories of that time were theories of comparative advantage that predicted that the largest gains were to be reaped by integrating countries with different comparative advantages. In the Ricardian trade theory such differences were explained by relative differences in technology. This is due to the fact that even when a country has an absolute productivity disadvantage in the production of every commodity it will still have a relative advantage in the production of some commodities. Consequently, economic integration will induce that country to export those commodities and to import the commodities in which it has a relative disadvantage. In the Heckscher-Ohlin-Samuelson (HOS) version of classical trade theory, comparative advantages are derived from differences in relative factor endowments. Even in the absence of any difference in technology, countries with a relative abundance of labour will export labour-intensive commodities and specialize in the production of them. The dominant prediction of these theories are welfare gains due to the increase of trade in different commodities (inter-industry trade) and the subsequent specialization of production into the commodities in which each country is best. Thus, aside from important political and historic reasons, starting de jure integration by means of the establishment of the European Coal and Steel Community (ECSC) in 1950 made sense economically, as the countries at hand had quite different natural endowments of coal and steel. This is different for the start of the European Economic Community (EEC) in 1958 and of the competing European Free Trade Association (EFTA) in 1959, as the countries involved had rather equal endowments and quite comparable levels of technology in manufacturing. Only in the cases of agriculture and transport services, would different climatic and geographical endowments produce the predicted

Trade During Economic Integration 187

welfare gains through the increase of inter-industry trade in these commodities between member states. For customs unions like the EEC, the Viner-Meade partial equilibrium theory for individual commodities furthermore predicted that welfare gains would be larger in the case of low external tariffs, but might turn out to be negative in the case of new high common external tariffs. In the first case, the positive welfare effects of trade creation between the member states would dominate the negative welfare effects of trade diversion from outside more efficient suppliers, towards inside less efficient ones. For the European Union empirical research in the 1960s and 1970s showed trade creation effects, estimated at US$9 to US$20 billion, and trade diversion effects, estimated at US$ 2.5 to 11.2 billion (see Balassa, 1975: 104). Trade diversion is defined as negative if trade with external countries increases because of a common external tariff that is lower than the average of the former separate countries’ tariffs.2 In the late 1970s the positive effect of external trade creation due to the GATT became dominant. Jacquemin and Sapir (1988) estimated double trade creation with regard to both internal and external EU trade in 70 per cent of the commodities, with external EU trade creation being even stronger than internal EU trade creation in most of the cases. Oosterhaven (1995) found comparable increases in internal and external openness of the EU for 1959–75, especially after 1965. It was not until the 1970s and 1980s, however, that a convincing scientific foundation for the economic integration of manufacturing Europe was provided by the then emerging new trade theories. As opposed to the classical trade theories, new trade theories assume increasing returns to scale and thus some kind of imperfect competition. The most general approach assumes monopolistic competition and a love of variety on the part of consumers (Dixit and Stiglitz, 1977). Applied to economic integration it predicts the growth of bilateral trade in the same product (intra-industry trade) by countries with comparable endowments and technology (Krugman, 1980; Dixit and Norman, 1980). The welfare gains are twofold. On the supply side welfare grows because of the exploitation of economies of scale by concentrating the production of each variety in a single country only. On the demand side welfare grows because of the availability of more varieties of each product compared to autarky. Other new trade theories assume countries with natural monopolies or oligopolies before economic integration. Given the large number of possible game theoretic oligopoly set-ups, almost any conclusion is possible. Both zero trade as well as intra- and inter-industry trade may be explained this way, while individual countries

188 The Evolution of the Core–Periphery Structure

may end up with negative welfare effects, although most set-ups predict positive welfare effects due to increased competition. In the 1990s Krugman (1991) and others (Fujita et al., 1999) formally integrated new trade theory with classical location theory (Isard, 1956) and labelled it inappropriately as new economic geography. It essentially adds transport cost and spatial mobility of labour to new trade theory, but it explicitly avoids assuming any difference in geographical endowments. Hence, the label new spatial economics would be more appropriate.3 In contrast with both classical and new trade models, adding transport cost leads to the absence of (factor) price convergence, while adding factor mobility leads to the possibility of the spatial concentration of all non-land bound production in a single region or country. In these models, economic integration is simulated as a reduction in the non-physical part of transport cost. The theory predicts spatial dispersion at both high and low levels of transport cost and spatial concentration of economic activity and people at intermediate levels of transport cost. When the assumption of labour mobility is replaced with that of multiple industries (Venables, 1996) the theory becomes more appropriate for international trade. Full or partial spatial concentration of (clusters of vertically related) industries is still a possible outcome, but agglomeration of non-land bound activity is no longer possible, as the assumption of spatial mobility of labour within a sector is replaced by that of sectoral mobility within a country. Consequently, this variant then predicts mixtures of intra- and inter-industry trade. Up till now, new spatial economics has shown little interest in the welfare effects of economic integration.4 One of the reasons might be that it is technically difficult to separate the (artificial) positive welfare effect of reducing transport cost from the (real) welfare effect of increasing variety, economies of scale and competition that result from more international trade. Domestic demand and input–output linkages do play a role in new spatial economics, but only as a ‘second nature’ force. They do not cause international trade and the spatial concentration of industry, they only multiply initial comparative advantages and economies of scale and variety. In Michael Porter’s qualitative integration of various economic and business theories, they also play a ‘first nature’ role as they represent two of his four primary causes for the broader concept of ‘competitive advantage’ (Porter, 1990). Having a large, strong and critical domestic demand, according to Porter, induces innovation and specialization of production, and thus creates a competitive advantage of domestic firms on foreign markets. The same holds for the presence of strong,

Trade During Economic Integration 189

innovative and competitive upstream and downstream firms, i.e. having clusters of industries with strong forward and backward linkages. Porter’s approach primarily explains the rise of inter-industry trade, as his third first nature cause is ‘factor conditions’ from classical trade theory. Only his fourth first nature cause opens up the possibility of creating competitive advantages in different segments of the same market, as differences in ‘firm strategy, structure and rivalry’ may induce firms from different countries to develop a competitive advantage in different segments of the same market. To summarize, economic integration may result in intra-industry trade as well as in the spatial concentration of individual industries and closely related clusters of industries. Besides, economic integration may be accompanied by specialization of domestic demand if that is the reason for the specialization of exports. Clearly, inter-country input–output tables are needed to study these impacts simultaneously, as such tables represent the only data source that consistently integrates the sources of the supply (i.e. domestic production 1 inter-country imports) and the sources of demand (i.e. domestic use 1 inter-country exports), per sector, per country.

8.3

EU inter-country input–output tables and model

The ‘ideal’ inter-country input–output table for the European Union contains full information on the sectoral and regional origins as well as on the sectoral and regional destination of each intermediate (z) and each final product (y) within the Union (Isard, 1951). The layout of such a table, with its country-by-country sub-matrices, is presented in Figure 8.1. Besides the transactions within the EU there are of course transactions with third countries (t). From a modelling point of view, these need not be fully disaggregated. Figure 8.1, however, shows the imports from third countries in the form of full intermediate input Zts and full final demand Yts matrices per sector of origin in ROW, where s indicates the EU country of destination. This detail on sectoral origin is necessary from an analytical point of view, as it is needed to be able to calculate the real technical coefficients of sector j in country s, which show the need for inputs from a specific sector i from all over the world per unit of output of sector j in country s, as follows: os ss s ts es s asij ¼ (zts ij þ zij þ zij )=xj ¼ (zij þ zij )=xj

ð1Þ

where i and j indicate the sectors of origin and destination, ts ¼ imports from third countries into s, os ¼ imports from other EU countries into s,

190 The Evolution of the Core–Periphery Structure

Z11 Y11

…

: Zr1 Yr1

Z1s Y1s

…

: …

:

Zrs Yrs

Z1s Y1s :

…

:

Zrs Yrs :

t1t

x1

:

:

trt

xr

:

:

Zr1 Yr1

…

Zrs Yrs

…

Zrs Yrs

trt

xr

Zt1 Yt1

…

Zts Yts

…

Zts Yts

ttt

me

V1 Vy1

…

Vs Vys

…

Vs Vys

vt

ve

(x1 y1)’

…

(xs ys)’

…

(xs ys)’

te

Figure 8.1 The ‘ideal’ full-information inter-country input–output table Source: Linden and Oosterhaven (1995).

ss ¼ domestic inputs, and es ¼ inputs from within the EU. In the case of exports to third countries a comparable detail about the sectoral destination of these exports is not necessary and one column per EU country or origin r (trt) is sufficient. The same holds for the transit of products from third countries through the EU to other third countries (ttt), which represents an important difference with the Asian-Pacific IO tables of IDE (see Oosterhaven et al., 2007). It is clear from Figure 8.1 that a simple aggregation over the indexes r and s will produce Zee and Yee, the core of the consolidated input–output table of the EU as a whole. Furthermore, Figure 8.1 clearly shows that an EU inter-country table provides many more opportunities for consistency checks than a consolidated EU table. Hence, even if one is not interested in the inter-country table as such, it would make sense to construct the consolidated table via the detour of the inter-country table. Such a procedure produces a more reliable consolidated table as compared to the case where it is constructed directly, i.e. without using the double-entry consistency checks at the EU country level (see Boomsma and Oosterhaven, 1992). Such a detour furthermore opens up the possibility of avoiding the inconsistency of adding import matrices measured in ex customs’ prices to domestic matrices measured

Trade During Economic Integration 191

in producers’ prices, as is done by Eurostat (1983, 1986; see Linden and Oosterhaven, 1995). Finally, of course, the inter-country input–output model offers many more analytical opportunities than the consolidated EU table, as will be shown in the following sections. The EU inter-country IO tables used in this chapter are constructed from Eurostat’s harmonized national IO tables that consist of four subtables and Eurostat’s import trade statistics (see Eurostat, 1990). The semi-survey construction method essentially proceeds along the following steps (see Linden and Oosterhaven, 1995, for details): 1. The IO sub-tables with primary inputs Vs and Vys , total output xs and total final demand ys are directly put into the corresponding submatrices and sub-vectors of Figure 8.1. 2. The IO sub-tables with domestic transactions measured in producers’ prices are directly put into the diagonal sub-matrices Zrr and Yrr of Figure 8.1. 3. The IO sub-tables with imports from ROW countries measured in ex customs’ prices are directly put into the third country import submatrices Ztr and Ytr of Figure 8.1. 4. The IO sub-tables with imports from EU countries measured in ex customs’ prices Zrs and Yrs are disaggregated per row by the EU country of origin by means of import ratios calculated from the Eurostat import statistics, and are then put into an auxiliary intra-EU trade matrix M. 5. The internal and external EU-transit trade columns from the subtables mentioned under 2–4 are distributed over the appropriate sub-tables or are directly put into the vector ttt in Figure 8.1. 6. The columns with the exports to ROW countries measured in producers’ prices are directly put into the ROW export columns trt in Figure 8.1. 7. The columns with the exports to the EU countries measured in producers’ prices tre are rescaled to match the overall total of the auxiliary EU-internal trade matrix M. The difference of about 1 per cent is added as a rescaling column to Figure 8.1. 8. The auxiliary matrix M measured in ex customs’ prices is iteratively rescaled by means of RAS such that its row totals match the rescaled columns with the exports to EU countries measured in producers’ prices, and the result is put in the corresponding sub-matrices Zrs and Yrs of Figure 8.1. The two crucial steps are 4 and 8. Step 4 adds the lacking spatial origin to the intra-EU import matrices and step 8 re-prices the intra-EU import

192 The Evolution of the Core–Periphery Structure

matrices from ex customs’ prices in producers’ prices. This last step is needed for two reasons. First, it is needed to produce an internally price-consistent consolidated EU table to replace the present inconsistent consolidated EU tables. Second, it is needed to correctly allocate the impacts of any change in final demand to the sectors that really produce the imported intermediate inputs. In the absence of step 8, impacts on the trade and transport sectors would be grossly underestimated, whereas impacts on especially the primary and secondary sectors would be systematically overestimated. Given the construction method, the EU inter-country IO models used in the next sections will have a hybrid character. The diagonal submatrices Arr of the inter-country input coefficient matrix A contain the full information needed for a truly inter-regional IO model of the Isard type. The off-diagonal sub-matrices Ars, however, only contain the more limited information used in a multiregional IO model of the Chenery-Moses-Polenske type,5 as the Ars are estimated by dividing M, rs os os with mrs ij ¼ (ti /t i ) zij , column-wise by the total input of the purchasing rs sectors. The A should be labelled ‘limited information’, as the total import trade ratios (tirs/t os i ) are applied uniformly across each row of the ‘imports from other EU countries’ matrices Zos and Yos.

8.4 Earlier inter-country input–output research on EU integration Using the above data in current prices, and a derived set in constant prices, two dissertations from the University of Groningen have analysed European economic integration in depth (Linden, 1998; Hoen, 1999). Here we will summarize the main results of these and some related studies as far as relevant for the study of the fragmentation and spatial concentration of production and the relative growth of intra-industry trade and intra-regional trade. First, we will discuss the changes in internal and external interdependence of the EU economies and sectors. Next, we will discuss the extent and changes in intra-industry trade. Then, we will relate these changes to the changes in the sectoral specialization of countries and the spatial concentration of industries within the EU. 8.4.1

Growing intra-EU and extra-EU interdependence

To study inter-country and inter-industry interdependence several methodologies may be used. First and most straightforwardly, one may investigate the inter-country IO table itself to directly observe the size of the direct linkages between sectors and countries. This is a good way

Trade During Economic Integration 193

to discover the largest clusters of industry, but the analysis will be biased towards the absolutely large sectors and countries. Using a feedback loop analysis of the total intermediate transactions by the old five EU countries, Sonis et al. (1993) show that the first two – forward and backward running – domestic feedback loops take up 90 per cent of the EU total, whereas the following series of inter-country feedback loops take up the remaining 10 per cent. In 1970 the first two – again forward and backward running – inter-country feedback loops contain all five countries. In 1975 the first two inter-country loops only contain the four core old EU countries, whereas the much smaller two complement loops contain the three then new EU countries, the UK, Ireland and Denmark, and the more peripheral old EU country Italy. In 1980 all then included seven EU countries (Ireland did not have an IO table for 1980) are again contained in two integrated inter-country feedback loops, showing the rapid integration of the new member states. Next, one may look at the purchase shares in the columns and the sales shares in the rows of the IO table in order to correct for the absolute size of the sectors and countries at hand. Using such data for the early 1959– 1965–1970 years of the EU-6,6 Oosterhaven (1995) showed a considerable increase in the trade shares in total supply and demand, especially in the manufacturing sector, and a faster increase of internal EU trade than of external EU trade. This is followed by a partial reversal, with stronger extra-EU trade growth than intra-EU trade growth for the old EU-6 over the period 1970–5. This reversal is due to both the oil price hike of 1973 and the already mentioned extension into the EU-9, which both induced external trade growth as regards the old EU-6. Finally, one may want to consider not only the direct size of backward and forward linkages, but also their further propagation by including indirect effects. This implies an analysis of inter-country multipliers. For backward linkages this has to be done by inspecting the columns of the Leontief-inverse, L ¼ (I  A)1. For forward linkages this is now mostly done by inspecting the rows of the Ghosh-inverse, rs r G ¼ (I  B)1, with brs ij ¼ z ij /xi ¼ inter-country output coefficients. This last method, first used by Beyers (1976) and Jones (1976), is superior to the old method of looking at the row totals of the Leontief-inverse (Rasmussen, 1956; Hirschman, 1958). When using the Ghosh-inverse a word of caution is needed. The standard quantity interpretation of the underlying supply-driven IO model suffers from extreme implausibility; cars may drive without gasoline and factories may work without labour (Oosterhaven, 1988, 1996), whereas the alternative view interprets the supply-driven model not as a quantity

194 The Evolution of the Core–Periphery Structure

model but as the value version of the cost-push Leontief price model (Dietzenbacher, 1997). Using the Leontief model, Oosterhaven (1995) shows for 1959–75 for the EU-6 that the smaller countries, Belgium and the Netherlands, and the goods-producing sectors have larger intra-EU spillovers than the larger countries and the services sectors. Furthermore, the intra-EU spillover effects show a systematic increase over time, with the exception of the Dutch fuel and power industry that showed a strong increase in domestic backward effects due to its shift from coal to natural gas. The total EUwide backward linkages do not show a clear pattern. For some sectors in some countries they increase; for some others they decrease. Comparable results are found for the 1965–85 period, for the EU-9 (Linden and Oosterhaven, 1995). The interesting addition is that Denmark, less than half the economic size of Belgium and the Netherlands, thus definitely an economically small country, has backward domestic linkages, even increasing for most sectors, as well as backward intra-EU spillovers, that are comparable in size to those of the large EU countries. So, some small countries may have relatively closed economies, whereas gateway small countries may have very open economies. Dietzenbacher (2002), looking only at 1985, goes into far more sectoral detail in comparing forward with backward linkages. Forward linkages to processing industries show a larger variation between sectors (0.16–1.60) than backward linkages to supplying industries (0.25–1.22). Moreover, the clear distinction between goods-producing industries and service industries disappears when forward linkages are studied. Sectors that primarily produce final consumption goods, for instance, have small forward linkages and comprise both goods and services industries, such as food, clothing, office machines, transport equipment, building and several service sectors. Sectors that produce primary and intermediate goods, such as agriculture, minerals, energy, paper, basic metals, chemicals, transport services, communication and financial services, all have large forward linkages. The hypothetical extraction method (Strassert, 1968) offers a methodology to study inter-industry and inter-country interdependencies in a more flexible way, as it allows for the extraction of any combination of transactions. Moreover, the method almost calls for making a distinction between absolute effects in millions of euros and relative effects in percentages of the extracted transactions. Again a word a caution is needed. This approach assumes that the extraction of intermediate goods and services is fully compensated by increased imports from ROW countries when the demand-driven Leontief

Trade During Economic Integration 195

model is used, whereas it assumes that these transactions are fully compensated by increased exports to ROW countries when the supply-driven Ghosh model is used. This may not be realistic, especially in the latter case, while the two assumptions combined of course contradict one another. Extending this method to the inter-country level, Dietzenbacher et al. (1993) again show that the domestic effects, especially for the large countries and for Denmark, are far larger than the inter-country spillovers. However, despite the relatively large importance of the inter-country spillovers of Belgium and the Netherlands, the inter-country spillovers of extracting each of the four large countries are larger in absolute terms. Going down to the sectoral level Dietzenbacher and Linden (1997) observe large differences between sectors, but note that each sector behaves in remarkably similar ways in all of the EU countries. Moreover each sector mainly influences its own domestic output and only occasionally that of one or two other sectors at most. Linden (1998) only analyses inter-country spillovers, but in great detail. Taking the sum of the forward and backward linkages, he finds the following rank order of the inter-country linkages of individual countries with the rest of the EU: BE (Belgium) > NL (the Netherlands) >> DK (Denmark) > FR (France) DE (Germany) UK (United Kingdom)

IT (Italy). The rank order of the inter-country linkages of the rest of the EU with the individual countries naturally follows a more or less reverse pattern, namely: DE > FR > UK IT NL BE > DK. For the other countries’ output levels extracting Germany, already before reunification, has the largest effect. These results hold for most of the period 1965–85, with some changes. All inter-country spillover effects increase systematically, except for 1970–5, while after 1975 the UK spillovers grew much faster than those of the other countries. Looking at the difference between forward and backward linkages, the net forward linkages had the following country rank order in 1985: NL > UK > BE DE FR > DK IT. Given the role of the Netherlands and the UK in the production of oil and natural gas, their large net forward linkages are not too surprising. The same holds for Denmark and Italy, whose manufacturing sectors are strongly oriented towards producing consumer goods. The position of the UK again changed most, namely from a small net buyer’s position in 1975 to a large net seller’s position in 1985. Extracting EU sectors from the 1985 EU inter-country IO models, Linden (1998) finds the largest effect when extracting the Belgian motor vehicle industry, which has a strong assembly character. Its removal from the demand-driven model leads to an output decrease in the rest

196 The Evolution of the Core–Periphery Structure

of the EU that is 120 per cent of that of the removed Belgian industry. The first non-Belgian sector is the Dutch motor vehicle industry, whose removal has an inter-country fallout of 61 per cent of its own output. When removed from the supply-driven IO model, the Belgian and Dutch basic metals sector and the Belgian chemical sector produce the largest relative inter-country fallout of about 110 per cent. Using a creative application of the field of influence method to track productivity impacts on the Leontief-inverse (Sonis and Hewings, 1992), Linden et al. (2000) distinguish propulsive, reactive and dependent sectors and four different types of spatial effects. Their empirical conclusions confirm and add to the above findings. Finally, looking at an even more detailed level, Linden (1998) finds that the metal industries form the largest inter-country cluster of mutually connected industries, with the transport equipment industries and the machinery industries in the different member states forming two satellite clusters. Within this overall cluster, removing the German basic metal industry has the largest impact on the related industries in the other EU countries. The combined chemical industries in the different countries form the second independent inter-country cluster, while agriculture and food industries form the third independent EU cluster of interdependent industries. The largest change over 1965–85 is the disappearance of the once strong coal cluster and the emergence of an inter-country oil and gas cluster. Thus, inter-country trade mainly occurs at the intra-sectoral level, i.e. on the diagonals of the inter-country trade matrices of Figure 8.1. All these observations regarding individual sectors suggest that intraindustry trade will be the dominant form of trade within the EU, but it needs to be noted that inter-country interdependence in this section is exclusively defined on the trade in intermediate goods. The trade in consumption goods and in capital goods has not yet been taken into account. In the next section we will summarize research that studies total trade and see whether intra-industry trade indeed dominates internal EU trade or not. 8.4.2

Growing intra-industry trade within the EU

The standard measure of intra-industry trade (Grubel and Lloyd, 1971) runs from zero to unity and equals one minus the standard measure of interindustry trade (Balassa, 1966). Hoen (1999) applies this Grubel–Lloyd index to the cells of the EU inter-country IO tables in constant prices, as follows: rs sr rs sr GLrs iq ¼ 1  j ziq  ziq j=ðziq þ ziq Þ;

ð2Þ

Trade During Economic Integration 197

with z rs iq ¼ exports of sector i in country r to sector or final demand category q in country s. This application implies that different sectors and different final demand categories are considered to buy different products, in fact 725, as the EU inter-country sub-matrices have 25  29 (¼ 25 sectors 1 4 categories of final demand) cells. Aggregation over q, i.e. aggregation over every bi-country sub-row of Figure 8.1, produces the traditional bilateral GL-index. Aggregation over both q and s ¼ 6 r, i.e. aggregation over the off-diagonal part of the rows of Figure 8.1, gives the traditional aggregate GL-index per sector per country. At the cell level of the EU inter-country IO table Hoen (1999) finds weighted average values of his disaggregate GL-index of 0.33–0.57, per country, and unweighted average values of 0.27–0.34. This indicates that the smaller types of goods have less intra-industry trade, and that inter-industry trade is dominant at this level of disaggregation. The traditional bilateral GL-index has weighted values of 0.41–0.68, per country, and the traditional aggregate GL-index has weighted values of 0.48–0.81. Each aggregation, of course, results in seemingly more intra-industry trade, with the aggregate GL-index showing a clear domination of intra-industry trade over inter-industry trade. Over 1970–85 all indexes slightly increase, except for Italy and the Netherlands. The amount of intra-industry trade not only differs between countries, but also differs between the different industries of origin i. The GL-indexes for all manufacturing industries are larger than average and those for agriculture, energy and all service industries far smaller than average. So, we see increasing intra-EU trade and a dominance of inter-industry trade at lower levels of aggregation. Does this imply that EU countries have become more specialized in certain industries? Or does the increasing share of intra-industry trade over time imply that the sector structures of the EU countries have converged? And, directly related to these two questions, what does this imply for the spatial concentration of single industries within the EU? 8.4.3

Increasing sectoral specialization of EU countries

The IO studies of specialization and concentration mostly use the following indexes for regional sectoral specialization and sectoral spatial concentration, respectively: X r r eu eu  SIr ¼ and i j xi = x  xi = x j 50% X ð3Þ r eu r eu  CIi ¼ r j xi =xi  x =x j 50%

198 The Evolution of the Core–Periphery Structure

where xri ¼ output, export, import, domestic demand, etc. of industry i in country r and where a blank stands for an aggregation over i. Both indexes run from zero to 100 per cent, which indicates a complete regional specialization in a single industry and complete spatial concentration of a singe industry.7 Oosterhaven (1995) investigates the specialization of intra-EU exports and ROW exports for the old EU-6 for 1959–75. He finds very mixed results: increasing and decreasing specialization as well as changes in the sectors of specialization. Italy’s early specialization in 1959 in intraEU exports, for instance, relates to agriculture, whereas its early specialization in ROW exports relates to market services. After de-specializing its EU exports, Italy strongly re-specialized in textiles and leather in 1975. However, Oosterhaven did not find a clear relation between the specialization of output and that of export. Hoen (1999) extends this analysis with Denmark using the constant price IO data for 1970–85. He finds increasing specialization in intra-EU exports for all countries, except for France and Belgium, which shows that increasing intra-industry trade does not per definition mean that export packages also become more alike. Linden and Oosterhaven (2000) investigate this issue in depth for 1965–85. The specialization of output increased in most countries, especially for 1965–75, while the sectoral pattern of specialization remained fairly constant for most of the EU-9 countries. Compared to the EU total, Denmark and the Netherlands have the most specialized sector structure, but still their SIr for value added for 31 sectors does not rise above 18 per cent and 16 per cent, respectively. France with an SIr of only 6 per cent resembles the EU sector structure most. Classical trade theory suggests a relation between output specialization and export specialization. Porter (1990) and Krugman (1991), however, argue that domestic demand preferences may equally well induce output specialization, which may then induce export specialization, reversing the causality. Table 8.1 shows the statistical relations between different types of specialization and the changes therein. Import specialization has been left out, as it equals the difference between output specialization and domestic demand plus export demand specialization. Domestic demand specialization has been split up into productive demand for intermediate and capital goods by firms, and consumptive demand by households and government. Almost all correlations between the levels of specialization, in the lower left part of Table 8.1, are highly significant. The first column furthermore shows that the correlation between the extent of output specialization and that of intra-EU exports is only slightly larger (0.87)

Trade During Economic Integration 199 Table 8.1 Pearson correlations between levels and changes in specialization, EU5, 1965–85 Levels of (lower left)

Changes in (upper right) VA

VA ¼ specialization of value added EU ¼ specialization of intra-EU exports ROW ¼ specialization of ROW-exports C.DEM ¼ spec. of consumptive demand P.DEM ¼ spec. of productive demand

EU

ROW

.121

.092

.873**

.475*

.804**

.756**

.771**

.817**

.563**

.667**

.448*

.638**

C.DEM .338

P.DEM .529*

.263

.186

.019

.095 .231

.482**

Notes: * Significant at 5%. ** Significant at 1%. Both two-tailed. Source: Linden and Oosterhaven (2000).

than that with consumptive demand (0.77). In addition to this, the top row shows that the changes in output specialization are only related to the changes in productive demand. The direction of causality, however, is two-way. Due to input–output relations between sectors, output specialization directly and indirectly determines the sectoral specialization in intermediate goods. On the other hand, specialization in intermediate goods forms a part of the specialization in total output. Furthermore, the upper right part of Table 8.1 shows that changes in both types of export specialization move in the same direction, despite the fact that rather different markets and different relative factor endowments are involved. Finally, contrary to the evidence on the correlation between levels, changes in consumptive demand specialization are at least weakly correlated with changes in output specialization (0.338), whereas changes in export specialization are not (0.121 and 0.092). Hence, the relative importance of export specialization and consumptive demand specialization for output specialization will not differ much. The correlation between the aggregate ratios, however, does not necessarily imply that the sectoral pattern of specialization is similar too, but it is, as is shown by Linden and Oosterhaven (2000). Finally, in view of its prominence in trade theory we pay some more attention to the development of export specialization by adding Figure 8.2.8 First, note that the level of intra-EU export specialization, which ranges from a low 15 per cent for France to a rather high 45 per

200 The Evolution of the Core–Periphery Structure 50 45 40 35 30 25 20 15 10 5 0 Germany

France

Italy

Netherlands 1965

1975

1970

1980

1975

1985

Belgium

U. Kingdom

Denmark

Figure 8.2 Development of specialization ratios for internal EU exports, 1965–85 Source: Linden and Oosterhaven (2000).

cent for Denmark, is more than twice as large as the level of specialization of output, which ranges from 6 per cent to 18 per cent for the same two countries. As regards levels, Figure 8.2 only suggests a very weak relation between the economic size of a country and its intra-EU export specialization. The intra-EU export specialization of the Netherlands and especially of Denmark are larger than average, but for Belgium they are comparable to those of the large countries.9 Over time the extent of intra-EU export specialization has increased more or less systematically. Before 1975, it decreased in Germany and Belgium; after 1975 it only decreased in Denmark. Thus, the EU countries are not only trading more, but they are also trading more specialized packages of goods. Note that Figure 8.2 does not yet suggest a maturing of the export specialization growth, since its increase is at least as large after 1975 as it was before that. This obviously seems to be at odds with the increase in intra-industry trade noted in section 8.4.2. We will return to this paradox in our summary of the earlier empirical findings. The levels and changes in ROW export specialization are quite comparable to those for the intra-EU exports, as follows from their correlation in Table 8.1. Nevertheless, there are differences as shown in Table 8.2. For Italy, the UK and Denmark the intra-EU exports are more specialized.

Trade During Economic Integration 201 Table 8.2 ROW export specialization as a percentage of EU export specialization, 1965–85 Ratio gross of VAT

Germany France Italy Netherlands Belgium United Kingdom Denmark Unweighted average

Ratio net of VAT

1965

1970

1975

1975

1980

1985

88 103 76 132 137

83 115 81 115 115

72 81 64 127 130

102

99

87

82 94 60 123 120 76 69 84

105 93 69 116 141 72 73 89

104 119 69 116 128 93 79 97

Source: Linden and Oosterhaven (2000).

For the Benelux, the ROW exports are more specialized. This means that the ROW exports of the three small countries, as opposed to their intraEU exports, are more specialized than those of the four large countries, as expected. The development over time is again of special interest. Before 1975, the intra-EU specialization increased somewhat more than the ROW specialization. This holds for all five EU countries. This may reflect the effects of the early customs union, generating a stronger internal than external increase in trade and specialization because of the removal of the internal trade barriers. After 1975, this trend reversed and ROW export specialization increased more than intra-EU export specialization. This may reflect an increasing openness of the EU with respect to the rest of the world, particularly as part of the GATT agreements. 8.4.4

Increasing spatial concentration of EU industry

Spatial concentration is the opposite side of sectoral specialization. If a country is very specialized in the production of some commodities, the sectors that produce those commodities should also be spatially concentrated in the country at hand. This also follows from the formulas of the two indexes in (3). The specialization index (SIr), however, emphasizes small countries and large sectors, whereas the concentration index (CIi) emphasizes large countries and small sectors. In the EU, small sectors like coal, lignite and leather are spatially most concentrated with CIs of 36 per cent, followed by small sectors like beverages and tobacco with CIs of 24 per cent and 21 per cent (Linden and Oosterhaven, 2000). The far larger services sectors are quite evenly

202 The Evolution of the Core–Periphery Structure

spread out, with public services having the lowest CI of 4 per cent. The spatial concentration of coal, leather and textiles grew in all five-year periods over 1965–85. The strongest growth in spatial concentration of industry occurred over the period 1965–75, followed by a more even spreading of several sectors over 1975–80, after which the spatial concentration of most industries picked up again over 1980–5. The pattern of spatial concentration did not change much. The production of coal is strongly concentrated in Germany and the UK, whereas demand for coal and export of coal is concentrated in Germany only. The production of basic metals, motor vehicles and machines is also concentrated in Germany; textiles, leather and wood are concentrated in Italy; oil, beverages, tobacco and other manufacturing in the UK. In general, the spatial concentration of production follows the spatial concentration of both domestic demand and foreign exports. The spatial concentration of exports, however, is far larger than that of either domestic demand or production.

8.4.5

Conclusions from earlier research

From the above overview of earlier studies, the following empirical conclusions may be drawn: 1. The specialization of intra-EU trade and ROW trade of the EU countries is strongly related, and both have grown at the cost of domestic transactions, with ROW trade growing faster than intra-EU trade after 1970. 2. Specialization of domestic demand and specialization of exports are both strongly related to the specialization of output. Changes in the specialization of output, however, are only related to changes in the specialization of productive domestic demand by firms. 3. The EU bilateral trade in intermediate goods predominantly connects the same sector, which seems to imply that intra-industry trade in intermediate goods may need a separate theoretical foundation from that in consumption goods (or capital goods). 4. Increasing export specialization and increasing spatial concentration of exporting industries occur along with increasing intra-industry trade. This seeming paradox may only be reconciled if import specialization is growing in the same commodities. This has not been studied explicitly, but when it appears to be confined to intermediate demand it offers a clear indication of outsourcing within the same industries.

Trade During Economic Integration 203

The conclusions 2-4 emphasize the importance of making a distinction between intermediate, consumption and capital goods when analysing the effects of economic integration on trade, as will be done in the next section.

8.5

Trends in European trade, 1975–95

The full inter-country IO tables for the years 1991 and 1995 from Policy Research in Antwerp (www.policyresearch.be) allowed us to extend the analysis of EU trade patterns in time. We chose to analyse a period of twenty years (1975–1995) such that it partly overlaps with the earlier analyses. In this period, Europe experienced a further integration in a political and economic sense. This section presents extensive results on changes in economic openness of EU countries and EU sectors in sections 8.5.1–8.5.4, while section 8.5.5 focuses on the question of whether these changes relate to inter-industry trade or intra-industry trade. 8.5.1 Changes in economic openness in intermediate goods and services First, we look at the trends in the origin of inputs used by the 25 industries that are evident in the tables for 1991 and 1995.10 Table 8.3 presents the results.11 They can be viewed as representing weighted averages of sectoral results, with sectors using large volumes of intermediate inputs receiving a larger weight than sectors using relatively few intermediate inputs. Table 8.3 clearly shows that the big economies (DE, FR, IT and UK) rely much more on domestically produced inputs than the small economies (NL and BE), which does not come as a surprise. The intermediate position of the economically smallest country (DK) again comes as a surprise. The economies that benefit from the relatively large share of intermediate inputs imported by the small countries, however, are not only the other EU-7 economies but also the ROW, as the smaller countries import larger shares of their inputs from third countries. All countries, except Italy, show a bell-shaped trend with regard to the share of inputs originating from the rest of the EU-7 and from the ROW. In most (but not all) countries the maximum share is found for 1980 or 1985, after which the import shares started to decline. In most countries the domestically produced inputs are becoming more important again, which is a reversal of the clear trend noted in section 8.4. This may be the result of less intensive trade across-the-board, but it may also be the result of an increased role of the services sectors in the economy

204 The Evolution of the Core–Periphery Structure Table 8.3 Origin of total intermediate inputs (%)* 1975

1980

1985

1991

1995

DE

Domestic Other six EU Third countries

85.6 6.0 8.5

81.4 7.4 11.2

82.6 7.5 9.9

81.3 7.9 10.8

88.8 4.6 6.6

FR

Domestic Other six EU Third countries

82.6 7.6 9.8

79.4 08.4 12.2

80.2 8.7 11.2

79.7 8.8 11.4

88.1 5.7 6.2

IT

Domestic Other six EU Third countries

80.7 6.4 12.9

78.5 7.5 14.0

80.4 7.2 12.4

81.2 7.0 11.7

79.9 7.8 12.3

NL

Domestic Other six EU Third countries

63.4 18.2 18.4

61.0 16.7 22.2

66.0 13.7 20.3

69.4 15.3 15.3

77.0 10.5 12.4

BE

Domestic Other six EU Third countries

57.0 26.3 16.7

52.2 26.1 21.7

47.8 33.7 18.5

49.9 29.8 20.3

51.0 29.1 19.9

UK

Domestic Other six EU Third countries

81.7 5.6 12.7

83.0 5.8 11.2

81.5 6.9 11.6

85.6 5.8 8.7

87.4 5.1 7.5

DK

Domestic Other six EU Third countries

73.2 11.1 15.7

71.9 12.0 16.1

72.7 12.2 15.1

75.7 11.1 13.3

77.2 8.8 14.0

Notes: *The country classification is as follows: DE ¼ (west) Germany, FR ¼ France, IT ¼ Italy, NL ¼ the Netherlands, BE ¼ Belgium, UK ¼ United Kingdom, DK ¼ Denmark. Source: Own calculations.

(and thus also in international trade, e.g. in the form of increasing imports and exports of producers’ services). 8.5.2

Sector-specific changes in economic openness

To investigate this hypothesis we next look at the level and the changes in the import ratios at the level of individual industries. To attain consistency in sector classifications over the IO tables for 1975, 1980, 1985, 1991 and 1995, we had to use a 25-sector classification. It contains two primary sectors, thirteen manufacturing sectors and ten services sectors. All results presented should be considered as weighted averages over the seven countries included, i.e. weighted by the sectoral gross output levels of each country.

Trade During Economic Integration 205 Table 8.4 Average self-sufficiency ratios (%), 1975–95, by sector* 1 2 3 4 5 6 7

AGRI ENER BMET MINR CHEM METP MACH

93 90 82 91 75 92 82

8 9 10 11 12 13

OFFM ELEC TREQ FOOD TEXT PAPE

82 83 80 89 84 93

14 15 16 17 18 19

RUBB OTHM BUIL TRAD LODG INTR

86 93 100 90 97 94

20 21 22 23 24 25

MATR AUTR COMM CRED MARS PUBS

69 92 99 99 98 98

Notes: *The sector classification in all tables comprises an aggregation of the R44 NACE-CLIO sector division of Eurostat, with: AGRI ¼ agriculture, forestry, fishery (1), ENER ¼ fuel and power (2–6), BMET ¼ ferrous and non-ferrous ores and metals (7), MINR ¼ non-metallic minerals (8), CHEM ¼ chemicals (9), METP ¼ metal products, except machinery and transport eq. (10), MACH ¼ agricultural and industrial machinery (11), OFFM ¼ office and data processing machines (12), ELEC ¼ electrical goods (13), TREQ ¼ transport equipment (14–15), FOOD ¼ food, beverages, tobacco (16–20), TEXT ¼ textiles, clothing, leather, footwear (21–22), PAPE ¼ paper, printing (24), RUBB ¼ rubber, plastic (25), OTHM ¼ other manufacturing (23, 26), BUIL ¼ building, construction (27), TRAD ¼ recovery, repair, wholesale, retail (28–29), LODG ¼ lodging, catering (30), INTR ¼ inland transport (31), MATR ¼ maritime and air transport (32), AUTR ¼ auxilary transport services (33), COMM ¼ communication (34), CRED ¼ credit, insurance (35), MARS ¼ other market services (36–37), PUBS ¼ non-market services (38–40). Source: Own calculations.

First, we look at the level of self-sufficiency by sector. The unweighted averages of the self-sufficiency ratios over the five years of analysis are shown in Table 8.4. The lowest value, 69 per cent, is (not surprisingly) found for maritime and air transport (20). The next lowest value is found for chemicals (5), a manufacturing industry. The non-transport services sectors all have very high levels of self-sufficiency, as expected. Although Table 8.4 provides useful, though unsurprising information on the level of economic openness by sector, it is more interesting to look at the changes over time. This is done in Table 8.5, which shows the growth rates of self-sufficiency levels over two ten-year periods. The upper left cell indicates that self-sufficiency in agriculture (1) decreased by 3.1 per cent from the average base of 93 per cent indicated in Table 8.4. Table 8.5 shows that the sectoral changes are quite different per period. In the early period (1975–85) most sectors showed decreasing levels of self-sufficiency as reported in section 8.4. This is particularly true for the primary and manufacturing sectors. The strongest decline is found for office machinery (8), which is a sector with a few multinational firms dominating global markets. A large decline is also found for chemicals (5), a sector which already had a relatively low level of self-sufficiency to start with. In many service sectors, the rates of decline were rather

206 The Evolution of the Core–Periphery Structure Table 8.5 Growth of self-sufficiency ratios (%), by sector* 1975–85 1985–95 1 2 3 4 5 6 7 8

AGRI ENER BMET MINR CHEM METP MACH OFFM 9 ELEC

3.1 7.6 8.4 2.2 9.2 0.6 4.4 11.1 1.1

0.6 8.2 0.1 1.9 5.5 0.9 4.6 13.7 2.3

1975–85 1985–95 10 11 12 13 14 15 16 17

TREQ FOOD TEXT PAPE RUBB OTHM BUIL TRAD

6.4 2.4 2.9 1.8 3.6 0.9 1.0 4.1

3.9 0.2 2.0 0.7 0.3 1.0 0.0 3.3

1975–85 1985–95 18 19 20 21 22 23 24 25

LODG INTR MATR AUTR COMM CRED MARS PUBS

2.4 4.1 30.3 0.1 0.5 0.3 0.2 2.6

0.5 0.7 9.9 0.9 0.0 0.1 1.4 1.2

Note: *See Table 8.4 for the sector classification. Source: Own calculations.

low, indicating that the notion of services being ‘untradable goods’ suggested by Table 8.4 is a relatively stable phenomenon. In the second period (1985–95), the results show a completely different pattern. In many sectors, positive growth rates of self-sufficiency ratios are found, which point towards a structural decrease in economic openness also at the sectoral level. The largest positive growth rates are recorded for sectors that experienced prominent negative growth rates in the early periods, such as energy (2), office machinery (8) and chemicals (5). Hence, the conclusion is that the reversal of the long-term EU trend towards greater economic openness is not caused by a change in the sector structure of EU imports, as this reversal is also found for the majority of the individual sectors. How could this reversal of signs come about? One explanation may be that we do not yet differentiate between ‘exports to the rest of the EU’ and ‘exports to third countries’. It might well be that exports to third countries started to decrease, for example because of increasing FDI (which allows firms to produce geographically close to consumer markets, in many countries at relatively low costs). Hence, the positive growth rates in the columns for 1985–95 need not be seen as evidence of an economic closure of the EU-7. Anyhow, it is not likely that the reversal of the trend towards more openness of the EU has caused the increased importance of service industries and service industries’ trade. This is confirmed if we repeat the analysis of Table 8.3 for tradable intermediate inputs only. We define tradable inputs in a very traditional sense, that is, as the output of two primary industries (1–2) and thirteen manufacturing industries (3–15). The results are shown in Table 8.6. Broadly speaking, Table 8.6 shows a similar picture to Table 8.3 as regards the differences between countries and the changes over time.

Trade During Economic Integration 207 Table 8.6 Origins of tradable intermediate inputs (%), 1975–95* 1975

1980

1985

1991

1995

DE

Domestic Other six EU Third countries

83.4 6.2 10.4

76.5 9.0 14.5

75.0 10.6 14.4

71.8 11.6 16.6

79.2 7.9 12.9

FR

Domestic Other six EU Third countries

76.9 9.6 13.4

71.8 11.0 17.2

70.6 12.8 16.7

69.9 12.9 17.2

79.6 9.4 11.0

IT

Domestic Other six EU Third countries

77.4 6.5 16.2

73.2 8.8 18.0

73.7 9.6 16.7

75.0 8.9 16.1

74.3 8.9 16.8

NL

Domestic Other six EU Third countries

54.3 22.0 23.6

49.6 20.9 29.5

55.1 19.9 25.0

56.7 24.7 18.6

56.8 22.9 20.3

BE

Domestic Other six EU Third countries

51.1 26.6 22.2

43.7 28.0 28.3

38.4 38.1 23.5

38.0 34.9 27.0

41.2 30.1 28.7

UK

Domestic Other six EU Third countries

76.6 6.3 17.1

77.6 7.4 15.0

75.4 8.7 15.9

79.3 7.7 13.0

74.4 9.3 16.2

DK

Domestic Other six EU Third countries

59.7 16.9 23.4

58.8 17.4 23.9

59.2 18.6 22.2

60.5 18.2 21.3

61.4 15.9 22.6

Note: *See Table 8.3 for the country classification. Source: Own calculations.

Concerning the latter, though, the maximum values of the bell-shaped patterns in the share of inputs bought from other EU-7 countries are mostly found for the more recent years (that is 1985 and 1991) if only tradables are taken into account. Structural change in intermediate input mixes thus seems to play a role in the explanation of the bellshaped patterns, but it cannot be held responsible for this feature. Another interesting difference between Tables 8.3 and 8.6 is that the share of imports from third countries clearly rose between 1991 and 1995, if only tradable inputs are considered. In this case, only France and Germany deviate from this stylized fact, whereas such a pattern is not visible if all intermediate inputs are considered. At least two developments in the international arena might be responsible for the increase in

208 The Evolution of the Core–Periphery Structure

third-country imports in recent years. First, the EU itself grew, after countries like Greece in 1981, Spain and Portugal in 1986, and Sweden, Finland and Austria in 1995, joined the EU. In this long-term study, these countries are considered ‘third countries’. If they integrated quickly before and after their accession, this might be reflected in the increasing shares of tradable inputs bought from third countries. Second, integration is not limited to the EU. Trade liberalization is an almost worldwide phenomenon that crosses continental borders. Consequently, more and more inputs might be bought from countries outside the EU, such as the Asian Tigers, who managed to produce high-quality products at relatively low cost. Looking only at Table 8.6, we find that intra-EU trade in general grew faster than domestic trade and trade with the ROW (if the input side is considered) in the early years of our analysis. Later on, trade with third countries grew faster. It should be borne in mind, however, that the patterns shown by the countries included in our analysis are not very general. Big economies like France and Germany show trends that are different from those found for other countries. Another interesting result reflected in Table 8.6 is the UK’s reliance on third-country imports, which is much stronger than for the other big countries. Apparently, the UK still maintains its relative focus on non-EU trade partners, when imports are considered. 8.5.3 Changes in economic openness regarding consumption goods In Table 8.7, we turn our attention away from the origin of intermediate inputs, towards the origin of consumption goods.12 For brevity and comparability, we only consider tradable consumption goods. If we compare the values for the shares of domestically produced consumption goods in Table 8.7 with the comparable shares for intermediate inputs in Table 8.6, we find that consumption goods were more prone to being bought in the country itself. In 1975, Germany is the only exception to this rule. In 1995, this difference is much less clear-cut. In this year, the domestic shares for consumption goods are lower than those for intermediate inputs in Germany, the Netherlands and the UK. Interestingly, but not surprisingly, the levels of the shares of goods bought from EU-7 suppliers compared to the shares bought from third countries are much larger for the old EU members (DE, FR, IT, NL and BE) than for its newer members, Denmark and again especially the UK. Table 8.7 furthermore reveals that the bell-shaped time patterns observed for the import shares for intermediate inputs are much less paramount for consumption goods. For two countries, Germany and

Trade During Economic Integration 209 Table 8.7 Origins of tradable consumption goods (%), 1975–95* 1975

1980

1985

1991

1995

DE

Domestic Other six EU Third countries

82.2 10.7 7.0

80.0 9.4 10.6

78.5 10.5 11.0

75.8 10.1 14.1

74.6 10.5 14.9

FR

Domestic Other six EU Third countries

87.6 7.2 5.2

81.1 9.3 9.6

82.2 9.7 8.0

81.3 8.1 10.5

80.2 8.2 11.6

IT

Domestic Other six EU Third countries

91.4 5.8 2.8

87.6 7.1 5.3

86.2 7.8 6.0

86.8 6.6 6.6

80.8 8.5 10.7

NL

Domestic Other six EU Third countries

65.8 25.1 09.2

63.4 21.1 15.5

60.5 24.0 15.6

57.2 27.0 15.8

56.1 27.5 16.4

BE

Domestic Other six EU Third countries

65.5 27.5 7.0

56.6 32.2 11.2

53.6 34.7 11.7

53.2 31.1 15.7

55.5 29.1 15.4

UK

Domestic Other six EU Third countries

81.5 7.3 11.1

79.0 7.4 13.6

65.3 22.2 12.5

66.9 15.0 18.1

69.8 9.0 21.2

DK

Domestic Other six EU Third countries

69.6 15.6 14.8

67.3 15.7 17.0

61.0 19.4 19.6

61.7 17.9 20.3

64.2 16.1 19.8

Note: *See Table 8.3 for the country classification. Source: Own calculations.

the Netherlands, the domestic share was even monotonically decreasing and for two other countries, France and Italy, a downward trend in selfsufficiency is obvious too. The bell-shape is only found for Belgium, the UK and Denmark, but the domestic shares in 1995 were still much lower than those for 1975. Hence, the reversal of the long-run increase in economic openness only applies for intermediate inputs. In all countries considered together, over time, consumers especially tended to buy more from third countries. This seems to be a manifestation of economic integration at a scale beyond the seven early EU countries included in this study. This effect was particularly strong for Italy, the country of which the citizens tended to spend more than 90 per cent of their consumption expenditures on domestic goods in 1975, but which

210 The Evolution of the Core–Periphery Structure

saw the share of third countries increase from less than 3 per cent to almost 11 per cent in 1995. For all other EU-7 countries, the growth of the share of third countries was less spectacular but equally clear. The net effect of generally decreasing domestic shares and increasing shares of third countries is a rather heterogeneous set of results for the imports of consumption goods from the rest of the EU-7. In Germany, for example, this share remained very stable over the twenty years studied. A similar result (with a little bit more, apparently unsystematic, variation) is found for the Netherlands and Italy. The other four countries show strong increases in imports from the EU-7 in the earlier years, followed by slight decreases in the later years. All in all, we do not find a clear difference between tendencies for large and for small countries, except for the much larger economic openness of the small countries, which in the case of consumption goods this time also holds for Denmark. 8.5.4

Changes in economic openness regarding capital goods

Table 8.8 repeats the analysis of Tables 8.6 and 8.7 for capital goods.13 A striking difference concerns the shares of domestic suppliers in the small countries, NL, BE and DK. These countries generally produced more than 50 per cent of their own intermediate inputs and consumption goods (Belgium is an exception in the case of intermediate inputs), whereas these shares are considerably below this value in the case of capital goods. For the big countries, the domestic shares for capital goods are also somewhat lower than for consumption goods. This does not hold for Germany, however, which underlines its international position as a producer of investments goods. The domestic shares for the three categories of goods distinguished do not differ very much for Germany. When the shares of third-country imports are compared with those from the rest of the EU-7, interesting differences between intermediate, consumption and capital goods may be observed. First, note that the shares of the rest of the EU-7 are generally either absolutely or relatively much smaller for the four larger countries than for the three smaller ones, but this is at least partly due to the fact that the rest of the EU-7 is smaller for the large countries than it is for the smaller countries. Second, note that in 1975, except for the UK, imports from the rest of the EU-7 are larger than imports from the ROW for both consumption goods and capital goods, whereas for imports of intermediate goods the reverse holds (except for Belgium). For imports of intermediate goods this size relation remains fairly constant over time, except for the Netherlands that started to import more

Trade During Economic Integration 211 Table 8.8 Origins of tradable capital goods (%), 1975–95* 1975

1980

1985

1991

1995

DE

Domestic Other six EU Third countries

81.6 9.9 8.5

82.2 8.5 9.4

80.9 8.0 11.1

78.1 8.7 13.2

76.6 8.3 15.1

FR

Domestic Other six EU Third countries

76.0 16.1 7.9

71.6 17.5 10.9

66.9 17.8 15.3

65.3 16.7 18.0

71.2 13.6 15.2

IT

Domestic Other six EU Third countries

73.6 17.5 8.9

69.6 19.1 11.3

78.1 12.2 9.7

81.8 10.9 7.3

68.7 15.7 15.6

NL

Domestic Other six EU Third countries

39.8 43.7 16.5

37.0 40.9 22.1

41.5 32.7 25.8

36.8 43.0 20.3

31.8 40.6 27.5

BE

Domestic Other six EU Third countries

49.2 41.0 09.8

38.2 43.6 18.1

33.2 41.1 25.6

34.9 46.6 18.5

37.2 42.7 20.2

UK

Domestic Other six EU Third countries

67.8 14.9 17.2

69.2 14.5 16.3

54.2 20.3 25.4

51.5 20.6 27.9

44.6 22.7 32.7

DK

Domestic Other six EU Third countries

35.8 37.1 27.1

49.2 27.7 23.1

41.9 30.5 27.6

45.2 31.1 23.6

45.1 26.3 28.7

Note: *See Table 8.3 for the country classification. Source: Own calculations.

intermediates from the rest of the EU-7 than from the ROW. For both consumption goods and capital goods this size relation changes over time. The dominance of imports of consumption and capital goods from the rest of the EU-7 over those from the ROW decreased for all countries for most of the periods distinguished. In most of the cases, especially for the larger countries of course, the imports from the ROW actually became larger than those from the rest of the EU-7. For most countries, the domestic shares in Table 8.8 appear to decline, especially for the UK, although there is not a single country for which this downward trend is monotonic. Again, the bell-shaped pattern that was found for intermediate inputs (i.e. initially declining domestic shares followed by increasing domestic shares) is not apparent for capital goods. It is only found for Belgium and might emerge in 1995 in France,

212 The Evolution of the Core–Periphery Structure

whereas the Netherlands even shows an inversed bell-shape. Denmark is an exception to the general pattern of declining self-sufficiency ratios, in the sense that it appears to have increased its share of home-produced capital goods, exclusively at the expense of the rest of the EU-7. Although the patterns of the shares of third countries in the provision of capital goods are rather erratic, we find that for most countries they tended to increase between 1975 and 1995. The exception to this stylized fact is Denmark, for which the share of third countries remained almost unchanged. We do not find remarkable patterns for the shares of capital goods delivered by the rest of the EU-7. We find a clear downward trend only for Denmark and for the UK we find a clear upward trend. Apparently, capital goods from Germany, France and the other countries included increasingly acted as substitutes for capital goods produced in the UK itself. 8.5.5

The relative importance of intra-industry trade by sector

In section 8.4.2, we mentioned that previous research into the nature of trade within the EU showed an increasing importance of intra-industry trade, relative to inter-industry trade. The aim of this and of the next section is to investigate whether or not this tendency has continued in more recent stages of EU integration. To this end, we use the Grubel– Lloyd (GL) indexes summarized in section 8.4.2 and described more extensively in Hoen (1999). A GL-index close to its maximum value of one indicates that almost all trade is of an intra-industry nature, whereas a value close to zero reflects a situation in which inter-industry trade is dominant. First, we present empirical results that illustrate to what extent GLindexes vary across industries. To this end, we computed the most detailed cell-level GL-indexes according to equation (2) and summed them over the inter-country parts of the rows of the IO tables, with weights determined by the magnitudes of the respective cells with intermediate inputs.14 Figure 8.3 clearly shows that the degree to which intra-industry trade is present is a very industry-specific issue. In the manufacturing sectors (3)–(7) and (8)–(14), the highest GL-indexes are found, with values up to about 0.7. For energy (2) and the services sectors (17)–(25) the indexes are generally much smaller. Intermediate values are found for agriculture (1), office machines (8) and other manufacturing (15), whereas the lowest values are found for building (16). These results corroborate Hoen’s (1999) findings. Hoen furthermore found that most industry-specific GL-indexes grew between 1970 and 1985, in particular in the manufacturing industries.

Trade During Economic Integration 213 0.8

0.6

0.4

0.2

0.0 1

2

3

4

5

6

7

8

9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 1975

1985

1995

Figure 8.3 Weighted averages of IO cell-level Grubel-Lloyd indexes by industry of origin* Note: *See Table 8.4 for the sector classification. Source: own calculations.

Figure 8.3 for 1975–95 sketches a more heterogeneous picture. For some industries, intra-industry became increasingly important. The most outstanding examples are office machinery (8), transport equipment (10), food (11) and paper and printing (13). For other industries, however, the importance of intra-industry trade decreased. We found considerable decreases for chemicals (5), electrical machinery (9) and rubber and plastics (14). In services, the GL-indexes vary quite a bit over time. In many cases, this is probably due to the low volumes of trade for services. Interesting results that might well reflect substantive increases in intra-industry trade are found for communication (22), financial services (23) and other market services (24). 8.5.6 The relative importance of intra-industry trade by country Next, we present the results of the cell-level Grubel–Lloyd indexes aggregated over all sectors and all countries of destination, i.e. aggregated by country of origin. Following Hoen (1999), we present two types of averages. The weighted averages attach more importance to levels or tendencies in large industries (i.e. with high trade volumes), whereas the arithmetic averages put more emphasis on levels or changes in relatively small industries. The results are shown in Table 8.9. When considering the weighted averages, France had the most intraindustry trade, although the differences with Germany, Belgium and the UK were very small. As Hoen (1999) already noted, Denmark did

214 The Evolution of the Core–Periphery Structure Table 8.9 Averages of IO cell-level Grubel-Lloyd indexes by country of origin (%)* Weighted averages

1975 1980 1985 1991 1995

DE

FR

IT

NL

BE

UK

DK

53.2 53.9 54.2 58.5 51.8

54.7 56.2 54.6 57.7 46.3

48.9 49.5 49.5 50.4 41.8

47.8 51.0 44.7 52.4 48.7

52.1 54.4 52.5 53.6 46.1

51.7 54.1 48.7 47.0 44.6

30.5 36.4 40.7 44.7 43.8

Arithmetic averages

1975 1980 1985 1991 1995

DE

FR

IT

NL

BE

UK

DK

28.7 33.4 32.3 34.4 33.6

21.6 24.7 22.9 23.5 31.8

21.1 24.2 23.2 23.4 23.0

23.6 28.7 23.4 24.7 25.5

27.7 30.6 29.9 32.3 32.3

24.2 25.2 25.8 27.3 30.2

19.1 24.3 25.3 27.3 26.3

Note: *See Table 8.3 for the country classification. Source: Own calculations.

not have much intra-industry trade in 1975 and 1980. Later on, this started to change. Actually, Denmark is the only country among the seven studied that shows a clear upward trend in the weighted averages of the GL-indexes. Most countries show the increases in the early periods, as already noted, but these are followed by marked decreases in intra-industry trade, especially in the last period. Comparison of the upper and lower panels of Table 8.9 reveals some interesting differences. As mentioned above, the lower panel emphasizes the GL-indexes of small industries in terms of trade volumes. The clearly lower values for the arithmetic averages illustrate that small industries generally show more inter-industry trade than intra-industry trade. Throughout the entire period, the highest arithmetic averages are found for Germany, followed by Belgium. The differences among the other countries are relatively small, with the exception of Denmark in the early phase. For this country, the strongly rising weighted averages do carry over to the arithmetic averages. Apparently, the increasing importance of intra-industry trade was not only confined to larger industries, but was an economy-wide Danish phenomenon. Looking at the development over time, Denmark is now joined by four other countries (DE, FR, BE and the UK), which also have rising intraindustry trade. Only for Italy and the Netherlands, do we find more or

Trade During Economic Integration 215

less stable unweighted average GL-indexes. Although our explanation of the differences between the weighted and unweighted averages is somewhat speculative, it is possible that the strongly increasing importance of intra-industry trade in some market services industries with little trade (see Figure 8.3) is responsible for this result.

8.6

Summary and conclusion

We have given an overview of both theory and empirical studies that relate the spatial concentration of industry (location theory) to the specialization of domestic demand (home market effect), the specialization of exports (inter-industry trade) and the growth of intra-industry trade (new trade theory). Most of the empirical studies are based on the inter-country input–output tables for the continuously extending European Union over the period 1959–85. The paradox of increasing specialization of exports and increasing intra-industry trade has to be solved by a growing specialization of imports in the same direction, which is an indication of increased outsourcing. The dominance of intra-industry trade in intermediates seems to confirm this. We have extended the overview with an empirical analysis of the intercountry EU tables for 1975–95, distinguishing intermediates from consumption and capital goods. The smaller EU countries have larger import shares and export packages that are a little more specialized than the larger countries. This indicates that, potentially, they may profit more from economic integration than the larger countries. Upon considering trade in intermediate inputs, we found that the long-term decreasing shares of domestic suppliers started to increase after 1985. The shares of third countries also increased, at the expense of inputs produced in the remaining six EU countries. We found this phenomenon for all countries, irrespective of their size. As long as we considered trade in all intermediate inputs, some of these patterns remained somewhat hidden, but they were clearly revealed upon considering trade in ‘tradable’ (agricultural and manufacturing products) intermediate inputs only. Interestingly, the results for consumption goods and capital goods did not yield the above-described bell-shaped relationship between time and shares of domestically produced goods. These shares almost monotonically decreased for consumption goods and showed a less regular, but also clear downward tendency for capital goods. The other EU-7 countries have not been the main ‘winners’ that gained shares. As with intermediate inputs, imports of consumption and capital goods from third countries also increased. This might well be due to

216 The Evolution of the Core–Periphery Structure

ongoing European integration, as an increasing number of countries have joined the European Union. Since we aimed at analyses regarding a relatively long period, our data allowed us to consider only seven of the oldest EU member countries. Consequently, if trade between France and Spain had intensified as a consequence of European integration, this would have led to more imports to France from a third country in our analysis. Moreover, an economic integration process on an even wider scale has been at work already for more than a decade. Due to decreased transportation cost, easier communication and policies towards international openness, Europe is part of an almost worldwide integration process. It is certain that this continuing global integration has affected trade patterns in our period of study. We found that intra-industry trade neither became much more important across industries, nor did we find an economy-wide decrease in importance vis-a`-vis inter-industry trade. Instead, some manufacturing industries showed marked increases, while others were characterized by reductions in their Grubel–Lloyd indexes. Probably, the most interesting result in this respect is that we found a substantial increase in intraindustry trade in market services industries, such as in communication and financial services. Upon considering countries, we concluded that most of them showed a decrease of intra-industry trade importance, when large trading industries are assigned large weights. This pattern is reversed for most countries if smaller and larger trading industries are given equal weights. The levels of importance of intra-industry trade did not vary strongly across countries, with the exception of Denmark in the first half of the period we considered. What insights can be gained concerning the ongoing integration processes in East Asia? Of course, many caveats apply when attempting to compare European integration in the (recent) past and the East Asian integration that is currently in full swing. Nevertheless, we will discuss some potentially important issues. First, we did not find that big countries benefited more from integration than small countries, or the other way round. Big countries produce more of their own intermediate inputs, consumption goods and capital goods than small countries do, but the European integration process did not lead to fundamentally changing relations between small and big countries. We do not have a priori reasons to believe that this will be different for East Asia, although at least one difference between East Asia and Europe at the start of its de jure integration should be considered. While the differences in terms of income per capita were already small in Europe in the 1950s, these are currently very big in East Asia. Intensifying

Trade During Economic Integration 217

integration is likely to yield convergence of income levels in East Asia. Consequently, trade among the countries involved will increasingly be of the intra-industry type. While no very marked economy-wide tendencies in Grubel–Lloyd indexes were found for Europe, this might turn out differently for East Asia. A shift from inter-industry-dominated trade to a more even representation of intra-industry and inter-industry trade would of course have implications for specialization patterns of individual countries. Second, we found that trade with third countries increased considerably over time. If this result were translated to the East Asian situation, it might mean that trade between countries that were early in integrating could be diverted to countries that join the integration process later. Another issue that should be taken into account is the continuing global integration that is simultaneously taking place with the more regionally oriented integration processes. The success of these regional processes is partly dependent on developments beyond the direct influence of parties involved in regional integration. Nevertheless, it is important to repeat a crucial difference between Europe and East Asia mentioned by Richard Baldwin in Chapter 3 of this volume: the European integration process has been characterized by strong supranational bodies (the European Union and its predecessors) and WTO discipline, while the East Asian ‘noodle bowl’ of trade agreements is much more fragile. Increasing de jure integration would enhance the power of the East Asian region, and increase opportunities for playing an important role in affecting global integration in favourable directions. Our results for Europe indicated that both intra-industry and interindustry trade remained important during the integration process. Since our industry analysis yielded a heterogeneous pattern (with some industries becoming more dependent on intra-industry trade, while for other industries the opposite was found), we feel that both types of trade should get sufficient attention in decision-making processes concerning further East Asian integration. It would be interesting to see whether or not the countries involved in East Asian integration have more or less comparable Grubel–Lloyd indexes. In the sample of European countries, the only country (Denmark) that relied much more on inter-industry trade than the other countries in the early stages of integration quickly caught up regarding the relative importance of intra-industry trade. This might happen in Asia, too. In our view, increasing intra-industry trade in market services industries requires special attention. We already found some clear tendencies in this respect, even though our data for Europe were not more recent than 1995.

218 The Evolution of the Core–Periphery Structure

In the meantime, the tradability of these services seems to have increased even faster, mainly due to technological progress in information and communication technology.

Notes 1. See, for example, Figure 1.1 in Chapter 1 of this volume, which indicates that income per capita differences between Japan and Laos exceed a factor of 100. According to the Penn World Tables (Heston et al., 2006), for the seven European countries this factor was less than 2.0 in 1950 and just 1.2 in 2004. 2. Trade diversion was mainly found for the small Benelux countries that had lower individual tariffs before 1958, and for chemicals, agriculture and agriculture-related manufacturing that got high common external tariffs after 1958 (see Linden, 1998: section 4.3, for an overview). The differences in the estimates are explained by using both different databases and different methodologies. Mayes (1978) showed that applying different methodologies to one and the same database produces a comparable range of effects, namely 9 to 21 billion for trade creation and 4 to 11.1 billion for trade diversion; this notwithstanding that using the same methodology on different data sets produces quite different individual outcomes. 3. One of the important findings of these ‘zero geography’ models is that history does matter (path dependency). Stelder (2005), however, shows that the opposite approach of ‘zero history’ and ‘real geography’ produces a better simulation of real world systems of central places. 4. As an exception, Brakman et al. (2002) do study whether an agglomeration equilibrium also represents a welfare optimum. Interestingly it does not. The relation is close, but not perfect. 5. See Oosterhaven (1984) for the entire family of both square and rectangular, regional and inter-regional IO tables and models. 6. The data used in this early study come from the Schilderinck (1984) EU inter-country tables for 1959–1975 (see Oosterhaven, 1995, for a discussion of these data). 7. In fact, the value of 100 per cent is only reached if the EU total is defined exclusive of, respectively, the country at hand or sector at hand (see Hoen, 1999: 210–11). See Hoen and Oosterhaven (2006) for a more general discussion and a comparison of the present additive measures with their multiplicative alternatives, known as the index of Revealed Comparative Advantage in international economics and as the Location Quotient in economic geography. 8. For comparability reasons, for 1975, two inter-country input–output tables are used: the first inclusive of Value Added Tax and exclusive of the UK and Denmark, the second exclusive of VAT and inclusive of the UK and Denmark (see Linden, 1998, for details). 9. Moreover, it should be taken into account that the exports of the larger countries weigh more heavily in the EU-average export package, which means that for definitional reasons they already have a less specialized export package.

Trade During Economic Integration 219 10. The 38-sector tables for 1975, 1980 and 1985 were aggregated to arrive at an identical sector classification. 11. Note that the rows denoted ‘Domestic’ refer to figures largely as ‘local content’ indicators by Kuroiwa in Chapter 4 of this volume, in the context of bilateral trade agreements. 12. We only consider the final demand category ‘consumption by households’; ‘government expenditures’, which also includes government investments, are not taken into account. 13. Only ‘Gross Fixed Capital Formation’ is considered. Although ‘Changes in Stocks’ are often formally seen as investment, we feel that this final demand category does not closely correspond to the physical capital goods we intend to study. 14. See Hoen (1999: 188). The results presented in this section are based on trade in intermediate inputs only.

References Balassa, B. (1966). ‘Tariff Reduction and Trade in Manufactures among Industrial Countries’, American Economic Review, 56: 466–73. Balassa, B. (1975). ‘Trade Creation and Trade Diversion in the European Common Market: an Appraisal of the Evidence’, in B. Balassa (ed.), European Economic Integration, Amsterdam: North-Holland. Beyers, W. B. (1976). ‘Empirical Identification of Key Sectors: Some Further Evidence’, Environment and Planning A, 8: 231–6. Boomsma, P. and J. Oosterhaven (1992). ‘A Double-entry Method for the Construction of Bi-regional Input-Output Tables’, Journal of Regional Science 32: 269–84. Brakman, S., H. Garretsen and M. van Marrewijk (2002). An Introduction to Geographical Economics, Cambridge: Cambridge University Press. Dietzenbacher, E. (1997). ‘In Vindication of the Ghosh Model: a Reinterpretation as a Price Model’, Journal of Regional Science, 37: 629–51. Dietzenbacher, E. (2002). ‘Interregional Multipliers: Looking Backward and Forward’, Regional Studies, 36, 2: 125–36. Dietzenbacher, E. and A. R. Hoen (1999). ‘Deflation of Input–Output Tables from the User’s Point of View’, Review of Income and Wealth, 44: 111–22. Dietzenbacher, E. and J. A. van der Linden (1997). ‘Sectoral and Spatial Linkages in the EC Production Structure’, Journal of Regional Science, 37, 2: 235–57. Dietzenbacher, E., J. A. van der Linden and A. E. Steenge (1993). ‘The Regional Extraction Method: EC Input–Output Comparisons’, Economic Systems Research, 5, 2: 185–206. Dixit, A. K. and V. D. Norman (1980). Theory of International Trade. Cambridge: Cambridge University Press. Dixit, A. K. and J. E. Stiglitz (1977). ‘Monopolistic Competition and Optimum Product Diversity’, American Economic Review, 67: 297–308. Eurostat (1983). National Accounts ESA, Input–Output Tables 1975. Luxembourg. Eurostat (1986). National Accounts ESA, Input–Output Tables 1980. Luxembourg. Eurostat (1990). External Trade Statistics: User’s Guide. Luxembourg. Fujita, M., P. R. Krugman and A. J. Venables (1999). The Spatial Economy, Cities, Regions and International Trade, Cambridge, MA: MIT Press.

220 The Evolution of the Core–Periphery Structure Grubel, H. G. and P. J. Lloyd (1971). ‘The Empirical Measurement of Intra-Industry Trade’, The Economic Record, 47: 494–517. Heston, A., R. Summers and B. Aten (2006). Penn World Table Version 6.2, Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania. Hirschman, A. O. (1958). The Strategy of Economic Development, New Haven: Yale University Press. Hoen, A. R. (1999). ‘An Input–Output Analysis of European Integration’, PhD, University of Groningen. Hoen, A. R. and J. Oosterhaven (2006). ‘On the Measurement of Comparative Advantage’, Annals of Regional Science, 40: 677–91. Isard, W. (1951). ‘Interregional and Regional Input–Output Analysis: a Model of a Space Economy’, Review of Economics and Statistics, 33: 318–28. Isard, W. (1956). Location and Space-Economy. Cambridge, MA: MIT Press. Jacquemin, A. and A. Sapir (1988). ‘European Integration or World Integration?’ Weltwirtschaftliches Archiv, 124: 127–39. Jones, L. P. (1976). ‘The Measurement of Hirschmanian Linkages’, Quarterly Journal of Economics, 90: 323–33. Krugman, P. R. (1980). ‘Scale Economies, Product Differentiation, and International Trade’, American Economic Review, 17: 950–9. Krugman, P. R. (1991). Geography and Trade. Leuven: Leuven University Press. Linden, J. A. van der (1998). ‘Interdependence and Specialisation in the European Union: Intercountry Input–Output Analysis and European Integration’, PhD, University of Groningen. Linden, J. A. van der and J. Oosterhaven (1995). ‘European Community Intercountry Input–Output Relations: Construction Method and Main Results for 1965–1985’, Economic Systems Research, 7, 3: 249–69. Linden, J. A. van der and J. Oosterhaven (2000). ‘Specialisation and Concentration ¨ cker and H. Herrmann (eds), Spatial in the European Union, 1965–85’, in J. Bro Change and Interregional Flows in the Integrating Europe, Berlin: Physica Verlag: 181–200. Linden, J. A. van der, J. Oosterhaven, F. A. Cuello, G. J. D. Hewings and M. Sonis (2000). ‘Fields of Influence of Productivity Change in EC Intercountry Input– Output Tables, 1970–80’, Environment & Planning A, 32, 7: 1287–1305. Mayes, D. G. (1978). ‘The Effects of Economic Integration on Trade’, Journal of Common Market Studies, 17: 1–25. Oosterhaven, J. (1984). ‘A Family of Square and Rectangular Interregional Input– Output Tables and Models’, Regional Science and Urban Economics, 14: 565–82. Oosterhaven, J. (1988). ‘On the Plausibility of the Supply-Driven Input–Output Model’, Journal of Regional Science, 28: 201–17. Oosterhaven, J. (1995). ‘Changing Specialisation and Interdependency of EC Economies 1959–1975’, Australasian Journal of Regional Studies, 1, 1: 51–64. Oosterhaven, J. (1996). ‘Leontief versus Ghoshian Price and Quantity Models’, Southern Economic Journal, 62, 3: 750–9. Oosterhaven, J., D. Stelder and S. Inomata (2007), ‘Evaluation of Non-Survey International IO Construction Methods with the Asian-Pacific Input-Output Table’. IDE Discussion Paper 114, Chiba: IDE/JETRO. Porter, M. E. (1990). The Competitive Advantage of Nations, New York: Free Press.

Trade During Economic Integration 221 Rasmussen, P. N. (1956). Studies in Intersectoral Relations, Amsterdam: NorthHolland. Schilderinck, J. H. F. (1984). ‘Interregional Structure of the European Community, Part II: Interregional Input–Output Tables, 1959, 1965, 1970 and 1975’, research memorandum, Tilburg University. Sonis, M. and G. J. D. Hewings (1992) ‘Coefficient Change in Input–Output Models: Theory and Applications’, Economic Systems Research, 4: 143–257. Sonis, M., J. Oosterhaven and G. J. D. Hewing (1993). ‘Spatial Economic Structure and Structural Changes in the European Common Market: Feedback Loop Input–Output Analysis’, Economic Systems Research, 5, 2: 173–84. Stelder, T. M. (2005). ‘Regions and Cities: Five Essays on Interregional and Spatial Agglomeration Modelling’, PhD, University of Groningen. ¨ r Bestimmung Strategische Sektoren mit Hilfe van Input– Strassert, G. (1968). ‘Zu Output Analysis’, Jahrbu¨cher fu¨r Nationalo¨konomie und Statistik, 182: 211–15. Venables, A. J. (1996). ‘Equilibrium Locations of Vertically Linked Industries’, International Economic Review, 37: 341–59.

This page intentionally left blank

Part III The Current Status of Indigenous Enterprises

This page intentionally left blank

9 Spillovers and Linkages between Local and Foreign Plants Kazuhiko Yokota

9.1

Introduction

International trade or foreign direct investment (FDI) inflow affects a local firm’s activities in many ways. Among these are linkage and spillover effects which have recently become important in the development of local firms as well as the country as a whole as globalization proceeds. Recent globalization has been led by various types of institutions which include free trade agreements (FTA) and economic partnership agreements (EPA). East and Southeast Asian countries have been deeply involved in globalization mainly though bilateral FTAs and EPAs. From 1992, the Association of South-East Asian Nations (ASEAN) started to formulate the ASEAN free trade agreement (AFTA) and the original five ASEAN member countries, i.e. Indonesia, Malaysia, the Philippines, Singapore and Thailand, have achieved AFTA’s targeted tariff reductions (0–5 per cent) in ASEAN trade in 2002. AFTA is just one example of their involvement in globalization. Although there have been many studies that analyse the impact of globalization or FTAs on the local economy, little research has been done on the changes in economic linkage between multinational enterprises (MNEs) and local firms. In most Asian, especially Southeast Asian countries, MNEs have a crucial role for their economic development. Research on the changing structure of linkages between MNEs and local firms is indispensable for understanding the development process of these countries. This chapter focuses on this issue. There are two types of linkage effects: forward and backward. Forward linkage effects between MNEs and local firms arise when an MNE sells its intermediate goods to local producers, while backward linkage effects arise when an MNE purchases input goods from local suppliers. It is generally expected in developing countries that backward linkage effects are 225

226 The Current Status of Indigenous Enterprises

more likely to occur than forward linkage effects. Javorcik (2004) analyses the effects of MNEs’ backward and forward linkages on MNEs’ and local firms’ productivity in Lithuania using plant-level data and suggests that the spillovers of knowledge from MNEs occur through mainly backward linkage effects. Lin and Saggi (2007) construct an oligopoly model that explains the spillovers through backward linkage.1 However, these linkage effects are not necessarily always positive. Negative linkage effects may exist when the arrival of MNEs leads to the exit of less productive local firms from the market, which in turn contributes to increased industrial productivity. From the local firm’s point of view, however, this is a negative linkage effect. There are a few empirical studies that have been done on this type of linkage effect. In addition to Javorcik (2004),2 Kugler (2006) finds the positive externality effect of demand linkage (FDI diffuses technological knowledge to local upstream suppliers) in Colombian manufacturing sectors. On the other hand, there is a great variety of literature on the trade/FDI impact on firms’ (local and MNEs’) productivity, such as Pavcnik (2002) for Chile, Fernandes (2007) for Colombia, Amiti and Konings (2005) for Indonesia, Topalova (2004) for India and Muendler (2004) for Brazil, to name just a few. There are mainly three channels for this impact. First, an increase in foreign competition fosters process innovation. Second, import input materials often contain foreign advanced technology. Hence, trading these materials reinforces technology transfer. Third, foreign pressure forces less efficient firms to exit from the market. Eventually, only efficient firms remain and an industry level productivity increases. This chapter contributes to the literature in an important respect. Focusing on the case of Thailand, I provide evidence that there are positive and negative linkage effects and spillover effects of MNEs depending on the type of industry. These findings confirm the results in the literature listed above, but at the same time suggest reservations about the existence of effects. Before proceeding with analysis, let us look at an overview of the industry characteristics of ASEAN. The structure of industries has changed over time in ASEAN. Table 9.1 shows some industry information about these countries. Except for the Philippines, the growth rates of the original member countries (Indonesia, Malaysia, Singapore, Thailand) after the Asian crisis are much lower than those before. On the other hand, many new member countries such as Cambodia, Laos and Vietnam have achieved high economic growth after the crisis. Table 9.1 also reveals another notable finding: that the manufacturing growth rate (fourth column) exceeds the GDP growth rate (third column) in all countries

Spillovers and Linkages between Local and Foreign Plants 227 Table 9.1 Industrialization in ASEAN Country

Period

Growth rate of GDP (%)

Growth rate of Manufacturing (%)

Share of Manufacturing (%)

Indonesia

1988–1996 1998–2005

7.8 4.1

10.7 4.9

23.4 27.8

Malaysia

1988–1996 1998–2005

9.0 5.2

13.1 6.9

26.4 31.1

Philippines

1988–1996 1998–2005

3.2 4.4

3.0 4.0

23.0 22.3

Singapore

1988–1996 1998–2005

8.5 5.1

7.5 6.7

25.2 24.7

Thailand

1988–1996 1998–2005

8.6 4.8

11.6 7.0

27.9 34.1

Brunei

1988–1996 1998–2004

1.4 2.8

0.0 3.6

43.6 41.2

Cambodia

1988–1996 1998–2004

N.A. 7.1

N.A. 16.7

9.0 18.6

Laos

1988–1996 1998–2004

6.9 6.1

16.2 9.4

12.5 18.4

Myanmar

1988–1996 2001–2003

N.A. 4.0

Vietnam

1988–1996 1998–2005

7.2 6.6

N.A. 8.3 5.2 10.3

9.0 20.1 15.2 19.6

Note: Figures are calculated based on 2000 constant price. Source: Asian Development Bank (2006), Key Indicators 2006.

after the crisis except for the Philippines. As a result of this, the share of the manufacturing sector expands for all countries except for the Philippines (fourth column). Although the impact of the Asian crisis was serious, this indicates that the manufacturing sector has been a leading sector for economic growth in many ASEAN member countries. Thailand is a typical example of a country that has achieved high economic growth led by liberalizing manufacturing sectors and also by FDI inflow in many sectors. However, there are few empirical studies on Thailand’s trade liberalization and FDI structure. Urata and Yokota (1994) analyse the impact of trade liberalization on the industry productivity in Thai manufacturing sectors and find positive effects of liberalization on productivity. Recently, Milner et al. (2004) studied the vertical FDI structure of Japanese MNEs in the manufacturing sector in Thailand and found that the Japanese FDI in Thailand is generally of a vertically integrated nature.

228 The Current Status of Indigenous Enterprises

In the following sections, I used the Thai plant-level data set to investigate these effects on local firms.3 This chapter is organized as follows. The next section provides industry characteristics and discusses the adjustment process of each industry group. Section 9.3 introduces the empirical model and discusses the results. Section 9.4 discusses the results of the previous section more carefully and the impact of FTA. Section 9.5 discusses the impact of FTAs on local plants and section 9.6 concludes the chapter.

9.2 Industrial structure and local–multinational linkages in Thailand 9.2.1

Categorization of industry

To show the results in a simpler fashion and to make analyses sharper, 2-digit international standard industrial classification (ISIC) industries which have 22 sectors4 were aggregated into four categories: resourcebased, machinery, labour-intensive, and metal and chemical. The results of the categorization are reported in Table 9.2. To categorize industries, three different ratios were calculated: resourceinput, machinery-input and labour-input. All information for this calculation comes from input–output (IO) tables compiled by the National Statistical Office (NSO), Thailand. The average of the ratios of three input–output tables of 1995, 1998 and 2000 was used. Resource-input ratio is defined as the ratio of resource inputs over total intermediate inputs. Resource inputs are composed of ISIC 3-digit codes from 001 to 044 including agricultural products (001–017), logging (025), forestry products (027), crude oil and natural gas (031), iron ore (032), tin ore (033) and fertilizer (037). Intermediate inputs include all agricultural, manufacturing and services sectors. This category includes coke, refined petroleum products (23), food products (15), wood products (20), rubber and plastic products (25), and tobacco (16) and so on. Industries categorized as resource-based have very high resource-input ratios ranging from 0.21 to 0.49. It is also notable that these industries have relatively lower machinery-input ratios as shown in the second column of Table 9.2. The machinery industry is composed of industries that have relatively higher machinery-input ratios, defined as the ratio of machinery inputs to the total intermediate inputs. The category of machinery that is used for this ratio includes general machinery (29), computer (30), electric machinery (31), TV, radio (32), precision (33), motor vehicles (34) and

Table 9.2 Categorization of industries Group

Industry

Resource

Machinery

Labour

23 15 20 25 26 16

Coke, refined petroleum products Food products and beverages Wood and products of wood Rubber and plastic products Other non-metallic mineral Tobacco products

0.81026 0.49483 0.42601 0.26524 0.20555 0.16361

0.00348 0.01514 0.03401 0.00803 0.03742 0.02201

0.04157 0.09567 0.16250 0.11933 0.18902 0.26821

29 30 31 32 33 34 35 36

Machinery and equipment n.e.c. Office, accounting and computing Electrical machinery Radio, television Medical, precision and optical Motor vehicles, trailers Other transport equipment Furniture manufacturing n.e.c.

0.00044 0.00000 0.00243 0.00123 0.00000 0.00644 0.00332 0.01128

0.74023 0.73412 0.70807 0.58214 0.51375 0.49820 0.47034 0.01899

0.05525 0.08677 0.12314 0.12669 0.14836 0.15422 0.10377 0.30627

3. Labour-intensive Industry

19 17 18 27 22

Tanning and dressing of leather manufacture Textiles Wearing apparel dressing Basic metals Publishing, printing and reproduction of records

0.08704 0.08608 0.00062 0.07748 0.00000

0.01719 0.01686 0.00647 0.01825 0.01767

0.18906 0.17478 0.17142 0.16538 0.16164

4. Metal and Chemical Industry

28 24 21

Fabricated metal products Chemicals and chemical products Paper and paper products

0.00122 0.04660 0.02809

0.01514 0.01693 0.01528

0.15151 0.12084 0.07870

1. Resource-based Industry

2. Machinery Industry

ISIC

229

Notes: Resource: resource inputs/total intermediate inputs, Labour: total wage payments/total intermediate inputs, Capital: machinery inputs/total intermediate inputs. Sources: Input–Output Tables, 1995, 1998, 2000, National Statistical Office, Thailand.

230 The Current Status of Indigenous Enterprises

other transportation equipments (35). As expected, the industries with high machinery-input ratios are the machinery industries themselves. Each industry included in this category has a very high machinery share ranging from 0.74 in machinery and equipment (29) to 0.47 in other transport equipment (35). The labour-intensive industry is categorized according to the labourinput ratio. Labour-input ratio is defined as the share of total labour inputs in industry j in total intermediate inputs. Labour input is defined as the total wage payments in industry j. The labour-intensive group includes furniture (36), leather manufacture (19), textiles (17) and wearing apparel (18). The final category is the metal and chemical industry. This group is characterized by lower resource ratios, lower machinery input ratios, and relatively lower labour-intensiveness. This category is named after the industries included in the group, such as basic metals (27), fabricated metal products (28) and chemicals and chemical products (24). 9.2.2

Direction of industrial adjustment

Using this categorization, the next task is to examine the direction of industrial adjustment, especially in output and productivity, by the four industry groups. To see the direction of the adjustments of industry’s outputs and productivity, the weighted productivity was decomposed into two parts: unweighted productivity and covariance between unweighted productivity and weight, i.e. the plant’s output share of the industry output as in Olley and Pakes (1996) and Pavcnik (2002). This decomposition was computed by year and industry group. The decomposition formula is the following: X j j X j j j j j j j j j sit pit ¼pt þ ðsit  st Þð pit  pt Þ ¼ pt þ ss sp corrðsit , pit Þ, ð1Þ ‘w t ¼ i j ‘wt

i

where is the weighted average of productivity in industry j and year t. j Productivity pit is labour productivity which is defined as value-added per j j production worker. sit and st are the plant’s share of the industry output of plant i, industry j, and year t, and an industry average of the variable. The last equation converts the variance to the correlation expression, denoting standard deviations of weight and productivity by ss and sp respectively. The covariance component represents and reveals the reallocation of output share and the plant’s productivity level. If the covariance is positive, we infer that more outputs in the industry are produced by more efficient plants. In other words, the reallocation of the resources from less to more productive plants in the industry occurs if the covariance term is positive.

Spillovers and Linkages between Local and Foreign Plants 231

Instead of covariance as used in other literature, the Pearson correlation index was used as well as the Spearman rank correlation that can test statistical significance. However, the interpretation is exactly the same as in the discussion above. Table 9.3 shows the results of the correlation between the plant’s productivity and output share in each industry group. As expected, all correlation coefficients are positive and almost all are statistically significant at 0.1 per cent level both in Pearson and Spearman rank correlations. This means that in each year and each industry group, resources had been reallocated from less to more productive plants in industry j in time t. Another notable finding from Table 9.3 is the importance of MNE presence in industry. In Table 9.3, the share of numbers and share of output indicate the share of MNE numbers in the total number of plants in its industry group, and the MNE’s output share in its industry group, respectively. Comparing 1996 and 2002, we notice that all industry groups reduce their shares in numbers but raise output shares, except for the labour-intensive industry. This means the emergence of a more oligopolistic situation in resourcebased, machinery, and metal and chemical industry groups. An astonishing fact is the high output share in the machinery industry. It is inferred from this fact that MNEs play an important role in the Thai manufacturing industries, especially the machinery industry group.

9.3 9.3.1

Spillover and linkage effects Estimation strategy

In this section, I specify the possible effects of a foreign plant on the local plant’s productivity. Two possible channels of technology spillover from MNEs to local firms are investigated: diffusion of technological knowledge in the same industry group, and the demand linkage effect where MNEs and local firms are vertically linked in the supply and demand for intermediate goods. In the case of knowledge diffusion, the extent of foreign presence in an industry affects the local firm’s productivity in the same industry. In the second case, demand linkage effects can be broken into two subeffects: backward and forward. Backward linkage effects occur when MNEs purchase inputs from local suppliers, in other words, when MNEs grow, local suppliers also grow. On the other hand, forward linkage effects occur when MNEs provide inputs to local (often final) producers. As mentioned in the introduction and as described in Javorcik (2004), Lin and Saggi (2007) and Kugler (2006), it is plausible that backward linkage

Table 9.3 Directions of resource reallocation

Local 1. Resorce-based Industry

2. Machinery Industry

3. Labour-intensive Industry

4. Metal and Chemical Industry

1998 MNE

Local

1999 MNE

Local

2000 MNE

Local

232

1996

2002 MNE

Local

MNE

Pearson Spearman Share of Numbers Share of Output

0.1996* 0.2514* 0.5632* 0.5216* 12.6%

0.2075* 0.3923* 0.7049* 0.5474* 7.6%

0.1761* 0.0727 0.7206* 0.5421* 6.9%

0.1665* 0.0854 0.6027* 0.3474* 7.7%

0.1316* 0.3625* 0.6961* 0.4765* 7.8%

33.3%

42.2%

52.7%

47.8%

49.4%

Pearson Spearman Share of Numbers Share of Output

0.3122* 0.3279* 0.5094* 0.4259* 29.4%

0.1472* 0.2375* 0.6086* 0.4287* 31.9%

0.3313* 0.2945* 0.6388* 0.5754* 22.2%

0.1601* 0.1759 0.438* 0.3206* 24.2%

0.2455* 0.3726* 0.5456* 0.5042* 27.7%

85.7%

90.0%

93.0%

91.6%

94.9%

Pearson Spearman Share of Numbers Share of Output

0.2450* 0.1900* 0.4631* 0.3657* 17.5%

0.2803* 0.2776* 0.6983* 0.5644* 14.1%

0.1853* 0.4638* 0.6191* 0.5714* 10.3%

0.1173* 0.1454 0.5641* 0.2716* 13.6%

0.1976* 0.2938* 0.6951* 0.4544* 11.9%

47.9%

47.5%

37.4%

41.8%

40.4%

Pearson Spearman Share of Numbers Share of Output

0.2740* 0.3237* 0.5362* 0.5665* 17.4%

0.2515* 0.3686* 0.6792* 0.6949* 15.9%

0.3436* 0.3792* 0.6885* 0.6279* 13.5%

0.2257* 0.0936 0.5168* 0.4012* 11.3%

0.2644* 0.3195* 0.6867* 0.6272* 13.8%

47.6%

48.5%

66.6%

45.3%

48.8%

Notes: * Represents 0.1 per cent statistical significance in Pearson correlation and Spearman rank correlation tests. Sources: Industrial Surveys (1999, 2000, 2001, 2003), National Statistical Office, Thailand.

Spillovers and Linkages between Local and Foreign Plants 233

happens more often in developing countries than forward linkage. I therefore focus on the backward linkage effect in this chapter. There are two directions of these two effects on local firms. First, a local plant’s output reduces due to being crowded out by foreign plants. Second, a local plant’s output increases due to linkage effects by foreign plants.5 Which effect is stronger crucially depends on the characteristics of the industry and this is an empirical issue. However, casual observations indicate that the first case occurs when MNEs sell the same products as local plants, and then MNEs take a large share of the market from local producers. It may often happen in the industry that MNEs compete with local producers for final goods, for example, food products, rubber goods, textiles, apparel, and metal product industries. Those can be basically categorized in resource-based, labourintensive, or metal industries. On the other hand, the second case often happens in the differentiated products or vertically linked industries, such as electric machinery, motor vehicles and computer industries that are categorized in the machinery industry in this chapter. In these sectors, MNEs purchase inputs from local suppliers to produce final goods. The spillover effect is captured by estimating the effect of foreign presence on the local plant productivity. I specify a foreign presence variable, MNEjt . MNEjt expresses the extent of foreign presence in sector j at time t and is defined as foreign equity participation in that sector. MNEjt is defined as follows: X foreign plant outputijt MNEjt ¼

i˛j

X

all plant outputijt

i˛j

Using the plant-level pooled cross-section data set, I estimate the following equation: pijt ¼ a0 þ aj dj þ at dt þ a1 MNEjt þ ða2 Þ0 Zijt þ eijt ,

ð2Þ

where pijrt is labour productivity of plant i in industry j in region r and in year t in logarithm. Labour productivity is defined as the ratio of the plant’s value-added to the number of production workers in that plant. MNEjt is calculated at each 3- and 4-digit ISIC level of industry j and year t, Zijrt is a vector of local plant’s and industry’s characteristics, such as plant age, plant size, plant’s trade activities, such as export and import and aj, at are industry effects, and year effects, respectively. The operating years of plants and its square terms are included to account for heterogeneity in productivity as in Fernandes (2007).

234 The Current Status of Indigenous Enterprises

Operating years and its square term are taken logarithm. Plant size is included in the estimation for controlling plant heterogeneity. The logarithm of the number of production workers in that plant is used for the proxy of plant size. Since labour productivity does not capture the contributions of capital, another important control variable for labour productivity is necessary. Capital-labour ratio is a proxy for the contribution of capital which is defined as the ratio of capital in terms of a million Thai baht to the number of production workers in the plant. Plant-level information about its external activities such as export and import is also important in accounting for its productivity. If the local plant exports its product to the world market, it indicates a strong plant’s competitiveness or that plant learns from exporting. On the other hand, if the local plant imports, it can learn new technology from the imported intermediate goods and from the interaction with the world market. The positive sign is expected for the estimated coefficient on foreign presence if spillover from the foreign plants to local plants exists. The expected signs of coefficients on capital-labour ratio, plant size, export and import are therefore positive. If labour productivity increases as a plant accumulates experience but that experience diminishes as a plant becomes old, the sign on plant age would be positive while the sign on age-squared would be negative. As I mentioned above, there are two main effects of MNEs on local plants: spillover and demand linkage effect. I implicitly assume that this spillover effect occurs within the same sectors. In this sense, the presence of MNEs affects local plants horizontally, as is described in Javorcik (2004). However, the spillover effect can affect local plants across sectors which I refer to as the linkage effect in the previous section. To examine the correlation between plant productivity and FDI across sectors, it is useful to consider the linkage effect of MNE. pijt ¼ b0 þ bj dj þ bt dt þ b1 Linkagejt þ ðb2 Þ0 Zijt þ eijt ,

ð3Þ

where Linkagejt is a proxy for a demand linkage effect defined as follows: ! X Linkage jt ¼ ln intermediate demandjht  MNEjt : h˛H

Intermediate demandjht is the total value (in logarithm) of intermediate demand of foreign plants in industry j for industry h and time t. In other words, the independent variable linkage captures the intermediate

Spillovers and Linkages between Local and Foreign Plants 235

demand for all sectors (denoted by H ) by the foreign plants in industry j. Demand linkage externality from MNEs to local suppliers occurs when MNEs increase their demand for intermediate inputs which in turn lead scale economies to local suppliers in the upstream. Linkagejt is calculated from the transaction matrices of IO tables for three years, 1995, 1998 and 2000. 9.3.2

Data

Plant-level data are compiled from the NSO, Thailand, which contains information on plant sales, employment, raw materials, equity, existence of export and import, and year of start-up operations, etc. The survey data of five years, 1996, 1998, 1999, 2000 and 2002, are used for the analyses. The 1996 data come from the census data while the other years come from survey data. The coverage of the survey varies from year to year. The final sample for estimation has about 35,000 observations. The NSO conducted surveys on manufacturing establishments by using the combination of stratified sampling and systematic sampling. The NSO stratified establishments in each province according to industry codes and the number of workers. Then samples were selected from each province-industry-worker stratum using systematic sampling. Industry categorization and some variables such as linkage effect are calculated from IO tables, compiled by the NSO. IO tables of 1995, 1998 and 2000 are used for this study. Each IO table is converted into 2-, 3- and 4-digit ISIC codes of the industry survey code (ISIC). I used IO tables of 1995 for 1996 industrial census, and 1998 IO tables for 1998 and 1999 industrial survey data, and IO tables of 2000 for 2000 and 2002 survey data respectively. This treatment lessens the reverse causality problem: the presence of MNEs raises or decreases the local plant’s productivity because of linkage effects; on the other hand, efficient labour attracts MNEs from abroad. All variables are deflated by appropriate price indices which come from the Bank of Thailand. Table 9.4 contains the summary statistics of variables used in estimation. Some notable features should be addressed. First, the metal and chemical industry has the highest average labour productivity, followed by the machinery industry. Second, averages of foreign presence and demand linkage are very high in the machinery industry. Third, the share of exports is relatively high in machinery and labour-intensive industries, while import share is the highest in machinery followed by the metal and chemical industry. In general, the machinery industry, on average, has a strong connection with the world market. The last notable feature of the variable is that the average output and average

236

Table 9.4 Summary statistics of variables Variable

Resource-based Industry Mean

Labour Productivity Foreign Presence 3-digit Foreign Presence 4-digit Demand Linkage 3-digit Demand Linkage 4-digit Capital-Labour Ratio Plant Size Plant Age Export Import Protection 4-digit Output Value-added

Machinery Industry Std. Dev.

Mean

Labour-intensive Industry Std. Dev.

Mean

Metal and Chemical Industry

Std. Dev.

Mean

Std. Dev.

0.41710 0.35265

2.32263 0.15934

0.49450 0.73711

1.79706 0.24940

0.22868 0.41617

0.73207 0.18499

0.62631 0.48435

4.28512 0.19659

0.33735

0.21461

0.72085

0.28194

0.40967

0.19438

0.47148

0.23105

6.28951

2.91091

13.31524

4.88158

6.92158

3.28553

8.70867

3.73795

5.72138

3.79341

12.61730

5.19825

6.74590

3.27704

8.13370

4.25056

0.85656 98.621 12.84157 0.21625 0.22328 0.22251 212.11 64.10

3.24417 270.574 10.6334 0.41170 0.41646 0.18981 1889.90 662.65

0.81646 192.917 11.11999 0.34876 0.55589 0.14916 577.35 155.83

2.31065 574.000 8.92147 0.47663 0.49692 0.11001 3435.59 1037.82

0.37119 147.798 10.74385 0.37001 0.36941 0.24651 114.18 35.45

1.41922 393.496 9.32393 0.48284 0.48268 0.30617 429.50 131.18

1.12976 70.594 12.37242 0.20164 0.42607 0.16058 219.38 62.02

4.32602 159.601 10.10208 0.40126 0.49454 0.07925 1329.16 541.78

Spillovers and Linkages between Local and Foreign Plants 237

value-added are the highest in the machinery industry. It is almost five times larger than the per capita output in labour-intensive industry. 9.3.3

Estimation results

Table 9.5 reports the estimation results of equation (2). Equation (2) is estimated using ordinary least squares (OLS). Table 9.5 is divided into four groups according to the categorization in Table 9.2. All estimations include capital-labour ratio, plant size measured as the logarithm of number of production workers, plant age which equals the logarithm of operating years of the plant, the square term of plant age, and status of international trade, that is, export and import, as explanatory variables. Trade status variables are binary, i.e. if the plant exports (imports), the variable takes 1 and zero otherwise. Year effect is included in all specifications. It may be assumed that aside from foreign presence, there is heterogeneity across industries which may affect plant’s productivity. Hence, the heterogeneity across industry should be taken into account when the equation is estimated. For removing this effect and controlling for unobservable shocks across industry, the industry effect is included in all estimated equations. Since foreign presence and industry effects are calculated at both 3- and 4-digit ISIC level, the possible correlation (multi-colinearity) between foreign presence and industry effect calculated at 3- and 4-digit ISIC level should be taken into careful consideration. Hence, I estimate four different specifications for each group: the first specification includes foreign presence at 3-digit ISIC level and industry effect at 3-digit ISIC level. The second estimation of each group has a combination of 3-digit foreign presence with 4-digit industry effect. The third specification has foreign presence calculated at 4-digit ISIC level as a dependent variable with 3-digit industry effect as an independent variable. The fourth specification includes 4-digit foreign presence and 4-digit industry effect. There are some notable results in Table 9.5. Among the four industry groups, only the machinery industry has positive and significant coefficients on foreign presence at both 3- and 4-digit levels. Coefficients on foreign presence at 3-digit in resource-based industries are negative and statistically significant. In other industry groups, such as labour-intensive and metal and chemical industries, coefficients on foreign presence are negative although insignificant. Coefficients on other independent variables have corrected (expected) signs in all industry groups and all specifications and almost all coefficients are highly significant. A positive coefficient on plant size indicates that a relatively large plant in terms of the number of employees makes

Table 9.5

Spillover effect and plant productivity

Foreign Presence 3-digit Foreign Presence 4-digit CapitalLabour Ratio Plant Size Plant Age Age-squared Export Import Year Effects Industry Effects 3-digit Industry Effects 4-digit N obsevations R-squared

(1)

(2)

0.1963** (0.0940)

0.1978** (0.0939)

Machinery Industry (3)

0.0637 (0.0559) 0.4496*** (0.0068) 0.1267*** (0.0098) 0.1485*** (0.0497) 0.0076 (0.0113) 0.2049*** (0.0330) 0.2034*** (0.0291) Yes Yes

0.4259*** (0.0069) 0.1032*** (0.0098) 0.1479*** (0.0492) 0.0049 (0.0112) 0.2978*** (0.0334) 0.1655*** (0.0293) Yes

0.4492*** (0.0068) 0.1281*** (0.0098) 0.1472*** (0.0498) 0.0073 (0.0113) 0.2035*** (0.0330) 0.2023*** (0.0291) Yes Yes

Yes 14355 0.3297

14355 0.3285

238

Resource-based Industry (4)

(5)

(6)

0.2558** (0.1305)

0.2343* (0.1308)

0.0764 (0.0626) 0.4259*** (0.0069) 0.1036*** (0.0098) 0.1489*** (0.0493) 0.0051 (0.0112) 0.2976*** (0.0334) 0.1662*** (0.0293) Yes

0.2352*** (0.0128) 0.0746*** (0.0171) 0.2621*** (0.0891) 0.0559*** (0.0208) 0.0645 (0.0537) 0.2187*** (0.0400) Yes Yes

Yes 14355 0.3304

14355 0.3286

0.2313*** (0.0129) 0.0759*** (0.0171) 0.2506*** (0.0891) 0.0518** (0.0208) 0.0561 (0.0540) 0.2208*** (0.0401) Yes

(7)

(8)

0.2577*** (0.0980)

0.203* (0.1248)

0.2351*** (0.0128) 0.0743*** (0.0171) 0.2549*** (0.0890) 0.0540*** (0.0208) 0.0633 (0.0537) 0.2179*** (0.0400) Yes Yes

0.2315*** (0.0129) 0.0761*** (0.0171) 0.2494*** (0.0891) 0.0517** (0.0208) 0.0558 (0.0540) 0.2207*** (0.0401) Yes

Yes 3577 0.1667

3577 0.1660

Yes 3577 0.1688

3577 0.1678

Labour-intensive Industry (9) Foreign Presence 3-digit Foreign Presence 4-digit CapitalLabour Ratio Plant Size Plant Age Age-squared Export Import

0.4022*** (0.0093) 0.0258** (0.0126) 0.4022*** (0.0642) 0.0758*** (0.0154) 0.3538*** (0.0388) 0.2716*** (0.0364) Yes Yes

(11)

0.1796 (0.1299)

0.3993*** (0.0093) 0.0235* (0.0126) 0.3878*** (0.0641) 0.0708*** (0.0154) 0.3731*** (0.0388) 0.2358*** (0.0366) Yes

7144 0.3114

(13) 0.1159 (0.1209)

0.1685 (0.1042)

0.0581 (0.1191)

0.4018*** (0.0093) 0.0261** (0.0126) 0.4006*** (0.0642) 0.0752*** (0.0154) 0.3554*** (0.0387) 0.2706*** (0.0364) Yes Yes

0.3990*** (0.0093) 0.0234* (0.0126) 0.3870*** (0.0641) 0.0707*** (0.0154) 0.3756*** (0.0388) 0.2356*** (0.0366) Yes

Yes 7144 0.3117

(12)

0.3165*** (0.0101) 0.1202*** (0.0141) 0.3917*** (0.0705) 0.072*** (0.0160) 0.1816*** (0.0462) 0.2545*** (0.0331) Yes Yes

Yes 7144 0.2942

7144 0.3108

(14)

(15)

0.1774 (0.1228)

0.3132*** (0.0102) 0.1213*** (0.0143) 0.3912*** (0.0706) 0.0729*** (0.0160) 0.1978*** (0.0464) 0.2549*** (0.0333) Yes

0.015 (0.0808)

0.0465 (0.1055)

0.3168*** (0.0101) 0.1203*** (0.0142) 0.3928*** (0.0706) 0.0722*** (0.0160) 0.1808*** (0.0462) 0.2550*** (0.0331) Yes Yes

0.3137*** (0.0102) 0.1214*** (0.0143) 0.3919*** (0.0706) 0.0730*** (0.0160) 0.1977*** (0.0464) 0.2555*** (0.0333) Yes

Yes 6279 0.2286

(16)

6279 0.2290

Yes 6279 0.2275

6279 0.2277

Notes: Dependent variable is labour productivity of plant. Standard errors are in the parentheses. ***, **, * indicate 1 per cent, 5 per cent, 10 per cent statistical significance, respectively.

239

Year Effects Industry Effects 3-digit Industry Effects 4-digit N obsevations R-squared

0.1992 (0.1228)

(10)

Metal and Chemical Industry

Table 9.6 Demand linkage effect and plant productivity

Demand Linkage 3-digit Demand Linkage 4-digit Capital-Labour Ratio Plant Size Plant Age Age-squared Export Import Year Effects Industry Effects 3-digit Industry Effects 4-digit N obsevations R-squared

(1)

(2)

0.011** (0.0053)

0.0113** (0.0053)

0.4496*** (0.0068) 0.1266*** (0.0098) 0.1485*** (0.0497) 0.0076 (0.0113) 0.2048*** (0.0330) 0.2034*** (0.0291) Yes Yes

0.4259*** (0.0069) 0.1032*** (0.0098) 0.1479*** (0.0492) 0.0049 (0.0112) 0.2977*** (0.0334) 0.1655*** (0.0293) Yes

Machinery Industry (3)

14355 0.3287

(4)

(5) 0.0156* (0.0080)

0.0041 (0.0032)

0.0051 (0.0037)

0.4491*** (0.0068) 0.1281*** (0.0098) 0.1472*** (0.0498) (0.0073) (0.0113) 0.2039*** (0.0330) 0.2021*** (0.0291) Yes Yes

0.4259*** (0.0069) 0.1035*** (0.0098) 0.1489*** (0.0493) 0.0051 (0.0112) 0.2976*** (0.0334) 0.1662*** (0.0293) Yes

Yes

14355 0.3299

240

Resource-based Industry

0.2352*** (0.0128) 0.0746*** (0.0171) 0.2620*** (0.0891) 0.0559*** (0.0208) 0.0647 (0.0537) 0.2188*** (0.0400) Yes Yes

Yes

14355 0.3305

14355 0.3284

(6)

(7)

0.0143* (0.0080)

0.2313*** (0.0128) 0.0758*** (0.0171) 0.2506*** (0.0891) 0.0518** (0.0208) 0.0562 (0.0540) 0.2209*** (0.0401) Yes

0.0137** (0.0056)

0.0105 (0.0072)

0.2348*** (0.0128) 0.0745*** (0.0171) 0.2536*** (0.0890) 0.0537*** (0.0208) 0.0638 (0.0537) 0.2172*** (0.0400) Yes Yes

0.2312*** (0.0129) 0.0762*** (0.0171) 0.2477*** (0.0891) 0.0512** (0.0208) 0.0559 (0.0540) 0.2201*** (0.0401) Yes

Yes

3577 0.1667

(8)

3577 0.1661

Yes

3577 0.1685

3577 0.1674

Labour-intensive Industry (9) Demand Linkage 3-digit Demand Linkage 4-digit Capital-Labour Ratio Plant Size Plant Age Age-squared Export Import

0.4021*** (0.0093) 0.0258** (0.0126) 0.4022*** (0.0642) 0.0757*** (0.0154) 0.3541*** (0.0388) 0.2716*** (0.0364) Yes

(11)

0.0010 (0.0078)

0.3992*** (0.0093) 0.0235* (0.0126) 0.3879*** (0.0641) 0.0708*** (0.0154) 0.3734*** (0.0388) 0.2358*** (0.0366) Yes Yes

7144 0.3112

(13) 0.0084 (0.0069)

0.0103* (0.0062)

0.0052 (0.0071)

0.4019*** (0.0093) 0.0261** (0.0126) 0.4007*** (0.0642) 0.0753*** (0.0154) 0.3549*** (0.0388) 0.2712*** (0.0364) Yes

0.3991*** (0.0093) 0.0234* (0.0126) 0.3872*** (0.0641) 0.0707*** (0.0154) 0.3751*** (0.0387) 0.2355*** (0.0366) Yes Yes

Yes

7144 0.3114

(12)

0.3163*** (0.0101) 0.1201** (0.0141) 0.3916*** (0.0705) 0.0720*** (0.0160) 0.1818*** (0.0462) 0.2544*** (0.0331) Yes

Yes

7144 0.3114

7144 0.3109

(14)

(15)

0.0119* (0.0070)

0.3130*** (0.0102) 0.1213*** (0.0143) 0.3913*** (0.0706) 0.0730*** (0.0160) 0.1978*** (0.0464) 0.2547*** (0.0333) Yes Yes

0.0005 (0.0044)

0.0035 (0.0059)

0.3169*** (0.0101) 0.1204*** (0.0142) 0.3921*** (0.0706) 0.0720*** (0.0160) 0.1813*** (0.0462) 0.2550*** (0.0331) Yes

0.3136*** (0.0102) 0.1214*** (0.0143) 0.3922*** (0.0706) 0.0731*** (0.0160) 0.1978*** (0.0464) 0.2554*** (0.0333) Yes Yes

Yes

6279 0.2295

(16)

6279 0.2299

Yes

6279 0.2271

6279 0.2279

Notes: Dependent variable is labour productivity of plant. Standard errors are in the parentheses. ***, **, * indicate 1 per cent, 5 per cent, 10 per cent statistical significance, respectively.

241

Year Effects Industry Effects 3-digit Industry Effects 4-digit N obsevations R-squared

0.0111 (0.0073)

(10)

Metal and Chemical Industry

242 The Current Status of Indigenous Enterprises

use of scale economies. A positive coefficient on plant age together with a negative coefficient on age-squared mean that the plant productivity increases as it ages, but its rate of growth diminishes. This can be observed in all industry groups with highly statistical significance except for resource-based industries. Although coefficients on plant age and agesquared are not significant, they have correct signs in resource-based industry. A positive sign on export and import are expected and estimation results confirm this. The export activities of plants are strongly correlated with high labour productivity in resource-based, labour-intensive, and metal and chemical industry groups. Machinery industry has a positive coefficient on exports but this is not statistically significant. On the other hand, coefficients on imports are positive with statistical significance in all industry groups and for all specifications. It simply means that the import activity of a plant is strongly correlated with high plant productivity no matter what the industry group is. This result also supports the idea of a spillover through imported intermediate goods channels. Another interesting question to ask is how demand linkage effect works. Table 9.6 summarizes the results and shows that demand linkage effect is observed in the machinery industry but not in other industry groups. Out of four specifications in the machinery industry, three specifications – columns (5), (6) and (7) – show positive and statistically significant demand linkage effects. These show that demand linkage is positively correlated with plant productivity. On the other hand, in resource-based, labour-intensive, and metal and chemical industry groups, demand linkage is negatively correlated with plant productivity. Tables 9.5 and 9.6 demonstrate that foreign plants have both positive spillover and linkage effects on local plants in the machinery industry, and negative spillover and demand linkage effects in resource-based industries. Import has a positive correlation with local plants’ productivity both in spillover and demand linkage effect channels in all industry groups. This finding, combined with the extent of import activity of a plant (Table 9.4), indicates that more competitive pressure from imports may affect linkage and spillover effects on local plants and may lead to a hypothesis that imported intermediate goods help increase local plants’ productivity.

9.4 Further discussion on spillovers and demand linkage effects, and trade liberalization In the previous section, I have shown that spillover and linkage effects exist in the machinery industry while there are negative linkage effects

Spillovers and Linkages between Local and Foreign Plants 243

in resource-based industries. Estimated coefficients also show a negative sign in labour-intensive and metal and chemical industries although they are not statistically significant. Why do empirical results show the opposite directions regarding the effects of spillover and linkage on plant productivity? One possible explanation is the following: in resource-based, labour-intensive, and metal and chemical industries, such as coke and petroleum products, food products, tobacco, wood products, rubber goods, textiles, wearing apparel, and paper products industries, foreign plants are more likely to compete with local producers in the final goods market than the machinery industry. In other words, the relationship between local and foreign plants is substituted. On the other hand, in machinery industries, such as electric machinery, motor vehicles and computer industries, foreign plants are more likely to purchase from local suppliers probably in the vertical production network. In this case, when foreign plants increase their purchases, local suppliers make use of scale economies and increase productivity. In other words, the relationship between local and foreign plants is complemented. The results in the previous section also indicate that trade status, especially imports, has a strong positive impact on local plant productivity in all groups and for all specifications. In this section I provide more detailed information on both linkage and spillover effects, and then I will discuss the impact of FTA. To examine the results of the previous section carefully, I calculated average values of variables used in the estimation. Each variable is averaged separately according to the type of ownership, i.e. local and foreign. Table 9.7 shows the average values of variables. The ratio in the table indicates the ratio of the average value of foreign to local plants. Combining Table 9.4 with Table 9.7, we have the following findings. The most notable finding is that the difference in output and valueadded per plant between local and foreign plants is much larger in the machinery industry than in other industries. Average output (valueadded) by foreign plants is 20 (17.6) times larger than that of local firms in the machinery industry, while foreign plant output (value-added) ranges from 4.7 (4.2) to 6.9 (5.4) times larger than local plants in resource-based, labour-intensive, and metal and chemical industry groups. The second notable finding is that an average foreign plant size measured by the number of production workers is 7.4 times larger than an average local plant in the machinery industry, a figure that is much greater than for other industry groups. The third notable finding is that about 42 per cent of local plants and more than 90 per cent of foreign plants import intermediate goods from

244

Table 9.7 Averages of variables by ownership Variable

Resource-based Industry Local

Foreign

Machinery Industry Ratio

Local

Foreign

Labour-intensive Industry Ratio

Local

Foreign

Ratio

Metal and Chemical Industry Local

Foreign

Ratio

Labour Productivity

0.36074 (2.1620)

0.980152.717 (3.4999)

0.32593 (0.9889)

0.92097 2.826 (2.9458)

0.20691 (0.7478)

0.355541.718 (0.6177)

0.46460 (4.0918)

1.516973.265 (5.1349)

Capital-Labour Ratio

0.7471 (2.4081)

1.9501 2.610 (7.5135)

0.5650 (2.0504)

1.40892 2.494 (2.8285)

0.3307 (1.1003)

0.6071 1.836 (2.5760)

0.7672 (2.9029)

3.1268 4.076 (8.4093)

Plant Size Plant Age

76.230 (205.743)

322.291 4.228 68.727 (571.713) (150.174)

507.197 (984.273)

7.380 115.758 (290.608)

334.498 2.890 56.026 (724.237) (118.142)

150.833 2.692 (285.218)

12.85 (10.5558)

12.75 0.992 (11.3826)

11.62 (8.9894)

9.85 0.848 (8.6217)

10.74 (9.4784)

10.74 0.999 (8.3707)

12.48 (10.1259)

11.78 0.944 (9.9534)

Export

0.16238 (0.3688)

0.754354.646 (0.4306)

0.16271 (0.3691)

0.81908 5.034 (0.3851)

0.29017 (0.4539)

0.835242.878 (0.3711)

0.12709 (0.3331)

0.612284.818 (0.4874)

Import

0.18070 (0.3848)

0.648573.589 (0.4776)

0.41851 (0.4934)

0.90318 2.158 (0.2958)

0.29717 (0.4570)

0.790382.660 (0.4072)

0.35913 (0.4798)

0.794742.213 (0.4041)

Output Value-added

138.00 (1263.12)

952.40 (4767.10)

6.901

88.06 (461.07)

1815.51 (6244.02)

20.616

73.63 (257.68)

350.50 (898.68)

4.760 134.80 (1068.81)

45.75 (625.15)

247.35 (940.81)

5.406

27.32 (112.85)

481.05 (1903.19)

17.611

24.05 (91.95)

101.92 (251.16)

4.238

39.57 (515.40)

685.22 (2225.48)

5.083

185.67 (655.30)

4.693

Spillovers and Linkages between Local and Foreign Plants 245

the world market which are much higher than the cases of resourcebased, labour-intensive, and metal and chemical industries.6 The last finding is that the average protection rate measured by the ratio of import tariffs and duties to the total imports at 4-digit ISIC is the lowest in the machinery industry which shows 14.9 per cent in all industry groups (this figure comes from Table 9.4). Coupled with the finding about import status, this suggests that the machinery industry enjoys the benefits of trade liberalization. However, this will be discussed in the next section. To explain the difference between the results of machinery and other industry groups on spillover and demand linkage effects, especially a large output difference between local and foreign plants, these insights should be taken into consideration. The most reasonable explanation of this difference in results is that the vertical production network prevails in the machinery industry while the horizontal competition dominates in resource-based and other industries. In order that the vertical network functions well in the industry, demand linkage should be strong and in order to have strong demand linkage, foreign presence should be large in that industry. Table 9.4 already reported that in the machinery industry, the average values of foreign presence and demand linkage are relatively large compared with the case of other industries. For example, foreign presence at 3-digit ISIC of machinery industry is 0.74 while 0.35 in resourcebased industries, 0.42 in labour-intensive industries, and 0.48 in metal and chemical industries. As for the demand linkage, an average linkage in the machinery industry at 3-digit ISIC is 13.3 while it is 6.3 in resource-based, 6.9 in labour-intensive, and 8.7 in the metal and chemical industries. Another important finding is the large difference in size between local and foreign plants in the machinery industry. This, combined with the vertical network hypothesis, shows how the large foreign plants dominate the lion’s share of the market, and have a linkage nexus with small upstream local suppliers in the machinery industry. Since local plants are so small (twenty times smaller than foreign plants in the machinery industry), they cannot survive if they compete with foreign plants in the final goods market. Instead of competing with foreign plants, local plants can survive to provide intermediate goods to downstream large foreign plants in the vertical production network. It is possible to assume that large and influential foreign plants transfer advanced technology and knowledge to small and upstream local plants through the expanding demand for intermediate goods. Since there is a large gap in output size between foreign and local plants in the machinery industry, it is interesting to ask how this gap

246 The Current Status of Indigenous Enterprises

Figure 9.1 Output difference and labour productivity (all local plants except for machinery) Sources: Industrial Surveys (1999, 2000, 2001, 2003), National Statistical Office, Thailand.

affects the local plant’s productivity in different industries. Casual observation suggests that a narrow gap between local and foreign outputs allows a small local plant to catch up with foreign plants while a large gap makes it difficult for a local plant to get benefits from spillovers. Figure 9.1 is a scatter diagram of local plants’ productivity and output difference between local and foreign plants in resource-based, labourintensive, and metal and chemical industries. Output difference is defined as the logarithm of the ratio of the local plant’s output to the average output of foreign plants in each industry calculated at the 4-digit ISIC level.7 Figure 9.2 is a scatter diagram of local plants’ productivity and output difference between local and foreign plants in the machinery industry. A positive correlation between two variables is observed although the slope in Figure 9.2 is flatter than that in Figure 9.1. From the above observation, in the machinery industry, the output difference between local and foreign plants is not as sensitive to the local plants’ productivity as other industry groups. Figures 9.1 and 9.2 clearly show a positive correlation between two variables. However, there are two interpretations of this fact: a catching-up hypothesis and an existence of scale economies in which the larger the size of output, the higher the labour productivity. It is difficult to distinguish between a catching-up hypothesis and the existence of scale economies from the information of these two graphs. Hence, I tested the catching-up hypothesis directly by regressing the labour productivity of each plant on output difference and the square

Spillovers and Linkages between Local and Foreign Plants 247

Figure 9.2

Output difference and labour productivity (local machinery plants)

Sources: Industrial Surveys (1999, 2000, 2001, 2003), National Statistical Office, Thailand.

term of output difference including year and industry effects. If a catching-up hypothesis holds, the coefficient on the difference is positive and the coefficient on the square term of difference is negative. In other words, labour productivity increases as output difference becomes smaller, but the rate of increase in productivity diminishes, i.e. the relation between productivity and output difference shows an inverted-U shape. Table 9.8 reports the estimation results. Salient findings are that all coefficients in all industries have expected signs and are highly statistically significant. The results confirm the hypothesis that a narrower gap in the size of plants is important for local plants to catch up with foreign advanced technology. The results also show that the coefficient on square term of output difference in the machinery industry is smaller than those in other industries. This indicates that in the machinery industry the slope of the curve is flatter than those in other industries. In other words, there is a weaker connection in the machinery industry between productivity and output gap than in other industries. Despite this finding, why does the machinery industry have a positive and strong spillover as well as linkage effects? All these observations suggest the following possible explanation: in the machinery industry, the output gap helps technology transfer in that a vertical production network between small local and large foreign plants prevails and spurs the knowledge spillover from a large foreign to a small local plant through backward linkage effects.

248

Table 9.8

Output difference and plant productivity Total sample (1)

Resource-based (2)

(3)

Machinery (4)

(5)

Labour-intensive (6)

(7)

Metal and Chemical (8)

(9)

(10)

Output 0.2443*** 0.2551*** 0.2747*** 0.2729*** 0.2029*** 0.2411*** 0.1789*** 0.1828*** 0.2828*** 0.3367*** Difference (0.0048) (0.0049) (0.0071) (0.0073) (0.0141) (0.0149) (0.0109) (0.0109) (0.0102) (0.0109) Square of 0.0236*** 0.0229*** 0.0196*** 0.0205*** 0.008*** 0.0038*** 0.044*** 0.0437*** 0.0196*** 0.0145*** Difference (0.0008) (0.0008) (0.0010) (0.0010) (0.0026) (0.0027) (0.0018) (0.0018) (0.0019) (0.0019) Year Effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Industry Yes Yes Yes Yes Yes Effects 3-digit Industry Yes Yes Yes Yes Yes Effects 4-digit N observations 31738 31738 14489 14489 3612 3612 7226 7226 6335 6335 R-squared 0.3709 0.3707 0.4072 0.4073 0.1856 0.1835 0.4461 0.4460 0.3631 0.3629 Notes: Dependent variable is labour productivity of plant. Standard errors are in the parentheses. *** indicates statistical significance at 1 per cent.

Spillovers and Linkages between Local and Foreign Plants 249

9.5

FTA and its impact on local plants

As Tables 9.5 and 9.6 show, export and import activities have strong impacts on plant productivity. In particular, imports have a positive and highly significant coefficient in all industry groups. On the other hand, Thailand has engaged in FTAs with China, Australia, New Zealand and India, and is planning to conclude EPAs with Japan and the USA.8 While Thailand has been involved in AFTA since 1992, the common effective preferential tariff (CEPT) scheme reached the targeted tariff rate (0–5 per cent) among ASEAN5 in 2002. Institutional agreements may spur the international trade of local plants and then raise plant productivity. Liberalization in international trade barriers, no matter whether institutional or natural, affects local plant activity. In this section, I discuss the relationship between FTA and local plant productivity. To estimate the correlation between FTA and plant productivity, I regress the local plant’s productivity in logarithm on the degree of protection that is measured as a share of import tax plus import duty in total imports in industry j. Protection is calculated at the 4-digit ISIC level.9 If industry j is more liberalized, that is, has a lower protection rate, plant productivity increases through the expansion of import and export. Thus, the expected sign of coefficient should be negative. Year effect and industry effect are included in the estimation. Table 9.9 summarizes the results. In the case of a total sample, the coefficient on protection is negative and statistically significant in column (1) but insignificant in column (2). Results with different industry effects do not show consistent evidence. For a total sample case, the correlation between trade liberalization and productivity is, therefore, ambiguous. In machinery and labour-intensive industry groups, coefficients on protection show negative signs and are statistically highly significant for both industry effects cases. It would be safe to say that liberalization by the FTA/EPA has a positive impact on productivity in machinery and labour-intensive industries. This confirms other related literature, listed in the introduction. On the other hand, in resource-based and metal and chemical industry groups, coefficients on protection are positive but some of them are statistically insignificant (columns (3) and (9)), indicating that trade liberalization does not have a definite effect on a plant’s productivity in these industries. In sum, it reveals that the impact of trade liberalization differs from industry to industry.

250

Table 9.9 Trade liberalization and plant productivity Total Sample (1)

Resource-based (2)

(3)

(4)

Machinery (5)

(6)

Labour-intensive (7)

Metal and Chemical (8)

(9)

(10)

Protection 4-digit 0.2821*** 0.0556 0.1297 0.4235*** 0.4915* 0.7306*** 0.6027*** 0.5159*** 0.2310 1.1048*** (0.0570) (0.0855) (0.0948) (0.1328) (0.2515) (0.2667) (0.0837) (0.1500) (0.2997) (0.3991) Year Effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Industry Effects Yes Yes Yes Yes Yes 3-digit Industry Effects Yes Yes Yes Yes Yes 4-digit N observations 31738 31738 14489 14489 3612 3612 7226 7226 6335 6335 R-squared 0.0420 0.0398 0.0353 0.0322 0.0549 0.0557 0.0715 0.0720 0.0333 0.0359

Spillovers and Linkages between Local and Foreign Plants 251

9.6

Conclusion

This study has focused on the empirical analysis of the effects of MNEs on local plants by using plant-level data from Thailand. More specifically, this chapter has examined the spillover and linkage effects of MNEs on local plants. The linkage effect occurs when MNEs purchase input materials from local suppliers (through backward linkage) or MNEs sell intermediate goods to local producers (through forward linkage). However, it is possible for MNEs to affect local plants negatively. This happens when the MNEs compete with local firms in the same market and MNEs take a large share of the market. Empirical results suggest that in the machinery industry, both spillover and linkage effects exist from MNEs to local suppliers while in resource-based industries, there are negative linkage effects. This chapter has also examined the relationship between the output gap of plants and plant productivity. The conclusion supports the casual observation in which narrow output gaps between local and foreign plants are strongly correlated with higher productivity in all industries. In other words, a local plant has more chances to catch up with advanced technology provided by foreign plants when the output gap is small. To explain why only the machinery industry has positive spillover and linkage effects, it is reasonable to refer to the vertical production nature of the industry. It may be assumed that small local plants produce and sell their products to large and downstream foreign plants in the machinery industry. In this case, small local plants can avoid direct competition with foreign plants in the final goods market. In addition, small local plants can benefit from the expanding demand of intermediate goods through the demand linkage effects. I also showed the FTA impact on local plants by examining the effects of trade liberalization. Results suggest that protection is strongly and negatively correlated with a local plant’s productivity in machinery and labour-intensive industries but not in resource-based and metal and chemical industries. This result suggests that FTA/EPA has a favourable impact on productivity improvement in machinery and labourintensive industries but the effects of FTA/EPA vary according to the type of industry. A simple implication derived from these findings is that establishing competitive suppliers in the upstream in an industry such as the machinery industry is crucial to bear fruits of FTA/EPA.

252 The Current Status of Indigenous Enterprises

Appendix Table 9.10

ISIC industry codes (Rev. 3)

2 digit

Division of Industry

15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

Manufacture of food products and beverages Manufacture of tobacco products Manufacture of textiles Manufacture of wearing apparel dressing and Tanning and dressing of leather manufacture Manufacture of wood and of products of wood Manufacture of paper and paper products Publishing, printing and reproduction of records Manufacture of coke, refined petroleum products Manufacture of chemicals and chemical products Manufacture of rubber and plastic products Manufacture of other non-metallic minerals Manufacture of basic metals Manufacture of fabricated metal products Manufacture of machinery and equipment n.e.c. Manufacture of office, accounting and computing Manufacture of electrical machinery Manufacture of radio, television Manufacture of medical, precision and optical Manufacture of motor vehicles, trailers Manufacture of other transport equipment Manufacture of furniture manufacturing n.e.c.

Acknowledgements I would like to thank the editors and two anonymous referees for helpful comments and suggestions. All remaining errors are, of course, mine.

Notes 1. Two important theoretical articles on backward linkage effects from MNEs to local firms are Rodriguez-Clare (1996) and Markusen and Venables (1999) which construct a model in the monopolistic competition framework. See Lin and Saggi (2007) for a recent survey of linkage effects. 2. Javorcik (2004) analyses the impact of MNE presence on total (MNEs and local) firms’ productivity, while this chapter focuses on the effects of MNE presence on the local firms. 3. I use plant-level, not firm-level, data from industrial surveys of Thailand. Although, to be precise, a foreign plant is not equal to an MNE, I use the terms ‘foreign plant’ and ‘MNE’ interchangeably hereafter.

Spillovers and Linkages between Local and Foreign Plants 253 4. See Appendix Table 9.10 for these sector classifications. I exclude sector 37: recycling, from my analyses, because some other main data sources do not have the recycling sector. 5. It is possible that a local firm’s output expands due to the exiting of inefficient other local firms after the entrance of MNEs into the market. However, exit data are not available for Thailand, so I have to exclude this possibility from analyses in this chapter. 6. Export (import) is a binary variable which takes 1 if the plant exports (imports), zero otherwise. Hence the average values for export and import in Table 9.7 show the percentage of the plants that export or import in the industry group.   Y 7. The output difference is calculated as ln Y i;j;t where Yi,j,t equals output of MNE;j;t

plant i, industry j, and year t, while YMNE,j,t equals an average of foreign plant outputs in industry j, at year t. 8. On 19 September 2006, a military coup took place. Under a new interim cabinet with Prime Minister Surayud Chulanont, many mega-projects and ongoing negotiations of EPAs may be shelved for at least one year. 9. I also use 3-digit data to calculate the share of import tax and import duty, but there is multi-colinearity between 3-digit protection and industry effects. Thus the results at 3-digit protection are not reported here.

References Amiti, M. and J. Konings (2005). ‘Trade Liberalization, Intermediate Inputs, and Productivity: Evidence from Indonesia’, IMF Working Paper. Fernandes, A. M. (2007). ‘Trade Policy, Trade Volumes and Plant-level Productivity in Colombian Manufacturing Industries’, Journal of International Economics, 71: 52–71. Javorcik, B. S. (2004). ‘Does Foreign Direct Investment Increase the Productivity of Domestic Firms? In Search of Spillovers through Backward Linkages’, American Economic Review, 94, 3: 605–27. Kugler, Maurice (2006). ‘Spillovers from Foreign Direct Investment: Within or Between Industries?’ Journal of Development Economics, 80: 444–77. Lin, P. and K. Saggi (2005). ‘Multinational Firms and Backward Linkages: a Critical Survey and a Simple Model’, in M. Blomstrom, E. Graham and T. Moran (eds), Does Foreign Direct Investment Promote Development? Institute for International Economics, Washington, D.C. Lin, P. and K. Saggi (2007). ‘Multinational Firms, Exclusivity, and Backward Linkages’, Journal of International Economics, 71: 206–20. Markusen, J. R. and A. Venables (1999). ‘Foreign Direct Investment as a Catalyst for Industrial Development’, European Economic Review, 43: 335–56. Milner, Chris, Geoff Reed and Pawin Talerngsri (2004). ‘Foreign Direct Investment and Vertical Integration of Production by Japanese Multinationals in Thailand’, Journal of Comparative Economics, 32: 805–21. Muendler, Marc-Andreas (2004). ‘Trade, Technology, and Productivity: a Study of Brazilian Manufacturers’, http://repositories.cdlib.org/ucsdecon/2004-06 Olley, S and A. Pakes (1996). ‘The Dynamics of Productivity in the Telecommunications Equipment Industry’, Econometrica, 64: 1263–97.

254 The Current Status of Indigenous Enterprises Pavcnik, N. (2002). ‘Trade Liberalization, Exit, and Productivity Improvement: Evidence from Chilean Plants’, Review of Economic Studies, 69, 1: 245–76. Rodriguez-Clare, A. (1996). ‘Multinatinals, Linkages, and Economic Development’, American Economic Review, 86: 852–73. Topalova, Petia (2004). ‘Trade Liberalization and Firm Productivity: the Case of India’, IMF Working Paper WP/04/28. Urata, S. and K. Yokota (1994). ‘Trade Liberalization Policies and Productivity Change: a Case of Thailand’, Developing Economies, 32, 4: 444–58.

10 Do Local Suppliers Benefit from Economic Integration? Evidence from Outward FDI from ASEAN Daisuke Hiratsuka

10.1

Introduction

In East Asia, goods, services, capital, investment and knowledge are moving more freely than before, and firms have expanded their economic activities across borders. What we have to pay attention to is that, in economic integration, segment markets are integrated and economies of scale work stronger than ever before. Thus, we need to ensure that benefits from economic integration are spread more evenly. There are two concerns, in particular, that arise for developing countries. One is whether or not developing countries can benefit from economic integration. Another is whether or not indigenous firms can benefit from economic integration. What is of particular concern among academics as well as producers and policy-makers is whether the indigenous enterprises in ASEAN countries have benefited from economic integration. This is a legitimate concern, since multinational enterprises (MNEs), with many tangible and intangible assets, have an advantage in production operations with economies of scale, but indigenous technical levels remain low in ASEAN countries (Hiratsuka, 2003). The previous chapter discussed this issue, from the viewpoint of linkages between MNEs and indigenous firms, and concluded that indigenous firms get the linkage effect when MNEs purchase input materials from local suppliers or MNEs sell intermediate goods to local producers. Meanwhile it is possible for MNEs to crowd out indigenous suppliers when both compete in the local market. It is uncertain whether indigenous firms in developing countries benefit from economic integration. How should this issue be examined? Outward foreign direct investment (FDI) from developing countries may be one of the measures to indicate the performance and capability of developing 255

256 The Current Status of Indigenous Enterprises

country enterprises in economic integration where border barriers are eliminated. Reduction of border barriers provides enterprises with the opportunity to reorganize their economic activities on various geographical scales. It is commonly known that MNEs have expanded their business activities all over the world but developing country enterprises are also expanding their business wings, organizing production and distribution networks on sub-region and region-wide scales. In fact, developing country transnational corporations (TNCs)1 are increasing in importance in the global economy. FDI from developing countries has been remarkable. In particular, these enterprises from Asia are emerging sources of FDI; in 2004 Asia and Oceania contributed more than four-fifths of outward FDI from developing countries (UNCTAD, 2005: 71). This phenomenon indicates that developing country Asian enterprises have succeeded in global market places, and have elevated their status from domestic market players to global ones. Among East Asian economies, South Korea and Taiwan have been major sources of outward FDI, and in recent years, Chinese firms have extended their sales networks to overseas, even as far as Africa through Chinese networks (UNCTAD, 2006: 117–21). These facts indicate that enterprises in Northeast Asian economies are innovative, and have entrepreneurship so that they can expand business beyond borders. In the context of economic development, this can be interpreted as gains due to a huge home market effect or the early start of industrialization. However, it is worrying that ASEAN enterprises, compared to those in Northeast Asia, have only rather weak linkages between MNEs and indigenous suppliers. With this awareness, this chapter aims to examine whether or not ASEAN country enterprises are extending their business activities to overseas markets within the context of economic integration where ASEAN country markets are being integrated through the ASEAN free trade area (AFTA), and with external countries such as China, Japan, South Korea, Australia and New Zealand through FTA. This chapter looks at how ASEAN indigenous enterprises are expanding beyond their borders, like the Northeast Asian enterprises, in economic integration where economies of scale work strongly. The next section will provide an analytical framework to discuss why TNCs in developing countries are increasing remarkably in the context of location theory. The ‘market-seeking’ and ‘efficiency-seeking’ forces of FDI, both major forces for attracting FDI in developed as well as developing countries, are discussed by incorporating the concepts of the home market effect and trade/transport costs (broadly defined)2 respectively. Section 10.3 presents the status quo of outward FDI from ASEAN.

Do Local Suppliers Benefit from Economic Integration?

257

Section 10.4 focuses on intra-regional FDI in ASEAN, and shows that intra-industry FDIs are growing in ASEAN and that ASEAN country TNCs have spread to the least developed countries (LDC) of Laos, Cambodia and Myanmar. The final section summarizes this study and presents policy implications.

10.2 10.2.1

What determines outward FDI? Transport costs

What does determine outward FDI? FDI is an issue related to the location of industry by firms. Krugman (1991) gives us an insight into the trade and investment activities of firms in economic integration where goods, capital and persons move easily across borders. Krugman argues that the interaction of market, transport costs and fixed investment costs determines the location of industry. Surprisingly, one businessman’s response to my question was perfect in the light of Krugman’s argument. He said that in deciding whether or not to expand his business to a new location (in this instance Shinzen) he would need to consider the expected market/ production size in China, transport costs from Johor, Malaysia to Shenzhen, and the difference in cost between a new fixed investment in Shinzen and an expansion of investment in Johor. At that time, his major customer, Hitachi Global Storage Technologies, was planning to start a hard disc drive production in Shinzen in early 2006 and was requesting its suppliers to come there, to reduce trade costs and time. Krugman’s hypothesis regarding the location of industry is based on the gravity framework in which economic size (market) and geographical distance (as a proxy of transport costs) play significant roles in trade and investment. Geographical distance determines trade and investment through transport costs (broadly defined). As defined by Anderson and Wincoop (2004), transport costs include all the costs incurred in getting goods to a final user other than the marginal costs of producing the goods.3 For purposes of estimation based on the gravity model, they divided transport costs into narrowly defined transport costs (both freight costs and shipping time costs), policy barriers (tariff and non-tariff barriers), other border-related barriers such as institutional, language and currency barriers, and domestic distribution costs. Transport costs (broadly defined) do play a significant role in determining the location of industry. If the transport costs of goods are too high, the industry will be located at the nearest large market, because firms wish to have lower transport costs. Indeed, Hattari and Rajan

258 The Current Status of Indigenous Enterprises

(2007) obtained the estimated results that common border and common language have a positive relationship with intra-Asia bilateral FDI flows, and distances have a negative one, meaning that transport costs affect outflows of FDI in the region.4 The presence of transport costs gives rise to the ‘home market effect’ where suppliers located near a large market can attain economies of scale and export the goods. In Krugman’s words, ‘the countries will tend to export those goods for which they have relatively large domestic markets’ (Krugman, 1980: 950). The supposition is that when export expands until total transport costs become large, suppliers will move their production facilities to a nearby overseas market, in order to reduce operating costs. Otherwise, suppliers would lose their market due to global competition. This can be interpreted as international trade being partly substituted with ‘market-seeking’ FDI. When transport costs (broadly defined) fall, what consequences can be expected for production location? Krugman and Venables (1995) argue that when transport costs fall to a point such that the advantage of low wages in the ‘periphery’ (unindustrialized economy) offsets the disadvantages of being remote from markets and suppliers of the ‘core’ (industrialized economy), manufacturing in the core will move to the ‘periphery’. To put it differently, ‘efficiency-seeking’ FDI will take place when transport costs fall until lower wages offset the disadvantages of other conditions. However, the industry tends to concentrate on specific countries/ cities. In a world where transport costs were zero, firms could locate anywhere they wished. However, even in this case, in reality, industries tend to concentrate in some specific cities in order to share externalities such as skilled human resources. For instance, software design is an industry in which transport costs are substantially zero, since software engineers can exchange information mutually and deliver products to the head office via the Internet instantly. Nevertheless, the industry concentrates in cities where there is a location advantage in terms of human resources. To be precise, the industry-specific TNCs will outsource some processes to developing countries where location advantages exist. 10.2.2

Driving forces and types of FDI

What are the driving forces of outward FDI? Roughly speaking, there are three primary drivers of outward FDI in the context of transport cost and home market: 1.

Driving force 1: to take advantage of resources in host countries (resource-seeking FDI);

Do Local Suppliers Benefit from Economic Integration?

2. 3.

259

Driving force 2: to take advantage of low wage rates (and, more generally, low factor prices) in host countries (efficiency-seeking FDI); Driving force 3: to have better access to the markets of host countries and nearby countries (market-seeking FDI).

‘Efficiency-seeking’ FDI can be regarded as one type of ‘resourceseeking’ FDI since it seeks abundant labour resources. However, since ‘efficiency-seeking’ FDI has special implications for the location of industry which differ from other ‘resource-seeking’ FDI, the two are distinguished here. When developed country TNCs set up their overseas production plants in developed countries where markets are large, the ‘market-seeking’ force works. A typical case is the European Union (EU) country TNCs that seek markets within the region. In contrast, when developed country TNCs set up their overseas production plants in developing countries, ‘efficiency-seeking’ forces are mainly at work because there is a large difference in wages between developed and developing countries. Strategic functions such as headquarters, research & development (R&D), and highly capital-intensive manufacturing processes are located in the home country where highskilled workers are available, while most of the manufacturing activities are located in host countries where wage rates are low. When a developing country has a large market, the two driving forces of ‘efficiency-seeking’ and ‘market-seeking’ FDIs work together so that developed country TNCs locate production in the host country (Fujita, 2007). A typical example is China where markets are large and wage rates are low, attracting both ‘efficiency-seeking’ and ‘market-seeking’ FDIs from developed countries. ‘Market-seeking’ FDIs are taken in goods for which transport costs are high due to high tariff rates. 10.2.3

‘Market-seeking’ FDI from ASEAN

We are primarily concerned with outward FDIs from developing countries, in particular, ASEAN countries. It seems there is no difference between developed country and developing country TNCs. ‘Marketseeking’ forces work very strongly for ASEAN TNCs. Thanks to the ‘home market effect’, giant ASEAN firms are very active in taking ‘marketing-seeking’ FDIs. A striking example of a market-seeking FDI is the Philippines giant, San Miguel. San Miguel operates breweries in Vietnam, Australia and Indonesia and maintains four breweries in China, including in Hong Kong. Overseas packaging facilities include a glass plant and metals-crown plants in China and Vietnam as well as plastics

260 The Current Status of Indigenous Enterprises

plants in China and Indonesia. There is also the Charoen Pokphand (CP) Group, a leading Thai agro-based company, which has expanded into Bangladesh, Cambodia, China, Hong Kong, India, Indonesia, Malaysia, Myanmar, Singapore, Taiwan, Turkey and Vietnam. The CP Group started seeds businesses, then through vertical integration, moved into feed mills, poultry, swine, duck breeding and processing. As opportunities developed, the CP Group extended its reach, first throughout Asia, and later throughout the rest of the world. It focused on the development needs of people, expanding not only its selection of foods but also the logistics and retail distribution links, such as convenience stores, supermarkets and hypermarkets as well as casual dining outlets, thus introducing its products on an international level.5 Not only big businesses but also small family businesses have succeeded in global market places, serving customers in Asia-Pacific and world wide. LKT, a semiconductor business equipment solution company, based in Penang, Malaysia, has opened a production facility in Thailand and service support centres in China, Costa Rica, Singapore, Thailand and the United States. Pentamaster, a semiconductor manufacturing automation solution company in Penang, has opened offices in Ireland, Germany and the United States to support customers and distributors in China, Taiwan, the Philippines and Singapore. These ASEAN TNCs, whether large or small and medium enterprises (SMEs), are expanding their businesses beyond their national boundaries, to neighbouring ASEAN countries, then to China, the United States, Australia and the United Kingdom – wherever they can find markets. 10.2.4

‘Efficiency-seeking’ FDI from ASEAN

ASEAN country TNCs are expanding their overseas business by ‘efficiency-seeking’ FDIs, in the way that developed country firms have done. Indeed, a large number of Singapore TNCs went to neighbouring economies. Malaysia and other country TNCs followed. Felda and KL Kepong, both Malaysian firms with objectives to attain economies of scale, went to Indonesia where factor prices are cheap to cultivate palm trees and import them back to Malaysia. Those products are mainly exported to China and other countries through their international distribution networks. Correspondingly, Thai sugar refinery TNCs cultivate and refine sugarcane in Laos where factor prices are low, and additionally, the language and the culture are similar. Refined sugar is imported back to Thailand to distribute all over the world. Usually these FDI host countries provide cheaper labour than FDI source countries. However, for ‘efficiency-seeking’ purposes, ASEAN TNCs usually

Do Local Suppliers Benefit from Economic Integration?

261

had overseas operation facilities in neighbouring ASEAN countries and did not go to African countries because transport and communication costs would be high due to the geographical distance. 10.2.5

‘Supplier following assembler’ FDI and other new types of FDI

FDI in intermediate goods is one of the characteristics of FDI in East Asia where so-called production fragmentation has developed. More specifically, multinational enterprises slice manufacturing processes into several stages and locate them in different countries according to location advantage, and/or they purchase parts from different countries, mainly from the region. When developed country TNCs, mostly multinationals, expand production-seeking to explore new frontiers of lower wage countries, local suppliers in the host countries follow foreign assemblers, moving to new frontiers. This type of FDI can be called ‘supplier following assembler’ FDI, which has both aspects of ‘efficiency-seeking’ and ‘market-seeking’ FDI in intermediate goods, and which reflects a close linkage between foreign assemblers and local suppliers. In ASEAN, communications equipment is an industry with a developed international production network. Figure 10.1 shows the international procurement of the hard disc drive assembler, Hitachi Global

Thailand SPNDLE MOTOR: NIDEC, NMB BA SE: W EARN S CARR IAG E: FUJIKUR A, SANEI FLEX CA BLE: NOK PIVO T: NMB SE AL: KO KO KU VCM: SMT, TDK TOP COVER: NOK, NHK PCBA: SS CI HG A:UTC HAS :UTC

Malaysia BASE: KENSEISHA PIVOT: NSK SPACER: KENSEISHA VCM:SHINETSU BASE: ASAHI CARD:SCI, CELESTICA TOP CLAMP: SCL, CELEST ICA, SOODE DI SK (Media): KOMAG

HEAD: GDL

Japan

Taiwan

Hong Kong FILTER CAP: GML

COVER: NHK DISK:HOY A SCREW: KATAYAMA SE AL: KO KO KU, TOKA I RAMP: DAICH I TO P CLAMP: BPI, SOODE LAT CH : NOK PLATE CASE: BRIDGESTONE LABEL: SANSAI FILTER: SSRETEK PCBA: SSCI SUSPENSION: NH K

Philippines DAMPING PLATE: IMEAS CO IL SUPPO RT: TO TOKU PCBA : IO NIX

Singapore

PIVOT: NSK PC ADP : JST DISK (Med ia): HOYA

Figure 10.1

China PCBA: GBM, SSCI CARR IAGE:TDK HG A: HSPC BASE: BPI HEAD: HG ST SUSPENSION:NHK

TOP CLAMP: G ML

COVER: KURODA, CHEUNGWOH SCREW: TIORMAC, SPURWAY

Mexico

USA DISK : HG ST HEAD: HGST SUSPENSION: HTI

Indonesia W.SUSPENSION: SUMITOMO VCM: SHINETSU PCBA: SOLECTION

Parts procurement of a hard disc drive assembler located in Thailand

Source: Compiled by author, based on interviews at Hitachi Global Storage Technology (Thailand) in August 2005.

262 The Current Status of Indigenous Enterprises

Storage Technology of Thailand (HGST). HGST purchases various parts and components from overseas. ASEAN local suppliers have participated in the production networks, although most of their first-tier suppliers are Japanese affiliates operating in East Asia, mainly in the ASEAN countries. The ASEAN local suppliers, for instance Engtek, a components manufacturer for the disc drive industry based in Penang, established factories in Dongguan in 1996, the Philippines in 1997, and Thailand in 1998 in order to meet the demands of its largest customer, Fujitsu, which has been aggressively expanding its market. Eventually, Engtek elevated its status from a mid-size Malaysian company to become a global player (ENG Teknologi Holding, 2000). 10.2.6

‘Function-seeking’ FDI

TNCs, both developed and developing, have explored location-specific advantages, which are different country by country, and more precisely, city by city. Singapore has offered strong financial service skills and excellent infrastructure and targets leading-edge offshore functions such as remote robotics management, health care and genetic diagnostics (A. T. Kearney, 2004). It has become one of the key hubs for regional headquarters (UNCTAD, 2004: 204). In fact, Malaysian, Thai and other ASEAN TNCs went to Singapore, and some to Hong Kong, to set up functions for international trade and financing. In this regard, research and development centre-seeking FDI has become a new phenomenon. Developing country firms are also setting up R&D activities abroad to access these foreign markets and centres of excellence (UNCTAD, 2005: 189). Singapore plays a role as a regional design centre, and Penang is a training centre for ASEAN countries where training costs are cheaper than in Singapore. Some ASEAN country TNCs went to the United States and India where knowledgeable human resources are available.

10.3 10.3.1

Pattern of outward FDI from ASEAN Outward FDI patterns

What features does outward FDI from ASEAN have? To answer this question, first of all, outward FDI patterns are examined as compared to inward FDI. Figure 10.2 shows outward and inward FDI patterns of EU15, NAFTA, East Asian developing economies, and ASEAN. The large marks indicate figures in 2004 and the small ones 2000, respectively. Interestingly, the

Do Local Suppliers Benefit from Economic Integration?

263

Inward stocks US$ billion

6,000

4,000

EU 15

East Asia NAFTA 2,000

ASEAN10 0

0

2,000

4,000

6,000

Outward stocks US$ billion

Figure 10.2 FDI inward and outward stock by region (2004, US$ billion) Note: East Asia includes ASEAN10, China, Japan, Korea, Hong Kong and Taiwan. Source: Compiled by author from UNCTAD FDI Online (http://stats.unctad.org/FDI/Table Viewer/tableView.aspx).

EU’s outward FDI stock was larger than its inward FDI stock and increased rapidly from 2000 to 2004. Outward FDI stock of the EU in 2000 was larger than that of NAFTA in 2004, while the inward FDI of both regions was at a similar level. The large EU outward FDI was due to intra-regional FDI. Intra-regional FDI in the EU amounted to 66 per cent of total EU FDI inflows in 2004, which was the same figure as intra-regional trade. How can the high intra-regional FDI be explained? The gravity model, in which location is determined by economic size (or market) and geographical distance (as a proxy of broadly defined transport costs), tells us that the EU would be the largest trade and FDI area because it has the second largest market in the world, and transport costs between the EU nations are low due to geographical proximity and reduction of border barriers in trade and investment, as well as the well-developed highway networks. NAFTA has a larger economic size than the EU. However, there are only three countries, and therefore both intra-regional trade and FDI flows of NAFTA are smaller compared to those of the EU. Looking at NAFTA, in 2000, its outward FDI was smaller than inward FDI. However, in 2004, its outward FDI became larger than its inward FDI. These facts indicate that outward FDI increases inward FDI as industrialization advances. This law seems to be applied to the ASEAN countries. For instance, Singapore is the largest FDI source country,

264 The Current Status of Indigenous Enterprises 180

Inward FDI stocks in 2004, US$ billion

Singapore 150

120

90

60

Thailand Malaysia Vietnam

30

Philippines 0

Indonesia 0

30

60

90

120

150

180

Outward FDI in 2004, US$ billion

Figure 10.3 Inward and outward FDI stocks of the ASEAN countries (2004, US$ billion) Source: Compiled by author from UNCTAD FDI Online (http://stats.unctad.org/FDI/TableViewer/tableView.aspx).

followed by Malaysia (Figure 10.3). However, the rest of the ASEAN countries are still minor sources of FDI. Outward FDI from Thailand and the Philippines is increasing but is still very small (Figure 10.4). 10.3.2

Geographical distribution of outward FDI from ASEAN

What geographical distribution patterns are observed for outward FDI from ASEAN? The geographical distribution pattern is subject to the gravity framework that market and geographical distance determine trade and investment. Figure 10.5 shows the share of FDI stock abroad from Malaysia by geographical destination and the total trade values of the destination country, which is a proxy of economic size. Malaysia took FDI to Singapore, 15.1 per cent of its total outward FDI, higher than that to the United States, which made up 14.1 per cent of the total. The same figure in Indonesia and Thailand is larger than those figures in Taiwan and Canada where their trade sizes are larger than Indonesia and Thailand. This evidence supports the fact that gravity forces, that is, geographical distance and market, are crucial determinants of outward FDI. In particular, geographical distance plays a significant role in the geographical distribution of outward FDI.

Do Local Suppliers Benefit from Economic Integration?

265

12,000 10,000

1980 1990

8,000 US$ million

2000 2002

6,000

2004 4,000 2,000 0

Indonesia

Malaysia

Philippines

Singapore

Thailand

2,000

Figure 10.4

Outward FDI flows of the ASEAN countries (US$ million)

Share of FDI stocks abroad, by geographical destination, % in 2004

Source: Compiled by the author from UNCTAD FDI Online (http://stats.unctad.org/FDI/ TableViewer/tableView.aspx).

16.0 Singapore

United States

14.0 12.0 10.0 8.0 Hong Kong

6.0 4.0 2.0 0

Indonesia China

Thailand Taiwan Vietnam 0 500

Canada

Japan

1,000

1,500

2,000

2,500

Total trade, US$ billion in 2004

Figure 10.5 Share of FDI stock abroad from Malaysia by geographical destination Source: Compiled by author: for FDI from UNCTAD and from IMF, Direction of Trade, CD ROM 2005.

Thailand provides another good example of developing country TNCs going to neighbouring countries. Outward FDI from Thailand to ASEAN countries amounted to 35 per cent of the total outward FDI of Thailand in 2004 (see Figure 10.6). Among them, FDI to Singapore accounted for

266 The Current Status of Indigenous Enterprises

70.0 60.0 50.0

(%)

40.0 30.0 20.0 10.0 0.0

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 East Asia, 62.0 United States, 12.8

ASEAN, 35.4 European Union, 6.8

Figure 10.6 Share of FDI stock abroad from Thailand by geographical destination (1995–2004, %) Source: Compiled by author from UNCTAD.

14.3 per cent, followed by the Philippines 7.1 per cent, Vietnam 4.9 per cent, Indonesia 3.5 per cent, Cambodia 2.0 per cent, Myanmar 1.5 per cent, Laos 1.4 per cent, and Malaysia 0.8 per cent. This indicates that, due to geographical proximity, ASEAN TNCs expanded to neighbouring countries where they have a comparative advantage in finding a market and performing cheap labour operations, due to low transport and communication costs with geographical proximity and similar language, culture and customs. This is the same reason that Japanese FDI went to Taiwan first, then to South Korea and Hong Kong in the 1960s and the 1970s for ‘efficiency-seeking’ and ‘marketseeking’: geographical proximity and the same language. Many Koreans, Hong Kong Chinese, and Taiwanese spoke Japanese at that time. It was in the late 1970s that Japanese firms expanded production facilities to Singapore where American semiconductors operated at that time and suppliers were concentrated. More importantly, Thai TNCs have extended their business to East Asian (ASEAN, China and Hong Kong) countries. The share of Thailand’s outward FDI directed to East Asia increased from 52 per cent of its total

Do Local Suppliers Benefit from Economic Integration?

267

70.0 60.0 50.0

(%)

40.0 30.0 20.0 10.0 0.0

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 East Asia

ASEAN

United States

European Union

Africa

Latin America

Figure 10.7 Share of FDI stock abroad from Malaysia by geographical destination (1995–2004, %) Source: Compiled by author from UNCTAD.

outward FDI in 1995 to more than 62 per cent in 2004. On the other hand, the share to the United States decreased from 23 per cent in 1995 to 13 per cent in 2004, and the share to the EU decreased slightly from 10 per cent to 7 per cent. The fact that 60 per cent of outward FDI from East Asia went to East Asia means that Thailand TNCs are doing business in neighbouring economies on a regional scale. This is also true for Malaysia and Singapore, as will be discussed later. 10.3.3

Diversification of geographical distribution

Outward FDI flow from Malaysia, more advanced than Thailand, shows a slightly different pattern to that of Thailand (see Figure 10.7). The share of outward FDI from Malaysia to ASEAN decreased remarkably from 32.3 per cent in 1995 to 23.2 per cent in 2004, and its share to East Asia also decreased from 58.8 per cent in 1995 to 34.2 per cent in 2004. On the other hand, the same figure to the United States increased from 9.1 per cent to 14.1 per cent in the same years. In addition, outward FDI from Malaysia to Africa grew from 1.2 per cent to 4.5 per cent.

268 The Current Status of Indigenous Enterprises

The decrease of outward FDI to ASEAN and East Asia is also seen in Singapore. The share of outward FDI from Singapore to ASEAN decreased from 32.2 per cent in 1994 to 21.9 per cent in 2003, and the same figure to East Asia decreased from 56.4 per cent to 47.0 per cent. Diversification in the direction of outward FDI in Malaysia and Singapore implies that developing country TNCs expand their business on a regional scale, but extend more on a global scale as they grow. A good example at the firm level is Ingenuity Solutions (Malaysia), which has targeted the knowledge base of developed countries such as the United States when investing in R&D abroad (UNCTAD, 2005: 182). The company has located its software development centre in India, while its representative office in the United States makes it possible to respond to the needs of its US-based customers.

10.4

Intra-regional FDI in ASEAN

ASEAN has launched the AFTA which, through a Common Effective Preferential Tariff (CEPT) scheme, is designed to attract FDI from abroad through improvement of the trade and investment environment in ASEAN. By the beginning of 2003, the ASEAN6, that is, the original six member states (Brunei, Indonesia, Malaysia, the Philippines, Singapore and Thailand) applied AFTA’s tariff level of 0–5 per cent to 98.4 per cent of a total of 44,060 products. ASEAN has determined to reduce the tariff level to 0 per cent by 2010. Vietnam is expected to achieve 0–5 per cent AFTA compliance in 2006, Laos and Myanmar in 2008, and Cambodia in 2010, and these four new members will eliminate all imported duties by 2012. The question then arises as to whether ASEAN country enterprises can expand their businesses beyond their home countries to within the ASEAN region. Which countries can be source and host, and in what industries are ASEAN country enterprises furthering their integration? 10.4.1

Who are sources and hosts?

Intra-regional FDI in ASEAN is not a new phenomenon. As discussed in the previous section, ASEAN TNCs have moved into neighbouring ASEAN countries. The question then arises as to which country is the source and which is the host of intra-regional FDI. In other words, what patterns does intra-regional FDI in ASEAN have? Tables 10.1, 10.2 and 10.3 summarize cross-border FDI, balance of payment base, within ASEAN countries between 1995 and 2003, compiled

Table 10.1 Intra-ASEAN FDI flows by source and host country (1995–2003, US$ million) Host country

BRN

CAM

INE

LAO

MAL

MYA

Brunei (BRN) Cambodia (CAM) Indonesia (INE) Laos (LAO) Malaysia (MAL) Myanmar (MYA) Philippines (PHL) Singapore (SIG) Thailand (THL) Vietnam (VET)

– – 55 – 205 – 4 1109 8 0

Total

1380

PHL

SIG

THL

VET

Total

– – – – – – – – – –

27 0 1 0 328 0 13 1401 140 0

– 0 – – 95 0 – 10 151 4

298 3 309 0 0 0 92 6082 186 38

– – 30 – 65 – 4 746 194 –

– – 39 – 85 – – 1086 29 –

176 5 3157 1 2830 45 43 0 303 13

0 9 43 5 134 2 29 5616 0 3

– 1 59 6 492 – 49 1708 382 –

448 17 3691 12 4234 48 234 17760 1393 58

–

1857

261

7009

1039

1239

6575

5840

2696

27894

Source country

Source: Compiled by author from ASEAN Secretary FDI Data, ‘Statistics of Foreign Direct Investment in ASEAN’, Seventh Edition, 2005.

269

270

Table 10.2 Share of intra-ASEAN FDI flows by source country (1995–2003, %) Host country

BRN

CAM

INE

LAO

MAL

MYA

PHL

SIG

THL

VET

Total

Source country Brunei (BRN) Cambodia (CAM) Indonesia (INE) Laos (LAO) Malaysia (MAL) Myanmar (MYA) Philippines (PHL) Singapore (SIG) Thailand (THL) Vietnam (VET) Total

– – 4.0 – 14.8 – 0.3 80.4 0.5 0.0

– – – – – – – – – –

–1.4 0.0 0.1 0.0 17.7 0.0 0.7 75.5 7.6 0.0

– 0.0 – – 36.4 0.0 – 3.9 58.1 1.6

4.3 0.0 4.4 0.0 0.0 0.0 1.3 86.8 2.7 0.5

– – 2.9 – 6.3 – 0.4 71.8 18.7 –

– – 3.1 – 6.9 – – 87.7 2.3 –

2.7 0.1 48.0 0.0 43.0 0.7 0.7 0.0 4.6 0.2

0.0 0.2 0.7 0.1 2.3 0.0 0.5 96.2 0.0 0.0

– 0.0 2.2 0.2 18.3 – 1.8 63.4 14.2 –

1.6 0.1 13.2 0.0 15.2 0.2 0.8 63.7 5.0 0.2

100.0

–

100.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

Source: Compiled by author from ASEAN Secretary FDI Data, ‘Statistics of Foreign Direct Investment in ASEAN’, Seventh Edition, 2005.

Table 10.3 Share of intra-ASEAN FDI flows by host country (1995–2003, %) Host country

BRN

CAM

INE

LAO

MAL

MYA

PHL

SIG

THL

VET

Total

Brunei (BRN) Cambodia (CAM) Indonesia (INE) Laos (LAO) Malaysia (MAL) Myanmar (MYA) Philippines (PHL) Singapore (SIG) Thailand (THL) Vietnam (VET)

– – 1.5 – 4.8 – 1.5 6.2 0.5 0.3

– – – – – – – – – –

6.0 0.0 0.0 0.0 7.7 0.0 5.7 7.9 10.1 0.0

– 0.2 – – 2.2 0.1 – 0.1 10.9 7.4

66.7 15.8 8.4 1.3 0.0 1.0 39.5 34.2 13.4 65.5

– – 0.8 – 1.5 – 1.6 4.2 13.9 –

– – 1.0 – 2.0 – – 6.1 2.0 –

39.3 28.1 85.5 11.2 66.8 95.4 18.6 0.0 21.8 22.4

0.0 52.3 1.2 39.4 3.2 3.5 12.3 31.6 0.0 4.4

– 3.6 1.6 48.1 11.6 – 20.9 9.6 27.4 –

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Total

4.9

–

6.7

0.9

25.1

3.7

4.4

23.6

20.9

9.7

100.0

Source country

Source: Compiled by author from ASEAN Secretary FDI Data, ‘Statistics of Foreign Direct Investment in ASEAN’, Seventh Edition, 2005.

271

272 The Current Status of Indigenous Enterprises

from the ASEAN Secretary FDI data, ‘Statistics of Foreign Direct Investment in ASEAN’, Seventh Edition, 2005. Singapore was the largest FDI source and host country of intra-ASEAN FDI and 63.7 per cent of source FDI was from Singapore. More than 34 per cent of outward FDI from Singapore was directed to Malaysia, followed by Thailand 32 per cent, Vietnam 10 per cent, and Indonesia 8 per cent. Singapore’s intra-regional FDI in ASEAN was concentrated in these four countries. Malaysia, Indonesia and Thailand were the second (15.2 per cent), third (13.2 per cent) and fourth (5.0 per cent) largest source of intraASEAN FDI, respectively. Interestingly, 66.8 per cent of outward FDI from Malaysia and 85.5 per cent from Indonesia were directed at Singapore. On the other hand, outward FDI from Thailand was directed at various countries: 27.4 per cent to Vietnam, 21.8 per cent to Singapore, 13.9 per cent to Myanmar, 13.4 per cent to Malaysia, 10.9 per cent to Laos, 10.1 per cent to Indonesia, and so on. Why have Malaysian and Indonesian enterprises sent so much FDI to Singapore? Singapore performs a function of intermediary trade where goods are transported from neighbouring countries to Singapore for trade and then shipped all over the world. Foreign trade payments are carried out in Singapore, and surplus dollars are operated there. For this function, neighbouring country enterprises have had international trade offices in Singapore. In recent years, Singapore has also taken on functions of R&D and design centres. In particular, Malaysian firms have set up centres of excellence there. For instance, Bogasari International (Indonesia, food processing) chose Singapore, in part due to the country’s favourable R&D incentive schemes for foreign investors (UNCTAD, 2005: 182). Malaysia’s labour force is on par with Singapore, so it has functioned as a training centre for ASEAN as training costs are cheaper there than in Singapore. 10.4.2

What industries are taking intra-ASEAN FDI?

What industries take in intra-ASEAN FDI? Table 10.4 compares approvalbased FDI inflows to ASEAN in the manufacturing industry from extraand intra-ASEAN by industrial sector during the period between 1999 and 2003, provided by the ASEAN Secretary. Unfortunately, the figures for Singapore are excluded due to a lack of available data. ASEAN’s intra-regional FDI has been characterized by ‘efficiencyseeking’ FDI, or more precisely, ‘supplier following assembler’ FDI in intermediate goods. A typical example is communications equipment. The communications equipment industry dominated the largest share of

Do Local Suppliers Benefit from Economic Integration?

273

Table 10.4 FDI inflow from extra- and intra-ASEAN in manufacturing industry (approved base) by industrial sector (1999–2003) Host country

Vslue of Projects (US$ millions)

% of Share by Industry

Sectors/ISIC Code

Extra-

Intra-ASEAN

Extra-

Intra-ASEAN

15 Food & Beverages 16 Tobacco Products 17 Textiles 18 Wearing Apparel 19 Tanning of Leather 20 Wood & Wood Products 21 Paper & Paper Products 22 Publishing, Printing 23 Refined Petroleum 24 Chemicals Products 25 Rubber & Plastics Products 26 Other Non-Metallic Mineral 27 Basic Metals 28 Fabricated Metal 29 Machinery & Equipment 30 Office Machinery 31 Electrical Machinery 32 Communication Equipment 33 Precision Instruments 34 Motor Vehicles 35 Other Transport Equipment 36 Furniture 37 Recycling Others

2,347 75 1,890 1,418 850 1,450 1,961 146 12,942 4,958 4,998 2,237 2,579 1,477 1,979 330 1,486 14,474 534 1,831 1,542 704 45 436

1,084 16 173 129 36 174 589 19 168 432 437 106 172 290 333 25 94 1,404 99 164 66 67 7 7

3.7 0.1 3.0 2.3 1.4 2.3 3.1 0.2 20.6 7.9 8.0 3.6 4.1 2.4 3.2 0.5 2.4 23.1 0.9 2.9 2.5 1.1 0.1 0.7

17.8 0.3 2.8 2.1 0.6 2.9 9.7 0.3 2.8 7.1 7.2 1.7 2.8 4.8 5.5 0.4 1.5 23.0 1.6 2.7 1.1 1.1 0.1 0.1

TOTAL

62,692

6,093

100.0

100.0

Note: Singapore is excluded. Source: Compiled by author from ASEAN Secretary FDI data, ‘Statistics of Foreign Direct Investment in ASEAN’, Seventh Edition, 2005.

intra-ASEAN FDI inflows, accounting for 23 per cent of total intra-ASEAN FDI inflows. Coincidentally, the share is the same as FDI inflow from the extra-ASEAN region. Many developed country TNCs in the communications equipment industry had moved production sites to Singapore first and then to Malaysia, Thailand, the Philippines and China, seeking cheap labour. Foreign assemblers moved small parts of manufacturing processes first, and then gradually moved other processes, finally moving the assembly process into Malaysia, Thailand and the Philippines. Foreign suppliers as well as local suppliers in the communications

274 The Current Status of Indigenous Enterprises

equipment industry have followed their customers to expand into those countries, in order to respond quickly to the needs of customers. Why did the communications equipment industry send the largest amount of FDI to the ASEAN countries? Communications equipment has a characteristic of low transport costs. Transport costs in parts and components related to communications equipment are only 2–3 per cent of purchased goods by air,6 which is lower than any other industry, since the cubic measurement and weight are very small. On the other hand, the sales prices of communication devices fall about 1.5 per cent every month. Considering this monthly fall in sales prices, transport costs amounting to 2–3 per cent of purchased part costs are not high. Interestingly, a low transport cost industry, such as the communications equipment industry, produces not only a large international trade but also a large amount of FDI. This evidence may suggest that FDI substitutes partly for trade. ASEAN firms in machinery and equipment and fabricated metal, which are backward linkage industries for communications equipment, are aggressively advancing beyond borders within the region, seeking cheap labour and markets with suppliers following assemblers. The food and beverage industry is a mixture of ‘efficiency-seeking’ and ‘market-seeking’. Palm oil and sugar are typical ‘efficiency-seeking’ products, but the beverage industry is ‘market-seeking’. The food and beverage industry of intra-ASEAN FDI reached US$1,084 million and accounted for 18 per cent of total intra-ASEAN FDI. Intra-ASEAN FDI in the food and beverage industry is concentrated in Indonesia at 56 per cent, Malaysia 29 per cent, Vietnam 15 per cent, and Thailand 5 per cent (see Table 10.5). The motor vehicle industry is one of the typical ‘market-seeking’ FDI. In particular, Malaysian automobile industries have spread abroad due to the ‘home market effect’. Proton, a Malaysian automobile assembler, expanded to Indonesia. Malaysian automobile parts supplier Ingress Autoventures has established a joint-venture company with Japanbased Katayama Kogyo Co. Ltd. in Thailand. Beginning with two facilities in Rayong in 2003, a third facility in Ayutthaya began operations in 2005. In order to expand and diversify the group’s operations, PT Ingress Malindo (IMV), a Malaysian automobile supplier, was formed in Indonesia in August 2003. The plant is located on the Jababeka Industrial Estate, a location prominent in this industry in the Cikarang Selatan area. Malaysian automobile supplier Auto Parts Holdings Sdn Bhd, a wholly-owned subsidiary of APM Automotive, signed a joint-venture agreement with P.T. Mekar Armada Jaya to manufacture and distribute

Table 10.5 Share of intra-ASEAN FDI flows in manufacturing industry (approved base) by industrial sector (1999–2003) Sectors/ISIC Code

Host Country Brunei Darussala

Cambodia

Indonesia

Laos PDR

Malayasia

Myanmar

Philippines

Thailand

Vietnam

ASEAN

15 Food & Beverages 16 Tobacco Products 17 Textiles 18 Wearing Apparel 19 Tanning of Leather 20 Wood & Wood Products 21 Paper & Paper Products 22 Publishing, Printing 23 Refined Petroleum 24 Chemicals Products 25 Rubber & Plastics Products 26 Other Non-Metallic Mineral 27 Basic Metals 28 Fabricated Metal 29 Machinery & Equipment 30 Office Machinery 31 Electrical Machinery 32 Communication Equipment 33 Precision Instruments 34 Motor Vehicles 35 Other Transport Equipment 36 Furniture 37 Recycling Others

0.2 0.0 0.0 6.4 0.0 0.0 0.0 0.9 0.0 0.0 0.1 2.0 0.4 0.9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.5 0.0 0.0

0.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

56.3 11.7 68.8 25.8 29.8 37.1 83.7 10.0 20.5 12.0 20.9 1.7 27.6 10.2 9.0 27.9 6.8 4.7 28.3 37.7 59.5 16.8 0.0 0.0

0.6 0.0 0.0 0.2 2.5 2.2 0.1 2.1 1.5 0.1 0.3 0.6 0.0 0.1 0.0 3.0 0.4 0.0 2.2 0.2 0.8 2.5 3.7 0.0

22.9 82.2 22.8 15.9 6.8 32.2 3.0 47.7 78.0 31.9 22.2 50.6 30.6 44.3 30.7 0.0 55.7 55.5 60.4 0.1 28.8 33.5 0.0 46.6

0.0 0.0 0.3 5.4 0.0 2.0 1.4 0.0 0.0 0.0 0.0. 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5.2 0.0 0.0

0.0 0.0 0.6 2.1 2.4 9.3 3.0 0.1 0.0 1.7 1.0 16.7 2.1 2.7 0.2 45.2 4.5 10.8 6.6 2.4 0.0 2.1 30.7 1.5

4.9 0.0 4.9 5.4 19.6 2.4 2.7 19.7 0.0 48.4 46.6 5.5 25.4 26.1 54.3 0.0 13.3 28.6 2.6 27.7 8.4 5.1 65.5 0.0

14.6 0.0 2.6 20.3 38.9 14.8 5.5 15.2 0.0 5.8 8.5 22.9 13.9 15.6 5.8 23.9 19.4 0.4 0.0 31.9 2.4 34.3 0.0 0.0

100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

TOTAL

0.3

0.0

30.2

0.4

33.6

0.4

4.3

21.2

9.0

100.0

275

Source: Compiled by author from ASEAN Secretary FDI Data, ‘Statistics of Foreign Direct Investment in ASEAN’, Seventh Edition, 2005.

276 The Current Status of Indigenous Enterprises

automotive bus and train seating products in Indonesia in 2001. It also entered into a joint venture with Hefei Winking Asset Co. Ltd. in 2002 to produce and distribute automotive seats, interior parts and metal components in China. Thai automobile supplier Thai Summit Group expanded into Malaysia. Thai motor cycle supplier New Chip Xeng moved into Laos in 1991. 10.4.3

Two-way intra-industry FDI in ASEAN

Looking at approval base FDI inflow data cross tabulation by industry and by country, it can be observed that two-way intra-industry FDI has emerged in ASEAN. ASEAN TNCs in the communications equipment industry have mutually invested. Malaysia advanced to Thailand, Indonesia and the Philippines; meanwhile Thailand expanded to Malaysia and the Philippines. Thus, a two-way intra-industry FDI between Malaysia and Thailand has arisen in the same way as between the EU countries. Two-way intra-industry FDI has developed well in the food industry where local suppliers have developed: between Malaysia and Thailand, between Indonesia and Malaysia, between Indonesia and Thailand, and between Thailand and the Philippines. Two-way intra-industry FDI has been observed in the automobile industry between Thailand and Malaysia. This intra-ASEAN FDI suggests that ASEAN country enterprises explore the markets of other countries in ASEAN and make use of the relative advantages of each location, such as labour force, resources, infrastructure, parts supplier or agglomeration of industries and investor friendly measures such as corporate tax exemptions. It can be said that there is a high possibility of further development of intra-industry FDI in industries in which indigenous firms are developing, if transport costs (broadly defined, including government procedures) continue to decline and capacity building for manufacturing is further enhanced. 10.4.4

FDI to ASEAN LDCs

ASEAN enterprises have expanded into the ASEAN LDC countries. For instance, Thailand invested US$151 million in Laos, which amounted to 10.9 per cent of intra-ASEAN outward FDI from Thailand, and 58.2 per cent of intra-ASEAN FDI into Laos between 1999 and 2003 (Tables 10.2 and 10.3). Thailand invested US$191 million during the same years. The ASEAN apparel industry went to Cambodia, Laos and Vietnam for the purpose of exporting to third countries. The food industry went to Cambodia, Laos, Vietnam and Malaysia for export to third countries and home markets. Similar language and culture, and geographical proximities are very important (see Figure 10.5).

Do Local Suppliers Benefit from Economic Integration?

277

Vietnam is increasing FDI since the Laos–Viet Nam Joint Venture was established in 2000 to extend loans to invest by Vietnam’s firms in Laos. Vietnam is increasing FDI in Laos and nine projects have been carried out in Laos so far.

10.5

Conclusion and policy implications

Outward FDI from developing countries is a proxy indicator to measure the important role developing country enterprises have played in the world market, and how they benefit from globalization where border barriers are reduced. This study finds that ASEAN enterprises have extended their business activities to East Asia, being regional and global players. First, Singapore has taken a lead, followed by Malaysia and Indonesia, although most of their FDI is directed at Singapore. Second, ASEAN country TNCs went to neighbouring countries at first, resulting in a high share of outward FDI within ASEAN and East Asia, and then to large markets such as the United States as they grew. They seem to take a step up from being local players, then ASEAN regional players, and finally global players. Through overseas production, even small players have grown to be global players. Third, there is a variety in outward FDI from ASEAN: ‘market-seeking’, ‘efficiency-seeking’ and ‘resource-seeking’ FDI. The most developed type is ‘efficiency-seeking’ FDI where ASEAN TNCs sought countries with cheap labour and land. They followed their customers when the customers expanded their facilities to lower wage countries, a phenomenon which can be called ‘supplier following assembler’ FDI. The communications equipment industry is a typical industry in which ASEAN suppliers followed multinational corporations. Fourth, thanks to the home market effect, some ASEAN enterprises become global players. Food and agrobased industries are typical examples of this. ASEAN enterprises have been very dynamic in seeking cheap labour and markets overseas. ASEAN local TNCs can develop together with foreign multinational firms. It may depend on the following two things: first, whether ASEAN can really establish a seamless production area or not. In this regard, tariff barriers are designed to be eliminated for substantially all goods, and border barriers will be reduced if e-custom clearance and single window inspection are completely implemented in all member countries by 2012 as scheduled, which will make it possible to apply and get approval for export and import procedures from customs offices and other related agencies. The elimination of trade barriers, particularly in the lower income countries of Cambodia,

278 The Current Status of Indigenous Enterprises

Laos, Myanmar and Vietnam, would provide business opportunities for not only multinational firms but also local ASEAN ones. Second is whether the low wage countries of Cambodia, Laos, Myanmar and Vietnam can implement investment friendly measures with a view to attractiveness for FDI.7 Production costs and transport costs can be reduced by enhancing capacity building in infrastructure (road, sea and air ports, communication, and so on), institutions (streamlining procedures for trade, investment and registration, deregulation of capital participation, tax privileges and other investment facilitation measures), human resources (skill labour training centres, laboratories and universities) and so on. The most important thing is the service spirit of government officers who approve certificates. These favourable measures for FDI will increase the number of ASEAN TNCs expanding into the LDC countries and enhance regionalization in investment as well as trade. ASEAN is not an exception to this phenomenon. In particular, Singapore-based enterprises are expanding their business opportunities to the global market. Malaysia follows this trend, increasing its presence remarkably not only in ASEAN but also in the rest of the world. ASEAN country TNCs are expanding their wings beyond their borders; first to neighbouring countries of ASEAN and then to the rest of the world. Advancing abroad, just like the Japanese and South Korea enterprises did in the past, many of these local enterprises are becoming global players.

Notes 1. A transnational corporation is a company that has an affiliate overseas, whereas a multinational enterprise is a big company that has several production and distribution overseas networks. 2. Broadly defined transport costs include freight costs as well as border-related costs such as tariffs, and other non-tariff barriers officially or unofficially. 3. Anderson and Wincoop (2004) use the term ‘trade cost’ which is favoured by international trade economists instead of ‘transport cost’ which a geographical economist would generally use. Both concepts are the same. 4. They estimated 13 economies of Newly Industrial Asia (Hong Kong, Korea, Singapore and Taiwan), China, ASEAN4 (Indonesia, Malaysia, the Philippines, Thailand), and South Asia (Bangladesh, India, Pakistan and Sri Lanka). 5. See the CP Group’s website (http://www.cpthailand.com/webguest/faq.aspx). 6. Filed study result by author. Normally, transport costs are covered by the buyer. 7. The success of some Asian economies is no coincidence. Active and coherent policies with a long-term vision are necessary (UNCTAD, 2005: 255).

Do Local Suppliers Benefit from Economic Integration?

279

References Anderson, James E. and Eric van Wincoop (2004). ‘Trade Costs’, Journal of Economic Literature, 42: 691–751. A.T. Kearney (2004). A.T. Kearney’s 2004 Offshore Location Attractiveness Index: Making Offshore Decisions, Chicago: A.T. Kearney. ENG Teknologi Holding (2000). SMI Globalization: 25 Years of Engtek Growth, Malaysia. Fujita, Masahisa (2007). ‘Formation and Growth of Economic Agglomerations and Industrial Clusters: a Theoretical Framework From the Viewpoint of Spatial Economics’, in A. Kuchiki and M. Tsuji (eds), The Flowchart Approach to Industrial Cluster Policy, Basingstoke: Palgrave Macmillan. Hattari, Rabin and Ramkishen Rajan (2007). ‘Intra-Asian FDI Flows: Trends, Patterns and Determinants’, paper presented at the international workshop, ‘Intra Asian FDI Flows: Magnitude, Trends, Prospects and Policy Implications’, organized by the Indian Council for Research on International Economic Relations, held in Delhi on 25–26 April. Hiratsuka, Daisuke (2003). ‘Competitive Strategy in the Six ASEAN Countries: Summary Findings from Individual Papers’, in I. Yamazawa and D. Hiratsuka (eds), ASEAN-Japan Competitive Strategy, Chiba-Makuhari, IDE-JETRO. Krugman, Paul (1980). ‘Scale Economies, Production Differentiation and the Pattern of Trade’, American Economic Review, 70: 950–9. Krugman, Paul (1991). Geography and Trade, Cambridge, MA: MIT Press. Krugman, Paul and Anthony J. Venables (1995). ‘Globalization and the Inequity of Nations’, Quarterly Journal of Economics, 110, 4 (November): 857–80. UNCTAD (2004). World Investment Report 2005: the Shift Towards Services, New York and Geneva: United Nations. UNCTAD (2005). World Investment Report 2005: Transnational Corporations and the Internationalization of R&D, New York and Geneva: United Nations. UNCTAD(2006). World Investment Report: FDI from Developing and Transition Economies: Implications for Development, New York and Geneva: United Nations.

This page intentionally left blank

Part IV Quantitative Implications of Economic Integration

This page intentionally left blank

11 Comparing Bilateral and Multilateral ASEAN10þ4 Free Trade Agreements: Possible Impacts on Member and Non-member Countries Michel Fouquin

11.1

Introduction

Since the 1990s, preferential trade agreements of all kinds (from free trade to customs unions, and regional community, from bilateral to regional) have spread all over the world. Almost every country is member of at least one agreement. Many agreements – at least the most important ones – are organized on a continental basis such as in Europe, America, Africa or, more recently, Asia. These regionalization trends have revived research on their economic rationale and on their compatibility with multilateral agreements. Asian movement towards regionalization – known as ASEAN10þ3, þ4 or þ61– is relatively new2 and may gain major importance as it involves Japan – one of the most advanced economies in the world – and the two biggest emerging countries in the world, China and India. These countries have contributed (Japan) or will contribute (China and India) to the shaping of a completely new world. Their economic integration with ASEAN and their possible impact on the world economy using a computational general equilibrium (CGE) model are the subjects of this chapter.3 In section 11.2 we will give a rapid overview of the literature on regional agreements and analyse the making of ASEAN; in section 11.3 we will define our baseline and scenarios; section 11.4 will be dedicated to the analysis of the main results for scenario 1; section 11.5 will present the results of scenario 2 and conclusions on scenario 3 as well as some further comments. 283

284 Quantitative Implications of Economic Integration

11.2 11.2.1

The open regionalism concept and AFTA The open regionalism concept

The traditional distinction introduced by Jacob Viner (1950) about the trade creation/trade diversion effect of preferential agreements is still valid. By definition, preferential trade agreements (PTA) are discriminatory and therefore they may tend to reduce welfare. If a firm located within a PTA zone produces and exports a liberalized item then it will benefit from the elimination (or reduction) of tariffs within the PTA, and the local consumer will benefit from reduced prices. This will improve the producer’s as well as the consumer’s welfare. On the other hand, if that producer is less efficient than outside producers and survives because of initial high tariffs, it may nevertheless benefit from tariff elimination by the other PTA members. This would occur if the difference in efficiency with external producers is lower than the remaining external tariff and/or if the resulting external tariff is higher than some of the previous ones. This will cause trade diversion and welfare losses for the consumer and global losses for producers. With the new international trade theory several developments have pursued the issue further. Krugman (1991) showed that, using a model of imperfect competition, in a world with high trade costs, continental bloc formation can be welfare improving. Frankel et al. (1993) estimated that 18 per cent was the threshold value for transportation costs to allow for positive welfare gains in continental bloc formation. They found that the observed value for intercontinental transportation was around 15 per cent, and they therefore concluded that regional discrimination was probably welfare reducing. However, research carried out on specialization shows that in the case of complementarity between the continental bloc members this should be welfare improving. In order to reduce the discriminatory nature of preferential agreements, and also probably to differentiate it clearly from the European experience, the concept of ‘open regionalism’ was developed and supported by US and Australian (and the Eminent Persons Group Report to the 5th APEC Ministerial Meeting in 1993) economists in the APEC project to create an Asia Pacific Free Trade Area by 2020. But no clear definition of open regionalism was available at that time. After the financial crisis of 1997–8 there was renewed interest in regional build-up. Wei and Frankel (1998) proposed a clear definition of open regionalism as a regional scheme which did not reduce global welfare. The argument is that regionalization should not be regarded as an isolated regional affair but as a global trend across the three continental regions.

Comparing Free Trade Agreements 285

Two proposals have been made to neutralize the negative impact of trade diversion. First is the Macmillan proposal which requires that there should be no volume changes in trade between the member and non-member countries after the formation of a bloc. This is achieved through a degree of external liberalization by the members vis-a`-vis the non-members. Given average parameters it assumes a reduction by 40 per cent of former external tariffs which might be politically unacceptable. Wei and Frankel (1998) argue that this line of reasoning does not take into account the fact that the non-member countries are also engaged in other continental trade agreements. Therefore, they propose to take into account the relative element of liberalization rather than the absolute value and they estimate that under a realistic hypothesis only 4 per cent further liberalization between members and non-members might be needed to produce Pareto improvement. According to its promoters, AFTA remains inspired by the idea of open regionalism. Among the issues addressed in this chapter is the extent to which Asian regional projects really fit into that. 11.2.2

The ASEAN free trade area

Institutional integration in East Asia originated with the creation of the Association of South East Asian Nations4 (ASEAN) in 1967. In 1992, ASEAN agreed to establish the ASEAN Free Trade Area (AFTA) through a Common Effective Preferential Tariff5 (CEPT) scheme, in order to attract foreign direct investment (FDI) from abroad. AFTA required that tariffs levied on a limited range of products (it excluded a wide range of sensitive products) traded within the region be reduced to no more than 5 per cent. Quantitative restrictions and other non-tariff barriers were also to be eliminated. There was also a cumulative rule of origin (RoO). Although originally scheduled to be achieved by 2008, the targets of AFTA were continually brought forward. In fact ASEAN suffers from two major problems: a lack of political will comparable to the Franco-German engagement in Europe and the fact that trade and investment within ASEAN10 are second to their relations with the US, Japan or even the EU25. In 1998, after the financial crisis, ASEAN leaders decided to accelerate the process of all tariffs on products in the inclusion list being reduced to a 0–5 per cent range, and further to zero. Moreover, they decided to enlarge the inclusion list. New members agreed to the same objective after a delay. By the beginning of 2002, the original ASEAN6 had met AFTA’s targets, and only 3.8 per cent of products in the CEPT inclusion list (see below), or 1,683 items out of 44,060, had tariff rates above

286 Quantitative Implications of Economic Integration

5 per cent. The simple average tariff on goods traded under the AFTA scheme was about 2.7 per cent in 2003, and 1.7 per cent in 2004. Vietnam is expected to achieve AFTA compliance in 2006, Laos and Myanmar in 2008, and Cambodia in 2010. ASEAN will eliminate all import duties by 2010 for the six original members and by 2015 for the new members. But there are still major limitations to an effective FTA. Each member country has the option of excluding a list of ‘sensitive products’ – mostly agricultural products and primarily rice products – from the agreement. Furthermore, each country has its own list of products, which means that this regional agreement is not actually a true single agreement but an agglomeration of bilateral agreements. 11.2.3 Long-term perspectives: from an ASEAN FTA to an ASEAN community Besides the liberalization of trade in goods, ASEAN has endeavoured to take the next steps to create a community. In October 2003, the ASEAN leaders agreed to achieve a dynamic, cohesive, resilient and integrated ASEAN community by 2020, by creating the ASEAN Security Community, the ASEAN Economic Community, and the ASEAN Sociocultural Community. The ASEAN Economic Community is to be a single market and production base. It aims to create a stable, prosperous and highly competitive ASEAN economic region by the year 2020, in which there is a free movement of goods and services, freer movement of capital, equitable economic development, and reduced poverty and socio-economic disparities. This ASEAN initiative was followed by projects involving ASEAN10þ3, then þ4 and lately þ6. The additional six countries were Japan, Korea, China, and India, Australia and New Zealand. However, the Kuala Lumpur Summit in December 2005, due to strong political friction among its participants, was unable to establish concrete measures towards the creation of an Asian community. Nevertheless the FTA fever continues to advance. 11.2.4

The ‘enlargement’ of the ASEAN FTA

ASEAN has extended its institutional integration instruments to other countries (see Figure 11.1 for a summary of relationships).6 First, in November 2002, ASEAN and China signed the framework agreement on Comprehensive Economic Cooperation that will establish the ASEAN-China Free Trade Area (ACFTA) within ten years. In November 2004, ASEAN and China agreed to establish ACFTA by 2010 for the original six ASEAN member states and China, and by 2015 for the newer ASEAN member states.

Comparing Free Trade Agreements 287 APEC Russia NAFTA SAFTA Nepal Bhutan Maldives Pakistan

AFTA ASEAN_10 ASEAN_6

Canada Japan United States

Indonesia Brunei* Malaysia Philippines

Mexico

Singapore* South Korea

Thailand Sri Lanka

China

Taiwan Peru

Hong Kong Chile* India Bangladesh

Vietnam Papua– New Guinea Cambodia Laos Myanmar

Australia

New Zealand*

CER

Implementation started

Negotiations started (without a framework agreement) Negotiations of a framework agreement * Members of the TPSEPA (Trans-Pacific Strategic Economic Partnership Agreement) Agreement signed

Figure 11.1

Asia-Pacific regional and bilateral trade agreements (June 2005)

Source: Fouquin (2007).

Second, in October 2003, ASEAN and Japan signed the framework for a Comprehensive Economic Partnership (CEP), and in November 2004, they agreed to start negotiations for an ASEAN-Japan CEP agreement by April 2005. Third, in November 2004, ASEAN and the Republic of Korea, as well as ASEAN and CER (the Closer Economic Relations agreement between Australia and New Zealand), agreed the overall framework to establish free trade areas. ASEAN and Korea agreed the ASEAN-Korea FTA in December 2005. Korea will liberalize Normal Track7 by 2010, while ASEAN6 will eliminate its tariffs for all tariff lines placed in the Normal Track no later than 2010, with flexibility to have tariff lines, not exceeding 5 per cent of all the tariff lines or as listed in an agreed schedule, eliminated no later than 1 January 2012. ASEAN and CER started negotiations in early 2005 and these could be completed by August 2008. With India, the framework agreement was signed in October 2003, and negotiations started in January 2004.

288 Quantitative Implications of Economic Integration

These different projects vary greatly depending on the partners’ development level. A South–South agreement, such as ASEAN-China, can be negotiated within the 1979 GATT ‘Enabling Clause’, which provides greater latitude to exclude products from the liberalization scheme between developing countries forming FTAs. For example, ACFTA allows exemption of tariff reductions for a sensitive list of products, up to a ceiling of 400 tariff lines at the HS 6-digit level and 10 per cent of the total import value, based on 2001 trade statistics for the ASEAN6 and China; and 500 tariff lines at the HS 6-digit level for Cambodia, Laos, Myanmar and Vietnam. In contrast, when one of the partners is a developed country, as in the case of a Japan-ASEAN FTA, the agreement must be notified under the GATT/WTO Article 24 and must comply with the requirement of sectoral comprehensiveness (the agreement should cover substantially all bilateral trade). Therefore there are more obstacles to this second kind of arrangement and China may have advanced quicker than Japan in its negotiations with ASEAN. The FTAs centred on ASEAN have different modalities and time frames and will lead to free trade areas covering a large part of East Asia. (There are also arrangements between countries in Northeast Asia, such as between China and Japan, between China, Japan and the Republic of Korea, and so on.) These efforts by ASEAN members are expected at some point to establish a base for East Asian FTAs as well as economic and political communities, due to the political nature of these agreements. The leaders of ASEANþ3 are open to an eventual enlargement of the East Asia trade region in which India, Australia and New Zealand could participate. This is probably an example of the ‘domino effect’ (Baldwin, 1993): exporters to regional blocs are strong pro-membership forces leading to enlargement of blocs which in turn intensifies pro-membership lobbying. A good example of that tendency is given by Australia which used to be opposed to Asian regionalization (see Garnaut, 2005). However, there is also the possibility that some countries will be excluded or left behind such as Taiwan, or South Asian countries like Bangladesh or Pakistan. This would be due to strategic reasons or to their low levels of development and to the fact that India, which is at the centre, has remained a relatively closed economy compared, for example, to China. These countries might suffer from negative diversion effects. To conclude, it may be recalled that East Asia has consistently maintained higher barriers to trade within Asia than with other partners, as can be seen from the estimates made by the ITC (International Trade Center) for the year 2002 (Table 11.1). In other words, East Asia, contrary

Table 11.1

Applied tariffs in East Asia, the EU and NAFTA, by sector, 2002

Importer

Ad valorem tariff equivalents (%) East Asia

Exporter Agriculture Light industry Food and beverages Textiles and clothing Transportation machinery Pottery products Chemicals Basic metals Mining products General machinery Electrical machinery Others Wood and paper Precision apparatus All products

NAFTA

EU-25

East Asia

NAFTA

EU-25

East Asia

NAFTA

EU-25

East Asia

NAFTA

EU-25

41.0 26.8 21.8 7.3 4.6 2.9 2.4 1.8 1.7 1.5 1.4 1.4 1.4 1.2

29.7 8.3 26.4 7.6 2.8 3.6 3.0 2.6 2.6 1.9 1.5 1.7 1.3 1.3

30.9 12.8 25.8 7.8 8.6 4.4 2.7 2.3 1.7 2.5 2.2 2.6 1.5 2.0

25.2 4.9 10.1 6.2 3.4 1.4 0.8 1.5 0.4 0.3 1.0 0.6 0.4 0.3

6.8 0.0 5.3 0.0 0.0 0.1 0.0 0.5 0.0 0.0 0.0 0.1 0.0 0.0

21.4 2.2 18.1 4.9 6.8 2.9 4.9 4.0 0.3 1.3 1.3 0.8 1.0 1.1

20.2 8.7 16.0 10.9 3.3 5.7 4.1 3.0 1.4 1.1 2.5 3.2 1.0 1.3

15.6 9.6 15.7 9.7 2.9 5.6 4.2 2.7 1.8 1.7 3.2 1.3 1.0 2.1

3.9 0.1 9.4 0.1 0.0 0.5 0.2 0.3 0.1 0.0 0.1 0.0 0.0 0.1

7.4

5.5

7.2

7.6

1.9

7.7

5.7

5.3

0.7

Source: Market Access Map, calculations by ITC (2003).

289

290 Quantitative Implications of Economic Integration

to other regions like the EU or NAFTA, tends to discriminate against its members. This is well illustrated for the most sensitive sectors: agriculture, textile clothing, and food and beverage industries. In contrast, there is no discrimination within East Asia for industrial products and no preferential treatment. A second conclusion might be that APEC’s failure to bring significant results may be due firstly to its non-continental content and secondly because of very divergent views on the real content of what was an open regional scheme. The US wanted clear and substantial engagements (‘legal bindings’) by country members while ASEAN countries preferred to engage in a limited and informal agenda corresponding to what is called the Musyawarch practice or ‘ASEAN way’, based on consensusbuilding. It transpired that these two views remained incompatible and the APEC project did not take off. A third conclusion could be that East Asia needs strong partnership with the rest of the world, but there is also a need for an Asian identity. The genuine need for an increased cooperation within Asian countries has been shown by the financial crisis, various natural disasters, health issues such as SARS and even the rise of China.

11.3 11.3.1

Quantitative model and possible scenario The MIRAGE model

Before presenting our results we first give a very short overview of the way MIRAGE and most CGE models work. CGE models refer to the economic model based on the theory of general equilibrium which assumes that if there is no distortion, all demand finds its supply at an equilibrium price and this is true on all markets for all goods and services simultaneously. Every agent adapts its behaviour as a consequence. A CGE model shows how the adjustment works, taking into account all the interaction between the different markets. These models are now commonly used for simulation purposes on international trade issues. We use an international database which relies on every national market to international trade through every bilateral flow. CEPII has decided to build a multi-sector, multi-region CGE model, also known as MIRAGE (Modelling International Relationships in Applied General Equilibrium), devoted to trade policy analysis. MIRAGE describes imperfect competition in an oligopolistic framework a` la Cournot. It accounts for horizontal product differentiation linked to varieties, but also to geographical origin (nested Armington–Dixit–Stiglitz

Comparing Free Trade Agreements 291

utility function). A new calibration procedure allows the available information on these aspects to be used efficiently. The modelling is done in a sequential dynamic set-up, where the number of firms by sector adjusts progressively, and where installed capital is assumed to be immobile, even across sectors. Capital reallocation therefore only results from the combined effects of depreciation and investment. It makes it possible to describe the adjustment lags of capital stock, and the associated costs. Compared to previous applied CGE trade models, MIRAGE has in addition three main distinctive features, aimed at improving the description of trade policies’ main transmission channels. • FDIs are explicitly described, with a modelling both theoretically consistent (with agents’ behaviour, and with domestic investment setting), and consistent with the empirical results about FDIs’ determinants and their order of magnitude; • a notion of vertical product differentiation is introduced, by distinguishing two quality ranges, according to the country of origin of the product; and • trade barriers are estimated by the MacMaps described above. As a result, MIRAGE is based on a description of trade barriers that, besides its precision, preserves the bilateral dimension of the information, in contrast to what is commonly done in applied modelling. Except for data on trade barriers, the model uses the GTAP 6 (Global Trade Analysis Project) database (see Dimaranan et al., 2002). This allows a wide flexibility in choosing the sectoral and geographical aggregations of MIRAGE that may be changed for each application. 11.3.2

The baseline assumptions and results

The simulation exercise starts with the introduction in MIRAGE of hypotheses reproducing the expected demographic, technological and macroeconomic changes which make the baseline. The initial levels of the skilled and unskilled labour force in each region of the model are those of the GTAP 6 database. The structure of the labour force (ratio of skilled and unskilled), however, is assumed to be constant over time. The growth rates of the labour force, GDP and savings rate in each region of the study are taken from the World Bank projections. The initial (2001) savings rate is taken from the GTAP 6 database. We then consider three sectors: agriculture, industry and services. The computation of the change in productivity in the three sectors is based

292 Quantitative Implications of Economic Integration

on the following assumptions: in the agricultural sector it is assumed to be constant and equal to 2.5 per cent; in the industrial sector it is assumed to be 2 points higher than the change in productivity in the service sector; changes in productivity in the third sector are calibrated in order to replicate the World Bank GDP growth rates. 11.3.3

Four scenarios for East Asian integration

There are two main types of regionalization in East Asia: the fragmented one, based on bilateral agreements, and the unified one, based on a plurilateral agreement for East Asia. Both types of agreement may be more or less liberal as they may include all products or exclude a list of ‘sensitive’ products. Therefore we have four possible scenarios. In scenario 1, we envisage a ‘hub and spoke’ scheme: ASEAN10 removes its tariffs vis-a`-vis China, India, Japan and Korea with reciprocal movements from these countries. There are no sensitive products and hence no exclusion list. Tariffs against third countries remain unchanged. Scenario 3 differs from scenario 1 only by the exclusion of a list of sensitive products from the liberalization scheme (but the baseline remains the same for intra-ASEAN trade). In scenario 2, we envisage a full FTA in which China, India, Japan and Korea not only remove their tariffs on imports from ASEAN10 members, but also remove their bilateral tariffs (e.g. China–Japan). ASEAN10 is doing the same. Scenario 4 differs from scenario 2 only by the exclusion of the same list of sensitive products as in scenario 3 from the liberalization process. As regards the list of exclusions (the extra-ASEAN sensitive products), we lack information, with the exception of China which has already notified its list of exclusions vis-a`-vis the ASEAN10, as well as, for example, Malaysia which has notified its exclusion list vis-a`-vis China.8 Otherwise the general principle adopted is to identify 10 per cent of the tariff lines to be defined as sensitive products, working as follows. We first replicate the eventual list of sensitive products within ASEAN before 2010, assuming that for each member, sensitive products within ASEAN should be sensitive vis-a`-vis the four new partners.9 We add to the former list the tariff lines that will not be bound by 2010, on the basis of the information available in 2005. If necessary we add the highest bound tariff lines. It should be noted that this methodology does not lead to artificially created protection for a country such as Singapore. We do not define the exclusion list by merging the different individual countries’ lists, but instead keep those lists at the country level. The averaging of the

Comparing Free Trade Agreements 293

protection opposed by ASEAN10 to a given partner (e.g. China) among the ten members is only a second step in our calculation. Reciprocally, India, Japan and Korea declare the same percentage of sensitive tariff lines; for these countries exclusions are defined by considering unbounded lines and the highest bound tariffs. As regards China, we extend the notified list to 10 per cent of its tariff lines, according to the previous principles, to adopt a unified treatment of tariff structures.

11.4 ASEAN10 bilateral FTAs with Japan, China, Korea and India 11.4.1

Main results

The macroeconomic results differ depending on whether we consider the volume impact of trade liberalization or the value impact. These gains are large compared to many results of similar studies. This is partly because we use detailed bilateral tariffs and estimates of non-tariff barriers, both of which are high within Asia (see Table 11.1). Table 11.2 shows the simulated results of scenario 1 where ASEAN10 removes its tariff vis-a`-vis China, Japan, Korea and India. ASEAN10, the biggest winner on all accounts, cumulates gains in volume (þ4.43 per cent of GDP) with gains in the real exchange rate (RER) (þ2.18 per cent), the total gain being 6.61 per cent. The gains in ASEAN exports are higher than those in imports and therefore are compensated by a real exchange rate appreciation. Japanese gains come next with þ0.75 per cent for GDP and 0.27 per cent RER appreciation. Korea shows a similar picture. For China there are only gains in volume (þ 0.84 per cent) while gains in value are nil. The Indian case offers the most contrasted results: gains in volume (þ2.19 per cent) are overshadowed by losses in value (2.37 per cent), making India the sole country to lose among the partners of ASEAN10. How can we explain that the most protected economy loses from the liberalization process? India’s real GDP increase is the second largest after ASEAN10. The first impact of liberalization is to stimulate a large rise in imports, while there are only small gains in exports. The small export gains are due to the relatively low initial tariffs of India’s partners. Therefore the balance of Indian trade tends to deteriorate and as a consequence its currency depreciates in order to rebalance its external trade. The depreciation is larger than the gains in volume, and so are the losses on the terms of trade. Another negative impact accrues from the depreciation of assets. In the end the welfare change is negative (see Table 11.3).

Macroeconomic impacts (in %, SC1) GDP (volume)

ASEAN-10 Japan Korea China India Hong Kong, Taiwan, Rest of East Asia South Asia EU-25 EFTA Russian Fed North Africa Rest of Europe USA Canada Mexico and Central America South America Australia and New Zealand Rest of the world Source: MIRAGE.

Real effective exchange

294

Table 11.2

Welfare

2010

2015

2020

2025

2010

2015

2020

2025

2010

2015

2020

2025

3.27 0.52 0.47 0.70 2.25 0.26

3.80 0.65 0.57 0.84 2.37 0.38

4.20 0.72 0.52 0.86 2.25 0.42

4.43 0.75 0.48 0.84 2.19 0.41

2.39 0.12 0.04 0.05 2.33 0.13

2.04 0.19 0.19 0.01 2.27 0.12

2.06 0.24 0.17 0.01 2.34 0.12

2.18 0.27 0.13 0.01 2.37 0.13

1.52 0.15 0.27 0.13 0.24 0.14

2.18 0.18 0.40 0.12 0.32 0.19

2.61 0.19 0.48 0.12 0.36 0.22

2.88 0.19 0.53 0.11 0.35 0.23

0.25 0.08 0.03 0.12 0.12 0.04 0.10 0.07 0.08

0.20 0.08 0.06 0.15 0.16 0.04 0.08 0.09 0.09

0.23 0.08 0.08 0.17 0.20 0.05 0.07 0.11 0.11

0.27 0.08 0.10 0.19 0.24 0.05 0.07 0.13 0.14

0.11 0.00 0.01 0.09 0.10 0.01 0.04 0.03 0.05

0.12 0.00 0.00 0.12 0.15 0.01 0.02 0.03 0.05

0.16 0.00 0.00 0.14 0.19 0.00 0.02 0.05 0.05

0.20 0.00 0.01 0.17 0.22 0.00 0.03 0.07 0.06

0.08 0.01 0.01 0.05 0.07 0.01 0.00 0.01 0.02

0.05 0.01 0.01 0.08 0.08 0.01 0.00 0.01 0.02

0.06 0.01 0.02 0.10 0.09 0.01 0.00 0.01 0.03

0.07 0.01 0.02 0.11 0.10 0.02 0.01 0.02 0.05

0.19 0.10

0.23 0.17

0.27 0.21

0.31 0.23

0.19 0.00

0.23 0.08

0.27 0.10

0.31 0.13

0.04 0.04

0.05 0.07

0.06 0.08

0.07 0.09

0.08

0.11

0.13

0.14

0.01

0.05

0.07

0.09

0.02

0.03

0.03

0.04

Comparing Free Trade Agreements 295 Table 11.3

Impact on welfare (US$ million, SC1)

Region ASEAN-10 Japan Korea China India Hong Kong, Taiwan, Rest of Asia South Asia EU-25 EFTA Russian Federation North Africa Rest of Europe USA Canada Mexico and Central America South America Australia, New Zealand Rest of the world

2010

2015

2020

2025

9,473.61 5,450.52 1,166.28 1,711.58 1,468.28 702.30

17,169.43 6,877.31 2,222.42 2,216.69 2,539.19 1,273.00

25,839.45 7,785.35 3,385.65 2,946.97 3,668.09 1,780.13

36,001.17 8,574.74 4,695.44 3,930.91 4,582.17 2,291.52

149.77 829.32 22.09 179.90 184.15 31.85 263.78 64.24 126.44

134.19 779.35 42.58 303.36 252.15 47.96 107.87 78.84 195.44

181.08 878.91 69.50 418.96 350.03 69.39 663.67 126.17 353.73

261.54 948.14 110.92 549.49 476.53 90.94 1,287.12 192.69 582.39

494.51 187.64 214.40

739.20 329.96 490.83

1,005.68 468.93 678.18

1,325.59 634.73 853.33

The same line of reasoning explains why welfare and GDP changes might have a different sign (India and Korea for example). Liberalization of trade in agricultural goods has a major impact on wages. Indian and Korean agriculture are highly protected so liberalization brings large decreases in price, in employment and finally in wages for a large part of the population which still works in that sector (at least in India). Are the value losses real? We can interpret these results as a consequence of being formerly closed: closeness encourages over-valuation of the currency. To some extent that over-valuation does not reflect the real competitiveness of India (which was clearly the cause of the financial crisis of 1991 and the reason for supporting economic reforms). So the loss on values should not be considered too negative for India. Maximizing volume growth is probably the first priority for a developing country rather than value maximization. The rest of the world loses in both volume and value. As access to Asian countries is reduced, these countries register a decline in their exports; due to the diversion effect, that decline induces a decline in output (between 0.41 per cent and 0.07 per cent) as well as a currency depreciation (between 0.31 per cent and 0 per cent).

296 Quantitative Implications of Economic Integration

Among the losers we find South America, South Asia, North Africa, the Russian Federation, Australia and New Zealand. They all lose out to the ASEAN hub. These zones are specialized in primary and agro-food products and suffer from the diversion effect more than industrialized countries. Other zones like Hong Kong and Taiwan are among the main losers due to their strong links to ASEANþ4 countries. Being excluded from the Asian hub represents a high price. One may dispute that result and conclude that these countries should be included in the hub. It is clearly the case with Hong Kong while for Taiwan there are political issues that may delay its regional integration. The impact on the EU25 and for the rest of Europe is the lowest; this is in part the result of a weak link between East Asia and Europe, this link being particularly weak for agro-food products in Asia. Similarly the US GDP is marginally affected (0.07 per cent), but contrary to the EU it also suffers currency depreciation (0.03 per cent) due to the impact on US agro-food exports. In general it seems that developed countries are less sensitive to Asian regionalization than developing countries. Our analysis up to now has been focused on the final year results (2025); we now look at the dynamic adjustment between 2010 and 2025. For most countries the impact on volume grows overtime. For ASEAN10 GDP it goes from 3.27 per cent the first year to 4.43 per cent in 2025, an increase of one-third. For the RERs, adjustment to the shocks appears to be immediate and often stays stable over time. These diverging evolutions show that real adjustment takes time while price changes are almost instantaneous; this seems to be a reasonable approximation of reality. This indicator of welfare summarizes the different impacts of the shocks on the supply side as well as on the demand side. ASEAN and Japan are the only two Asian zones to benefit in terms of welfare. ASEAN gains are valued at US$36 billion (Table 11.3) while those of Japan are valued at US$8.6 billion and those of the US at US$1.3 billion. The rest of the world loses – Korea around US$4.7 billion and India US$4.6 billion. 11.4.2

Impacts on trade in agricultural goods

As expected, ASEAN10 is among the biggest winners on the export side (see Table 11.4). Their agricultural export volumes increase by 35 per cent in 2010 and 30 per cent by 2015 compared to the baseline projection. India comes first in terms of percentage but second in absolute value. As India was one of the most protected economies (50 per cent tariffs), it also faces high tariffs for its exports (24 per cent on average). Its world

Table 11.4 Impact on agricultural trade (in %, SC1) Agricultural exports (volume)

ASEAN-10 Japan Korea China India Hong Kong, Taiwan, Rest of East Asia South Asia EU-25 EFTA Russian Fed North Africa Rest of Europe USA Canada Mexico and Central America South America Australia and New Zealand Rest of the world

Agricultural imports (volume)

2010

2015

2020

2025

2010

2015

2020

2025

34.73 8.93 5.56 9.54 43.25 0.32 5.50 0.01 0.05 0.28 0.25 0.27 0.59 0.07 0.19 0.27 0.39 0.99

35.12 7.04 4.57 11.06 43.66 0.76 6.96 0.10 0.11 0.16 0.24 0.40 0.63 0.01 0.19 0.27 0.40 0.99

33.12 5.35 3.75 13.00 46.45 0.49 8.48 0.16 0.15 0.36 0.20 0.50 0.64 0.00 0.16 0.24 0.51 1.03

30.34 4.07 2.97 15.31 49.11 0.05 10.25 0.24 0.19 0.68 0.16 0.62 0.67 0.03 0.11 0.17 0.54 1.11

23.13 17.56 24.51 3.31 47.84 0.80 0.99 0.10 0.07 0.31 0.43 0.06 0.47 0.17 0.12 0.61 0.85 0.54

23.61 19.64 24.85 3.43 50.43 1.25 0.85 0.18 0.15 0.42 0.62 0.08 0.48 0.26 0.17 0.83 1.37 0.65

23.98 20.45 24.35 3.44 50.45 1.37 0.91 0.20 0.18 0.48 0.72 0.08 0.46 0.35 0.21 0.94 1.51 0.68

24.16 20.66 23.65 3.40 50.06 1.40 1.03 0.21 0.21 0.52 0.83 0.08 0.47 0.43 0.23 1.04 1.63 0.70

Source: MIRAGE.

297

298 Quantitative Implications of Economic Integration

exports increase from 43 per cent in 2010 to 50 per cent in 2025. Most of that increase is made on ASEAN10 markets. The bilateral flow more than doubles. China comes in third place with a 15 per cent increase. Japan and Korea’s low levels of exports increase only marginally. Among other zones with very small gains (less than 1 per cent) are Australia, New Zealand, Hong Kong, Taiwan and South America. They benefit from the real exchange rate appreciation of ASEAN (see below). Other zones lose market share because of a diversion effect. First, South Asian countries lose 10 per cent in favour of ASEAN10 and India10 but not to China with which it has very limited trade relations. This is probably the most critical issue since South Asia covers one of the poorest regions of the world, highly dependent on agriculture. The EU25 loses 0.24 per cent and the US 0.67 per cent. These losses are not big, but we should realize that for the EU it comes on top of an already relatively low share of the fast growing ASEAN markets. Therefore it is the lost opportunity that is much more important than these direct losses. The EU and the US may want to get a slice of the market, but the EU25’s high level of discrimination against ASEAN will have to be alleviated to do so. Except for China, whose late WTO accession in 2003 forces its tariffs down before our shock simulations, Asian imports increase very fast. The 24 per cent increase in ASEAN10 imports is large but less than its export gains. This reflects the rather low initial protection of that zone (inclusion of Singapore in ASEAN10 tends to lower the average tariff rate) compared to the protection it is facing. India registers a record 50 per cent variation in its imports, corresponding to a reduction to zero of its more than 50 per cent tariff rate vis-a`-vis its Asian partners, and 57 per cent for ASEAN. ASEAN is the main benefactor. For example, Indian imports of fats from ASEAN grow more than four-fold which corresponds nearly to a US$9 billion increase. Other Asian countries are being ousted. Japan and Korea register similar import increases of around 20 per cent. In three cases out of four it means that their balance of trade is deteriorating. For South Asia the decline in imports (1.03 per cent) is very far from balancing its export decline (10.3 per cent). The rest of the world – European countries excepted – generally reduces its imports more than it reduces its exports. In the US case (a 0.67 per cent decline in imports compared to a 0.47 per cent export decline) this was a consequence of a relative increase in Asian producer prices. Indeed, the decline in Asian consumer prices, due to the decline

Comparing Free Trade Agreements 299

in tariffs, produces an increase in production volumes, and therefore an increase in producer prices, and, in the case of ASEAN, an exchange rate appreciation (see below). In general all European countries reduce their exports and their imports as well by small margins. Their losses in exports are the result of the diversion effect while their import decline is more a consequence of higher Asian prices. EU25 countries reduce their imports less than their exports (0.21 per cent compared to 0.24 per cent) thereby deteriorating their trade balance. 11.4.3

Impact on production and prices

Overall, agro-food production increases for ASEAN10, while it decreases by 4.4 per cent for India, reflecting initial differences in protection and competitiveness. For the rest of the world, production decreases generally by less than 0.3 per cent, except in China and Korea where it decreases by 0.6 per cent, i.e. twice the decrease for Japan (0.27 per cent) (see Table 11.5). Price and production changes generally move in the same direction: where production increases, prices tend to increase. ASEAN production grows between 3.8 per cent in 2010 to 8.3 per cent in 2015; ASEAN prices grow by 2.8 per cent. Indian prices decline by 2.6 per cent and its production by 4.4 per cent. Some differences appear between Japanese and Korean trends: the decline in Korean production is twice as large as Japan’s. So is the Korean price decline. This is a consequence of structural differences between the two countries. Although both maintain high protection in the agricultural sector, they are in different positions. The share of agricultural output in the economy is larger in Korea than in Japan, and since Korea is a small economy compared to Japan, it is much more open. The combination of the two factors makes Korea far more sensitive to an agricultural shock than Japan. South Asia combines a 0.24 per cent decline in production with a 0.22 per cent decline in prices. The impact on the EU25 is limited, being nil on prices, and weak on production (0.05 per cent). The USA is a little more affected, with 0.06 per cent and 0.01 per cent, respectively. 11.4.4

Trade in industrial goods

Two factors influence industrial trade: the elimination of tariffs between ASEAN10 and its four partners and the changes in agricultural goods trade, which have to be compensated to assure equilibrium. Movements in the industrial goods trade balance generally compensate the

300 Quantitative Implications of Economic Integration Table 11.5

Impact on production and prices (in %, SC1) Agro-food production (vol)

ASEAN-10 Japan Korea China India Hong Kong, Taiwan, Rest of East Asia South Asia EU-25 EFTA Russian Fed North Africa Rest of Europe USA Canada Mexico and Central America South America Australia and New Zealand Rest of the world

Agro-food production price

2010

2015

2020

2025

2010

2015

2020

2025

3.79 0.00 0.30 0.32 2.91 0.07

7.00 0.12 0.04 0.46 4.08 0.10

7.93 0.20 0.35 0.52 4.31 0.18

8.34 0.27 0.61 0.56 4.42 0.22

4.13 0.62 1.30 0.08 2.81 0.07

2.99 0.60 1.25 0.02 2.48 0.04

2.82 0.55 1.24 0.04 2.55 0.03

2.81 0.52 1.25 0.08 2.60 0.03

0.05 0.03 0.02 0.09 0.01 0.00 0.00 0.01 0.01

0.15 0.01 0.07 0.18 0.01 0.04 0.04 0.06 0.02

0.20 0.03 0.10 0.24 0.01 0.06 0.05 0.09 0.04

0.24 0.05 0.12 0.28 0.01 0.07 0.06 0.12 0.05

0.13 0.00 0.01 0.10 0.07 0.00 0.04 0.02 0.04

0.14 0.00 0.00 0.12 0.12 0.00 0.02 0.02 0.04

0.17 0.00 0.00 0.13 0.16 0.00 0.01 0.03 0.04

0.22 0.00 0.00 0.15 0.19 0.00 0.01 0.05 0.05

0.08 0.16 0.19 0.22 0.17 0.20 0.22 0.26 0.18 0.13 0.11 0.11 0.06 0.03 0.05 0.06 0.01 0.04 0.07 0.09 0.02 0.06 0.08 0.10

Source: MIRAGE.

movements on agricultural goods. ASEAN10 increases in imports are higher (þ6.6 per cent) than export increases (þ5 per cent). The same is true for China with 6.2 per cent and 3.4 per cent respectively (see Table 11.6). For the same reason, Japan and Korea improve their position on industrial goods, compensating for their losses on agricultural goods with an increase of 3.5 per cent and 3.4 per cent respectively on the export side and 1.9 per cent and 2.2 per cent for their industrial imports. India is the exception and has its industrial imports increasing (28 per cent) more than its exports (20 per cent). In the case of India, increases in other products such as primary products exports compensate for this. Generally speaking, industrial exports for the rest of the world decrease more than their imports; this is due, for the most part, to a diversion effect, with Japan and Korea benefiting from a better access to the ASEAN markets.

Comparing Free Trade Agreements 301 Table 11.6 Trade in industrial goods (in %, SC1)

ASEAN-10 Japan Korea China India Hong Kong, Taiwan, Rest of East Asia South Asia EU-25 EFTA Russian Fed North Africa Rest of Europe USA Canada Mexico and Central America South America Australia and New Zealand Rest of the world

Industrial exports (volume)

Industrial imports (volume)

2010

2010

2015

2020

2025

2015

2020

2025

4.34 4.91 5.06 5.07 7.00 6.60 6.53 6.58 2.39 2.94 3.25 3.47 1.01 1.35 1.66 1.94 2.82 3.20 3.31 3.38 1.49 1.79 2.00 2.21 2.18 2.88 3.22 3.41 4.09 5.06 5.71 6.22 17.34 19.04 19.62 20.25 20.60 23.80 26.01 28.10 0.94 1.74 2.05 2.21 0.55 0.55 0.54 0.54 0.17 0.28 0.23 0.25 0.69 0.27 0.29 0.04 0.17

0.08 0.48 0.49 0.35 0.97 0.34 0.63 0.34 0.25

0.15 0.59 0.66 0.39 1.17 0.38 0.79 0.49 0.32

0.17 0.69 0.81 0.42 1.37 0.42 0.91 0.57 0.38

0.29 0.09 0.08 0.20 0.33 0.04 0.22 0.08 0.15

0.13 0.11 0.13 0.25 0.42 0.02 0.11 0.17 0.20

0.07 0.10 0.15 0.27 0.51 0.01 0.05 0.23 0.24

0.06 0.09 0.15 0.29 0.59 0.00 0.01 0.26 0.26

0.95 1.35 1.57 1.76 0.49 0.68 0.79 0.87 0.88 1.22 1.42 1.58 0.44 0.48 0.48 0.49 0.09 0.20 0.23 0.24

0.03

0.03

0.05

0.08

Source: MIRAGE.

In the EU25 case, the reduction in exports is significant (0.69 per cent) while reduction in imports is low (0.09 per cent). Consequently equilibrium will be achieved through service exports. 11.4.5

Terms of trade changes

As previously stated, movements in real exchange rates are used in the MIRAGE model as a tool to balance current accounts. ASEAN10, Japan and Korea showed RER appreciation of 2.2 per cent, 0.3 per cent and 0.1 per cent respectively. These movements, which go through prices to affect terms of trade (Table 11.7), explain a relative decline in export competitiveness vis-a`-vis the rest of the world, in order to compensate for their gains within the region. But terms of trade changes are also the result of many price changes such as those affecting product prices. Table 11.8 shows that the most protected products such as rice or sugar get higher price rises.

302 Quantitative Implications of Economic Integration Table 11.7

Terms of trade (in %, SC1) Terms of trade

ASEAN-10 Japan Korea China India Hong Kong, Taiwan, Rest of Asia South Asia EU-25 EFTA Russian Fed North Africa Rest of Europe USA Canada Mexico, Central America South America Australia, New Zealand Rest of the world

2010

2015

2020

2025

1.40 0.00 0.13 0.44 3.24 0.25 0.16 0.03 0.03 0.08 0.21 0.00 0.09 0.03 0.06 0.23 0.14 0.01

1.29 0.04 0.12 0.49 3.28 0.19 0.09 0.02 0.04 0.10 0.25 0.02 0.05 0.04 0.07 0.25 0.19 0.00

1.28 0.03 0.15 0.51 3.33 0.19 0.10 0.02 0.04 0.11 0.28 0.02 0.03 0.04 0.07 0.27 0.21 0.00

1.35 0.02 0.20 0.53 3.37 0.20 0.13 0.02 0.03 0.12 0.33 0.03 0.03 0.05 0.08 0.29 0.23 0.00

Source: MIRAGE.

On the other side of the spectrum industrial goods register a decline in prices, notably all the mechanical sector, from the steel industry to the automotive industry, reflecting the effect of scale economies. Indian RER depreciates by 2.4 per cent and China’s by a minor 0.1 per cent. As a consequence, terms of trade tend to deteriorate when RER depreciates. This is true in all cases, except for Korea where RER appreciation coincides with a deterioration of its terms of trade, to be explained later. 11.4.6

Major changes in bilateral trade

As expected, the most important changes are registered in the trade within East Asia, and more precisely for ASEAN countries that are at the centre of the scenario (Table 11.9). On the ASEAN export side India appears to be the major market for ASEAN’s primary products (natural gas in the case of ASEAN) and food fats which are also among the main comparative advantage (their exports largely exceed their imports) of ASEAN. In 2003 India remained with South Asia a closed market for primary goods (with an MFN tariff of 44 per cent) in contrast to most

Comparing Free Trade Agreements 303 Table 11.8 Impact on world prices (in %, SC1) World prices for developing countries

Rice Sugar Fishing ForWoodPap Dairy LiveAnMeat VegFruits CerealsOil FoodFats BevTobacco BusinSer FinanceIns AnProdWool OthSer FibersCrop TexClothLe TrT Com Primary Electronic Machinery Chemicals TrspEqNec MotorVeh OthMetal OthManuf MetalProd FerMetal

2010

2015

2020

2025

2.74 3.41 1.67 0.76 2.89 1.10 0.43 1.05 1.38 0.88 0.41 0.42 0.46 0.49 0.27 0.03 0.23 0.18 0.23 0.27 0.18 0.41 0.08 0.39 0.05 0.01 0.13 0.19

2.55 1.82 1.38 0.96 1.42 0.77 0.52 0.95 0.78 0.59 0.56 0.42 0.44 0.39 0.31 0.13 0.21 0.17 0.17 0.18 0.08 0.12 0.00 0.12 0.01 0.08 0.15 0.19

2.67 1.58 1.35 1.04 1.15 0.78 0.61 0.83 0.68 0.56 0.56 0.44 0.46 0.40 0.34 0.18 0.20 0.18 0.16 0.16 0.05 0.04 0.01 0.08 0.04 0.13 0.21 0.24

2.83 1.45 1.41 1.15 1.02 0.82 0.71 0.70 0.69 0.58 0.57 0.47 0.47 0.43 0.35 0.25 0.20 0.19 0.16 0.15 0.05 0.02 0.02 0.06 0.08 0.17 0.27 0.30

Note: Developing countries have a different basket of imported goods and export prices are different. Source: MIRAGE.

other zones which have reduced their external tariffs on their imported inputs (but which may have imposed large domestic taxes). The second market for ASEAN is China for chemicals, machinery and electronic goods. The third market is Japan for sugar, textiles and clothing and dairy products. These results show a consolidation of the different comparative advantages of each partner. For Japan, which is the second winner of the scenario, its export gains towards ASEAN are concentrated in the motor vehicle industry and in

304

Table 11.9

Major changes in bilateral trade (US$ million, SC1)

Fifteen major variations in millions dollars Decreases

Increases Products

Exporter

Import country

Variation

Products

Export country

Import country

Variation

Primary Chemicals MotorVeh FoodFats Machinery TexClothLe Electronic Sugar TexClothLe Electronic CerealsOil TrspEqNec Machinery Dairy TexClothLe

ASEAN ASEAN Japan ASEAN ASEAN ASEAN China ASEAN ASEAN ASEAN ASEAN China Japan ASEAN ASEAN

India China ASEAN India China China ASEAN Japan Japan China Korea ASEAN ASEAN Japan India

22247.24 14139.07 12310.15 11374.44 9520.15 7915.07 7003.43 6387.81 5830.71 5745 5293.68 5278.85 5220.82 5136.95 4520.61

Electronic Machinery MotorVeh Electronic TexClothLe Primary Chemicals Primary TexClothLe Machinery TexClothLe TrspEqNec Chemicals Chemicals FerMetal

ASEAN ASEAN ASEAN ASEAN ASEAN ASEAN Korea ASEAN Japan Japan China ASEAN ASEAN Japan ASEAN

ASEAN ASEAN ASEAN Japan ASEAN Japan China ASEAN China China Japan ASEAN ASEAN China ASEAN

8806.29 6564.74 3600.79 2266.75 1752.5 1729.32 1642.74 1513.47 1509.65 1475.58 1400.3 1055.03 1047.44 1000.34 951.3

Source: MIRAGE.

Comparing Free Trade Agreements 305

the machinery sector. For China the products benefiting most from regionalization are first the electronics industry vis-a`-vis ASEAN, Japan and Korea and then transport equipment other than cars.

11.5 11.5.1

An Asian single market? Main results

This scenario assumes that there is a single global agreement between the ASEAN10þ4 countries (even if there are differing time patterns to achieve that goal, depending on the development level of the partners). Compared to scenario 1, the scenario 2 hypothesis roughly increases the effects of the regionalization process by 50 per cent on average but with very large differences depending on the country or sector concerned (Table 11.10). For example, the gains of ASEAN10 are reduced by around 40 per cent compared to those of scenario 1 and those of its four partners are twice as high. The main factor behind ASEAN10’s weakening is that ASEAN has to face its partners’ competition in third markets: Chinese competition in the Japanese market or Korean competition in the Chinese market etc. Therefore ASEAN10 loses some market shares. All Asian FTA members are winners, but the gains are more equally shared than in scenario 1: India with a 4 per cent increase (2.2 per cent in scenario 1), ASEAN and Korea both with 3.5 per cent (4.4 per cent and 0.5 per cent respectively in scenario 1), Japan with 2.46 per cent (instead of 0.75 per cent in scenario 1), and China with 1.9 per cent (instead of 0.8 per cent in scenario 1). The rest of the world loses, with Hong Kong and Taiwan as the main losing zone with 0.83 per cent in GDP instead of 0.41 per cent, South Asia following with 0.48 per cent instead of 0.27 per cent, South America with 0.41 per cent compared to 0.31 per cent, and Australia and New Zealand with 0.37 per cent compared to 0.23 per cent. The EU25 loses only 0.11 per cent instead of 0.08 per cent. As mentioned earlier, real exchange rates tend to appreciate in these kinds of scenarios when GDP increases. In general, a GDP increase results partly because export increases are stronger than import increases and so the balance of trade needs to be re-established through RER appreciation. India is the sole exception to that rule here with an RER depreciation of 2.73 per cent against a 2.4 per cent depreciation in scenario 1. The same basic explanation as for scenario 1 results remains valid here. Major changes in RER concern Japan and Korea. Their currency appreciates by 1.61 per cent and 1.73 per cent instead of 0.27 per cent

306

Table 11.10 Macroeconomic results (% change, SC2) GDP (volume)

ASEAN-10 Japan Korea China India Hong Kong, Taiwan, Rest of East Asia South Asia EU-25 EFTA Russian Fed North Africa Rest of Europe USA Canada Mexico and Central America South America Australia and New Zealand Rest of the world Source: MIRAGE.

Real effective exchange rate

Welfare

2010

2015

2020

2025

2010

2015

2020

2025

2010

2015

2020

2025

2.62 1.99 4.22 2.13 3.87 0.58

3.03 2.28 3.92 2.18 4.12 0.79

3.33 2.40 3.65 2.07 4.07 0.84

3.50 2.46 3.47 1.94 4.04 0.83

1.80 1.20 1.82 0.16 3.16 0.37

1.54 1.41 1.65 0.15 2.88 0.38

1.53 1.53 1.68 0.06 2.80 0.40

1.60 1.61 1.73 0.03 2.73 0.43

0.99 0.37 1.91 0.35 0.44 0.28

1.43 0.41 1.64 0.27 0.37 0.42

1.75 0.43 1.31 0.23 0.31 0.47

1.98 0.45 1.04 0.21 0.23 0.49

0.40 0.10 0.03 0.23 0.20 0.08 0.08 0.03 0.07 0.29 0.26 0.16

0.42 0.11 0.06 0.24 0.25 0.10 0.06 0.03 0.06 0.33 0.30 0.21

0.45 0.11 0.09 0.27 0.29 0.11 0.06 0.05 0.08 0.37 0.33 0.24

0.48 0.11 0.12 0.32 0.33 0.11 0.06 0.07 0.10 0.41 0.37 0.28

0.34 0.00 0.02 0.24 0.17 0.00 0.03 0.01 0.03 0.29 0.19 0.08

0.39 0.01 0.02 0.25 0.21 0.01 0.01 0.03 0.00 0.32 0.22 0.10

0.42 0.01 0.01 0.28 0.24 0.02 0.02 0.02 0.00 0.36 0.25 0.12

0.45 0.01 0.00 0.32 0.27 0.02 0.03 0.00 0.00 0.40 0.28 0.15

0.12 0.01 0.04 0.12 0.09 0.02 0.00 0.03 0.02 0.07 0.10 0.05

0.12 0.02 0.06 0.16 0.11 0.03 0.00 0.03 0.04 0.09 0.13 0.09

0.12 0.02 0.08 0.19 0.13 0.04 0.00 0.04 0.06 0.10 0.15 0.10

0.13 0.02 0.09 0.22 0.14 0.04 0.01 0.04 0.08 0.11 0.16 0.12

Comparing Free Trade Agreements 307

and 0.13 per cent in scenario 1. If we add volume and value change we find an increase of 4.07 per cent for Japan and 5.20 per cent for Korea. The Chinese currency remains almost stable. For the rest of the world the decline in real gross product is compounded by their currency depreciation. As matter of fact, South Asia, Hong Kong and Taiwan, South America, Australia and New Zealand are the main losers in this scenario. Canada and the US countries minimize their losses while the EU25, EFTA and Russia remain a little ahead of America. As for welfare gains, ASEAN continues to improve its welfare more than Japan. It appears that ASEAN is able to improve its position on agricultural goods over time. China and India fare worse in this scenario, probably because Japan and Korea benefit more from the reduction of industrial tariffs. 11.5.2

Impacts on trade in agricultural goods

In terms of the impact on trade in agricultural goods (Table 11.11), Chinese agricultural exports are nearly tripled! Korea’s exports increase by more than half and even Japanese exports grow by almost 20 per cent, but these increments in percentages do not mean much as their initial level was very low. Losses for the rest of the world are also twice as high, except for South Asia. Losses for the EU25 go from 0.24 per cent in scenario 1 to 0.99 per cent in scenario 3, mostly the result of the diversion effect. On the import side, the impacts are much less impressive compared to scenario 1 as the only change is that there are more possible partners than in scenario 1. But, in that scenario ASEAN10 was in a position to take full advantage of the liberalization process. The major difference is for Korean imports, which jump from a 23.7 per cent increase in scenario 1 to a 72.8 per cent increase in scenario 2. China becomes the main winner with a ‘cereals oil’ gain of US$13 billion and a US$2 billion gain for rice. These results are not shown here in more detail due to limitations of space. 11.5.3

Impact on production and prices

Apart from ASEAN10, the scenario 2 impacts on production and prices (Table 11.12) are more important than in scenario 1, being two to three times higher, although the comparison between the two scenarios is based on small numbers in scenario 1. In Korea’s case, production increases by 12 per cent while there was a decline in scenario 1 and prices decreased by 8 per cent. In terms of prices

308

Table 11.11

Trade impacts on agricultural goods (%, SC2) Agricultural exports (volume)

ASEAN-10 Japan Korea China India Hong Kong, Taiwan, Rest of East Asia South Asia EU-25 EFTA Russian Fed North Africa Rest of Europe USA Canada Mexico and Central America South America Australia and New Zealand Rest of the world Source: MIRAGE.

Agricultural imports (volume)

2010

2015

2020

2025

2010

2015

2020

2025

26.91 21.55 49.09 96.16 65.25 0.91 6.55 0.64 0.71 0.11 1.09 1.08 2.83 0.79 1.07 0.23 0.96 1.87

29.30 19.25 53.35 124.92 70.69 1.37 7.78 0.78 0.86 0.06 1.10 1.17 2.96 0.93 1.08 0.34 1.16 1.94

29.01 18.24 55.23 156.66 78.77 0.85 9.30 0.87 0.95 0.44 1.08 1.29 2.87 1.03 1.03 0.42 1.10 2.05

27.69 18.41 57.03 198.20 88.09 0.11 11.12 0.99 1.03 0.98 1.09 1.48 2.82 1.16 1.01 0.56 1.06 2.21

16.91 26.36 72.18 10.75 52.53 2.25 1.70 0.54 0.53 1.20 0.81 0.41 0.99 0.66 0.42 1.32 1.90 0.94

16.92 28.58 77.01 9.36 55.25 2.99 1.82 0.61 0.60 1.34 1.02 0.51 0.93 0.72 0.40 1.59 2.28 1.07

17.21 29.09 75.22 8.08 55.11 3.06 1.87 0.60 0.62 1.38 1.11 0.51 0.85 0.75 0.40 1.69 2.24 1.10

17.50 28.93 72.81 7.03 54.53 2.94 1.90 0.57 0.62 1.40 1.17 0.47 0.80 0.79 0.38 1.77 2.21 1.13

Table 11.12 Impact on production and prices (in %, SC2) Agro-food production (vol)

ASEAN Japan Korea China India Hong Kong, Taiwan, Rest of Asia South Asia EU-25 EFTA Russian Federation North Africa Rest of Europe USA Canada Mexico, Central America South America Australia, New-Zealand Rest of the world

Agro-food production prices

2010

2015

2020

2025

2010

2015

2020

2025

3.02 0.62 12.94 0.28 2.96 0.11 0.15 0.06 0.21 0.58 0.06 0.05 0.08 0.22 0.08 0.25 0.16 0.09

5.92 0.92 14.59 0.45 3.92 0.50 0.28 0.13 0.31 0.85 0.08 0.10 0.14 0.43 0.15 0.38 0.37 0.19

6.98 0.99 13.19 0.59 4.05 0.62 0.33 0.17 0.35 0.89 0.08 0.12 0.15 0.51 0.17 0.41 0.39 0.23

7.58 0.99 11.64 0.69 4.10 0.67 0.36 0.20 0.37 0.90 0.08 0.13 0.15 0.57 0.19 0.44 0.38 0.27

3.08 0.06 7.68 0.71 3.41 0.34 0.39 0.03 0.02 0.33 0.17 0.02 0.07 0.08 0.09 0.31 0.24 0.12

2.15 0.31 8.79 0.52 2.92 0.28 0.41 0.02 0.00 0.26 0.19 0.02 0.02 0.01 0.01 0.28 0.25 0.12

2.01 0.49 8.41 0.32 2.85 0.26 0.43 0.01 0.01 0.25 0.22 0.02 0.01 0.00 0.00 0.30 0.24 0.13

2.01 0.58 7.92 0.13 2.79 0.26 0.46 0.01 0.01 0.28 0.25 0.02 0.02 0.01 0.00 0.34 0.26 0.16

Source: MIRAGE.

309

310 Quantitative Implications of Economic Integration

it seems that the decline in agricultural imported inputs prices translates into a decline in production prices for the food industry. These changes are impressive and result from the initial closed nature of the Korean market (as well as the Japanese market). A total removal of tariffs implies a very big shock where Korea switches to a net surplus in its bilateral trade with Japan. For ASEAN10, production still increases by 7.58 per cent against 8.34 per cent, while prices increase by 2.01 per cent against 2.81 per cent in scenario 1. For the EU25, the decline is four times the decline registered in scenario 1 (0.20 per cent compared to 0.05 per cent) with almost no impact on prices. It is one of the weakest declines in the world with the US at 0.15 per cent. 11.5.4

Trade in industrial goods

Percentage changes in the trade in industrial goods may seem modest compared to those seen for agriculture but they are significant nevertheless, notably because the volume of trade is much more important (Table 11.13). This is the case in particular for developed countries such as Japan or Korea where industrial products cover more than 95 per cent of their trade. In relative terms, Korea, Japan and China multiply their gains respectively by less than five (from 3.4 per cent to 15.1 per cent), a little less than three and more than two. India goes from a 20 per cent increase in scenario 1 to a 29 per cent increase in scenario 2. ASEAN10 sees its former gains reduced from 5.07 per cent to 3.49 per cent; these changes clearly reflect intense competition from the other four partners in the regional agreement. Other countries tend to lose by 1 or 2 per cent, except for Hong Kong and Taiwan who are the main losers with a total loss of 4.8 per cent. The impact on EU25 exports is much more important than in scenario 1 with a decline of 1.32 per cent compared to 0.69 per cent and it is also higher than in the case of agriculture. On the import side, the changes are greater: Indian import changes are greatest with an increase of 43 per cent instead of 28.10 per cent in scenario 1, followed by China with almost a 20 per cent increase against 6.22 per cent, and next is Korea with an 11.92 per cent increase against 2.21 per cent for scenario 1. Opening the markets of Asian countries for industrial products appears to be even more fruitful than the opening of the market for agricultural goods (except naturally for ASEAN). All the non-ASEANþ4 countries reduce their imports, including from the EU25, due to a slower growth, but less than their exports.

Table 11.13 Impact on industrial trade (in %, SC2) Industrial exports (volume)

ASEAN-10 Japan Korea China India Hong Kong, Taiwan, Rest of East Asia South Asia EU-25 EFTA Russian Fed North Africa Rest of Europe USA Canada Mexico and Central America South America Australia and New Zealand Rest of the world

Industrial imports (volume)

2010

2015

2020

2025

2010

2015

2020

2025

3.48 8.41 12.40 7.08 27.33 2.82 1.29 0.84 0.77 0.92 1.47 0.95 0.88 0.15 0.45 1.84 1.61 0.76

3.67 9.33 14.56 8.02 28.37 4.21 1.25 1.11 1.08 0.99 1.83 1.22 1.26 0.25 0.50 2.24 1.86 0.90

3.65 9.67 14.96 8.28 28.49 4.61 1.12 1.23 1.27 1.02 2.00 1.28 1.44 0.32 0.55 2.44 2.00 0.93

3.49 9.78 15.14 8.24 28.96 4.82 0.98 1.32 1.43 1.05 2.16 1.29 1.58 0.35 0.57 2.59 2.12 0.94

5.79 7.30 10.17 16.64 34.36 1.65 1.09 0.49 0.45 0.79 0.86 0.47 0.75 0.23 0.48 1.15 1.13 0.49

5.59 7.89 10.66 18.28 37.73 1.54 1.04 0.51 0.51 0.75 0.98 0.49 0.60 0.22 0.52 1.39 1.16 0.50

5.59 8.15 11.28 19.12 40.17 1.49 0.96 0.48 0.52 0.77 1.05 0.48 0.53 0.23 0.53 1.49 1.16 0.48

5.64 8.31 11.92 19.71 42.83 1.46 0.89 0.44 0.52 0.81 1.11 0.44 0.48 0.22 0.53 1.57 1.17 0.45

Source: MIRAGE.

311

312 Quantitative Implications of Economic Integration Table 11.14 Terms of trade (% change, SC2) Terms of trade

ASEAN Japan Korea China India Hong Kong, Taiwan, Rest of Asia South Asia EU-25 EFTA Russian Federation North Africa Rest of Europe USA Canada Mexico, Central America South America Australia, New-Zealand Rest of the world

2010

2015

2020

2025

0.90 1.39 0.67 0.71 3.99 0.55 0.32 0.05 0.05 0.16 0.29 0.02 0.17 0.05 0.10 0.40 0.41 0.10

0.82 1.51 0.61 0.84 3.92 0.46 0.28 0.04 0.05 0.15 0.31 0.01 0.11 0.04 0.10 0.39 0.39 0.09

0.82 1.50 0.81 0.91 3.85 0.45 0.28 0.03 0.05 0.17 0.35 0.00 0.09 0.04 0.11 0.40 0.39 0.09

0.87 1.48 0.97 0.94 3.82 0.46 0.29 0.02 0.05 0.20 0.39 0.01 0.08 0.04 0.10 0.42 0.41 0.09

Source: MIRAGE.

11.5.5

Terms of trade changes

ASEAN10, Japan and Korea improve their terms of trade while China’s and India’s terms deteriorate (Table 11.14). ASEAN does worse than in scenario 1, while Japan and Korea do much better, in spite of a decrease in international industrial prices. With better access to new markets, Japan and Korea see their export prices increase relatively while they benefit from the decreases in international prices. ASEAN loses ground to its partners. On agricultural goods this is for two reasons: first, it appears that international prices are lower than in scenario 1; second, new competition reduces its capacity to increase its export prices. India and China lose more with respect to their terms of trade than in scenario 1 due to intensification of competition. More generally, it appears that for international prices (Table 11.15) there is a tendency to get lower prices than those in scenario 1. Manufacturing prices in particular are declining which tend to introduce a deflating trend in the world economy. These are the results of two main developments. Reduced tariffs bring automatically lower prices for the consumer, and increased production and trade tends to allow for more scale economies and therefore reduced prices.

Comparing Free Trade Agreements 313 Table 11.15 World prices for developing countries (% change, SC2) Impact on world prices for developing countries

MetalProd FerMetal OthManuf Electronic Chemicals OthMetal Machinery MotorVeh TrspEqNec TexClothLe Primary Com TrT CerealsOil FibersCrop OthSer FinanceIns FoodFats BusinSer AnProdWool BevTobacco VegFruits LiveAnMeat ForWoodPap Dairy Fishing Sugar Rice

2010

2015

2020

2025

0.29 0.33 0.13 0.05 0.13 0.06 0.05 0.48 0.00 0.15 0.10 0.09 0.15 0.47 0.13 0.30 0.25 0.76 0.21 0.52 0.67 0.28 0.81 0.68 2.54 1.29 2.96 1.83

0.30 0.32 0.23 0.19 0.15 0.13 0.04 0.18 0.08 0.18 0.07 0.08 0.10 0.31 0.17 0.23 0.25 0.40 0.31 0.41 0.45 0.35 0.50 0.74 1.17 1.08 1.51 1.61

0.40 0.38 0.32 0.25 0.25 0.18 0.11 0.15 0.10 0.13 0.06 0.08 0.08 0.20 0.18 0.22 0.26 0.31 0.32 0.36 0.40 0.42 0.51 0.74 0.92 1.04 1.32 1.69

0.51 0.45 0.38 0.31 0.28 0.24 0.17 0.14 0.14 0.04 0.05 0.07 0.08 0.11 0.17 0.22 0.26 0.30 0.32 0.32 0.39 0.49 0.54 0.79 0.80 1.06 1.20 1.81

Note: Developing countries have a different basket of imported goods and export prices are different. Source: MIRAGE.

11.5.6

Major changes in bilateral trade

One major result of a multilateral liberalization (multilateral is used here as opposed to bilateral as in scenario 1 which relates to a sum of four bilateral agreements centred on ASEAN, while here we talk of twentyfour bilateral agreements) is that every possible bilateral link benefits from accrued trade flows. This can be seen in Table 11.16: in other words the simulation gives results that benefit all Asian partners. China–Korea relations are noticeable as Korea increases its textile exports

314 Quantitative Implications of Economic Integration Table 11.16 Major changes in trade (US$ million, SC2) Bilateral Major Variations (scenario 2) Major Increases Primary TexClothLe CerealsOil Chemicals Machinery Electronic MotorVeh FoodFats TexClothLe Electronic FoodFats Machinery Electronic TexClothLe MotorVeh Sugar

Major Decreases ASEAN Korea China ASEAN Japan China Japan ASEAN China China Korea ASEAN China Japan Japan ASEAN

India China Korea China China EU25 ASEAN India Japan ASEAN Japan China Japan China China Japan

21 074.4 18 913.3 13 555.6 12 665.4 12 085.7 11 699.1 11 474.2 11 362.3 9 283.1 8 765.1 7 773.3 7 511.1 7 421.5 7 292.8 7 257.1 6 296.2

Electronic TexClothLe Machinery Electronic Machinery Machinery MotorVeh TexClothLe TexClothLe Primary Machinery TrspEqNec Electronic MotorVeh OthManuf Electronic

ASEAN ASEAN Japan Korea EU25 Korea EU25 EU25 EU25 ASEAN ASEAN Korea Korea Korea EU25 ASEAN

EU25 EU25 EU25 EU25 China EU25 China China Japan Japan EU25 EU25 ASEAN EU25 India Japan

               

3,979.7 3,529.8 2,994.0 2,505.0 2,421.6 2,311.3 2,007.7 1,465.7 1,449.2 1,431.7 1,402.7 1,391.6 1,228.2 1,216.1 1,204.0 1,157.9

Note: the table should be read as follows: ASEAN export to India US$ 21 billion more in primary products. Source: MIRAGE.

to China by US$19 billion, while China increases its cereal oils exports by US$14 billion. Among the declining flows, the story is even clearer, as the EU25 loses out on both directions: as an exporter it loses to China in machinery, motor vehicles and textile clothing; as an importer, the EU25 imports less from ASEAN in electronics and textiles. There are only three intra-Asian cases where ASEAN loses; two cases are on the export side and are connected with Japan in primary goods and in electronics (in that case it is to the profit of China, US$7.5 billion). One case is on the import side again, in electronics, to the detriment of Korea. Trade gains are quite different between scenario 1 and scenario 2: in scenario 2, ASEAN’s export gains are lower, while gains between non-ASEAN partners are much higher. There is a substitution of non-ASEAN countries increasing their gains to the detriment of ASEAN countries. 11.5.7

The two other scenarios and general conclusions

As described earlier, scenario 3 is a hub and spoke kind of scenario, like scenario 1 but with a limited liberalization. Most industrial goods are

Comparing Free Trade Agreements 315

liberalized while many agricultural goods are not when classified as sensitive (or highly sensitive) products. Indeed, except for surplus countries, Asian countries are very likely to resist full liberalization on agricultural products and in particular for some highly sensitive products such as rice and processed rice products, and sugar. Japan and Korea, which are otherwise among the most open countries, are also expected to resist any change in these two areas. On the other hand, there is no bilateral talk between China and Japan, and any discussion seems difficult between Japan and Korea; therefore, scenario 3 could be considered as the most plausible scenario. Limited liberalization naturally reduces the impact of Asian integration on the world economy and notably for East Asian countries. For ASEAN10, which is at the centre of this scenario, the impact on GDP, RER and welfare is reduced by 46–47 per cent for the first year (2010) and 36–38 per cent in the final year of the simulation, compared to scenario 1. Given the fact that in scenario 1, the impact on GDP and welfare increases over time, the diminishing gap between scenario 3 and scenario 1 for GDP implies that the longterm impact of limited liberalization remains quite important. One reason for such convergence is related to the declining role of agriculture in ASEAN economic activities. One may conclude that with the passage of time and with the development of the ASEAN countries, the resistance to agricultural liberalization will diminish, but Japan and Korea (not to mention the European countries) show how difficult it may be. In the case of scenario 4 similar changes appear with scenario 2. In conclusion, up until 2001, the year of China’s entry to the WTO, there have been many discussions about regional integration within the Asia Pacific region, but until that date very little actual progress was made. On the one hand, the US promoted a large ‘open regionalism’ such as the APEC forum including the notion of the Pacific Basin Community. This project failed to make progress when the discussion focused on sensitive issues such as the will of the US to include agricultural goods liberalization in the project and get firm commitments. On the other hand, the Asian countries failed to create what Malaysia called an Asian caucus, because the US and its allies, Korea and Japan, refused to support such an initiative. ASEAN, as a kind of compromise, was enlarged to include the four countries of Indochina and started to implement a limited but effective FTA between its members. Then came the financial crisis in 1997–8, followed by SARS, the tsunami and now the avian flu crisis. All these events showed the weakness of regional cooperation. China, as an emerging trading power, will in the long term challenge the US in Asia and therefore is looking for more regional

316 Quantitative Implications of Economic Integration

integration. On the trade issue its arguments are clear: it is one of the less protected economies of Asia, even more open than Japan on agriculture. Its trade balance is negative with almost all its regional partners, and access to China is vital for them, as China brings monetary stability and dynamism to that zone. There are naturally a lot of problems to be resolved before we can talk of a regional FTA in the region, and there are several road maps to achieve that. We have explored a variety of possible issues: the creation of a fully multilateral and global (including agricultural) East Asian FTA, or the creation of an ASEAN hub with a number of bilateral agreements with every member of the region, as well as a weak version of these two projects with limitations for so-called ‘highly sensitive’ products. Agriculture is a central problem, but there is an almost general consensus within the region that agriculture should be treated separately. For many ASEAN countries, notably for the new members, agriculture still represents a major source of employment for many poor people. This is also true for China and more so for India. Even in Japan and Korea the rice question has become a kind of cultural and social identity matter rather than an economic problem. Singapore and Hong Kong are the exceptions. However, we have shown that it would be in the clear interest of ASEAN10 to include agriculture, so we cannot exclude the possibility that there will be strong pressure within East Asian countries to include part of agriculture in any FTA agreement. The main results of each scenario can be synthesized according to two macroeconomic measures which indicate the benefits for each country or zone in terms of its GDP value (summing GDP growth in volume with RER changes). It is clearly in the interest of ASEAN10 to have separate bilateral negotiations within the region and to include agricultural products (scenario 1). This will give ASEAN easier access to its main natural partners and – as it is not directly in competition with them – this will bring the greatest benefits to ASEAN. Southeast Asia is specialized in agricultural and food production, which are in short supply elsewhere, and ASEAN is potentially very competitive. The main problem would be the lagging countries such as Vietnam or even worse, Myanmar. Korea has, in contrast, the largest interest in negotiating a global agreement (this is also the Korean president’s position), but excluding sensitive products (scenario 4). This will give Korea greater access to the Chinese and Japanese markets for its industrial products and also to the rest of Asia. A bilateral agreement with ASEAN including sensitive

Comparing Free Trade Agreements 317

products would have tough consequences on agriculture, surpassing the gains on industrial products and would therefore be unacceptable; the worst scenario for Korea is scenario 2. Japan is a winner in any case, even and especially if the FTA includes a liberalization of agriculture, in which case its benefits are much greater than its losses. The most advantageous scenario by far is scenario 2. Japan, which until the late 1990s was a strong advocate of multilateral agreements negotiated under the auspices of the GATT, has changed its policy towards regionalization for two reasons: first, the world trend towards regionalization as exemplified by NAFTA and the EU25, and second, China’s extremely active economic diplomacy within Asia raises the risk of Japan becoming isolated. Japan, however, still has a serious handicap with its inability to come to terms with its post-war legacies. Therefore it cannot promote a multilateral agreement within the region. Its farmers seem able to resist any change in its agriculture policy. So it is more likely that Japan will continue its strategy of negotiating bilateral agreements within the region which, in the end, is the worst scenario for Japan from a strictly economic point of view. China is leading the regionalization process for political as well as for economic reasons. For political reasons, China wants to become a leader of the emerging Asia, and for economic and strategic reasons it needs to secure its vital supplies of raw materials. Japanese security is taken care of by the US, whereas China has to fend for itself. Being a latecomer to the WTO, it has had to engage in a radical reshuffling of its customs as well as its tariffs which are the lowest among Asian developing countries, so this gives China considerable room for manoeuvre. The best scenario for China is scenario 1. However, in that case it appears that China will become the focal point of the zone as its weight makes it the first partner to be dealt with. The other partners may find this difficult to accept from a diplomatic point of view. For India, the major task is to discard its traditional protectionist policy, which is one of the most restrictive in the world and therefore the shock could be devastating in social terms. Therefore it is probably reasonable to think of a much more gradual involvement of India in the process of liberalization, and to define a more restrictive list of products to be more or less excluded from the liberalization process. A global agreement with limitations for sensitive products (scenario 4) would be the best from an economic and social point of view but from a purely economic point of view scenario 2 is better.

318 Quantitative Implications of Economic Integration

In the end we see that ASEAN10þ4 countries have diverging interests. If only economic factors were taken into account, a simple average of their preferred scenarios gives scenario 2 as the number one scenario for that region, and scenario 4 as number two, i.e. a multilateral agreement excluding sensitive products would be the second preferred scenario. However, this result does not mean that these are the more plausible scenarios as political factors have not been taken into account. Other Asian countries such as Taiwan and Hong Kong, which supposedly are excluded from these regional integration scenarios, suffer from a trade diversion effect. Not surprisingly, the worst scenario for these countries is the global scenario in which ASEAN10, China, Japan, Korea and India maximize their gains as a whole. For the EU25, the consequences are limited and almost nil in terms of welfare. However, considering the EU25’s position as already very weak in terms of market shares, it will become even weaker in Asia. Asia is the largest economic zone in the world and the most dynamic. This makes it more worrisome for the EU and calls for a strong reaction, in other words to engage in negotiations with ASEAN10. The most difficult part of a deal for the EU25 would be the agricultural issue with ASEAN. Progress made at the multilateral level would greatly improve the EU’s position. The preferred scenario for the EU25 would be scenario 3, a bilateral agreement within Asia excluding sensitive products, and the worst scenario would be scenario 2. There is a relative homogeneity of interest in the region. If a bilateral agreement is possible within ASEANþ4 then it may also be easier for the EU25 to try to negotiate bilateral agreements with these countries or zones. The US has different interests: its favourite scenario should be scenario 4, a multilateral East Asian scenario but excluding sensitive products. As the US produces primary products, it is better for it to keep its market access on an equal footing with Asian producers. Finally, the rest of the world is rather close to European positions on scenario 3.

Notes 1. The þ3 are Japan, South Korea and China; the þ6 include India, Australia and New Zealand which were invited to the Kuala Lumpur Summit in December 2005. 2. Baldwin (1997) noted that ‘Regionalism in Asia . . . would certainly be an important event, but has yet to happen.’

Comparing Free Trade Agreements 319 3. Most CGE studies on Asian integration are centred almost exclusively on its impact on the participants: see Kawasaki (2003), Gilbert et al. (2001), Fukase and Martin (2001), Otsubo (2005). 4. Established in 1967 by five founding members: Indonesia, Malaysia, the Philippines, Singapore and Thailand. It was later joined by Brunei when it gained independence from Great Britain. The initial objective at the time of the Vietnamese war was to foster regional stability and to promote political and economic cooperation. 5. The Common Effective Preferential Tariff (CEPT). The CEPT is the mechanism by which tariffs on final goods traded within the ASEAN region, which meet a 40 per cent ASEAN content requirement, were to be reduced to 0–5 per cent by the year 2002/3 (2006 for Vietnam, 2008 for Laos and Myanmar, and 2010 for Cambodia). ASEAN members have the option of excluding products from the CEPT in three cases: (1) temporary exclusions; (2) sensitive agricultural products; (3) general exceptions. Temporary exclusions refer to products for which tariffs will ultimately be lowered to 0–5 per cent, but which are being protected temporarily by a delay in tariff reductions. For a limited number of sensitive agricultural products, the deadline will be extended to 2010. In an agreement that has yet to be fully spelled out, the process of tariff reduction on these products began between 2000 and 2005, depending on the country and the product. General exceptions refer to products which a country deems necessary for the protection of national security, public morals, the protection of human, animal or plant life and health, and the protection of articles of artistic, historic or archaeological value. Approximately 1 per cent of ASEAN tariff lines fall into this category. The CEPT scheme was to cover nearly 98 per cent of all tariff lines in ASEAN by the year 2003; by then, the only products not included in the CEPT scheme were to be those in the general exceptions category and sensitive agricultural products. In the longer term, the ASEAN countries have agreed to enact zero tariff rates on virtually all imports by 2010 for the original signatories and 2015 for the four newer ASEAN members. 6. Since June 2005, two agreements with South Korea have been signed: one with Singapore in August 2005 and one with EFTA in September 2005. 7. Normal track means that 90 per cent of all the tariff lines and 10 per cent of the total value of imports from Korea or from the ASEAN member countries as a whole, as appropriate, based on 2004 trade statistics. 8. Lists of sensitive products of each one of ASEAN members have been taken from countries’ declarations posted on the ASEAN website. 9. The exception here is Malaysia, for which the sensitive products vis-a`-vis China are replicated vis-a`-vis India, Japan and Korea. 10. In the geographical breakdown used here, India is separated from South Asia which should be named as ‘Rest of South Asia’.

References Baldwin, Richard E. (1993). ‘A Domino Theory of Regionalism’, NBER Working Paper Series, no. 4465. Baldwin, Richard E. (1997). ‘The Causes of Regionalism’, The World Economy, 20, 7 (November).

320 Quantitative Implications of Economic Integration Bandara, S. and W. Yu (2003). ‘How Desirable is the Asian Free Trade Area? A Quantitative Economic Assessment’, The World Economy, 26, 9: 1293–1323. Dimaranan, Beatrix and R. A. McDougall (2002). Global Trade Assistance and Production: the GTAP 5 Data Base. Center for Global Trade Analysis, Purdue University. Feridhanusetyawan T. (2005). ‘Preferential Trade Agreements in the Asia-Pacific Region’, IMF Working Paper, WP/05/149. Frankel, Jeffrey, Ernesto Stein and Shang-jin Wei (1993). ‘Continental Trading Blocs: Are they Natural or Super Natural?’, NBER Working Paper, no. 4588, December. Fukase, E. and W. Martin (2001). ‘Free Trade Area Membership as a Stepping Stone to Development: the Case of ASEAN’, World Bank Discussion Paper, no. 421, February. Garnaut, Ross (2005). Australia, US and China: Open Regionalism in an Era of Bilateral FTAs, Melbourne: Asialink. Gilbert, J., R. Scollay and B. B. Bora (2001). ‘Assessing Regional Trading Arrangements in the Asia-Pacific’, Policy Issues in International Trade and Commodities Services, no. 15, UNCTAD. Graham, E. and P. Krugman (1989), Foreign Direct Investment in the United States, Washington, D.C.: Institute for International Economics. Kawasaki, K. (2003). ‘The Impact of Free Trade Agreements in Asia’, RIETI Discussion Paper Series, no. 03-E-018. Krugman P. (1991). ‘The Move Towards Free Trade Zones’, paper presented at the ‘Policy Implication of Trade and Currency Zones’ symposium sponsored by the FRB of Kansas City. Otsubo, S. (2005). ‘Computational Analysis of the Economic Impacts of Japan’s FDI in Asia’, Forum of International Development Studies, 28. Viner, Jacob (1950). The Customs Union Issue, New York: The Carnegie Endowment for International Peace. Wei, Shang-jin and Jeffrey A. Frankel (1998). ‘Open Regionalism in a World of Continental Trade Blocks’, IMF Staff Papers, 45, 3.

12 How Will ASEANþ3 Integration Accelerate Investment? A CGE Analysis Ken Itakura

12.1

Introduction

Economic integration in East Asia progressed during the 1990s and has rapidly continued to evolve towards a free trading area after entering the twenty-first century. Sixty-six bilateral free trade agreements (FTAs) exist between the ASEAN member countries, Japan, China and Korea (ASEANþ3). These account for 85 per cent of possible combinations of bilateral free trade agreements within the ASEANþ3 countries (Kimura et al., 2006). Led by the ASEAN Free Trade Agreement (AFTA) in the early 1990s, negotiations and implementations of FTA in the region developed rapidly in subsequent years. As of 2006, the remaining 15 per cent needed to complete the possible combinations among the ASEANþ3 includes the bilateral trade liberalizations of large economies such as Japan, China and Korea. Given the size of their economies, the remaining FTAs will have potential impacts on trade, production and investment in the three countries, as well as in the ASEAN member countries. Although trade liberalization of the ASEANþ3 may bring more trade, production and investment to the region, the ASEAN countries have expressed concerns that potential economic benefits from the integration could mostly be channelled to the large economies, especially to fast-growing China. They are more concerned over the investments coming from abroad, since the economic growth of ASEAN has been fuelled by attracting foreign investment to manufacturing industries. One of the major issues that the ASEAN countries need to consider is whether the investment from abroad will be diverted outside of ASEAN, once the ASEANþ3 FTA have been established. The aim of this study is to analyse the potential economic impacts of the ASEANþ3 FTA by using a multi-sector multi-region recursively 321

322 Quantitative Implications of Economic Integration

dynamic Computable General Equilibrium (CGE) model. To capture the impacts on inter-sectoral linkages within a country and international trade flows across regions, it is natural to use a multi-sector multi-region comparative static CGE model. The widely used platform for this type of CGE analysis is the Global Trade Analysis Project (GTAP) database and modelling framework (Hertel, 1997; Dimaranan and McDougall, 2002). Accounting for the dynamics arising from capital accumulation, the Dynamic GTAP model (Ianchovichina and McDougall, 2001) is utilized for this study to shed some light on the impact of the ASEANþ3 FTA on investment. In section 12.2, the basic mechanism of investment theory which is assumed in the Dynamic GTAP model is explained in greater detail. The data aggregation and simulation settings used in this chapter are discussed in section 12.3. Results from the ASEANþ3 FTA simulations are reported in section 12.4, and concluding remarks follow in section 12.5. From the simulation results of the ASEANþ3 FTA, we will see in section 12.4 that establishing the ASEANþ3 FTA can bring economic benefits in terms of increasing GDP and investment in the participating countries of the FTA. In fact, the simulation results show that all of the ASEANþ3 countries attract more foreign investment. These findings from the simulation results imply that the ASEANþ3 FTA have the potential to accelerate investment in the region, and that all the countries in the FTA would have gains from integration by committing to freer trade.

12.2

Investment theory in the Dynamic GTAP model

Ianchovichina and McDougall (2001) extend the comparative static framework of the standard GTAP model originally developed by Hertel (1997) to the comparative dynamic framework by incorporating international capital mobility and capital accumulation. In the standard comparative static GTAP model, capital can move across industries within a country (or a region), but not across borders. For long-run analysis to be more realistic, the Dynamic GTAP model needs a mechanism to capture incentives to invest in different regions, consequently allowing international capital mobility and capital accumulation. The Dynamic GTAP model preserves all the features of the standard GTAP model such as constant return to scale production technology, perfectly competitive market, and product differentiation by origin which is known as the Armington assumption (Armington, 1969), while enhancing the investment theory in the model to incorporate international capital mobility and ownership. The Dynamic GTAP model uses the

How Will Integration Accelerate Investment?

323

standard GTAP database (Dimaranan and McDougall, 2002) supplemented with foreign income data from the IMF Balance of Payments statistics to track international capital mobility and foreign wealth. In the Dynamic GTAP model, each of the regions is endowed with fixed physical capital stock owned by domestic firms. The physical capital is accumulated over time with new investments. This dynamics is driven by the net investment, which is sourced by the regional household savings. Regional households own indirect claims on physical capital in the form of equity. There are two types of equity: equity in domestic firms and equity in foreign firms. The household directly owns the domestic equity but only indirectly the foreign equity. For accessing equity in foreign firms, the household must own shares in a portfolio of foreign equities provided by the ‘global trust’ which is assumed to be the sole financial intermediary for all foreign investments. Values of the household equity holdings in domestic firms and in the global trust evolve over time, and the household allocates its entire savings for investment. Collecting such investment funds across regions, the global trust reinvests the funds in firms around the world and offers a portfolio of equities to households. The sum of the household equity holdings in the global trust is equal to the global trust equity holdings in firms around the world. The savings in one region will be invested directly in domestic firms and indirectly in foreign firms through the global trust, which in turn reinvests in all regions. The dynamics arising from positive savings in one region are related to the dynamics from the net investment in other regions. At the global level, it must follow that all the savings across regions are completely invested in both home and overseas markets. In theory, incentives for investments or equity holdings are governed by rates of return, which will be equalized across regions if the capital is perfectly mobile. However, this equalization of rates of return seems unrealistic, at least in the short run. Further, there exist well-known empirical observations of so-called ‘home bias’ in savings and investment, equity holdings by households, and capital flows. Home bias refers to the observed fact that the domestic market is preferred to the foreign market. This empirical observation suggests that the capital is not perfectly mobile, thus leaving varying rates of return across regions. The Dynamic GTAP model allows inter-regional differences in rates of return in the short run, which will eventually be equalized in the longer term. This may be regarded as a realistic approach, but it calls for a mechanism to allocate equity holdings of the households and the global trust in a way consistent with the observed data. Differences in rates of

324 Quantitative Implications of Economic Integration

return are attributed to the errors in investor expectations about the future rate of returns. However, the errors in expectation are gradually adjusted to the actual rate of return as time elapses. Eventually the errors are eliminated and a unique rate of return across regions can be attained. Therefore, the model assumes that perfect capital mobility applies only in the longer term, but not in the short run. Investment is the result of a gradual movement of expected rates of return to equality across regions, but the expected rate of return may differ from the actual rate of return due to errors in expectations. Explicit modelling of the ownership of regional investment allows for determination of the accumulation of wealth by foreigners. In addition, the ownership of domestic and foreign assets can also be tracked. Income accruing from the ownership of these foreign and domestic assets can then be appropriately incorporated into the total regional income. Participating in FTA could lead to more investment from abroad. Trade liberalization tends to make prices of goods from a participating country of FTA cheaper due to reducing tariffs, and then creating increase in demand for the goods. Responding to the increased demand, production of goods expands in the exporting country. The production boost is attained by employing more intermediate inputs, labour, capital and other primary factor inputs. As a consequence, these increased demands for production inputs raise the corresponding prices, wage rates and rental rates. Higher rental rates are translated into higher rates of return, attracting more investment from both home and foreign countries. These are the basic processes which translate FTA positively to investment.

12.3

Data aggregation and simulation design

This section explains the settings for data and simulations for the analysis of the ASEANþ3 FTA. First, we describe aggregation and extension of the GTAP database (Dimaranan and McDougall, 2002), and then take an overview of the basic set-up of the simulation analysis as developed by Walmsley et al. (2002). This basic set-up provides a reference for the simulation analysis of the FTA. 12.3.1

Data aggregation

The data set used in this chapter is the GTAP database, version 5.0 (Dimaranan and McDougall, 2002). The GTAP database covers 66 regions in the world with 57 sectors, and corresponds to the year 1997. Even though it might be possible to conduct a simulation at the fully

How Will Integration Accelerate Investment?

325

disaggregated level of the 66 regions and the 57 sectors, its computational cost will be enormous and simply impractical. Therefore, it is necessary to aggregate the GTAP database up to some smaller sets of regions and sectors. The 66 regions are aggregated to 19 regions (Table 12.1), and the 57 sectors are aggregated into 23 sectors (Table 12.2).1 For the Dynamic GTAP model, the GTAP database (Dimaranan and McDougall, 2002) is extended by incorporating information on foreign income payments and receipts. This extension of the GTAP database enables the Dynamic GTAP model to determine household equity holdings in domestic firms and foreign firms, and allocation of firms’ ownership between domestic households and foreign households.

Table 12.1 Aggregation of regions Regions

GTAP regions

Japan Singapore Indonesia Malaysia Philippines Thailand Vietnam Korea Taiwan Hong Kong China USA Canada Mexico AusNzl CSAmerica

Japan Singapore Indonesia Malaysia Philippines Thailand Vietnam Korea Taiwan Hong Kong China USA Canada Mexico Australia, New Zealand Central Am., Carib, Colombia, Peru, Venezuela, Argentina, Brazil, Chile, Uruguay, Rest of South America Austria, Belgium, Denmark, Finland, France, Germany, United Kingdom, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, Sweden, Switzerland, Rest of EFTA Bangladesh, India, Sri Lanka, Rest of South Asia Hungary, Poland, Rest of Cent. Eur., Former S.U., Turkey, Rest of Mid-East, Moroc., Rest of N. Africa, Bots., Rest of SACU, Malawi, Moz., Tanz., Zam., Zimb., Other S. Africa, Uganda, Rest of Sub-Saha. Afr., Rest of World

WEuro

SAsia ROW

326 Quantitative Implications of Economic Integration Table 12.2

Aggregation of sectors

Sectors

GTAP sectors

Rice Grains Other crops

Paddy rice; Processed rice Wheat; Cereal grains nec Veg., fruit, nuts; Oil seeds; Sugar cane; Fibres; Crops nec; Wool, silk-worm cocoons Cattle, sheep, goats; Animal products nec; Meat products nec Raw milk; Veg. oils and fats; Dairy; Sugar; Food products nec; Bev & tobacco Forestry Fishing Coal; Oil; Gas; Minerals nec Textiles; wearing apparel Leather products Wood products; Paper products, publishing Petroleum, coal products; Chemical, rubber, plastic prods; Mineral products nec Ferrous metals; Metals nec; Metal products Motor vehicles and parts Transport equipment nec, Machinery and equipment nec, Manufactures nec Electronic equip. Electricity; Gas; Water; Business services nec; Recr. and oth. services; Dwellings Construction Wholesale/retail trade Transport nec, Sea transport, Air transport Communication Financial services nec, Insurance PubAdmin/Defence/Health/Education

Meat Other food Forestry Fish Extract Texwap Leather Paperwood Chemical Metal Auto Machinery Electrequip Other services Construction Trade Transport Comm Insfinance Pubservice

12.3.2

Simulation design: baseline

The purpose of the baseline scenario for the Dynamic GTAP model is to establish the likely evolution of trade, production, consumption, investment and other variables under a set of assumptions. The baseline scenario used in this chapter is based on the work by Walmsley et al. (2002). Simulation of the ASEANþ3 FTA is defined as a policy scenario against the baseline scenario. Thus, results from the policy scenario are obtained in the form of deviations from the baseline scenario, by comparing the counterfactual policy scenario against the baseline scenario. The baseline scenario contains information about macroeconomic variables as well as expected policy changes. In addition to the projections

How Will Integration Accelerate Investment?

327

for macroeconomic variables, the expected policy changes included in the baseline scenario are those which are already agreed and legally binding (e.g. Uruguay Round commitments and China’s WTO accession, ASEAN’s Common Effective Preferential Tariff scheme). Bilateral FTAs within the region are also included in the baseline scenario, such as Japan-Singapore FTA, New Zealand-Singapore FTA, Singapore-Australia FTA, USA-Singapore FTA, Thailand-Australia FTA, Thailand-New ZealandFTA, Japan-Mexico FTA, Korea-Singapore FTA, and Japan-Malaysia FTA. In sum, the baseline scenario is based on two sets of projections: the projections for macroeconomic variables and the projections about policies. In the next section, policy simulations of the ASEANþ3 FTA are examined by imposing additional policy changes to the baseline scenario. Bilateral tariff cuts between ASEANþ3 countries are the main components in policy experiments, differing from the baseline scenario. While the projection of population growth is maintained exactly the same as in the baseline scenario, all the other macroeconomic projections may be influenced by the tariff cuts. Of course, GDP and investment will respond to the additional policy shocks of the tariff cuts, generating deviations from the baseline scenario. By formulating the policy experiments in this way, potential economic impacts of the ASEANþ3 FTA can be identified and measured in deviations from the baseline scenario.

12.3.3

Simulation design: policy

The tariffs and subsidies are key variables in simulating the trade liberalization between ASEANþ3 countries, since they play a crucial role as external shocks given to the Dynamic GTAP model simulation. Average tariff rates are reported in Table 12.3, and which is a matrix of trade-weighted, bilateral tariffs between country pairs. Relatively high average tariff rates are observed in several ASEAN countries, such as Vietnam, Thailand and Malaysia. The average tariff rates of Vietnam and Thailand on imports from Japan, China and Korea are relatively high. It can be anticipated that removal of tariffs by the ASEANþ3 FTA would create large responses in economic variables for Vietnam and Thailand, because all tariffs between ASEANþ3 countries are completely eliminated.2 Tariffs reported in Table 12.3 cover the merchandise tariffs, which exclude potential barriers on service sectors. The ASEANþ3 FTA may include liberalization of service trade, and estimates of barriers on service trade are necessary components for the FTA simulation. However, it is difficult to obtain such estimates because data on services trade and

328 Table 12.3

Average tariff rates (2010, %)

Japan

Korea

China

ASEAN

Singapore

– 4.4 4.6 1.9 0.0 1.1 0.3 3.7 5.2 3.8

7.4 – 11.9 2.6 0.0 3.1 3.0 4.9 4.5 8.3

6.5 7.0 – 5.1 4.7 3.0 5.6 4.3 8.6 3.0

5.3 7.2 7.4 2.0 2.5 3.6 1.4 1.1 1.6 4.8

0.0 0.0 0.1 0.1 – 0.1 0.1 0.2 0.0 0.2

Indonesia

Malaysia

Philippines

Thailand

Vietnam

7.6 15.6 7.0 1.0 0.2 – 2.5 0.8 2.7 2.2

0.0 2.4 6.6 2.3 1.9 5.2 – 0.6 2.2 20.3

6.7 6.5 10.9 1.1 0.9 3.6 1.9 – 3.4 7.4

17.0 12.5 15.1 4.6 4.7 4.4 5.1 3.8 – 0.5

15.6 17.2 24.2 14.1 12.9 7.6 17.3 4.8 17.8 –

Source: Author’s simulation results.

potential barriers are scarce and inaccessible. Bilateral merchandise tariffs and subsidies on traded goods and services are completely removed in the policy experiment of the ASEANþ3 FTA. Degrees of such removals of the tariffs consist of exogenous shocks given to the Dynamic GTAP model. All the shocks are given in 2010, when the counterfactual FTA takes place. This specification of policy experiment is a standard setting for simulating likely effects of trade liberalization.

12.4

Simulation results3

Potential economic impacts of the trade liberalization between the ASEAN countries, Japan, China and Korea are simulated with the Dynamic GTAP model by completely removing bilateral tariffs and subsidies on goods and services in 2010. Simulation results are obtained for the period 2010–20. Most of the results are reported in terms of

How Will Integration Accelerate Investment?

329

14.0 13.0 12.0 11.0 10.0 9.0 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Japan

Singapore

Philippines

Korea

Indonesia

Thailand

China

Malaysia

Vietnam

Figure 12.1(a) FTA effect on GDP (% point differences from baseline) Source: Author’s simulation results.

percentage point differences between the baseline scenario and the policy scenario. By focusing on such deviations from the baseline, the likely effect of the ASEANþ3 FTA can be measured. Figure 12.1(a) reports the impact of the ASEANþ3 FTA on GDP from 2010 to 2020, measured in percentage point difference from the baseline scenario. All the countries participating in the FTA show an increase in GDP relative to the baseline scenario. Vietnam and Thailand’s GDP indicate the largest impacts of the FTA. Vietnam’s GDP rises to 12.8 per cent in 2015 and falls to 8 per cent by 2020, as compared to the baseline, while Thailand’s GDP shows a continuous increase resulting in 8.8 per cent. Korea’s GDP in 2020 is about 3 percentage points higher than the baseline, and the Philippines (1.5 per cent) and Indonesia (1.4 per cent) also show modest increases in GDP. The effect on GDP resulting from the FTA can be attributed to accumulation of physical capital. All of the ASEANþ3 countries except Malaysia show higher capital accumulation than the baseline scenario, as shown in Figure 12.1(b). Physical capital stock in Vietnam and Thailand increases about 12 percentage points above the baseline, followed by Korea (6.4 per cent). As we can see, the rise and fall in Vietnam’s GDP also correspond

330 Quantitative Implications of Economic Integration 29.0 27.0 25.0 23.0 21.0 19.0 17.0 15.0 13.0 11.0 9.0 7.0 5.0 3.0 1.0 1.0

2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Japan

Singapore

Philippines

Korea

Indonesia

Thailand

China

Malaysia

Vietnam

Figure 12.1(b) FTA effect on capital accumulation (% point differences from baseline) Source: Author’s simulation results.

to the changes in physical capital stock. For Malaysia, accumulation of physical capital results in slightly lower (0.3 per cent) than the baseline. Looking at the FTA effect on the capital accumulation from different points of view, the impact of the FTA on investment is reported in Figure 12.2. Investments in most of the countries are positively affected by the ASEANþ3 FTA. Vietnam’s investment shows a large swing as inferred from the changes in capital stock. Even if the investment resulting from the FTA becomes lower than the baseline, it does not necessarily become a ‘negative’ investment, depleting physical capital stock. This point can be discerned by looking at Figures 12.3 and 12.5. The hump shape of capital accumulation in Vietnam in Figure 12.1 reflects the effect of the FTA on investment shown in Figure 12.2, changing the large positive difference from the baseline into the negative. However, from Figure 12.3, we can see that both in the baseline and the policy scenario Vietnamese investment increases by about 80 per cent in 2020 from 2010. For Malaysia in Figure 12.2, the FTA impact measured by the percentage point differences does not seem significant, but the level of investment increased about by 50 per cent in the baseline

How Will Integration Accelerate Investment?

331

100 80 60 40 20 0 20 40 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

Figure 12.2

Japan

Singapore

Philippines

Korea

Indonesia

Thailand

China

Malaysia

Vietnam

FTA effect on investment (% point differences from baseline)

Source: Author’s simulation results.

and the policy scenario (Figure 12.4). Rather, physical capital continues to grow over time because of the investment stream. Liberalized trades in ASEANþ3 stimulate production activities in Vietnam, generating more demands for intermediate inputs and production factors. Higher rates of return in Vietnam, reflecting the increase in rental rates, attract larger investment. As investment continues to flow in, accumulating physical capital stock for production, rental rates in Vietnam begin to fall, and then lead the rate of return decline. Unlike other countries, Vietnamese investment shows a downturn. The reason for the big swing of investment in Vietnam may be as follows. Firstly, the magnitude of tariff reduction rates in Vietnam is the largest among the simulated countries. Secondly, the economic size of Vietnam is relatively small. Due to these combined factors, positive effects as well as rebound of FTA or zero tariffs on the Vietnamese economy are expected to be large.

332 2.5

2.0

1.5

1.0

0.5

0.0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Baseline Policy

Figure 12.3 Investment in Vietnam for baseline and policy scenario (2010 ¼ 1.0) Source: Author’s simulation results.

1.7

1.4

1.2

0.9

0.7 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

Baseline Policy

Figure 12.4

Investment in Malaysia for baseline and policy scenario (2010 ¼ 1.0)

Source: Author’s simulation results.

How Will Integration Accelerate Investment?

333

120 100 80 60 40 20 0 20 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Japan

Singapore

Philippines

Korea

Indonesia

Thailand

China

Malaysia

Vietnam

Figure 12.5 FTA effect on foreign ownership of physical capital (% point differences from baseline) Source: Author’s simulation results.

In the Dynamic GTAP model investment is basically sourced by domestic and foreign households’ savings. Therefore, change in investment can be traced back to the change in equity ownerships of the domestic and foreign households, indicating who is providing the investment funds. Figure 12.5 shows the ASEANþ3 FTA effect on foreign ownership. Increase in investment is funded by domestic and foreign households by holding the equity. All the countries indicate more foreign ownership than the baseline, meaning that more foreign investments are coming into the countries. Most notably, foreign ownership of capital stock has almost doubled in Thailand. The higher rate of return to investment resulting from the FTA attracts more investment from abroad to Thailand. This links back to the discussion in section 12.2 about the home bias in equity holdings by domestic households. The higher rate of return also attracts investments from domestic households.

334 Quantitative Implications of Economic Integration

90 80 70 60 50 40 30 20 10 0 10 20 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

Figure 12.6

Japan

Singapore

Philippines

Korea

Indonesia

Thailand

China

Malaysia

Vietnam

FTA effect on export volume (% point differences from baseline)

Source: Author’s simulation results.

As tariffs and subsidies fall as a consequence of the FTA, increased demands for goods and services from FTA member countries stimulate more exports and imports in the regions. For the simulation period of 2010–20, Figure 12.6 illustrates the increased export volume from the FTA member countries, and Figure 12.7 shows the imports volume (see also Tables 12.4 and 12.5). Although all the aggregate volumes of exports and imports are expanded for the FTA member countries as compared to the baseline, breaking down the aggregate volume by sector reports both expansion and contraction by the end of the simulation period in 2020.

12.5

Concluding remarks

The simulation results of the ASEANþ3 FTA show that establishing the FTA can increase the participating countries’ GDP as well as investment.

How Will Integration Accelerate Investment?

335

50 45 40 35 30 25 20 15 10 5 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

Figure 12.7

Japan

Singapore

Philippines

Korea

Indonesia

Thailand

China

Malaysia

Vietnam

FTA effect on import volume (% point differences from baseline)

Source: Author’s simulation results.

Increases in investment observed for all the ASEANþ3 countries indicate more capital accumulation that can account for higher GDP. Additionally, the simulation results indicate that more foreign investment is attracted to the region for a higher rate of return due to the FTA. These findings from the simulation results imply that the ASEANþ3 FTA has a potential for accelerating investment in the region and that all the countries in the FTA can gain from integration by engaging in free trade. It should be noted, however, that simulation results may change by taking different set of assumptions and available data for further exploration into consideration. Recent developments in FTA cover a wider range of negotiations beyond liberalizing trade via reducing tariffs and subsidies, such as facilitation of trade and investment, customs clearance procedures, mutual recognition, and so on. These relatively new elements in negotiations will no doubt influence the simulation results once they are captured by quantitative information.

Effect on export volume by sector in 2020 (% point differences from baseline)

Rice Grains Other crops Meat Other food Forestry Fish Extract Texwap Leather Paperwood Chemical Metal Auto Machinery Electric equip. Construction Trade Transport Comm. Insfinance Pubservice Other services

336

Table 12.4

Japan

Korea

China

Singapore

Indonesia

Malaysia

Philippines

Thailand

Vietnam

0.3 0 0.6 0.6 0.3 0.2 0.1 0.6 5 1.2 0.3 2.8 2.6 1.4 2.6 1 0.1 0 0.4 0.1 0.1 0 0.1

1.4 0.4 1 21.4 4.3 0 0.2 0.4 16.2 16.8 1.9 3.3 3.4 0.4 1.5 2 0 0.8 1.1 1 0.8 0.5 0.6

3.9 3.6 1 1.1 1 0 0 0.2 4.5 3.2 0.7 0.8 0.3 0.6 1.8 5.5 0 0 0.4 0.1 0 0 0

20 3 9.7 3.8 11.5 0.3 0.5 0.7 4.6 6.3 1.9 7.4 5 3.5 2.2 1.7 0 1.2 0.5 0 0.5 0.1 1.3

0.2 0.1 0.7 0 1.2 0.3 0 0.6 9.1 18.7 3 1.4 7.6 4.8 11.6 14.6 0 0 0 0.2 0 0 0

55.4 11.2 4 4.5 4.9 0.6 0.3 4.3 5.1 79.9 2 2.9 2.2 3.6 3.6 0.6 0 1.3 1.8 0 0.2 0.4 1.1

1.1 0.2 6.7 0.4 0.6 0.1 0.2 0.8 16.3 37.6 0.9 0.7 3.4 23.5 10.9 5 0 0.1 0.4 0.3 0.2 0 1.1

2.4 0.1 0 0.7 1.3 0 0.4 0.1 5.8 1.9 1 4.6 9.3 2.4 15.5 23.1 0 0.3 0.7 2.2 0.2 0.1 0

15.9 2.7 9.3 4.9 8.4 1.9 0.5 1.7 188.1 112.6 10 2.5 2.3 9.1 20.2 24.6 0 0.2 5.2 0.1 4.9 0.5 1.1

Source: Author’s simulation results.

Table 12.5

Effect on import volume by sector in 2020 (% point differences from baseline)

Rice Grains Other crops Meat Other food Forestry Fish Extract Texwap Leather Paperwood Chemical Metal Auto Machinery Electric equip. Construction Trade Transport Comm. Insfinance Pubservice Other services

Japan

Korea

China

Singapore

Indonesia

Malaysia

Philippines

Thailand

Vietnam

186.3 0.8 6.1 12.9 4.2 1.2 1.4 3 20.2 23.4 5.4 5.1 7.9 4.5 6.9 7.1 3.6 3.1 2.7 3.5 3 3.8 3.7

26.4 13.7 20.4 13.4 6.3 4 16.4 5.8 27.9 24.4 11.6 9.7 9.7 11 13.4 9 5 5.6 3.7 3 9.2 6.7 3.5

155.3 4.5 4.3 4.2 3.6 0.7 3.1 0.8 11.6 12.9 2.3 5.3 3.9 13.1 4.1 2.9 2.3 1.5 1.3 1.7 1.4 1.2 1.2

3.9 8.7 2.4 1.7 1.9 1.8 0.6 8.4 2.7 4.4 1.8 1.8 2.9 0.3 0.5 0.8 2.1 2 1.7 0.2 1.6 1.9 1.2

5.7 2.3 4.3 3.2 4.8 2.4 3.9 2 29.9 3.3 5.4 5.3 9.9 20.8 4.8 4.7 0.1 0.1 0.5 0.6 0.9 0.1 0.4

428.9 6.6 33.7 38.6 10.2 5.9 4.8 5.8 20.7 36.8 9.4 7.7 2 23.2 3 0.2 1.7 3.9 1.8 2.4 2.7 4.8 0.2

121.3 8.7 19 7.1 5.6 9.6 8.1 0.8 22.9 36.1 7.3 5 7.1 7.5 4.8 4.2 2.1 0.4 0.4 0.4 0.8 0.3 2.2

30.8 7 11.6 16.5 10.2 3 10.7 13 47.3 51 10.6 18.1 22 42.4 20.9 23 19.1 2.8 0.2 2.7 0.5 2.3 0.6

82.3 13.2 38.5 36.6 16.7 33.4 12.9 3.5 126.2 57.3 23.7 16.2 11.6 5.4 4.6 6.9 3.4 5.1 4.3 2.6 2.6 4.6 5.8

Source: Author’s simulation results.

337

338 Quantitative Implications of Economic Integration

Notes 1. The countries involved in the ASEANþ3 FTA are Singapore, Indonesia, Malaysia, the Philippines, Thailand, Vietnam, Japan, China and Korea. While the GTAP database does not distinguish Brunei, Cambodia, Laos and Myanmar, the other ASEAN member countries, these four economies comprise only a small part of the total ASEAN GDP, so that errors associated with the missing ASEAN countries are anticipated to be small. 2. Table 12.3 also displays intra-ASEAN average tariffs for 2010. These tariffs reflect the expected policy changes assumed in the baseline scenario by ASEAN’s Common Effective Preferential Tariff (CEPT) reduction programme. 3. The General Equilibrium Modelling Package (GEMPACK) developed by Harrison and Pearson (1996) is used in this chapter for numerical simulation. All the simulation results discussed here are obtained by the GEMPACK economic modelling software.

References Armington, P. S. (1969). ‘The Geographic Pattern of Trade and the Effects of Price Changes’, International Monetary Fund Staff Paper, 16, 2: 179–99. Dimaranan, B. V. and R. A. McDougall (eds) (2002). Global Trade, Assistance, and Production: the GTAP 5 Database, Center for Global Trade Analysis, West Lafayette, Purdue University. Harrison, J. W. and K. R. Pearson (1996). ‘Computing Solutions for Large General Equilibrium Models Using GEMPACK’, Computational Economics, 9: 83–127. Hertel, T. W. (ed.) (1997). Global Trade Analysis: Modeling and Applications, New York: Cambridge University Press. Ianchovichina, E. I. and R. A. McDougall (2001). ‘Theoretical Structure of Dynamic GTAP’, GTAP Technical Paper, 17. Kimura, F., A. Kuno, H. H. Lee and H. S. Huh (2006). ‘Economic Integration in East Asia and Asia-Pacific and International Commercial Policies in Japan and Korea’, Mita Gakkai Zasshi, 99, 2. Walmsley, T. L., B. V. Dimaranan and R. A. McDougall (2002). ‘A Base Case Scenario for the Dynamic GTAP Model’, GTAP Application Paper, 417.

Index NB: ASEAN – Association of South East Asian Nations ACFTA see ASEAN-China FTA (ACFTA) Africa 256, 296 AFTA (ASEAN free trade area) 63, 97, 285–6, 321 de facto/de jure integration and 7–8, 29, 32–5, 37 local suppliers and 256, 268 noodle bowl syndrome and 47, 49, 56–7, 71 preference margins 58–9, 61 rules of origin (RoO) and 84–5 spillovers/linkages 225, 249 threshold for local content 89 utilization rates 57–8, 58 agglomeration economies 158, 160 defined 144–5 aggregate GL–index 197 agricultural goods 296–9, 297, 307, 308 AICO (ASEAN Industrial Cooperation) scheme 32–3, 41n, 108 air transportation 125, 129–30, 129, 134, 136, 150 AKFTA see ASEAN–Korea FTA (AKFTA) Amiti, M. 226 Anderson, J. E. 15, 150, 166–7, 171, 257 Ando, M. 6, 50, 56 architecture, forming 10–11, 29–30 Armington assumption 322 Asahi Maxima (Laos) Co. Ltd, case study 176–9, 176 ASEAN common aim of 148 cumulative local content 85, 87, 89: automotive sector 107, 107; electronics 104, 104; manufacturing 97–8, 97, 101 industrial location and 11–12 industrialization in 227 intra-ASEAN trade 55, 57, 59, 59 normal Track tariff lines 287

preferential trade arrangements (PTA) 6–7 regional trade arrangements 9–10 structure 77 tariff-cutting and 45 value added-based local content 100 ASEANþ3 3, 29, 283, 286, 288 cumulative local content: automotive sector 116n; electronics 115n; manufacturing 114n, 115n deeper integration prospect 39, 73, 74 regionalism and 10–11 see also under free trade agreements (FTAs) ASEANþ3þ3 see ASEANþ6 ASEANþ4 (ASEAN10þ4) see under free trade agreements (FTAs) ASEANþ6 3, 6, 10–11, 29, 283, 286 preferential tariff schemes and 8 ASEAN Common Investment Climate Initiative 9 ASEAN Economic Community 286 ASEAN Industrial Cooperation (AICO) scheme 32–3, 41n, 108 ASEAN Security Community 286 ASEAN Socio-cultural Community 286 ASEAN Summit meeting (2005) 10 ASEAN-China cumulative local content 89, 100 automotive sector 107, 107 electronics sector 104, 104 manufacturing 97–8, 97, 101 ASEAN-China FTA (ACFTA) 1, 97, 286, 288 de facto/de jure integration and 7, 34, 37 noodle bowl syndrome and 46, 47, 49, 63, 71 rules of origin (RoO) and 84–5 ASEAN-China Summit (2000) 63

339

340 Index ASEAN-Japan CEP agreement 287 ASEAN-Japan Research Institute 162 ASEAN-Japan trade 134, 135, 136 air transportation/ containerization 136 cumulative local content: automotive sector 116n; electronics 115n; manufacturing 114n, 115n machinery/electrical equipment 137 see also under free trade agreements (FTAs) ASEAN-Korea FTA (AKFTA) 1, 7, 34, 37, 48–9, 71, 287 ASEAN-Korean cumulative local content automotive sector 116n electronics 115n manufacturing 114n, 115n Asia Pacific Economic Cooperation (APEC) 11, 284, 290, 315 noodle bowl syndrome and 49, 57, 59, 61–2, 77 Asia Pacific Free Trade Area 284 Asian crisis 5, 33, 226–7, 284–5, 315 Asian Manufacturing Matrix 50, 51–2, 54–5, 55, 61 Asian Tigers 208 Australia 39 ASEANþ4 FTAs and 288, 296, 298, 305, 307 CER and 287 ‘open regionalism’ and 284 outsourced production in 14 per capita GDP 2 SMEs and 260 see also ASEANþ6 Australia-Thailand FTA 249 Austria 208 automotive sector 104–7, 274, 276 import content 105, 105: changes in 106 integral architecture in 108 local content 104–5, 105: ASEANJapan, ASEAN-Korean, ASEANþ3 cumulative 116n; ASEAN/ ASEAN-China cumulative 107, 107; changes in 105

Malaysia/Thailand 153–6 structural changes 106 avian flu crisis 315 Baba, T. 107 backward linkage 193–5, 231, 274 Balance of Payments statistics (IMF) 323 Balassa, B. 196 Baldwin, R. E. 12, 143–4, 160, 288 Banda, O. G. D. 48 Bangkok 158 Bangladesh 14, 288 Beijing 153 Belgium 194–6, 198, 200–1, 203, 208–11, 213–14 benefits, economic integration 14–16 Beyers, W. B. 193 bilateral factors cumulation 83, 113n economic partnership agreements (EPAs) 225 GL-index 197 trade changes 71, 302–5, 304, 313–14, 314 see also under free trade agreements (FTAs) ‘bindings overhang’ 74 ‘Bogor Goals’ (APEC) 57 Bohai Bay Rim 151 Boomsma, P. 190 border-related barriers Laos 170–2: case study 172–9 Bowen, J. 150 Brand-to-Brand Complementation (BBC) 108 Brunei 2, 33, 48, 87 see also ASEAN entries Cambodia 11, 14, 226, 286 see also ASEAN entries; CLMV countries Canada 15, 77, 307 capital accumulation, FTA impact 330 capital goods 211 EU openness and 210–12 Caribbean 121 Carruthers, R. 121

Index 341 catching-up hypothesis 246–7 CER see Closer Economic Relations (CER) CGE (computational general equilibrium) model 283, 287, 321–38 Change in Tariff Classification (CTC) 83, 84 Chenery-Moses-Polenske type IO model 192 Cheong, I. 84 Chile 84 Chile-Korea FTA 84 China 143 ASEANþ4 FTAs and 298, 307, 310 de facto/de jure integration 2–4, 9–14, 31–2, 39 exports 150 industrial agglomeration 151, 152, 153 Laos and 163, 168–9, 170 local content: automotive sector 104; manufacturing and 93–4 local suppliers and 256, 257, 259, 260 noodle bowl syndrome and 45–6, 51–7 passim, 62–9 passim: domino effect 49, 63–4 per capita GDP 157 transport and 121, 125–6 see also ASEANþ3; ASEANþ4; ASEANþ6; ASEAN–China entries; factory economies; Japan-China trade China National Tourism Administration 125 China-Japan FTA 9 China-Thailand FTA 249 Chonburi 4 Clark, X. 119 CLMV countries 2, 12, 87, 162–3 tariffs 33, 288 ‘closed architecture’ FTAs 39 Closer Economic Relations (CER) agreement 287 ‘cold chain’ 122 ‘collective action’ 72, 73

commodity-wise reciprocity 33 common borders, FDI and 258 Common Effective Preferential Tariff (CEPT) 58, 84, 162, 249, 268, 285, 319n Common Investment Climate Initiative (ASEAN) 9 common language, FDI and 258, 266, 276 communication technologies 125, 147, 150 developments in 160 communications costs 266 equipment 261, 272–4, 276 comparative advantage 50, 51, 126, 186 ‘competitive advantage’ 188 components, characteristics 107–8 Comprehensive Economic Cooperation 286 Comprehensive Economic Partnership (CEP) 287 computational general equilibrium (CGE) model 283, 287 see also under free trade agreements (FTAs) computer sector 59 computer-aided product design (CAD) 122 confidence building 74–6 consumption goods 209 EU openness and 208–10 containerization 129–30, 129, 136, 148, 149, 150 main routes/feeder services 154 core–periphery structure 11–13, 143–61 among cities 12–13 among countries 11–12 ‘flying geese pattern’ 158–61 new economic geography (NEG) and 144–7 policy implications 18–19 regional integration 148–58 costs courier service 171 fixed 170–2, 171 international trade 166–7

342 Index costs – continued natural geographic conditions and 119 shipping 121, 166 transport 51, 108 cross-border FDI 268 cumulation 113n rules 83–5 cumulative local content 85, 87, 89 automotive sector 107, 107, 116n electronics 104, 104, 115n manufacturing 97–8, 97, 101, 114n, 115n currency barriers 170 crisis (1997) 5, 33, 226–7, 284–5, 315 de facto integration 3–6, 30–2 factory Asia 3–5 regional integration drive forces 5–6 de jure integration 6–11, 30–2, 186 architecture for 10–11 geographical boundary 38–40 political will 9–10 rules of origins (RoO) and FTA 8 ‘shallow’ integration 9 in terms of trade 32–6 trade liberalization 6–8 de minimis rule, rules of origin (RoO) and 83, 84 Deardorff, A. V. 164 decomposition analysis 110–12 direct input-based 110–11 value added-based 111–12 demand linkage 234–5, 242 Denmark 193, 195, 198, 200, 203, 209–10, 212–14 Department of Domestic and Foreign Investment (DDFI) (Laos) 172 diagonal cumulation 83–4, 85 Dietzenbacher, E. 194–5 Dimaranan, B. V. 322–5 direct linkages, IO tables and 192 discrimination 15, 63–4, 70, 74 dispersion, NEG and 147 disputes 65 Dixit, A. K. 187 domestic demand 188

domestic feedback loops 193 domino theory 63–4, 65, 70 Duranton, G. 120 duty-drawbacks 33, 41n, 55, 58, 69, 79n duty-free MFN treatment 69 Dyer, J. H. 108 Dynamic GTAP model 325–8, 333 East Asia Summit (EAS) 3, 10–11 see also ASEANþ6 East Asian Economic Community (EAEC) 56 East Asian Manufacturing Matrix 68 East Asia’s Challenges (research project) 2 East Asia’s De Facto Economic Integration (Hiratsuka and Kimura) 15 economic partnership agreements (EPAs) 225 economies of scale 246 ‘efficiency-seeking’ FDI 258, 259, 260–1, 266, 272, 274 electrical equipment sector 59, 131, 137 air transportation and 130 electronics industry ASEAN import content 102 import content 102: changes in 103 local content 101–2, 102: ASEANJapan, ASEAN-Korean, ASEANþ3 cumulative 115n; ASEAN/ASEANChina cumulative 104, 104; changes in 102 modular architecture in 108 procurement of parts and components 108–9 structural changes 103 technological changes 102 Eminent Person Group Report 284 Enabling Clause 9, 33, 74, 75, 288 endogenous growth theory models 147 Europe 12, 39, 52, 67, 73 European Coal and Steel Community (ECSC) 186 European Economic Community (EEC) 72, 186–7 European Free Trade Association (EFTA) 76–8, 186, 307

Index 343 European Union (EU) 118, 143, 259 ASEANþ4 FTAs and 285, 296–9 passim, 301, 310, 317–18 EU-3 6 EU-6 193–4, 198 EU-7 203, 206, 208, 210–12 EU-9 193–4, 198 EU-15 6, 262–3 integration and 11, 12, 15, 31 noodle bowl syndrome and 60, 67, 74–5, 76 European Union (EU) structural change 184–218 growing interdependence 192–6 growing intra-industry trade 196–7 inter-country input–output tables 189–92 intra-export specialization 198–201, 200 location theory and 186–9 policy implications 19–20 sectoral specialization 197–201, 199, 200 spatial concentration 201–2 trade trends 203–15 Eurostat 191 Evans, C. 121 export -oriented industries 108 specialization, intra-EU 198–201, 200 volume, FTA impact 334, 336 Export Processing Zones (EPZs) 55, 69 factor cost savings 124 Factory Asia 3–5, 45, 49, 59, 74 interdependence in 65–6, 66 management 70 unilateralism and 50–6 factory economies 50, 52, 55, 62, 68 Fan, C. C. 151, 153 FDI see foreign direct investment (FDI) Fernandes, A. M. 226, 233 financial crises 61–2 Finland 208 ‘flying geese’ pattern 143, 158–61 food and beverage industry 274 Foreign Direct Investment Committee (Laos) 172 foreign direct investment (FDI) 206

AFTA and 9, 31–2, 285 ASEAN intra-regional 268–77, 269, 270, 271, 273, 275 ASEAN LDCs 276–7 ‘efficiency-seeking’ 258, 259, 260–1, 266, 272, 274 ‘function-seeking’ 262 geographical distribution 264–7, 265, 266, 267–8, 267 industries receiving 272–6, 273, 275 inward, Laos 169 Japanese 125, 126, 227, 266 Malaysia 159, 264 ‘market-seeking’ 258, 259–60, 266, 274 outward 257–62: from ASEAN 262–8; patterns 262–4, 263, 264 ‘resource-seeking’ 258–9 sources/hosts 268, 272 spillovers/linkages and 225–7, 234 ‘supplier following assembler’ 261–2, 261, 272 trade facilitation 37 two-way intra-industry 276 types/driving forces 258–9 vertical 163, 164, 165, 169 foreign ownership 333 forward linkages 193–5, 231, 233 Fouquin, M. 163 fragmentation 3–4, 36–7, 36, 120, 124, 164–5 France 195, 198–9, 203, 207–9, 211, 213–14 Frankel, J. 284–5 free trade agreements (FTAs) 1–2, 29, 283–319 ASEAN enlargement 286–90 ASEANþ3 35: CGE analysis 321–38, capital accumulation 330, data aggregation 324–6, 325, 326, export volume 334, 336, foreign ownership 333, import volume 335, 337, investment impact 331, 332, simulation design/results 326–7, 327–8, 328–34, 329, tariff rates 327–8, 328

344 Index free trade agreements – continued bilateral 7–8, 9, 34–5, 37, 38, 225: ASEANþ4 293–305, agricultural goods trade 296–9, 297, bilateral trade changes 302–5, 304, industrial goods 299–301, 301, macroeconomic impacts 294, production and prices 299, 300, 303, terms of trade changes 301–2, 302, welfare impact 295; ASEAN-Japan 1, 34, 37–8, 46, 48–9, 64, 71, 288, networks 29, 30 comprehensiveness of 36 discriminatory features 38 local plant impact 249 MIRAGE model 290–2 multilateral: ASEANþ4 305–18, agricultural goods impact 307, 308, bilateral trade changes 313–14, 314, industrial goods trade 310, 311, macroeconomic impact 306, production/ prices impact 307, 309, 310, 313, terms of trade changes 312, 312 plurilateral 7–8, 35, 38–9, 41–2n policy implications 21–2 resource allocation and 16 rules of origin (RoO) and 8, 83–4, 84 spillovers/linkages and 225, 243 trade facilitation and 9 see also ASEAN-China FTA (ACFTA); ASEAN-Korea FTA (AKFTA); noodle bowl syndrome Free Trade Area for the Asia Pacific (FTAAP) 11 Freudenberg, M. 15, 60 FTA see free trade agreements (FTAs) Fujidenki 4 Fujikawa, K. 88 Fujita, M. 13, 51, 120, 124, 145, 147, 188 Fulita, M. 164 full cumulation 84, 85 ‘function-seeking’ FDI 262 Garnaut, R. 288 GATT see General Agreement on Tariffs and Trade (GATT)

General Agreement on Tariffs and Trade (GATT) 33, 67–76 passim, 187, 201, 317 Enabling Clause 9, 33, 74, 75, 288 general machinery sector 59 geographical factors advantage 122–4 distance 257 proximity 266, 276 trade costs and 119 Germany 195–6, 200, 202–3, 207–8, 210, 212–14 Ghosh model 193, 195 GL see Grubel-Lloyd (GL) indexes Global Trade Analysis Project (GTAP) 322, 324–5 Dynamic model 322–4, 325–8 gravity model 167, 257, 263 Greece 208 Grubel, H. G. 196 Grubel-Lloyd (GL) indexes 196–7, 212, 213–14, 213, 214 GTAP see Global Trade Analysis Project (GTAP) Hamaguchi, N. 51, 124, 145 Hanson, G. H. 164 hard disc drives (HDDs) 3–4 Harrigan, J. 121 Hattari, R. 257 headquarter (HQ) economies 50, 51, 52, 62, 63, 68 Heckscher–Ohlin–Samuelson (HOS) trade theory 186 Hertel, T. W. 322 Hewings, G. J. D. 196 ‘highly sensitive’ products 316–17 Hiratsuka, D. 4, 62, 165–6, 255 Hirschman, A. O. 193 Hoen, A. R. 196–8, 212–13 ‘hollowing out’ 50–2 ‘home bias’ 323 ‘home market effect’ 185, 258–9, 274 Hong Kong 14, 262, 266 ASEANþ4 FTAs and 296, 298, 305, 307, 310, 318 ‘hollowing out’ 50–2 see also headquarter (HQ) economies horizontal competition 245

Index 345 Hummels, D. 120, 164, 166–7 Hungary 12 hypothetical extraction method 194 Ianchovichina, E. I. 322 import content 101 automotive sector 105, 106 changes 96–7 electronics 102 manufacturing 94, 95, 98, 99 import volume, FTA impact 335, 337 import-substitution 33 India 39, 262, 288, 296, 302 outsourced production in 14 see also ASEANþ6; free trade agreements (FTAs) India-Thailand FTA 249 indigenous enterprises 13–14 Indonesia 66, 225–6 import content 101 local content: automotive sector 104; manufacturing and 93; value added-based 98 outsourced production in 14 tariffs 33, 47, 69 see also ASEAN entries; factory economies; Southeast Asia Indonesia-Malaysia preferential tariff structure 47 industrialization ‘flying geese pattern’ 158–61 Laos 168–70, 168, 169 vertical production networks and 164–6 industry agglomeration 145–7, 160 categorization 228, 229, 230 industrial goods 299–301, 301, 310, 311 infant, protection 31 ISIC codes 235, 252 information technology 122, 150 developments in 125 Information Technology Agreement (ITA) 35 Ingenuity Solutions (Malaysia) 268 in-plan import–export 172 input–output (IO) linkages 160, 188 input–output (IO) tables

Asian-Pacific (IDE) 190 Chenery-Moses-Polenske type 192 Isard-type international 86 Thailand 228, 235 input–output (IO) tables, Asian international 52, 82, 85–9, 92 automotive sector 105, 106, 107, 116n electronics 102, 103, 104, 115n manufacturing 93, 96, 98, 100, 101, 114n, 115n sector classification 90–1 input–output (IO) tables, inter-country (EU) 185, 203 Grubel-Lloyd index and 196–7 ‘ideal’ 189, 190 model 189–92 research 192–203 semi-survey construction method 191–2 integral architecture, automotive industry 108 ‘integrated world economy equilibrium’ 31 inter-continental transportation 150 inter-country trade clusters 196 inter-sectoral level 196 multipliers 193 interdependence EU 192–6 in factory Asia 65–6, 66 inter-firm production fragmentation 4 inter-industry trade 185, 186–7, 189 intermediate goods/services 145–6, 261, 272 EU openness and 196, 203–4, 204–8, 204, 207, 210–11 spillovers/linkages and 234–5, 242, 245 international standard industrial classification (ISIC) 228, 233, 237, 245, 249 industry codes 235, 252 international trade fixed costs 170–2, 171 theory 31

346 Index International Trade Center (ITC) 288 International Transportation Handbook (Ocean Commerce Ltd) 148 inter-regional trade 150 intra-AFTA trade 63 intra-ASEAN trade 55, 57, 59, 59 intra-EU trade 118, 208 intra-firm production fragmentation 3–4 intra-industry trade 185–6, 187 China and Japan 128–9 European Union (EU) 196–7: importance 212–13, 213–15, 213, 214 intra-NAFTA trade 118 intra-regional trade 67, 73, 124, 148, 263 world trade share 119 investment theory, Dynamic GTAP model 322–4 IO see input–output entries Ireland 193 Isard, W. 188 ISIC see international standard industrial classification (ISIC) Italy 193–8 passim, 200, 203, 208–10, 214 Jacquemin, A. 187 Japan comparative advantage and 50 de facto/de jure integration 2, 5–6, 9–10, 14, 15, 39 disk-drive assembly and 65–6 domino theory and 63–4 exports 150 foreign direct investment (FDI) by 125, 126, 227, 266 FTAs see under free trade agreements (FTAs) ‘hollowing out’ 50–1 local content 101–2, 104, 108: value added–based 98, 100 plant placement 51 tariffs and 68, 78 triangle trade and 52 see also ASEANþ3; ASEANþ6; ASEAN–Japan entries; headquarter (HQ) economies;

Japan-China FTA 9 Japan-China trade 125–6, 130, 134 air transportation/ containerization 129–30, 129, 134 machine/electrical equipment 131 textile industry 130, 132–3, 134 Japan-Korea FTA 9, 49, 64, 71 Japan-Malaysia FTA 34, 48 Japan-Singapore FTA 48, 84 Japan-Thailand EPA 249 Japan-Vietnam FTA 48 Javorcik, B. S. 226, 231, 234 JETRO surveys 9, 58 on FTA utilization 8 Jones, L. P. 193 ‘just-in-time’ delivery 108, 125 factory, East Asia as 70 inventory management 121 production 5 Kearney, A. T. 262 Keola, S. 164 Khon Kaen 14 Kimura, F. 6, 50, 56, 321 knowledge spillover 147 Konings, J. 226 Korea 256, 266 domino theory and 63–4 FTAs see under free trade agreements (FTAs) hard disc drives (HDDs) and 3 ‘hollowing out’ of economy 50–2 import content, manufacturing 96 local content: automotive sector 104; manufacturing and 93; value added-based 98 regional trade arrangements and 9–10 tariffs and 69–70, 287 trade with China 71 WTO and 78 see also ASEANþ3; ASEANþ6; ASEAN-Korea entries; headquarter (HQ) economies Korea-Chile FTA 84 Korea-Japan FTA 9, 49, 64, 71 Korea-US FTA 64

Index 347 Krugman, P. R. 2, 11–12, 118, 143, 257–8, 284 European Union (EU) and 187–8, 198 vertical production networks 166, 169 Kuchiki, A. 55 Kugler, M. 226, 231 Kuroiwa, I. 150, 184–5 Kwon, K. D. 84 labour cost differential 147 Laos 226 AFTA compliance and 286 APEC and 11 border-related barriers 170–2, 171: case study 172–9 indigenous enterprises and 14 industrialization 168–70 inward FDI 169 production fragmentation and 4 trade volume 168 see also ASEAN entries; CLMV countries Laos–Vietnam Joint Venture 277 LDCs (least developed countries) 167, 276–7 ‘lean retailing’ management 121 least developed countries (LDCs) 167, 276–7 Leinbach, T. R. 150 Leontief model 193, 194–5 liberalization 285 of agriculture 317 limited 314–15 trade 6–9, 16, 242–8, 249, 250, 321, 331 light industry sector 59 Lima˜o, N. 118–9, 167 Lin, P. 226, 231 Linden, J. A. van der 191, 192, 194–6, 198, 199–201 linkages see spillovers and linkages Lloyd, P. J. 196 local content AFTA rules of origin (RoO) 85–6 automotive sector 104–7, 105 calculation of 85–9 change in (C–LC) 94–6

decomposition analysis 110–12 direct input-based 86–8 electronics 101–2, 104 factors affecting 107–9 manufacturing 93–101, 94, 95 policy implications 18 sector changes 89, 93, 95 sector classification 90–1 value-added based 88–9, 98, 99, 100, 111–12 see also cumulative local content local enterprises 255–78 FTA impact 249 intra-regional FDI 268–77, 269, 270, 271, 273, 275 outward FDI 257–62: from ASEAN 262–8 policy implications 21, 277–8 local–multinational plant linkages 228–31 location industry 52, 121, 257–8 Japanese plant 51 MNEs 11–12, 155–6 theory 185, 188 logistics 122 milk-run style of 5 ‘low language barrier’ 170 lubricants, fuels and oil sector 59 Luxembourg 201 machinery sector 59 air transportation and 130 Japan-ASEAN trade 137 Japan-China trade 131 Macmillan proposal, trade diversion and 285 Mahathir, Malaysian Premier 56–7 Malaysia 158 automotive industry 153–6 disc-drive assembly and 65–6 FDI and 159, 264 FTAs see under free trade agreements (FTAs) hard disc drives (HDDs) and 3–4 import content 94 indigenous enterprises and 13–14 industrial location and 52

348 Index Malaysia – continued local content: automotive sector 104–5; electronics sector 101–2; manufacturing 94 MNCs and 165 spillovers/linkages 225–6 tariffs 33, 69 see also ASEAN entries; factory economies; Southeast Asia Malaysia-Indonesia preferential tariff structure 47 Malaysia-Japan FTA 34, 48 manufacturing FDI inflow 273, 275 growth, total 62 import content 94–7, 94, 95, 98, 100: changes in 99 Japanese FDI 125, 126 local content 93–101, 94, 95: ASEAN-Japan, ASEAN-Korean, ASEANþ3 cumulative 114n, 115n; ASEAN/ASEAN-China cumulative 97–8, 101; changes in 99; value-added based 98, 100 structural changes 96, 100: technological 96, 100 Marcouiller, D. 167 market magnification 160 ‘market-seeking’ FDI 258, 259–60, 266, 274 Marrakech accord (1994) 73 Marshall, A. 147 Matsumura, F. 88 McCallum, J. 15 McDougall, R. A. 322–5 McKendrick, D. G. 166 mega ports 148 MERCOSUR, common aim of 148 Mexico 77–8, 121 MFN see most-favoured-nations (MFN) Milner, C. 227 Ministry of Foreign Affairs (MOFA) 10 MIRAGE (Modelling International Relationships in Applied General Equilibrium) model 290–2, 301 MNEs see multinational enterprises (MNEs)

modular architecture, in electronics industry 108 most-favoured-nations (MFN) NEARME priorities and 74 principle 9 tariffs 6, 30, 35, 55–9 passim, 56, 63–78 passim motor vehicles see automotive sector Muendler, M.-A. 226 multilayered supporting industries 145 multinational companies (MNCs) 163, 165 multinational enterprises (MNEs) 153 de facto/de jure integration and 2, 3, 13–14, 32, 38 industrial location and 11–12, 155–6 local suppliers and 255–6 spillovers/linkages 225–7 Thailand 231, 233–5 Myanmar AFTA compliance and 286 APEC and 11 outsourced production in 14 production fragmentation and 4 see also ASEAN entries; CLMV countries NAFTA (North American Free Trade Agreement) 3, 6, 76–7, 84, 118, 148, 317 imports from 15 outward FDI 262–3 Nakhon Rachasima 13 national borders, international trade and 15 NEARME see New East Asian Regional Management Effort (NEARME) NEG see new economic geography (NEG) Netherlands 194–5, 197–8, 200–1, 203, 208–10, 212, 214 New East Asian Regional Management Effort (NEARME) 72–8 institutions 72–4 longer-term goals 76–8 management priorities 74–6

Index 349 new economic geography (NEG) 143, 144–7, 161, 188 new spatial economics 188 new trade theory 185, 187–8 New Zealand ASEANþ4 FTAs and 296, 298, 305, 307 CER and 287 East Asia trade and 288 per capita GDP 2 see also ASEANþ6 New Zealand-Thailand FTA 249 newly industrialized economies (NIEs) 118, 150 noodle bowl syndrome 45–80, 79n current status 47–64 fragility/tensions, sources 64–72 graphic representation 48 New East Asian Regional Management Effort (NEARME) 72–8 policy implications 17 regionalism 49–64 Nordas, H. K. 121 Norman, V. D. 187 North Africa 296 North American Free Trade Area (NAFTA) 67 North American markets 39 see also Canada; United States of America (USA) Northeast Asia (NEA) 39, 97, 125, 256, 288 import contents in manufacturing 94–6 local content and 101–2, 104–5 political will 9–10 Olley, S. 230 Oosterhaven, J. 187, 190–1, 193–4, 198, 199–201, 201 ‘open architecture’ FTAs 39 ‘open regionalism’ 40, 57, 315 concept 284–8 openness, economic capital goods 210–12 consumption goods 208–10 intermediate goods/services 203–4, 204–8, 210–11

original equipment manufacturers (OEMs) 14 out-planned import 172 Pakes, A. 230 Pakistan 288 Pan-European Cumulation System (PECS) 74 parts characteristics 107–8 procurement 261 Paulmier, T. 15, 60 Pavcnik, N. 226, 230 Pearson correlation index 231 Penang Skill Development Centre (PSDC) 155 petroleum derivatives sector 59 Philippines 225–7 indigenous enterprises and 14 local content: automotive sector 104–5; electronics sector 101–2 manufacturing sector import content 94 outward FDI and 264 tariffs 33, 47, 69 see also ASEAN entries; factory economies; Southeast Asia Philippines-Singapore preferential tariff structure 47 Poland 12 policy implications 16–22 industrial 108–9 political will 9–10 Porter, M. E. 188–9, 198 Portugal 12, 77, 208 preference margins 60 preferential trade agreements (PTA) 6–7, 8, 47, 284 see also Common Effective Preferential Tariff (CEPT) production fragmentation 3–4, 36–7, 36, 120, 124, 164–5 prices, FTA impact 299, 300, 303, 307, 309, 310, 313 see also vertical production networks

350 Index productivity demand linkage effect 240–1 local plants 246, 247, 248 spillover effect 238–9 trade liberalization and 250 purchase shares 193 ‘quasi-regionalism’, tariff-cutting and 55 Rajan, R. 257 Rasmussen, P. N. 193 regional integration 148–58 drive forces 5–6 gap of development 156–8 industrial agglomeration 151–6, 152 new economic geography (NEG) and 144–7 transportation development 148–51, 149 regionalism 49–64 delayed 49, 56–61 domino effect 63–4, 70 preparation for 61–2 see also New East Asian Regional Management Effort (NEARME) ‘replenishment hub centres’ 5 resource reallocation 16, 232 ‘resource-seeking’ FDI 258–9 ‘reverse domino’ effects 70 Ricardian division of labour 4 rules of origin (RoO) 7–8, 33–6, 82, 83–5 ASEAN uniformity 48 cumulative 285 in East Asia 84–5 in EU 75 in FTAs 8, 83–4, 84 Russian Federation 296, 307 Saggi, K. 226, 231 sales shares 193 Sapir, A. 187 SARS 315 Scott, A. J. 151, 153 sector analysis 101–9 changes 89, 93, 95 classification 90–1

specialization (EU) 197–201, 199, 200 -specific changes in EU openness 204–8 self-sufficiency 205–6, 205, 206, 212 ‘shallow’ integration 9 shipping 120, 121 see also containerization Singapore 13, 225–6 hard disc drives (HDDs) and 3–4 ‘hollowing out’ of economy 50–1 import content in manufacturing 94, 96 inward FDI and 262 local content: automotive sector 105; electronics 101; manufacturing and 94 non-East Asia FTAs 39 outward FDI 263 per capita GDP 2 tariffs 33, 69 value added-based local content 98 vertical production and 165–6 see also ASEAN entries; headquarter (HQ) economies; Southeast Asia Singapore-Japan FTA 48, 84 Singapore-Philippines preferential tariff structure 47 small and medium enterprises (SMEs) 260 Sonis, M. 193, 196 South America 296, 298, 305, 307 South Asia 296, 298–9, 302, 305, 307 regional integration and 16 Southeast Asia 102, 125, 150, 158, 162, 316 South–South agreements 71, 288 ‘spaghetti bowl’ phenomenon 35 Spain 12, 208 spatial factors concentration (EU) 201–2 mobility of labour 188 structure, NEG and 147 ‘special and differential treatment’ (SDT) rules (EFTA) 77 specialization EU sectoral 197–201, 199, 200 export, intra-EU 198–201, 200 vertical 3

Index 351 spillovers and linkages 225–53 effects of 231–42, 238–9, 240–1 FTA impact 249 inter-country 194–5 policy implications 20–1 Thailand local–multinational 228–31 trade liberalization and 242–8, 249, 250 ‘Statistics of Foreign Direct Investment in ASEAN’ (ASEAN) 272 Stiglitz, J. E. 187 Storper, M. 120 Strassert, G. 194 structural changes automotive sector 106 electronics 103 manufacturing 96, 100 technological 96, 100, 102, 106 ‘supplier following assembler’ FDI 261–2, 261, 272 supply-side integration, disputes and 46 Suzuki, M. 164 Sweden 208 system fragility 70 emerging pressures 70–2 tensions, sources of 64–72 Taiwan 256 ASEANþ4 FTAs and 288, 296, 298, 305, 307, 310, 318 ‘hollowing out’ of economy 50–2 import content 101 Japanese FDI and 266 local content: automotive sector 104–5; electronics 101–2; manufacturing and 93–4; value added-based 98 outsourced production in 14 regional integration and 16 tariffs and 69 vertical production and 165 see also headquarter (HQ) economies tariffs applied EU, NAFTA 289 bound and applied 66–8, 68 cutting 55, 56

discrimination 15, 63–4, 70, 74 FTA impact and 327–8, 328 most favoured nations (MFN) 55, 56, 74 ‘overhang’ 66–8, 68–70 reduction 118 Technical Process (TP) 83, 84 technological factors changes 96–7, 102 structural changes: automotive sector 105, 106; electronics 102; manufacturing 96, 100 transportation advances 123 temperature-controlled logistic system 122 textile industry 132–3 Thailand 3–4 automotive industry 153–6 disc-drive assembly and 65–6 EPAs 249 FTAs see under free trade agreements (FTAs) indigenous enterprises and 14 industrial location and 52 Laos and 168–70, 172 local content in electronics sector 101–2 local–multinational linkages 228–31 manufacturing sector 155: import content 94 MNCs and 165–6 outward FDI and 264 per capita GDP 158, 160 spillovers/linkages 225–8, 245 tariffs 33, 69 TNCs and 266–7 see also ASEAN entries; factory economies; Southeast Asia Thailand-Australia FTA 249 Thailand-China FTA 249 Thailand-India FTA 249 Thailand-Japan EPA 249 Thailand-New Zealand FTA 249 Thailand-United States of America (USA) EPA 249 TNCs see transnational corporations (TNCs)

352 Index Tokyo Coil Engineering, case study 172–6, 175 Topalova, P. 226 Toyota, in Thailand 4 trade bilateral flows 71, 72 and FDI facilitation 37 structural changes 96, 102 see also liberalization trade theory classical 186, 198 new 185, 187–8 new international 284 Ricardian 186 training costs 272 transnational corporations (TNCs) 256, 258–60, 262, 265–8, 273, 276 transparency 74–6 transportation costs 6, 51, 170: agglomeration and 145; communication equipment 274; crossborder 15; defined 257; NEG and 147; outward FDI and 257–8, 266; trade theory and 188; vertical production networks 166–8 development of 125, 160 infrastructure 121–2, 161 inter-continental 150 inter-regional air cargo 150 machinery sector 59 technology 120, 123, 134 transportation, timeliness 118–38 air/containerization 129–30, 129 development of 122–4 Japanese trade and 124–38, 126, 127–8 policy implications 18 ‘triangle trade’ 50, 52 trilateral FTAs 35, 39 tsunami 315 Turkey 14 UNCTAD 256, 262, 264, 266, 268, 272 unilateralism accelerated 49, 56–61 rampant 49, 50–6

United Kingdom (UK) 193, 195, 200, 202–3, 208–9, 211–14 SMEs and 260 United States of America (USA) 3, 5, 11 ASEANþ4 FTAs and 284–5, 290, 296, 298–9, 307, 315, 317–18 Canada, trade with 15 local content 101–2: value addedbased 98, 100 local suppliers and 262, 267–8 MNEs 14 NAFTA and 77, 148 noodle bowl syndrome and 52, 57, 65, 67, 73 preference margins 60 SMEs and 260 transportation and import sensitivity 120–1 United States of America (USA)-Thailand EPA 249 Urata, S. 227 US-Korea FTA 64 utilization rates, AFTA 57–8, 58 value chain geographical expansion of 124 management 125 Value Content (VC) 83, 84 Venables, A. J. 2, 11–12, 118–19, 188, 258 vertical production networks and 166–7, 169 vertical production networks 162–82 industrialization through 164–6 Laos, border-related barriers 170–2: case study 172–9 Laos, fixed costs and 170–2, 171 Laos industrialization 168–70, 168, 169 policy implications 19 Thailand 245 transport costs 166–8 vertical specialization 3 Vietnam 226 AFTA compliance and 286 customs clearance and 9 FDI and 11 Laos and 168–9, 170 outsourced production in 14

Index 353 production fragmentation and 4 tariffs and 69 see also ASEAN entries; CLMV countries; factory economies; Southeast Asia Vietnam-Japan FTA 48 Vietnam–Laos Joint Venture 277 Viner, J. 14, 284 Viner–Meade partial equilibrium theory 187 Walmsley, T. L. 324, 326 Wei, S. 284–5 welfare, FTA impact 295 Whalley, J. 48 Wincoop, E. van 15, 150, 166–7, 171, 257

world prices 303, 313 world trade 146 World Trade Organization (WTO) 1, 45, 73–8 passim, 298, 315, 317 discipline 66–8, 68–70, 217 integration process 11 trade liberalization and 9, 16 world’s factory 3–5 WTO see World Trade Organization (WTO) Wyplosz, C. 12, 143–4, 160 Yokota, K. 227 ‘zero geography’ models 218n Zhu Rongji, Chinese Premier 49, 63