Industrial Clusters in Asia: Analyses of Their Competition and Cooperation

Industrial Clusters in Asia Analyses of Their Competition and Cooperation Edited by Akifumi Kuchiki and Masatsugu Tsuji...

Author: Kuchiki | Akifumi Tsuji | Masatsugu

30 downloads 1197 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Industrial Clusters in Asia Analyses of Their Competition and Cooperation

Edited by Akifumi Kuchiki and Masatsugu Tsuji

Industrial Clusters in Asia

This page intentionally left blank

Industrial Clusters in Asia Analyses of Their Competition and Cooperation Edited by Akifumi Kuchiki and Masatsugu Tsuji

© Institute of Developing Economies (IDE),JETRO 2005 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph 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, 90 Tottenham Court Road, London W1T 4LP. 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 2005 by PALGRAVE MACMILLAN Houndmills, Basingstoke, Hampshire RG21 6XS and 175 Fifth Avenue, New York, N. Y. 10010 Companies and representatives throughout the world PALGRAVE MACMILLAN is the global academic imprint of the Palgrave Macmillan division of St. Martin’s Press, LLC and of Palgrave Macmillan Ltd. Macmillan® is a registered trademark in the United States, United Kingdom and other countries. Palgrave is a registered trademark in the European Union and other countries. ISBN-13: 978–1–4039–4350–7 ISBN-10: 1–4039–4350–8 This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data Industrial clusters in Asia : analyses of their competition and cooperation / edited by Akifumi Kuchiki and Masatsugu Tsuji. p. cm. “Project … coordinated and ﬁnancially supported by the Institute of Developing Economies (IDE), Japan External Trade Organization (JETRO)”–Pref. Includes bibliographical references and index. ISBN 1–4039–4350–8 (cloth) 1. Industries–Asia. 2. Competition–Asia. I. Kuchiki, Akifumi. II. Tsuji, Masatsugu. III. Ajia Keizai kenkyujo (Japan) HC415.D5I53 2005 338.6⬘042⬘095–dc22 2005040231 10 14

9 13

8 12

7 11

6 10

5 09

4 08

3 07

2 06

1 05

Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham and Eastbourne

Contents

List of Tables

vi

List of Figures

ix

Preface

xi

List of Contributors 1

Introduction Akifumi Kuchiki and Masatsugu Tsuji

Part I 2

xii

Industrial Clusters in Vietnam

Competition and Cooperation: Vietnam Adam McCarty, Richard Record and James Riedel

Part II

1

27

Industrial Clusters in China

3

Industrial Clusters in the Tianjin Area Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu

111

4

A Flowchart Approach Akifumi Kuchiki

169

5

Chinese Automotive and Parts Industries Masatsugu Tsuji and Wu Quan

200

Part III

Industrial Clusters and IT Industries

6

Can Malaysia’s Growth Be Sustainable? Yumiko Okamoto

225

7

Bangalore’s Software Cluster Aya Okada

244

8

The Local Economy: Lessons from Japan Rika Nakagawa

278

9

Japan: Remedies to Activate Local Cities Takuo Imagawa

299

Index

319

v

List of Tables 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11 2.12 2.13 2.14 3.1

3.2 3.3 3.4 3.5 3.6 3.7 4.1 4.2 4.3 4.4 4.5 4.6

Vietnam total and manufacturing GDP by ownership: 1995 and 2001 Enterprises in Vietnam: a snapshot in 2003 Vietnamese export composition: 1995–2002 (percentages) Vietnam’s trade-related infrastructure International relative competitiveness in Asia, 2001 Comparative educational indicators, selected economies, 1990 Vietnamese licensed FDI projects January–July 2003: top ﬁve provinces Vietnamese gross exports January–July 2003: top ﬁve provinces Examples of some world-class clusters Vietnamese enterprise dataset used for GIS mapping and analysis: labour and turnover Vietnamese enterprise dataset used for GIS mapping and analysis: ownership type List of GIS maps produced of Vietnam Cluster index of Vietnam – all 12 industries Perceived role of clusters, selected Western and East Asian economies Comparison of the development characteristics of the electronic information industry in six Chinese provinces and cities, 2001 Comparison of the degree of agglomeration in the electronic information industry in six provinces and cities in 2001 Electronic industrial clusters in Tianjin Output and output value of main products in the green battery base in Tianjin Distribution of Tianjin’s bicycle enterprises Changes in the percentage of the output of bicycles in Tianjin compared with that in all China Changes of annual export volume of Tianjin’s bicycles Tianjin’s industrial zones Infrastructure of Tianjin Economic Development Area Tax system of Tianjin Economic Development Area Total output of the Chinese automobile industry, 1993 Total output of the Chinese automobile industry, 2001 Keiretsu companies of Toyota in Tianjin vi

32 33 36 37 38 41 43 43 54 61 62 63 91 104 116

117 120 135 155 157 158 181 183 184 190 191 192

List of Tables vii

4.7 4.8 5.1 5.2 5.3 5.4 5.5 5.6 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 7.1 7.2 7.3 7.4 7.5 7.6 7.7 8.1 8.2 8.3 8.4 8.5 8.6 8.7

Japanese companies related to Toyota in Tianjin Summary of industrial cluster policy, Tianjin and northern Vietnam Major Chinese automobile assemblers Number of automobile assemblers and parts suppliers in concentrated regions of China Eight major ﬁrms in the Chinese automotive sector, 1995 Concentration ratios in the automotive industry: four-country comparison Localization ratio of Santana Major parts manufacturers in Shanghai Regional shares of foreign and domestic investment in Malaysia between 1993 and 2002 Market shares of Malaysia in major developed countries RCA index of Malaysia Spearman’s rank correlation coefﬁcients of the RCA indexes between Malaysia and some of its major trading partners Spearman’s rank correlation coefﬁcients of the market shares between Malaysia and China in electronics products Intra-industry trade (IIT) indexes between Malaysia and China Economic impacts of the MSC in Malaysia Occupational grouping in Malaysia Average annual software labor cost comparisons in selected countries, 1999 Indian software ﬁrms’ global market penetration Locational distribution of headquarters of India’s top 700 IT software and services ﬁrms, 2000 Growth of employment and revenue per employee in the Indian software industry IT-related jobs in demand in India Higher education institutions offering courses related to computers, all India and Karnataka Growth in intake in IT-related courses in Karnataka Number of manufacturing factories in Japan, 1985–2000 Number of industrial zones in China Per capita income in each Japanese prefecture Growth rates of per capita income in each Japanese prefecture Details of the Japanese industrial cluster plan 2001 Projects of industrial cluster planning in each Japanese region Details of support for the ICT industrial cluster in Okinawa

193 197 202 203 206 207 211 212 228 231 232 235 235 236 238 240 250 251 253 258 259 263 265 280 282 284 286 288 289 292

viii List of Tables

9.1 9.2 9.3

9.4 9.5

IT industry accumulation in Japan: summary statistics (city data) Estimation results of IT industry accumulation in Japan: basic model (accumulation factor analysis) Estimation results of IT industry accumulation in Japan: inﬂuences of education facilities (education and proximity analysis) Industrial classiﬁcations of the IT industry in Japan – 1: accumulation factor analysis Industrial classiﬁcations of the IT industry in Japan – 2: educational facilities and proximity analysis

308 309 310

311 312

List of Figures 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11 2.12 2.13 2.14 2.15 2.16 2.17 2.18 2.19 2.20 2.21 2.22 2.23 2.24 2.25 2.26 2.27 2.28 3.1 3.2 3.3 3.4 4.1 4.2 4.3

Opinions as to whether intellectual property is well protected, selected economies Hotelling model: competition along a line Porter’s ‘diamond of advantage’ Dot density map of Vietnam – all 12 industries Dot density map of Vietnam – automobiles Dot density map of Vietnam – cement Dot density map of Vietnam – electronics Dot density map of Vietnam – garments Dot density map of Vietnam – motorcycles Dot density map of Vietnam – paper Dot density map of Vietnam – rice Dot density map of Vietnam – seafood Dot density map of Vietnam – software Dot density map of Vietnam – steel Dot density map of Vietnam – sugar Dot density map of Vietnam – textiles Vietnam northern region commune-shading map – garments Vietnam northern region commune-shading map – paper Vietnam northern region commune-shading map – seafood Vietnam northern region commune-shading map – steel Vietnam northern region commune-shading map – textiles Vietnam southern region commune-shading map – garments Vietnam southern region commune-shading map – seafood Vietnam southern region commune-shading map – steel Vietnam southern region commune-shading map – textiles Seafood clusters in the Mekong Delta Hourly labour costs, selected developing countries Vietnam provincial codes Hierarchical structure of regional competitive power in China Local production system of bicycle enterprises in Wangqingtuo Town, Tianjin City Tianjin annual bicycle production as a percentage of that of all China Competitive advantage of enterprise clusters Industrial cluster formed by an anchor ﬁrm Value chain from design planning to marketing Efﬁciency in production only ix

40 46 50 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 100 106 114 152 158 159 171 173 173

x List of Figures

4.4 4.5 4.6 4.7 4.8 4.9 4.10 4.11 4.12 4.13 5.1 6.1 6.2 6.3 6.4 7.1 7.2 8.1

Optimization for customer’s satisfaction Measures of value chain management: an example Core competence (selection and concentration) Cost structure Average cost Average cost, brand marketing and R&D Toyota’s value chain network in China Toyota’s cluster in Tianjin Consulting services procedures in Tianjin Cross-docking at work Production structure of the Shanghai automobile industry Number of Japanese ﬁrms established in ASEAN Countries at the end of 2000 Foreign investment in the electronics industry of Malaysia on the application basis Flow of FDI into China and ASEAN Numbers of world-class companies in the MSC by nationality, 2003 Growth of software exports from Bangalore Structures of the software and services cluster in Bangalore Total number of ICT companies in Okinawa, May 2003

173 173 173 175 176 177 179 182 186 189 216 227 230 233 239 255 257 291

Preface This volume is the product of a research project entitled ‘Industrial Clusters in Asia: Analyses of their Competition and Cooperation’ which was coordinated and ﬁnancially supported by the Institute of Developing Economies (IDE), Japan External Trade Organization (JETRO). Industrial clusters have attracted considerable attention in recent years but little literature exists on comprehensive analyses of Asian clusters. We have paid attention to industrial cluster polices in Asia as growth strategy. Firms in China are supplying many manufactured goods and clusters in China are bases for the manufacturing industries. Firms in East Asia including ﬁrms in Japan are competing with ﬁrms in China. Therefore we analyzed industrial clusters in Asia and tried to ﬁnd out a prototype model of industrial cluster policies in Asia. We are grateful to all the members and counterparts who contributed to our project. We would also like to thank all the individuals and organizations that accepted our visits and shared with us their precious time and information. We also wish to express our sincere thanks to Mr Kazuo Katsume, Mr Yuichi Takayama, Professor Susumu Sanbonmatsu, Ms Momoko Kawakami and Mr Ko Yukawa for their lectures for our project. Any opinions expressed in this book are those of the contributors and not their organizations. We hope this publication will contribute to the academic circle and toward further understanding of industrial clusters in Asia. AKIFUMI KUCHIKI Executive Vice-President, JETRO

xi

List of Contributors Zhang Donggang

Nankai Institute of Economics

Takuo Imagawa Institute for Information and Communication Policy, Ministry of Public Management, Home Affairs, Posts and Telecommunications Akifumi Kuchiki

Institute of Developing Economies (IDE-JETRO)

Adam McCarty

Mekong Economics

Rika Nakagawa

Institute of Developing Economies (IDE-JETRO)

Aya Okada University

Graduate School of International Development, Nagoya

Yumiko Okamoto Wu Quan

Doshisha University

Graduate School of Economics, Osaka University

Richard Record James Riedel

Mekong Economics

Johns Hopkins University

Luo Ruoyu

Doctoral student, Nankai Institute of Economics

Xie Siquan

Nankai Institute of Economics

Masatsugu Tsuji University Bai Xuejie

Osaka School of International Public Policy, Osaka

College of Economic and Social Development

xii

1 Introduction Masatsugu Tsuji and Akifumi Kuchiki

Industrial clusters in Asia: analyses of their competition and cooperation Why ﬁrms agglomerate has been a question in economics ever since the time of Marshall and Hotelling. In the 1990s, the success of Silicon Valley in California and Route 128 in the Boston area as IT clusters, among others, brought new insights to this ﬁeld. Since then, owing to the development of new theoretical models such as those of Krugman and of Porter the notion of cluster has become the most important in terms of competitiveness and strategy for economic development. In order to set up policy measures to foster clusters, attempts have been made to identify the strategic factors involved, for example Porter’s (1) demand condition, (2) competitive condition, (3) factor condition and (4) support condition. Cluster projects have been implemented in developing as well as developed countries, and it is said that the total number to date worldwide is more than 500. Traditional or old industrial clusters such as Nishi-Mikawa (Eastern Aichi prefecture) and Yawata City in Japan, Manchester in the UK, or Silicon Valley in California, although the last belongs to the new IT industry, were formed through the competitive process. Anchor ﬁrms decided their location through optimizing behaviour by comparing economies of scale and transaction costs, and others agglomerated close to them. On the other hand, recent economic development in East Asia has been initiated and promoted mainly by foreign direct investment (FDI). Here, foreign ﬁrms selected their locations by considering the availability of not only cheap unskilled labour but also skilled labour and other economic resources. As a result, East Asia was referred to as the ‘growth center in the world’ in the early 1990s, then the ‘factory of the world’ in late 1990s. Behind this success story, we can ﬁnd government support to attract ﬁrms to the regions. Good examples are industrial parks and export-processing zones in China and Vietnam. In China there are hundreds of these zones in Zhejiang and Jiangsu provinces in the Yangtze River Delta, for example, 1

2 Introduction

established by all levels of government, local as well as central. They offer generous subsidies and tax exemption for proﬁts, and provide infrastructure such as electricity, telecommunications, highways and water supply. In this sense, regions in East Asia have been competing with each other to attract not only foreign but also domestic ﬁrms. All understand that without agglomerating there is no hope for industrial development. On the other hand, ﬁrms established in one country have been shifting to others, and this implies that the network of outputs, inputs and other business activities such as R&D expands to other areas. Such activities are a natural outgrowth of globalization as the new areas concentrate and specialize while at the same time their interdependence increases more and more. Although, in reality, the competition among East Asian regions is emphasized, tighter interdependence should be noticed. The aim of this book is to analyse competition and collaboration among East Asian industrial clusters, and to envision their future. In Industrial Agglomeration: Facts and Lessons for Developing Countries (2003), we compared agglomerations mainly in Japan, the USA, Italy and some East Asian countries. This book, on the other hand, focuses on East Asia, which has been attracting FDI and developing as a centre of industrial agglomeration, with the result that the production structure in the world has been transforming dynamically. This book aims to analyse this world trend and to provide the strategy framework that is required not only for Japanese local governments but also for ﬁrms to survive global competition. Brieﬂy, the conclusions of this volume are as follows. The success of these industrial agglomerations depends on (1) infrastructure (highways, ports, electricity supply, and so on), (2) institutional framework (legal systems, participatory actors, coordination among actors, and so on), and (3) government support in terms of laws, taxation and ﬁnance. As shown in this volume, the role of government is essential. In developing countries, the market mechanism to foster clusters does not function well owing to regulations, bureaucracy and the existing legal framework. Moreover, there is no proper market for capital and human resources. In such a case, the government must complement market mechanisms and prepare the above three conditions. The cases of northern Vietnam and China provide good examples of collaboration in these three areas working well. This book is contributed by the following three teams: the US Team which analyses the case of Vietnam by recent theory, the China Team which examines the case of Tianjin, China, and the Japan Team which studies the case of Japan and other Asian economies. The text is divided into parts as follows: Part I provides a case-study of clusters in Vietnam, utilizing geographic information systems approach in particular. Part II presents a case-study of the Tianjin cluster by the Chinese Team, and attempts to make clear how clusters in Tianjin with various industries have been formed. Part II also includes two chapters by the Japan Team: Chapter 3

Masatsugu Tsuji and Akifumi Kuchiki 3

analyses the Tianjin cluster by applying a ﬂowchart analysis to the Toyota production plant located there, and Chapter 4 the Chinese automobile industry, which examines how Chinese automobile assemblers and parts suppliers choose their locations. Part III consists of four chapters by the Japan Team: Chapter 6 focuses on the Malaysian electronics industry and analyses its competitiveness; Chapter 7 looks at Bangalore, India, and examines its local labour market as a basis for competitiveness; Chapter 8 introduces policy measures taken by the Okinawa prefectural government in promoting ICT cluster, and highlights the importance of policy coordination with central government; and ﬁnally Chapter 9 examines how ITrelated ﬁrms choose their locations and suggests a strategy for a local region to vitalize its IT industry.

PART I: INDUSTRIAL CLUSTERS IN VIETNAM Chapter 2: Competition and Cooperation: Vietnam In both developing and developed countries, certain industries tend to agglomerate in one area in order to take advantage of knowledge creation and spillovers, access to inputs, human and social capital, lower costs and lower barriers to market entry in clusters. For example, in the USA, the information technology industry is concentrated in Silicon Valley in California, while the international ﬁnancial market is based mainly in London and the semiconductor industry can be found in Hsinchu Park in Taiwan. Vietnam is no exception to clustering. As we note on p. 56, The Old Quarter in Hanoi is famous for its specialist shops and microhousehold enterprises. Many of today’s street specializations date back to the establishment of Hanoi’s merchant guilds in the thirteenth century. Lan Ong Street is famous for herbs and traditional medicine, Lo Ren Street for blacksmiths and metalworking, Hang Gai Street for silk, Hang Bac Street for jewelry, and Hang Dau and Cau Go Streets for leatherwear and shoes. Chapter 2 attempts to answer questions as to why ﬁrms choose some locations and not others for their factories and ofﬁces. Studies have identiﬁed four broad categories for industrial agglomeration, described on p. 102: 1. clusters where locally specialized items are produced or local product districts; 2. clusters where a large core ﬁrm has many subcontracting or parts makers surrounding it or the so-called industrial castle town; 3. clusters in large cities where lots of basic production processes are treated or urban processing clusters; and, 4. government-led industrial parks and estates often seen in developing countries.

4 Introduction

The chapter focuses on the fourth type of cluster – industrial zones (IZs), industrial parks (IPs) and export-processing zones (EPZs). The zones are considered a proxy for public goods in a developed country, that is in a developed country, entrepreneurs would expect to have quality infrastructure and institutions, such as electricity, roads, telecommunications, water supply and sewerage, and so on. Our study revealed that the agglomeration of foreign-invested ﬁrms in Vietnam has been in large part owing to the infrastructure and administrative advantages available in the IZs. Besides the development of government-led industrial parks, other forms of clustering are also emerging in Vietnam. Therefore, the chapter builds on the previous study, but pays particular attention to the ﬁrst three categories of industrial clustering by using a geographic information systems (GIS) approach to locational analysis. Chapter 2 begins with an overview of the industrial competitiveness of Vietnam, before surveying the economic theories that explain industrial clustering and then and reviewing the international empirical evidence of the theories. The chapter then looks at some of the recent work on cluster engagement in Vietnam, after which GIS analysis is applied to analyse 12 key industrial sectors of the Vietnamese economy. The chapter then reviews the roles of both business associations and also IZs, IPs and EPZs as potential ‘cluster facilitator’ mechanisms. It concludes with industrial policy recommendations. Vietnamese competitiveness and the environment for business Without a doubt, Vietnam has been an economic growth success story for a decade now. The transition to a market economy has occurred comparatively smoothly and with relatively strong growth and economic stability throughout most of the 1990s, despite institutional weaknesses and a policy bias against the private sector. However, as we note on p. 31, Vietnam remains a low-income country and the public sector still dominates the industrial economy of Vietnam, contributing some 50 per cent of GDP. The state’s share of output in the manufacturing sector has fallen somewhat over recent years from 59 per cent in 1995 to 51 per cent in 2001. Private domestic ﬁrms, by contrast, accounted for only about 9 per cent of manufacturing value added (GDP). Legal reforms in recent years have, however, greatly encouraged private sector development, and in early 2002 the status of the private sector was enhanced when a plenum of the ruling party resolved that the private sector is ‘an integral part of the national economy’. Stronger ofﬁcial endorsement of the private sector embodied in recent policy changes has helped build investor conﬁdence. Biases in favour of the public sector remain, but Vietnam is now beginning a period of rapid growth led by domestic businesses, rather than by FDI and commodity exports as occurred during the 1990s. The private sector’s contribution to national investment has risen from 21 per cent in

Masatsugu Tsuji and Akifumi Kuchiki 5

2000 to 29 per cent in 2002. Through international trade, Vietnam has begun to capitalize on its principal area of industrial comparative advantage – low labour costs. Currently, Vietnam’s garment industry is the second most important market after the EU, and the country is now the secondlargest shrimp exporter to Japan, after Indonesia. However, Vietnam’s competitive potential is weakened by state monopolies (or cartels), which have kept costs high in areas such as air transport, coastal shipping, railways, power and telecommunications. As we note, ‘Vietnam is one of the most expensive countries in the world for both incoming and outgoing international telephone calls’ (p. 36). Such infrastructure problems largely restrict FDI and new business activity to specially built investment zones. In the absence of private provision, government services tend to be driven by what ofﬁcials think is important to supply, rather than driven by demand. Theoretical literature review Until quite recently, economists have paid only limited attention to the location of economic activity and to the choices ﬁrms and households make to decide where to produce or consume. Early works by Alfred Marshall and Harold Hotelling laid some basic premises for looking at the role of geography in economic decision-making. Marshall (1920) did some speculating on the subject with his observation that: ‘Industries tend to cluster in distinct geographic districts, with individual cities specializing in production of narrowly related sets of goods.’ In more modern parlance, as summerized on p. 47, Marshall argued that industrial districts arise because of: • knowledge spillovers (‘the mysteries of the trade become no mysteries; but are as it were in the air’); • the advantages of thick markets for specialized skills; and • the backward and forward linkages associated with large local markets. Hotelling (1929) expanded on this idea by looking at location as a question of ﬁnding a stable equilibrium based on the location of the market under highly constrictive assumptions. The work of Paul Krugman, Masahisa Fujita and Anthony Venables began a renaissance in geographically based economic analysis, which they dubbed ‘the new economic geography’. The foundation of this analysis rests largely on monopolistic competition modelling with an emphasis on the interaction of increasing returns, transport costs and the movement of productive factors to understand the latent forces of economic agglomeration. First, economies of scale and transportation costs imply that ﬁrms prefer to locate close to a large market. Second, workers are also consumers, and would thus prefer to live close to a large market.

6 Introduction

Michael Porter of the Harvard School of Business developed an alterative approach to economic geography. Porter deﬁnes clusters as: geographic concentrations of interconnected companies and institutions in a particular ﬁeld … They include suppliers of specialized inputs, components, machinery and services. Clusters also often extend downstream to channels and customers and laterally to manufacturers of complementary products … ﬁnally many clusters include governmental and other institutions. This school of thought also stresses that nations succeed not in isolated industries, but in clusters of industries connected through vertical and horizontal relationships. This system has an implicit concept based on a ‘critical mass’ of speciﬁc inﬂuences that will improve knowledge, share resources and stimulate creativity, innovation and entrepreneurship. ‘Porter-style’ clusters in developing countries tend to be shallow and to rely primarily on foreign components, services and technology. Industrial clusters in Vietnam Agglomeration can occur at many different levels, serve large or small areas, and take place in a variety of industries. This study looks at 4400 enterprises engaged in a wide range of sectors – automobiles, cement, electronics, garments, motorcycles, paper, rice, seafood, software, steel, sugar and textiles – and their potential for clustering. Within these 12 sectors, 3 (garments, seafood and textiles) have been chosen for further analysis. Using a basic deﬁnition of clustering based on ﬁrm density, our sample identiﬁes the garment sector to have the greatest tendency to cluster, followed by textiles, rice, seafood and paper. Our results show that there are some signs of industrial agglomeration in Vietnam; however, these results must be taken into perspective with the transitional nature of the Vietnamese economy. The ability of a particular cluster to form depends primarily on an environment conducive for private enterprises, that is institutional frameworks. Economic reform has been the principal driver of private-sector growth, and future efforts should be focussed on further implementation of the 2000 Enterprise Law. If clusters are dynamic centres of growth, then understanding the development, management and stability of clusters is a matter of great signiﬁcance for the future of Vietnam’s economy. Especially Japan, Korea and other leading East Asian economies can provide constructive examples for Vietnam. These countries demonstrate that a strategy of preserving core cultural values while being open to new ideas can be extremely successful in a knowledge-driven global economy. State-funded technological catch-up has decreased in importance, while private enterprises have become the international drivers of technological development.

Masatsugu Tsuji and Akifumi Kuchiki 7

The reforms necessary for the long-term and stable growth of the Vietnamese economy were noted at the Vietnam Business Forum (9 December 2002) and are listed on p. 100 as: • introduce greater competition to reduce costs closer to the regional levels in sectors such as telecommunications, power and shipping; • ease restrictions on technology transfers; • make business policies and regulations more transparent in their formulation as well as using consistent enforcement to enhance predictability for investors; • broaden the currently narrow tax base that puts a disproportionate burden on honest taxpayers and slows the growth in employment; • accelerate deregulation and implement the Enterprise Law aggressively outside the main urban areas; • correct misconceptions on the function of the private sector and enhance its role; and stress the importance of encouraging corporate social responsibility. While Vietnam has taken steps in the right direction, there are still substantial institutional changes needed for continued growth. The lesson for Vietnam is that the government is no longer the principal actor in driving economic and technological development. Its role has changed from setting the stage and creating a conducive enabling environment to freely allowing entrepreneurship and innovation.

PART II: INDUSTRIAL CLUSTERS IN CHINA Chapter 3: Industrial Clusters in the Tianjin Area Development and spread of industrial clusters and studying-type regions in Tianjin The strong power of industrial clusters and studying-type regions has been displayed in global competition. It is the best choice for supporting development of enterprises and strengthening competitive power of enterprises. As noted on p. 112, a so-called industrial cluster comprises both the industries that gather within a certain region and their supporting system. Porter generalized this supporting system as having the following four aspects: demand condition, competition condition, factor condition and support condition. Sustained competitive advantage can be produced only by mutual support and coordinated development of these four aspects. Industrial clusters include upstream industrial sectors that provide raw materials and means of production for core industrial sectors and those industrial sectors each of whose production technology affects each others (p. 112). To build mature industrial clusters, good regional economic environment is needed.

8 Introduction

Good regional economic environments should include at least the following six factors. 1. Rich technological resources and its distribution media. Universities and private and government research institutes can provide industrial sectors with the research achievements they need to improve their competitive position, through fast and efﬁcient distribution media. 2. Human resources with high adaptability and organizations for development and training. Economic development requires that regions provide lots of professional personnel and a skilled contingent of workers for enterprises. Also, ‘through basic occupational education, higher education and continuous in-service education, regions should give human resources the knowledge and skill that allow them to adapt to competition and changes’ (p. 113). 3. Abundant fund sources. From fo. 192: The competitive power of regional industries cannot do without the strong support of ﬁnancial organizations. The functions of these ﬁnancial organizations include raising enough funds for economic development and establishing investment mechanisms that can fully reﬂect guidance for market competitive power. 4. Favourable macroeconomic environment with efﬁcient management. As shown on p. 113 this kind of management level is embodied in two aspects: one is that macroeconomic regulations can guide enterprises effectively and at the same time do not destroy the independence and competitive relation of enterprises; the other is that the tax system is beneﬁcial to competition, and at the same time meets the needs of economic development and the public. 5. Well-developed infrastructure for software and hardware. The efﬁciency of regional economic activities is, to a large extent, affected by the condition of infrastructure. 6. Fair-scale groups of customers with discriminatory ability. ‘Views of these customers and their requirements for products can reﬂect the basic trend of market changes and should be regarded as the basis for product innovation by enterprises’ (p. 114). There are many mature industrial clusters, such as the German chemical industry, the Italian shoe industry, the Swiss pharmaceutical industry, the semiconductor industry in the USA and Japan, the mobile phone industry in Scandinavia and so on. In Latin America, South Asia and a few African countries, industrial clusters, such as Leon shoe manufacturing in Mexico,

Masatsugu Tsuji and Akifumi Kuchiki 9

Santa Catarina ceramic tiles in Brazil and the Nnewi automobile components and parts cluster of Nigeria have attracted people’s attention and developed to some extent. In Guangdong, Jiangsu, Zhejiang and other provinces in China, some traditional industries have also shown the characteristics of agglomeration. These industries include the garment industry, shoe manufacturing industry, hosiery industry and toy industry. Tianjin is demonstrating the fruits of its decades-long development by producing large-scale foreign-funded enterprises in the complete processing, manufacturing, and heavy industry areas. In the process of industrial agglomeration, high-tech industrial clusters and traditional industrial clusters have their respective characteristics in shaping process, features and development trends. In Tianjin, industrial clusters of high-tech industries such as electronic information and automobiles and traditional industries such as bicycles and garments have appeared in an embryonic form. The network system of study between industrial clusters has also appeared. Affected and encouraged by developed industrial clusters around the world, in the interaction and competition of domestic industrial clusters, Tianjin hopes to form its own industrial clusters with advantages and studying-type regions of high dynamic ability. Industrial clusters in the Tianjin Area In the process of industrial agglomeration, high-tech industrial clusters and traditional industrial clusters have their respective characteristics of formation processes, features and development trends. The success of the hightech industries depends on: (1) infrastructure (highways, ports, electricity supply and so on); (2) institutional frameworks (legal systems, participatory actors, coordination among actors and so on); and (3) government support (or foreign assistance) in terms of laws, taxation and ﬁnance. As shown in Chapter 2, the case of northern Vietnam provides good examples where collaboration in these three areas worked well. There exist the following industrial clusters in Asia: southern China, the Chu-Chiang River Delta in China, Johor, Selangor and Penang in Malaysia, Hanoi and Ho Chi Minh in Vietnam, Bangalore in India, Zhong Guan Cun in China, Hsinchu Science Park in Taiwan, and Leam Chabang in Thailand. As new industries, here we chose the electronic information industry in Tianjin and the battery industry with its special feature of green environmental protection. For traditional industries, Chapter 3 chose the automobile manufacturing industry in Tianjin and the bicycle manufacturing industry. The new motive force for the development of the electronic information industry in Tianjin In recent years, the electronic information industry in Tianjin has developed very quickly and taken on a healthy look. The rapid formation and

10 Introduction

development of the electronic information industry in Tianjin has the following obvious characteristics. 1. Exogenous industrial clusters. Electronic information clusters’ exogeneity is evident from their dependence on foreign funding sources. The increasingly large scale of foreign funds in the electronic information industry, and the ratio of foreign funds in the electronic information industry to gross funds. 2. Industrial clusters have taken on the structural characteristics of a ‘single core’. The average scale intensity of Tianjin’s enterprises is comparatively high; enterprises are of relatively large scale, and ‘single-core enterprises of industrial clusters are large-scale transnational corporations’ (p. 118). 3. Rooting of industrial clusters and the foundation of local production network. Industrial clusters are primarily rooted in local socio-cultural characteristics and institutional advantages. These play an important role in the further development of high-tech industrial clusters. 4. Industrial clusters with product chain. ‘Within electronic information industrial clusters in Tianjin, economic bodies connect with each other to form an inner value chain’ (p. 119). It is a typical mode of product chain cluster. In this chain, three kinds of enterprises share work and cooperate with each other to form a product chain cluster with comparative power. Some problems still exist in Tianjin’s electronic information industry: the industry has comparatively strong dependence; institutional factors restrict the further development of industrial clusters; the local production network has not embedded deeply; the single-core structure of industrial clusters is fragile and unstable; and there is no knowledge connection which is favourable for innovation. Regarding these characteristics and problems, Chapter 3 proposes a solution to the development of the electronic information industry in Tianjin in order to improve its competitive advantage and bring the clustering effect into full play. The solution is as follows. (1) Tianjin’s electronic information industrial clusters should combine industrial advantage, policy advantage, low cost advantage and location advantage. (2) The structure of industrial clusters should develop from ‘single-core’ to ‘multiple-core’ to form a competitive pattern within the region and to transfer and improve electronic information industrial technology. (3) Local industrial clusters with self-strengthening mechanisms should be cultivated. Product chain clusters should be turned into innovation chain clusters so as to change regional comparative advantages into competitive advantages. (4) The process of reform for state-owned enterprises in the electronic information

Masatsugu Tsuji and Akifumi Kuchiki 11

industry and to further complete institutional network within industrial clusters should be speeded up. (5) The electronic information industry should develop toward specialization, interconnection, internationalization and clustering through the combination of technology, institute and market. (6) Investment should be promoted by favourable policy, sound infrastructure, convenient trade environment and improving the ecological environment in regional clusters. The rise of the green battery industry in Tianjin The Tianjin municipal government has listed new energy (taking the green battery as the dominant factor) in three pillar high-tech industries, so called pillar industries. Tainjin New-Tech Industrial Park also looks on new energy as a key ﬁeld of development. These measures have led to the rise of the green battery industry. Tianjin has distinct advantages in developing this industry as follows: (1) The green battery industry in Tianjin has strong R&D facilities and a high production capacity. (2) Types of products are complete and the production chain has appeared in an embryonic form. (3) The industry has a vast range of prospects and great potential. Tianjin has been successful in opening up the market at home and abroad. After ten years’ development, the green battery industry has achieved a certain scale. However, as a growing industry, it still has some problems that demand prompt solutions. Through analysis of the basic situation and industrial structure of the green battery industry, Chapter 3 makes suggestions for guidance. Development environment, characteristics and direction of the automobile industrial clusters in Tianjin On p. 141 we summarize the situation of Tianjin’s automobile industry: The automobile industry in Tianjin is an important part of the national industry. Especially in the twenty years since reform and opening-up, the automobile industry in Tianjin has both risen and fallen. It is now facing the opportunity of starting a new undertaking. After decades of development, automobile industrial clusters have appeared in an embryonic form with several important development characteristics. These characteristics include: (1) The accumulated development basis of the industry provides an advantage in terms of human resources. (2) The unique product positioning under the national industrial policy has gained some pre-emptive advantages. (3) The industrial agglomeration of which Toyota is seen as the centre exists together with the original industrial agglomeration. (4) Tianjin’s automobile industrial clusters participate in the formation of a domestic oligarchic market structure through its alliance with China’s FAW Group Corporation. (5) Because the industry is in the key stage of change from comparative advantage to competitive advantage, the competitive advantage of the industry is not outstanding. (6) The competitive power of

12 Introduction

enterprises manufacturing automobile components is in decline on the whole, but some enterprises have suddenly come to the fore. (7) The localization system of the R&D of products has yet to be formed. (8) The means by which government affects industrial development lies in facing the requirements of change. Beginning from the background of the automobile industry at home and abroad, Chapter 3 analyses the industry’s development characteristics; next it raises the construction direction of the automobile industry in Tianjin in order to strengthen and/or sustain competitive advantage and improve the clustering effect of this industry. The construction direction includes the following aspects: (1) constructing two interrelated systems of the production chain; (2) medium-scale and small-scale enterprises with quality management and innovative ability being the main factor in maintaining the competitiveness of the value chain; (3) promoting the change of the value chain from comparative advantage to competitive advantage and even to dynamic competitive advantage; (4) government plays a positive role in cultivating industrial dynamic competitive advantage. Analysis of the formation and development of bicycle enterprise clusters in Tianjin Tianjin is the birthplace of China’s bicycle industry. Here the industry has experienced both rise and fall. In this process, unique bicycle enterprise clusters have been formed gradually, the clustering effect playing a signiﬁcant role in encouraging industrial development. Taking the formation characteristics, factors and development mechanism of bicycle enterprise clusters in Wangqingtuo as the starting-point, Chapter 3 analyses the development, structural features and competitive advantage of bicycle enterprise clusters and the structure of bicycle industrial clusters in Tianjin. The structural characteristics of Tianjin’s bicycle industry are summarized on p. 156: first, the vertically integrated, rigid mode of production in stateowned enterprises has been adaptable neither to market demand nor to technical improvement. These enterprise clusters are, rather, the substitute for such a mode of production. Moreover, they are located in suburbs where it is easy to obtain a ﬂexible labour force. Also, the convenient conditions of transport and communication make it possible for enterprises to communicate with each other in commodities, service, information, labour force and technology. Second, medium-scale and small-scale enterprises with a relatively high degrees of specialization have formed clusters, and the relationship between enterprises in the same industry or between related enterprises is one of dynamic competition and cooperation. This kind of relationship helps clusters to maintain their energy and competitive power. From the above analysis, Chapter 3 concludes that the formation of Tianjin’s bicycle enterprise clusters is the result of many factors, such

Masatsugu Tsuji and Akifumi Kuchiki 13

as geography, economy, society, humanism, institutions, technology, the market and history. It has not resulted simply from the action of enterprises, nor was it able to appear in a short time. Enterprises within the clusters have an obvious coordination effect. Once this kind of organizational structure is formed in a region, other regions have difﬁculty in copying it quickly. It should be considered a signiﬁcant factor in the continuous improvement of the competitive power of Tianjin’s bicycle enterprise clusters. Conclusion: to cultivate Tianjin’s new industrial clusters In the process of industrial agglomeration, both high-tech industrial clusters and traditional industrial clusters have their respective characteristics in shaping processes, features and development trends. Regarding these four industrial clusters as our research object, through contrast and analysis on shaping process, features and development trends, Chapter 3 raises future characteristics and development trends for industrial clusters in Tianjin. The development trends are as follows: (1) A developed interrelated industrial cooperation system in which foreign-funded enterprises are the centre. Tianjin’s industrial clusters regard large foreign-funded enterprises as the centre. This type of cluster, however, does not yet enjoy the beneﬁts of an ‘interrelated connection of innovation networks.’ Rather, with local enterprises still dependent on core enterprises for funding and technology, these clusters still experience a hierarchical division of labour within the value chain. This is one reason why these clusters are inherently fragile, and why they are still largely at the bottom of or outside the global production system. (2) Being good at study, imitation and innovation, and to bring dynamic ability into full play. Through the alliance, medium and small-scale enterprises may form a certain kind of scale economy and maintain and strengthen their own competitive advantage of technology and also improve their core and dynamic ability. (3) With the tie of the market, industries in urban areas and industries in rural areas should be combined to form a rational structure of priorities. As described on p. 165, so far, Tianjin has formed a pattern of industrial clusters with the characteristic of distribution according to priorities in urban and rural areas. The enterprise clusters in capital-intensive or technology-intensive industries gather in economic–technological development areas, free-trade zones, new technical industrial parks and some development areas in counties. Traditional industrial clusters have appeared in some industrial parks in villages and towns. Relying on new technology and new concepts of management, these traditional industrial clusters are being reformed into new industrial clusters. (4) Support by a developed international supply chain and market network. On p. 167 we argue that cooperation in the international supply chain is the best choice for entry into the international market. However, the ﬁnal goal is not the internationalization of enterprises. The ﬁnal goal of

14 Introduction

an enterprise is to occupy market share and gain proﬁts. The market network can collect rich market information for enterprises and provide unblocked channels for selling. (5) Positive enterprising concept and clustering culture. It is our brief that ‘Tianjin’s future industrial clusters of the new type should maintain and develop a coordinated and open clustering culture. In addition, future industrial clusters in Tianjin should incorporate a foreign advanced culture and cultivate compound competitive advantage’ (p. 167).

Chapter 4: A Flowchart Approach The economic development of East Asia can be explained by an industrial cluster, or industrial cluster policy, and industrial agglomerations such as the export-processing zone in Kaousing, Taiwan, established in 1965, the free-trade zone in Penang, Malaysia, in 1971 and the export-processing zone at Tan Tuan near Ho Chi Minh, Vietnam, in 1993 are typical examples. Industrial Parks near Shanghai developed by central government as well as all levels of local government are other examples. Chapter 4 makes clear the conditions required for forming a cluster in developing countries, and attempts to present a new method referred to as a ﬂowchart approach. The author applies this method to analyse the Toyota Motor Corporation cluster in Tianjin, China, as a case-study. The ﬂowchart approach proposed in Chapter 4 is used to extract the factors contributing to a cluster and to explain how these affect each other so as to form a cluster. The four factors proposed in this chapter are: (1) industrial zones, (2) capacity-building, (3) anchor ﬁrms, and (4) related ﬁrms. These are the key to industrial clustering in implementing industrial cluster policy. Capacity building means facilitation of physical infrastructure, institutional reform, human resource development, and preparation of living conditions. Institutional reforms include streamlining investment procedures and establishing one-stop services. The cross-dock logistics introduced in Shanghai of a just-in-time system contribute to value chain management but weaken an incentive to form an industrial cluster in Tianjin. The following three factors have positive effects on expanding an industrial cluster: a keiretsu relationship between an anchor ﬁrm and its related ﬁrms, scale economies, and modularity. Chapter 4 presents two interesting theoretical models for analyzing the decision of an anchor ﬁrm to construct a new production plan in a cluster, and on the scope of the value chain, namely, a ‘regional value chain network’. The author shows that, in the former, the crucial elements for the decision are scale economies and external economies of the anchor ﬁrm and that these are related to the size of the market (the amount of production), while those in the latter are the international vertical or horizontal division of labour and of parts, and outsourcing strategy.

Masatsugu Tsuji and Akifumi Kuchiki 15

After the theoretical formulation, the chapter applies the ﬂowchart analysis to Toyota in Tianjin. The ﬁrst step in forming an industrial cluster is to ﬁnd the market for products. Here scale economies in China are crucial to forming an industrial cluster in Tianjin. The market in the case of Tianjin is domestic. The next step is to determine whether there are saucers for ﬁrms to build plants. Tianjin has (1) the Tianjin economic development area of 130 km2 and other industrial zones established in 1992. Tianjin can also offer to ﬁrms as an incentive a good port, a physical infrastructure to be part of (2) capacity-building. Tianjin Port is effective in forming an industrial cluster. Tianjin’s government has been criticized for slowness in reforming laws and regulations but it is now improving. Living conditions in Tianjin are favourable to foreign investors such as the Japanese since there is a Japanese department store and many Japanese restaurants. Tianjin offers to ﬁrms as a further incentive many university students as human resources. Toyota started to operate in October 2002 as (3) an anchor ﬁrm for a cluster in Tianjin. Toyota merged with Daihatsu and decided to establish a subsidiary, partly owing to Daihatsu’s subsidiary located in Tianjin, though it had several candidate cities to invest in. Japan’s keiretsu system was effective in inviting (4) related companies of Toyota. Inviting an auto assembly company as an anchor ﬁrm is effective in implementing industrial cluster policy, since a car is composed of some ten thousand components. Component ﬁrms move into a cluster where an anchor ﬁrm exists if they can have surplus by building a plant. Cross-dock logistics are characteristic of Toyota’s Tianjin cluster. The just-in-time system there collects components in Shanghai and transports them to Tianjin. The logistics reduces transportation costs from Shanghai to Tianjin and weakens ﬁrms’ incentive to move into Tianjin. After discussing these factors, Chapter 4 concludes that the crucial factor in the decision to build a factory in Toyota’s case was economies of scale.

Chapter 5: Chinese Automotive and Parts Industries This chapter analyses how Chinese automotive assemblers and their parts manufacturers choose their locations. There are approximately 120 automobile assemblers and 2201 registered parts suppliers, and the characteristic of locations is that concentrated areas of the industry show a wide dispersion all over the country. The economic grounds of this are the rational behaviour of each ﬁrm to maximize proﬁts subject to various constraints such as costs of transportation, transactions and information, but also the size of the market. The more ﬁrms concentrate in a particular region, the larger the information ﬂow such as managerial know-how and technological innovation, and all participants there receive beneﬁts from this. This cumulative process attracts more economic resources to the region.

16 Introduction

In addition to the economic factors, Chapter 5 emphasizes political factors in the Chinese automotive industry. The location of assembly plants was chosen mainly from the viewpoint of national security in such a way as to scatter factories around the country. Those locations of the assemblers became a core to attract other parts manufacturers, and this process created the agglomerate regions of the automotive industry such as Jiansu, Sichuan, Zhejian, Jilin and Shangdon provinces, and Shanghai and Beijing cities. In 1987, the Chinese central government designated six large motor assembly plants as ‘key ﬁrms’ targeted for expansion with administration support. The six automotive corporations were given different market goals. Three were allocated to the large passenger car sector and three to the small passenger car sector according to the displacement, and the plan was referred to as ‘three big, three small’. According to this plan, China’s main vehicle manufacturers would be the First Auto Works (FAW) in Changchun, the Second AutoWorks (SAW) in Hubei and the Shanghai Vehicle Factory, which formed a joint venture with Volkswagen. The three small players would be joint ventures of Beijing Jeep, Guangzhou Peugeot and Tianjin Daihatsu. However, the plan was expanded to ‘three big, three small, two mini’ under pressure from the Central Military Commission. As a result, the scales of production of the ‘eight major large automotive ﬁrms’ were very small, while other governments in the world imposed a strict concentration policy on the automotive industry. These eight key companies controlled only 28.7 per cent of total volume of terminal assembly plants and 7.5 per cent of components plants. Although they accounted for 66 per cent of vehicles production, their individual scale was very small by international comparison, with an average output volume of only 27,466, far below the efﬁcient level of production, which is estimated to be between 250,000 and 300,000 vehicles. The eight majors accounted for less than half of total assets and investment made in the automotive sector in 1995, which indicates that the Chinese central government was unable to invest strategically. The political issue can be found in the solution of the above problems. In order to obtain a sufﬁcient level of production, the Planning Committee and the Ministry of Machinery Industry (MMI) determined to restrict severely the number of automobile plants. MMI had worked hard on sorting out the chaos in new entry acquisition to make some sense of its role in the broad development of the Chinese automobile industry. MMI, however, ran into much difﬁculty in performing its functions in control and coordination. Not only did its policy preference differ from the rational interests of most of the local governments and the defence industry, which have a bureaucratic stature equal or higher than MMI, but also its ﬁnancial weakness rendered it unable to enforce its authority effectively. Moreover, this was due also to the governance of local governments in seeking their own strategy to develop their economy, and they made use of its political

Masatsugu Tsuji and Akifumi Kuchiki 17

power to nurture the automotive industry according to their local industrial policy. This resulted in a larger number of automotive assemblers and an inefﬁcient level of production. Next, Chapter 5 focuses on the city of Shanghai as a case-study, which shows the remarkable development of the automotive and its parts industry. Almost half of parts manufacturers in China are located in Shanghai and its nearby provinces such as Jiangsu and Zhejiang. The success lies in not only the historical background of the manufacturing industry and the largest amount of foreign direct investment (FDI), but also the organization of municipal governments such as the Shanghai Localization Ofﬁce (SLO), the Shanghai Municipality Trans-bureau Localization Cooperation Ofﬁce (SMTLCO), and the Municipal Automotive Small Group (MASG). The bureaucracy of the Chinese government is so strong that the different bureaux have their own authority and jurisdiction. They cannot surpass the authority of other bureaux. The organizations of Shanghai municipal government used strong leadership to nurture and promote its automotive industry. Major Japanese automotive assemblers such as Toyota and Honda have already established plants in China, and Nissan will be producing soon. Chapter 5 brieﬂy examines their relationship with their parts suppliers and make a comparison with that in Japan, and from ﬁeld research on Toyota in Tianjin concludes that current Japanese automobile assemblers cannot exploit merits of agglomeration through the hierarchical production structure, since the amount they produce is far less than that required for economies of scale.

PART III: INDUSTRIAL CLUSTERS OF IT INDUSTRIES Chapter 6: Can Malaysia’s Growth Be Sustainable? This chapter focuses on the Malaysian electronics industry, which has been successfully forming an industrial agglomeration since the mid 1980s, as a result of which Malaysia has been a leading producer of electronics products in the international market. However, owing to the recent rapid emergence of China as an industrial power, it may not be possible for the Malaysian electronics industry to sustain its current industrial growth. In order to analyse these issues, the chapter examines empirically the international competitiveness of the Malaysian electronics industry, using various analytical measures. With these methods, it analyses rigorously the Malaysian and Chinese electronics industries in the global market, examining whether the two industries are competitive or complementary. Finally, the chapter investigates whether the new rise of the ICT industry in Cyberjaya in Malaysia may constitute another industrial cluster to sustain the country’s growth.

18 Introduction

The Malaysian electronics industry is dispersed rather than concentrated all over the country, so that, the regional distribution of foreign and domestic investment tends to be dispersed among such states as Selangor, Penang, Perak, Johor, Negeri Sembilan, Melaka, Kedah, Pahang and Sarawak. This is shown by examining the Gini coefﬁcients of foreign and domestic regional investment patterns in 1993 and 2002, which are 0.67 and 0.50, respectively, values signiﬁcantly lower than those of China and Thailand, for example. The recent ﬂow of FDI into the Malaysian electronics industry went up enormously in 2000, after a short drop during the 1997 Asian crisis. The characteristics of the recent ﬂow of FDI are aiming for the expansion and/ or diversiﬁcation of existing investment and projects, and concentrating on the production of high value added products such as plasma TVs. These seem to reﬂect the high level of conﬁdence prevalent among foreign investors in Malaysia. Chapter 6 analyses the competitiveness of the Malaysian electronics industry by calculating the RCA index, which indicates the comparative advantage of each commodity within a country compared with the world trade structure. The author shows that there is no change in that Malaysia continues to specialize in the production of electronics products in general, but some products such as ofﬁce machines and computers raise the index, while others like radios and sound recorders tend to decline. This means that the industrial competitiveness of Malaysia shifts away from the production of low value added products to that of high value added ones. Despite the closure of some establishments and the shifting of their operations to lower-cost production countries such as China and Vietnam, many existing ﬁrms continue to engage in expansion or diversiﬁcation of their investment projects in the production of relatively high value added products. Chapter 6 also focuses on whether China is a threat to Malaysia, that is, whether the emergence of China as an industrial power may have a signiﬁcant impact on the Malaysian electronics industry. The background of this argument comes from the ﬂow of FDI into both China and ASEAN. It is very clear that ASEAN cannot compete with China in terms of the absolute value of FDI inﬂow. Moreover, the gap seems to have become even wider in recent years, as the entry of China into the WTO turned around the declining trend of FDI ﬂow into that country, while the ﬂow of FDI into ASEAN remains stagnant. In order to answer the above issue, the chapter again utilizes the RCA index. The procedure is as follows. The RCA indexes are calculated for Malaysia and China for electronics products. These indexes are ranked for each country respectively and the Spearman’s rank correlation coefﬁcient between the rankings of the two countries is calculated. If they show a high positive correlation, their trade structure is very competitive. That means that the emergence of China as an indus-

Masatsugu Tsuji and Akifumi Kuchiki 19

trial power may have a signiﬁcant impact on Malaysia. On the other hand, if the coefﬁcient is near zero or negative, their trade structures are rather dissimilar and could be complementary to some extent. The results of the analysis for the Japanese market show that in 1996 the coefﬁcient was a little bit high (0.453), but statistically insigniﬁcant, while in 2000 the coefﬁcient became even lower and close to zero (0.134). Considering the high degree of involvement of MNCs in the electronics industry in Malaysia and China, the Chapter 6 concludes that MNCs tend to produce different types of electronic products in China and Malaysia, that is, China and Malaysia can be complementary in the electronics industry.

Chapter 7: Bangalore’s Software Cluster It is widely recognized that capital, technology, information and human resources embedded in a region are essential factors for ﬁrms to agglomerate there. This chapter examines as a case-study Bangalore in India, one of the leading centres of software clusters in the world, by analyzing (p. 244): the conditions under ‘which knowledge-intensive industrial clusters in developing countries build competitiveness, with particular focus on the dynamics of the local labour market. In other words, this chapter examines the characteristics of the local labour market, such as an institutional mechanism for promoting innovation and strengthening the competitiveness of its software industry’. In the 1990s, Bangalore emerged as the largest software cluster in India, achieving tremendous growth particularly since 1997, mainly through exports of lower-end software and IT-enabled services to the US market. In more recent years, leading IT global players such as Motorola, HP, TI and IBM have started outsourcing their R&D to Bangalore, allowing the cluster to gradually move up the value chain by engaging in higher-end services such as embedded systems. Lower-end IT-enabled services have also grown rapidly. A question posed by this chapter is: Why did this phenomenal export growth occur, particularly in Bangalore? The answer is an afﬂuent, skilled, IT-related labour force with proﬁciency in English in that region. As of 2001, approximately 80,000 IT professionals were working in Bangalore, accounting for about 20 per cent of the IT professionals in the whole country. Chapter 7 analyses the factors to explain this in the demand and supply sides of the labour market, the latter including educational institutions in the region. In 1998, Bangalore established the Indian Institute of Information Technology (IIIT) as a joint initiative between government and industry. The IIIT-Bangalore was located right in the centre of the International Technology Park (ITP), in order to improve the quality of training in IT-related courses and train IT professionals to meet industry needs. In addition, there were (as of 2001) 106 engineering colleges in Karnataka: two government-run and the rest privately funded. Private engi-

20 Introduction

neering colleges in Karnataka accounted for 19 per cent of the 553 all private technical colleges nationwide. These institutions contributed greatly to the IT-skilled workforce for the software industry in the region. Moreover, in-ﬁrm training is a principal means to develop IT professionals’ skills in Bangalore. Because technology changes so quickly, ﬁrms need to keep upgrading their skill sets, by retraining their employees in ‘hot skills’. The average life of technical skills is only three years. Major ﬁrms frequently update the skills proﬁle of their employees through in-ﬁrm training, in technical skills, domain knowledge and communication skills. It is said that ﬁrms spend about 5 per cent of total labour costs on in-ﬁrm training, and on average employees spend 10 to 12 per cent of their work time on such training. Another characteristic of the local labour market for IT is found in its high mobility. Chapter 7 discusses the fact that the average IT professional works only 2 to 5 years at one ﬁrm. During the boom years between 1997 and 2001, the average labour turnover rate rose to 22 to 25 per cent, sometimes even up to 30 per cent, as engineers moved frequently to other ﬁrms offering higher salaries. After the IT industry slowed down in 2001, the industry average for labour turnover declined to 8 per cent. The high turnover rate in search of higher remuneration brought about positive effects for the skills development of workers. Workers in the software and services industry are in a fast career trajectory. The author of Chapter 7 found an interesting fact through ﬁeld research that in only 4 to 5 years software engineers reach the top level, becoming project leaders. Software design engineers must have at least 5 years of experience, while software programmers must have at least 3 years or more of experience, while junior engineers doing coding need only 1 year. Because of this fasttrack career development, most leading ﬁrms have performance-driven and merit-based compensation schemes. This again causes high mobility in the local labour market. The dynamism of the local labour market comes also from spin-offs, a phenomenon similar to that in Silicon Valley. Many workers in large MNCs such as HP and Motorola and in leading Indian ﬁrms leave to set up their own ﬁrms. Former Wipro employees, for example, have set up more than 100 companies with no assistance from the ﬁrm. These IT professionals use the knowledge and skills they acquired in large ﬁrms in their new startup ﬁrms. The relatively low capital requirements for such new startups, along with the growth of venture capital across the country, have facilitated this trend. Chapter 7 also analyses successfully a so-called virtuous cycle that generated knowledge management in the region. The process was formed through the following mechanisms: (1) the state government quickly responded to the enormous growth in demand for skilled IT manpower through its manpower planning polices; (2) high intra-ﬁrm transnational

Masatsugu Tsuji and Akifumi Kuchiki 21

mobility as well as inter-ﬁrm mobility has helped ﬁrms to quickly deploy and mobilize the skill sets they require; (3) the emergence of local labour market intermediaries in terms of stafﬁng agencies as well as the industrywide collaboration in carrying out a periodic compensation survey helped facilitate the working of the local labour market; and (4) both MNCs and leading Indian ﬁrms actively promoted in-ﬁrm training to upgrade the skills of the workforce in their internal labour markets. These altogether have led the cluster to better mobilize the skills to meet the global demand and move up the value chain, and thus build competitiveness.

Chapter 8: The Local Economy: Lessons from Japan Since the Plaza Accords in 1985, many Japanese manufacturing companies have moved production bases from Japan to overseas locations such as Southeast Asia and East Asia because of the sharp appreciation of the yen. In the 1990s, Japanese manufacturing companies shifted investment to factories to China owing to the low costs for production and a global management strategy. These drastic changes in ﬁrms’ investment strategies, coupled with the economic deterioration following the bursting of Japan’s ‘bubble economy’, have had a serious negative impact on the Japanese local economy. As a consequence, not only per capita incomes but also the number of factories in regions have decreased. For the latter, from 1985 to 1990, the number of factories decreased at the annual rate of –0.1 per cent; from 1990 to 1995 it was –2.2 per cent, and from 1995 to 2000 it was worse still, at –2.4 per cent. Chapter 8 makes an attempt to analyse policy measures taken by local governments as well as central government, to cope with industrial transformation and globalization as stated above, by focusing particularly on IT and information, communications and technology (ICT) policies. The chapter analyses the case of Okinawa Prefecture, which aims to promote the IT industry through collaboration with the central government, and attempts to draw lessons from the case of Japan in order to contribute to stable IT development in the local economy. The chapter introduces the Industrial Cluster Plan of 2001, and learns how the local government has attracted ﬁrms to its prefecture in order to form industrial clusters. The Industrial Cluster Plan of 2001 consists of the following three core programs: (1) to support ﬁrms to collaborate with universities; (2) to assist technical innovations in the industrial sector; and (3) to establish institutions for promotion of startups and entrepreneurs. The government prepared about ¥476bn in funds for the plan. Following the lead of the central government, the Okinawa government has implemented policies to promote and invite the ICT industry to the prefecture. Chapter 8 summarizes these into the following ﬁve policies: (1) Financial support for ICT companies established in Okinawa to reduce

22 Introduction

telecommunication costs, (2) Development of the ICT industrial cluster. Since the Japanese government laid a special telecommunication cable between Tokyo and Okinawa in order to reduce costs for telecommunications, costs for telecommunications between Okinawa and Tokyo have become inexpensive as compared with to or from elsewhere. The government of Okinawa has taken full advantage and provided support for data centres with funds up to 50 per cent of the total cost incurred in transferring an existing facility to Okinawa. (3) Enhancement of job opportunities for the young. (4) Human resource development. This includes training courses for engineers to learn advanced IT knowledge and for people who are willing to work for call centres. (5) Tax deduction and tax exemption for ICT ﬁrms. As for policy achievements, the chapter shows that as of May 2003 there are 34 call centres, 17 software companies, 10 information service companies, 13 contents companies and 6 companies in other categories. As lessons learned from the experiences of Okinawa, Chapter 8 points out that local governments should make a plan for forming industrial clusters in an effective way by cooperating with the central government; in other words, local governments should not duplicate policies implemented by the central government, but should supplement the disadvantages of their programs with the policies of the central government. The following two points are, however, suggested by the author for further development of the ICT cluster. First, local government should stimulate competition and promote cooperation among companies in the cluster. In the case of Okinawa, the prefectural government played an important role in inviting ICT companies to the area; however, it did not act as a promoter for competition and cooperation in the cluster. Second, the local government should keep a balance between support for vocational training and for new business. The government committed itself to support human resource development; however, it has not paid much attention to support for new businesses or venture businesses. To stimulate the local economy, it is also important to help entrepreneurs start new businesses. Thus, local governments need to maintain these balances.

Chapter 9: Japan: Remedies to Activate Local Cities Silicon Valley and Multimedia Gulch in the USA, Bit Valley in Tokyo and Sapporo Valley in Hokkaido are typical examples of IT clusters. As of March 2002, approximately 30 per cent of establishments engaged in IT business were located in Tokyo. It is common all over the world that the IT industry tends to agglomerate in the larger cities. This chapter makes an attempt to explain empirically the factors that determine IT-related ﬁrms to agglomerate to a certain region. This issue attracts interest not only from academia but also from policy-makers, since while they have been

Masatsugu Tsuji and Akifumi Kuchiki 23

implementing so many projects to revitalize regions by promoting or attracting IT-related ﬁrms to them, in reality most of them seem to have failed. In this sense, the analysis of this chapter thus bears an important message. The chapter identiﬁes common characteristics that attract IT-related ﬁrms to a region by comparing the agglomerated regions listed above. In so doing, the author examines the supply and demand of information services. From the supply side, IT requires skilled engineers. Potentially skilled workers are most likely to exist near high-level educational facilities, and existing research also shows that industrial accumulation occurs near universities. This can be explained by the nature of information, that is, IT is related to embodied information such as experience accumulated in individuals which is difﬁcult to transfer. Face-to-face meetings play an important role to sharing such information. An industrial innovation depends on the resources of fundamental scientiﬁc knowledge of universities, research institutes and so on. Firms, therefore, concentrate their resources of related research into areas where universities and research institutes are easy to access. Regarding the demand side, it seems to be common all over the world that the IT industry tends to agglomerate in larger cities. Population density is a factor related to the size of consumer demand. In addition, another source of demand is other ﬁrms. Thus, existing industries are considered to increase demand for corporate services from the IT industry, and also to be a measure of economies of scale or industrial diversity as an element of accumulation. This is referred to as the accumulation effect. In addition to supply and demand factors, Chapter 9 emphasizes social amenities as an important factor for the location of IT ﬁrms. Since talented high-skilled workers are key in this high-tech industry, they demand dense information exchange between them, which brings about positive information and knowledge spillovers. A good quality of social amenities around the workplace may also be related to location, as a stimulator of face-to-face communication. Lastly, government policy both central and local an important element in attracting IT industry enterprises. After identifying the expansionary variables, the chapter constructs an econometric model to examine the effects of the above variables on the growth of the IT-related industries in the region, which are classiﬁed into the following three categories: (1) software business, (2) informationprocessing and (3) the internet. The result of the empirical analysis is summarized as follows. Empirical estimation proposes that the IT industry tends to accumulate in an area where there are many people, many establishments and many educational facilities. In particular, it can be shown that availability or easy access to educational facilities is the most inﬂuential factor. This is consistent with the idea that the IT industry is likely to rely heavily on knowledge workers to promote innovation.

24 Introduction

The policy recommendations derived from Chapter 9 are therefore such that if local governments wish to attract IT industry they should have highquality educational facilities such as universities and graduate schools as a minimum requirement. This requirement may become increasingly important in the information or knowledge economy of this century.

References Hotelling, H. 1929. ‘Stability in Competition’, Economic Journal, Vol. 39, pp. 41–57. Kagami, M. and M. Tsuji (eds) 2003. Industrial Agglomeration: Facts and Lessons for Developing Countries. Institute of Developing Economies–JETRO, Chiba-shi, Japan. Marshall, A. 1920. Principles of Economics. London: Macmillan (8th edn).

Part I Industrial Clusters in Vietnam

2 Competition and Cooperation: Vietnam Adam McCarty, Richard Record and James Riedel

Background to the study The previous JETRO–IDE study carried out in 2002, titled ‘Industrial Agglomeration: Facts and Lessons for Developing Countries’, analysed industrial agglomeration across Japan, Korea, China, Vietnam, Italy, the USA and Mexico. The study attempted broadly to answer questions as to why ﬁrms choose some locations and not others for their factories and ofﬁces. A key question was: Why in particular do we see so much high-technology agglomeration, when in theory new developments like the internet should supersede constraints such as distance? The 2002 study covered a broad spectrum of cluster types, from heavy industry clusters in Japan, through small and medium-sized enterprise clusters in Italy, to high-tech software clusters in the USA. From these existing clusters, lessons were learned for application to the developing countries covered in the study (namely Korea, China, Vietnam and Mexico). According to the 2002 project, industrial agglomerations can be grouped together into four broad categories: 1. clusters where locally specialized items are produced or local product districts; 2. clusters where a large core ﬁrm has many subcontracting or parts makers surrounding it, or the so-called industrial castle town; 3. clusters in large cities where lots of basic production processes are treated or urban processing clusters; and 4. government-led industrial parks and estates, often seen in developing countries. The 2002 study concluded that the success of these industrial agglomerations depends on: 1. infrastructure (highways, ports, electricity supply and so on); 27

28 Competition and Cooperation: Vietnam

2. institutional frameworks (legal systems, participatory actors, coordination among actors and so on); and, 3. government support (or foreign assistance) in terms of laws, taxation and ﬁnance. The Vietnam chapter in the 2002 study, titled ‘Agglomeration of Exporting Firms in Industrial Zones in Northern Vietnam: Players and Institutions’, focused on the fourth type of cluster – industrial zones (IZs), industrial parks (IPs) and export processing zones (EPZs). Development of such zones began in earnest in Vietnam in 1994, and at present much of Vietnam’s foreign direct investment is channelled into such zones. The study looked at zones as being a proxy for the kind of public goods normally provided in a developed country, that is in a developed country entrepreneurs would expect to have infrastructure and institutions such as electricity, roads, telecommunications, water supply and sewerage and so on. In countries such as Vietnam, such facilities are provided to a much higher standard in zones than elsewhere, and by a number of different actors including the public sector, but also the quasi-public sector and the private sector. For example, Japanese Ofﬁcial Development Assistance (ODA) has ﬁnanced the construction of such zones in Vietnam. The 2002 study compared practices in Vietnam with other regional countries (Thailand, Malaysia and the Philippines), and applied the theory of endogenous growth with quasi-public goods to show the positive impact of IZs on economic growth. The 2002 Vietnam study also looked closely at the role of Japanese ﬁrms in developing IZs across the region, and then at their impact on particular IZs in northern Vietnam. The paper summarized the advantages of two successful IZs in Northern Vietnam as having: 1. 2. 3. 4. 5. 6.

a favourable location; a good standard of education in the labour supply; abundant cheap labour; good infrastructure; existing support industries; and straightforward administrative procedures.

The paper also examined the transport infrastructure and some of the vocational and technical educational institutions in the locality. The study concluded that the infrastructure and administrative advantages for ﬁrms choosing to locate in the IZs were signiﬁcant and that these factors were contributing towards the agglomeration of foreign-invested ﬁrms and the promotion of economic growth in northern Vietnam. International experience suggests that much of the innovation that enhances growth and expands trade is likely to occur in industrial clusters

Adam McCarty, Richard Record and James Riedel 29

in a few major cities that have a strong research capacity and a full suite of other amenities. Agglomeration economies in these major cities can feed the growth impulse generated by industrial clusters. Clusters form as a result of multiple factors. Among these are policy measures that inﬂuence the entry of entrepreneurial ﬁrms: the availability of skills, infrastructure and amenities and also the presence of institutions that have a signiﬁcant role in innovation. The direction of causality between clustering and industrial growth is debated. Questions remain as to whether clusters can really be ‘created’ and therefore drive industrial growth and development, or whether clusters develop merely as a naturally occurring side-effect of successful economic development. Clustering is associated with speciﬁc localization economies resulting from the presence of many ﬁrms drawn from a single industry or industry value chain. Clustering spurs innovation and allows ﬁrms to beneﬁt from the availability of diverse and speciﬁc skills, shared infrastructure, cross-fertilization of ideas and learning, knowledge spillovers and faceto-face contacts. Many clusters already ﬂourish in East Asia, and produce everything from buttons to electronics and cane furniture and are a source of continuous, incremental innovation; however, few high-tech clusters exist outside Japan. Industrial clusters in developing countries have recently become a focus of attention because clustering seems to help ﬁrms to overcome growth constraints, upgrade productive techniques and compete more effectively in distant markets. Experience suggests however that it is not desirable for governments to form industrial polices deliberately to build high-tech clusters. A common approach across East Asia involves investing heavily in transport and communications infrastructure, serviced land and research facilities, and providing incentives for high-tech industries to locate in the designated area. Cities can strongly inﬂuence production and productivity by serving as centres for the clustering of industries and services serving domestic and global markets. They are also crucibles for stimulating consumption, developing tastes, spreading new consumer habits and trends, and engendering different lifestyles. As the original ‘tiger economies’ of East Asia have begun to discover, high savings and investment, and greater participation in the workforce can launch a country’s industrialization, and for a time exports can be an important source of growth and indeed the driver of economic development. Ultimately, however, domestic consumer spending, especially by urban consumers, is vital to sustain stable growth. Moreover, sophisticated and cosmopolitan urban consumers can push local producers into reﬁning their offerings and provide the market for the testing and launching of new products and technologies. Experience suggests that openness and the freedom of movement of skilled labour is vital for creativity. Linkages offering opportunities for collaboration with ﬁrms in other clusters can support innovation, and the

30 Competition and Cooperation: Vietnam

circulation of human capital from overseas supports creative innovation. Many of the East Asian economies have beneﬁted from the great diaspora of skilled workers employed in the large North American and European clusters where they have acquired valuable intangible assets, skills and human capital, and contacts, as well as more tangible ﬁnancial wealth. In Hong Kong, Singapore and Taiwan, overseas Chinese have in many cases been the nuclei for cluster development. Provided the institutional environment is hospitable, Vietnam could similarly beneﬁt and learn from the experience and expertise of overseas Vietnamese (Viet Kieu) participating in clusters abroad. Industrial clusters, ‘competitiveness’ and ‘the new economic geography’ are all rather new topics with respect to Vietnam’s industrial development. This study of industrial clusters in Vietnam will begin in the next section with an overview of Vietnamese competitiveness, the environment for business and the recent reforms directed towards the private sector. The section after that will revisit the theories that lie behind industrial agglomeration and dispersion, then the next sections discuss some of the international evidence of successful clustering, and look at some of the recent work on cluster engagement in Vietnam. Next, a detailed analysis of potential clusters will be conducted using the techniques of geographic information systems (GIS) for 12 key sectors of the national economy. A look at business associations and their role as possible ‘cluster facilitators’ through enterprise cooperation in Vietnam will follow, then the discussion will revisit the role of IZs,/IPs and/EPZs as ‘cluster facilitators’, as covered in the JETRO–IDE study in 2002. The scope of analysis will be widened to include zones in the south of Vietnam – the nation’s industrial heartland, and to include FDI from all countries. The ﬁnal section will make conclusions and industrial policy recommendations based on the ﬁndings of the preceding sections.

Vietnamese competitiveness and the environment for business The macro economy There is no doubt that Vietnam’s economic growth story in the last decade can be called a success. GDP growth rates have been among the highest in the world: 6.1 per cent in 2000, 5.8 per cent in 2001 and 6.4 per cent in 2002 (ADB 2003). The transition to a market economy has occurred comparatively smoothly and with relatively strong growth and economic stability throughout most of the 1990s. From a closed economy, the ratios of FDI and exports to GDP have increased rapidly to levels that are high by international standards. Vietnam has signed bilateral trade agreements with many OECD member countries (most recently with the USA in 2002), is a member of AFTA and is seeking WTO membership. Poverty incidence has fallen from more than 70 per cent in the mid 1980s to about 37 per cent in

Adam McCarty, Richard Record and James Riedel 31

1998, and about 29 per cent in 2002 (according to the government’s ofﬁcial poverty line). Most employment growth has been generated in the informal and private sectors. This success was achieved despite institutional weaknesses and a policy bias against the private sector. Nevertheless, Vietnam remains a low-income country (annual GNI per capita US$ 410). Human development indicators are relatively high, given per capita income, but market institutions in Vietnam remain underdeveloped. Until the mid 1980s, the state actively discouraged private business activity. Most private enterprises and cooperatives in Vietnam are small or medium-sized enterprises (SMEs), as are about 70 per cent of all state enterprises. The public sector still dominates the industrial economy, contributing some 50 per cent of GDP (Table 2.1). The state’s share of output in the manufacturing sector has fallen somewhat over recent years from 59 per cent in 1995 to 51 per cent in 2001. This change is a result of the increasing presence of foreign-invested enterprises in the manufacturing sector, rather than in any increase in manufacturing by the domestic private sector. The share of capital absorbed by the state sector is presumably large, but precise estimates are difﬁcult to obtain owing to the difﬁculty of valuing state assets in a transitional economy. An indirect measurement is the share of domestic credit consumed by the state sector. The outstanding back debt of SOEs amounts to 90tn dong, or about US$6bn – some 40 per cent of total domestic credit (World Bank 2002). How much of this is nonperforming is unknown. The share of new credit extended to SOEs, while declining, still represents some 25 per cent of new loans. The almost total dominance in the banking sector of the big four state-owned commercial banks also makes it difﬁcult to assess whether new loans to SOEs are extended on market terms. The World Bank described recently how ‘Vietnam’s accounting standards afford some scope for discretion in what to count as proﬁts, making them a poor indicator of performance, and making performance of SOEs generally difﬁcult to assess’ (World Bank 2002). Coming ASEAN trade liberalization will strip away the protective tariffs under which many inefﬁcient SOEs in ‘heavy’ industries such as cement and steel have previously enjoyed protection, exposing them to more cost-competitive rivals in the region. International evidence suggests that barriers sustained by historical connections between loss-making SOEs and state-owned banks helps keep lossmaking enterprises aﬂoat and divert capital away from more efﬁcient producers. Research on Taiwan shows that the turnover of ﬁrms – entry and exit – has contributed to the changing mix of industry, technological gains and advances in productivity because of the differential between those entering and those leaving the market (Aw et al. 1997). In 2001, private domestic ﬁrms, although growing rapidly from a low base, still accounted for only about 9 per cent of manufacturing value

32 Competition and Cooperation: Vietnam Table 2.1 2001

Vietnam total and manufacturing GDP by ownership: 1995 and

Share of total GDP

Share of manufacturing GDP

1995

2001

1995

2001

100.0

100.0

100.0

100.0

Public (state and collective)

50.3

49.1

59.1

51.0

Private (total) Household enterprises Private companies

44.4 40.3 3.1

40.2 32.6 7.6

28.7 26.0 2.7

29.0 20.0 9.0

6.3

10.7

11.4

21.0

Total

Foreign-invested companies Source: General Statistical Ofﬁce (2002).

added (GDP) (Table 2.1). Compared with its peers, Vietnam has quite an underdeveloped private sector in terms of the contribution made by private companies to the economy as a whole, particularly in manufacturing. Even in China, the equivalent of private companies – the township and village enterprises – account for a larger share of manufacturing value added. Much of this is a direct result of the past embedded institutional and policy bias against the private sector. Approval by the National Assembly of the Private Enterprise Law and Company Law in 1990, and amendments to the 1992 Constitution to recognize the long-term role of the private sector, were important turning points in reducing ofﬁcial discrimination against the private sector. A key far-reaching reform was the enactment and implementation of the new Enterprise Law in 2000. Recently substantial changes have been introduced to address remaining constraints, including a small and mediumsized enterprise (SME) decree, simpliﬁcation of business licensing, easing of restrictions on private involvement in export trade, providing domestic and foreign investors with similar incentives, relaxation of registration and minimal capital requirements. The result has been a rapid acceleration in new business registrations, currently at a rate of 1600 per month (World Bank 2002). Domestic investors set up 22,000 new enterprises in 2002, with a total investment capital of US$3bn – signiﬁcantly more than the US$1.33bn in new FDI pledges. A snapshot of enterprises in Vietnam is shown in Table 2.2. The government’s SME decree of November 2001 allowed SMEs to enjoy ‘preferential treatment policies’. Early in 2002, the status of the private sector was enhanced when a ruling party plenum resolved that the private sector is ‘an integral part of the national economy’. Stronger ofﬁcial

Adam McCarty, Richard Record and James Riedel 33 Table 2.2

Enterprises in Vietnam: a snapshot in 2003

Private businesses (registered under the Enterprise Law) Household businesses (registered under Decree no. 02) Cooperatives (registered under the Cooperative Law) SOEs (registered under the State-Owned Enterprise Law) Farm households Foreign-invested ﬁrms (registered under the Foreign Investment Law)

100,000 2,400,000 15,000 5,200 10,000,000 3,000

Sources: www.sme.com.vn and the German Technical Cooperation (GTZ).

endorsement of the private sector embodied in recent policy changes has helped build investor conﬁdence. International donor support to the enabling environment for business has included assistance directed at: 1. providing a stable macroeconomic and institutional environment; 2. improving the policy and regulatory environment directly related to SME development; and 3. strengthening institutions and policy implementation to promote SME development. Despite recent progress, much remains to be done. Consultations on business issues are still on an ad-hoc basis, although the Vietnam Chamber of Commerce and Industry (VCCI) does play a large role. Recent reforms have not been consistently implemented in all sectors and provinces. The legal basis for establishing business associations is still not totally clear. (The role of business associations as ‘cluster facilitators’ will be discussed in Chapter 8.) In terms of the policy environment, critical constraints no longer relate to the policies adopted at the national level, which are now reasonably positive, but to the implementation by the various agencies of government below the national level and by the state-owned organizations serving the private sector. Progress is often (at least partially) undermined by administrative inertia. National institutions responsible for ensuring compliance with business regulations remain weak, and there is limited understanding of the intentions of the new policies at the middle and lower levels of government. Outside donors can inﬂuence the policy agenda. In some cases, donor support has facilitated and improved the formulation and implementation of government-initiated measures to improve the enabling environment (for example the Enterprise Law). In other cases, donor support has resulted in the issuing of new policy documents (the Comprehensive Poverty

34 Competition and Cooperation: Vietnam

Reduction and Growth Strategy – CPRGS) or regulations (the SME decree) that – while consistent with government policy – are unlikely to have been issued in that form without donor support. The impact of donor conditional lending in these areas is yet to be seen. Disbursements of the World Bank Poverty Reduction Support Credit (PRSC) are tied to CPRGS implementation; similarly the Asian Development Bank ties budget support to policy reforms in such areas as SME development and governance. The ‘watchword’ for such policy reforms is always ‘national ownership’, and this is crucial for such assistance to be effective. The private sector is now experiencing lively growth, a lagged response to policy reforms of recent years that improved the business environment and access to bank credit. The next step is to identify the remaining obstacles and the most productive, feasible interventions. Biases remain, but it may be that Vietnam is now beginning a period of rapid growth led by domestic businesses, rather than by FDI and commodity exports as it was in the 1990s. The emerging private sector is becoming more important as a source of investment and its contribution to national investment is rising fast (from 21 per cent in 2000 to 29 per cent in 2002). Vietnam desperately needs a vibrant private sector to create jobs for the labour market and raise incomes for the poor in the coming decade. The country’s youthful population sees some 1.2 million new labour market entrants every year. The public sector simply cannot absorb such numbers. Given the relatively high capital intensity of state-owned enterprises, compared with their private sector counterparts, the cost of expanding their employment base is simply unaffordable. If anything, employment in stateowned enterprises will contract over the next few years, as a result of restructuring and increased competition in product markets. Hence the expansion of small and medium-sized enterprises is the only reasonable alternative to generate off-farm employment in rural areas, and to give earning opportunities to new migrants in urban areas. Trade The collapse of the Soviet Union and Vietnam’s preferential Eastern bloc (Council for Mutual Economic Assistance) trading network in the late 1980s prompted Vietnam to search for new export destinations. A series of trade reforms, including the unifying of previous multiple exchange rates in 1989, propelled Vietnam into world markets. Initially exporters were extremely successful in penetrating world markets in a number of key commodities. Vietnam is now the world’s second-largest rice exporter; this represents a signiﬁcant turnround given previous food shortages and faminelike conditions in the mid 1980s. Similarly Vietnam is among the world’s largest coffee and seafood exporters. It also exports sizeable quantities of rubber, pepper and cashew nuts. In fact in some ways one might say that Vietnam has been too successful in expanding commodity exports –

Adam McCarty, Richard Record and James Riedel 35

the collapse of world coffee prices in recent years is largely a result of Vietnam’s unregulated and large-scale entry into the lower end of the coffee market. In addition oil exports continue to play an important part in the nation’s export portfolio, contributing approximately 20 per cent of exports throughout the late 1990s. More recently, Vietnam has begun to capitalize on its principal area of industrial comparative advantage – low-cost labour. Garment and footwear exports, which account for about a third of Vietnam’s non-oil export revenue, increased by respectively 32 and 12 per cent in 2002. Garment exports to the USA increased 16-fold between 2001 and 2002, pushing total exports to the USA above US$2bn, and making it the second most important market after the European Union. Garments may well soon overtake oil as Vietnam’s primary source of foreign exchange earnings. The industry has emerged as one of the key sources of employment generation, particularly in HCMC Binh Duong and Dong Nai provinces, and has attracted large numbers of migrant workers from rural areas. With some 80 per cent of the national population still based in rural areas, where underemployment is a continuing problem, it seems likely that there will continue to be a plentiful supply of willing migrant workers. There seems little likelihood that Vietnam’s comparative labour cost advantage, a key driver in what is a notoriously ﬂighty multinational industry, will be eroded in the near future. Nevertheless, there is scope for Vietnam to ‘move up the value chain’ in the garment sector. Only cutting, trimming and sewing are done on a truly large scale, many of the more valuable inputs in the garment manufacturing process being added in other countries. Vietnam is now the second-largest shrimp exporter to Japan, after Indonesia (Vietnam Economic News 2003b). Exports of seafood to the USA increased 28 per cent in 2002, despite US anti-dumping complaints, which have made market access more difﬁcult for Vietnam. Vietnam has also recently begun to make inroads into the higher-margin commodity exports markets, producing a higher proportion of arabica coffee beans instead of lower-margin robust beans, and producing more high-value seafood exports such as king prawns to specialist premium markets like Taiwan and Japan. The composition of Vietnam’s exports in recent years is shown in Table 2.3. While Vietnamese exports to the US market are encouraging and give credible cause for future optimism, sales to China have not fared so well. The dollar value of exports to China declined by 8 per cent in 2001 and by some 5 per cent in 2002. This poor performance casts doubt on the competitiveness of the Vietnamese economy, not least because other countries in the region have been able to increase their exports to China over 2002. This is of real concern, as the importance of China’s economy in both the region and globally is growing rapidly. Trade with China could counterbalance the impact on Vietnam of a slowdown in other regions. Perhaps

36 Competition and Cooperation: Vietnam Table 2.3

Vietnamese export composition: 1995–2002 (percentages)

Primary products Rice Coffee Sea products Oil and petroleum Manufactured products Garments Leather and footwear Electronics Handicrafts

1995 1996 1997 1998 1999 2000 2001

2002

9.5 10.6 8.3 19.7

11.7 4.6 8.9 18.3

9.5 5.4 8.6 15.6

10.9 6.3 9.2 13.2

8.9 5.1 8.4 18.1

4.6 3.5 10.2 24.2

4.1 2.7 12.3 21.9

4.4 1.9 12.2 19.5

8.3 3.8 – 1.3

15.7 7.2 – 1.1

16.4 10.7 4.8 1.3

15.5 11.0 5.3 1.2

15.1 12.0 5.1 1.5

13.1 10.1 5.4 1.6

13.6 10.8 4.2 1.6

16.4 11.1 3.1 2.0

Source: General Statistical Ofﬁce.

China, more than any other country, is seen as the model for Vietnam. Hence China’s entry to the WTO has accelerated Vietnamese policymakers’ efforts to follow. However, the evidence of Vietnam’s poor competitiveness compared with China underlines the dangers of following China’s reform process too slowly. In many ways Vietnam needs to be ahead of China. China’s development is a resounding force that cannot be excluded from global economic and investment decisions, whereas recent drops in FDI ﬁgures underline the contrasting fact that the small market of Vietnam can quite easily be ignored. Similarly, while Vietnam has a sizeable domestic market to act as a cushion to the export cycle, it is nothing like the cushion China enjoys. Competitiveness State monopolies (or cartels) in key service industries have kept costs very high in areas such as air transport, coastal shipping, railways, power and telecommunications. This goes some way towards undermining the competitiveness of the Vietnamese economy, and can also act as a disincentive to investors. For instance, the telephone installation fee is US$93 in Hanoi, compared with US$66 in Bangkok, US$33 in Jakarta and US$28 in Beijing. Vietnam is one of the most expensive countries in the world for both incoming and outgoing international telephone calls. A three-minute call to Japan costs US$6.9 from Hanoi, more than twice as much as from Bangkok, Jakarta or Beijing (JETRO 2002). Table 2.4 displays some of the weaknesses in Vietnam’s trade-related infrastructure. The above table presents some fairly damning evidence against Vietnam’s internal and external trade infrastructure, and underlines to an even greater extent Vietnam’s recent export achievements in spite of all the above.

Table 2.4

Vietnam’s trade-related infrastructure Ports/ airports

Air freight

Inland waterway shipping

Road freight

Railway freight

Power

Telecoms

Market structure

State cartel

State monopoly Small barge: private sector

Large barge: public sector

Private and public trucks

State monopoly

State monopoly

State monopoly

Availability of service

Reasonable

Good

Good

Good

Good

Unreliable with voltage ﬂuctuations

Reasonable

Price

High

Not available

Depends on location

Competitive

Adequate

Adequate

Very high

Productivity/ reason

Low • Insufﬁcient container handling capacity in ports. • Insufﬁcient airfreight facilities in airports

Low Low • Insufﬁcient • Insufﬁcient cold storage management • Delays due to leading to outdated insufﬁcient security dredging and checks lack of safety • Unsatisfactory enforcement facilities and handling techniques

Overall impact on trade

Varies by region Negative and product

Varies by product and region

Very low Very low • Very high • Delays in power losses new line installation • Restrictions on internetbased business tools

Neutral

Negative

Negative

Negative

37

Sources: Adapted from Hopkins (2002) and Nomura Research Institute (2002).

Low Very low • Underused • Insufﬁcient road railway development capacity and maintenance

38 Competition and Cooperation: Vietnam

Sophistication and productivity is inﬂuenced strongly by the quality of the business environment. Firms cannot employ advanced logistical techniques unless a high-quality transportation infrastructure is available. The infrastructure problems described above are a signiﬁcant reason why so much foreign direct investment and new business activity is taking place in purpose-built investment zones. (Chapter 9 examines further the role of such zones in enabling industrial cluster development.) Key elements in the modern logistics of trade, including e-commerce and web pages, face stringent regulations that are difﬁcult to justify from an economic perspective. Greater private participation in infrastructure services could help address these concerns. In the absence of private provision, government services in Vietnam tend to be driven by what ofﬁcials think is important to supply, rather than by the demands of exporters (Pham Van Thuyet 2002). Other countries in the region (principally Singapore and Malaysia) have made considerable efforts to invest in worldclass infrastructure, partially as a means to encourage inward investment and partially to facilitate and nurture high-tech, high-value ‘knowledge’ industry clusters. Clearly Vietnam has some way to go in this area. Such sentiment is reinforced by the competitiveness statistics published by the World Economic Forum (WEF) (Table 2.5). The WEF produces two indices, the growth competitiveness index (GCI) and the current competitiveness index (CCI). The GCI aims to measure the capacity of the national economy to achieve sustained economic growth over the medium term. It

Table 2.5

International relative competitiveness in Asia, 2001

Economy

GCI (macro) Rank

CCI (micro) Rank

2000 GDP per capita (PPP adjusted)

Singapore Taiwan Hong Kong Korea Malaysia Thailand China Philippines India Vietnam Sri Lanka Indonesia Bangladesh

4 7 13 23 30 33 39 48 57 60 61 64 71

10 21 18 28 37 38 47 54 36 62 57 55 73

23,000 17,223 24,448 17,311 8,924 6,489 3,953 3,956 2,403 1,974 3,512 3,014 1,561

Source: World Economic Forum (2001).

Adam McCarty, Richard Record and James Riedel 39

looks at the macroeconomic sources of GDP per capita growth, and generates predictions of the ability of a country to improve its per capita income over time. The CCI examines the microeconomic bases of a nation’s GDP per capita and provides insights into the level of GDP per capita that is sustainable in the long term. The index is made up of two sub-indices, the quality of the national business environment and the degree of company sophistication. Both indices are constructed using information and opinions of business leaders from surveys (around 4600 respondents). The GCI (macro) index is constructed of sub-indices covering technology, public institutions and the macroeconomic environment. The CCI (micro) index is constructed of sub-indices covering company operations and the quality of the national business environment. Indices such as those shown in Table 2.5 are purely subjective, and at ﬁrst glance it might be tempting to dismiss the competitiveness indices as being a function of per capita GDP, that is to say that competitiveness (or productivity) rises with living standards. However, the table does raise some interesting points, most notably how Vietnam is considered to be relatively more competitive in the macro index than the micro. In fact in the breakdown of the CGI (macro) index, as would be expected, Vietnam scores very badly on technology and public institutions, but is propped up somewhat by low labour costs and a stable macroeconomic environment. Improved cluster formation and activity could be the key to improving micro-level competitiveness. Knowledge clusters Spending on research and development (R&D) is one of the key drivers of long-term sustainable economic growth; it is the key to productivity growth and the expansion of the production frontier. However, for many developing countries such as Vietnam, signiﬁcant medium-term growth can be achieved through the adaptation, assimilation and application of existing foreign technology. Nevertheless, while R&D does not ensure innovation or rapid growth rates, such spending is a necessary condition for long-term growth. In the absence of an adequate volume of R&D outlay, a country will most likely not ascend the technology ladder. The experience of the OECD countries, especially that of the USA, suggests that the private and social returns from R&D can be substantial. Private returns have been estimated in the range of 10 to 20 per cent, with social returns ranging from 20 to 60 per cent (Griliches 1992). If 25 per cent is the median social rate, then a permanent increase in R&D spending equivalent to 2 per cent of GDP could raise the aggregate growth rate by 0.5 per cent. Empirical work indicates the high potential value of investing in R&D, and the creation of an environment conducive to R&D is clearly one of the principle prerequisites for the formation of knowledge and high-tech clusters. Just as important is the creation of an environment where innovators

40 Competition and Cooperation: Vietnam Figure 2.1 Opinions as to whether intellectual property is well protected, selected economies: 7 = strongly agree, 1 = strongly disagree 7

Survey score

6 5 4 3 2

K U

SA U

Ja pa n ng ap or e Si

Vi et na m In do ne si a C hi na Ph ilip Ko pi ne re s a, R ep .o f Th ai la nd M al ay si H a on g Ko ng Ta iw an

1

The score reported is the mean assessment of a survey of an economy’s business people. (Source: Global Competitiveness Report 2000 by World Economic Forum, copyright © 2000 by the World Economic Forum. Used by permission of Oxford University Press, Inc.)

and entrepreneurs have the freedom and incentives to pursue different avenues of R&D, without the fear that the fruits of their endeavors will be expropriated or lost owing to a poor intellectual property regime. On this measure, Vietnam scores rather badly (see Figure 2.1 for a subjective ranking of countries). While this is an imperfect measure, it reﬂects the distance Vietnam has to travel before would-be innovators can be assured that the gains from their own investment in R&D will accrue to themselves. In addition, the lack of an intellectual property regime is a considerable impediment to knowledge sharing and intellectual clustering, as entrepreneurs are necessarily protective of their own invested capital. While this lack of an effective intellectual property regime might not be considered such a big deal for a developing country such as Vietnam that is still ‘catching up’, it does do enormous damage to creative, media and research-based industries and undermines the prospect of high-tech cluster development – currently the most valuable type of cluster. Commentators have suggested that Vietnam could develop a software development industry along the lines of Bangalore in India. However, the disincentive effects of Vietnam’s poor intellectual property regime are likely to act against this; for example, currently almost no major international software manufacturers provide product support in Vietnam. However, merely clamping down in intellectual property rights abuses is not really an adequate solution. The price of the full version of Microsoft Windows XP Professional is in the region of US$400.1 Assuming a per

Adam McCarty, Richard Record and James Riedel 41 Table 2.6

Comparative educational indicators, selected economies, 1990

Vietnam Thailand Philippines Indonesia Low human development All developing countries

Real GDP/capita (USD, PPP)

Mean years schooling for 25+ years

Adult literacy (%)

1,100 3,986 2,303 2,181 1,110 2,170

4.6 3.8 7.4 3.9 2.3 3.7

88 93 90 82 49 65

Source: United Nations Development Programme (1993).

capita GDP in the USA of around US$30,000, then the cost of Windows XP represents approximately 1.3 per cent of annual per capita income. With a per capita GDP in Vietnam of around only US$400, such a cost represents an astonishing 100 per cent of annual per capita income. Were the price of Windows XP in Vietnam to be of the same per centage of per capita GDP, then its price would need to be around just over ﬁve dollars. This still represents a considerable premium over the current ‘pirate’ versions of Windows XP (it sells for about 60c. for a copied version), but it does seem likely that a signiﬁcant proportion of customers would be prepared to pay a premium of this magnitude for licensed software with a warranty. (A good example where something like this takes place is in publishing, where the Oxford University Press distributes English language training materials in Vietnam at substantially lower prices than in the West, but with lower print quality, and hence the market for photocopied (pirated) English language textbooks is quite small. Similarly popular novels such as the Harry Potter series are divided into sections and distributed separately on cheaper paper.) Clearly the universal provision of high-quality education is also a factor in the development of ‘knowledge’ industries. Table 2.6 shows how Vietnam compares very favourably with neighbouring countries with higher levels of GDP per capita. However, in common with other East Asian education systems, a bias towards the sciences and schooling based on rote learning means that Vietnamese graduates may know a large number of facts, but often do not have the skills employers demand. Nor does such a system encourage creative problem-solving, risk-taking or an entrepreneurial mentality. Investment Another factor affecting the successful development of clusters is the ease of entry (and exit) for ﬁrms into a given industry. In small, underdeveloped economies such as Vietnam, the hurdles for starting a new business or

42 Competition and Cooperation: Vietnam

BOX 2.1 The Binh Duong experience Binh Duong was a poor province before it separated from Song Be in early 1997. Today, it is one of the most successful in terms of economic growth, foreign direct investment, exports and employment. Between 1997 and 2001, the industrial output of the foreign-invested sector grew 45.9 percent per year. The corresponding ﬁgures for the private domestic sector and the state sector were 29.7 and 8.0 percent respectively. Exports grew 27.9 percent per year during the same period. Based on a local standard, the poverty headcount fell from 8.7 percent in 1996 to 3.0 percent in 2002. The determination of the local authorities to develop the business environment, and to foster trust between government and business, may be one of the main reasons for this success. There are regular opportunities for local leaders to meet, discuss and solve the problems faced by the business community. The people’s committee has established a consulting council for investment under the management of its vice-chairman. This council meets every Thursday to help investors and businesses overcome the difﬁculties they face. To attract foreign direct investment, a simpliﬁed registration procedure was established, allowing licences to be granted within three days. If consultation with central government agencies is needed, licences should be granted in no more than 15 days. A one-stop-shop policy was also adopted, allowing licences to be obtained within 7 days for private enterprises, and within 5 days for household enterprises. Individuals and businesses can transfer land-use rights faster than regulated by the existing legal framework. The process is 5 days shorter, on average, than stipulated in current regulations. Sources: World Bank (2002); Steering Committee for State Administrative Reform (2002).

expanding an existing one into new markets can be cumbersome and slow. Regulatory barriers are the largest obstacles, but other constraints include weak entrepreneurship, labour market constraints, limited organizational skills, unwillingness to dilute family ownership, poor infrastructure and insufﬁcient capital, especially risk capital. New businesses are being established at a rate of 1600 per month; it still takes some 10 steps and up to 68 days in some provinces to register one. Local government is the key factor determining the local business environment for private enterprises. The legislation now exists (the Enterprise Law), but the key, as with so many laws in Vietnam, is implementation and regulation at the local level. Binh Duong province (see Box 2.1) is an excellent positive example of how provincial authorities can make a difference. Binh Duong, and neighbouring Dong Nai and Ho Chi Minh City have emerged as the industrial and export powerhouse of Vietnam, attracting the lion’s share of FDI. These three provinces together attracted some 50 per cent of the total FDI into Vietnam in the ﬁrst half of 2003 (Table 2.7). Similarly, the same ﬁve provinces (HCMC, Binh Duong and Dong Nai in the south, plus Hanoi and Hai Phong in the north) provide the dominant

Adam McCarty, Richard Record and James Riedel 43 Table 2.7 Vietnamese licensed FDI projects January–July 2003: top ﬁve provinces Rank

Province

Total FDI investment, Jan.–July 2003 (US$1,000)

Share of total FDI Investment, Jan.–July 2003 (%)

1 2 3 4 5

Ho Chi Minh City Binh Duong Hai Phong Dong Nai Hanoi

207,547 103,679 89,487 77,790 47,671

26.7 13.3 11.5 10.0 6.1

Total (all provinces)

778,178

Source: Vietnam Economic News (2003).

Table 2.8

Vietnamese gross exports January–July 2003: top ﬁve provinces

Rank

Province

1 2 3 4 5

Ho Chi Minh City Hanoi Binh Duong Dong Nai Hai Phong

Exports, Jan.–July 2003 (US$1,000) 5,018,284 866,603 819,095 806,742 303,137

Source: Vietnam Economic News (2003).

contribution towards Vietnam’s exports (Table 2.8). For both FDI and export volumes, province in the sixth place is a signiﬁcant way behind the one in ﬁfth. The export data underlines Ho Chi Minh City’s role as the economic and industrial capital, generating almost six times more export revenue than Hanoi. One of the most striking characteristics of Vietnam’s FDI inﬂows has been the relatively low, although rising, share of FDI in light, labour-intensive manufacturing – Vietnam’s apparently principal area of comparative advantage. Most FDI has gone into heavy, capital-intensive (and labour unintensive) industry and sectors such as oil, construction, transportation, communications and property (STAR 2003). Much of this has been to serve the temporarily lucrative and protected domestic market, a form of import substitution where particular heavy industries have beneﬁted from protection. Falling tariffs and Vietnam’s entry into multilateral trade organizations will likely spell the end of this FDI phase. At present, while FDI accounts for a signiﬁcant share of GDP (13.3 per cent), industrial production (35.5 per cent) and domestic investment (18.5 per cent), it accounts for only a minis-

44 Competition and Cooperation: Vietnam

cule share of employment (about 400,000 jobs in a workforce of 40 million). This apparent misallocation of resources could be the reason for Vietnam’s relative lack of competitiveness compared with China. The changing nature of Vietnam’s FDI portfolio is now, however, having more of an impact on export-oriented light manufacturing. Between 1995 and 2000, the share of FDI in manufactured exports increased from 8.8 per cent to 33.8 per cent (Athukorala 2002). This trend is a reﬂection on the increasingly important role of other East Asian developing countries as foreign investors in Vietnam.

Theoretical literature review Until quite recently, economists have paid only limited attention to the location of economic activity and to the choices ﬁrms and households make to decide where to produce or consume. Most economic models are utterly ‘distance-less’ (space is not a factor), in the sense that little or no attention is paid to the friction between space and distance. Much international trade theory disregards the effect of transportation costs and only fairly recently have ‘iceberg’2 models been used widely as a means to incorporate the costs of movement on trade. Alfred Marshall (1920) witnessed in early twentieth-century England how industries such as textiles and garments manufacturing clustered in and around Lancashire, steel making in Shefﬁeld and so on. He commented: ‘Industries tend to cluster in distinct geographic districts, with individual cities specialising in production of narrowly related sets of goods.’ However, economic theory in Marshall’s days lacked the tools to conduct more detailed analysis. The basic problem with doing theoretical work in economic geography has always been rooted in the role of increasing returns to scale. Economists have been historically reluctant to abandon the world of perfect competition and constant returns. While such a world might be mathematically ‘pure’, pursuing this to its logical conclusion one would expect to ﬁnd an economy characterized by ‘backyard capitalism’ where every household or small group produces most items for itself. While there might be regional differentials in climate, soil conditions and natural resources to explain some differences in human location, this is simply not enough to paint a realistic picture of the real economy – densely populated manufacturing zones and thinly populated farming zones. The spectacular concentration of some industries (Silicon Valley, Hollywood and so on) cannot be explained by a constant-returns world; there must be some kind of self-reinforcing logic – increasing returns. This issue has been sidestepped somewhat in economic theory until recently through a number of theoretical ‘tricks’, such as simply taking the existence of cities as given, or representing increasing returns in a sort of ‘black box’ way as localized production externalities.

Adam McCarty, Richard Record and James Riedel 45

The difference between a constant returns to scale production technology and an increasing returns production technology can best be seen through Cobb–Douglas production functions as shown in equation (1): Y1 = AKα Lβ

where α + β = 1

(1)

Y1 in equation (1) deﬁnes output with a standard Cobb–Douglas production function with constant returns to scale, K represents capital, L represents labour and A represents all other elements contributing towards output (which can be taken as total factor productivity to include such factors as technology, management skills, organization and so on). Taking increasing returns to scale as an assumption, the production function can be modiﬁed as in equation (2): Y2 = Aγ Kα Lβ

where α + β + γ > 1

(2)

Y2 in equation (2) now deﬁnes output with increasing returns to scale; the sum of the superscripts must by deﬁnition be greater than 1 for this to be true. The inclusion of a γ term takes into account the role of increasing returns to not only capital and labour, but also the ‘other elements’. This would be intuitive, as ﬁrms that cluster may adopt new technology and improve productivity faster if there are other ﬁrms nearby to copy than they perhaps would have otherwise. The introduction of space or distance into economic modelling creates problems, but also solves some others. First, without increasing returns to scale, there is no reason for producers to incur transportation costs (the backyard capitalism model). Second, without a spatial model, increasing returns and efﬁcient pricing are incompatible; marginal cost would be below average cost. The existence of competition is also thrown into doubt; taken to its logical conclusion, inﬁnitely increasing returns would see the world dominated by one producer or at least one producer in each industry sector. However, with the introduction of space and time into the increasing-returns world, ﬁrms are likely to face a U-shaped average cost curve. Initially ﬁrms beneﬁt from returns to scale, pushing average costs down as output expands, but eventually the only way to increase output further is to ship the incremental product further. At some point the increasing costs of transportation outweigh and begin to reverse the gains from increasing returns. Using the iceberg model, then, the size of a ﬁrm is limited by its ability to reduce costs faster than the rate of ‘melting’. Hotelling (1929) introduced a microeconomic model for space and transport costs. In it one supposes that there is a beach along which sunbathers are evenly distributed, and on this beach are two ice-cream vendors, each wanting to locate to maximize sales. Every sunbather purchases one ice-

46 Competition and Cooperation: Vietnam Figure 2.2

Hotelling model: competition along a line

a

a′′

a′

b′

b

cream per day and in that sense demand is perfectly inelastic. The sunbathers buy their ice-creams from the nearest vendor. The optimum location is hence fairly obvious – the vendors must locate respectively one-quarter and three-quarters of the way from one end of the beach (at points a and b in Figure 2.2). However, even if the two agents start in these locations, they have strong incentives to shift their positions and move towards each other. In the extreme case, the vendor beginning at point a will move to point a⬘, thus capturing all of the market to the left of point b, three-quarters of the beach. Such a movement of course will not come without a response from the other vendor, who starting at point b will hop over the vendor at point a⬘. The vendors will keep on hopping back and forth until eventually both end up next to each other at the midpoint of the beach (points a⬙ and b⬘). This result is the only stable equilibrium; however it is not the optimum solution, as the average trip length to buy an ice cream is twice as long as in the optimum solution. While this is only a very simple model, and there are a considerable number of simplifying assumptions, it does show how competition can drive economic agents to colocate, or in other words to cluster. In more recent times, since the early 1990s, essentially two schools of thought have emerged on the role of location as a factor in economic activity. The ﬁrst has become known as the ‘new economic geography’, advocated chieﬂy by Paul Krugman and Anthony Venables. Seminar papers by Krugman and Venables (1990) and Krugman (1991) led to the development of research in this area. These theories are based ﬁrmly upon monopolistic competition tools and emphasize the interaction of increasing returns, transport costs and the movement of productive factors to understand the latent forces of economic agglomeration. Clusters are seen as being primarily geographic in nature and are considered to be a naturally occurring phenomenon. The second ‘school of thought’ is that advocated chieﬂy by Michael Porter, and has emerged from business school theory. It focuses more on looking at how like-minded ﬁrms in a given value chain can cooperate to advance their interests and improve competitiveness on a national or regional level. While clusters may be geographic, the primary focus is on social and institutional networks among ﬁrms. Porter advocates that cluster formation can and should be a tool of national industrial policy in order to improve competitiveness and productivity.

Adam McCarty, Richard Record and James Riedel 47

Krugman and Venables Marshall’s three-point classiﬁcation for the returns to spatial concentration made in 1920 still stands today. In more modern parlance, he argued that industrial districts arise because of: 1. knowledge spillovers (‘the mysteries of the trade become no mysteries; but are as it were in the air’); 2. the advantages of thick markets for specialized skills; and, 3. the backward and forward linkages associated with large local markets. All three of the above are clearly operating in the economy of today. The ﬁrst two elements are intuitively very convincing but difﬁcult to model, hence much recent theory has focused on the third, ‘linkages’ element. The story is essentially one of producers choosing locations that have good access to markets and supplies of intermediate goods. However, locations where a concentration of producers are already operating for whatever reasons are likely to already have good access to markets and well-developed intermediate supplies. Hence ﬁrms agglomerate to take advantage of existing backward and forward linkages. As a result of such linkages, a spatial concentration of production, once established, is likely to persist over time and indeed grow at the expense of smaller production centres. The knowledge spillovers argument is essentially one where increasing returns associated with knowledge, not the accumulation of capital and labour, are the key forces driving the growth of a production concentration. Much of the theory developed under the umbrella of the ‘new economic geography’ was conceived against the background of increasing international economic integration, principally in Europe (the EU) and North America (NAFTA). However the same backdrop of increasing trade, falling tariff barriers and falling transportation costs can be applied to the case of a developing economy where improved transport infrastructure can have a dramatic effect on internal trade. The new economic geography hypothesis is that such a background can lead to the endogenous formation of an ‘industrial core’ and an ‘agricultural periphery’. This has profound implications for industrial structure and the location of economic activity across regions. The ‘dynamic effects of economic integration’ model developed by Krugman and Venables considers an economic environment of one economy with two regions (the North and South of Vietnam could be taken as theoretical regions). Within this economy there are two goods: ‘agriculture’ and ‘manufacturing’. Consumer preferences are identical across all agents and both regions (all producers are also consumers). A constant proportion of expenditure is allocated to manufacturing and agriculture. Manufacturing consumption is composed of a large number of varieties of different manufactured goods. Consumers have a ‘love of

48 Competition and Cooperation: Vietnam

choice and variety’. For endowments the model considers only two factors of production: ‘farmers’ and ‘workers’, where only farmers can produce agricultural goods and only workers can produce manufactured goods. The farmers cannot choose to become workers and vice versa. Hence the economy is endowed with a ﬁxed number of farmers and a ﬁxed number of workers. Crucially farmers are assumed to be immobile (each region is endowed with a given number of farmers), and workers are assumed to be mobile (they will move to whichever region has a higher real wage). Agricultural goods are produced by farmers under constant returns to scale and perfect competition. Manufactured goods are produced by workers under increasing returns to scale and imperfect (monopolistic) competition. Transportation is considered to be costless for agricultural goods. Iceberg transportation costs for manufactured goods are applied where only a fraction τ (where 0 < τ < 1) of each unit shipped arrives. 1/τ captures the size of transportation costs or trade barriers. Hence the model demonstrates a number of dynamic effects that can be analysed to see under what conditions manufacturing activity agglomerates in a single region (the ‘core’), leaving only agriculture in the ‘periphery’. The forces for agglomeration are twofold. First, economies of scale and transportation costs imply that the ﬁrms would like to locate close to a large market (backward linkages similar to the Hotelling model). Second, consumers’ love of variety in manufactured goods and transportation costs imply that workers would like to live close to a large market (forward linkages). Together these two forces imply cumulative causation. The force acting to disperse economic activity acts through immobile farmers and transportation costs, implying that ﬁrms have an incentive to move out of the core to serve the peripheral agricultural market. The net outcome of these two sets of forces determines whether or not agglomeration takes place. Three parameters are important in determining whether or not agglomeration occurs: 1. the size of transportation costs (1/τ); 2. the share of consumers’ expenditure on manufacturing; and 3. the size of scale economies. A number of effects can be seen when there is a reduction in the size of transportation costs and/or trade barriers. Such changes can be seen in the context of regional integration on an international level, or regional integration on a national level through the ‘opening-up’ of a transitional economy such as Vietnam’s. Doi Moi in Vietnam has seen an improved environment for business, encouraging internal private sector trade and therefore improving parameter (1); and increased expenditures on manufactured goods as incomes have risen – parameter (2). Together, parameters (1) and (2) reinforce parameter (3) – economies of scale.

Adam McCarty, Richard Record and James Riedel 49

When transportation costs are large (τ→ 0), manufactured goods will in general be produced in both regions (to serve immobile farmers). As transportation costs fall (τ rises), it becomes feasible for manufacturing ﬁrms to agglomerate in the ‘core’ (where they beneﬁt from large market size) and transport manufactured goods to the agricultural periphery. There is a critical value for transportation costs (1/τ) below, which a core–periphery pattern of industrial development arises. When transportation costs fall to zero (τ→ ), the location of economic activity becomes indeterminate and location is irrelevant. If the two regions are identical, then which region becomes the core and which the ‘periphery’ may depend on historical accident. Krugman (1991) describes how trade and economic activity often take place as a result of more or less ‘arbitrary specialization based on increasing returns, rather than an effort to take advantage of exogenous differences in resources or productivity’. Certainly there is considerable evidence to support this ‘randomness’ theory; many industrial clusters in the West appear to have become dominant for little or no apparent reason. They maintain their dominance from the region’s recognition as a dominant region of a particular industry or sector. This could explain the fact that communications technologies such as the internet have not so far undermined the existence of the world’s major industrial clusters. The key conclusion is that the effects of reduced transportation costs (or tariffs) in models with agglomeration economies may have substantial implications for the location of economic activity and industrial structure in different regions. Such models give us considerable help in understanding the latent forces that lead to the development of industrial clusters. Porter Porter’s (1980 and 1998) analysis takes a much more micro-level analysis of clusters by looking at the ﬁrm-level effects and drivers towards agglomeration, rather than an analysis of the macroeconomy. Clusters and cluster strategies are seen within the context of improving ﬁrms’ (and thereby a nation’s) productivity and competitiveness. He deﬁnes clusters as: geographic concentrations of interconnected companies and institutions in a particular ﬁeld … They include suppliers of specialised inputs, components, machinery and services. Clusters also often extend downstream to channels and customers and laterally to manufacturers of complementary products … ﬁnally many clusters include governmental and other institutions. (Porter 1980) An industry cluster therefore includes members from every part of the value chain including all supporting industries such as packaging, advertising, transporters and so on. In addition to government and institutions,

50 Competition and Cooperation: Vietnam

one could also add business associations and social and collective organizations (social capital). Most clusters include: 1. 2. 3. 4. 5. 6. 7. 8.

end-product or service companies; suppliers of specialized inputs, components, machinery, services; ﬁnancial institutions; ﬁrms in related industries; ﬁrms in downstream industries (channels and customers); producers of complementary products; specialized infrastructure providers; specialized training, education, information, research (often universities, institutes and/or think tanks); 9. Vocational training providers; and 10. Standard setting agencies. Porter also stresses how nations are not able to succeed in isolated industries, but in clusters of industries connected through vertical and horizontal relationships. This is similar to the linkages argument ﬁrst articulated by Marshall. Porter further goes on to describe how the phenomenon of industrial clustering is so pervasive that it appears to be the central feature of advanced national economies. Cluster analyses, strategies and competitiveness initiatives use the ‘Porter diamond of advantage’ shown in Figure 2.3 to summarize the linkages and Figure 2.3

Porter’s ‘diamond of advantage’

Chance

Firm strategy, structure and rivalry

Factor conditions

Demand conditions

Related and support industries (value chain) Government

Adam McCarty, Richard Record and James Riedel 51

related factors considered to be most important in developing a successful cluster. The Porter diamond is essentially a logical structure for looking at the factors that determine the success (or failure) of a business or cluster of businesses. The government is outside the diamond (exogenous) and considered to be a factor only as a facilitator. There is no government–business partnership or the like, nor is there any scope for government activity within the diamond. Crucial to the formation of an effective cluster is the idea of ‘critical mass’. A cluster should include a critical mass of representation from the value chain so as to improve knowledge, share resources and stimulate creativity, innovation and entrepreneurship. The interconnections and spillovers within a cluster often inﬂuence productivity growth more than the actual scale of individual ﬁrms. Nevertheless, ‘Porter-style’ clusters in developing countries tend to be shallow and to rely primarily on foreign components, services and technology. Firms in such locations most often integrate vertically, producing not only their own components but even back-up electricity and such services as well. Similarly, cluster development in developing countries tends to be impeded by low local education and skill levels, weaknesses in technology, and lack of access to capital and poorly developed institutions. Hence in developing countries, a disproportionate amount of the country’s economic activity is often concentrated around large capital cities, and reﬂects the absence of infrastructure and the lack of available suppliers. Ultimately the greatest clash between the schools is derived from the use of Porter ideas to advance national ‘competitiveness’. Porter deﬁnes competitiveness as: Sustainable increases in productivity leading to improved standards of living for the average person. Porter’s work has been often used to further mercantilist and strategic trade arguments, particularly in developing countries such as Vietnam, that trade is a ‘zero-sum’ game. The idea that export productivity needs to be constantly driven up to combat imports and increase foreign exchange earnings is an alluring one, but ultimately ﬂawed. Ricardian comparative advantage theories show that trade is always and everywhere beneﬁcial, even when a country has no absolute cost advantages. If Porter clusters are not used to support industrial nationalism or mercantilist views, then in many ways Porter clusters are the practical counterparts of the more theoretical results of the economic geography models. A crucial condition to both the Krugman–Venables approach and that advocated by Porter is the idea that smaller ﬁrms clustered together can also gain the same (or at least similar) beneﬁts of economies of scale and size that just one larger ﬁrm (or a reduced number of larger ﬁrms) would gain from producing the same product. A cluster of ﬁrms, however, should in theory not be constrained by the limited dynamism inherent in a single larger enterprise acting as a local monopolist.

52 Competition and Cooperation: Vietnam

What are clusters and why are they important? International evidence In the present era of rapid and ongoing technological change, the costs of communications and transportation have plummeted. Foreign travel, once considered to be the exclusive pursuit of the wealthy, is now normal and commonplace. Such changes have led many to speculate upon the ‘death’ of distance, arguing that face-to-face interaction is no longer necessary for business; the telephone, email and the internet make interaction possible between people, enterprises and organizations from anywhere in the world. However, with the ongoing international trend towards tariff reductions, free trade and regional integration (ASEAN, EU, NAFTA and so on) we have seen increasing agglomeration in many industries. In terms of the Krugman–Venables model discussed in Chapter 3, the internet is a tool for reducing transportation costs and edging τ closer towards 1. In 1991, Krugman wrote: ‘The most striking feature of the geography of economic activity is … concentration.’ Developing East Asia has demonstrated signiﬁcant comparative advantage in high-tech products, and in many East Asian economies manufacturing has been concentrated in industrial parks in major urban centres. Principally this advantage has been conﬁned to product assembly, with most upstream research and development done in the more developed economies and with many intermediate goods and support services sourced from the more developed economies. While such downstream activities can provide valuable learning opportunities to assimilate technology and business practices, the real rewards come from being at the innovative frontier – designing, developing and introducing new products or services. This has to be the long-term aim of all developing countries in East Asia. There is increasing evidence that economies with dynamic knowledgeintensive industries clustered in a few urban venues are most able to move up the value chain. Porter and Stern (2001) noted: Clusters offer potential advantages in perceiving both the need and the opportunity for innovation. Equally important, however, are the ﬂexibility and capacity clusters can provide to act rapidly to turn new ideas into reality. A company within a cluster can often more rapidly source the new components, services, machinery, and other elements necessary to implement innovations. Local suppliers and partners can and do get involved in the innovation process, especially when participants are located nearby. Reinforcing these advantages for innovation is the sheer pressure – competitive pressure, peer pressure, customer pressure, and constant comparison – that is inherent within a cluster. This essentially is a continuation of the knowledge spillovers argument. Going back to Marshall, the principal other factors are the backward and

Adam McCarty, Richard Record and James Riedel 53

forward linkages and the thick labour markets that are inherent within a cluster of ﬁrms beyond a certain critical mass. Firm-level proximity can also act as a competitive force for preventing costs from rising too fast, and this is particularly important in an R&D environment. Why are clusters important? To spell it out in more detail, clusters increase productivity and efﬁciency by allowing: 1. efﬁcient access to inputs, employees (thick markets) and information; 2. ease of coordination across ﬁrms (there is a role for business associations here); 3. rapid diffusion of best practices; and 4. ongoing performance comparisons vs. local rivals and therefore quick productivity enhancement and cost control through the rapid spread of industrial best practices. Clusters also stimulate and enable innovations and commercialization through: 1. enhanced ability to perceive innovation opportunities; 2. presence of multiple suppliers and/or institutions to assist in knowledge creation; 3. locally available resources – ease of experimentation; and 4. lower barriers to commercialize new products, services, ideas and business models because of available skills and suppliers. Because knowledge is their lifeblood, clusters thrive on the density of informal networks made possible by proximity and to some extent further reinforced by communications technology. These two factors help develop horizontal relationships, a form of social capital and trust that induces the circulation of knowledge and is the foundation of good governance within producer networks. This social capital is essentially the glue that holds clusters of all types together. World-class clusters Table 2.9 gives examples of some of the most famous clusters around the world. The Silicon Valley cluster is probably the world’s most famous and valuable. An analysis of the key factors contributing to its successful development (Cohen and Fields 1999) found seven principal interactive forces: 1. strong research universities; 2. government policy (in particular defence related research for the US government was of vital signiﬁcance); 3. venture capital ﬁrms;

54 Competition and Cooperation: Vietnam Table 2.9

Examples of some world-class clusters

Silicon Valley

USA

Information communications technology

Bangalore

India

Information communications technology

Las Vegas

USA

Casinos

Sweden

Sweden

Forestry/wood products

London

UK

Financial services

Zurich

Switzerland

Private banking

Hollywood

USA

Films

Baden-Württemberg

Germany

Automobiles

Scotland

UK

Biotechnology

Hsinchu Park

Taiwan

Semiconductors

Antwerp

Belgium

Diamonds

4. 5. 6. 7.

law ﬁrms; business networks; stock options; and thick labour markets.

The combination of such forces has provided the technical knowledge, the ﬁnancial backing, the social capital, the incentives and the supply of specialized labour that has made Silicon Valley such a success. The irony now is that the costs of living and doing business in Silicon Valley and its surrounding area are among the world’s highest. This is a considerable disincentive for ﬁrms to operate in the region and for skilled workers to choose to locate there. The same can be said of London, where very high property prices and poor local transport infrastructure ought to drive economic actors away. The fact that this is not happening underlines the power of increasing returns that accrue from agglomeration; such beneﬁts outweigh other negative factors. These clusters are well beyond their critical mass point and hence the forces of agglomeration are selfreinforcing. Of much more importance for Vietnamese policy-makers is how clusters begin. Most international examples of successful clusters started in areas favoured by history and tradition. Development of clusters has generally taken off with the appearance of a cluster engine, such as a major university or research institute, a particularly dynamic ﬁrm, or an equivalent transnational corporation. Singapore, for example, initially climbed the technology ladder through knowledge and technology transfers brought in

Adam McCarty, Richard Record and James Riedel 55

by TNCs attracted by government incentives. In most cases cluster development is as much about what governments don’t do as about what they do. Limiting regulation and red tape can do considerably more to assist than crude and mistimed interventions by governments. However, there are cases where government facilitation has played the key role in guiding successful cluster development. The model developed in Baden–Württemberg, Germany’s industrial heartland, is a good early example of how public initiatives can create a system with multiple layers of institutes and research clusters that work with private companies and local chambers of commerce to meet market needs ranging from the identiﬁcation of technological needs to the adaptation of technology, customized development, original research and training. Hsinchu Park in Taiwan was formed as a result of the deliberate government industrial policy. Starting with the creation of the Industrial Technology Research Institute in 1973, the government founded Hsinchu Park with the speciﬁc aim of copying the Silicon Valley model. Government initiatives to encourage the institute to obtain technical capability and human resources through Chinese–American ﬁrms in Silcon Valley triggered the very successful development of Taiwan’s semiconductors industry. The presence of transnational corporations in Hsinchu Park, chieﬂy Philips, also assisted in speeding technology transfer along. The Hsinchu Park story again underlines the tremendous untapped resources and unused potential Vietnam has in its large overseas Vietnamese population in California. Conversely, the ﬁnancial services industry in the City of London is an excellent example of how light regulation can advance cluster development. London rose to pre-eminence in international ﬁnancial services in the eighteenth and nineteenth centuries for a number of reasons, including the extent and strength of the British empire, Britain’s strong seafaring tradition and interest in trade, and a low-tax, laissez-faire business environment. In 2003, London is still the world’s largest market for international equities, bonds, commodities and a number of derivative markets. However, all but a few of the major investment banks and brokerage houses operating in London are headquartered there. Most have head ofﬁces in New York, Zurich, Frankfurt or Tokyo. Deutsche Bank, for example, employs more staff in London than in Frankfurt. Financial services are now Britain’s largest export. Much of the London-based trade in securities is conducted by the London operations of different foreign banks dealing in third-country securities. London achieved dominance in the eurodollar markets in the 1970s precisely because regulation was heavy-handed in New York, and the eurodollar markets migrated to London. Even with the advent of the Euro, and the European Central Bank being based in Frankfurt, the largest and dominant share of Euro spot and futures trading takes place in London, currently outside the Euro area.

56 Competition and Cooperation: Vietnam

Special markets often require special conditions. A culture of neutrality and, secrecy, with strong laws protecting privacy and a standofﬁsh regulatory culture are the principal reasons for Switzerland’s position as the world’s leading centre for private banking.

Do clusters exist in Vietnam? The evidence (including a Vietnam literature review) Chapter 1.1 ‘The Rediscovery of Geography’, in Fujita et al.’s (2001) The Spatial Economy, begins with an interesting piece of anecdotal evidence on enterprise location and agglomeration in London: Around the corner from the English National Opera lies St Martin’s Court, a short street occupied mainly by sellers of second-hand books and prints. It is a reasonable location for such shops, but there are no doubt other locations that would serve just as well. Why, then, have the shops’ owners chosen to be there? To be near each other. No doubt there is some interesting story about how that cluster of book and print shops originally became established, but what sustains it now is a sort of circular logic: Potential customers come to St Martin’s Court because they expect to ﬁnd a range of shops to browse in, and shops locate there because they know they will have access to a large pool of potential customers. One could begin with an almost identical paragraph describing the location of economic activity in Hanoi. In fact there are numerous examples of such street-level clusters. The old quarter of Hanoi is famous for the specialization of shops and micro or household enterprises. Many of today’s street specializations date back to the establishment of Hanoi’s merchant guilds in the thirteenth century. Lan Ong Street is famous for herbs and traditional medicine, Lo Ren Street for blacksmiths and metalworking, Hang Gai Street for silk, Hang Bac Street for jewellery, Hang Dau and Cau Go Streets for leatherwear and shoes. There is no speciﬁc advantage accrued by sellers and manufacturers of, say, traditional medicine from being on Lan Ong street. So why are they located there? The answer again is that shops’ owners deliberately choose to be near to each other. Sellers locate in Lan Ong street because that is where customers will come when they are looking for traditional medicine, and customers will come to Lan Ong because they know that is where the sellers are located. This phenomenon pervades almost every economy in the world. Agglomeration – the clustering of economic activity, reinforced by a circular logic occurs also at a multitude of levels, from the Old Quarter of Hanoi serving the Vietnamese capital, to international specialized economic regions, such as Silicon Valley in the USA (software) or the City of London in the UK (ﬁnancial services) that serve global markets.

Adam McCarty, Richard Record and James Riedel 57

However, questions remain as to why some shops or activities cluster while others do not. Why do shops selling for example shoes or electrical goods cluster and agglomerate, while say bakeries or shops selling pho (noodle soup) disperse? Through networking among the various economic ‘actors’, a cluster deepens and accentuates the urban agglomeration effects closely correlated with the productivity of cities. Other things being equal, the beneﬁts of agglomeration are derived from the existence of increasing returns to scale, communication externalities, lower transaction and search costs, and higher returns to human capital as the stock of such capital in an urban location rises. The answer to the pho question is perhaps a result of transaction and search costs. One of the key factors in explaining why some markets cluster and others do not is the level of consumer knowledge with respect to that product market. pho is essentially a commodity good, there is little or no innovation among pho sellers, and consumers know what pho will taste like when they decide that that is what they would like to eat. One does not generally need to ‘shop around’ when purchasing pho. Hence the transactions and search costs rise to prominence; consumers do not want to travel far. Transportation costs (as the Krugman – Venables model shows) are a crucial factor in the formation of all clusters, but for differing reasons. With the greater economic stability inherent in a large economy and the effect of a decline in transport costs, clustering rather than dispersion is further encouraged. Fuijita and Thisse (2002) observe: Low transport costs with sufﬁcient product differentiation push economic agents towards agglomeration. The reason is that product differentiation relaxes price competition and consequently allows ﬁrms to attract more consumers when they are clustered than when any ﬁrm chooses to stand alone. This reveals another argument for dispersion in the market for pho, lack of product differentiation. Without that product differentiation, the only means of competing is through price. Spatial economics and the analysis of clusters are fairly new concepts in Vietnam. Only a limited number of studies have been carried out. Below are details of some cluster analyses that have recently taken place in addition to the 2002 JETRO study. VNCI project The Vietnam Competitiveness Initiative (VNCI) is a pilot project funded by the United States Agency for International Development (USAID) designed to test the cluster competitiveness theories in Vietnam. The project investigated approximately 20 potential industry clusters during 2002. That research led to a decision to offer assistance to 4 clusters in 2 cities: software/ICT clusters in both Hanoi and Ho Chi Minh City, a fruit cluster in

58 Competition and Cooperation: Vietnam

Ho Chi Minh City, and a ﬁne arts ceramics cluster in Hanoi – Bat Trang village. The VNCI then analysed each of the four clusters to identify weaknesses in their competitiveness, formulate strategies, implement initiatives and remove constraints to growth in the local, regional and global economies. Software/ICT cluster (VNCI 2003c) The emergence of high-tech ICT/software clusters in developing countries such as India should act as a considerable incentive to other developing countries. The Bangalore ICT cluster demonstrates that successful knowledge clusters can develop outside Europe, the USA or Japan. Hence there are now excellent opportunities for Vietnam to develop an ICT sector achieve signiﬁcant growth. Unfortunately the VNCI software/ICT cluster strategy found that Vietnam has done little to capitalize on these opportunities. The benchmarking analysis carried out indicated that Vietnam possesses strengths in affordable, high-quality labour, a low geopolitical risk to investment and close proximity to the large growth economies of China and India. However, limitations include comparatively expensive high-bandwidth links, poor English language and business skills, and weak international marketing and promotion. Furthermore there is limited commitment at the national level towards e-government and very weak intellectual property safeguards. Such failings, and without a clear commitment to change, mean that Vietnam is not achieving its potential in this sector. For this to change, there needs to be considerable investment in workforce skills, experience, business management and telecommunication infrastructure. In addition, institutional commitments are needed to improve the business environment for knowledge industries and innovation. Fruit cluster (VNCI 2003b) Vietnamese fruit exports jumped from an average of around US$50m. in the mid 1990s to over US$200m. in 2000 and US$300m. in 2001 on the back of a favourable climate, good access to markets and perhaps most importantly of all – low-cost labour in what is a very labour-intensive industry. However, when exports fell by a third in 2002 the need for improved competitiveness was demonstrated. The VNCI fruit cluster strategy concluded that unsophisticated post-harvest technology, poor cooperation between production elements, a low-quality product, poor infrastructure support, high production costs and a lack of management skills were all hindering sustainable growth in the industry. Hence foreign competition and tougher global standards requirements have done damage to the Vietnamese fruit industry. Vietnam’s current low-quality, low-cost, low-proﬁt-margin fruit strategy is not viable for long-term growth. VNCI developed a strategy to raise competitiveness in the sector through the formation of a fruit sector cluster, increased government support through

Adam McCarty, Richard Record and James Riedel 59

training and improved technology, and development of premium branding for Vietnamese fruit. Ceramics cluster (VNCI 2003a) Vietnam’s ceramics industry is worth some US$110m. per year, of which Bat Trang contributes around 40 per cent. However, this ﬁgure is dwarfed by China – home to about 60 per cent of the world’s output. Bat Trang is a cluster in the traditional sense; it is a village on the outskirts of Hanoi renowned for pottery and ceramics production. Production occurs at the household, micro and small-sized enterprise level, with 83 per cent of families living in the village producing these crafts. Production of ceramics in Bat Trang has been ongoing for 1000 years. Bat Trang could be classiﬁed according to the 2002 JETRO study as a type (1) cluster – one where locally specialized items are produced or a local product district. The top end of the ceramics production in Bat Trang is considered to be among the best in the world, but there is a wide difference between the top- and bottomquality wares produced in the village. Highly skilled artisans are undermined by a village-level reputation for low-quality cheap products. Much of this is a result of almost no product standardization, testing tools, and process knowledge and training. Additionally, Bat Trang producers have very little access to and knowledge of foreign customers’ tastes and requirements. Market signals on popular designs and consumer trends are therefore very slow to reach producers. Again, the existing strategy is to compete on low cost. Bat Trang is an excellent example of a cluster that now ﬁnds itself constrained by the limitations of the informal networks that previously facilitated cluster development. Among other things, the VNCI strategy to institutionalize cluster networks and improve the competitiveness of Bat Trang ceramics highlighted the need to: 1. 2. 3. 4.

improve marketing (both internationally and locally), foster quality improvements, improve the design and R&D processes, and plan more at the village level.

Do clusters exist in Vietnam? A GIS approach In this section, Geographic Information Systems (GIS) software has been used to produce maps and analysis testing for the existence of economic clusters in a selection of sectors on a national, regional and local level in Vietnam. The working hypothesis was, not surprisingly, an expectation of ﬁnding signiﬁcant clustering of industries and subsectors around Hanoi to Haiphong in the north, and all around Ho Chi Minh City in the South. More interestingly, the explanation for such clustering was hypothesized to be more related to access to land and infrastructure, rather than to labour or other variables.

60 Competition and Cooperation: Vietnam

Data The data used for the analysis is a subset of enterprise data taken from the 2001 Vietnam Industry Census collected by the Industry Department of the General Statistical Ofﬁce (GSO), Ministry of Planning and Investment. The Industry Census is comprised of data for every ﬁrm registered centrally under 14 ownership classiﬁcations (see Table 2.11 below for details). The total number of enterprises in the data set comes to approximately 100,000 (see Table 2.2 above for a snapshot of the breakdown of enterprises in Vietnam in 2003). The data set excludes household enterprises and farming households. It also excludes most small and medium-sized enterprises (SMEs), which are registered with the provincial level authorities (Departments of Planning and Investment) and not at the central level. This is reﬂected in the large average number of employees recorded for ﬁrms in the data set (see Table 2.10 below). Obviously it also excludes unregistered ﬁrms, ﬁrms in the informal sector and those newly established since 2001. For this analysis, 12 key manufacturing and processing sectors were selected from the data set on the basis that these were the most interesting for cluster analysis and were most likely to accrue beneﬁts from agglomeration. They were also the most interesting from a trade point of view. The following sectors were chosen: automobiles, cement, electronics, garments, motorcycles, paper, rice, seafood, software, steel, sugar, and textiles, comprising some 4400 enterprises. Within these 12, 13 sectors (garments, seafood and textiles) have been chosen for further analysis. These 3 were chosen because there are enough ﬁrms in Vietnam (and in our data set) to carry out analysis. Furthermore, they are also the heavily exporting sectors where there are the most advantages from knowledge spillovers in areas such as sales and marketing, access to new markets, production and/or processing technology, quality control, management, hygiene standards in the case of seafood processing, and product design in the cases of garments and textiles. Each entry in the database contains the following information: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

company name; address; province code; district code; commune code; telephone number; type of enterprise (see classiﬁcations above); principle sector of activity; number of employees; and annual turnover.

Each province in Vietnam is assigned a standard three-digit code; Hanoi for example is 101, Ho Chi Minh City is 701. See the Appendix for a full map

Adam McCarty, Richard Record and James Riedel 61 Table 2.10 Vietnamese enterprise dataset used for GIS mapping and analysis: labour and turnover Sector

No. of ﬁrms

Total no. of Average employees no. of in the sector employees per ﬁrm

Total annual turnover in the sector (US$m.)

Average annual turnover per ﬁrm (US$m.)

Revenue (turnover) per employee (US$)

Total (12 sectors)

4,402

669,099

152

8,243

1.9

12,320

Automobiles Cement Electronics Garments Motorcycles Paper Rice Seafood Software Steel Sugar Textiles

23 76 106 800 74 513 1,396 472 93 186 93 570

9,574 39,588 14,873 242,552 12,082 40,207 22,022 90,589 1,662 29,405 25,019 141,526

416 521 140 303 163 78 16 192 18 158 269 248

683 808 573 747 773 484 734 1,503 12 589 295 1,042

29.7 10.6 5.4 0.9 10.5 0.9 0.5 3.2 0.1 3.2 3.2 1.8

71,339 20,410 38,526 3,080 63,979 12,038 33,330 16,591 7,220 20,031 11,791 7,363

of Vietnam with province codes. Within each province, each district has a 2-digit code and within each district, each commune has another 2-digit code. Hence each ﬁrm in the data set is assigned a 7-digit code allowing its location to be geographically mapped down to the commune level. Of the total number of ﬁrms in the data set, 1027 (23.3 per cent) are based in Ho Chi Minh City, 345 (7.8 per cent) are based in Hanoi. Table 2.10 summarizes the data set used for this GIS analysis. The relatively large average size of the ﬁrms for the 12 sectors selected here (average of 152 employees and annual turnover of US$1.9m.) underlines the bias towards larger ﬁrms through using centrally collected data. SMEs in these sectors are likely to be registered at the provincial level, if at all. The 3 sectors chosen for detailed GIS analysis (garments, seafood and textiles) employ the largest numbers of people in the 12 sectors. The column displaying revenue (turnover) per employee gives an indication of the relative capital intensity of each industry. The very high ﬁgures recorded for automobile and motorcycle manufacturing (US$71,339 and US$63,979 respectively) demonstrate the large amounts of ﬁxed capital employed in these industries. Table 2.11 shows the same information as in Table 2.10, but disaggregated by ownership type. Even with the central data, the number of private companies (whether private enterprises or two-or-more-member limited-

62

Total annual turnover by ownership type (US$m.)

Average annual turnover per ﬁrm (US$m.)

Revenue (turnover) per employee (US$)

162,974 1,148

1,907

13.4

11,701

Locally owned SOEs

237

128,228

541

1,167

4.9

9,101

Cooperatives

272

14,193

52

58

0.2

4,087

0.3

15,172

1

Centrally owned stateowned Enterprises (SOEs)

2 3

Total no. of employees by ownership type

142

Firm ownership type

No. of ﬁrms

Average no. of employees per ﬁrm

Table 2.11 Vietnamese enterprise dataset used for GIS mapping and analysis: ownership type

4

Private enterprises

1,942

39,941

21

606

5

Partnerships

0

0

0

0

6

One-member limitedliability companies (state-owned)

3

136

45

1

0.3

7,353

7

One-member limitedliability companies (private)

32

2605

81

50

1.6

19,194

8

Two-or-more-member limited-liability companies (state owned)

18

10,074

560

23

1.3

2,283

9

Two-or-more-member limited-liability companies (private)

1,114

128,411

115

1,121

1.0

8,730

Joint-stock companies with state-funded capital

68

42,401

624

321

4.7

7,571

Joint-stock companies with

96

18,100

189

168

1.8

9,282

367

91,404

249

1,030

2.8

11,269

82

25,860

315

1,670

20.4

64,578

29

4,772

165

121

4.2

25,356

1.9

12,320

10 11

0

non-state-funded capital 12

100 % foreign-invested companies

13

Joint ventures between SOEs and foreign-invested enterprises

14

Joint ventures between non-state-owned enterprises and foreign invested enterprises Totals

4,402

669,099

152 8,243

Adam McCarty, Richard Record and James Riedel 63 Table 2.12 List of GIS maps produced of Vietnam No.

Area

Industry

Type

Filenames*

01

National

All 12

Dot density

dotdens_all_lr

02

National

Automobiles

Dot density

dotdens_automobile_lr

03

National

Cement

Dot density

dotdens_cement_lr

04

National

Electronics

Dot density

dotdens_electronics_lr

05

National

Garments

Dot density

dotdens_garment_lr

06

National

Motorcycles

Dot density

dotdens_motorcycle_lr

07

National

Paper

Dot density

dotdens_paper_lr

08

National

Rice

Dot density

dotdens_rice_lr

09

National

Seafood

Dot density

dotdens_seafood_lr

10

National

Software

Dot density

dotdens_software_lr

11

National

Steel

Dot density

dotdens_steel_lr

12

National

Sugar

Dot density

dotdens_sugar_lr

13

National

Textiles

Dot density

dotdens_textile_lr

14

Northern region

Garments

Commune shading

comm._hanoi_garment_lr

15

Northern region

Paper

Commune shading

comm._hanoi_paper_lr

16

Northern region

Steel

Commune shading

comm._hanoi_steel_lr

17

Northern region

Textiles

Commune shading

comm._hanoi_textile_lr

18

Southern region

Garments

Commune shading

comm._hcmc_garment_lr

19

Southern region

Seafood

Commune shading

comm._hcmc_seafood_lr

20

Southern region

Steel

Commune shading

comm._hcmc_steel_lr

21

Southern region

Textiles

Commune shading

comm._hcmc_textile_lr

22

Mekong delta

Seafood

Cluster density

cluster_seafood_lr

*Both high- and low-resolution maps have been produced. The low-resolution maps are refered to here to economize on ﬁle size.

liability companies), dwarfs all other categories including the state sector. However in terms of employment, the SOEs (centrally and locally owned) are still signiﬁcant, accounting for almost half of total employment for this sample. Centrally owned SOEs tend to be much larger than all other ownership types with an average of 1148 employees per ﬁrm. The joint venture sector (between foreign-invested enterprises and SOEs) displays revenue per

64 Competition and Cooperation: Vietnam

employee of almost US$65,000 per year, the highest ﬁgure for this sample. This would be an indication of the high capital intensity of these ﬁrms; virtually the only way for foreign ﬁrms to participate in sensitive heavyindustry sectors such as steel and cement is through joint ventures with a state-owned partner. GIS maps Table 2.12 lists the GIS maps that have been produced for this analysis. They include national dot density maps produced for each of the 12 industries, and then one aggregated map. Commune shading maps have been produced for selected industries, ‘zoomed in’ to the Hanoi–Hai Phong corridor and the Ho Chi Minh City–Binh Duong–Dong Nai triangle, the two principal zones of industrial activity in Vietnam. A further cluster density map has been produced for seafood ﬁrms in the Mekong Delta region. Methodology for cluster mapping Identifying clusters can be done in many different ways. It is for instance possible to look at occurrences of ﬁrms belonging to a speciﬁc sector in an area, such as a province, district or commune. Another approach is carrying out an input–output analysis to study the value added at various intervals of the value chain in vertically integrated clusters, an approach that needs considerable amounts of data. When focusing on the spatial analysis of clusters, cluster analysis comes down basically to identifying spatial patterns. The easiest and most commonly used technique to do this is mapping the occurrences and visually inspecting the map. Quantitative analysis of spatial patterns is a ﬁeld covered by quantitative geography and often involves substantial data processing and complicated statistical techniques. An indicator often mentioned for spatial clustering is Ripley’s K-statistic, which basically compares the pattern of occurrences with a random pattern. If parameters of occurrences are relevant (that is the turnover of ﬁrms) it is possible to carry out spatial regression analysis, which includes routines for analyzing spatial dependence and heterogeneity (similar to, but more complicated than, routines for analyzing temporal dependence, which are standard in any statistical package). Many other methods exist of discovering something about the existence of spatial clusters, though for most methods one needs to deﬁne the areas being analysed (local level, regional level) and a cluster needs to be deﬁned. In the case of ﬁrm location: How many ﬁrms together are required to form a cluster? At what distance is a ﬁrm included in a cluster? How close do ﬁrms need to be for knowledge spillovers to occur, 1 km, 5 km, 10 km? Ultimately these questions are subjective in nature. For statistical analysis, cluster parameters have to be set arbitrarily beforehand. It should be noted that the total number of ﬁrms mapped is just 3940, less than the 4402 ﬁrms in the data set. Since 2001 some province, district

Adam McCarty, Richard Record and James Riedel 65

and commune codes have changed as a result of mergers, boundary changes and splitting of provinces and so on. Hence not all the enterprises in the data set could be matched to the maps. Some 462 ﬁrms have therefore been ‘lost’. Note on the geography of Vietnam Frequent references are made throughout this analysis to provinces, districts and communes. These are the principal administrative boundaries used in Vietnam. Provinces (tinh) are the largest division; there are a total of 61 in Vietnam. Hanoi, Hai Phong, Da Nang and Ho Chi Minh City are cities (thanh pho) with province status, and included in the 61. Each province is divided into districts (huyen in rural areas, quan in urban areas). There are just over 700 districts in Vietnam. Each district is further broken down into communes (xa) in rural areas and wards (phuong) in urban areas. There are some 10,000 communes and wards in Vietnam. Throughout this chapter, ‘commune’ should be taken as including both communes and wards. National GIS maps: Figures 2.4 to 2.16 Conclusions from the national dot density maps The above maps conﬁrm the locational hypotheses, particularly the bipolar development pattern of Vietnam. With the exception of the software enterprises, access to labour did not appear as a likely explanation for most distributions. More important for developing countries, for both import-substituting and export industries, is access to land and infrastructure (leading to urban markets and ports). Maps on a national scale for a country as large as Vietnam have only rather limited value. It is difﬁcult to see more than the broad patterns of economic activity in the country. Nevertheless, some interesting observations can be made. First, and perhaps most obviously, there is a split between the agricultural and commodity production industries (rice, seafood and sugar) and the manufacturing industries (automobiles, cement, electronics, garments, motorcycles, paper, software, steel and textiles). The agricultural industries are much less concentrated than the manufacturing industries. Second, there also appears to be a split between the relatively ‘new’ light manufacturing industries and the ‘older’ heavy industries. Firms in the electronics, garments, motorcycles and software industries seem to be even more closely located to the twin poles of Hanoi and Ho Chi Minh City, than the heavier and older industries such as cement, paper, steel and textiles. The ownership structure of these industries could well be a factor in these cases also. The heavy industries tend to be dominated by the large state-owned enterprises (SOEs; grouped together as general corporations – a

66 Figure 2.4

Dot density map of Vietnam – all 12 industries

One dot is marked for each ﬁrm in the 12 industry sectors. Firms are mapped according to commune. That is, each represents one ﬁrm in a given commune. In order that dots do not ‘stack’ in the same commune and remain hidden from view, they are randomly dispersed within each commune. The map shows the broad pattern of economic activity across Vietnam: industry is polarized in the northern and southern delta regions with small amounts of activity along the coast. Hanoi and Hai Phong can be clearly seen as centres of industrial activity. Ho Chi Minh City and the surrounding provinces show the greatest concentration of dots. The inclusion of rice as an industry sector is something of a distortion, as it appears to give undue emphasis to industry in the Mekong Delta, the traditional ‘rice bowl’ of Vietnam. Some small clustering can be seen in the provincial capitals along the coast.

67 Figure 2.5

Dot density map of Vietnam – automobiles

This shows only the location of the 19 automobile ﬁrms in Vietnam that could be mapped. Very few conclusions can be drawn from such a map, given that the data set is so small. The 19 include the various component suppliers within the automotive industry supply chain. There are three areas where these ﬁrms are located: Hanoi, Hai Phong and Ho Chi Minh City. Automobile assembly takes place on only a very limited scale in Vietnam. Transnational corporations with a presence in Vietnam include GM Daewoo, Daihatsu, Toyota, Ford, Isuzu and Daimler Benz.

68 Figure 2.6

Dot density map of Vietnam – cement

The locations of the 66 cement ﬁrms are mapped by commune location. Most striking is the relative dispersion of cement plants. The production of cement requires large amounts of (heavy) limestone, mined principally in the northern mountainous areas of Vietnam. High transport costs, as the Krugman–Venables model suggests, means that ﬁrms are likely to choose to locate closer to the market. Hence there is a tendency for plants to be located close to supplies of raw materials and serve the local market in that vicinity. There does appear, however, to be some evidence of concentration in the Hai Phong–Hai Duong area. Hai Phong has traditionally been a centre for cement production; the ﬁrst factory was established in Hai Phong by the French colonial rulers of Vietnam in 1899.

69 Figure 2.7

Dot density map of Vietnam – electronics

This shows the location of 96 electronics ﬁrms. Virtually all the ﬁrms are clustered very tightly in and around either Hanoi or Ho Chi Minh City. This is perhaps unsurprising given that the electronics industry in Vietnam is still rather undeveloped. Electronics ﬁrms tend to require large amounts of semi-skilled labour, of the sort prevalent in the major cities. Consumer electronics ﬁrms in Vietnam producing for export will most likely choose to locate close to transport infrastructure. Firms producing for the domestic market will also most likely choose to locate close to the major cities, given that in general it is primarily the comparatively wealthy urban population that can afford to purchase electrical goods.

70 Figure 2.8

Dot density map of Vietnam – garments

This shows the location of 721 garments ﬁrms. The garments industry is comparatively new to Vietnam. It has experienced rapid growth in the last ten years and is now Vietnam’s largest export sector after oil, accounting for some 16.4 percent of exports in 2002 (STAR 2003). The industry has generated substantial employment and much of the export growth has been as a result of foreign direct investment from other Asian countries. The garments industry is a historically ‘light-footed’ industry, moving from country to country. This is largely due to the industry’s limited working capital requirements and large unskilled labour requirements. The strongest concentration of ﬁrms appears to be in and around Ho Chi Minh City. There also appear to be clusters in Hanoi and Hai Phong. Clustering in this industry is analysed in greater depth further on in this chapter

71 Figure 2.9

Dot density map of Vietnam – motorcycles

This shows the locations of the 68 motorcycle industry ﬁrms. There appears to be high concentration in Ho Chi Minh City and Dong Nai in the south, and in Hanoi and Hai Phong in the north; serving the principle population centres. The motorcycle industry in Vietnam produces almost exclusively for the domestic market and is characterized for the most part by foreign invested or joint venture assembly plants with local components suppliers clustered around the major plants. For motorcycle production, Vietnam has mandatory ‘local content’ requirements. Premium motorcycles are imported already assembled.

72 Figure 2.10

Dot density map of Vietnam – paper

This shows the locations of 429 paper ﬁrms. Ho Chi Minh City and Hanoi appear to have the largest concentrations of ﬁrms, with secondary clusters in Da Nang and Hai Phong.

type of super SOE holding company), with foreign investors allowed to participate in these sectors only through joint ventures with these SOEs. Decisions on plant location may be constrained by non-economic factors such as local patronage, efforts to develop the regions by providing jobs in deprived areas, tradition or pure inertia. Private enterprise, smaller ﬁrms and 100-per-cent foreign-invested enterprises account for a much larger

73 Figure 2.11

Dot density map of Vietnam – rice

This shows the location of 1322 rice ﬁrms. This is the sector (of the 12 analysed) with the largest number of ﬁrms. There is tremendous evidence of clustering in the Mekong Delta region. Obviously rice is grown best in places where there is ﬂat land, good irrigation and water supply, and good soil quality. The Mekong Delta has been Vietnam’s ‘rice bowl’ for generations and is the area that has driven rice production up in the last 15 years to make Vietnam the world’s second-largest rice exporter. Rice now counts for some 4.4 percent of national exports (STAR 2003). Looking at this map one might assume erroneously that rice is grown only in the Mekong Delta. In fact it is grown nationally but mostly on a household scale. Only in the Mekong are there sufﬁcient production surpluses for rice growing to become an industry. Hence in this region there exist a large number of ﬁrms registered at the central level, thus skewing the map somewhat.

74 Figure 2.12

Dot density map of Vietnam – seafood

This shows the location of 414 seafood ﬁrms. It is not surprising at all that these ﬁrms are located along the coastal zones and in the Mekong Delta region. The largest single grouping appears to be in Ho Chi Minh City, with a secondary cluster in Da Nang. The seafood processing industry is another industry that has experienced lively export growth in the last 10 years. In 1995, Vietnam’s seafood exports were less than rice or coffee exports. In 2002, seafood accounted for 12.2 percent of total exports, more than coffee and rice put together. Clustering in this industry is analysed in greater depth further on in this chapter.

75 Figure 2.13

Dot density map of Vietnam – software

This shows the location of 92 software ﬁrms. The software industry in Vietnam is still very underdeveloped and plays only a limited role in the national economy. As one would expect, software ﬁrms require highly skilled labour of the sort that can only be found in the major urban centres. Hence software ﬁrms are located almost exclusively in either Hanoi or Ho Chi Minh City.

share of activity in light manufacturing. Choices on plant location may therefore be freer and made more frequently on pure economic grounds. The following section examines speciﬁc industry clusters on a closer scale in the two dominaant industrial areas of Vietnam: the Hanoi–Hai Phong corridor and the Ho Chi Minh City–Binh Duong–Dong Nai triangle.

76 Figure 2.14

Dot density map of Vietnam – steel

This shows the location of 156 steel ﬁrms. Steel production appears to be concentrated in Ho Chi Minh City in the south, Da Nang in the central region, and Hanoi and Hai Phong in the north. There also appears to be a steel cluster in Thai Nguyen, north of Hanoi. In common with the cement industry, steel is protected by high tariff barriers in Vietnam. In the ﬁrst years following on from Vietnam’s ‘opening up’, there was a large amount of foreign investment into the steel sector, principally through joint ventures. These investments were made almost exclusively for the purpose of producing for the local market, where prices have tended to be higher than in other regional economies. Import substitution of this kind has had something of a distorting effect, challenging ﬁnite resources of capital into industries where Vietnam does not exhibit any comparative advantage. Steel is an interesting case in that high transportation costs are likely to be a driver towards smaller-scale regional production close to enduse markets. However, this factor is undoubtedly outweighed by the enormous economies of scale gained from steel melting, thereby driving concentration.

77 Figure 2.15

Dot density map of Vietnam – sugar

This shows the locations of 88 sugar-producing ﬁrms. As one would expect for agricultural production, ﬁrms are dispersed most likely according to where sugar cane is grown. There is very little evidence of concentration, although slightly higher ﬁrm densities can be seen in the provinces of the Mekong Delta and in particular in An Giang.

78 Figure 2.16

Dot density map of Vietnam – textiles

This shows the location of 469 textile ﬁrms. Again, Ho Chi Minh City and Hanoi have the largest concentration of ﬁrms, but ﬁrms are not conﬁned exclusively to these cities, large numbers being located in the surrounding provinces. Given that textile ﬁrms are part of the same value chain as garment ﬁrms, it is not surprising that the location patterns for the two industries are quite similar. However, there do appear to be slightly more textile ﬁrms in the rural provinces than garment ﬁrms. Vietnam has a much longer history of producing textiles (previously for the CMEA markets), than garments. Hence the state plays a greater role in the textiles industry, and this may explain the relative dispersion, as plants may have been located in given regions for non-economic reasons. Clustering in this industry is analysed in greater depth further on in this chapter.

79

Northern and southern region GIS sector maps: Figures 2.17 to 2.25 Figure 2.17

Vietnam northern region commune-shading map – garments

This is a commune-shading map for garments ﬁrms in the region surrounding Hanoi. The insert in the top right-hand corner is the zoomed-in view for the centre of Hanoi. There do not appear to be any striking indications of industrial clustering in this map, although the highest densities of garments ﬁrms can be seen in Ba Dinh and Dong Da districts of central Hanoi. As perhaps one would expect, ﬁrms tend to be located along the major road arteries. The largest number of ﬁrms outside the city centre of Hanoi seems to be located along National Highway 5 running east from Hanoi towards Hai Phong.

80 Figure 2.18

Vietnam northern region commune-shading map – paper

This shows the location of paper ﬁrms according to number of ﬁrms per commune in and around Hanoi. Interestingly, the locations of paper ﬁrms in central Hanoi appear to be in communes other than those containing the garment ﬁrms as shown in northern and southern region GIS sector maps. There does seem to be some evidence of clustering along the continuation of National Highway 1A going northeast out of the centre of Hanoi, in Gia Lam district. Hai Ba Trung district in central Hanoi appears to have a higher than average density of paper ﬁrms. Finally Hai Phong also has evidence to some extent of a paper cluster.

81 Figure 2.19

Vietnam northern region commune-shading map – seafood

Here are shown the locations of seafood ﬁrms in the north of Vietnam. The most striking feature of this map is actually the apparent lack of seafood ﬁrms in the region. There are a very small number of ﬁrms in Hanoi. Only two small clusters can be seen in the region, and neither is particularly signiﬁcant, one in Hai Phong and another further down the coast in Thai Binh province. It is, however, interesting to compare this map with Figure 2.23 showing seafood sector enterprise location in the south of Vietnam.

82 Figure 2.20

Vietnam northern region commune-shading map – steel

This shows the location of steel ﬁrms. Given the large amounts of pollution generated by steel plants, it is hardly surprising that there are almost no ﬁrms located in the centre of Hanoi. Steel plants require very large amounts of ﬁxed capital and hence are generally few in number, but large in size. There is a small steel cluster northeast of Hanoi in Gia Lam district. However, the most striking clusters can be seen in Hai Phong. Three communes located adjacent to river ways in Hai Phong City show quite profound evidence of steel clustering. Two communes in particular (in Hong Bang district on the south side of the Cam river and in Thuy Nguyen district on the north side) show signiﬁcant evidence of high density clustering with more than ﬁve steel ﬁrms per commune. It seems highly likely that proximity to river and seaways for transportation are signiﬁcant in the steel industry.

83 Figure 2.21

Vietnam northern region commune-shading map – textiles

While Hanoi and Hai Phong have a number of textile ﬁrms, the communes with the highest densities (ﬁve or more ﬁrms per commune) can be seen outside the two major urban areas. One such cluster can be seen on the southwest outskirts of Hanoi, just inside Ha Tay province, close to Ha Dong town. Travelling south approximately 40 km an additional cluster can be seen in Van Dinh town, still in Ha Tay. Southeast from there a cluster can be seen at the major road intersection at Phu Ly town, Ha Nam province. East of that point (in the bottom right section of the map), more clusters can be seen in Nam Dinh and Thai Binh provinces. It is remarkable that there are three high-density areas of textile ﬁrms in Thai Binh province; one of these is located actually in Thai Binh town itself, but the remaining two clusters are in communes in comparatively rural districts.

84 Figure 2.22

Vietnam southern region commune-shading map – garments

This shows how large the garment industry is in Ho Chi Minh City and the surrounding areas. The critical mass of garments ﬁrms appears to be in the western districts of the city, namely Binh Chanh, Tan Binh, Go Vap, Hoc Mon and District 12. Almost every commune in these districts has garments ﬁrms, and several have ﬁve or more ﬁrms. An interesting commune with a high density of garment ﬁrms can be seen south of the city centre in District 7, in the loop of the Saigon river. Garments ﬁrms are located in all directions leading out of Ho Chi Minh City, but there are two noticeable areas to the northwest and northeast of the city with secondary garments clusters. These two areas respectively are Cu Chi district of HCMC, and Bien Hoa town, Dong Nai province.

85 Figure 2.23

Vietnam southern region commune-shading map – seafood

This map demonstrates clearly how much larger the seafood industry is in the southern region compared with the north. The locations of seafood enterprises actually in Ho Chi Minh City are located in broadly the same industrial dstricts as garment ﬁrms, but there appears to be a small bias somewhat towards the southwest. This may be a reﬂection of the role the Mekong Delta (southwest of HCMC) plays in the industry. Leading southwest about 20 km out of the city along National Highway 1A, an additional cluster can be seen in Ben Luc town, Long An province. Further to the southwest, signiﬁcant evidence of seafood clustering can be seen in My Tho, one of the larger towns in the Mekong Delta and capital of Thien Giang province. Further investigations of seafood clustering in the Mekong Delta can be seen later on in this chapter.

86 Figure 2.24

Vietnam southern region commune-shading map – steel

This shows the locations of steel ﬁrms in and around Ho Chi Minh City. Activity in the steel industry appears to be conﬁned to four main areas, three in HCMC in Tan Binh and Binh Chanh districts and another southwest near Ben Luc town in Long An province. Ho Chi Minh City does not appear to have the same steel sector highdensity clustering exhibited in Hai Phong, in the north of Vietnam (see Figure 2.20).

87 Figure 2.25

Vietnam southern region commune-shading map – textiles

The distribution of textiles ﬁrms seems to follow a similar pattern to that of garment ﬁrms in the south (see Figure 2.22), with a concentration in the west of the city and secondary clusters to the north. This would make sense given that textiles and garments ﬁrms manufacture as part of the same supply chain. High-density communes can be seen again in Binh Chanh district, Tan Binh district, District 7 and District 8. Secondary clusters can be see to the north of the city in three locations. Similar to the garments ﬁrms, there are secondary clusters in Cu Chi (northwest) and Bien Hoa (northeast). There is a large secondary cluster directly north of HCMC in Thu Dau Mot town, Binh Duong province. One other cluster can be seen due west of the city centre on the far side of the Nha Be river in an apparently rural commune in Dong Nai province, close to National Highway 51 running towards Vung Tau.

88 Competition and Cooperation: Vietnam

Seafood enterprises in the Mekong Delta We have used density maps to analyse clusters in the Mekong Delta region, where seafood ﬁrms are shown as a circle (see Figure 2.26). The shade of the circle denotes the density of ﬁrms per unit area surrounding that location. A high density of ﬁrms indicates a cluster. The parameters that can be changed are the area that should be taken into account, represented by r, the radius of a circle surrounding a location and the ﬁeld on which the density is based (count, employment, turnover and so on). In addition the

Figure 2.26

Seafood clusters in the Mekong Delta

Adam McCarty, Richard Record and James Riedel 89

way the density is computed can be changed, resulting in more or less ‘smoothed’ maps. The interpretation of the ﬁgures on the density map is as follows (Tim Ormsby, ESRI knowledge base: http://www.esri.com): When you calculate a density grid from a point theme, you specify a search radius. The radius deﬁnes a circle, which is applied to each grid cell in turn as it is processed (the circle is centred on each grid cell). All points from the point theme that fall within the circle are counted and summed. The sum is then divided by the area of the circle. The result is the grid cell’s density value, in square units of measure. For instance, suppose you specify a search radius of 0.5 miles. Suppose also that for a particular grid cell, three points (car thefts or whatever) fall within its 0.5-mile-radius circle. The density value for that cell, in square miles, is three divided by the area of the circle (3 divided by πr squared). πr squared in this case is ((0.5 × 0.5) × 3.14) or 0.785. The density value for the grid cell is therefore 3.82 (3/0.785). So what does 3.82 mean? Not 3.82 thefts in that grid cell. It means approximately: If you marked off any square mile that included that grid cell, you would probably ﬁnd that four auto thefts had occurred in that square mile. If you marked off a bunch of different square miles that included that grid cell, some of them would have only 2 or 3 car thefts occurring in them. Some would have 5 or 6. The average number would be 3.82. ‘Any square mile that includes that grid cell’ is not a strictly accurate formulation (it really depends on the size of the search radius), but it makes it easier to grasp the idea. The point is that the density value reﬂects the occurrences in a circle drawn around each cell, not the occurrences within the cell itself. To avoid this lengthy explanation you can change the map categorization to broad generalized categories: low clustering–medium clustering–high clustering, as has been done in the seafood maps above. Interpretation of maps Three maps can be seen analyzing seafood clusters in the Mekong Delta and surrounding region. Though the ‘measurement for occurrence’ (count of ﬁrms, sum of employment, or sum of turnover per community) has been varied, the general cluster pattern remains unchanged for the most part. In the seafood maps, on looking at the number of ﬁrms, a cluster can be found in HCMC and on Phu Quoc Island. However, if we look at the turnover or number of employees Phu Quoc disappears as a cluster and Soc Trang and Ca Mau appear in the Mekong Delta beside HCMC. If we looking at the number of ﬁrms, however, these large ‘clusters’ are formed by only a few ﬁrms, so obviously these are very large ﬁrms and it is some-

90 Competition and Cooperation: Vietnam

what difﬁcult to really speak of there being clusters. We may conclude that we need to take several issues into account when deciding what is a cluster. Perhaps low clustering (in the seafood maps) and low values in the other maps should actually be interpreted as ‘no clustering’, as small clusters might be generated by a single large ﬁrm. The question that arises again here is: How many ﬁrms do we need to speak of a cluster? For example, a large seafood-processing ﬁrm may be the largest and principal seafood-processing ﬁrm in a province. The location of that ﬁrm, in terms of turnover per community or sum of employment, may register as an area with comparatively high industrial density. One ﬁrm should surely not be a cluster. However, the single large ﬁrm may support a large number of suppliers and intermediate ﬁrms in the value chain. Such ﬁrms may be registered at the provincial level, or not at all, in which case they would be missing from our data. This would ﬁt with the type (2) cluster identiﬁed in the 2002 JETRO study: one where a large core ﬁrm has many subcontracting or parts makers surrounding it or the so-called industrial castle town. Similarly, Phu Quoc island, a cluster in terms of the large number of (small) ﬁrms only, could be a cluster following type (1) of the 2002 JETRO study: one where locally specialized items are produced or local product districts. Cluster analysis The cluster index (Table 2.13) was computed as follows: 1. First, density maps were generated. The density was computed for 500 by 500-meter grid cells for the total area of Vietnam, using a search radius of 5 kilometres. This gives for each cell a density per square kilometre, based on an area of πr squared (= 78.5 km2). The distance of 5 km can be interpreted as a small region: the longest distance between two ﬁrms to be considered in the same cluster is in this case 10 km. (Note: that smaller values for r do not most likely substantially change the outcomes of these computations; a smaller r will result in higher density values for sectors with ﬁrms that are located nearby and lower density values for sectors with larger distances between ﬁrms.) 2. Second, each grid cell (with a density value) has been classiﬁed in the categories 0, 0.25–0.50, 0.50–0.75 and so on. After the reclassiﬁcation for each category the number of grid cells was computed. 3. As the density is affected by the number of ﬁrms in a sector, the values were weighted (divided) by the total number of ﬁrms. 4. Finally, for clearer ﬁgures the weighted clustering numbers are computed as a fraction of the clustering of all ﬁrms together. In this way it is possible to compare the relative clustering of sectors.

Table 2.13

Cluster index of Vietnam – all 12 industries

Classes

0.25–0.50

0.50–0.75

0.75–1.00

1.00–1.25

1.25–1.50

1.50–2.00

2.00–3.00

3.00–5.00

5.00–10.00

Count

1.000

1.000

1.000

1.000

1.000

1.000

1.000

1.000

1.000

Textiles Garments Seafood Cement Rice Steel Sugar Software Paper Motorcycles Automobiles Electronics

2.441 1.656 2.890 8.534 1.961 3.443 6.262 1.098 2.248 3.891 6.288 2.464

0.825 0.556 0.503 – 0.805 0.255 – 2.910 1.256 – – 1.291

0.943 1.212 0.300 – 0.911 – – 1.967 1.892 – – –

1.427 1.822 3.123 – 0.644 – – – 1.182 – – –

0.305 1.569 – – 0.028 – – – – – – –

– 3.415 – – – – – – – – – –

– 2.161 – – – – – – – – – –

– – – – – – – – – – – –

– – – – – – – – – – – –

Source: Calculated from the industry dataset.

91

92 Competition and Cooperation: Vietnam

Conclusions The garment sector has the greatest tendency to cluster in our sample, followed by textiles, rice, seafood and paper. On visual inspection of the maps, this seems quite reasonable. The cluster index is, however, still quite considerably affected by the number of ﬁrms, so one has to be careful in drawing strong conclusions, especially when comparing different sectors where there are large variances in the numbers of ﬁrms included in the data set. The fact that there are more ﬁrms may be shown as an indication of clustering.

The role of business associations in facilitating cluster formation and development Clustering can act as a signiﬁcant driver for the formation and development of business alliances and organizations. The presence of many different producers in one place facilitates the establishment of formal associations. Such associations can assist by way of advice or information, or they may offer regulatory services and help industries develop codes of conduct, establish set business practices, or arbitrate in commercial disputes. In addition, clustering promotes the development of distinctive business cultures and etiquette in particular places, thus helping the constituent ﬁrms to communicate with and understand one another. Business associations can be the institutional glue that binds ﬁrms within a cluster together, through lobbying, providing business support services, gathering statistics, hosting events and interfacing with government authorities. They can help institutionalize cluster linkages. There are approximately two hundred business associations in Vietnam. They range from the large and highly inﬂuential Vietnam Chamber of Commerce and Industry (VCCI) to sector-speciﬁc associations such as the Vietnam Leather and Footwear Association (LEFASO), right down to smaller regional ‘business clubs’, cooperative alliances and ‘young entrepreneurs’ clubs’. Such associations actually provide rather less services to their members than their respective equals in other countries. But this can hardly be unexpected given the historical underdevelopment of the private sector in Vietnam. A survey of enterprises and business associations in 2002 by the Mekong Project Development Facility (MPDF 2002) found that business associations in Vietnam performed fairly well in the following areas: 1. providing information on policies and legal issues to members; 2. making policy and legal recommendations on behalf of their members; and 3. providing or organizing short-term training courses.

Adam McCarty, Richard Record and James Riedel 93

However, the survey also revealed that business associations were considered to be still relatively weak at: 1. providing technical advice and consulting; 2. providing information on markets; and 3. providing assistance in assessing both domestic and export markets. The constraints inhibiting Vietnamese business associations from providing high-quality professional services highlighted by the survey included: a general lack of resources, a lack of capacity and lack of professional staff, and the lack of a clear legal framework for the activities of business associations. Such limitations are in many ways a result of Vietnamese associations’ poor ability to raise funds through membership dues, collection of service fees and fundraising. Reform in this area was one of the principal recommendations of the study. Other recommendations included the need to ‘professionalize’ activities and services, and the need for the Government to streamline the legal and regulatory environmental with respect to business associations. Describing the Silicon Valley and Route 128 (Massachusetts) high-tech clusters in the USA, Saxenian (1994) wrote: The nature of business culture – particularly the openness to new ideas, the ﬂexibility of business arrangements permitting easy entry of ﬁrms into an industry, and the richness of social capital – sets the pace of technological advance and the emergence of business models relevant to the times. ‘Social capital,’ although a somewhat unfamiliar term in Vietnam, is actually very much in evidence. Vietnamese mass organizations have traditionally played a strong role in both society and business. Grassroots organizations such as the Women’s Union and the strong tradition of cooperatives and cooperative alliances all contribute to the ‘richness’ of Vietnam’s social capital. The fabric of these organizations represents tremendous potential for Vietnam to also adopt a dynamic business culture of the kind described by Saxenian. In Vietnam, business associations could ﬁll an important gap in helping ﬁrms to enter new export markets. The global marketplace is a relatively new place for Vietnamese ﬁrms to compete. Vietnamese ﬁrms have expanded rapidly into new export markets in East Asia (including Japan), the European Union and most recently the USA (following the signing of the US–Vietnam BTA accord in 2000). However, the ‘success’ of the various product groups’ export boom has been somewhat mixed. The manifold ‘success’ in many sectors of Vietnamese manufacturing exports to the USA (garments exports to the USA rose by some 1700 per cent between 2001 and

94 Competition and Cooperation: Vietnam

2002; STAR 2003) has to be contrasted with the apparent ‘failure’ of coffee and catﬁsh, to name two examples. These two product sectors’ export story can be described as failures in the sense that with coffee, apparent naivety with respect to world markets caused Vietnam to collapse world prices and ﬁnd that even as export volumes rose, revenues fell. Oversupply and lower prices than anticipated have resulted in signiﬁcant difﬁculties for large numbers of coffee farmers in the central highlands. The ‘small country’ assumption and horizontal supply curve did not apply to Vietnam in this case. Clearly there could have been a role for better and ‘smarter’ industry regulation to anticipate Vietnam’s impact on world markets. Similarly Vietnam is largely wedded to the production of robosta, the coffee bean of lowest value. More strategic application of industry regulation could drive up quality and help farmers move to higher-yielding products. Equally in the case of catﬁsh, an extremely rapid expansion of export quality production channelled to the US market prompted the US Department of Commerce in 2003 to impose punitive restrictions on Vietnamese catﬁsh, ban retailers from marketing Vietnamese catﬁsh as ‘catﬁsh’ – forcing them to market the produce (in Vietnamese) as basa or tra, and label Vietnam a ‘non-market’ economy. The anti-dumping charges were based largely on very weak economics, and dumping has clearly not taken place; the reality is that production (principally labour) costs are several multiples lower in the Mekong Delta than in the southern USA. However, again, a better understanding of world markets and the ‘intricacies’ of trade policy could have helped Vietnam avoid such a situation. While the size of the Vietnamese catﬁsh export market is in actual fact very small, the impact of the dispute was much larger in terms of the coverage it received in the media and among policy-makers. A stronger, smarter and more savvy business association might have advanced the interests of the catﬁsh industry cluster and lobbied more effectively against the punitive tariffs. Informal actions and institutions played a key role in the rapid expansion of household and micro businesses in Vietnam during the early Doi Moi period, and have underpinned much of the recent employment growth in Vietnam. This is not unique to Vietnam. In many developing market economies, especially in Asia, transactions are conducted with limited reliance on formal contracts and a binding enforcement mechanism. Instead, parties to a contract rely on relationships to enforce or adjust commitments, and/or to address non-performance of commitments. Penalties for non-performance are linked to reputation and opportunities to engage in future business. However, as ﬁrms grow the limitations of informal institutions include: 1. The need to deal with ‘outsiders’ and ‘foreigners’ such as bankers and traders increases as the scale of business grows – either party may doubt that informal mechanisms will result in transparent and equitable outcomes.

Adam McCarty, Richard Record and James Riedel 95

2. Opportunities to enter contracts depend on established relationships based on past business contacts. Thus, informal institutions have a bias against new entrants with negative consequences for equity and efﬁciency. 3. The formal ﬁnancial sector often requires formal contractual arrangements to reduce transaction costs, and risks, in assessing and enforcing debt repayment. 4. Informal arrangements can breakdown during general economic downturns. Borrowers continue to operate loss-making businesses longer than they should, losses increase, and ﬁnancial problems can spread to suppliers and creditors. These limitations are particularly pronounced in Vietnam because business networks – and informal ‘rules of the game’ – that played a pivotal role in ﬁnancing private investment elsewhere in East Asia are less developed in Vietnam. Such limitations could well be mitigated through increased activity by business associations in helping to ‘formalize’ the activities of businesses. Clusters could well be the basis upon which groups of small businesses enter into more formal contractual arrangements with customers, suppliers, partners and ﬁnancing institutions (especially those abroad), while still maintaining the ‘insider’ status of ﬁrms within the cluster. Relying on informal contacts and instructions limits dramatically the pool of potential business partners. Failing to grow beyond informal networks is one of the reasons why household enterprises fail to grow into successful small and then medium-sized enterprises.

The role of IZs and EPZs in facilitating cluster formation and development Increased factor mobility throughout the East Asian region has played a critical role in stimulating the development of clusters in several regional economies. One of the implications of increased factor mobility is the now heightened competition for inward foreign direct investment (FDI). The development of EPZs and IZs has been an important tool in many developing countries, not least China, in attracting such investment. If the attractiveness of a site to any ﬁrm seeking to invest is dependent somewhat on the presence of other ﬁrms, then polices that encourage investment may be a crucial factor in facilitating the development of clusters. The key conclusion is that today’s factor mobility does not guarantee that economic growth will take place in a geographically balanced manner. Overview of foreign direct investment in Vietnam Vietnam’s FDI inﬂows have fallen signiﬁcantly since 1996. The poor state of Vietnam’s trade infrastructure is no doubt a contributing factor to this

96 Competition and Cooperation: Vietnam

(see Table 2.4 above), but in reality Vietnam has gone through something of an FDI ‘boom and bust’; as the world’s newest market opened up, too much FDI ﬂooded in in the mid 1990s. Many changes have been made to the FDI business environment to try and encourage and facilitate further investment. Changes included adjustments to foreign investment laws and substantial alterations in the domestic business environment. These changes have ensured an increased role for and strengthening of the private sector, highlighted by the new Enterprise Law in 2000, which has made it easier to establish and register a private enterprise with an assurance of equal recognition by the government. Trade agreements also continued to be signed, most notably the US BTA in 2002. However, despite these substantial changes for the better, and an improvement in foreign investor sentiment, FDI has not returned to the levels seen in previous years and appeared in the ﬁrst half of 2002 to be in substantial decline, some 55 per cent below the previous year’s ﬁgures (Freeman 2001). In many ways these changes have simply mirrored a general liberalization throughout Asia as well in other developing countries and reﬂect a level of freedom many investors feel is simply a minimum to secure their investments. Though Vietnam’s legal system may be more in tune with FDI needs, other requirements such as local skills, hard and soft infrastructure, services, supply networks, institutions and property rights are not. Equally, Vietnam’s general business environment remains hindered by bureaucracy and red tape, a major continuing complaint of inward investors. On top of this, FDI itself has substantially changed. Previous ‘greenﬁeld’ FDI activity, which was a large source of FDI in Vietnam in the past, where a company comes and establishes operations, has been in decline around the world over the last few years. In its place has grown increased merger and acquisition activity, a source of investment Vietnam has found hard to attract owing to its restrictive enterprise and share ownership laws. On the whole, given similar FDI policies in other developing countries, especially in Asia, Vietnam should perhaps consider more radical changes in its current foreign investment strategy. This will be necessary if Vietnam is to achieve the US$12–16bn target it has set itself for foreign private capital inﬂows between 2001 and 2005. Vietnam needs a more considered list of targets, showing an understanding of where the country has the ability to make best use of its comparative advantage and where it requires not only investment but know-how and the other attributes that FDI might bring. In terms of Vietnam’s position in the regional and global value chain for transnational corporations (TNCs), recent studies have highlighted the division of high-value elements of the chain such as research and development to Japan, sales and marketing to Singapore and production and assembly to second- and third-tier Asian countries such as the Philippines and Vietnam (Giroud 2002). Not only in leather and footwear (as discussed

Adam McCarty, Richard Record and James Riedel 97

in the previous chapter) is Vietnam limited to a small, low value added section of the value chain. Regional integration, measured by the level of input sourcing and production exporting to ASEAN, appears to be of a minimal level, with integration of TNC and FDI production at more of a global than a regional level. Data so far indicate that as little as 5 per cent of exports and inputs are sent to and received from other ASEAN countries (Giroud 2002). While ASEAN investors are much in evidence in Vietnam, export processing is more often than not destined for the EU, Japan and now increasingly the USA. The ASEAN Investment Area (AIA) aims to raise this and promote greater levels of regional integration. Domestic linkages also appear to be at a basic level, with foreign producers sourcing mainly low-technology inputs from ASEAN local suppliers. This is despite the view within TNCs that local suppliers have improved many areas of performance, such as quality and delivery efﬁciency, but are still considered to have poor technological capabilities. The Giroud survey also found that TNCs in Vietnam rate the level of their employees’ competency and know-how very highly, with many TNCs complaining that they suffered from human capital ‘bottlenecks’ with the supply of human capital not keeping up with ﬁrms’ requirements. This development of a skilled workforce via foreign investors’ transfer of knowledge can have beneﬁcial knock-on effects for local ﬁrms as they employ and learn from TNC staff. The same survey found that the presence of TNCs has not hindered the development of locally based industries in the sectors where FDI is prevalent. Many domestic ﬁrms are able to compete with foreign investors in the local market and are ‘poaching’ trained staff to help develop their businesses. This would imply some sector-wide beneﬁcial effects from agglomeration; local ﬁrms are strengthened by the presence of transnational investors, rather than supplanted by them. The Giroud Vietnam survey ﬁndings illustrate an interesting meso picture. While on the one hand FDI may not bring about a great degree of regional integration, linkages may remain low-ended unless measures are taken to improve technical know-how of suppliers as well as ensuring quality. Human resource development and competitive effects do appear to provide more non-tangible beneﬁts in addition to par dollar value of foreign direct investments by TNCs. The question is how Vietnam can ensure its ﬁrms take full advantage of the development of an FDI workforce and capitalize on these spillover effects to help industrial clusters develop further. IZs and EPZs Government-led industrial parks and estates are the fourth type of cluster identiﬁed in the 2002 JETRO study, and are the source of much economic activity and evidence of industrial clustering in Vietnam. The example of

98 Competition and Cooperation: Vietnam

the ﬁrst wave of economic development in East Asia (the ‘Asian tigers’) shows how modern technologies, skills and management practices can be assimilated through FDI and the presence of TNCs at the hub of special economic zones. In a developing country such as Vietnam, where public goods (infrastructure, utilities and so on) are not provided universally to a high standard, export-processing zones (EPZs), industrial zones (IZs) and industrial parks (IPs) can ‘ﬁll the gap’ for foreign investors by providing an area where such facilities are provided to a high standard. Companies may choose to produce within an EPZ to take advantage of exemptions from customs duties for equipment, raw materials and commodities imported into the zones, and for ﬁnished goods and products exported from the zones, subject to speciﬁc provisions regulating EPZs. All the production within an EPZ must be exported. IZs have been developed to offer tax advantages for establishing factories within the zones. Companies can produce within an IZ for the domestic market or for export. The companies pay no duties when importing raw materials, if the end products are exported. The ﬁrst EPZ was established in Vietnam in 1991. According to ﬁgures published in 2000 by the US Department of State, Vietnam has 63 IZs and 3 EPZs. As of the same time there were 914 licensed enterprises in the zones with total registered capital of US$7.8bn, including 569 foreign enterprises with US$6.4 bn in registered capital. About 40 per cent of the registered capital was disbursed at that time. In 2000, the IZs and EPZs attracted 262 projects with US$900m. registered capital. In 2000 the IZs and EPZs accounted for 25 per cent of gross industrial output and 16 per cent of total export value. Investors opting to locate in such areas also receive tax incentives and preferential treatment. Projects located in EPZs and IZs enjoy the following investment incentives, among others: 1. exemption from export tax on ﬁnished products exported directly to foreign markets; 2. lower production tax rates; 3. exemption from tax on proﬁts for the ﬁrst few years of activity; 4. the ability to carry forward cumulative losses for several years to offset against proﬁts; and 5. Low tax rates on proﬁts remitted abroad. In general the incentives are more generous for ﬁrms to locate in EPZs, which focus almost exclusively on exporting, than in IZs, where ﬁrms produce for both domestic and export markets. The procedures for the granting of investment licenses can be somewhat simpler if investing through an EPZ or IZ. EPZ and IZ management com-

Adam McCarty, Richard Record and James Riedel 99

mittees have full authority delegated by their respective provincial authorities to issues licenses. Companies submit license applications and pay taxes directly to the management committees. Clearly there are a number of formal beneﬁts that can accrue to foreign investors choosing to site their plants in EPZs and IZs. However, with respect to domestic enterprises, things are not always so clear. There can be something of a disincentive for ﬁrms to locate in designated zones, as this forces them to ‘display’ their activities to the authorities. Many domestic ﬁrms are, perhaps understandably, reluctant to advertise success. Some commentators, including Dapice et al. (1999), have also stressed that provincial authorities need to do more than just build IZs and EPZs, and then watch industry develop. Dapice et al. contrasted the cases of various zones in Vietnam. The Nomura IZ opened in Hai Phong in 1995 and attracted only 5 tenants in the ﬁrst 5 years of operations. Domestic private ﬁrms complained that Nomura IZ was too expensive (many ﬁrms did not require such high-quality infrastructure) and in a poor location (not close to plentiful supplies of unskilled labour). Dapice contrasted Nomura with the success of IZs in Dong Nai province in the south of Vietnam. Here the authorities were able to provide a range of zoned areas from just-cleared land, to semi-developed areas, right up to the high-quality infrastructure sites similar to Nomura. Again this highlights the need for IZs and EPZs to capitalize on Vietnam’s comparatively low-cost labour and light manufacturing advantages, not work against them. If an EPZ or IZ attract only foreign ﬁrms then the scope for knowledge transfers and overspills to take place between foreign and local ﬁrms is severely constrained.

Conclusions: What kind of cluster model for Vietnam? Industrial policy recommendations Competitiveness, productivity and long-term growth Competitiveness is relative, and politically the competitiveness of China is a major concern in Vietnam. China’s entry to the WTO, free trade between China and ASEAN in 2010, and its lead in many areas of reform is pushing Vietnam’s own reform process. That process has sped up in recent years, relative to the 1990s, and is likely to accelerate in coming years, even in the more difﬁcult areas of state enterprise and banking reform. On the external front, there is concern that Vietnam’s inherent structural weaknesses have the potential to undermine its competitiveness in external markets. The high level of tariff and non-tariff protection on many manufacturing and assembling activities will soon decline or vanish once Vietnam starts complying with its regional and international obligations such as AFTA and USBTA in the short run and WTO in the long term. Labour costs in Vietnam are low (see Figure 2.27), even compared with those of China, but utility and infrastructure costs are higher and productivity in general is low.

100 Competition and Cooperation: Vietnam

Hourly labour wages (US$)

Figure 2.27

Hourly labour costs, selected developing countries

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2

Vi et na m

Su In bdi Sa a ha ra n Af ric a Ba ng la de sh

Sr

iL

an ka

hi na C

nd Ph ilip pi ne s

Th ai la

er ic a Am

S.

M

al ay s

ia

0

Data are based on interviews with industry experts carried out by J.E. Austin Associates. The ﬁgure for South America is the minimum wage, actual levels vary across countries in the continent. The ﬁgure for Sub Saharan Africa is an average for the region. Source: J.E. Austin Associates (2001).

In fact labour costs in Vietnam are less than half those of China. However, low wages are not a strong basis upon which to base competitiveness or industrial development. Similarly, abundant natural resources, government incentives or cheap currency are not long-term sources of advantage. At best they offer a limited ‘window of opportunity’ for Vietnam to climb the productivity ladder. Sustainable increases in productivity are the only way to permanently improve living standards. Cluster building can be the driver of productivity growth. The high costs of doing business in the Vietnam were noted at Vietnam Business Forum (9 December, 2002). Participants highlighted the need to: 1. introduce greater competition to reduce costs closer to the regional levels in sectors such as telecommunications, power and shipping; 2. ease restrictions on technology transfers; 3. make business policies and regulations more transparent in their formulation as well as using consistent enforcement to enhance predictability for investors; 4. broaden the currently narrow tax base that puts a disproportionate burden on honest taxpayers and slows the growth in employment; 5. accelerate deregulation and to implement the Enterprise Law aggressively outside the main urban areas; 6. correct misconceptions on the role of the private sector and enhance its role; and the importance of encouraging corporate social responsibility.

Adam McCarty, Richard Record and James Riedel 101

Thus the nature and pace of reform in Vietnam is itself an important macroeconomic variable in explaining short-run growth performance, much more so than in more developed economies with efﬁcient markets where the growth returns to reforms are less signiﬁcant. From this perspective, there are still large growth gains to come from trade liberalization, from ﬁnancial sector and state enterprise reform, and from business deregulation. Many international donor agencies are actively trying to support the development of SMEs in Vietnam. Donor and government interventions to support a given ﬁrm or sector will always accelerate development of the given ﬁrm or sector. The real question is: Are such interventions sustainable and do they justify the costs? While there is no doubt that Vietnam needs a dynamic private sector, not least to provide jobs for the growing labour market, there are real questions as to how the private sector should be developed. The evidence to date suggests that reform is the principal driver of economic growth in Vietnam. The long-term growth prospects remain good for Vietnam, irrespective of ﬂuctuations in the external environment. This is because central planning placed Vietnam’s economy well away from its production possibilities frontier (growth potential), and ongoing reform is a steady process of moving towards full protection. Thus while Vietnam’s factor and product markets remain weak by international standards, they are much stronger than they were only ﬁve years ago and are improving every year. This process unleashes potential, increases productivity and generates ‘once-off’ supply responses to institutional reforms such as the Enterprise Law. Provided these reforms are implemented successfully at the local level, the further development of a healthy business enabling environment will act as a considerable spur to growth. Vietnam’s long-term development may therefore be pictured as the progressive blossoming of an economy that had been restricted by central planning regulations. As those regulations are removed, economic activity and growth responded – led by booming household and agriculture sectors. In the 1990s, foreign investment and rapid export growth added more speed. Then, after a relatively slower period, a new phase of growth led by an emerging private corporate sector may be beginning. Hurdles do however remain. While Vietnam’s low labour costs are a source of regional comparative advantage in the production of certain labour intensive exports, low productivity and the high other costs such as utilities and poor quality infrastructure diminish this advantage. Such issues of competitiveness are likely to leave Vietnam more exposed in the years ahead given trade commitments to AFTA and possible accession to the WTO. Given continuing private sector development, combined with ongoing institutional reforms and macroeconomic stability, a decade or more of high growth rates may well be expected.

102 Competition and Cooperation: Vietnam

Revisiting the conclusions from the 2002 JETRO study Summarized again, the 2002 study described how agglomerations can be grouped together into four broad categories: 1. clusters where locally specialized items are produced or local product districts; 2. clusters where a large core ﬁrm has many subcontracting or parts makers surrounding it or the so-called industrial castle town; 3. clusters in large cities where lots of basic production processes are treated or urban processing clusters; and 4. government-led industrial parks and estates often seen in developing countries. There certainly seems to be some evidence of type (1) clustering in Vietnam. The case of ceramics produced by a large number of small ﬁrms in Bat Trang village, as discussed by the VNCI project, would serve as a good example here. The seafood clusters analysis in the Mekong Delta (see Figure 2.26) also revealed some evidence of a large number of small producers on Phu Quoc Island. Phu Quoc is famed throughout Vietnam for the production of highly pungent ﬁsh source. There also seems to be some evidence of type (2) clustering; the steel industry in Hai Phong would be an example for this case. Perhaps also the large number of textiles clusters seen in the provinces south of Hanoi and the seafood clusters seen in Soc Trang and Ca Mau in the Mekong Delta would fall under this classiﬁcation. The heavy incidence of garments manufacturing in the west of Ho Chi Minh city and northeast and northwest satellites (Bien Hoa and Cu Chi) could be considered under both cluster types (3) and (4). So all four types of cluster are evident in Vietnam to some degree. The 2002 JETRO study also concluded that the success of industrial agglomerations depended on three main factors: 1. infrastructure (highways, ports, electricity supply, and so on); 2. institutional frameworks (legal systems, participatory actors, coordination among actors, and so on); and 3. government support (or foreign assistance) in terms of laws, taxation and ﬁnance. Certainly infrastructure is of vital importance. The basic contention of the Krugman–Venables core and periphery model shows the importance of transportation costs the quality of infrastructure being the key factor in lowering that cost. The evidence suggests that the quality of Vietnam’s state-dominated provision of infrastructure and facilities is fairly poor and expensive. This has obvious knock-on effects with respect to productivity.

Adam McCarty, Richard Record and James Riedel 103

However, these things take time and signiﬁcant progress has already been made in the provision of basin infrastructure. Institutional frameworks are clearly signiﬁcant. The evidence does suggest that reform has been the principal factor in generating private sector growth. The 2000 Enterprise Law put much of the basic legislation is in place now, and remaining efforts will need to focus on implementation. In terms of participatory actors, Vietnam has a strong basis with its existing network of business associations, cooperatives and alliances. There is a long history of collection action and cooperation. However, any notion of institutions needs really to take into account the strength of the rule of law and perceptions with respect to corruption. Transparency International,3 the world’s leading NGO concerned with ﬁghting corruption, publishes an annual ‘Corruption Perceptions Index.’ The highly regarded index ranks 102 countries and reﬂects the perceptions of business people and country analysts, both resident and non-resident. In the 2002 index, Vietnam is ranked at number 85, lower (considered more corrupt) than China (59), Thailand (64) and the Philippines (77), but higher (considered less corrupt) than Indonesia (96). Cambodia and Laos were not surveyed. Finally the notion of government support ﬁts closely with the institutional frameworks concept. Special IZs and EPZs have been quite a successful means of encouraging investment through tax breaks and ﬁnancial incentives, but the evidence shows that the costs and quality of services provided in such zones need to be graded to cater for a wider range of would-be investors. The case of the provincial authorities in Binh Duong (see Box 2.1) highlights the importance of the attitude and approach of the local-level authorities in dealing with the emergent private sector. There does seem, however, to be one more factor that needs to be added to the ‘menu’ above, namely the environment for risk-taking and entrepreneurship. This is the ﬁnal key that is perhaps needed to ensure that clusters, particularly of the more valuable types, can be formed. While there is some evidence of economic agglomeration in Vietnam, clusters of the kind seen in countries such as the USA or Japan have yet to emerge. The GIS maps conﬁrm the locational hypotheses, particularly the bipolar development pattern of Vietnam. With the exception of the software enterprises, access to labour did not appear as a likely explanation for most distributions. More important for developing countries, for both import-substituting and export industries, is access to land and infrastructure (leading to urban markets and ports). The World Economic Forum’s The Global Competitiveness Report 2000 included a survey of business people in 50 economies. The respondents were asked to assess the extent to which they agreed with the following statement: ‘Clusters are present in most international industries and include not only suppliers, but specialized institutions such as university research programs and training providers.’ The mean response for each

104 Competition and Cooperation: Vietnam Table 2.14 Perceived role of clusters, selected Western and East Asian economies Economy Finland United States Germany Singapore Taiwan Japan Hong Kong Korea, Rep. of Malaysia China Philippines Thailand Vietnam Indonesia

Ranking 1 2 3 4 5 8 11 17 25 28 35 38 39 46

Source: Global Competitiveness Report 2000, World Economic Forum, copyright © 2000 by the World Economic Forum. Used by permission of Oxford University Press, Inc.

economy was calculated and the lowest possible ranking was 50. The results for selected economies can be seen in Table 2.14. The list of top countries is not surprising, although the position of Taiwan at number 5 in the rankings ought to act as some inspiration to Vietnam. Vietnam compares reasonably favourably with its peers in ASEAN (Philippines, Thailand and Indonesia), but rather disappointingly trails China by some distance. If clusters are dynamic centres of growth, then understanding better how they might be created and managed in Vietnam is a matter of great signiﬁcance for the future. The experience of the leading economies of East Asia, especially Japan and Korea, should serve Vietnam well. The evidence from these countries suggests that a culture that is open to ideas, but is also successful in conserving core values, is most likely to thrive in a global economy where knowledge is the dynamic driving force of economic development. Questions as to the extent of the role of the state in Vietnam’s market economy abound. The example of Japan’s post-1945 industrial development is particularly relevant. Japan’s state-oriented and cooperative drive towards modernity achieved astonishing results in allowing Japan to emerge from the ashes of wartime destruction to achieve equal status with the world’s leading industrial nations by the 1980s. However, Japan’s economic stagnation over the last ten years now sends a different message to its East Asian neighbours. The role of the state in directing technological catch-up is much diminished. Private enterprises are now the international drivers of technological development. While research in universities and in government and private institutes makes a major contribution to global

Adam McCarty, Richard Record and James Riedel 105

R&D, in East Asia and across the world ﬁrms do the dominant share of applied and basic research. Increasingly, technological advance rests with researchers spurred by a highly competitive environment that rewards excellence, and it depends on large, research-oriented corporations, universities and private ﬁnanciers willing to support research. Malaysia serves as another warning to Vietnam in this regard. Malaysia’s high-proﬁle drive towards the ‘K’ or knowledge economy was launched amid much fanfare in the late 1990s with the setting-up of the ‘multimedia technology corridor’ and city of Cyberjaya. The results from the investments in infrastructure have been very disappointing. The world’s largest investment fund (Calpers) has withdrawn all investments from Malaysia, saying the lack of press and political freedoms undermined Malaysia’s longterm prospects (BBC News, 29 May 2003). Perhaps the most important requirement for the successful development of a knowledge-based cluster economy is an environment where there is freedom of expression, opportunity and thought, where individuals are truly able to experiment, innovate and become entrepreneurs. Vietnam has some distance to travel in this regard. The lesson for Vietnam is that government is no longer the principal actor in driving economic and technological development; its role has changed to one of setting the stage and creating a conducive enabling environment to freely allow entrepreneurship and innovation.

Acknowledgements The authors would like to thank Mekong Economics Ltd for their assistance towards completing this report – in particular, for the GIS inputs of Dr Joost Buurman and for the research inputs of Scott Robertson. Mekong Economics Ltd is a private Vietnamese company, supporting the development of a competitive market-based economy in Vietnam and the Greater Mekong sub-region. Established in 2000, Mekong Economics specializes in economic research, consulting and Ofﬁcial Development Assistance (ODA) projects and programs (http://www.mekongeconomics.com).

Appendix: Vietnam provincial codes Each of the enterprises included in the GIS maps and analysis was assigned a 7-digit geographical location code. The ﬁrst 3 digits represent the province, the next 2 digits represent the district, and the last 2 represent the commune. The map (Figure 2.28) shows the provincial codes for Vietnam. These are standard and assigned by the Ministry of Science, Technology and Environment (MOSTE). Owing to changes in province, district and commune boundaries, a number of enterprises in the 2001 data set were allocated codes which no

106 Figure 2.28

Vietnam provincial codes

Adam McCarty, Richard Record and James Riedel 107

longer exist. Hence out of the total 12-sector data set of 4402 ﬁrms, only 3932 enterprises were mapped and analysed.

Notes 1. From Amazon.com. 2. ‘Iceberg’ transportation costs: only a fraction of each unit shipped arrives. 3. http://www.transparency.org

Bibliography and references Asian Development Bank (ADB). 2003. Asian Development Outlook 2003. Manila. Athukorala, Prema-Chandra. 2002. Foreign Direct Investment and Manufacturing Exports: Opportunities and Strategies. Washington, DC: World Bank. Aw, Bee Y., X. Chen and M.J. Roberts. 1997. ‘Firm Level Evidence on Productivity Differentials, Turnover and Exports in Taiwanese Manufacturing’, Working Paper no. 6235, National Bureau of Economic Research. Cambridge, MA. Cohen, Stephen S. and Gary Fields. 1999. ‘Social Capital and Capital Gains in Silicon Valley’, California Management Review, 41(2), pp. 108–30. Dapice, David O., Brian J.M. Quinn, Pham Anh Tuan, Bui Van and Ho Dang Hoa 1999. Trade and Industrial Development Strategies for the Ha Long Bay Region. Development Discussion Paper no. 683, Harvard Institute for International Development. Cambridge, MA, March. Freeman, Nick J. 2001. ‘Understanding the Decline in Foreign Investor Sentiment Towards Vietnam During the 1990s’, Asia Paciﬁc Business Review. Autumn. Vol 8, no. 1, pp. 1–18. Fujita, Masahisa, Paul Krugman and Anthony Venables. 2001. The Spatial Economy – Cities, Regions, and International Trade. Cambridge, MA: MIT Press. Fujita, Masahisa and Jacques-François Thisse. 2002. Economics of Agglomeration. New York: Oxford University Press. Giroud, Axèle. 2002. Vietnam in the Regional and Global TNC Value Chain. Draft paper, DFID Globalisation and Poverty Research Programme Workshop, Hanoi, September. Griliches, Zvi. Nov. 1992. ‘The Search for R&D Spillovers’, Scandinavian Journal of Economics Vol. 94 (Supplement 1992), pp. 29–47. General Statistics Ofﬁce. 2002. Statistical Yearbook. Statistical Publishing House, Hanoi. Hopkins, B. 2002. The Impact of Vietnam’s Infrastructure on Exporters: A Survey. Background paper, World Bank study: Vietnam Exports: Policy and Prospects. Hanoi. Hotelling, Harold. 1929. ‘Stability in Competition’, Economic Journal. Vol. 39 pp. 41–57. Japan External Trade Organization (JETRO). 2002. Industrial Agglomeration: Facts and Lessons for Developing Countries. Tokyo: Institute of Developing Economies, JETRO. J.E. Austin Associates. 2001. Survey of Competitiveness in Developing Countries http://www.jeaustin.com/index.html Krugman, Paul. 1991. Geography and Trade. Leuven University Press. Krugman, Paul and Anthony Venables. 1990. ‘Integration and the Competitiveness of the Peripheral Industry’, in C. Bliss and J. Braga De Macedo, (eds), Unity With Diversity in the European Economy: The Community’s Southern Frontier. Cambridge University Press.

108 Competition and Cooperation: Vietnam Marshall, A. 1920. Principles of Economics, (8th edn). London: Macmillan. Mills, Edwin S. and Bruce W. Hamilton. 1994. Urban Economics, (5th edn). New York: HarperCollins. Porter, Michael E. 1980. The Competitive Advantage of Nations. New York: Free Press. Porter, Michael E. and Scott Stern. 2001. ‘Innovation: Location Matters’, MIT Sloan Management Review, 42(4), 28–36. Pham Van Thuyet. 2002. ‘Vietnam: Regulatory Environment and Support Services for Export Development’, background paper to the World Bank study Vietnam Exports: Policy and Prospects. Hanoi. Saxenian, Anna Lee. 1994. Regional Advantage: Culture and Competition in Silicon Valley and Route 128. Cambridge, MA: Harvard University Press. STAR – Vietnam. 2003. An Assessment of the Economic Impact of the United States – Vietnam Bilateral Trade Agreement, Annual Report for 2002. Hanoi: National Political Publisher. Steering Committee for State Administrative Reform 2002. Impact of Public Administrative Reform on Socio-Economic Development in Binh Duong province, unpublished manuscript. Hanoi. United Nations Development Programme. 1993. Human Development Report. New York: Oxford University Press. United States Department of State. 2000. FY 2001 Country Commercial Guide: Vietnam. Washington, DC: US Department of State. Vietnam Administrative Atlas. 2002. Hanoi: Cartographic Publishing House. Vietnam Competitiveness Initiative (VNCI). 2003a. Bat Trang Ceramics Competitiveness Strategy, report prepared by J.E. Austin Associates for USAID. Hanoi. Vietnam Competitiveness Initiative (VNCI). 2003b. Fruit Cluster Strategy, report prepared by J.E. Austin Associates for USAID. Hanoi. Vietnam Competitiveness Initiative (VNCI). 2003c. Software/ICT Cluster Strategy, report prepared by J.E. Austin Associates for USAID. Hanoi. Vietnam Economic News. 2003. Tuesday, 5 August 2003 pp. 12 World Bank. 2002. Vietnam Development Report 2003 – Vietnam: Delivering on its Promise, World Bank in collaboration with the Asian Development Bank, Vietnam Consultative Group Meeting, Hanoi, December 10–11. World Economic Forum. 2000. The Global Competitiveness Report 2000. New York: Oxford University Press. World Economic Forum. 2001. The Global Competitiveness Report 2001. New York: Oxford University Press.

Part II Industrial Clusters in China

This page intentionally left blank

3 Industrial Clusters in the Tianjin Area Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu

Introduction Regional industrial clusters in evolution In this time of economic globalization, economic regionalization develops rapidly. The reason is that the region has been an agglomeration centre for enterprise clusters, qualiﬁed personnel, funds, markets, information, infrastructures, important regional studies and networks as well as a tie of integration. Economic globalization beneﬁts from the development of transportation and information technology. A well-developed network of information and transportation in modern society makes it possible for one industrial sector to very conveniently gain material support from a distant region and then sell its products to more distant markets. However, as we have observed, economic globalization has not successfully reduced the distinction among regions, though the economic level of each region has taken on a dynamic look. Instead, many superior industrial sectors are still tightly connected with geographical features. The Competitive Advantage of Nations (1980), written by Michael E. Porter, an American scholar, listed some lasting industries, such as chocolate in Switzerland, ﬂowers and plants in the Netherlands, computers in the USA and so on. All of these industries have the characteristic of regional agglomeration. Both economic globalization and regionalizing are trends that can be accomplished without conﬂict. Taking technological innovation as the motive force and regarding transnational corporations as the carrier, economic globalization is, essentially, a worldwide adjustment of industrial setup. In this process, developed countries turn to innovation-driving and developing countries speed up industrialization so that economic resources can be redistributed on a large scale around the world. The distinction of economic development levels and the speed of different regions (including various regions within developed countries) is usually directly related to the uniqueness of their industrial setup. Moreover, the economic vigour of 111

112 Industrial Clusters in the Tianjin Area

a region is ﬁrst displayed by a group of superior industrial sectors. Research in economics shows that traditional theory, which emphasizes only the comparative advantage of natural resources and other cost factors, cannot explain the development and changes of regional economy or how each superior industrial sector in a region has its own complicated and dynamic supporting system. A so-called industrial cluster comprises both the industries that gather within a certain region and their supporting system. Porter generalized this supporting system as having the following four aspects: demand condition, competition condition, factor condition and support condition. He considered that these four aspects form the paints of a diamond shape that interrelate and interact. He particularly emphasized that a competitive advantage on the basis of only one or two aspects of the diamond has been proved to be unstable because other competitors may easily surpass it and that prolonged competitive advantage can be produced only by mutual support and the coordinated development of all four aspects. Component factors of industrial clusters Industrial clusters include upstream industrial sectors that provide raw materials and means of production for core industrial sectors and those industrial sectors each of whose production technology affects the others. That is not to say that each region must have a complete system. The role of upstream industrial sectors can be displayed by a well-developed international purchasing network, and this kind of network has become a symbol of developed regions. The blending trend of correlated industries is increasingly obvious during the evolution process of modern industries. For example, the computer, TV sets, communication and software industries are merging into an organic whole. Such a phenomenon also appears in industries like agriculture, food, medicine, textiles, plastics, chemistry and garments. Thus, it is difﬁcult to distinguish which industry is upstream and which downstream. Through combination of production technologies, numbers of new industries have been formed. Concerning regional economic environment, a favourable regional economic environment should have at least, six key factors in order to forcefully participate in the competition of the global economy. These factors are as follows: 1. Rich technological resources and appropriate distribution media. Technology is one of the most fundamental motivations that promote the development of a regional economy. The effect of technology on the development of a regional economy is as follows. First, technology is a key factor for increasing output and improving productivity. The purpose of technological improvement is to provide new material, new energy resources and new methods of exploiting energy. Second, technology can change the input–

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 113

output structure of production. In addition, technology can also change the existing condition and relative value of regional factors through substitution among factors. Third, technology may raise the competitive power of a regional economy. Both producing new products and producing old products by new methods can enhance the relative competitive power of a regional economy. The source of technological resources includes universities, national labs and scientiﬁc research institutes in enterprises. They can offer research achievements that are necessary for industrial sectors to seek a more consolidated competitive position. These achievements can be timely and efﬁciently transferred into industrial sectors through distribution media. 2. Human resources with high adaptability and organizations for development and training. One of the difﬁculties that underdeveloped countries and regions are facing is the shortage of human resources with high adaptability. In quantity, there is surplus labour in underdeveloped countries and regions. Nevertheless, modern industry needs not only ordinary physical but also qualiﬁed labour with professional skill and high adaptability. Economic development requires that regions provide lots of professional personnel and skilled workers for enterprises. Besides, through basic occupational education, higher education and continuous in-service education, regions should give human resources the knowledge and skill that allow them to adapt to competition and changes. 3. Abundant sources of funds. The competitive power of regional industries cannot do without the strong support of ﬁnancial organizations. The functions of these ﬁnancial organizations include raising enough funds for economic development and establishing investment mechanisms that can fully reﬂect guidance of competitive market power. 4. Favourable macroeconomic environment with efﬁcient management. This kind of management level is embodied in two aspects: one is that macroeconomic regulations can guide enterprises effectively and at the same time do not destroy the independence and competitive relation of enterprises; the other is that the tax system is beneﬁcial to competition, and at the same time meets the needs of economic development and the public. 5. Well-developed infrastructure of software and hardware. The efﬁciency of regional economic activities is, to a large extent, affected by the condition of infrastructure. Therefore, if a region wants to reach its goal of economic development, it must arrive at a comparatively high level of communication facilities, transport and storage facilities, environmental protection facilities, living facilities that can attract qualiﬁed personnel and a friendly environment in term of public policy and the administration of justice and so on.

114 Industrial Clusters in the Tianjin Area

6. Groups of customers with discriminatory ability. Views of these customers and their requirements for products can reﬂect the basic trend of market changes and should be regarded as the basis for product innovation by enterprises. Three hierarchies of regional competitive power In the three hierarchies of regional competitive power (Figure 3.1), the competitive power of enterprises is the highest and it is the ﬁnal manifestation of regional economic competitive power. However, the competitive power of an individual enterprise does not necessary depend on an industrial cluster. With a unique resource advantage or a special nucleus ability, an enterprise may keep its favourable market position for a long time. The competitive power of an industrial cluster is expressed as the competitive power of the chief enterprise within it. Besides resource condition, nucleus ability and dynamic ability within the enterprise, this kind of competitive power is supported by other component factors in a cluster, including an orderly competitive relationship among its enterprises, which provides the external motive force for sustained innovation by them a developed international supply chain which ensures that enterprises can easily purchase raw material, components which are up to international quality standard, and a worldwide market network which can collect plenty of market information and supply unblocked market channels for the enterprises. A platform for regional economic development must offer the necessary condition for the development of industrial clusters, and at the same is the Figure 3.1

Hierarchical structure of regional competitive power in China

Competitive power of enterprises

Resource advantage

Nucleus ability

Dynamic ability

Competitive power of industrial clusters

Interior competition

Supply chain

Market network

Government policy

Service industry

Innovation system

Infrastructure

Human resource

International status

Platform of Economic development

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 115

basic factor of regional economic competitive power. The components of this development platform decide whether the chief industrial clusters in a region are rooted, that is, whether the regional environment has an active effect on running costs, opening up markets and innovation efﬁciency for each level of enterprises within the industrial clusters. Industrial clusters in Tianjin After decades of development, the industrial system in Tianjin is making a show of its distinguished characteristics with large-scale groups of basically foreign-funded enterprises, complete processing and manufacturing industry, and heavy industry. High-tech and capital-intensive industries, such as electronic information and automobiles, and traditional industrial clusters, such as bicycles and garments, have appeared in an embryonic form. In the process of industrial agglomeration, both high-tech industrial clusters and traditional industrial clusters each have their respective characteristics in shaping the process, features and development trends. Among high-tech industries in Tianjin, we chose the electronic information industry and the green battery industry for environmental protection. Among traditional industries there, we chose the automobile manufacturing industry and bicycle manufacturing industry. These four industrial clusters are representative in Tianjin. Taking them as our research object we can, through comparison and analysis, understand their respective characteristics, shortcomings and development trends, and thus help to improve the effect of industrial clusters, strengthen industrial competitive power and cultivate Tianjin’s new industrial clusters.

New motive force in the development of Tianjin electronic information industry In recent years, the electronic information industry in Tianjin has developed rapidly. The ratio of the electronic information industry’s output value to the gross industrial output value of the whole city increased from 18 per cent in 1998 to 24 per cent in 2001. The ratio of its value added to industrial value added over the same period rose from 14.5 per cent to 24.6 per cent in. The proportion of its pre-tax proﬁts in the gross pre-tax proﬁts of Tianjin has remained at about 30 per cent for a long time. In 2000, the gross industrial output value, value added and proﬁts of Tianjin’s electronic information industry respectively reached ﬁfth, third and second positions, respectively, in the national information industry. In 2000, total sales of the electronic information industry in China were RMB 556.4bn. Total proﬁts were RMB 33.9bn. and the average proﬁt rate was 6 per cent. In the same year, total sales of the electronic information industry in Tianjin were RMB 57.158bn. Total proﬁts were RMB 6.82bn. and its proportion in total proﬁts of the electronic information industry of China was 20.1 per cent.

116 Industrial Clusters in the Tianjin Area

Gross Industrial Output Value of Electronic Information Industry (100m. Yuan)

718.76

948.23

Ratio of Electronic Information Industry to Gross Industrial Output Value (%)

24.4

31.4

14.5

22.2

Ratio of High-tech Industry to Gross Industrial Output Value (%)

31.3

41.99

23.8

Ratio of Electronic Information Industry to High-tech Industry (%)

80.5

69.25

54.69

Ratio of foreign-funded enterprises to output value of electronic information industry (%)

94.3

87.0

Zhejiang

Jiangsu

Guangdong

Shanghai

Beijing

Tianjin

Table 3.1 Comparison of the development characteristics of the electronic information industry in six Chinese provinces and cities, 2001

1013.09 3109.91 1070.92 329.00

–

9.1

4.2

–

–

–

–

–

–

78.3

76.6*

48.1

The ratio of foreign-funded enterprises to the electronic information industry in Jiangsu is the ratio of foreign-funded enterprises to the value added of the electronic information industry. Sources: Related data of above designated size industrial enterprises in Statistical Yearbook of Provinces and Cities in 2002 and 2002 China Statistical Yearbook.

The proﬁt rate was 11.9 per cent, twice the national level and the highest in China. A comparison of six provinces and cities whose electronic information industry has the largest scale and developed most quickly shows that the healthy and rapid development of Tianjin’s electronic information industry leads the whole country and has become the ﬁrst-pillar industry of Tianjin (Table 3.1). Characteristics of electronic information industrial clusters in Tianjin Regional industries can be compared and measured by the location quotient (LQ), which expresses the comparative advantages or relative special agglomeration of a certain industry in a certain region. Also, the LQ is the

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 117

Jiangsu

3.46

1.54

2.36

1.08

0.44

3.29

3.27

1.50

2.44

1.2

0.52

Zhejiang

Guangdong

2.6

Specialization rate of value added, SVu

Beijing

Location quotient, LQ

Tianjin

Shanghai

Table 3.2 Comparison of the degree of agglomeration in the electronic information industry in six provinces and cities in 2001

Sources: Compiled from Statistical Yearbook of Provinces and Cities, 2002; 2002 China Statistical Yearbook.

basis for judging the local economic base, local industrial resources, local economic advantages and industrial activity tendencies. The specialization rate of value added (SVu) can express the degree of specialization in the electronic information industry. For Tianjin, these two indicators are respectively 2.6 and 3.29 (Table 3.2), ronking it respectively second and ﬁrst of six provinces and cities. This shows that the electronic information industry in Tianjin has a comparatively high degree of specialization and agglomeration and has formed a certain scale of industrial clusters. Speciﬁcally, the electronic information industrial cluster in Tianjin has the following characteristics: Exogenous industrial clusters The motive force shaping industrial clusters comes from the entry of foreign-funded enterprises into the electronic information industry. The exogeneity of electronic information industrial clusters is expressed in the following aspects. The ﬁrst is the dependence on foreign funds. Here, the dependence on foreign funds is indicated by the ratio between the volume of the contract foreign funds of the electronic information industry and the volume of its gross contract investment. In 2001, major electronic industries in the Tianjin Economic-Technological Development Area took up 90 per cent of the whole electronic information industry in Tianjin. The data from the technological development area can reﬂect the export-oriented characteristics of the electronic information industry in Tianjin, the ratio of the volume of the contract foreign funds of the industry to the volume of its contract investment having been 93.01 per cent in 1998, 89.62 per cent in 1999, 94.37 per cent in 2000 and 93.71 per cent in 2001. These ratios have shown strong dependence on foreign funds. The second aspect is the increasingly large scale of foreign funds in the electronic information industry. This indicator is expressed by the ratio of the accumulated contract foreign funds invested in the industry in the

118 Industrial Clusters in the Tianjin Area

economic and technological development area to the total accumulated contract foreign funds of all industries in the area. This ratio was 26.4 per cent in 1998, 26.03 per cent in 1999, 39.3 per cent in 2000, and 39.9 per cent in 2001. This trend tells us that foreign funds are highly concentrated in the electronic information industry. The third aspect is the ratio of foreign funds in the electronic information industry to gross funds, which the following data reﬂect. The ratio of Hong Kong, Macao and Taiwan and foreign-invested economy to the gross output value of Tianjin’s electronic information products manufacturing industry was 95.4 per cent in 2000 and 94.3 per cent in 2001. The ratio of Hong Kong, Macao and Taiwan and foreign-invested economy to scale income was 95.4 per cent in 2000 and 95.1 per cent in 2001. Thus foreign funds have taken up a large percentage. Industrial clusters with the structural characteristics of a ‘single core’ First of all, let us see the average scale intensity of enterprises that is expressed by the ratio of the gross output value of the electronic information industry to the numbers of enterprises. In 2001, Tianjin’s ratio was 2.68. In the same year, Beijing’s 3.26, Shanghai’s 2.43, Guangdong’s 2.19, Jiangsu’s 1.55 and Zhejiang’s 0.69. Tianjin ranked second, and its enterprises had a very large scale of production. Second, regarding the scale structure of enterprises, large-scale enterprises took up 91.37 per cent of the gross output value of the electronic information industry and 91.02 per cent of sales income in 2000. These ratios were respectively turned into 91.05 per cent and 91.13 per cent in 2001. Large-scale enterprises occupy centre stage. Third, ‘single-core’ enterprises of industrial clusters are largescale transnational corporations. The gross sales of Motorola in Tianjin were 57.8 per cent of those of Tianjin’s electronic information industry in 2000. In the same year, its proﬁts were equal to 45 per cent of the gross proﬁts and 74.63 per cent of proﬁts of the whole electronic information industry in Tianjin. The proﬁt rate was 15.3 per cent, which was higher than 11.9 per cent, the average proﬁt rate of the whole electronic information industry in Tianjin. In 2001, the gross sales of Motorola in Tianjin took up 63.33 per cent of gross sales of Tianjin’s electronic information industry. Regarding investment structure, investment from the USA has occupied a comparatively large part of the electronic information industry (37 per cent) because of Motorola. After the USA comes Korea. Investment structural proportions from Hong Kong, Japan and Taiwan also show the structural feature of a single core. Rooting of industrial clusters and foundation of local production network The rooting of industrial clusters is, in fact, the social and cultural characteristics within the clusters and the institutional source of industrial cluster advantages. It plays an important role in the further development of high-

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 119

tech industrial clusters. Tianjin is a famous old industrial base in China. The development in Tianjin of state-owned enterprises has laid a solid industrial foundation, trained a large number of skilled industrial technical workers and a created 9 manufacturing industry culture, an industrial atmosphere and an institutional environment there. If foreign funds can be regarded as the direct motive force for the formation of electronic information industrial clusters, the industrial culture and institutional environment founded on the basis of the original industry are the potential motive force for such formation. The interaction and integration of the two kinds of force form a large-scale, enterprise-centred local production network where numbers of auxiliary specialized enterprises share work and cooperate with each other and where related enterprises and downstream enterprises develop in concert. Meanwhile, the interaction and integration give the economic activities of enterprises within industrial clusters their rooting in the local network. The foundation of the local production network in the economic information industry in Tianjin is realized by the interaction and integration of foreign-funded enterprises and state-owned enterprises. First, stateowned enterprises speed their steps of institutional innovation and technological improvement through pooling capital and cooperation. Second, in order to lower costs and improve market reaction ability, transnational corporations begin to seek local suppliers and sub-suppliers. At the same time, state-owned enterprises extend their industrial chain by cooperation with transnational corporations. This kind of industrial cluster with rooting has supplied a sort of glue for agglomeration that makes many enterprises stick together. It is beneﬁcial both for building a regional innovative environment and for giving enterprises deep roots in the local area. Industrial clusters with a product chain Within the electronic information industrial clusters in Tianjin, the economic bodies connect with each other to form an inner value chain through the supply and manufacture of accessories and the production, processing and service of upstream and downstream products. It is a typical clustering mode of product chain. In this chain, three kinds of enterprises share work and cooperate with each other to form a product chain cluster with competitive power (Table 3.3). Main problems of electronic information industrial clusters in Tianjin Relatively strong dependence This kind of dependence is shown through the following two aspects. First, dependence on funds. The percentage of foreign-funded enterprises in Tianjin changed from 89.9 per cent in 1995 to 94.3 per cent in 2001, which was far higher than the national average level (73.76 per cent) and ranked ﬁrst. Besides, the investment scale of foreign funds keeps increasing.

120 Table 3.3

Electronic industrial clusters in Tianjin

Type

Contract manufacturers

Characteristics

Upstream semiconductor/ key parts and union production ﬁrms

Downstream components and parts and union manufacturers

Features of Enterprises

TechnologyIntensive

InvestmentDriving

Market-Driving

Competitive Power

Mastering Key Technology and Industrial Criteria

Comparatively High productive capacity, Advanced Technology of Manufacturer and Considerable Technical Strength

Varied Products

Additional Value of Products

Highest

Participation in the Distribution of Additional Value

Relatively Low

Market Share

Considerable, with Own Network of Sale and After-sales Service

Building Stable Market by Longterm Contract with OEM Customers

Market Divided in Detail

Production Mode

Integration and OEM Manufacture by Contract Manufacturer

Digesting Design Technology of Union Firms and Producing Products According to Requirements of Given Market (EMS)

Enlarging Scale and Carrying on Plural Production

Status

Leading Enterprise Which Guides Industrial Development

Important Link in Industrial Chain

Edge

Market Entry Barrier

Patent Right and Capital Volume

Service and Quality/ Cost Advantage

Not High

Representative Enterprises

Taking Foreignfunded and Foreignholding Enterprises as the Dominant Factor, Motorola, Samsung Electronics, Mitsumi and Panasonic

Taking State-owned and State-holding Enterprises As the Dominant Factor, 754 Plant, 712 Plant, 764 Plant and 609 Plant

A Fairly Large Number of Joint Ventures and Stateowned and Stateholding Enterprises of Medium and Small Size

Sources: Tianjin Zhonghuan Electronic Information Group Co. Ltd; Tianjin Optical Electrical Communication Corporation (754 Plant); Enterprise Survey Group of Tianjin Municipal Statistical Bureau.

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 121

Second, the strong dependence on external funds leads to the technological dependence of the electronic information industry in Tianjin. The development of the electronic information industry in Tianjin relies, in a large degree, on pooling capital with foreign businessmen and manufacturing complete sets of equipment for foreign-funded enterprises and foreignholding enterprises. Although the technology and management of the electronic information industry in Tianjin have been improved to a certain extent through pooling capital and manufacturing accessories, foreignfunded and foreign-holding enterprises still dominate in the key technology for semiconductors and overall product design. These foreign-funded and foreign-held enterprises gain the greatest additional value. In contrast, local enterprises get only the spread of peripheral technology and management knowledge, and their ability for self-development and design is limited. Institutional factors restricting further development The development of electronic information industrial clusters in Tianjin has a strong ‘path dependence’. As an initial condition for the formation of industrial clusters, basic industrial environment and institutional environment that have been created by investment accumulation, technology and human capital accumulation of original state-owned enterprises play a fundamental role in the formation of local production network. However, the further development of industrial clusters needs further innovation in the institutional network. The three main reasons are as follows. First, stateowned enterprises have small-scale, insufﬁcient investment and slow-speed development. Regarding average scale intensity of large and medium-sized state-owned enterprises in the electronic information industry, this indicator was 0.34 in 2001, and it is far below the industrial average of the whole city (2.68). Concerning input, in 2002 there were altogether 8 ratiﬁed technical reform projects of state-owned enterprises and the total investment value was 149m., which was 14.7 per cent of gross investment value for technical reform of the whole industry, which was RMB 1,012m. Severe insufﬁciency of input to state-owned enterprises restrains their further development. Second, state-owned enterprises have inadequate input in technical development, low ability, weak competitive power and brain drain, and lack inner motive force for reform. From the viewpoint of intensive characteristics of industrial factors, the LPK (expressed by numbers of the employed absorbed by ﬁxed net assets) value of large-scale and medium-sized state-owned enterprises in the electronic information industry in Tianjin was 0.115 in 2001. In contrast, the average value of the electronic information industry in the same year was 0.051. This tells us that state-owned enterprises in the electronic information industry have obvious labour-intensive characteristics and have fallen short in technology. Third, many enterprises in the electronic information industry are

122 Industrial Clusters in the Tianjin Area

enterprises solely owned by the state. Their shortcomings, such as rigidity in system and many chiefs in management and separation, limit the development of enterprises. As a key factor in institutional network, the further reform and development of state-owned enterprises in the electronic information industry in Tianjin will play a signiﬁcant role in sustained and stable development of industrial clusters. Local production network not embedded to a large extent In electronic information industrial clusters in Tianjin, foreign-funded enterprises are the important source for forming the local production network. However, in the local production network of electronic information industrial clusters in Tianjin, foreign-funded enterprises have a comparatively low degree of localization in the commercial connection of high-tech products. Regarding the product market and the factor market, the connection of foreign-funded enterprises is mainly national and international, and their technical spread is quite limited. This is reﬂected in the following aspects: on the one hand, some key components and parts and raw materials are imported mainly from foreign countries; on the other hand, with the coming of union production ﬁrms, ﬁrms producing accessories for key components move so that foreign-funded enterprises have a weak connection with local enterprises. Although local enterprises producing accessories are increasing, they are concentrated mainly in sectors whose additional value is low. Therefore, though the local production network of the electronic information industry in Tianjin has been founded and become a regional knot in the global network of transnational corporations, it is still in the production sector with low additional value and a peripheral position in the global network and has not been deeply inlaid in the global production system. Fragility and instability of the ‘single-core’ structure of industrial clusters One of the extremely important characteristics of electronic information industrial clusters in Tianjin is the structural one of having a ‘single core’. Such a characteristic is a necessary stage in the formation and development process of electronic information industrial clusters in Tianjin. However, in the long run, industrial clusters with it are easily affected by political and economic changes in the mother country of single-core enterprises and hence have some risk. Meanwhile, the electronic information industry is a technology-push industry. Furthermore, the market share of each singlecore enterprise in a single-core industrial cluster directly inﬂuences the development of the industrial cluster as a whole, and thus industrial clusters have market risk. Within a region, the development of leading industries is, in fact, monopolistic and lacks motive force in transferring technology to local enterprises because single-core enterprises dominate in the key technology. This makes the technical spread of foreign-funded

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 123

enterprises limited, restricts technological improvement of local enterprises, causes further dependence of local enterprises on core enterprises, forms a vicious circle and ﬁnally leads to the fragility and instability of industrial clusters. Lack of a knowledge network connection helpful to innovation Electronic information industrial clusters in Tianjin are more or less typical product chain industrial clusters. Enterprises are connected only through sharing work in the product chain. The technology overﬂow effect and the study effect of foreign-funded enterprises are not clear. Because of the lack of a sound cooperation mechanism, universities and research institutes in Tianjin have failed to become an important overﬂow source of industrial innovation in electronic information. In addition, compared with the connection founded on the basis of products, enterprises are short of formal connection based on common technology. In particular, the information and knowledge spread mechanism that is helpful to innovation is lacking between foreign-funded enterprises and local enterprises. In order to protect a monopoly on advanced technology, foreign-funded enterprises usually manage to avoid too quick an overﬂow of key technology to a local area. Their technical innovation and reform have a vertical connection only with mother enterprises and have no horizontal connection with local enterprises. In such product chain industrial clusters, the cooperating R&D mechanism and interaction mode among enterprises and among enterprises, universities and research institutes have not been found. Thus the competitive advantage of industrial clusters cannot be brought into full play. Solutions for development of electronic information industrial clusters in Tianjin According to the above analysis, the future development of the electronic information industry in Tianjin should start with the following: Combining industrial advantage, policy advantage, low-cost advantage and location advantage, and exactly positioning the electronic information industry in Tianjin according to regional competitive advantage The advantages of the electronic information industry in Tianjin are reﬂected mainly through the following aspects: 1. Industrial advantage. Electronic information industrial clusters in Tianjin have appeared on a certain scale, and formed industrial agglomerations to relatively good effect. Moreover, the industrial clusters have acquired some mutual dependence with transnational corporations. The electronic information industry has a comparatively high ability in manufacturing accessories and a relatively complete industrial chain.

124 Industrial Clusters in the Tianjin Area

2. Policy advantage. In the development process of the electronic information industry, Motorola and Samsung settled down in Tianjin entirely thanks to the open, and favourable policy advantage of the area. 3. Low cost advantage. As a famous old Chinese industrial base, Tianjin has a large number of skilled technical workers, a complete training system at each level and low labour force costs. These together compose the low cost advantage of the electronic information industry in Tianjin. 4. Location advantage. Tianjin is located in the centre of the Bohai Sea economic region where China’s city clusters, industrial clusters and port clusters concentrate most densely. This region has become a key investment region preferred by foreign businessmen. While making foreign investment strategy, foreign businessmen mainly consider a production base, an R&D base and an operation base. Taking the comparative advantage of Tianjin into account, foreign businessmen may locate their R&D centres and operation headquarters in Beijing and build their production base in Tianjin. The development direction of electronic information industrial clusters in Tianjin may be rationally formulated according to this basic layout. Using the development of the structure of industrial clusters from single-core to multiple-cores to form a competitive pattern within the region and to transfer and improve electronic information industrial technology With its technical advantage, transnational corporations in electronic information industrial clusters in Tianjin are in a monopolistic position. Industrial clusters with this kind of structure are not helpful for the technological improvement and industrial structure upgrade of local enterprises. This kind of single-core structure should be changed through two channels. One is that, regarding investment promotion, transnational enterprises are still the dominant factor. Tianjin should continue to keep Motorola’s central position, attract more foreign funds from Samsung and Japanfunded enterprises, and meanwhile seek investment from Southeast Asia and the developed regions of Europe and America so as to promote competition among transnational corporations. The other is that Tianjin should cultivate leading state-owned and state-holding enterprises in the electronic information industry as well as state-owned enterprises which can compete with transnational corporations. Cultivation of local industrial clusters with self-strengthening mechanisms and turning product chain clusters into innovation chain clusters so as to change regional comparative advantage into competitive advantage If the electronic information industry in Tianjin wants to realize its industrial escalation and technological improvement through a change from OEM to ODM, enterprises must have quite high technological design abil-

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 125

ity. To gain this kind of ability depends on the foundation of a selfstrengthening mechanism that can maintain sustained innovation in technology. This mechanism can be realized through the following channels: strengthening personnel and information communication among universities, research institutes and industries, building studying-type plants, and encouraging further development of a ‘production–study–research’ cooperation mechanism through the intervention of local government. Through these channels, the mechanism of regional knowledge ﬂow and innovation study can be strengthened so as to change comparative advantage into competitive advantage. Speeding up the steps of reform of state-owned enterprises in the electronic information industry and further completing institutional network within industrial clusters Since Tianjin is an old industrial base, its state-owned enterprises make up a large percentage of the electronic information industry. These stateowned enterprises lack competitive power and motive force, and hence they are in an unfavourable position in competition. Thus, to speed up the development of state-owned enterprises has become an important link in further completing institutional networking within electronic information industrial clusters. The development of state-owned enterprises can be divided into two cases. First, after intense market competition, some large and medium-sized state-owned enterprises have rapidly improved their technical grade and quality level. The asset quality and the technical level of these enterprises have been pushed up a gear. For example, built in 1960, the Zhonghuan Semiconductor Co. Ltd, with more than 900 employees, imported advanced technology, equipment and management methods to take it to international level. However, depending just on imports, the company was always at the stage giving low priority to technology. In order to change this kind of situation, it increased input in R&D and carried on self-exploitation and self-innovation. In 2002, the input to R&D reached 3.63 per cent of the total output value of that year. Today, more than 90 per cent of the products are researched and exploited by the company itself. The company has its own intellectual property right and its output and sales have increased by 50–60 per cent. The quality of products keeps improving. To this kind of enterprise should be given a policy environment equal to that of foreign-funded enterprises. Every favourable policy that foreign-funded enterprises enjoy should also be given to this kind of enterprise. In addition, we should support it to realize asset reorganization, to perfect the share reform of pluralistic investment bodies, and to build a competitive environment that is equal to that of foreign-funded enterprises. Second, numbers of medium-sized and small state-owned enterprises are market-driving enterprises. The blind enlargement of their scale of production may cause too much production capacity, and vicious price com-

126 Industrial Clusters in the Tianjin Area

petition, and lead ﬁnally to their quitting the market. Government should carry on regulation, adjustment and reorganization on a case-by-case basis. As regards small-scale enterprises producing components and parts, if they have not been brought into the production unit of electronic components and parts, they may be sold so that those enterprises with some advantage can make use of venture capital risk to develop. The electronic information industry developing towards specialization, interconnection, internationalization and clustering through a combination of technology, institute and market Government can improve the technical ability of local enterprises that manufacture accessories through encouraging foreign-funded enterprises to supply drawings, training and other services. By offering favourable loans to enterprises, government may stimulate accessory manufacturing enterprises to raise their self-innovation ability so as to integrate with transnational corporations in technology. Government should stop its direct micromanagement of enterprises and turn to making policy, regulating industries, stimulating investment and constructing public facilities. In addition, it should speed up the marketing process of state-owned enterprises in order to realize integration with transnational corporations. To develop industries requires the analysis and correct understanding of the structure of both the domestic and the international market. Those low additional value products whose technology is mature and widespread should not only occupy the domestic market but also expand their market share internationally. Through accepting high-tech products from foreign countries, domestic enterprises should produce foreign famous-brand products and enter the international market with ‘Made in China’, stride forward toward the great goal of an international manufacturing base and realize market integration. With these three kinds of integration the electronic information industry in Tianjin can develop towards specialization, interconnection, internationalization and clustering. Advancing from promoting investment by favourable policy, sound infrastructure and convenient trade environment to promoting investment by improving the investment environment of enterprises in regional clusters In promoting foreign investment, the favourable policy whose main content is to reduce tax on enterprises and allow them to retain more proﬁts has played an important role in the earlier stage of Tianjin’s opening-up and reform. However, this method has increased costs and its appeal to foreign funds is weakened not only now that, foreign funds policies of various developing countries and regions tend to be liberalized, but also and especially because the favourable degree of policy to various domestic regions has increased. Taking global competition into consideration, transnational corporations have changed their forms of operation

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 127

around the world. First, many transnational corporations and consortia have transferred their production base of electronic information products to the Chinese mainland. They have set up research centres in China and realized the localization of R&D and service. Second, with the development of technology and the globalization of markets, enterprises should try their best to give all their attention to core competitiveness. As for activities outside of core competitiveness, enterprises should adopt specialized outsourcing. In order to make efﬁcient use of foreign funds and speed up the development of the electronic information industry in Tianjin, investment promotion policy must be adjusted according to the investment targets of transnational corporations. Changing from the competition of ‘initiating enterprises’ to that of ‘improving investment environment’, the electronic information industry in Tianjin will ﬁnally go on its way, from dependence to independence, through improving the soft and hard environment which attracts transnational corporations, including industrial chains, enterprise clusters and the enterprise commercial environment of regional economic bodies.

Rise of the green battery industry in Tianjin A ‘green’ battery is a new type of environmental protection battery, of high quality pollution-free, which has been put into use in recent years or is being researched and exploited. Green battery production is generally recognized as high-tech industry that has rich technological content and great market potential, and its development should be given priority. National Projects 863 and 973 listed green batteries and their relative materials as a key item and gave them full backing. The Tianjin municipal government considers new energy and resources (taking green batteries as the dominant factor) as one of the three pillar high-tech industries which will be developed as a key point. Here the pillar industry means the core industry in China’s economic development. Also for Tianjin New Technological Industrial Park new energy and resources are regarded as a priority ﬁeld. Basic situation of the industry In China, Tianjin has obvious advantages in developing the green battery industry. The industry in Tianjin has reached a certain level not only in R&D but also in scale of enterprise. It has a great variety of products and it has successfully opened up markets at home and abroad. Strong power of R&D and high ability of production So far, Tianjin leads in the research, development and production of green batteries in China. The Eighteenth Institute of the China Electronic Technological Corporation (original the Eighteenth Institute of the

128 Industrial Clusters in the Tianjin Area

Ministry of Information Industry), Nankai University, the Tianjin University, Tianjin Lishen Battery Joint-Stock Co. Ltd, the Tianjin Peace Bay Power Co. Ltd, the Tianjin Jinneng Battery Co. and so on represent the research and production level of the green battery industry in China. The Tianjin Peace Bay Power Co. Ltd, the Tianjin Lishen Battery Joint-Stock Co. Ltd, the Lantian Power Co., the Tianjin Jinneng Battery Co., the Lantian Sanyo Co. and some others are the key enterprises of the industry in Tianjin. Besides these, the Tianjin Tong Yee Industrial Co. Ltd is a peripheral enterprise. The Tianjin High-Tech Industrial Park is where green battery enterprises are concentrated. This region is close to many colleges and universities, such as Nankai University and Tianjin University, which have sound technological and teaching resources. Against this backdrop, Tianjin is actively pushing ahead with its program as the green battery base. The Eighteenth Institute of China Electronic Technological Corporation was founded in 1958. It is a comprehensive institute of chemical and physical power as well as a military electronic institute. It was the earliest institute in the national battery industry and has the largest scale and most complete kinds of products. It has successfully researched more than 30 series and about 400 kinds of power products that are widely used. The Tianjin Peace Bay Power Co. Ltd was founded in September, 1996. It mainly produces and sells the communication-use Ni–MH battery, the power Ni–MH battery and its downstream products. With overall capital investment of nearly US$80m., the company is equipped with fully computerized automatic Ni–MH battery production lines. The R&D centre, grouped with pairs of experts from various academic ﬁelds, keeps pace with the fast development of world battery technology. And through the achievement of ISO9001, the company demonstrates its commitment to quality. The company is the largest Ni–MH battery manufacturer in China. It is the industrialization model base (conferred by the Ministry of Science and Technology and the Ministry of Information Industry in January, 1999) of the Ni–MH battery. The company has the most advanced production facilities and production technology for the Ni–MH battery in the world and has formed a production capacity of 60 million small-sized Ni–MH batteries and 200 million A h Ni–MH battery plates per year. Peace Bay plays a leading role in the R&D for communication and power batteries. With its own intellectual property rights, the corporation has realized signiﬁcant technological innovation and is considered to have reached the world’s most advanced level. Moreover, the company is a main undertaker of the Ni–MH battery item in National Project 863. The corporation produces various types of Ni–MH rechargeable battery cells and packs, including types AA, AAA, 4/3A, D, F6, F6M, C1U cylindrical and prismatic cells and also 100-A-h, 50-A-h, 18-A-h, 8-A-h power use batteries and packs. These products provide excellent rechargeable and envir-

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 129

onmentally friendly power sources for the Walkman, digital cameras, camcorders, two-way radios, cordless phones, laptop computers, measurement instruments and meters, power use vehicles and e-bikes, and so on The corporation also produces disposable lithium cell batteries, Li-ion batteries, zinc–air batteries, nickel–zinc batteries (under development), battery power management circuits, battery chargers, e-bikes and so on. The Tianjin Lishen Battery Joint-stock Co. Ltd, in Tianjin Huayuan Hi-Tech Industry Park and with a total occupying area of 30,000 m2 and a total built area of 24,000 m2, is sponsored jointly by eight shareholders, namely the Tianjin Jinneng Investment Corp., the Tianjin Lantian Power Sources Corp., the Tianjin Huaze Co. Ltd, the Tianjin Communication Service Corp., the SDIC Electronics Company, the Tianjin Taimen Investment Consultation Co. Ltd, the Tianjin Hi-Tech Holding Group Co. Ltd and the Tianjin Science and Technology Developing Investment General Corp. Its registered capital is RMB 193m. and its total investment has reached RMB 720m. Taking the Tianjin Institute of Power Sources as its strong R&D support, Lishen is mainly engaged in the R&D and production of Li-ion batteries, and has become the leading LIB and LIP producer in China. Lishen was founded in December 1997. With technological support from the Eighteenth Research Institute, it carries on R&D and production of green high-energy batteries and is a Li-ion battery production company with the largest investment scale and highest technology level. Lishen has independently completed the R&D for polymer batteries and has a production line that can produce 300,000 polymer Li-ion batteries per year. So far, the whole production capacity of the company has reached 50 million Li-ion batteries, the products including polymer, soft-package and automobile-use power Li-ion batteries. Thanks to Lishen’s advantages of an industrialized scale for Li-ion batteries and product technology, it was elected as a chief unit for the ‘Promotion Committee of Mobile Telephone Battery Localization’, a subsidiary committee of the Information Industry Ministry. Entrusted by chief departments of state, Lishen drafted the ﬁrst ‘National Standard of Cell Phone-Use Li Battery’, which was put into practice in July 2001. In 1998, with approval from the State Development Planning Commission, the Lishen Li-ion battery industrialization project was formally launched with the aim of a 5 million annual production capacity of cylindrical cells. After successive three-phase investment and expansion, Lishen has realized an annual production capacity of 50 million cells. The advanced automatic equipment can make high-consistency products consisting of more than 30 speciﬁcations, ranging from cylindrical LIB to prismatic LIB, laminated LIB and LIP. The quality and performance of Lishen cells are on the same level as Japanese products, and some types of cells have exceeded their rivals. Meanwhile, Lishen pays much attention to

130 Industrial Clusters in the Tianjin Area

quality management, and has proudly obtained the certiﬁcations ISO9002, ISO14001, CE and UL. The success of these third-party certiﬁcations has paved a brilliant road for Lishen to march into international and domestic markets. So far, Lishen has realized volume supplies to transnational corporations like Motorola ESG, TDI and Samsung, making it an aggressive competitor to its Japanese counterparts. Lishen has been designated as the ﬁrst supplier by major Chinese mobile phone manufacturers and is steadily shipping cells to them. Founded in 1992, the Tianjin Lantian Power Sources Co. is a full funded company of the Eighteenth Institute of the China Electronic Technological Corporation. The company engages mainly in technological development and industrialized scale production of a new type of green environmental protection battery. In 2000, Tianjin Lantian Power Sources Co. began to build the production line that can produce 10 million power-driven tooluse Ni–MH batteries of high-powered Sc type. The lead–acid battery department produces 72,000 large and medium-sized maintenance-free lead–acid storage batteries and 260,000 small maintenance-free lead–acid storage batteries, which add up to 30,000 kW h. Thermoelectric cooling module products have reached an advanced level compared with the same type of products around the world. The company is China’s leading researcher on power materials. As the main sponsor, it founded the Tianjin Lantian Hightech Power Sources Joint-stock Co. Ltd in April 2000. The capital of this latter company is RMB 85.4m. It is a professional manufacturer engaged in the research, development, production and sales of rechargeable batteries, including Li-ion batteries, Ni–MH batteries, NiCd batteries and sealed lead–acid batteries. The company has more than 500 employees, with an R&D staff of more than 150, including 10 national specialists. All senior management personnel of the company have the Executive Master of Business Administration (EMBA) title. The company has ﬁled 6 patents and 6 patents are pending. Certiﬁed with ISO9001:2000, CE and UL, the company is trying its best to satisfy every customer with high quality, competitive prices and good service. The Tianjin Lantian Sanyo Energy Co. Ltd was founded in March 1999. Its total investment is US$34.60m. its registered capital US$14.64m. The company imported ﬁrst-class technology and production equipment for Ni–MH batteries from the Japan Sanyo Electrical Motor Co. Ltd. It mainly produces Ni–MH batteries with large capacity and high reliability. The annual production capacity of the company may reach 30 million batteries. The Tianjin Lantian Electric Source Co. was invested in solely by the famous Chemical and Physical Electric Source Institute – formerly the Eighteenth Institute of the Ministry of Electronic Industry. The institute is engaged in research, development and production of chemical and physical electric sources, and most of its products come up to the ﬁrst standard in China, with some of them reaching international standards. Its technical

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 131

competence in research, development and production of nickel–hydrogen batteries is on the ﬁrst footing in China. The Sanyo Motor Co. is a worldfamous transnational group. Its technical competence and technology equipment for the production of nickel-hydrogen batteries are world-class. The Japan Toyota Business Co. is a comprehensive enterprise under the Japan B Toyota Co., engaged mainly in business and various services. Being jointly invested in by these three parties, the Tianjin Lantian Sanyo Co. Ltd embodies the alliance of power and power. Utilizing the technical competence of investing parties, following the newest development of technical competence of nickel–hydrogen batteries in the world, the Tianjin Lantian Sanyo Electric Source Co. Ltd will produce high-quality products to meet the demands of our age. The Tianjin Jinneng Battery Technological Co. Ltd was jointly funded by the Tianjin Jinneng Investment Company, the state-invested high-tech pioneering company of the State Development Company, Tianjin Elec-Mech International Trade Inc. and Nankai University. Having imported ﬁrst-rate production and testing equipment from the USA, the company has a production capacity of 5 trillion watts’ worth of solar cells. Complete types of products and rudimentary industrial Chain In Tianjin, the production of green batteries is connected closely with market demand. Besides meeting the needs of the present market, green battery companies develop new products continuously. Meanwhile, the agglomeration effect of the green battery industry becomes more and more remarkable. Some auxiliary industrial items of green energy and resources have settled in Tianjin and an industrial chain is gradually being completed. According to the industrial situation, green battery products can be divided into three hierarchies. The ﬁrst includes communication-use liquid Li-ion batteries and polymer Li-ion batteries, Ni–MH batteries, maintenancefree lead–acid batteries and so on. These batteries have realized production in batches and have acquired some market share. The second grouping includes amorphic Silicon solar cells and power Ni–MH batteries; they are in the process of construction. The third includes Zn-air batteries, fuel cell and power-force Li-ion batteries, which are at the R&D stage. Besides the abovementioned, there are some accessories of the green battery industry, such as thermoelectric cooling modules and battery materials. The Lishen Company is a production enterprise for the Li-ion battery. Its product quality has reached the level of that of Japan, the world leader. Indeed, some products have even surpassed Japanese batteries. The Lantian Power Sources Company has researched and exploited a series of batteries of large capacity. The production line for the Li-ion battery can produce 3 million batteries per year. Enterprises producing Ni–MH batteries include Peace Bay, Tianjin Lantian Power Sources and Lantian Sanyo. The Ni–MH battery that was

132 Industrial Clusters in the Tianjin Area

independently researched and developed by Peace Bay has reached the world’s most advanced level and supplies an ideal power source for partly power-driven bicycles, fully power-driven bicycles, motorcycles, powerdriven cars and aviation. The production capacity of Lantian Sanyo is 40 million Ni–MH cells and 10 million batteries and its production level is the best in the world. Maintenance-free lead-acid battery. The Tianjin Lantian Power Sources Company’s 12 V, 3 A h battery has a 10-year history of exports to the USA and dominates the market for mowers. Its lead–acid battery with high power has been used in Beijing’s trolleybuses. In 2001, the annual production capacity of Lantian Power Sources Company was 110,000 and production value was RMB 23.4m. Now the company is preparing to extend its production scale so that its production capacity can reach 80,000 kW h. Thermoelectric cooling module. The technology for thermoelectric cooling modules is a kind of new refrigeration technology without any pollution or noise. The products of the Lantian Company have reached the world’s advanced technology level. The company has developed more than 40 series of products and 80 per cent of them are exported. At present, the annual production capacity of thermoelectric cooling modules is 500,000. Battery materials. The Lantian Power Sources Company has the production technology of cobalt that is necessary to produce LiCoO2. In battery materials, the production scale of cobalt has reached 100 tonnes per year. The company has developed and produced dozens of battery materials of the highest quality. And new projects the company is preparing to undertake include a high-quality, low-cost production line for lithium cobalt oxide, with an investment of RMB 20m. The ﬁrst phase of the project is to build a production line that produces 400 tonnes of lithium cobalt oxide per year to meet the needs of the domestic market. Vast range of market prospects and large development potential It is reported that the total market value for batteries and storage batteries was US$28.5bn from April 1999 to March 2000. Of this, the total market value for batteries was US$10.1bn, the total market value for lead–acid batteries US$11.8bn and the total market value of small-type storage batteries US$6.5bn. It is also reported that, in the small-type battery market, new NI–MH batteries, Li-ion batteries and polymer Li-ion batteries have increased by threefold in the 10 years since 1990. Moreover, the market demand of Li-ion and polymer Li-ion batteries will go on rising. Within China, the national economy is developing rapidly and the information industry is making giant strides. This has spurred the improve-

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 133

ment of chemical power sources within the country. Accordingly, new products are constantly emerging and the applied ﬁeld of battery power source is widening continuously. In particular, the research and development for wireless computing batteries is now in the ascendant. The domestic market demand for electronic facilities, such as mobile telephones and portable computers, is increasing. Especially with the development of the communication industry, the sales volume of mobile phones keeps rising. Correspondingly, demand for batteries, as an accessory, will also rise. Besides the accessory market, the capacity of the products that substitutes battery is still large because of the special national conditions. At present, China is a leading production and trade country for semiconductor materials and its market scale ranks ﬁfth. These factors have all contributed to an increased market opportunity for the green battery industry. According to the type and purpose of green batteries, products can be divided into small Li-ion batteries, small Ni–MH batteries, power-use Li-ion batteries, power-use Ni–MH batteries and maintenance-free lead–acid batteries. In the ﬁeld of green battery industry the products of Tianjin’s green battery industry have considerable competitive power. In 2001, the production capacity of cylindrical LIB and prismatic LIB cells was 6 million and 110 million, respectively. Their actual outputs were 600,000 and 18.8 million. In 2001, the production capacity of the Lishen Company reached 5 million cylindrical LIB cells and actual output was 500 thousand. The percentage of production capacity and the percentage of real output in the country as a whole were 83 per cent and 83 per cent, respectively, which ranked ﬁrst in the country. In the same year, the production capacity of the Lishen Company reached 10 million prismatic LIB cells and actual output was 1.13 million. The percentage of production capacity and the percentage of actual output as a percentage of the country as a whole were 9 per cent and 6 per cent, respectively, which ranked third in the country. In the period of the Ninth Five-Year Plan, the Tianjin Lantian Power Sources Co. undertook the ‘Research on the Power System of the Li-ion Battery for Electrically Operated Automobiles’ project, at the behest of the Ministry of Science and Technology. The company researched and developed 55-A-h Li-ion batteries and an knowledge management system. The project is cutting-edge and the Li-ion power researched by the company shows good overall quality and has a vast range of prospects. In 2001, the annual production capacity of the Tianjin Peace Bay Co. Ltd was 40 million batteries and actual output 22 million, which respectively took up 25 per cent and 18 per cent of those of the whole country. The annual production capacity of Lantian Sanyo Co. Ltd was 30 million and actual output was 22 million, which respectively took up 19 per cent and 18 per cent of those of the whole country. In output, both the Tianjin Peace Bay Co. Ltd and the Lantian Sanyo Co. Ltd ranked second.

134 Industrial Clusters in the Tianjin Area

The Sanyo Company is the world’s largest production enterprise of Ni–MH batteries. It is preparing to import six of the world’s most advanced production lines so that its output can reach 200 million in 2004 and its sales income can reach RMB 600m. Lead–acid batteries are the only kind of battery that is produced in batches for electrically operated vehicles. The technical level of the Tianjin Lantian Power Sources Co. is the highest in China. As already stated, the company’s high-efﬁciency lead–acid battery has been used in Beijing’s trolleybuses. Moreover, its 12-V, 3-A-h battery has some market share in the USA. Shortcomings and problems After 10 years of development, the green battery industry in Tianjin has a certain scale. However, as a growing industry, it still has some problems that demand prompt solutions: 1. The industrial chain needs further improvement and industrial clusters remain to be developed. The industry has not formed a rational enterprise group structure that is adaptable to market competition. The whole level of the industry and the emergence of the market need improvement. 2. Compared with international enterprises, green battery enterprises in Tianjin have inadequate R&D input and insufﬁcient technological innovation. 3. In spit of advanced equipment and a high degree of automation, production enterprises have a low degree of ﬂexibility of production. Fewer types of products, high costs and the slow speed in meeting market needs are the main reasons why production energy cannot be brought into full play. 4. In the long run, the production capacity of domestic battery materials is limited. This, plus quality problems, means that production is controlled by foreign suppliers and that generous proﬁts are earned by foreign battery material companies. Although the enterprises in Tianjin are producing batteries, there are a lot of problems that haven’t been completely solved. Industrial construction By 2005, Tianjin New High-Tech Industrial Park will be the largest green battery base in China and the green battery industry will have become a pillar high-tech industry. By then, Tianjin will have become a region where domestic high-energy battery enterprises concentrate most densely and types of battery production are most complete. By 2005, a group of highenergy battery industrial clusters, such as Peace Bay, Lishen, Lantian Power Sources, Jinneng Battery and Lantian Sanyo, will have been formed within

Table 3.4

Output and output value of main products in the green battery base in Tianjin (10,000/RMB 100m.)

Type of product

2001

2002

Output

Output value

Output

Output value

Li-ion batteries

1,500

0.25

1,500

Ni–MH batteries

4,561

3.1

7,852

Amorphic silicon

0

0

0

2003

2004

2005

Output

Output value

Output

Output value

1.0

5,000

9.55

5,000

9.55

6.9

12,150

0

2.5 MW

10.9 0.625

21,080 5 MW

19.0 1.25

Output 5,000 3,000 10 MW

Output value 9.55 31.0 4.0

batteries Source: Tianjin Statistical Yearbook 2003.

135

136 Industrial Clusters in the Tianjin Area

the base. Furthermore, the production capacity will have reached 50 million Li-ion batteries, 300 million Ni–MH batteries and 10 MW amorphic silicon solar cells. Sales income of leading products in the base will have reached RMB 5.7bn. and, including other battery enterprises and accessory manufacturing enterprises, this income will have reached RMB 10bn. The output and output value of leading products within the green battery base are as shown in Table 3.4. The product power targets for 2005, are for Li-ion batteries to have a market share of 30 per cent, for Ni–MH batteries will try to have a market share of 80 per cent and for maintenance-free lead–acid batteries a market share of 30 per cent. Regarding the amorphic silicon solar cell and power Ni–MH battery plant that is being built, the ﬁrst phase of the project should be shortened and the second phase should start as soon as possible in order to enable large-scale production to take place. In the case of the Zn air battery, fuel battery, power-use Li-ion battery and battery materials that are being researched and developed, trial production should begin soon so as to turn achievements into productive forces. Furthermore, a complete battery industrial chain, as well as a world-class R&D institute, should be built within the green battery base. The leading position of technology should be maintained in order to preserve development strength of the base. Tianjin should build up as an inﬂuential trade centre related to the battery industry, making it a growing point that can push sustained economic development. Industrial suggestions In order to promote the development of the green battery industry, both enterprises and related departments of government should take some active measures. Enterprises should do as follows: 1. Speed up the steps of development and set market position. Enlarge the scale of enterprises and try to occupy a market share of 30 per cent through cooperation, mergers and acquisitions in order to have the decisive right to speak for the industry. 2. Strengthen the technological development power and increase the central competitive power of enterprises. In order to be adaptable to the rapid changes of the market, enterprises should have comparatively strong research power. So far, main enterprises in the base have built the ‘R&D Centre’. In future, enterprises will still need to increase R&D input and try to keep their technological advantage. 3. Strengthen international cooperation, including developing joint ventures and cooperate on research. Improve technology of enterprises and gain the competitive advantages in markets at home and abroad. Establish effective institutes so as to attract, cultivate and keep qualiﬁed personnel.

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 137

4. Widen the channels for raising funds and strengthen the actual power of enterprises. Green batteries are to an industry that the country encourages to develop and advance. The enterprises in the green battery base should actively ask for special projects and seek support for government bonds to create favourable conditions for their development. The government should do as follows: 1. At municipal level, in both ﬁnance and organization, support enterprises in the base to develop fuel batteries. Government should also decide the research and development direction for fuel batteries and suggest R&D objects that focus on PEMFC and SOFC. 2. Encourage the development and production of upstream and downstream products for batteries. Adopt an open strategy and favourable policies that will attract battery material companies and technological development companies. Meanwhile, battery companies should also extend their products to downstream products, such as power-driven toys, power-driven bicycles and power-driven automobiles. 3. Provide services for investment and raising funds, services for personnel exchange and training, services for information network, services for technological property right trade, services for returned personnel exchange, services for technical consultation, services for social consultation and support services for government purchase. 4. Strengthen the connection base for the green battery industry, mobile phones, digital audio-visual equipment, electronic mobile commercial systems, economical cars and motorcycles. 5. Strengthen environmental protection legislation, protect the green energy industry, increase propaganda and report to green energy (mainly including Ni–MH batteries and Li-ion batteries) enterprises and encourage the use of Ni–MH batteries within society. 6. Departments of technological supervision, industrial and commercial management and customs should devote major efforts to attacking fake commodities, purify the market environment, regulate market order and standardize market activities in order to create a fair competitive environment for enterprises. 7. Provide support on equipment import tariffs, enterprise income taxes and project construction taxes and fees. 8. Build nationwide green battery trade centres and green battery joint research centres including research institutions and enterprises.

Development environment, characteristics and direction of automobile industrial clusters in Tianjin China has a comparatively high tax rate on imported automobiles, so its automobile industry does not have a high degree of participation in inter-

138 Industrial Clusters in the Tianjin Area

national competition. Since reform and opening up to the outside world, the development of the automobile industry has been promoted by foreign funds. From the world’s angle, China’s automobile industry is one of the industries with the highest degrees of internationalization. Accordingly, to probe into the development of the automobile industry in Tianjin we must begin by examining the domestic and international background. Domestic and international background of the development of automobile industrial clusters in Tianjin International background The automobile industry has a long life-cycle and will not decline. However, for it to stay this way, constant changes in both organization and technology are required. The following international background of the development of the automobile industry reﬂects this trend: 1. During the 1990s, mergers and the reorganization of American, European and Japanese enterprises have formed an international market structure of oligopoly. In that period, the world’s automobile industry formed the ‘6 + 3’ oligarch competition pattern in which American, European and Japanese enterprises occupied the leading position. Reorganization was not limited to automobile enterprises. Mergers and reorganization also occurred in the components industry so as to form the ‘16 + others’ pattern for components. Global large-scale reorganization take the chain of investment, R&D, production, purchase and sales of the automobile industry across borders to realize optimized distribution throughout the world. The design and exploitation of products no longer simply copy conform to domestic modes and begin to meet diversiﬁed needs. This global adjustment has completely changed automobile industry strategy from scrambling for markets in developed countries to scrambling for markets worldwide. Therefore, the development of an automobile industry in any developing countries faces serious challenges. 2. The main players in the auto industry scramble for markets in developing countries in the form of exports and direct investment. Seen from the standpoint of the global automobile consumption market, we ﬁnd that the economic situation in the USA, EU and Japan is not optimistic, and that consumption is generally in a slump. In sharp contrast to that, the prospects of the automobile industry in developing countries are rising. Especially in some Asian countries such as China and Korea, automobile sales keep increasing. Take China as an example. In the past ten years, the average annual growth of the domestic automobile market was 15 per cent, of which the average annual growth of the car market was 24 per cent. At present, the average proﬁt rate of the automobile manufacturing industry is about 10 per cent. Therefore, to enter

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 139

3.

4.

5.

6.

rapidly into this expanding auto market has become the common strategy of the transnational automobile oligarchs. This has created much ﬁercer competition among the main players for the markets of the developing countries. The scramble is expressed in the form of investment and trade. Moreover, directly exporting automobiles and intermediate products to developing countries is another way of entering the market. The automobile industry has developed into an industry with typical inner trade characteristics. Global reorganization and the scramble for markets in developing countries, which are led by the transnational automobile magnates, have pushed the automobile industry to the present typical inner trade characteristic. The rapid development of trade within transnational corporations forces more and more intermediate products to enter become the mainstream of the automobile trade and makes the ‘nationality’ of automobile products vague. The automobile industry uses its ability of continuously exploiting products to prolong its ‘mature period’. Strategy reorganization and the operation of trade and investment cannot ensure proﬁts are earned continuously. In order to ‘thrive and never to decline’, the automobile industry must continuously carry on R&D for new products. At present, the funds needed for R&D on new products are increasing, mainly for the following reasons. First, diversiﬁed market demand and increasingly ﬁerce competition have spurred the pursuit of new products; it is the introduction and exploitation of new products that usually decides success in competition. Second, regulations on environmental protection, safety and energy saving are more rigid in many countries. Development and integration with new technology has become a new focus of industrial competition. The exploitation of new products is expressed by constant integration with new technology. Another such focus is the automobile industry introducing the latest achievements in high-tech ﬁelds, such as new materials, new energy and micro-electronics. Meanwhile, by means of exploitation, the international automobile industry makes use of information to carry on technological cooperation. The system of the international supply chain is increasingly open. In order to lower the costs of production and R&D, the international automobile industry has begun to abandon the traditional organizational form of the vertically integrated division of labour to reduce the production and stocking of components and parts by a large margin, to increase outsourcing and to realize the global purchase of components. The partner relationship of the vertical integration of the division of labour between automobile enterprises and component enterprises has gradually been displaced by a network organizational mode that regards a contract as a tie. Thus the international supply chain is increasingly open and is realizing global distribution.

140 Industrial Clusters in the Tianjin Area

Domestic background The history of the development of China’s automobile industry is not short. However, almost all the signiﬁcant changes took place after reform and opening-up to the outside world, even in recent years. Furthermore, it is expected that China’s automobile industry will experience more radical changes in the future. The trends of the automobile industry are as follows: 1. With China’s entry into the WTO, the protected position of the automobile industry has gradually weakened and its degree of participation in international competition has risen. Protections on imports and exports, retail, ﬁnancial services, entry restrictions and other aspects will be cancelled or lessened, and the industry will have to face ﬁerce international competition, a rigorous test for its enterprises. 2. The market structure of a group oligarchy that takes transnational automobile enterprises as the centre is being formed. After China entered the WTO, ‘acquisitions and mergers’ in China’s automobile industry surged forward and the dynamics of reorganization are unprecedented. Through reorganization and integration at home and abroad, international automobile manufacturers’ 6 + 3 and component manufacturers’ 16 + others entered China in all aspects. The group oligarchy has what we call the ‘3 + 9’ pattern, being headed by three giant groups. 3. With economic growth and development, the formation of a high- and medium-income class has cultivated a rapidly growing automobile consumption market, pushing the development of the industry. In addition, policies such as tax and fee reduction or exemption and automobile credit will appear. All this has led both transnational automobile enterprises and domestic automobile manufacturers to have optimistic expectations of future economic growth. 4. Competition between local governments on policy that fosters the automobile industry has become ﬁercer. As the resources of the production, purchase and selling network of the automobile industry achieve global distribution, local governments grow into major players. The suddenly formed 3 + 9 pattern of China’s automobile industry cannot do without the support of local governments. To lay a solid foundation, various regions are competing in setting out automobile industrial policies. They regard the automobile as a pillar industry and suggest cultivating and developing component industrial clusters. 5. The market for automobile accessories continuously raises its nationalization rate. At present, China has more than 700 foreign-funded enterprises manufacturing automobile accessories. Almost all the world-famous enterprises manufacturing automobile accessories have built joint ventures or foreign-funded enterprises in China. Among them, the nationalization rate of leading automobile brand names has been above 80 per

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 141

cent and the rate of the nationalization of cars has shortened from 6–8 years to 3–4 years. Development characteristics of automobile industrial clusters in Tianjin The automobile industry in Tianjin is an important part of the national industry. Especially in the twenty years since reform and opening-up to the outside world, the industry in Tianjin has both risen and fallen. It is now facing the development opportunity of starting an undertaking once again. After decades of development, automobile industrial clusters in Tianjin have appeared in an embryonic form and taken on development characteristics. Mastering the development characteristics is the prerequisite of cultivating sustained competitive advantages for the automobile industry in Tianjin. The development characteristics of automobile industrial clusters in Tianjin are as follows. The accumulated development basis of the industry forms advantage of human resources. The mandatory distribution mode of resources in the period of China’s planned economy made Tianjin an important machine-building base. Both the ﬁrst tractor and the ﬁrst mini station wagon in the history of New China were born in Tianjin. Also, Tianjin formed a production system of components and parts in order to coordinate with the production of automobiles. In particular, Tianjin accumulated some technological and human resources, which played a signiﬁcant role in its becoming a production base of light-duty cars in the 1980s. In the special development environment of the country, the industrial development basis that has been formed in history is a two-edged sword. On the one hand, the close production system leads to too much dependence on a single strength (assembly production enterprises). This kind of dependence prevents the production system from having any dynamic competitive power. On the other hand, once the development environment changes, human resources, as the active main body in the development, may form a kind of innovative strength outside the original production system and thus become a main source of competitive advantage. The unique product positioning under the national industrial policy has gained some pre-emptive advantages The development of the automobile industry in China shows obvious signs of government guidance. In particular, the state restrains the product positioning of enterprises through production entry restrictions. The development of the automobile industry in Tianjin began from the production of machines for agriculture and light-duty lorries and formed a distinct product dislocation together with the traditional FAW and Shanghai

142 Industrial Clusters in the Tianjin Area

Automobile Group corporations. In order to consolidate this kind of product dislocation, when Tianjin began the production of cars in the 1980s, the industrial policy of the state positioned enterprises in Tianjin to produce economical cars differing from the car enterprises such as the FAW and the Shanghai Automobile Group. The rapid development of the automobile industry in Tianjin from 1985 to 1995 is attributable to the increase in demand for economical cars. However, with the state loosening its restrictions on product positioning, economical cars have become a competitive focus in China’s automobile market. And the uniqueness of the product positioning of the automobile industry in Tianjin has begun to disappear. The industrial agglomeration that regards Toyota as the centre exists together with original industrial agglomeration The automobile industrial clusters in Tianjin are composed of two different parts. One is the Tianjin Automobile Group Corporation, a state-owned joint-stock company, including assembly and component enterprises of Xiali cars, Huali cars, Yanpai light-duty lorries and Sanfeng station wagons. The other is a foreign-funded enterprise cluster that takes Toyota as the centre. At the time Toyota planned to build automobile production enterprises in Tianjin, the main enterprises manufacturing automobile components and parts in the Toyota system began to enter Tianjin. Hence, as soon as with its Toyota entered Tianjin automobile production enterprises, the enterprise clusters of automobile accessories including tens of enterprises were rapidly formed around it. Enterprises of key components and parts, such as engines, air conditioners and seats have played a signiﬁcant role in enhancing the vigour of Tianjin’s enterprises manufacturing components and parts. What exists together with the new enterprise agglomeration of Toyota is the enterprise agglomeration within Tianjin Automobile Group. At present, it has a longer history and a more stable relationship and greater competitive pressure. The ﬁxed trade relationship between automobile enterprises and component enterprises has weakened the competitive environment for component enterprises. Therefore, when the pulling power of automobile enterprises is weakened, the enterprises manufacturing components and parts will land themselves in a predicament and have no power of inﬁltrating into other enterprises. This is also the reason why the scope of another enterprise cluster is gradually widened. Participation in the formation of a domestic oligarchic market structure through alliance with the FAW Competition within China’s automobile industry is fundamentally in the process of change from monopolistic competition to oligopoly. However, the usual market structure for the automobile industry in a developed

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 143

country is an oligopoly and the rate of market centralization is 95 per cent or so. Hence, with the increasing degree of market competition in China’s automobile industry, the general trend is to change toward the market structure of oligopoly, and for the existing automobile enterprises it is an advisable choice to participate in this. On 14 June, 2002, the signing ceremony for the joint reorganization of the FAW and the Tianjin Automobile Group Corporation was solemnly held in the Great Hall of the People in Beijing. The reorganization took the form of a share transfer. The Tianjin Automobile Group Corporation held 84.97 per cent of the share rights of the Tianjin Automobile Xiali Stock Co. Ltd. At the ceremony, the Tianjin Automobile Group Corporation transferred 60 per cent of the 84.97 per cent share rights to the FAW, namely 50.98 per cent of the total capital stock of the Xiali Company. Meanwhile, the Tianjin Automobile Group Corporation transferred all the 75 per cent of Chinese share rights that were held by the Huali Company, one of its subsidiary companies, to the FAW Group Corporation. The alliance of the Tianjin Automobile Group Corporation and the FAW Group Corporation made use of the tie of assets to connect the FAW Group Corporation and the Toyota Company of Japan. Thus the three parties participated in the competition in the formed oligarchic market structure as a whole. For Tianjin’s automobile industry, possible participation in the competition in the formed an oligarchic market structure is no guarantee of occupying a position in the competition. Only the dynamic competitive power of automobile industrial clusters can decide the future market competitive position of Tianjin’s automobile industry. To acquire this kind of dynamic competitive power requires hard work by enterprises and effective action by government. Because the industry is in the key stage of change from comparative advantage to competitive advantage, the competitive advantage of the industry is not outstanding. Comparative advantage is mainly expressed as low production cost, caused by resource endowments, geographical position and the availability cheap of factors production. Low production costs are the ﬁrst manifestation of industrial competitive power. Comparative advantage is easy to imitate and change and cannot be sustained for a long time. If comparative advantage can be said to be a kind of cost advantage, then competitive advantage is expressed as a kind of technological advantage. An enterprises can keep its sustained competitive advantage if its competitors cannot imitate its technology or if it can innovate faster than or at least as faster as its competitors can. The development of Tianjin’s automobile industry is attributable to its comparative advantage. Particularly under the effect of national industrial policy, unique product positioning has made Xiali cars the main brand

144 Industrial Clusters in the Tianjin Area

purchased by individuals in China. Economical cars are the main area on which competition in the future automobile market will focus and the bright prospect of market demand and fewer entry barriers attracts more and more enterprises to enter into their production. In addition, overall price competition directly attacks the price advantage of Tianjin’s products. All of these factors lead to the change from seeking comparative advantage to seeking competitive advantage. Whether this change can be realized will directly affect the competitive power of Tianjin’s automobile industry. Accordingly, the automobile industry in Tianjin takes cultivating technological advantage as the breakthrough and its change from comparative advantage to competitive advantage is in a crucial stage. The competitive power of enterprises manufacturing automobile components is in decline on the whole, but some enterprises suddenly come to the fore Relying mainly on the Tianjin Automobile Group, the enterprises manufacturing automobile components were very prosperous and had considerable strength in the production of accessories. In the middle 1980s, the market occupation rate of automobile components in Tianjin ranked second in the country and its proﬁts ranked third. As Germany’s Volkswagen, America’s General Motors Corporation and other companies respectively founded joint ventures in Shanghai, a group of enterprises manufacturing components rapidly emerged around that city. These enterprises began to grow in a comparatively tough market and the super-critical customers made them feel the market pressure. For Tianjin, however, the product positioning of the Tianjin Automobile Group has not played a positive role in improving the competitive power of component enterprises. In addition, no supercritical customers have applied pressure, so there has been no strengthen of industrial competitive power. The comparatively close enterprise clusters manufacturing components have no opportunity to accept incentives from the outside world. The innovative ability of the network system of enterprises has not been cultivated at all and the network system lacks any dynamic ability that is adaptable to the changes in environment. Correspondingly, the competitive power of enterprises has begun to decline. However, when the competitive power of the original enterprises manufacturing components and parts began to decline, some enterprises manufacturing components and parts outside the system were, with their high-quality products, inlaid in the system of the international supply chain, and successfully played the role in the supply chain of transnational corporations, gaining a sustained competitive advantage with their technology. At the beginning of their development, these enterprises had managed to squeeze into the relatively open supply chain which the transnational corporations founded in Chinese enterprises in order to seek development opportunities, just because they had no way to inﬁltrate the system of the supply chain of the Tianjin Automobile Group Corporation. This has, from

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 145

another angle, shown that a market environment with pressure is the essential condition for cultivating the competitive power of enterprises. The localization system for R&D of products has not been formed Generally speaking, Tianjin’s automobile industry has not formed its localized R&D system. Xiali cars, the main product of the Tianjin Automobile Group, were developed on the basis of advanced technology and production equipment that were imported from Japan. Owing to its insufﬁcient R&D ability and single type of product, the Tianjin Automobile Group has not been able to rely on product innovation to create new market value. Accordingly, the enterprises manufacturing components and parts in the system ﬁnd it impossible to have a corresponding R&D ability. Japan’s Toyota Company has brought improvement in production management and operation concepts to Tianjin’s automobile industry. However, the effect of technological spread has not appeared at the most crucial technological link. This may, on the one hand, be attributed to the conservative attitude of Japanese enterprises on technical transfer. On the other hand, it may be because Tianjin has not formed a localized R&D environment that takes qualiﬁed personnel and mechanism as a prerequisite. Enterprises and industries cannot gain sustained competitive advantage without their continuous improvement from the bottom of the value chain to the top. To establish a localized R&D system with competitive power is, beyond doubt, a way to indicate an intention to reach the top. The means by which government affects industrial development are facing the requirements of change In the process of cultivating industrial clusters with competitive power in a region, the function of government is to create a environment for regional development. Objectively speaking, there are two parts that are interrelated and at the same time separated in Tianjin’s automobile industrial clusters. One is the Tianjin Automobile Group, a state-owned stock company, which includes more than ﬁfty enterprises. Owing to historical evolution, the management methods of government have yet to adapt to the market. This part of the cluster is a closed system with a high ability to produce accessories. The vigour produced by institutional change has yet to be released in Tianjin Automobile Group. In particular, human resources are still limited by the system. However, with the alliance with the FAW Group, the closed system has begun to weaken and the management methods of government are facing the requirements of change. The other part of Tianjin’s automobile industrial clusters is the foreigninﬁltrated enterprise clusters centred on the Tianjin Toyota Company. As restrictions on national industrial policy are loosened, the degree of foreign inﬁltration is rising. Because of the special position of this part, the effect of government on it is close to the methods employed under conditions of a

146 Industrial Clusters in the Tianjin Area

complete market economy. Hence, in order to keep and extend the contributions of these enterprises to local economic development, government must start creating a regional development environment that can attract enterprises – that is, begin making a new economic platform. Construction direction of automobile industrial clusters in Tianjin The above description of the development characteristics of automobile industrial clusters in Tianjin is an objective account of the present situation. Judged by the standard of competitive power and especially by the standard of constructing industrial clusters with dynamic competitive power, Tianjin’s automobile industrial clusters have many aspects that need to be improved. The construction directions are as follows. To construct two interrelated systems of value chain As the value chain of the industry becomes increasingly open and international, industrial competition has changed from competition between enterprises to competition between value chains. The capability of being inlaid in a value chain with international competitive power is an essential prerequisite to measuring and improving the competitive power of enterprises. To take the present situation of Tianjin’s automobile industry, two interrelated systems of value chain may be constructed. These are the system of value chain in which the Toyota Company is the main part and the system of component value chain that faces large international corporations. As the value chain of the industry becomes increasingly open and international, transnational corporations have become the central part that maintains the value chain. Toyota Company is an international tycoon enterprise in the automobile industry. So far, part of its value chain has been extended to Tianjin. This is a rare opportunity for Tianjin’s enterprises manufacturing components and parts. However, the system of value chain of which Toyota Company is the main part has not been completely formed. It has two kinds of manifestation. On the one hand, the length of the value chain is limited. In particular, the activities that can produce great value have not appeared at all. On the other hand, only a few local enterprises manufacturing components can be inlaid in the value chain in which Toyota is the main part, and they tend to be further reduced. What exists together with the system of value chain in which the Toyota Company is the main part is another system of component value chain that faces the large international corporations. So far, the latter chain has not been formed and owing to the generally weakened competitive power of Tianjin’s enterprises manufacturing components. From the point of view of cultivating automobile industrial clusters, these two kinds of system of value chain supplement and complement each other. Enterprises are the main part in constructing the system of

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 147

value chain, and innovation is the essential power in constructing the value chain. For government, maintaining the development environment favourable to innovation is more or less manifestation of its functions. In the above two different value chains, the functions of government are different. In the value chain in which Toyota Company is the main part, government should bring its function of being an information centre into full play. In the component value chain which faces large international corporations, the function of government is to create a looser development environment for privately owned enterprises and medium- and small-sized enterprises, so that environment is no longer a barrier to the development of enterprises. To further participate in the competition of oligopoly market structure in the form of the whole industrial cluster Nowadays, market competition is gradually changing from competition between enterprises to competition between value chains and even competition between industrial clusters. The number of oligarchs in the market structure is not purely the number of enterprises. Instead, it is the number of enterprise alliances that appear in the form of value chains and industrial clusters. If Tianjin’s automobile industry wants to occupy a certain position in the market structure of oligopoly in China in the future, it must participate in ﬁerce market competition in the form of the industrial cluster as a whole. At present, the automobile industrial clusters in Tianjin have several factors that are favourable to improving competitive power. One is the existence of Toyota, another is the accession of the FAW Group and a third is development advantages such as human resources, which are accumulated by enterprises like the Tianjin Automobile Group after a long period. Of course, these favourable factors are not necessary to bring overall competitive advantages. Institutional restrictions may be a barrier and competition within industrial clusters is also an essential condition. Therefore, from the angle of government, to loosen institutional restrictions is a prerequisite in order to facilitate the overall relationship of competition and cooperation between automobile industrial clusters. Medium-scale and small-scale enterprises with quality management and innovative ability are an important factor in maintaining the competiveness of the value chain The automobile industry by its nature contains a division of labour. In order to improve the quality of the automobile, the quality of components, raw materials, production machinery and equipment must be improved at the same time. In countries with a developed automobile industry, the medium- and small-scale enterprises producing components and parts that have competitive advantages play a signiﬁcant role.

148 Industrial Clusters in the Tianjin Area

The value chain composed of large numbers of medium- and small-scale enterprises with quality management and innovative ability is the kind of value chain that can retain dynamic ability. In order to construct such value chains, Tianjin needs to take institutional changes and development environment optimization as the breakthrough. Through institutional changes, the potential of enterprises can be spurred onwards under the pressure of market competition, and enterprises can bring into play the full effect of the survival of the ﬁttest. For those medium- and small-scale enterprises manufacturing components outside the present system, optimizing the development environment is crucial to improving their competitive power. At present, though those enterprises are not numerous, they have a fairly high technological ability and quite a strong innovative ability and can participate in the system of supply chain of large international corporations. To promote the change of value chain from comparative advantage to competitive advantage and even to dynamic competitive advantage There is no need for reticence in recognizing that the market competitive power of Tianjin’s automobile industry comes mainly from its comparative advantage, namely the price advantage caused by low costs. However, the advantage of low costs and low price is the advantage that is the easiest to displace and surpass. Hence, for Tianjin’s automobile industry, it is extremely urgent to realize the change from comparative advantage to competitive advantage caused by technology and brand. At present, Tianjin’s automobile industry has some favourable conditions for creating competitive advantage. The ﬁrst is that its advantage of the commodity brand is increasing. With the settlement of Toyota and the alliance of the FAW and Tianjin Automobile, the product structure of cheap automobiles that was represented by Xiali has been changed. The value of Toyota’s international brand and the intangible assets of the FAW must be further exploited. The automobile industry of Tianjin should make use of the high quality these brands have to form the quality requirements for those enterprises manufacturing components so as to force the enterprises to change from seeking cost to seeking technology and quality. In this process, those enterprises should create the brand effect for components and also use brand to enhance the improvement of technical ability. Next comes getting the automobile industry to acquire the competitive advantage of technology. In the present situation, it is not realistic for enterprises to complete this kind of improvement by themselves. Accordingly, government must bring into full play its corresponding functions to create a mechanism favourable to technological overﬂow. For industrial clusters, dynamic competitive advantage is a kind of manifestation of higher ability. It takes comparative advantage as the prerequisite. However, from the angle of industrial development in the long run, it is about making sure to research the dynamic ability that makes industrial clusters

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 149

adapt to changes in the development environment at the same time as creating competitive advantage. In this aspect of cultivating industrial dynamic competitive advantage, government is able to develop its ability to the full. Government plays a positive role in cultivating industrial dynamic competitive advantage The dynamic competitive advantage of industry refers to the ability by which regional industry can adapt itself to changes in the development environment and keep its sustained competitive advantage. The main embodiment of this kind of advantage is whether to form a regional innovation environment. In creating a regional innovation environment, government has full scope for its talents. Government functions can be divided into two types. One is to invest in building a good regional transport system, an information communication network, a platform of information exchange between enterprisers and an attractive natural and social environment. Regarding the automobile industry, Tianjin should consider guiding enterprises to forming open, non-governmental associations and studying the methods of connecting enterprises with technology, like those the Toyota Company has formed in Japan, in order to push forward technical communication and spread it among enterprises. The other is to provide data and information services for enterprises and industries, to perform legal supervision, to establish industrial regulations, to enhance the combination of government, production and study, and to develop education to supply the qualiﬁed personnel needed by the industry. Regarding the automobile industry, long-term plans for automobile fairs that are just in their beginning stage should be made. Government should promote the foundation of the Tianjin Toyota Automobile Research Institute, improve ability for cultivating present qualiﬁed personnel in automobile technology, provide special fund support for R&D activities for medium- and small-scale enterprises and offer measures supporting the utilization of academic achievements in colleges, universities and research institutes. In brief, in cultivating industrial dynamic competitive advantage, government should ﬁrst make clear its positive attitude and then take practical measures to ensure its inﬂuence.

Analysis of the formation and development of Tianjin’s bicycle enterprise clusters Appearance of the problem Tianjin is the birthplace of the bicycle industry in China. In the 1970s and 1980s, Tianjin’s bicycle industry was a very prosperous one and became the pillar industry of Tianjin. From 1950 to 1990, it realized proﬁts of RMB2.42406bn in all and tax of RMB1.68681bn, in total RMB4.11097bn. Tianjin’s bicycle industry made great contributions to the economic devel-

150 Industrial Clusters in the Tianjin Area

opment of Tianjin. However, with the development of the reforms, of the early 1990s, the state-owned bicycle enterprises were landed in a predicament and began to go into deﬁcit. The goods were piling up in stock and could not be sold. The state-owned bicycle enterprises stopped production and the employees were laid off and had to sell their assets. In sharp contrast to state-owned enterprises, privately owned bicycle enterprises developed on the basis of supplying accessories to state-owned bicycle enterprises. By the end of 2000, there were 655 bicycle enterprise. Among these 655 bicycle enterprises, there were more than 283 privately owned ones that had a 74.67 per cent market share. The annual output of bicycles by those privately owned enterprises was 13 million, or more than 25 per cent of the total national output. After Guangdong and Jiangzhe, Tianjin was third in bicycle production. Tianjin’s privately owned bicycle enterprises have, once again, pushed forward the economic development of Tianjin. In spatial distribution, a new form of organization has appeared, namely, the specialized enterprise agglomeration in which medium- and small-scale enterprises for the main part have appeared spontaneously in suburbs. These medium- and smallscale privately owned enterprises can be found in the districts of Xiqing, Dongli, Wuqing, Beichen and Dagang. Eight enterprise clusters have appeared in embryonic form. Through specialized division and cooperation, these enterprise clusters have not only improved the competitive power of Tianjin’s bicycle industry but also have effectively solved local employment problems. The historical process from rise to fall and to rise again forces us to further consider some questions, including how to improve regional competitive advantage through increasing the innovative ability of local enterprise clusters, how to make use of the advantage of agglomeration to realize the structural escalation of local industries, how to make different enterprises in the same industry gather together to bring the clustering effect into full play and improve industrial competitive power through division and cooperation, and how government should position itself during the process of fostering medium- and small-scale enterprises and in creating a regional environment of innovation. This part of the chapter attempts to make a preliminary probe into the above questions through positive research and theoretical analysis, so as to provide new ideas for the economic and social development strategy of Tianjin. Case analysis: development of bicycle enterprise clusters in Wangqingtuo Town Formation and characteristics of bicycle enterprise clusters in Wangqingtuo Town Wangqingtuo Town, which is famous for bicycles, is situated in Wuqing district, Tianjin City. Between 80 and 90 per cent of the local workforce are

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 151

employed in occupations that relate to bicycles. The bicycle industry in Wangqingtuo began in 1994. At that time, there were only a few bicycle assembly enterprises in the whole town and there were no enterprises that produced components and parts. However, by 1995, there were more than 20 enterprises that produced 1.2 million bicycles in all. With a workshop production mode, low-grade accessories, a ﬂexible sales mechanism and low price, these enterprises grasped the opportunity to alter the original situation and ﬁnally displace many well-known state-owned bicycle enterprises. In the bicycle world, this phenomenon is called the ‘Tianjin Phenomenon’. In order to promote the development of the privately owned economy in the bicycle industry and to regulate competitive activities, Wangqingtuo founded ‘two associations’ (the Administrative Commission of the Privately Owned Economic Area and the Association of the Bicycle Industry) and ‘two centres’ (the Service Centre of Quality Management of the Bicycle Industry and the Service Centre of Bicycle Accessory Supply). Through them, the function of government and sectors can be brought into full play. Service promotes development and management ensures quality. To date, the bicycle industry in Wangqingtuo has entered a new stage in which it competes with others in brand, input reform and development speed. The bicycle enterprise clusters in Wangqingtuo have the following characteristics: 1. Most bicycle enterprises are family enterprises and not large-scale. These family enterprises belong to individuals and have clear property rights. 2. These enterprises, being individual, lack competitive power. However, as a whole, they have become a bicycle and component production base that is quite inﬂuential in Tianjin and even in the whole country. 3. These small enterprises that are independent and meanwhile interrelated have formed an industrial chain that is difﬁcult to imitate and copy. This kind of industrial chain based on the division of labour and cooperation can produce new productive forces, overcome the shortcomings of the weakness of individual enterprise and improve overall scale strength. 4. The bicycle enterprise clusters have displayed of characteristics dynamic change. Because of fewer barriers to entering or leaving clusters, the clusters in Wangqingtuo are in dynamic development. In recent years, competition in the domestic bicycle industry has been very ﬁerce and proﬁt margins have become smaller and smaller. However, the bicycle industry in Wangqingtuo continues to develop rapidly. Tracing the source, we ﬁnd that it is the production organization mode of the bicycle industry network that supports its rapid development (Figure 3.2).

152 Industrial Clusters in the Tianjin Area Figure 3.2 Local production system of bicycle enterprises in Wangqingtuo Town, Tianjin City Transport industry

Raw material suppliers

Firms producing components

Bicycle firms

Customers

Agents of distribution flywheels

Industrial associations and administrative centres

Local governments

Analysis of the formation factors and development mechanism of bicycle enterprise clusters in Wangqingtuo 1. Condition of location. Internal relationships of economic activities and the external economic pursuit of agglomeration make the regions which have won in the initial location competition more attractive to other enterprises and thus lead to the spatial agglomeration of the same enterprises. Once this kind of agglomeration happens, it will self-strengthening through circulation and accumulation. Wangqingtuo, 20 kilometres from Tianjin, is where the Jingtong Road meets the Jinbao Expressway and the supplementary route of the Jingjiu Railway. The geographical location on the outskirts of a large city makes it easy for Wangqingtuo to accept the resources, information and technology spread from Tianjin. The factory director of the Aosheng Bicycle Factory, the ﬁrst to apply for a state production licence, has made his fortune by assembling bicycles. In the period of the market monopoly by state-owned bicycles, he assembled bicycles whose brands were Red Flag and Pigeon. In the early 1990s, when state-owned bicycle factories were in decline, he saw a market demand for bicycles and built his own bicycle factory in 1994. However, unlike the Aosheng factory director who anticipated the pro-

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 153

spects within the bicycle market, most other bicycle factory directors are laymen in the ﬁeld. Never the less, they have developed their own enterprises by making use of their location advantage, technology, and the administrative background of original state-owned bicycle enterprises. The development of Wangqingtuo’s bicycle industry from 4 enterprises to more than 230 is attributable to its location advantage. 2. Factors of technology and market. Uncertainly over both market conditions and the technical path to follow have displaced the once endogenous and exogenous ﬁxed market conditions. Characterized and diversiﬁed market demand and technical improvement have placed the vertically separated mode of production and organization increasingly in the ascendent and produced a ﬂexible mode of production. The rigid mode of production and the vertically integrated mode of organization of Tianjin’s original state-owned bicycle enterprises lacked efﬁciency, effect and ﬂexibility. Thus, their disintegration and the rise of privately owned medium- and small-scale enterprises became inevitable, as the formation of the bicycle enterprise clusters in Wangqingtuo that have replaced them clearly shows. The development of Wangqingtuo’s bicycle enterprises have in turn promoted the production of accessories, while on the other hand the production of accessories has pushed the development of bicycles forward. Thus, a virtuous circle has been formed. 3. Factor of institution. The economic effect of a country is decided by its basic institutional arrangements, institutional structures, institutional environment and institutional trends. The formation of bicycle enterprise clusters in Wangqingtuo can be said to be the epitome of the institutional changes in the process of the economic development of China. Wangqingtuo’s bicycle industry began in 1994 when the marketoriented economic system in China was in development. At that time, the market system was not perfected and popular consumption demand was changing from simplistic product demand under a planned economy to diversiﬁed choice. Because people in Wangqingtuo have been aware of the diversiﬁed consumer demand for bicycles, they, with the original mode of production, have used their own position in the market, having completed their original capital accumulation, to start their own undertakings. If the market vacuum between a planned economy and a market-oriented economy can be said to have presented a crucial opportunity to the development of Wangqingtuo’s bicycle industry, granting production licences to bicycle enterprises have made it prosperous. The production of bicycles is not allowed without production licence granted by the Light Industry Bureau of the state. However, to acquire a licence takes a long time, that is at least half a year from the application. In addition, not all bicycle enterprises can achieve the requirements for a licence. Taking local interests

154 Industrial Clusters in the Tianjin Area

into consideration, local governments have adopted the granting of temporary licenses that ensure the production continuity of enterprises. The temporary period is usually from 3 to 6 months. During the period of the rise of the bicycle enterprises, our country has adopted an encouraging attitude toward the development of privately owned enterprises. Local governments have played a signiﬁcant role in promoting the development of privately owned enterprises, lowering operation costs within the privately owned economy and improving the outer environment for the development of privately owned enterprises. 4. Factor of social culture. The new economics of institutions holds that the institutional effect of ideology and cultural mentality plays a positive role in lowering trade costs, promoting economic development and maintaining social stability. The formation of bicycle enterprise clusters in Wangqingtuo has proved this. Wangqingtuo has a comparatively high terrain, sandy loam and no rivers. It has no favourable natural conditions for agriculture. Thus, people in Wangqingtuo are engaged in buying and selling agricultural products, selling eggs, manufacturing garments and producing bicycles. In this process, their sense of commodity has been strengthened. Furthermore, people in Wangqingtuo have formed their own network of social connections in the trade of commodities, and their economic activities are rooted in this net. Such rooting has played an important role in regional economic development. 5. Factor of capital. Here capital includes ﬁnancial capital, social capital and human capital. After years of development, Wangqingtuo’s bicycle industry has passed the primary stage of the original accumulation of capital and each enterprise now has the more or less idle ﬁnancial capital that is necessary for enlarging production and improving the industrial structure. Social capital refers mainly to social connections. Because of relationships with neighbors, relatives, friends or partners in business, managers of enterprises believe in each other to a large extent and often gather to exchange market information and to make market forecasts and judgments. This kind of belief and cooperation have become the basic condition which allows innovative factors such as knowledge and information to ﬂow rapidly in a region, and meanwhile have strengthened the regional accumulation of knowledge and information as well as the regional innovation function. This is an important condition, on which Wangqingtuo’s bicycle enterprises can quickly develop. Human capital in Wangqingtuo is expressed mainly as people’s sense of market, enterprising spirit and characteristic ability to bear hardships.

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 155

Shortcomings and problems 1. Wangqingtuo’s bicycle industry has insufﬁcient technical innovation within industrial clusters and less connection with colleges, universities and research institutes, so its innovation comes mainly from imitation and self-design. 2. Competition focuses on price. Many enterprises have not attached great importance to improving quality and perfecting style. This will lead to vicious competition and ﬁnally affect overall production efﬁciency and the outer scale economy of the industry. 3. The bicycle industry in Wangqingtuo lacks famous-brand products. It should, via setting an overall brand, namely a regional brand, combine this regional brand with a product brand to improve the market competitive power of local products. 4. The industry has not enough connection with other industrial clusters in Tianjin. This will lead to low-level repetitive construction and is not favourable to the overall development of bicycle industrial clusters in Tianjin. Wangqingtuo’s enterprise cluster should ﬁnd its exact position among Tianjin’s bicycle industrial clusters and develop itself in a planned way.

%

Bicycle enterprises

%

Distribution of Tianjin’s bicycle enterprises Components and parts enterprises

Table 3.5

Wangqingtuo

112

32.7

143

43.4

255

36.9

District of Wuqing

143

41.7

105

31.8

248

35.9

Beichen District

56

16.3

46

14

102

14.8

Dongli District

48

14

67

20.3

115

16.7

Dagang District

21

6.1

3

1.0

24

3.4

Xiqing District

32

9.3

36

10.9

68

10.0

Jinnan District

14

4.1

13

4.0

27

3.9

Hebei District

20

5.8

10

3

30

4.3

Other regions

57

16.6

33

10.3

90

13.1

Regions

Total enterprise

%

The above numbers are counted according to the latest statistical data from the Tianjin Bicycle Association in July 2002, so they are different from the data published in newspapers.

156 Industrial Clusters in the Tianjin Area

Development and structural features of Tianjin’s bicycle enterprise clusters In addition to Wangqingtuo’s bicycle enterprise clusters, Tinajin has formed bicycle enterprise clusters in Wangqinzhuang industrial park, Xinhou industrial park, Dabizhuang industrial park and Wangzhuang industrial park in Beichen district, Yangzhuangzi industrial park, the Xinbao Bicycle Company and Nanhe industrial park in Xiqing district, Dongli district and Dagang district. The distribution of bicycle enterprises is shown in Table 3.5. The enterprise clusters just listed have something in common with Wangqingtuo’s bicycle enterprise clusters. First, the vertically integrated, rigid mode of production in state-owned enterprises has not been adaptable neither to the diversiﬁed and characterized market demand nor to technical improvement. These enterprise clusters are, rather, the substitute for such a mode of production. Another is that these enterprise clusters are located in suburbs where it is easy to obtain a ﬂexible labour force. Also, the convenient conditions of transport and communication make it possible for enterprises to communicate with each other in commodities, service, information, labour force and technology. Second, medium- and small-scale enterprises with a relatively high degree of specialization have formed clusters, and the relationship between enterprises in the same industry or between related enterprises is one of dynamic competition as well as cooperation. This kind of relationship helps clusters to maintain their energy and competitive power. However, there are some differences between Wangqingtuo’s enterprise clusters and other enterprise clusters in terms of inner structure. Wangqingtuo’s enterprise clusters are in the primary stage of the development of enterprise clusters, that is, scattered agglomeration. As regards in inner structure, they belong to the form of Marshall’s industrial region. Within the clusters are purely medium- and small-scale enterprises. The scale of privately owned enterprises is small. The number of employees in several large enterprises is just 100 or so. The annual output of these enterprises is 200,000–300,000. Moreover, each enterprise focuses on a single function in the production chain. Their intermediate products are sold in Tianjin and the ﬁnal products are sold in the whole country. This kind of agglomeration of enterprises that belong to different links of the industrial chain and adopt family management has provided enterprises many opportunities of formal or informal face-to-face communication and cooperation. Meanwhile, it has effectively lowered the trade costs and material costs which are inevitable when enterprises seek production factors and market information. In contrast to this, other enterprise clusters are in the stage of core development and belong to the structure of ‘leading enterprises + network’. The so-called ‘leading enterprises’ refer to those enterprises that have a comparatively large scale, a comparatively high output, comparatively high

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 157

quality, comparatively advanced technology and some inﬂuence in the whole country. For example, the Fushida Bicycle Enterprise in Dongli’s bicycle enterprise clusters has adopted the group system of management, production and selling. It has carried on independent operation and ﬁnally entered the high-grade market. Fushida has become a shining star in Tianjin’s bicycle industry. The Jigang Bicycle Company, which is in the bicycle enterprise clusters in Nanhe industrial park, Xiqing district, is the ﬁrst enterprise in Tianjin to pass ISol9001 in 1997. At present, its annual output is 500,000, half of which are exported. Its bicycles, brand name ‘Black Horse’, are a famous brand in China’s bicycle industry. The Kelin Bicycle Co. Ltd in Dagang Economic–Technological Development Area is one of the top 100 enterprises in China. In 2001, its output reached 600,000 and its ‘Saike’ bicycles dominated the domestic market and were exported to Korea, Japan, Southeast Asia and Europe. The model effect and technical overﬂow effect of these leading enterprises has improved the overall strength of the enterprise clusters. The ‘leading enterprises’ play the role of horizontal support in production, vertical tie-in selling, and innovation and guidance in R&D for key technologies. Medium- and small-scale enterprises and family factories form networks and realize the socialization of production, interrelated organization and marketing circulation. On the one hand, these enterprises open up the market by themselves and on the other hand, they need to depend on the leading enterprises and networks within the region because of their limited market ability. They are independent of and at the same time dependent on each other. Analysis of competitive advantage of Tianjin’s bicycle enterprise clusters The changes in competitive power of Tianjin’s bicycles can be reﬂected, to some extent, by changes in the output and export volume of bicycles. From Tables 3.6 and 3.7 and Figure 3.3, we can see that Tianjin’s bicycle industry has experienced a tortuous course of development in both output and export volume. Output rose from 2.2517 million in 1978 to 5.6317 million in 1985. However, the percentage in the total output of the whole country fell from 26.4 per cent in 1978 to 17.4 per cent in 1985. In addition, the export volume also kept falling. In the 1990s, the annual export volume Table 3.6 Changes in the percentage of the output of bicycles in Tianjin compared with that in all China Year

1978

1980

1985

1990

1995

1998

2000

Output (10,000)

225.17

293.24

563.17

536.69

576.36

310.6

1300

26.4

22.5

17.4

17.1

12.9

13.4

%

44.1

158 Industrial Clusters in the Tianjin Area Table 3.7

Changes of annual export volume of Tianjin’s bicycles

Years

1980

1981–88

1990–97

1998

1999

2000

2001

Export volume (10,000)

19

10

5

6.1

23

63

110

Sources: Survey Report of Tianjin Bicycle Association; Tianjin Statistical Yearbook; China Statistical Yearbook.

Figure 3.3

Tianjin annual bicycle production as a percentage of that of all China

50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% 1978

1980

1985

1990

1995

1998

2000

was 50,000 or so and the export value was not more than US$2m. In contrast, the bicycle industry in Guangdong, Jiangsu, Shanghai and Zhejiang developed rapidly and the export volume increased greatly. Hence, Tianjin was left behind. After 1998, the percentage in the total output in the whole country and in exports went up again. In 2000, the output of Tianjin’s bicycles reached 13 million, or 44.1 per cent of the total output of the whole country. At that time, Tianjin’s output ranked second in the country. The export volume of Tianjin’s bicycles rose from 630,000 in 2000 to 1.1 million in 2001. This gave Tianjin an important position in both the domestic and the international market. Tianjin is accordingly known as ‘the bicycle industrial base with the vastest range of prospects’. In 2001, the First International Exhibition of Bicycles, Assistance-Bicycles and Accessories of North China, held jointly by the Tianjin Bicycle Industrial Association, the National Quality Control and Test Centre and the International Exhibition Centre, achieved great success and Tianjin’s position as the ‘base’ in production and sales was beyond doubt. Tianjin’s bicycle industry has great economic energy and competitive power because of the formation of bicycle enterprise clusters that are composed of medium- and small-scale privately owned enterprises. These enterprises have both a comparative advantage (price ﬂexibility) and a

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 159 Figure 3.4

Competitive advantage of enterprise clusters

Direct economic factors: production cost, trade cost, varied products, regional sale and competitive position in market

Competitive advantage of enterprise clusters

Indirect economic factors: institutional advantage, regional innovation system and region of studying type

competitive advantage (characteristic ﬂexibility) when they participate in competition in the form of clusters. The competitive advantage of enterprise clusters is reﬂected in two aspects (Figure 3.4). Let us analyse the aspect of direct economic factors, shown uppermost in the ﬁgure. In Tianjin’s bicycle enterprise clusters, the specialized division of labour has improved production efﬁciency, and the relative concentration of geographical location has allowed enterprises to obtain materials from local sources. It has reduced purchase cost, storage cost, circulating funds and the cost for changing processing links. Meanwhile, the close location and the networking founded on the common basis of social culture have lowered trade expenses for enterprises and have at the same time saved the time and money enterprises have to spend in searching for production factors and market information. Moreover, this kind of interrelated organizational structure forms an outer scale economy so as to meet the diversiﬁed and characterized demands of the market. The industrial environment within clusters makes enterprises in clusters more adaptable to the market environment and its changes. With this kind of industrial environment, enterprises in clusters can make use of the feedback mechanism between the diversiﬁcation of products and the occupation rate of the market to gain market power. Seen from the aspect of indirect economic factors, shown below in the ﬁgure, we ﬁnd that the micro basis of Tianjin’s bicycle enterprises is basically non-state-owned enterprises or privately owned enterprises, which have the advantage of constituting a ﬂexible system. With the development of clustering, industrial organization is increasingly strengthened and networking for medium service is being perfected day by day. All of these factors lead to the formation of an overall coordination advantage for the clusters. The enterprises within the clusters get the relative market information about advanced technology, advanced accessories and equipment

160 Industrial Clusters in the Tianjin Area

supply much earlier than those outside the clusters. Within clusters, information, knowledge and best practice can spread rapidly in the region, and thus the creative ability of enterprises and institutes has been improved and a regional innovation system formed. The interrelated organization structure within clusters makes various knots in the network believe in each other and continuously accumulate knowledge through interaction and study. The region of studying type is thus formed. The regional innovation system and the formation of the studying type region increases the innovation ability of Tianjin’s bicycle enterprise clusters. From the above analysis, we conclude that the formation of Tianjin’s bicycle enterprise clusters is the result of many factors, such as geography, economy, society, institutions, technology, the market and history. It has not resulted simply from the action of enterprises, nor was it able to appear in a short time. Enterprises within the clusters have an obvious coordination effect. Once this kind of organizational structure is formed in a region, other regions have difﬁculty in copying it quickly. It should be considered a signiﬁcant factor in the continuous improvement of the competitive power of Tianjin’s bicycle enterprise clusters.The bicycle industry should catch hold of favourable opportunity to upgrade the industry arrangement by the formation of industrial clusters. We have, from the micro angle, analysed the formation and characteristics of Tianjin’s bicycle clusters and the competitive advantage those clusters have brought. Tianjin’s bicycle industry should grasp the opportunity of economic globalization and the opportunity that Taiwan, Europe, America and Japan transfer their labour-intensive industry to the outside in the industrial structure adjustment. Also, as regards industrial agglomeration, structural adjustment and outward development, Tianjin’s bicycle industry should also make use of the clustering advantage to improve its own level through constructing bicycle industrial clusters. Structure of Tianjin’s bicycle industrial clusters In terms of the development of the bicycle market, China is the biggest country. China is a vast territory and its economic development is uneven. Meanwhile, diversiﬁed demand has appeared in the international market, its trend increasingly obvious. Because Tianjin’s bicycle market is diversiﬁed, Tianjin’s bicycle industry should, according to different characteristics of various enterprise clusters, be directed rationally and develop coordinatedly in order to construct a comparatively complete interrelated organizational structure, that is, to form the structural mode of ‘large core enterprises + leading enterprises + network of medium- and small-scale enterprises’. The ‘large core enterprises’ refer to these world-famous bicycle enterprises. The ‘leading enterprises’ are local enterprises that are advanced in scale, technology and management. The ‘network of medium- and small-scale enterprises’ refers to a kind of network which is composed of

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 161

specialized medium- and small-scale enterprises and public institutions, such as local education and training institutions, research institutions, ﬁnancial institutions, institutions of medium service, industrial associations and government. Industrial clusters with this kind of structure have four functions: • Regional system of ﬂexible production – ﬂexible changes take place in production mode, mode of organization within enterprises and relationships between enterprises. • Region of studying type – various bodies of action including enterprises in the region, commercial institutes and research institutes promote regional development and innovation of enterprises through study and cooperation. • Regional innovation system – an innovation system is formed within the region and rooted in the local social and cultural environment. The interaction between the innovation system and the local social and cultural environment promotes the perfection of the self-regulation and innovation function of the regional production system. • Open and interrelated organization structure – specialized enterprises which are independent within industrial clusters depend on the connection of knots in the network and their common interests to form a network connecting the globe with the locality through the interactions between enterprises and close network connections outside the region. Some thoughts and suggestions Thoughts on high-tech industry and traditional industry When high-tech and knowledge-intensive industries are mentioned, we usually think immediately of the information industry, of the biotechnology industry and others. When the adjustment and escalation of industrial structure are spoken of, we think at once of the escalation from labourintensive industries to technology-intensive industries. And we know, Tianjin’s bicycle industry is considered a ‘sunset industry’. However, the rise of an industry with low technology makes us reconsider our attitude toward traditional industries. For the background of economic globalization, economic growth and employment in each region presents us with a more serious challenge. Does that stereotyped practice of developing hightech industries conform to reality? The rise of Tianjin’s bicycle industry should make us realize that there are no sunset industries but sunset enterprises. When the capital (we call it agglomeration capital) produced by enterprise agglomeration is added to traditional labour-intensive industries, the traditional industries at this time have become knowledge-intensive industries. The reason is that once traditional industries without high technical content agglomerate in some region, the improved specialist division of labour and cooperation ability have signiﬁcantly increased the technical

162 Industrial Clusters in the Tianjin Area

content of those traditional industries. By this means, traditional industries are gradually changed into knowledge-intensive industries with a certain technical level and their advantage of increasing returns is obvious. Therefore, instead of giving all of our attention to high-tech industries and industries with high value added, should we rathers, from a new angle of view, reconsider the position of traditional industries that can efﬁciently solve the problem of unemployment such as the bicycle industry? Is it a mistake for government to use its policy strength intervene in economic operation merely, to support the key industries and key enterprises which it subjectively considers promising, and to pay no attention either to superior industries produced by agglomeration, or to enterprise clusters which have established a sound pattern of cooperation? Thoughts on industrial parks The development of industrial clusters is a social process changing from production of Fordist type to specialized production. In this process, social capital is formed, institution is innovated and the importance of local government is greatly improved. Tianjin’s bicycle enterprise clusters are formed by non-governmental force. To this kind of enterprise cluster, government should give conscious guidance and support to promote their rapid growth and proliferation. Meanwhile, government should make use of the outstanding industrial characteristics of spontaneous agglomeration and should cultivate the kinds of factors and environment which industrial development needs to actively support the conception, implantation, gestation, birth, and growth of specialized industrial clusters. Further, government should promote the perfection of such factors and environment through the development of industrial clusters. In fact, the construction of high-tech industrial parks and economic–technological development areas is a kind of means by which industrial clusters can be developed. However, the facts prove that it is often not successful. The reason, we think, is as follows. The vitality of industrial clusters lies in their rooting and innovation ability. When foreign technology and enterprises cannot be rooted in the local historical, social and cultural environment as well as the institutional environment, their innovative ability will inevitably be restricted. When we shift our view from the construction of high-tech industrial parks and economictechnological development areas to the construction of industrial parks with distinctive features, we have, in fact, realized that characterized regions have the greatest potential. At present, Tianjin is building industrial parks. Industrial parks are the extension of the concept of enterprise clusters. Hence, when we choose industries for industrial parks, we should draw lessons. Those enterprises that cannot be rooted in a region should not be chosen, and we should pay attention to industries with distinctive features. From our survey, we have found that each district and each town is con-

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 163

structing industrial parks and they have formulated various kinds of favourable conditions and measures to attract new enterprises. However, they must avoid getting into the prisoner’s dilemma, where various regions compete to attract enterprises only to ﬁnd that the value brought by those enterprises that cannot be rooted in a region is not as great as that brought by local enterprises. To construct enterprise groups and to develop medium- and small-scale enterprises The rise of Tianjin’s bicycle industry is to be attributed to the formation of groups of medium- and small-scale enterprises that are privately owned. However, in the process of our survey, we have heard two voices. The departments concerned hold that Tianjin’s bicycle industry has a scattered layout, disorderly brands and small scale, that this industry of Tianjin should construct an enterprise group to regulate its enterprise members and should unite the brands. Only in this way, they say, can the industry speed up its technical improvement, promote the adjustment of product structure and strengthen its market competitive power. The other voice is that of most privately owned enterprises and industrial associations. Their views are different from the above. Through practical survey and analysis, we think that when constructing enterprise groups we should realize that resources have their own price and in the market-oriented economy their occupation has to be paid for. If resources cannot bring returns, they are a burden instead of a beneﬁt to the occupiers. Therefore, enterprises need to pay costs when they extend their scale and occupy more capital. If returns brought by new resources cannot compensate for the costs of the new resources, this kind of capital expansion lowers the market competitive power of enterprises. This is why some superior enterprises feel a heavy load after a large-scale merger and expansion. Tianjin’s bicycle industry makes use of agglomeration of medium- and small-scale enterprises and constructs different forms of network to improve regional and industrial competitive advantage. We should pay more and more attention to this industry. Meanwhile, the development mode of medium- and small-scale enterprises in Tianjin’s bicycle industry has provided a new idea for the development of Tianjin’s medium- and small-scale enterprises generally – namely, that through agglomeration we can improve the overall production efﬁciency of medium- and small-scale enterprises, lower operation costs and display ﬂexible operations and scale economies so as to strengthen the competitive advantage of enterprises.

Conclusion: cultivating new type local industrial clusters in Tianjin An industrial cluster may have one of two typical forms: either a vertically integrated or vertically controlled cooperative system of enterprises or an

164 Industrial Clusters in the Tianjin Area

interrelated cluster composed of medium- and small-scale enterprises. As we have analysed, after the decades of the development of Tianjin’s industrial system, high-tech industries and technology-intensive and capitalintensive industries, such as electronic information, green batteries for environmental protection, and automobile and traditional industrial clusters, such as bicycles and garments appeared in embryonic from. On this basis, the characteristics and development trend of Tianjin’s new industrial clusters should be as follows. A developed interrelated industrial cooperation system in which foreign-funded enterprises are the centre Unlike the industrial clusters composed of medium- and small-scale enterprises in Taiwan, Guangdong and Jiangzhe, Tianjin’s industrial clusters take large foreign-funded enterprises as their centres. So far, in Tianjin, there have been more than 80 enterprises producing accessories for Motorola, more than 20 enterprises producing accessories for Samsung and tens of enterprises producing accessories for Toyota. Strictly speaking, this kind of industrial cluster is still in the hierarchy of the division of labour in the product chain and enterprises within clusters have not formed the interrelated connection of innovation networks. Local enterprises depend on core enterprises for funds and technology and this will lead to the fragility and instability of industrial clusters. On the whole, the present local industrial clusters are still at the bottom of or outside the global network and have not been embedded in the global production system. In view of the above facts, industries with a comparatively long industrial chain, such as electrical household appliances and automobiles, should, with the backing of an international supply chain and with the advantage of resources, attract transnational corporations to come to local regions to build industrial clusters which depend on each other instead of clusters which carry on vicious competition. In addition, those industries should also promote local medium- and small-scale enterprises to move towards key technical links. They should continuously improve themselves in their administrative systems, provide transnational corporations with powerful networks of accessory production, equipment maintenance and sales service, and lower their overall trade cost. Moreover, those industries should cultivate the developed networks in which enterprises can communicate and cooperate with each other, form local industrial clusters with a self-strengthening mechanism, realize the counter-escalation of industry, draw support from the expansion (such as external investment and exploitation of new products and new markets) of transnational corporations to embed themselves in the international industrial value chain and play a signiﬁcant role in the pattern of the international division of labour with their strong and stable industrial clusters.

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 165

To be good at study, imitation and innovation and to bring dynamic ability into full play Medium- and small-scale enterprises in technology-intensive industries that are represented by medicine, electronics, information and automobile accessories may, according to their own development strategy, form an alliance for technical R&D with other enterprises or the ﬁxed relationship of cooperation. They should imitate advanced technology and make innovations in order to lower the costs and risks of product exploitation. Through an alliance, medium- and small-scale enterprises may form scale economies, maintain and strengthen technical competitive advantages of their own and improve core and dynamic abilities. Compared with individual enterprises, industrial clusters can better reﬂect the essence of competition and are more capable of cultivating the competitive advantage of local industries. Competition is ﬁrst reﬂected in local clusters of the same kind, such as the enterprise clusters within the Tianjin Automobile Group Company and the foreign-funded enterprise clusters whose centre is the Toyota Automobile Corporation. Competition is then reﬂected in the competition between local industrial clusters, enterprise clusters or enterprises and those at home and abroad. For example, the enterprise clusters in Zhejiang, whose centre is Jili, imitated motors produced by Tianjin Toyota and thus caused competition. Last but not least, enterprises within clusters maintain moderate competition. This kind of orderly competition between enterprises within clusters has provided the outer motive force for the continuous innovation of enterprises. With the tie of market, industries in urban areas and industries in rural areas should be combined to form a rational structure of priorities So far, Tianjin has formed a pattern of industrial clusters with the characteristic of distribution according to priorities in urban and rural areas. The enterprise clusters in capital-intensive or technology-intensive industries, such as electronics, automobiles, high-precision machinery processing, new energy and resources, and the biomedicine and chemical industry gather in economic–technological development areas, free-trade zones, new technical industrial parks and some development areas in counties. Traditional industrial clusters including bicycles, garments and textiles have appeared in Beichen, Dongli, Xiqing and Wuqing, as well as in some industrial parks in villages and towns. Relying on new technology and new concepts of management, these traditional industrial clusters are being reformed into new traditional industrial clusters. From the angle of development conditions, this kind of distribution according to priorities is not rational. The industrial clusters of metropolis industry (non-polluting processing industry) have not been formed and some problems still exist in the above two kinds of clusters. Some industrial clusters are not spontaneously formed by the market. Instead, they are

166 Industrial Clusters in the Tianjin Area

formed by administrative orders or by direct transplantation from foreign countries. There are even some clusters whose members simply settle themselves down in parks with no regard to gains and losses. These clusters have no specialized division of labour, no scale returns and no ﬁxed industrial characteristics. Labour-intensive industrial clusters, such as bicycles and garments, are mainly a group of privately owned medium- and small-scale enterprises. Although they regard the market interests as a tie, they are short of a platform for product innovation and R&D. Nowadays, characterized and diversiﬁed market demand has become the mainstream. If these clusters have no high-level plans and arrangements for product design, technological process and qualiﬁed personnel training, they face the pressure of competition R&D and innovation day by day. In central urban districts, tertiary industries such as commerce, catering and real estate are the main body in most of the industrial function areas. Industries have in part formed areas where commerce and business concentrate. However, industrial clusters of metropolis industry are lacking in central urban districts. In addition, the service industries for producers, such as logistics, accounting, lawyers, consultation and so on are not developed, not to mention agglomeration returns. These three different parts are a process of gradual development and cultivation and they should be improved in order to maintain the sustained competitive power of the industrial clusters of Tianjin’s industry. Moreover, these three clusters are different from the usual scattered structure. Instead, they are united on the global market and form interrelated industrial clusters which can extend to the world. Support by a developed international supply chain and market network In the system of a developed international supply chain, upstream and downstream enterprises cooperate in production, goods supply and operation. They divide labour and share information as well as risks, seeking a competitive advantage that cannot be displaced in the ﬂuctuations of the industrial environment. However, this kind of ideal level is difﬁcult to reach. It’s reported that only 36 per cent of American enterprises are willing to share information on production and selling and information on demand changes with their cooperative partners. The situation in China is more serious. If self-interest and risk are taken into consideration, enterprises are usually not willing to frankly exchange information on production and selling. Besides, they lack the basis on which to believe in each other and instead cooperative partners suspect each other. All of this means that upstream suppliers are much too optimistic or pessimistic about downstream suppliers, product supply falls short of demand or else exceeds its

Zhang Donggang, Xie Siquan, Bai Xuejie and Luo Ruoyu 167

and so the whole supply chain loses its market competitive advantage and this ends up with the overall cooperative relationship of the alliance of production and selling in a state of disintegration. Cooperation in the international supply chain is the best choice for entry into the international market. However, the ﬁnal goal is not the internationalization of enterprises. The ﬁnal goal of an enterprise is to occupy market share and gain proﬁts. The market network can collect rich market information for enterprises and provide unblocked channels for selling. Hence, enterprises must build complicated global production networks, lower production costs by basing themselves on local resources, keep close to market demand, support relative assisting industries, perfect the local logistics network which is composed of operation, storage, delivery and information, and meanwhile extend their sales network to the world. The positive pioneering concept and cluster culture A distinguishing characteristic of an industrial cluster is its rooting in local areas. This kind of rooting is usually difﬁcult to copy. Thus, clusters are usually close and exclude people, things and enterprises that do not belong to their group. This kind of closed culture restricts the growth and expansion of industrial clusters. Therefore, Tianjin’s future industrial clusters of the new type should maintain and develop a coordinated and open clustering culture. In addition, future industrial clusters in Tianjin should incorporate a foreign advanced culture and cultivate compound competitive advantage. So far, many industrial clusters in the Tianjin economic–technological development area have formed and cultivated the unique culture of foreign-funded enterprises. Here are some examples. The enterprise culture with European and American characteristics, represented by the Tianjin Motorola Company, pays attention to personal encouragement. The enterprise culture with Japanese characteristics, represented by the Tianjin Toyota Automobile Company, has regard for rules and hierarchies and emphasizes employees’ orderliness, preciseness, discipline and collectivism. The enterprise culture which borrowed from Hong Kong, Taiwan, Korea and Singapore, whose representative is the Tianjin Tingyi International Food Co. Ltd, maintains loyalty and dedication in relations to the family network and absorbs the essence of European and American enterprise culture as well as Japanese enterprise culture, so it has strong afﬁnity with other ﬁrms. Besides the culture of foreign-funded enterprises, the local enterprise culture of Tianjin has rich characteristics including patriotism, the spirit of enterprise and the spirit of human unity. Those old ideas and concepts that have been discarded by history need to be rejected in the process of the integration of Chinese and foreign advanced cultures.

168 Industrial Clusters in the Tianjin Area

Bibliography and references China Bureau of Statistics. 2002. 2002 China Statistical Year Book. Beijing: China Statistics Press. Chu Xiaoping, Li Hua. 2002. ‘The Overview on the Theories of Small Enterprise Clusters’, Academic Research, issue 5, pp. 25–9. Jin Xiangrong and Zhu Xiwei. 2002. ‘The Startup and Development of Specialized Industrial Regions’, Economic Research, issue 8, pp. 18–24. Li Yonggang and Zhu Qing. 2003. ‘Preliminary Analysis on the Development of Zhejian Small Enterprise Clusters’, Review of Finance and Economics, issue 5, pp. 56–60. Luo Ruoyu. 2002. ‘Analysis on China’s Enterprise Clusters among Different Regions’, Nankai Economic Research, Issue 6. Porter, Michael E. 1980. The Competitive Advantage of Nations. New York: Free Press. Tianjin Municipal Bureau of Statistics. 2002. 2002 Tianjin Statistical Year Book. Beijing: China Statistics Press. Tianjin Municipal Government. 2003. The Future Development Project on Tianjin Green Battery Industry. Tianjin Municipal Government. Wang Jici. 2001. Innovative Spaces: Enterprise Clusters and Regional Development. Beijing University Press. Wang Yingjun, Zhou Liqun, Bai Xuejie and Luo Ruoyu. 2003. The Report on Tianjin New Industries. Nankai University. Xie Siquan. 2003. ‘Institution Reform and Industry Development A Case-Study on the Tianjin Bike Industry’, Theory and Modernization, issue 6, pp. 54–59. Xie Siquan and Huang Jiuli. 2003. ‘The Institution Reform and Accommodation with Demand Constraint: The Case of the Tianjin Bike Industry’, Nankai Economic Research, issue 3, pp. 22–32. Zhou Liqun and Luo Ruoyu. 2003. ‘The Characteristics of Tianjin Telecommunication Enterprise Cluster and Suggestions for Future Development’, Tianjin Social Science, Issue 5 pp. 81–5.

4 A Flowchart Approach Akifumi Kuchiki

Introduction Policy for forming industrial clusters – industrial cluster policy – plays an important role in the development of any region in East Asia. Such policy is followed not only in Japan by the Ministry of Economy, Trade and Industry but also in most countries in the area, such as Malaysia and Singapore. Silicon Valley in the USA and Bangalore in India are well-known success stories of development in information technology. Agglomeration theory and cluster theory explain that an industrial cluster is effective in generating external economies and reducing transportation costs. But we need to be clear about what the required conditions are for forming a cluster, particularly when we are discussing the growth strategy for developing countries. Porter (1998) studied industrial cluster policy as a growth strategy. According to Fujita and Thisse (2002), the central issues of spatial economics are to explain the mechanism for forming an agglomeration, and linking an industrial agglomeration to innovation. Porter (1998) considered the latter more important than the former, and examined how an industrial cluster brought innovations by using his diamond approach in cluster theory. There are two kinds of explanation of the mechanism of the formation of an agglomeration: one is in terms of the factors involved, while the other is to discuss which equilibrium point becomes a core industrial agglomeration under certain conditions in the case of multiple equilibria. The latter case of equilibrium theory is mathematically easy to analyse; we can ﬁnd the equilibrium analysis of urban agglomeration in Nishikimi and Hamaguchi (2003). Many papers, such as that of Sonobe and Kawakami (2001), have explained industrial agglomeration in a country by econometric methods. Kuchiki (2003) found that the conditions for forming new clusters in northern Vietnam were (1) industrial zones, (2) the capacity-building of physical infrastructure and institutional reforms in investment procedures and 169

170 A Flowchart Approach

(3) anchor ﬁrms in the manufacturing industry – showing that industrial zones together with the combination of infrastructure and institutions played a crucial role in an industrial agglomeration and that the sogoshosha (see below) of Japanese trading corporations played an important role in inviting Japanese manufacturing ﬁrms to the industrial zones. The anchor ﬁrm of the cluster, Canon, a Japanese ﬁrm, exported its products. This case was called the Canon effect. This chapter will generalize the typical pattern of forming industrial agglomerations in East Asia by illustrating another case. The purpose of this chapter is to build a framework with which to analyse industrial cluster policy. The framework is a ﬂowchart that can be applied to the analysis of a region from the view point of industrial cluster policy. We will make clear which factors are crucial to forming an industrial cluster. We will take into account both the role of governments in making industrial cluster policy and that of multinational ﬁrms, whose value chain management determines the deployment of plants; industrial cluster policy should be consistent with value chain management. By taking a ﬂowchart approach to industrial cluster policy, we will show how ﬁrms are agglomerated. The approach chosen for this chapter is to study the case of Toyota’s investment in Tianjin, China. Toyota has adopted cross-dock logistics to supply components from Shanghai to Tianjin so as to achieve a local content of 80 per cent; its transportation costs depend on these logistics. We will also refer to Kuchiki (2003), who discussed the roles of industrial zones in forming industrial clusters in East Asia after 1986. We propose a simple calculation method to ﬁnd the critical factors needed to form an industrial cluster. We consider the variables in transportation costs, the external effects of agglomeration, the costs of building a new plant and scale economies. Having examined Toyota’s cluster in Tianjin, we conclude that the three factors key to forming industrial clusters in implementing industrial cluster policy are (1) industrial zones, (2) capacity-building and (3) anchor ﬁrms. Capacity-building functions by combining the facilitation of physical infrastructure, institutional reform, human resource development and preparation of living conditions. Institutional reform includes one-stop service and the deregulation of investment procedures. Toyota’s cluster is not like that of Canon’s cluster in northern Vietnam in the procuring of components. The cross-dock logistics of the Toyota just-in-time system contribute to value chain management to inhibit the formation of an industrial cluster. We found two factors had positive effects in forming an industrial cluster: a keiretsu relationship (see below) between an anchor ﬁrm and its related ﬁrms, and scale economies. The two stages in the formation of an industrial cluster are as follows: I. An anchor ﬁrm moves into an industrial zone if conditions on (1) the industrial zone and (2) capacity-building hold.

Akifumi Kuchiki 171

II. Its related ﬁrms move into the industrial zone, provided the minimum requirement obtains that the anchor ﬁrm attains sufﬁcient production for the scale economies to apply. Note that an industrial zone is not enough to attract an anchor ﬁrm without the prospect of capacity-building, and that infrastructure is not enough to satisfy the condition of capacity without institutional reform. Figure 4.1

Industrial cluster formed by an anchor ﬁrm Market with saucers

Domestic

(1)

(2)

Export

Export-processing zone

Industrial zone

Capacity-building (a) Infrastructure (b) Institutions (c) Human resources (d) Living conditions

(3)

Anchor firm

(4)

Related firms

Industrial cluster

Regional economic growth

172 A Flowchart Approach

In Figure 4.1 we summarize this process as a ﬂowchart. Sometimes an anchor ﬁrm and its related ﬁrms take their decisions at the same time, and in these cases we do not need a ﬂowchart. At other times the related ﬁrms do not follow the anchor ﬁrm’s decision and there is no agglomeration. In such cases our ﬂowchart approach is needed to give priority to measures of industrial cluster policy. The ﬂowchart is crucial to the success of policy implementation, since budgets are limited. An action plan is needed to implement the ﬂowchart. The ﬂowchart approach of Kuchiki (2003) was applied to evaluate a project of the Japan Bank for International Cooperation in northern Vietnam (see JBIC 2004). Our ﬂowchart approach is practical and can be applied to developing countries as a growth strategy by checking whether conditions (1) and (2) in Figure 4.1 are present in sufﬁcient strength to invite anchor ﬁrms and their related ﬁrms. The next section of the chapter explains how a ﬁrm deploys plants according to value chain management. The third section presents a ﬂowchart approach to the forming of an industrial cluster and examines what conditions are critical to formation. The fourth section is a case-study of Toyota’s investment in Tianjin, China. A procurement process is part of Toyota’s value chain management and cross-dock logistics have a negative effect on the formation of the industrial cluster. Finally, we focus on decision-making in a component ﬁrm, analysing in detaie the behaviour of Denso, a keiretsu company related to Toyota. The remainder of this chapter is a case-study of Toyota’s cluster in Tianjin and tries to generalize a pattern for forming a cluster. In it we will build a prototype model of a ﬂowchart approach to industrial cluster policy, but the prototype model has many variations. Note that to justify the generalization more cases are needed in order to study various types of industrial cluster policy; the ﬂowcharts for biotechnology clusters and information technology clusters for example may well be different from those for industrial clusters in electronics and automobiles.

Value chain management This section explains how a ﬁrm deploys a plant from the point of view of value chain management, according to our interpretation of value chain management as seen in Porter (1998). Value chain Under free competition, private companies cannot survive without competitive advantages. Value chain management and the establishment of core competence can effect such an advantage. We will now explain the mechanism, using Figures 4.2 to 4.6.

Akifumi Kuchiki 173 Figure 4.2

Value chain from design planning to marketing

Design and planning

Procurement

Assembly

Marketing

D

P

A

M

Figure 4.3

Efﬁciency in production only Procurement

Figure 4.4

Marketing

A

Optimization for customer satisfaction

D

P

Figure 4.5

A

Customer satisfaction

M

Measures of value chain management: an example D

Alliance

Figure 4.6

P

A

Outsource

Outsource

M

Mergers and acquisitions

Core competence (selection and concentration)

D

P

A

Core

Outsource

Outsource

M

Mergers and acquisitions

As shown in Figure 4.2, a value chain is a whole process of research and development, design, the procurement of parts and components, assembly and marketing, including aftercare service in order to generate value for customers. Production functions in economics focus on assembly and production. The objective of private companies is to maximize proﬁt on the basis of production functions. But Figure 4.2 shows that the maximization is based on the value chain of the whole process of design, procurement, assembly and marketing. The situation here is different from that of production-only given in textbook economics and shown in Figure 4.3.

174 A Flowchart Approach

Figure 4.4 makes clear that the optimization of a value chain is to maximize competitive advantage for customer satisfaction, as we explain below. This is different from the Japanese characteristic of product-out, which means producers think that they can sell their products if the products are of good quality. Producers are satisﬁed that their product, made according to its particular characteristics, is of the best quality. As shown in Figure 4.5, it is not always the optimal case for one company only to take part in the whole process. US multinational corporations take strategies of mergers and acquisitions, alliances, and outsourcing, which is why they enjoy advantages in cost performance over Japanese companies. A company that has a competitive advantage in the long run must have a core competence as part of its value chain. It cannot have competitive advantages over the whole value chain, and must select a core competence and focus on it. Examples of value chains in the entertainment industry are games, television broadcasting and character goods in the case of sports, and movies, theme parks and musical shows in the case of Walt Disney. We may summarize the component parts of of a value chain as follows: Design and plan → Procurement of Parts and Components → Assembly → Marketing (logistics and aftercare service). Strategies for value chain management are as already mentioned: (1) outsourcing, (2) M&A (mergers and acquisitions) and (3) alliances. IBM puts an emphasis on marketing and R&D in its value chain. Ford and GM each started an internet business for marketing from the main process of assembly and production (see Figure 4.6). Sony is interested in ﬁnancial sectors including banking and insurance. Nissan (Japan) merged with Renault (France). A French manager, who changed the Japanese subcontracting system, and its employment systems, makes important decisions. A merger of Zaibatsu, between the Sumitomo and Mitsui banks, was announced in 1999. Competitive advantage The role of foreign direct investment is crucial for developing countries if they are to generate employment opportunities. Michael Porter uses the concept of competitive advantage instead of comparative advantage to explain this situation. Our understanding of his deﬁnition of competitive advantage is as follows: VCM denotes value chain management, and independent variables of competitive advantage are VCM and core competence. Thus: 1. Competitive advantage = f1 (VCM, core competence). And VCM depends on cost performance, quality and speed, and must attain the minimum cost. Value chain management by MNCs requires value chain networking in Paciﬁc basin countries.

Akifumi Kuchiki 175

2. VCM = f2 (cost performance, quality, speed). Core competence can be obtained by the generation of a brand name, or 3. Core competence = f3 (brand name). Brand marketing depends on commercial considerations as well as culture and tradition: 4. Brand marketing = f4 (commercial, culture and tradition), where brand marketing (for example televisions, personal computers and mobile phones) is one of the crucial measures in order to have a competitive advantage. A product consists of a function and a brand name. We can illustrate it as follows. On the one hand a necktie without a brand name costs US$10, and on the other a designer necktie costs US$200. The difference of US$190 is generated by the brand name. A personal computer with an Intel component Figure 4.7

Cost structure Total cost

Cost

Proportional cost (wages, materials)

Capital equipment

R&D

Fixed cost

Brand marketing

0

Quantity

176 A Flowchart Approach

can be sold well. It is necessary for a company to invest and have strategies in order to diffuse the brand name, as shown in Figure 4.7. Costs of branding strategy are ﬁxed costs since investment in branding strategy is the same as that in capital goods. Figure 4.8 shows that the larger the scale of a company, the lower the average cost. Figure 4.9 illustrates a case where a company outsources its processes of procurement of parts and components, and assembly. Consumers buy a commodity for function and brand name. That is why brand marketing is crucial for MNCs to survive in international competition. We will now explain competitive advantage in terms of cost, using Figure 4.9. Total cost is equal to proportional cost plus ﬁxed cost. Here proportional costs are wages, materials, etc. Fixed costs are ﬁxed equipment, design and development, and costs for brand marketing. For a country, it is also important to obtain a competitive advantage over other countries. Michael Porter indicates the following four factors in determining the competitive advantage of a country: 1. resource endowments (natural resources, artiﬁcial resources, capital, and infrastructure), 2. clustering of related industries and local suppliers, 3. hard competition among companies, and Figure 4.8

Average cost

Average cost

Capital equipment

R&D

Brand marketing Proportional cost

Akifumi Kuchiki 177 Figure 4.9

Average cost, brand marketing and R&D

Average cost

R&D

Brand marketing

4. a large-scale domestic market with high-end customers. This chapter discusses how value chain management can enhance an industry’s competitive advantage; only some of the items discussed in it overlap with the items indicated by Porter’s four factors. Global optimization and regional optimization of multinationals in the network We will show a typical case of regional optimization under the conditions described above under ‘Value chain’ and ‘Competive advantage’. The examples discussed here are taken from the actual economy. For instance, Tianjin, China, is regarded as an automobile cluster, and this section examines the role of Tianjin in the Chinese industry’s network formation. In the optimization of a value chain network, a process of assembly and manufacturing does not exist independently from the processes of design, procurement and marketing. In other words, a value chain must be considered comprehensively, that is, as is consisting of: 1. research and development, 2. procurement of parts and components,

178 A Flowchart Approach

3. assembly and production and 4. aftercare service. That is, D, P, A and M in Figure 4.2. In studying the global optimization of a supply chain as part of value chain management, the point to be considered is whether the Asian strategy, European strategy and American strategy can each, for the time being, as a ﬁrst step, be examined independently. We will explain why a solution for the global optimization of value chain management is the same as those for optimization independently by Asia, America and Europe. One objective of value chain management is to reduce costs, thus ensuring an ‘international competitive advantage’. The important factors for cost reduction are the tax system (such as customs duty), transportation costs (from Asia to Mexico), labour costs and stock costs (in particular, interest). Customs duty is greatly inﬂuenced by (1) the industrial policies of a country, (2) regional cooperation and (3) international economic cooperation. The examples are as follows: 1. preferential tariff rates to introduce foreign direct investment, 2. free-trade agreements that will be activated for regional cooperation in East Asia, and 3. liberalization policy of the World Trade Organization (WTO). These elements all had an inﬂuence on the determination of the customs duty by each country and by each region, which also inﬂuenced the costs of ﬁrms. This concept is applicable to East Asia also. The elements are: ﬁrst, improvement of laws that positively affect direct investment in each country, including Thailand and Korea after the Asian currency crisis; second, as for regional cooperation, early implementation of the reduction of customs duty within the area by the early realization of the Asian Free Trade Area (AFTA). Any ﬁrm when forming a globally optimal value chain network in order to maintain a competitive advantage must consider as a precondition whether a tariff system is inﬂuenced by these factors. Even when value chain management is global, multinational ﬁrms will be found to optimize regionally where transportation costs are too high to procure components from another region or regions (the regions in this context being the Asia-Paciﬁc region, the EU, and America). Then the AsiaPaciﬁc region is a size to be optimized locally. The objective of value chain management is to maximize competitive advantage for customer satisfaction in Asia. We illustrate a value chain network in Figure 4.10. In this example, the value chain network strategies of M&As, alliances and outsourcing are

Akifumi Kuchiki 179 Figure 4.10

Toyota’s value chain network in China

D

P

1 Japan

1 Japan

1

2

–

2 Tianjin

3

–

3

4

–

4 Shanghai

1 = domestic,

2 = M&A,

A

–

3 = alliance,

M

–

1

–

2 Tianjin

2

–

3

–

3 China

4

–

4

–

4 = outsourcing (cross-dock logistics)

taken among the Asian countries. The core competence of this company exists in design and planning. Recently, US companies have focused on a process of marketing and aftercare service. Optimal size of a region of value chain management Here we illustrate an optimization of the value chain network in Asia alone. CA denotes competitive advantage. An objective function to be maximized is as follows: CA = f (Dj,n1, Pk,n2, Al,n3, Mm,n4)

where j, k, l, m = 1, 2, 3 or 4

and each of n1, n2, n3 and n4 in Asia represents a name of its location such as Tianjin, Shanghai or Japan, where 1, 2, 3 or 4 denotes domestic production, mergers and acquisitions, alliances and outsourcing, respectively. Optimization of the value chain network means that we evaluate all the cases of combination, and choose the highest rate of CA. When multiple solutions are available, a chief executive ofﬁcer (CEO) of a multinational chooses one of them. A value chain network of Toyota is shown in Figure 4.10. The value chain network has two characteristics. First, it is different from the model of intra-ﬁrm trade that is internationally a vertical or horizontal division of labour of parts and components in Asian-Paciﬁc nations. Second, due to the ﬁrst characteristic, outsourcing strategy is a key factor to reducing cost. Contract manufacturers play important roles, which are also different from those of subcontractors with close relationships to parent companies of multinationals. The management system is changing from an internationally horizontal division of labour by intra-ﬁrm trade to a regional value chain network. As will be shown later, Toyota decided to

180 A Flowchart Approach

build a plant in Tianjin and procure components from Shanghai, Japan and Tianjin in order to maximize competitive advantage by using the crossdock logistics that will be explained below.

A ﬂowchart framework A typical economic growth model in East Asia consists of the following factors: (1) industrial zones: port, roads, (2) capacity-building the facilitation of (a) physical infrastructure, (b) institutional reform, (c) human resource development, and (d) preparation of living conditions (3) anchor ﬁrms, and (4) related ﬁrms. These factors form a cluster and enhance regional economic growth. Figure 4.1 is a ﬂowchart of industrial cluster policy as generally applied in East Asia. First, we consider the market for product sales by taking into consideration (1) industrial zones. There are two types of market, domestic and foreign. The export-processing zone in Kaousing in Taiwan (established in 1965) was the ﬁrst in Asia. The free-trade zone in Penang, Malaysia, (1971) and the export-processing zone at Tan Tuan near Ho Chi Minh City, Vietnam, (1993) have the same anchor ﬁrm as in Kaousing. Products in most developing countries are exported since their populations are small and incomes are relatively low in the early stages of development. It is usual for developing countries to have to establish export-processing zones to attract foreign direct investment. But industrial zones are not enough to invite anchor ﬁrms; to form an industrial cluster, capacity-building (item (2) above), which as we have noted, consists of the factors of physical infrastructure, institutions, human resources and living conditions, is also needed. Kuchiki and Yamada (1997) showed that 100 per cent ownership of capital and tax reduction or exemption was crucial to success in inviting foreign direct investment in East Asia as part of institutional reforms. Cheap labour in the 1980s in East Asia as an initial condition was also a crucial factor for success as part of human resource development. Human resources are classiﬁed as unskilled labour, skilled labour or professionals such as accountants and researchers. Living conditions are facilities like hotels, international schools, banks, shopping, entertainment and hospitals. An anchor ﬁrm (3) with its related ﬁrms (4) forms a cluster. Matsushita’s Panasonic division is a Japanese electronics ﬁrm in Shah Alam in Selangor, Malaysia, and with its related ﬁrms is a typical case of an industrial cluster in East Asia. Two relevant terms here are sogoshosha,1 Japanese for a trading ﬁrm playing the role of gate-keeper of a closed business circle, and keiretsu, a ﬁrm that plays a key role in inviting Japanese investors into industrial zones in East Asia (see below, p. 188). Sogoshosha ﬁrms that established industrial zones in Asian countries such as Thailand and Malaysia asked their keiretsu ﬁrms to be tenants, and formed agglomerations of Japanese ﬁrms in the industrial zones.

Akifumi Kuchiki 181

In summary, industrial zones or export-processing zones (1) and capacitybuilding (2) are conditions to invite anchor ﬁrms (3) as step I of industrial cluster policy. Then, as step II, their related ﬁrms (4) build plants near the anchor ﬁrm, provided sufﬁcient conditions for the related ﬁrms hold. These steps are shown in Figure 4.1. In the ﬁfth section we will show that in the case of Toyota’s related ﬁrms one of the sufﬁcient conditions is scale economies.

Step I: Toyota’s cluster in Tianjin Tianjin city has saucers to accept anchor ﬁrms and their related ﬁrms (see Table 4.1). The city is 120,000 km2 in size, has 9.19 million people, the largest artiﬁcial port in China, and is under the direct control of the central government. Industrialization has been accomplished by establishing industrial zones all over the city. The GDP growth rate of the city is 12.6 per cent. The value of the contracted amount of foreign direct investment in Tianjin is US$5.81bn and corresponds to that of Vietnam. We apply the ﬂowchart of Figure 4.1 to Tianjin Toyota’s case in Figure 4.11. In the implementation of industrial cluster policy, the three factors of (1) industrial zones, (2) capacity-building and (3) anchor ﬁrms are the key to forming industrial clusters. There are three special characteristics as follows. First, Toyota’s market is domestic, not foreign. Second, Tianjin Table 4.1

Tianjin’s industrial zones

Name

Rank

Year

Tianjin New Technology Industrial Park

Nation

1988

22

Tianjin Economic Development Area (TEDA) Tianjin Port Tax-Free Zone Tianjin Export-Processing Zone Tianjin Tanggu Ocean Hi-tech Development Zone National Tianjin Beichen Science Park Tianjin Wiquing Hi-tech Development Zone Tianjin Hangu Economic Development Zone Tianjin Dagang Economic Development Zone Tianjin Dongli Economic Development Zone Tianjin Xiqing Economic Development Zone Tianjin Jinnan Economic Development Zone Tianjin Baodi Economic Development Zone Tianjin Ninghe Economic Development Zone Tianjin Jinghai Economic Development Zone Tianji Jixian Economic Development Zone

Nation Nation Nation Nation Nation Nation City City City City City City City City City

1984 1991 2000 1992 2002 1991 1992 1992 1992 1992 1992 1992 1992 1992 1992

130 7 2 24 17 25 2 2 2 8 14 10 2 14 5

Source: Sato (2003).

Planned area (km2)

182 A Flowchart Approach Figure 4.11

Toyota’s cluster in Tianjin Market

Domestic

(1) Industrial zone

(2) Capacity-building (a) Infrastructure (b) Institutions (c) Human resources (d) Living conditions

Value chain management

Cross-dock logistics

(3) Anchor firm

(4) Related firms

Cluster

Regional economic growth

has a problem of capacity-building in fostering entrepreneurs. Third, crossdock logistics as a just-in-time method related to value chain management affected the formation of an industrial cluster in Tianjin. The Toyotatsusho Corporation collects components in Shanghai and transports them to Tianjin. Tianjin Toyota needs to neither buy nor import; nonetheless its local content of components in China was 80 per cent in 2002. 1. Industrial zones The city has seven industrial zones which the central government approved to establish. These are Tianjin New Technology Industrial Park, Tianjin Economic and Technological Development Area (TEDA) established in 1984, Tianjin Port Tax-Free Zone, Tianjin Export-Processing Zone, Tianjin Tanggu Ocean Hi-tech Development Zone, National Tianjin Beichen Science Park and Tianjin Wiqing Hi-tech Development Zone. There are many industrial development zones in Tianjin, as Table 4.1 shows. Japanese ﬁrms in the rubber, glass, steel, non-ferrous metal, machinery, electric appliances, transport equipment and precision equipment industries total 106, and 53 per cent of them have invested in development zones (see Sososha 2003, for a list of Japanese ﬁrms’ investments in China). The TEDA at the national level is the largest among all of these.2 It is located in the southeast of the city, about 45/km from downtown and 140 from Beijing. The TEDA enjoys easy access to North China, Northeast

Akifumi Kuchiki 183

China and Northwest China. The Beijing–Tianjin–Tanggu expressway runs through the TEDA and divides it into two parts, with the ﬁnancial, trading and residential area of 8.5 km2 to the south and an industrial park spanning acreage of 24.5 km2 to the north. The TEDA is 130 km2 in size and its pillar industries are electronics and electricals, food and beverages, pharmaceuticals and machinery. The TEDA is the largest saucer in Tianjin to accept foreign ﬁrms. 2. Capacity-building This sub-section explains (a) infrastructure, (b) institutions and (c) human resources. (a) Physical infrastructure The port is one of the most attractive facilities in Tianjin. Other physical infrastructures are sufﬁcient to invite foreign ﬁrms. As the largest freight terminal in North China, the harbour has an annual handling capacity of over 100 million tonnes, and 2 million TEUs (20-foot equivalent units) can be handled here each year, which will increase to 5 million TEUs by 2010. The port has freight exchanges with over 300 ports in 160 countries and regions around the world. With 47 container liner routes and nearly 200 international container liner routes, it is one of the world’s container hubs. Table 4.2

Infrastructure of Tianjin Economic Development Area

The current water supply capacity of TEDA is 180,000 tonnes per day, and the goal is 400,000 tonnes per day. Power supply capacity is 200,000 kV A by double circuit feeding. TEDA can provide dedicated power supply circuits to big users or those who have special needs Sewage treatment capacity is 100,000 tonnes per day. Gas supply capacity is 87,000 cubic meters per day. The capacity for treating the electroplating waste liquids is 1,000 tonnes per day. The liquids can be treated on site or carried away by specialized vehicles. All specialized companies in the infrastructure have been ISO 9002 certiﬁed. ISDN and IP telephone services are available in TEDA. In 1998, Motorola successfully had a videophone meeting for ﬁve hours. This meeting was made possible by three dedicated lines of 2B+D ISDN connected with Motorola’s functional departments around the world. It was also the ﬁrst meeting of its kind ever held in China. The 8M-DDN dedicated line provided by China Telecom is very advanced. Motorola once rented a 6M dedicated line in 2000. The TEDA ‘Urban Wide-Band Network Project’ was put into implementation in October of 2000. 1-10M broadband and 10-100M services are available. Source: http://www.teda.gov.cn/englishnew/invest/inf.htm

Tax system of Tianjin Economic Development Area

184

Table 4.3

Productive enterprise Advanced-technology- Export-oriented oriented enterprise enterprise

Energy, communications and construction project

Tax rate (%)

15

15

15

Enterprise income tax

Preferential treatment

Enterprise may be granted a reduction in income tax by half for 3 more years(and pays at a tax rate of 10% if the tax rate after reduction is less than 10%). See Note (1).

Enterprise whose export value exceeds 70% of its output value of the same year pays enterprise income tax at a rate of 10% after reduction.

Sino-foreign equity joint venture with an operational period of over 15years in the ﬁeld of harbor and wharf construction, upon its application as well as approval of the taxation authorities of the local province, autonomous region or municipality, may be granted 5 years of tax exemption and another 5 years of 50% tax reduction, commencing in the year in which the enterprise begins to make proﬁts. See Note (2).

Income tax on remittance of proﬁts gained by a foreign investor

Tax rate (%)

10

10

10

Preferential treatment, See Note (3).

Same as left

Same as left

Same as left

Notes: 1. According to state regulation, an enterprise remains advanced-technology-oriented after the exemption reduction period is over. 2. ‘5 years of tax exemption and another 5 years of 50% tax reduction’. 3. A foreign investor without an establishment within China pays withholding tax at an after-reduction rate of 10% on his interest, retails, royalties, or other earnings from TEDA, except for those exempt from the tax according to law; one who provides funds or equipments under favorable conditions or transfers advanced technologies may, once approved, receive extra tax deduction or tax exemption treatment. Source: http://www.investteda.org/en/ﬂfg/tzzctx_ss.html

Akifumi Kuchiki 185

With its 12-m-deep, two-way navigation channels, the port serves North China, Northeast China and Northwest China as the primary gateway for imports and exports. The port is used by Toyota to import components and export products. Table 4.2 illustrates the infrastructure of the TEDA, which provides a water supply, a power supply, sewage treatment, a gas supply, communications and so on. (b) Institutional reform Preferential treatment such as tax exemption or reduction and streamlined investment procedures in Tianjin are crucial to inviting foreign investors and are almost the same as those of other cities in Guangzhou and Shanghai. Table 4.3 shows the preferential tax treatment of the TEDA’s productive enterprises such as advanced-technology-oriented enterprises, export-oriented enterprises and energy, communications and construction projects. The rate of enterprise income tax for advanced-technologyoriented enterprises is 15 per cent. An advanced-technology-oriented enterprise may be granted a reduction in income tax by half for 3 more years. It is intended that investment within the TEDA should be streamlined, and Figure 4.12 shows the TEDA’s consulting services procedures. The TEDA Economic Development Bureau (EDB) and Investment Promotion Centre (IPC) function to accept foreign investors at the point of entry. The main processes are as follows: to make an on-site trip to the TEDA, to do feasibility studies, to decide investment, to apply for work permits and visas, to arrive in the TEDA and to register establishment of the enterprise. Until 2000 the TEDA had followed a policy of inviting large-scale ﬁrms, but this policy changed and it started inviting small- and medium-scale ﬁrms to provide supporting industry, partly because of Toyota’s involvement in 2002. The TEDA encourages the following automobile component industries: brake assemblies, driveshaft assemblies, transmissions, diesel fuel pumps, pistons, bulbs, hydraulic tappets, sliding bearings, aluminum radiators, bumpers, car air-conditioners, locks, back millers, power windows, unit meters, motors, lights, die-castings for automobile use and so on. Under this policy component ﬁrms are invited to supply components to assembly ﬁrms in Beijing, Tianjin and other cities near Tianjin. The number of registered foreign ﬁrms in the TEDA is more than 3000, but large-scale ﬁrms with more than US$20 million in total investment number only 200 (see Sato 2003). The policy of supporting small- and medium-scale ﬁrms has gone as follows since 2000. First, exhibitions are held to give small- and medium-scale component ﬁrms the opportunity to negotiate with large-scale assembly ﬁrms. Second, the TEDA Credit Guarantee Centre was established in 2002 to provide loans to small- and medium-scale ﬁrms. Third, TEDA Venture Capital provides loans to venture ﬁrms. Fourth, a committee was established to help enhance the productivity of small- and medium-scale ﬁrms. Fifth, an industrial area speciﬁcally for small- and medium-scale ﬁrms will be established.

186 A Flowchart Approach Figure 4.12

Consulting services procedures in Tianjin TEDA Economic Development Bureau (EDB) Investment Promotion Center (IPC) Introduce TEDA investment environment Make an on-site trip to TEDA EDB/IPC makes trip arrangements Do feasibility studies EDB/IPC provides related helpful data

Select the residence of the proposed enterprise

Decision is made to invest

EDB/ICP assist

Apply for allowances

Select the residence of the proposed enterprise Apply for allowances

Register establishment of the enterprise EDB/IPC provides suggestions Apply for work permit and visa

EDB/IPC introduces professional lawyers and accounting firms to assist

Arrive in TEDA

EDB/ICP assist

Look for an apartment

Accompanied children get in school

Register establishment of the enterprise

Fit in the local community

Source: TEDA Administrative Commission.

(c) Human resource development We consider the three kinds of labour noted above, that is unskilled, skilled and professionals such as accountants and researchers. Tianjin is attractive in having many university students and almost unlimited cheap labour.

Akifumi Kuchiki 187

What Tianjin lacks in the ﬁeld of human resources is entrepreneurs who will engage in capacity-building to invite foreign investors. However, there are 20 universities and colleges in Tianjin, boasting 79,581 students, of whom 73,630 are undergraduates and 5951 postgraduates. There are 26,150 staff members, of whom 9589 are full-time teachers, including 1093 professors, 3142 associated professors, 3674 lecturers, and 1060 assistant lecturers (1997 ﬁgures). 3. Anchor ﬁrms Toyota is an anchor ﬁrm agglomerating automobile component ﬁrms in the Tianjin area. In 1980 Japan surpassed the USA for the ﬁrst time to become the world’s top auto-producing country. At around that time, the Chinese authorities become interested in Toyota as an investor and joint venture partner. However, Toyota continued for the meantime its strategy of exporting to China, setting up sales networks but refusing to manufacture locally. According to Water (2003), we may summarize Toyota’s history of manufacturing in China as follows. The company took a multi-faceted approach to establishing a base for passenger car production in China. First, it had set up an extensive sales network and a considerable number of service facilities in China, and in 1995 it invested US$132m. to set up the Toyota China Technology Centre in Tianjin. This facility was used to train parts and components suppliers, to support dealers and service facilities, and to provide Toyota motor vehicle repair centers with factory-sponsored technical training. The second move was to begin a series of targeted investments in China’s automotive industry. Between 1995 and 1999, Toyota invested more than US$321m. in a joint venture engine plant with the Tianjin Automotive Group. The Tianjin Automotive Group was a producer of mini cars, trucks and buses that also produced the Daihatsu Charade under a licensing agreement. In addition, Toyota established a facility to produce 150,000 Toyota 1.3-litre engines per year, and three major automotive component ventures for the production of universal joints, castings and steering components in Tianjin. In May 2000 Toyota established the Tianjin Toyota Motor Company in a joint venture with the Tianjin Automotive Xiali Company, Daihatsu’s joint venture with a Chinese ﬁrm, to produce Toyota’s small-car model, the Vios. Toyota had been negotiating with Tianjin for several years for a joint venture and also assisted the Tianjin Automotive Group to increase production to 150,000 Charades per year to qualify as a key producer in China’s automotive sector. Since then, the newly formed Tianjin Toyota Motor Company has established several ventures with Chinese automotive suppliers and has induced some of its major Japanese suppliers to move operations to the Tianjin area to more effectively support Toyota’s new small-car venture.

188 A Flowchart Approach

In February 2002 the FAW Group acquired a majority stake of 51.9 per cent in the Tianjin Automotive Group. About a year later, in mid-summer 2003, Toyota and the FAW Group agreed to launch four Toyota models by 2005, the Land Cruiser Prado and the Crown. Meanwhile, the joint production of the Toyota Vios small-car in Tianjin has been unable to keep up with demand simply because of inadequate production capacity. Tianjin Toyota started operating in 2002 as an assembly ﬁrm.The TEDA is building capacity for small- and medium-scale ﬁrms to supply components to Toyota. Toyota’s local contents ratio at the Tianjin plant is 80 per cent.3 Tianjin is developing an industrial cluster as a group of many industrial zones including the TEDA.

Step II: the related ﬁrms A keiretsu,4 already mentioned, is a loose conglomeration of ﬁrms organized around a single bank for their mutual beneﬁt, and playing an important role in forming an industrial cluster. The ﬁrms sometimes, but not always, own equity in each other. Part of the formation, symbolized by vertical relationships, is the manufacturers’ keiretsu integrating a pyramid of suppliers and component manufacturers in one structure. The large, horizontal keiretsu spans a wide range of industries, including banking, insurance, steel, trading, manufacturing, electricals, gas and chemicals. The companies within it, whether they are customers higher up in the pyramid or subcontractors, behave as one ﬁrm: giving loans, technology, development costs, long-term supply agreements etc. This is known as the one-set principle. Agglomeration depends on the behaviour of components industries and anchor ﬁrms. In the case of Tianjin, Toyota took opted for crossdock logistics, which is why components ﬁrms need not establish subsidiaries in Tianjin. We will now see how this works. And we will use the case of Denso, a typical component supply ﬁrm in the keiretsu of Toyota, to see how scale economies were crucial to establishing a plant producing components (in this case, air-conditioners) in Tianjin. Cross-dock logistics Cross-dock logistics5 allow component ﬁrms to avoid establishing plants in Tianjin by reducing the transportation costs of components from Shanghai (see Figure 4.13). Toyota uses such means to transfer parts and components over this route. The Shanghai branch of Toyotatsusho, Toyota’s keiretsu company, is responsible for the logistics. We may see how the component industry in Shanghai developed in the 1990s from Tables 4.4 and 4.5. Table 4.4 shows output value by region of automobiles, engines and components in 1993. Jilin, Hubei and Shanghai are in order of output value of automobiles. Sichuan, Guangxi, Jiangsu, Beijing and Liaoning are in order of output value of engines. Shanghai, Sichuan, Hubei and Zhejiang are in

Akifumi Kuchiki 189 Figure 4.13

Cross-docking at work

Shipping

Sorting

Receiving

Source: Gue (2001).

order of output value of components. The situation regarding automobiles and engines in 1993 is almost the same as that in 2001. It is noteworthy that Sichuan dropped in the output of automobiles. The most striking change between 1993 and 2001 is that Shanghai’s output value of components rose by ﬁve times. It is further clear, by taking Jiangsu and Zhejiang into consideration, that the Shanghai region has formed an agglomeration of components. Shanghai, Jiangsu and Zhejiang are in order of output value of components from the highest in 2001. We can ﬁnd that Tianjin’s Toyota has an incentive to reduce the transport costs by collecting components in Shanghai and transporting the components packed by cross-docking from Shanghai to Tianjin. It is cheaper for Tianjin’s Toyota to import components from Shanghai than to produce them in Tianjin. The cross-dock logistics therefore have a negative effect on cluster forming in Tianjin.

190 A Flowchart Approach Table 4.4 Yuan)

Total output of the Chinese automobile industry, 1993 (million

Commodities Province

Engines

Parts & components

5,810 6,854 1,368 512 188

379 1,120 136 30 0

1,299 1,592 866 235 216

Liaoning Jilin Heilongjian

4,662 16,523 656

1,114 23 471

1,523 820 395

Shanghai Jiangsu Zhejiang Anhui Fujian Jiangxi Shandong

10,674 6,134 1,171 974 938 3,474 2,403

247 1,329 478 128 0 214 585

4,527 1,776 2,112 647 392 656 1,298

Henan Hubei Hunan Ghangdong Guangxi Hainan

787 14,995 744 5,045 3,199 0

24 22 376 26 1,430 0

919 2,492 1,594 833 476 11

7,189

2,173

3,056

Tianjin Beijing Hebei Shanxi Inner Mongolia

Sichuan

Automobiles

Source: China Automotive Industry Yearbook 1994.

Toyota’s keiretsu ﬁrms and other related ﬁrms in Tianjin are shown in Tables 4.6 and 4.7, respectively. They make clear the following three characteristics. First, all of the ﬁrms except two Toyota Motor Corporations are of either joint venture type or technical assistance type. Of the Japanese ﬁrms, 33 per cent are wholly owned and the others are joint ventures with Chinese ﬁrms. It follows that capacity-building by local ﬁrms as partners of joint ventures is key to attracting foreign ﬁrms. Tianjin is said to be lacking in entrepreneurship compared with other regions in China, which is one of the reasons why it is slow to introduce foreign direct investment compared with southern China and the Shanghai region (my interview to a staff member of Marubeni Corporation in Tianjin, 28, August 2002). Second, all the keiretsu ﬁrms and related ﬁrms except Daihatsu in Tables 4.6 and 4.7

Akifumi Kuchiki 191 Table 4.5 Yuan)

Total output of the Chinese automobile industry, 2001 (million

Commodities Province Tianjin Beijing Hebei Shanxi Inner Mongolia

Automobiles

Engines

Parts & components

3,841 5,065 669 436 598

661 106 120 159 0

2,053 863 2,278 969 219

Liaoning Jinlin Heilongjian

8,627 58,242 4,060

1,855 65 1,859

6,116 3,639 1,926

Shanghai Jiangsu Zhejiang Anhui Fujian Jiangxi

43,099 8,882 2,204 6,466 4,668 8,238

299 454 0 2 0 0

22,733 7,891 7,676 1,766 1,777 680

Shandong Henan Hubei Hunan Guangdong Guangxi Hainan

4,503 2,662 47,359 3,423 12,480 4,352 0

333 0 0 89 287 1,869 0

3,543 2,299 5,265 2,201 2,820 1,315 27

4,492

774

2,272

Sichuan

Source: China Automotive Industry Yearbook 2002.

started operating in and after 1995 (Daihatsu operated in 1985 independently of Toyota and later merged with Toyota). Third, Toyota’s keiretsu ﬁrms such as Denso and Aishin started operating in 1995, 1996, 1997 and 1998. The three related ﬁrms invested in the form of technical assistance. To take Aishin’s inlet modularity as an example, it set up ﬁve components such as throttle bodies, inlet pipes and variable inlet bulbs. Modularity increases measurement capacity and transportation costs, letting keiretsu ﬁrms establish a new plant in a cluster. Conditions sufﬁcient for building a new plant at a cluster In order to consider whether an anchor ﬁrm and its related ﬁrms will form and develop a cluster we take scale economies and external economies of the anchor ﬁrm into consideration.

Keiretsu companies of Toyota in Tianjin

Technical assistance Technical assistance Joint venture Technical assistance Joint venture Joint venture Joint venture Joint venture Joint venture Wholly owned Joint venture Joint venture Joint venture Joint venture Wholly owned Joint venture Joint venture Technical assistance Joint venture Joint venture Wholly owned

1985,01 1994,09 1995,04 1995,10 1995,12 1995,12 1995,12 1996,03 1996,05 1997,02 1997,06 1997,07 1997,07 1997,12 1998,02 2000,06 2000,10 2001,02 2001,02 2001,05 2001,07

Source: Toyota’s Activities in China, Toyota, April 2002.

Daihatsu Motor Co. Ltd Aishinseiki Co. Ltd Toyotatsusho Corporation Denso Corporation Toyota Motor Corporation Denso Corporation Toyota Gosei Co. Ltd Denso Corporation(asumo) Toyota Motor Corporation Toyota Motor Corporation Aishinseiki Co. Ltd Toyota Motor Corporation Denso Corporation Denso Corporation Toyota Motor Corporation Toyota Motor Corporation Toyota Gosei Co. Ltd Denso Corporation Tokairika Co. Ltd Aishinseiki Co. Ltd Toyota Motor Corporation

192

Table 4.6

Technical assistance to Daihatsu Disc calipers Copper plate prints Distributor Constant-velocity joint Alternator, scooter Brake hoses Small motors Engines Material for casting of c.v. joints Clutch parts, brake parts Steering, propeller shafts Electric parts Car air-conditioners Consulting on car New compact car Weather strips Air cleaners Switches Door locks, window winders, hinges, etc. Support of Toyota’s marketing strategy in China

Table 4.7

Japanese companies related to Toyota in Tianjin

Joint venture Joint venture Joint venture Joint venture Technical assistance Technical assistance Technical assistance Technical assistance Joint venture

1995,10 1995,12 1995,12 1995,12 2000,08 2000,08 2000,09 2001,05 2002,02

Araco Corporation Aisan Industry Co. Ltd Tokai Rubber Industries Ltd Fujitsu Ten Limited. Chuo Spring Co. Ltd Futaba Industrial Co. Ltd Chuo Spring Co. Ltd Maruyasu Industries Co. Ltd Futaba Industrial Co. Ltd

Panels, doors, roofs Fuel supply devices, throttle bodies, carburettors Engine mountings, water hoses Audio products Control cables Exhausts Coil springs Brakes, fuel pipes Body parts

Source: Toyota’s Activities in China, Toyota, April 2002.

193

194 A Flowchart Approach

Hypothesis: Development of an industrial cluster depends on demand – the quantity of production of the anchor ﬁrm. If an anchor ﬁrm expands production because of an expected expansion of sales in China, then its related ﬁrms will establish a new plant in the cluster. Independent variables of transportation costs are distance, cross-dock logistics and modularity in the following. That is: t = f (distance, cross-dock logistics, modularity) We assume that an increase in transportation costs caused by technological progress of modularity depends on management and technology: m = g (management, technology) It is well known that externality depends on a number of related ﬁrms in a cluster. Notations are as follows: t = transportation costs per unit, m = increase in transportation costs per unit caused by technological progress of modularity, e = external economies, s = movement costs for a related ﬁrm to establish a new plant, and q = quantities of production We assume that cross-dock logistics do not change, that the technology of modularity does not change, that movement costs are ﬁxed and that the number of related ﬁrms is ﬁxed. Then, we can conclude that the quantities of production decide the inequality. Related ﬁrms’ decisions depend on the quantities of production of an anchor ﬁrm. A related ﬁrm compares the total transportation costs, tq, with the movement costs, s, as follows: s < tq

or

s
The difference between transportation costs and movement costs reduces the proﬁts of the ﬁrm. An increase in transportation costs caused by new modularity technology is crucial to deciding whether a related ﬁrm builds a new plant at a cluster as follows: s < mq

or

s
Expected gains from external economies, e, in Tianjin can be deducted from the movement costs, s, and the total costs of not building a new plant are the transportation costs, tq and mq: s – e < tq + mq

Akifumi Kuchiki 195

One of the sufﬁcient conditions for a related ﬁrm to build a new plant in a cluster is in the following: s – e < q = 150,000 t+m In the next subsection we illustrate how the quantities of production are crucial to determining whether a ﬁrm builds a new plant. The number of automobiles was 150,000 and the minimum size of the average costs of production of a car is decreasing due to technological progress in car production.6 Denso’s case This subsection shows how Denso decided to establish a joint venture subsidiary with a Chinese corporation in Tianjin to produce air-conditioners, partly because of Toyota’s expansion plan for automobile production in China, which contributed to the development of a cluster in Tianjin. The company produces air-conditioners for automobiles from two factories in Japan, one in the city of Kitakyushu in the south and the other in Nishio in central Japan. The products are supplied to factories located within 400 km of the company’s plants. Denso chose to establish a plant in Kitakyushu in 1993, ﬁrst because its anchor ﬁrm Toyota had established a factory there in 1992; and second because living conditions – schools, shopping centres, hospitals, and so on – there better than those in other cities. Toyota however built from the ground up an in-house capability in electronic components, buying less from Denso and scaling down its dependence on the latter company (Ahmadjian 2000). The case of Denso in fact reﬂects the onset of the collapse of the keiretsu system; it sells components not only to Toyota but also to other ﬁrms such as the Mitsubishi Automobile Company and Matsuda who are not in the keiretsu of Toyota. Denso divides China into four areas: northern, eastern, southern and west-southern. It has four ﬁrms in Tianjin and one ﬁrm in Chongqing, and plans to have a supplier system for supplying Tianjin’s products with the other areas’ products by making the Tianjin ﬁrm the base and establishing satellite plants in southern China. The Denso plant located in Yantai, China, supplies air-conditioners to GM in Yantai and Toyota in Tianjin. Denso will produce air-conditioners in Tianjin if Toyota increases automobile production to more than 150,000 per year. Scale economies are crucial to Denso’s case (interview by the author on 24, August 2002 in Yantai). Denso’s staff in Yantai have said that Denso planned to establish a plant in Tianjin if Toyota announced expansion of automobile production. On 28, August 2002, Toyota announced that a second plant in Tianjin would produce 30,000 to 50,000 per year of a large model of automobile. Toyota started to produce compact cars called Corolla in 2004. By 2005,

196 A Flowchart Approach

with various types of automobiles, Tianjin will be a base large enough to produce 550 000 cars eventually. On 9, October 2003 Denso announced that it would establish a new plant in Tianjin at the cost of 1.5bn, to start operating in 2005 as a joint venture with a Chinese corporation (Nikkei newspaper, 10, October 2003). Denso decided to invest in Tianjin when the requisite condition of scale economies mentioned in the last section to attain its minimum average cost was established.

Conclusions We have built the framework of a ﬂowchart approach to industrial cluster policy and applied it to Toyota’s cluster in Tianjin. We have proposed that four factors – (1) market and industrial zones, (2) capacity-building, (3) anchor ﬁrms and (4) related ﬁrms – are the key to successful industrial cluster policy. Capacity-building means the facilitation of physical infrastructure, institutional reform, human resource development and the preparation of living conditions. Institutional reforms include one-stop service and deregulation of investment procedures. We have applied a prototype model of the ﬂowchart approach to Toyota’s cluster in Tianjin. The sequence was that Toyota ﬁrst, decided to invest in Tianjin, then second it asked its related ﬁrms to build plants in the region. One of the related ﬁrms, Denso, waited to decide to invest in Tianjin until Toyota announced it would expand its production to 550,000. We had two lessons from the case of Toyota’s cluster. First, at step I, the cross-dock logistics of Toyota deterred its related ﬁrms from joining an industrial cluster, since this procurement process was part of value chain management, a just-in-time system of efﬁcient logistics that decreased transportation costs from Shanghai to Tianjin. Second, at step II, Denso, a Japanese keiretsu ﬁrm of Toyota, waited to decide on its investment in Tianjin until Toyota announced it would expand its production to 550,000 units in China. Scale economies are a sufﬁcient condition for the related ﬁrms to join the anchor ﬁrm’s cluster. The market for the products of an anchor ﬁrm can be domestic or foreign. Toyota’s market is domestic, since China has a large amount of local demand. Governments can apply our ﬂowchart approach to other regions as growth strategy in Japanese Ofﬁcial Development Assistance (ODA) by ordering measures of industrial cluster policy. Table 4.8 and Figure 4.11 summarize this chapter. The ﬁrst step in forming an industrial cluster is to ﬁnd a market for the product. Here scale economies in China were crucial to forming an industrial cluster in Tianjin. The market in the case of Tianjin was domestic. The next step depends upon whether there are saucers for ﬁrms to build plants. Tianjin has (1) the Tianjin Economic Development Area (TEDA) of 130 km2 established in 1984 and other industrial zones established in 1992. Tianjin can offer as an incentive to ﬁrms a good port, physical infrastructure as part of (2) capa-

Akifumi Kuchiki 197 Table 4.8

Summary of industrial cluster policy, Tianjin and northern Vietnam

(1) Market and saucers (2) Capacity-building (a) infrastructure

Figure 4.11 Tianjin

Tables and ﬁgures

Figure 4.1 Northern Vietnam

EDZ TEDA

Table 4.1

EPZ TLIP, NHIZ

Port, roads, etc.

Table 4.6

Investment procedures Universities

Table 4.7, Figure 4.13 20 universities

Shopping etc.

Good

Route 5 and Haiphong port Investment procedures Cheap labour, universities Hospitals, schools

(3) Anchor ﬁrm

Cross-dock logistics

Toyota

Canon

(4) Related ﬁrms

Keiretsu

Tables 4.4, 4.5

Keiretsu

(b) institutions (c) human resources (d) living conditions

EDZ= economic development zone; EPZ = export processing zone; NHIZ = Nomura Haiphong Industrial Zone; TEDA = Tianjin Economic Development Area; TLIP = Thang Long Industrial Park.

city-building. Tianjin port is thus effective in forming an industrial cluster. Tianjin also offers ﬁrms as an incentive in human resources, students attending 20 universities and colleges, satisfying another condition of capacity-building. Toyota started to operate in October 2002 as (3) an anchor ﬁrm for a cluster in Tianjin. It merged with Daihatsu and decided to establish a subsidiary partly due to Daihatsu’s subsidiary located in Tianjin, though it had several candidate cities to invest in. Japan’s keiretsu system was effective in attracting (4) companies related to Toyota. Inviting an auto assembly company as an anchor ﬁrm is effective in implementing an industrial cluster policy, since a car is composed of ten thousand components. Component ﬁrms move into a cluster where an anchor ﬁrm exists if they can have proﬁts by building a plant. Cross-dock logistics are characteristic of Tianjin Toyota’s cluster. It is a just-in-time system of collecting components in Shanghai and transporting them to Tianjin. The logistics reduce transportation costs from Shanghai to Tianjin and weaken ﬁrms’ incentive to move into Tianjin. Kuchiki (2003) applied our ﬂowchart approach to a case-study in northern Vietnam. In this case products were exported to market. Saucers in the region were (1) the industrial zones of Nomura Haiphong Industrial Zone and Thang Long Industrial Park in Hanoi. Physical infrastructures as part of (2) capacity-building were the key to success in forming industrial clusters in Hanoi and Hochiminh City. The construction of National Highway

198 A Flowchart Approach

Route 5 and the rehabilitation of Haiphong Port, mainly under the Ofﬁcial Development Assistance commitments of the Japan Bank for International Cooperation, were crucial to attracting foreign investors by combining infrastructures with institutional reforms such as streamlining investment procedures and establishing one-stop services. Hospitals, international schools and entertainment facilities as part of living conditions were also needed to invite foreign investors. Vietnam has the advantage of both cheap labour and skilled labour at universities. Canon (3), a Japanese anchor ﬁrm, chose not China but Hanoi to build a plant to produce inkjet printers in order to avoid the risk of investing in China, country risk being one of the most important factors for ﬁrms in determining where to invest. A lot of Canon’s related keiretsu component ﬁrms have become residents of the industrial zones in northern Vietnam. Other related ﬁrms are also agglomerating there. We have shown that our approach is effective in analysing both cases and determining their characteristics. However, in order to generalize the approach it may well be necessary to increase the number of case-studies.

Notes 1. Sogoshosha: Homepage.mac.com/naoyuki_hashimoto 2. National Economic and Technological Development Zones (2003), (http://www.china.org.cn/english/SPORT-c/75871.htm). 3. 80 per cent: (http://www.1ban.co.jp/tachibana/geppo_bn). 4. Keiretsu: (http://www.investorwords.com/cgi-bin/getword.cgi?2691, http://www.rotman.utoronto.ca). 5. Cross-docking logistics: (http://www.0085.co.jp/news/newspage, http://www.schaferbros.com/warehouse/cross_dock.htm). 6. A Denso staff member told us that the number is from 100,000 to 150,000 according to my interview in the Economic Analysis Project of IDE-JETRO in 1998.

References Ahmadjian, C.L. 2000. Keiretsu, Governance, and Learning, Working Paper no. 76, Institute of Industrial Relations, University of California, Berkeley. China Association of Automobile Manufactures. 1994. China Automotive Industry Yearbook 1994, Tianjin. China Association of Automobile Manufactures. 2002. China Automotive Industry Yearbook 2002, Tianjin. Fujita, M. and J.F. Thisse. 2002. Economics of Agglomeration: Cities, Industrial Location, and Regional Growth. Cambridge University Press. Japan Bank for International Cooperation (JBIC). 2004. Report of Post-Evaluation on Yen Loan Projects 2003. Kevin, R.G. 2001. Cross-Docking: Just-In-Time for Distribution, Graduate School of Business and Public Policy, Naval Postgraduate School, Monterey, CA 93943, 8 May.

Akifumi Kuchiki 199 Kuchiki, A. and K. Yamada. 1997. ‘Lessons from Japan: Industrial Policy Approach and the East Asian Trial’, in L. Emmerij (ed.), Economic and Social Development into the 21st Century, Inter-American Development Bank, pp. 359–93. Kuchiki, A. 2003. ‘Agglomeration of Exporting Firms in Industrial Zones in Northern Vietnam’, in M. Kagami and M. Tsuji (eds), Industrial Agglomeration, Institute of Developing Economies-JETRO, Chiba-shi, Japan. Nishikimi, K. and N. Hamaguchi. 2003. ‘Urbanization and Agglomeration’, Development Economics (in Japanese). Yuhikaku. Publishing Co., Ltd, Tokyo, pp. 75–92. Porter, M.E. 1998. On Competition. Harvard Business School Press, Boston, MA. Sato, K. 2003. ‘Investment Climate in Northern Region in China’, Shinkin Central Bank Monthly Review, pp. 5–6. Sonobe, T. and M. Kawakami. 2001. ‘Economic Development and Industrial Location’, Ajia Keizai, IDE-JETRO, Chiba-shi, Japan pp. 2–19, 42 (1). Sososha Co., Ltd, Tokyo. 2003. List of Japanese Firms Invested in China. Toyota Motor Corporation. 2002. Toyota’s Activities in China April 2002, Tokyo. Water, A. 2003. The Japanese Automobile Industry in China (http://www.jpri.org/ WPapers/wp95.htm).

5 Chinese Automotive and Parts Industries Masatsugu Tsuji and Wu Quan

Introduction It is a fact that automotive assemblers and parts suppliers are located in the same regions, also that this heavy localization is a common characteristic of the automotive industry worldwide. In the case of the Toyota Group, the location of its headquarters and factories and those of the major member companies of Kyohokai are situated in Aichi Prefecture, or more preciously, 80 per cent are in Aichi Prefecture and of these 55 per cent are located in the Nishi-Mikawa district, in the eastern sector of the prefecture. Moreover, nearly half are found in Toyota City. The reason for this is that the parts suppliers are directly tied up with Toyota and have located themselves close to its factories in Toyota.1 The reasons why automotive assemblers and their parts suppliers are located thus are explained by the theory of information initiated by Coase (1937; see also Williamson 1989, for example). According to this theory, the closeness of two types of players is determined by either the transaction costs or the information structure of the system to prevent opportunism and bounded rationality. Another theory relevant to this issue is spatial economies, as expounded by, inter alia, Krugman 1991 and Fujita, et al. 1999. According to this, localization of a particular industry can be explained by economies of scale and transportation costs: if scale economies prevail at the ﬁrm level, and transportation costs are relatively small, then economic resources are attracted to the region where factories are located, and this accelerates the attraction of even more resources to the region. This cumulative process will increase the advantage of the region further, and the industrial cluster is realized. The automobile industry seems to ﬁt the above ‘core–periphery model’, since it shows increasing returns to scale, and the transportation costs of most of the parts as well as the automobile are not large. This theory, however, cannot provide a consistent explanation for why the particular region was selected at the initial stage. 200

Masatsugu Tsuji and Wu Quan 201

In the case of Toyota, there was no need for it to choose to locate its headquarters in Koromo City.2 Regarding the localization of automotive assemblers and their parts suppliers, an appropriate theory has not yet been found. There are various types of automotive parts from heavy, simple parts to high-tech ones, the latter being quite light compared with price, so that transportation costs do not reﬂect the location of production. In reality, those parts are collected from various places all over the world in the age of globalization. For developing countries, which aim to promote the automobile industry, it is important to make a decision on the locations of automotive assemby plants and also those of parts suppliers. This chapter analyses the location and the relationship among automotive assemblers and their parts suppliers, taking China as a case-study and trying to explain how the locations of automotive assemblers were decided and how the relations between them were formed. In the latter, we discuss extensively the case of Shanghai. The outline of this chapter is as follows. The next section presents brieﬂy the current situation of the Chinese automobile industry. The third section provides the historical background of the industry, and tries to explain how the policy was formed to determine the location of automotive assemblers. China is basically a planned economy, and even economic policy is determined based not only on economic but also on political factors. The latter point is emphasized in the fourth section. While the growth of the automotive industry has been impressive, we will explain how the automobile policy of the Chinese government has failed. Following this, other topics regarding the relationship among automotive assemblers and parts suppliers in terms of location will be discussed in the ﬁfth section. This section is a case-study of the performance of automobile industrial organization in Shanghai, whose automotive industry is the most successful in China. The success is due to its local industrial policy, which became the basis of the potent automotive conglomerate ranked number one in the country. A brief conclusion will be provided in the sixth and ﬁnal section in relation to Japanese automotive assemblers, such as Toyota and Honda, and their parts manufacturers based on ﬁelds research.

Current concentration of automobile and parts industries Localization of the industry Currently there are approximately 120 automotive manufacturers, and 140 motorcycle manufacturers, which are widely spread over the country, as already explained. Due to increasing competition as well as guidance and regulation by the central authorities, small-scale manufacturers cannot maintain their production. The list is summarized in Table 5.1. Among them, the most important assemblers are Shanghai Volkswagen, the First

202 Chinese Automotive and Parts Industries Table 5.1

Major Chinese automobile assemblers Vehicles

Country of Headquarters

Production

Sales

Location

Shanghai Volkswagen Automotive Co. Ltd

Germany

278,890

301,095

Shanghai, Jilin

FAW-Volkswagen Automotive Company Ltd

Germany

191,695

207,700

Changchun

Shanghai GM Automobile Industry Co. Ltd

USA

111,623

110,763

Shanghai

Tianjin Xiali Automobile Company Ltd

Japan

89,921

86,976

Tianjin

Dongfeng Peugeot Automotive Company Ltd

France

84,378

85,088

Hubei, Wuhan Sichuan,

Changan Suzuki Automobile Co. Ltd

Japan

67,846

65,018

Chongqing Guangdong,

Guangzhou Honda Automobile Co. Ltd

Japan

59,024

59,151

Guangzhou

Jieli Group Corporation

China Germany

47,370

43,500

Guangzhou Jilin,

China FAW Group Corporation

China

30,165

26,634

Changchun Jiangsu,

Yue Jin Motor Group Corporation

Italy

23,026

23,002

Nianjin

Name

Source: Data from China Automotive Industry Year Book 2003.

Automotive Company and the Dongfeng Motor Corporation. The major manufacturing regions are Shanghai, Changchun, Taijin, Beijing and Guangdong. The major automotive ﬁrms collaborate with foreign assemblers which have high technological capabilities. Regarding automotive parts suppliers, since there are so many scattered throughout the country, it is almost impossible to obtain complete data. According to the China Automotive Industry Year Book 2003, approximately 2201 parts manufacturers are registered, among them 352 afﬁliated to foreign companies. The latter occupy the major share in ﬁelds such as ABS systems, EMS, air conditioning parts, and turbochargers, while Chinese parts manufacturers produce most of the basic and simple parts.3 Locations with high concentrations of automotive assemblers and their parts suppliers are, as summarized in Table 5.2, the provinces of Jiansu, Sichuan, Zhejian, Jilin and Shangdon, and the cities of Shanghai and Beijing. The characteristics of the localization of parts suppliers can be explained by the traditional policy of dispersion of the automotive industry, which

Masatsugu Tsuji and Wu Quan 203 Table 5.2 Number of automobile assemblers and parts suppliers in concentrated regions of China Location

No. of ﬁrms

Provinces Jiangsu Sichuan Zhejiang Jilin Shandong Hunan Hubei Liaoning

191 172 163 134 131 131 127 122

Cities Shanghai Beijing

104 96

Source: Data from China Automotive Industry Year Book 2003.

will be discussed in detail later. Each province or region tends to have every kind of automotive parts manufacturer in the region in order to secure the regional economy. Another reason is that the location of automotive assemblers and parts suppliers was spread widely throughout the country, and in turn this made for little interregional movement of parts, on account of the high transportation costs.4

Historical background of the Chinese automobile industry Early production and policies: 1949–65 In the face of the heavy reliance on imported vehicles, the construction of an indigenous automotive manufacturing industry came to be a top priority for the newly established Communist government. The reasons are clear. First, to help revive the economy, the efﬁcient transportation of rural products by mechanized trucks would be vital to stimulate agricultural development. Furthermore, with the beginning of the cold war and military threats to China on several fronts, an automotive industry was required for enhancing military mobility. Only one year after the founding of the new China, therefore, the Chinese moved quickly to create their own large-scale automotive industry. In 1950, they approached the USSR for support in planning an auto factory, and the Ministry of Heavy Industry commissioned a ‘preparatory ofﬁce’ to guide development of the automotive industry. By 1951, the Chinese government had chosen Changchun in northeast China as the site of their First Automotive Works (FAW). It became one of the ‘three big’ major automobile enterprises existing since then.

204 Chinese Automotive and Parts Industries

To develop their nascent freight transportation capabilities, the central government decided to manufacture trucks exclusively at their new government-owned and managed plant. The ﬁrst product would be a 4-tonne Soviet model, the ZIS 150, and the target was to produce 30,000 vehicles per year. The Finance and Economic Commission of the Government Administration Council approved such initial decisions. In 1953, the government acknowledged the importance of the automobile industry by formally establishing the ‘Automotive Industry Administration Bureau’ under the First Ministry of Machinery Building, and the construction of the FAW began, and opening ofﬁcially in 1956. By the mid-1950s, the government recognized the need for a small number of passenger cars, to serve top-ranked ofﬁcials. In 1958, the FAW manufactured its ﬁrst Honqi model limousine, a luxury vehicle based on Daimler Benz’s 220 model sedan. In the same year, the Shanghai Automobile Assembly Plant began manufacturing its Phoenix passenger car with funds from the Shanghai municipal government. Actual production, however, was quite small; in 1960, the whole nation produced only 98 of its own cars; in 1961, following the shortcomings of the ‘political struggle’ (1958–60), it produced only 5 units. To compensate for the small number in domestic production, China imported an average of some 1000 passenger cars per year from 1954 to 1965, mostly from Eastern Europe. Chinese automotive production policy ﬂuctuated during the ﬁrst decade of the new China. The early 1950s were marked by emphasis on the promotion of workers’ skills and technological innovation at the manufacturing level. With the arrival of Soviet experts in the mid-1950s, however, the Chinese government made efforts to adopt foreign production methods and experiences. The FAW adopted widely the Soviet method of production through technology absorption; the ‘political struggle’ again reversed the attitude toward automotive innovation.5 China shifted on a national scale towards reliance on their own working styles, and campaigned against the dominance of foreign technologies and equipment. The ‘political struggle’ delayed the development of the Chinese auto industry. The policy swings guiding automotive development roughly parallel the political clashes among China’s top leadership over the country’s ﬁrst ﬁfteen years. The failure of the political struggle was reﬂected in the automotive industry’s swing back to technological reliance in the early 1960s. The automobile industry under the revolution era: 1966–76 As relations with the Soviet Union deteriorated in the 1960s, China became worried about a foreign attack on the largest vehicle plant in Changchun. In 1965, the newly formed China National Automotive Industrial Corporation approved the creation of the Second Automotive Works (SAW), and the next year the Chinese decided to build a new truck factory in the relatively isolated mountainous region of Shiyan in Hubei province.

Masatsugu Tsuji and Wu Quan 205

The SAW factory was to put into operation the ideals of self-reliance, and would depend on indigenous technology and the experience of other Chinese manufacturers. The production of passenger cars received scant attention during this period. Such vehicles were still intended primarily for high-level ofﬁcials, since private car ownership and operation were essentially prohibited; passenger cars accounted for only about 1 per cent of Chinese production, while trucks and other vehicles were prioritized. In 1966, the FAW introduced a more spacious model of the Honqi luxury vehicle, and the First Ministry of Machinery Building approved a plan to enlarge the Shanghai Automobile Assembly Plant’s ‘Shanghai Brand’ cars in 1972. Still, passenger car production dropped from 302 in 1966 to 196 in 1970, and imports virtually vanished. Although overall production growth was slow, the revolutionary years emphasized the spirit of geographical self-sufﬁciency, as mentioned above, and led to a spiralling growth in the number of automotive manufacturers throughout the country. In 1964, there were 417 factories producing trucks, passenger cars, motorcycles and fundamental automotive parts; by 1976, the number reached more than 1950. The number of automotive manufacturers experienced a similar jump, and nearly every province came to boast automotive plants. There were some 50 factories capable of serial production of trucks or other vehicles, however only four were reported to be able to manufacture more than 10,000 units per year. Some plants produced only a few thousand or a few hundred vehicles annually, and this showed signiﬁcant diseconomies of scale in many geographical areas. (for a rough comparison, US auto manufacturers in the 1960s aimed at some 60 cars per hour per assembly line, and an annual plant capacity of 200,000 to 400,000 units; at the time, this represented an economy of scale). The automobile industry during the reform era: 1978–94 During the reform era, which started in 1978, the Chinese automobile industry developed rapidly. In 1978, China produced only 149,000 motor vehicles. However, output reached 1.35 million in 1994, an almost tenfold increase during 17 years. Today the Chinese automobile industry is noticeable in another aspect: multinational enterprises have become a signiﬁcant source of capital and technology. As of 1995, the accumulated foreign direct investment in the automobile industry amounted to US$10.5bn, accounting for 7.89 per cent of China’s total foreign direct investment during this period. Furthermore, almost all passenger cars were produced in joint ventures or under licence with foreign automotive manufacturers. In 1995, joint ventures accounted for 25.3 per cent of total automotive sales, and almost half the proﬁts in the sector. Despite these achievements, the Chinese automobile industry suffered from a number of serious problems by the mid-1990s, which were mainly as follows: (1) the automotive corporations were excessively fragmented;

206 Chinese Automotive and Parts Industries

(2) the parts industry lagged behind; (3) all automotive corporations had not yet arrived at an optimal level of production; (4) cost reduction did not progress; and (5) Chinese companies were not conscious of international competition. When we look at the Chinese automobile industry in retrospect, we ﬁnd that policy in this period lacked concreteness, and the superintendent agencies that set excessive targets lacked macroeconomic managerial capability. The foremost problem however was the extremely fragmented automotive sector: the 1.45 million motor vehicles produced were spread among 122 independent automotive ﬁnal assembly plants; thus the average plant produced only around 12,000 vehicles, one of the lowest in production volumes in the world. In 1987, the Chinese central government designated six large motor assembly plants as ‘key ﬁrms’ targeted for expansion with administration support. The six automotive corporations were given different market goals. Three were allocated to the large passenger car sector and three to the small passenger car sector, and the plan was referred to as ‘three big, three small’. According to this plan, China’s main vehicle manufacturers would be the FAW in Changchun, the SAW in Hubei and the Shanghai Vehicle Factory, which by then had formed the joint venture with Volkswagen. The three small players would be joint ventures of Beijing Jeep, Guangzhou Peugeot and the Tianjin Automotive Corporation, which had a licensing agreement with Japan’s Daihatsu Motor Corporation. However, the plan was expanded to ‘three big, three small, two mini’ under pressure from the Central Military Commission. The two ‘minis’ in this plan were enterprises under the Central Military Commission, namely the Changan Manufacturing Company and the Guizhou Aircraft Auto Company. Consequently, their proﬁt was subject to the military. The scales of production of the ‘eight major large automotive ﬁrms’ were very small, while other governments in the world imposed a strict concenTable 5.3

Eight major ﬁrms in the Chinese automotive sector, 1995 No. of assemblers

No. of suppliers

Vehicle volume

Book value of assets (¥bn)*

Eight majors

35

129

960,618

Auto industry

122

1,732

1,452,697

115

23.13

Shares of eight majors

28.70%

7.50%

66.10%

42%

32%

* Value of ﬁxed assets evaluated in original prices. Source: Data from China Automotive Industry Yearbook 1996.

48.57

Investment in 1995 (¥bn) 7.38

Masatsugu Tsuji and Wu Quan 207

tration policy in the automotive industry.6 These eight key companies controlled only 28.7 per cent of the total volume of terminal assembly plants and 7.5 per cent of components plants. Although they accounted for 66 per cent of vehicles production, their individual scale was very small by international comparison: their average output volume was only 27,466, far below the efﬁcient level of production, which is estimated to be between 250,000 and 300,000 vehicles. The eight majors accounted for less than half of total assets and investment made in the automotive sector in 1995, indicating that the Chinese central government was unable to invest strategically (see Table 5.3).

Issues of localization in the Chinese automobile industry Small scale of production The undercapacity of major Chinese companies is best illustrated by a comparison with other developing countries like Brazil and Korea during the 1970s and 1980s, because the two countries were at that time similar to China today. Korea attempted to enter the automotive sector at a time when the multinational ﬁrms had already established their dominant positions. Korea implemented a very strict concentration policy to develop a ‘national passenger car’ starting in the 1970s. And the policy acquired signiﬁcant success both for automotive production and for exports. Their successful experience indicated the importance of automotive concentration policy; it is a necessary condition for developing countries to establish their own automobile industry. Table 5.4 Concentration ratios in the automotive industry: four-country comparison (%) Top-one-ﬁrm ratio

Top-two-ﬁrms ratio

Top-three-ﬁrms ratio

Brazil 1959 1970

24.8 56.1

42.7 74.3

60.6 91.2

Japan 1960 1975

32.1 33.7

56.1 63.6

65.1 72.8

Korea 1975 1986

54.6 71.3

77.7 88.6

96.4 97.9

China 1985 1995

19.2 12.6

38 23.6

43 33.3

Sources: Data on Brazil, Japan, and Korea calculated from World Motor Vehicle Data. Data on China calculated from China Automotive Industry Yearbook 1995.

208 Chinese Automotive and Parts Industries

The most straightforward measure of the degree of concentration is the concentration ratios of the top one, top two and top three ﬁrms in automotive output (see Table 5.4). Political decentralization and dispersion of the automobile industry The importance of passenger car production has been recognized by the increase in imports of passenger cars in the 1980s. With the reform, therefore, not only various branches of the central bureaucracy, but also local governments started establishing automobile plants under their jurisdictions. The initiative came ﬁrst from local authorities that obtained special trading status as well as discretion in disposing of a limited amount of the foreign exchange required for international trade. The Hainai Special Economic Zone, for example, had rights to import goods without duty. In addition, local governments had most of the ﬁscal revenue by 1994, and the power to regulate local banks. These entitled local authorities to ﬁnance subordinate enterprises easily. These local governments included the three municipalities of Beijing, Shanghai and Tianjin. Some local authorities established afﬁliated organizations such as provincial automotive companies so as to support the subordinate automotive manufacturers to modernize their plants. Hubei and Liaoning provinces at least also implemented these policies. Given the tariff protection and stern administrative barriers, the automobile sector produced higher proﬁts than other manufacturing industries. The rise can be seen in ﬁrms related to national defence, which had began recently to produce motor vehicles or parts so as to complement military expenditures; for example, one automobile plant under the Central Military Commission, ChanHe, submitted ¥140m. from proﬁts to military research institutes in the 1990s. With so many parties launching projects, duplication became inevitable. For example, between 1982 and 1985, four tyre production projects were introduced. By all accounts, they probably occurred exclusive of one another, because the local authorities in Shanghai and in Guangzhou, Liaoning and Heilongjiang provinces obviously initiated these projects without talking to one another or to the central bureaucracy. The transaction between Heilongjiang province and Italy’s Pirelli was the only project supported by the central authorities, with the China National Chemical Construction Corporation, an import–export arm of the chemical industry that had supervisory authority over production, involved in technological acquisitions for the plant. Failure of industrial policy in the automobile industry Such duplication created not only inefﬁciency due to the relatively small scale of production, but also obstacles for technological transfer. Since resources were scattered, both the scale of the projects and the subsequent volume of production were limited. It is obvious that the ability of the recip-

Masatsugu Tsuji and Wu Quan 209

ients of technological transfer would increase with the scale of production according to the learning curve. Knowing this, governments reinforced local protection in production and procurement. As a result, Beijing, Tianjin, Shanghai, Jiangsu and Shangdon, where many automotive manufacturers were located, achieved annual outputs of 50,064, 20,403, 32,730, 12,165 and 24,474 vehicles respectively in 1985, while 14 other provinces made only between a thousand and several thousands of vehicles. In 1986, three assemblers in Beijing, including Beijing Jeep, Nanjing Auto Works and Tianjin Auto Works, produced 58,392, 20,403 and 21,542 units respectively. Shanghai Volkswagen made about 5000 saloons, whereas Guangzhou Peugeot managed only 790 cars and 592 pickup trucks. In fact, most of the 130 auto assemblers in China built only hundreds of vehicles each that year. In order to solve the above problems, from the mid 1980s, policies for the automobile industry were aimed at upgrading production scale and improving productivity. In order to obtain a sufﬁcient level of production, the Planning Committee and the Ministry of Machinery Industry (MMI) determined to restrict severely the number of automobile plants. MMI had worked hard on sorting out the chaos in new entry acquisition to make some sense of its role in the broad development of the Chinese auto industry. Despite its ofﬁcial capacity, however, the MMI ran into much difﬁculty in performing its functions in control and coordination. Not only did its policy preference differ from the rational interests of most of the local governments and defence industry, which have a bureaucratic stature equal or higher than the MMI, but also its ﬁnancial weakness rendered it unable to enforce its authority effectively.

Localization of the automobile industry in the Shanghai area Historical background of Shanghai’s automobile industry Shanghai before the revolution saw signiﬁcant industrial growth over the thirty years following World War I. The city was known for its trading enterprises: some 40 per cent of China’s import and exports passed through Shanghai in the 1920s and 1930s. Annual industrial growth from 1919 to 1939 reached nearly 14 per cent, and textile and light machinery production took the lead. The municipal banking sector competed with wellestablished foreign ﬁnancial ﬁrms in the international treaty settlements, though Chinese ﬁnancial companies and other business people beneﬁted from foreign capital and expertise. Following the Communist takeover in 1949, Shanghai’s heavy industrial sector developed quickly. The textile industry held the largest share of Shanghai’s 1949 output, constituting 62.4 per cent; light industry accounted for 24 per cent, and heavy industry only 13.6 per cent. During the decade after the establishment of the People Republic, however, heavy industry achieved more than 40 per cent every year, accounting for 52% of the city’s

210 Chinese Automotive and Parts Industries

industrial output in 1959. Textiles had fallen to only 27 per cent of total industrial production by that year. During this period, several important automotive ﬁrms were established, including the Shanghai Automotive Electric Motor Factory, the Shanghai Automotive Transmission Shaft Factory and, most importantly, the city’s ﬁrst passenger car factory, the Shanghai Automotive Assembly Plant, later known as the Shanghai Car Plant (SCP), which became the partner of Germany’s Volkswagen in later years. All of these were founded in 1958. In addition, the cooperation among various manufacturers allowed the SCP to produce its ﬁrst car, the Phoenix, in 1958. The technology and the level of skills at the factory were extremely limited, so that production remained at only 100 vehicles each year. By 1964, SCP’s Phoenix had built a reputation and thereafter was referred to as the ‘Shanghai Brand’. The development of Shanghai’s heavy industry continued in the late 1960s and early 1970s with concerted efforts to modernize the steel industry. Based on Shanghai’s many years of commitment and investment in heavy industry, in addition to its large population as a potential market, it is not surprising that foreign automotive companies would consider the city as the prime partner for joint ventures. In 1985, after a trial production, the Shanghai municipal government agreed to establish a joint venture with Volkswagen. VW held 50 per cent of the newly established company, named Shanghai Volkswagen (SVW). On the other hand, the Shanghai Automotive Industrial Corporation (SAIC) under the municipal government, which had tight control over the Shanghai Car Plant, took 25 per cent of the venture. The Bank of China’s Shanghai Trust and Consultancy Corporation owned 15 per cent of SVW, and the Chinese Automotive Industrial Corporation held 10 per cent. The initial capital of the new company was US$40m. and it located its factories in Anting, Jiading Qu, a suburb of Shanghai. VW promised to provide advanced technologies. The contract was to run for 25 years.7 SVW stared with an annual production of 1733 units and 1752 employees, and 10 years later they reached 160,000 and 9318, respectively. The SAIC’s passenger car output accounts for about 50 per cent of overall domestic production, which reached 278,890 units in 2002. In addition, the content percentage of local parts reached nearly 100 per cent that year, the highest in Chinese automotive industry. Thus, with nearly 20 years of development, the SAIC has achieved success. Localization of the Santana and parts industry in Shanghai The development of the automotive and parts industries in Shanghai is closely related to the localization of SVW’s Santana. At the beginning, because of the level of Chinese parts manufacturers, the knockdown system was adopted, and therefore the ratio of localization was quite low. With the support of Volkswagen, Shanghai made efforts to localize its parts, and this

Masatsugu Tsuji and Wu Quan 211 Table 5.5

Localization ratio of Santana

Year

Production (10,000 vehicles)

Ratio of localization (%)

1985 1986 1987 1988 1989 1900 1991 1992 1993 1994 1995 1996

0.8 1.04 1.55 1.56 1.85 3.5 6.5 10.0 11.5 16.0 20.0

3.60 3.99 5.70 13.05 31.04 60.09 70.37 75.33 82.20 84.50 86.50 90.00

Source: Ikeya (1997:158).

stimulated its parts industry to grow, so that eventually Shanghai become the most centralized area of automotive parts production.8 Table 5.5 indicates the process of localization. Automotive parts suppliers located in the suburbs of Shanghai; Anting, Jiading Qu, is a typical example, now referred to as ‘China’s Detroit’. Table 5.6 lists some of the major parts manufactures in Shanghai. With the process of localization, the supply chain network of Shanghai has also been forming.9 Local industrial policy of Shanghai Here we attempt to analyse the factors that make Shanghai’s automotive and parts industries a success in terms of industrial policy. The ‘six pillar industries’ in the Shanghai area, announced by the Shanghai municipal government, are the automotive industry, the communications equipment manufacturing industry, the power plant equipment manufacturing industry, the petro-chemical industry, the steel industry and the electric appliances industry. Among them, the automobile industry was given top priority. The success of the Shanghai’s automotive and parts industries lies in the organizations of its municipal bureaucracy and a package of policy to promote and nurture the industry, which can be referred to as the ‘local industrial policy’. Those factors are summarized as follows. Shanghai Automotive Industry Company (SAIC) The automotive industry focused on passenger car and automotive parts production. The target for passenger automobile production was to strengthen the production of Volkswagen, especially to establish a manufacturing centre for high-grade passenger cars. To this end, the Shanghai

212 Chinese Automotive and Parts Industries Table 5.6

Major parts manufacturers in Shanghai

Name of ﬁrm

Number of employees

Sales (¥m.)

10,957 99 913 500 392 793 400 661 726 94 365

36,265 488 226 306 380 456 961 871 731 104 69

60 81 39 * 559

64 50 25 * 247

410 123

340 52

60 81 60 82

64 50 16 9

85 160 247

97 97 19

Anting, Jiading Qu Shanghai Volkswagen Automotive Co. Ltd Shanghai Boch Parts Co. Ltd Shanghai Shendeman Forging Co. Ltd Shanghai Henan Air Condition Equipment Co. Ltd Shanghai Honghu Silencer Works Automotive Parts Division of Honghu Machinery Works Shanghai Automotive Brake System Co. Ltd Shanghai Caiaifu Steering Wheel Co. Ltd Shanghai Koito Headlamps Co. Ltd Shanghai Kangdi Automotive Brake Co. Ltd Shanghai Feizhong Automotive Fittings of Machine Co. Ltd Qingpu Qu Shanghai Kang’di’tai’ke Automotive Parts Co. Ltd Shanghai San’da Automotive Fittings of Machine Co. Ltd Shanghai Tai’hou Automotive Fittings of Machine Co. Ltd Shanghai Jia’lin Automobile Co. Ltd Shanghai Sa’ke’si Module of Dynamic Automotive Parts Co. Ltd Shanghai Shen’ya Sealing Parts Co. Ltd Shanghai He’da Automotive Fittings of Machine Co. Ltd Baoshan Qu Shanghai Zhong’xing Car Suspension Co. Ltd Shanghai Zhong’xu Spring Making Co. Ltd Shanghai Zhong’wei Spring Making Co. Ltd Shanghai Bao’shan Yuan’zheng Huan’wei Machinery Co. Ltd Shanghai Huo’fu-Automotive Locks Co. Ltd Shanghai Qian’wei Wood Works Shanghai Xiang’ming Axle Works *= unknown. Source: China Automotive Industry Year Book 2003.

Automotive Industry Company (SAIC) was established to target collaboration with the world’s top 500 companies, and its objective was to achieve sales of ¥100bn, and proﬁt of ¥10bn per year. The goals of production apart from passenger cars were 10,000 trucks 5500 large buses, 1 million motorcycles, and 20,000 tractors. Another important aim was to foster the automotive parts industry. The investment ratio of the passenger car projects and parts industry projects was about 50–50 per cent. The policy aimed to develop passenger car pro-

Masatsugu Tsuji and Wu Quan 213

duction as well as parts production simultaneously. Moreover, in order to impose market competition on former parts enterprises and promote cost cutting and quality upgrading, the Shanghai municipal government gave a lot of ﬁnancial support and implemented tax exemption polices. In order to foster Shanghai’s automotive parts manufacturing industry for high-grade passenger cars, the SAIC acquired lots of beneﬁts through the policy, since the Chinese side of joint ventures were subordinate to the SAIC. The upgrading of the ability of its subsidiaries could greatly accelerate the development of Shanghai’s automotive industry. Municipal Automotive Small Group (MASG) In 1985, although there was still no nationwide automobile industry policy, the Shanghai municipal government established the Municipal Automotive Small Group (MASG). The MASG took the lead in coordinating the city’s automotive production. Lead by Mayor Zhu Rong Ji (the future prime minister), the group included Vice-Mayor Huang, who succeeded Zhu as mayor in 1991. Huang had wide expertise and was interested in the industry, and he extended this concern during his tenure as mayor. Beijing and Guangzhou had small groups of automotive companies in the early 1990s. The small groups included one or more vice-mayors and representatives from the cities’ economic, planning and international trade committees, as well as representatives from their automotive industrial corporations. For instance, the Beijing group had 15 people in the fall of 1990, and Shanghai about 20. These small groups tried to chart a general course for the local industry, and to provide a forum for coordination among the different government bodies. The MASG in Shanghai had no obstacles to interfere with its management over the automotive sector. The MASG had absolute power to regulate the ﬁnance of companies in the city; they depended to a great extent on bank loans, which were under the control of the municipal economic committee, and the MASG could decide how much in funds each factory would receive. The rest of the MASG consisted of representatives from the municipal planning, economic, international trade and construction commissions, and the leader of the SAIC. The members came from various administrations. Each of the municipal bodies played an important role in SVW’s development: namely, the international trade commission helped its foreign exchange, and the economic and planning commissions charted the long-term localization process. The importance of the automotive industry was evident in the composition of some of these committees. Under the MASG, two bodies were speciﬁcally designed to support the SVW project. The Shanghai Localization Ofﬁce (SLO) consulted venture managers during the localization process. A Shanghai Santana Commonwealth, which included local research institutes and universities, was

214 Chinese Automotive and Parts Industries

designed to coordinate city-wide efforts to improve the production of the German-designed vehicle. Shanghai Localization Ofﬁce Under the direction of MASG, the Shanghai Municipality Trans-bureau Localization Cooperation Ofﬁce (SMTLCO) was founded. The SMTLCO played an important role in the localization of the Shanghai automotive industry. The bureaucracy of the Chinese government has always been strong, and the different bureaux have their own authority and jurisdiction. They cannot surpass the authority of other bureaux without the permission of upper leaders. The automotive parts factories could not cross the administrative border to directly provide parts to Shanghai Volkswagen. Under the direction of the SMTLCO, the supply of parts could be delivered without a divisional limitation. In other words, as the name of the SMTLCO showed, it was the ﬁrst time automotive parts supply relations were grouped outside bureau restriction in China. The establishment of the SMTLCO seemingly promoted the pace of the SAIC’s localization, but the construction of a trans-bureau to promote a company’s project had a signiﬁcant impact on the development of Shanghai’s automobile industry. And the trans-bureau’s organizations greatly assisted the SAIC in its efforts to merge the passenger car assembly factories and components supply factories in the Shanghai area. After the foundation of the SMTLCO the importance of localization was realized. In addition, the name was changed to the Shanghai Localization Ofﬁce (SLO) in 1987, and the vice-mayor of Shanghai, Wang, was inaugurated as the chief. This change upgraded the ofﬁce’s administrative level to a vice-municipality, higher than a bureau. The Shanghai Localization Ofﬁce can exclude interference from other parts of the administration. From then on, the localization of automotive parts was accelerated. Formation of the supply chain network Prior to a joint conclusion with VW, a conference about the localization of parts and the selection of a passenger car model was held in Shanghai. The Shanghai Municipal Automobile Small Group and the Chinese Automobile Industry Company (CAIC) held the conference, and imposed a large-scale technical remodeling schedule from 1984. Parts manufacturers that either possessed large production capability or owned high-quality products were invited to participate in the conference. The SAIC contacted more than 400 parts factories in the Shanghai area, inspected them and then selected latent partners from among them. Following the new schedule, new technology and equipment and other related items were introduced from abroad. Under the instruction of the MASG the and SLO, the SAIC established its own automotive supply chain system. The SAIC expanded the system to

Masatsugu Tsuji and Wu Quan 215

several auto-related groups. The supply chain system consisted of the Shanghai VW group, which is the main joint venture between the SAIC and VW, the SAIC group, and other concerned groups such as the Electronic Bureau group, the Transportation Bureau group, the Spinning Industry Bureau group and other agency groups. Parts manufacturers under SAIC accounted for only a low percentage in the early stages, but soon increased their presence. However, with the promotion of the MASG, the SAIC accelerated the process of mergers in the Shanghai area. In addition, the SAIC parts suppliers manufactured important components – suspensions, electronics and bodies. Other groups mainly produce multi-use parts. During the process to solve technological difﬁculties, the SAIC parts suppliers formed China’s primary parts manufacturers with many secondary parts suppliers, and this was similar to the case of the Japanese parts industry. The parts suppliers afﬁliated with the SAIC had at their disposal technology from VW, and were gradually absorbing technologies. Accordingly, the formation of an efﬁcient production method and managerial knowhow were important meaning in the development of the Chinese automotive industry. In this process, related agencies in Shanghai, such as the MASG and the SLO, played a substantial role. Expansion of the supply chain network nationwide In 1988, with the deliberation between Shanghai municipality and concerned agencies in the central government, one-third of components’ production and sale had been shifted to the Ministry of Aircraft group. The goal of this policy was to convert the local sourcing by the SAIC to national sourcing. Prior to the ﬁrst half of 1995, there were 281 parts factories which contracted with Shanghai VW or the SAIC. Some 167 parts factories in the Shanghai area accounted for 59.4 per cent. Moreover, of the 167 factories only 28 enterprises were subordinated to the SAIC, and they occupied about 35 per cent of the total localization rate. The other parts factories were subordinate to 14 different bureaux or general companies that had the same administrative level as a bureau, and those factories occupied 15 per cent of the total localization rate. Out of the Shanghai area, ﬁrst, the China Automotive Industrial Corporation (CAIC) group concentrated on the production of important items related to engines, electronics, and bodies. These important parts needed advanced technology, which required time to accumulate. Thus, CAIC group ﬁrms are thought to be primarily parts manufacturers. Moreover, if we investigate the items manufactured by the former Ministry of Aircraft and Spacecraft (MAS) group, they concentrated on safety-related and multi-use parts. This was due to the poor accumulation of automotive technology, and the group’s short history since entering into the automotive industry. The other suppliers accounted only for a very low percentage. Among these, the Ministry of Machinery Industry (MMI) group chieﬂy

216 Chinese Automotive and Parts Industries

manufactured subcomponents for main parts. In contrast, the General Vessel Corporation and the General Petrochemical Corporation group produced mainly multi-use products. Suppliers among the group known as the ‘Allied Forces’ were specialized as special automotive parts suppliers or multi-use parts manufacturers. Production structure of the Shanghai automotive industry A feature of the SAIC’s supply system is that the suppliers in the Shanghai area, or the SAIC group, serve as the most important part in the overall supply chain system. In addition, the CAIC group supplemented the feeble section of the SAIC group. They constitute the primary parts manufacturers, and serve as a key point for the success of the SAIC. On the other hand, the MMI group and the MAS group specialize in their own respective advantageous products. And, because of the poor technological quality, the MMI group and the MAS group supply the supplementary product for the SAIC and the CAIC. They therefore, constitute secondary parts manufacturers. Figure 5.1 depicts the structure of the SAIC’s supply chain network, which is similar to the Japanese hierarchical production structure. From the viewpoint of employment maximization in a certain area, the policy-promoted division-of-labour structure lacks rationality. But from the aspect of competition predominance it works rationally. As a result, the policy contributed substantially to the formation of a nationwide supply network that positioned the SAIC at the core. Furthermore, it contributed immensely to the growth of localization and the output of SAIC, compared with Tianjin and Guangzhou municipalities which compelled automakers to acquire supply only from their own respective jurisdictions. In sum, the success so far is due to local industrial policy, and had Shanghai municipality government not established the trans-bureau organizations, the development of the Shanghai automobile industry could not have acquired support that crossed different administrative borders, and the project could not have progressed smoothly.

Figure 5.1

Production structure of the Shanghai automobile industry Assemblers Primary parts manufacturers

Secondary parts manufacturers

Shanghai VW SAIC, CAIC and other groups

MAS, MMI and other groups

Masatsugu Tsuji and Wu Quan 217

Concluding remarks Economic as well as political factors Thus we may analyse how Chinese automotive assemblers and their parts manufacturers choose their locations. Economic factors as well as political factors are important in China. The former imply that locations are decided as a result of rational behaviour to maximize their proﬁt subject to various constraints, which include not only costs such as transportation, transactions and information but also the size of the market. By achieving a greater amount of sales, automobile assemblers tend to ﬁnd their locations close to populated areas, for instance. After the decision of automotive assemblers, parts suppliers tend to concentrate around the assemblers’ factories so as to minimize transportation cots. The more ﬁrms concentrate on the particular region, the larger the information ﬂow such as managerial know-how and technological innovation, and all participants there receive beneﬁts from this. This cumulative process attracts more economic resources to the region. However, the story does not end as economic theories suggest. In addition to the economic factors, politics are also important for the Chinese automotive industry, as discussed in the previous sections. At the beginning, the location of assembling plants was chosen mainly from the viewpoint of national security so as to scatter factories around the country. Those locations of the assemblers became a core to attract other parts manufacturers, and this process created the agglomerate regions of the automotive industry such as Jiangsu, Sichuan, Zhejiang, Jilin and Shandong provinces, and Shanghai and Beijing cities. The situation is quite similar to that suggested by the core–periphery model. The difference between the conclusion the theory suggests and the reality in China, however, is quite clear. The Chinese situation was mainly created by political factors. In this chapter, we have emphasized the role of local governments in the above process. After they obtained freedom to seek their own strategy to develop their economy, they made full use of its political power to nurture the automotive industry according to their local industrial policy. The aftermath was a larger number of automotive assemblers and their inefﬁcient level of production in comparison with those in other economies. Success factors in Shanghai Among Chinese regions, however, Shanghai shows remarkable development in the automotive and parts industry. Today, almost half of parts manufacturers are located in Shanghai and its nearby provinces such as Jiangsu and Zhejiang. The success lies in its historical background of manufacturing industry before the establishment of the People’s Republic of China. We have emphasized the part played by the organization of municipal governments such as the Shanghai Localization Ofﬁce (SLO), the

218 Chinese Automotive and Parts Industries

Shanghai Municipality Trans-bureau Localization Cooperation Ofﬁce (SMTLCO) and the Municipal Automotive Small Group (MASG). The bureaucracy of the Chinese government is so strong that the different bureaux have their own authority and jurisdiction. They cannot surpass the authority of other bureaux. The organizations of the Shanghai municipal government took a strong lead in nurturing and promoting its automotive industry. In addition to the political reasons, the fact that the Shanghai area has attracted the largest amount of foreign direct investment (FDI) so far has made it the most heavily agglomerated area in China. Characteristic of its FDI is the variety of industries, especially those dependent on high technology such as electronics, biotechnology and communications. One of its industrial parks is known as ‘Silicon Valley in Shanghai’. The origins of foreign-afﬁliated ﬁrms are the USA, Europe, Japan, Hong Kong and Taiwan. This is different from the Guangdong area, where Hong Kong and Taiwanese ﬁrms mostly dominate. Because of this, domestic ﬁrms have been receiving the beneﬁts of transferring technologies and managerial know-how. This is the most important effect of industrial agglomeration. Japanese automotive and parts manufacturers in China Let us now discuss examples of Japanese automotive assemblers and parts suppliers in terms of their locations. In Japan, they have a close relationship by way of forming a hierarchical production structure, and major Japanese automotive assemblers such as Toyota and Honda have already established plants in China, while Nissan will be producing soon. Let us brieﬂy examine their relationship with their parts suppliers. The Toyota Motor Corporation, for example, started its production in Tianjin in 2002, with an annual production of 30,000.10 It is now producing 50,000 vehicles. The current supply chain network is as follows: engines are supplied by its own nearby factory, which is the same as in Japan. As for the other parts, 75 per cent are supplied by Japanese-afﬁlated manufacturers, while Chinese ﬁrms supply 25 per cent. Toyota purchases 70 per cent of parts from ﬁrms in the Tianjin area, with 20 per cent from Shanghai and 10 per cent from Chongqing. Parts from Shanghai are collected by trucks dispatched by Toyota, and take 36 hours to reach Tianjin.11 Due to the small amount of production, Toyota cannot ask them to ﬁnd a location close to its factories. Let us take another example of a secondary parts supplier located in Tianjin, some 40 minutes by road from the Toyota plant, which is almost the same as its factory in Toyota City, Japan. It delivers parts eight times a day, the same as in Japan. The amount of production of parts is adjusted to that of Toyota’s output, that is 5000 vehicles monthly. It carries two days, inventories of materials, since some of them are produced by its allied ﬁrm located in near Shanghai.

Masatsugu Tsuji and Wu Quan 219

Let us take another example of a Japanese automotive assembler. Honda in Guangdong produced approximately 60,000 vehicles in 2002.12 Its assembling factory handles 1100 parts, which are supplied by approximately 100 ﬁrms; of these, 60 per cent is produced by its group suppilers located in China, while 30 per cent comes from local ﬁrms, the rest from other East Asian countries such as Malaysia, Thailand and the Philippines. Since Honda’s parts suppliers are so scattered, it carries two-day inventories. The reason that the ratio of local parts is lower than that of Toyota in Tianjin is its location in Guangdong, where the parts industry is less agglomerated. It can be said in summary that at present Japanese automobile assemblers cannot exploit the merits of agglomeration through the hierarchical production structure, since the amounts they produce are far below the threshold. Further issues As discussed so far, the Shanghai area became the centre of automotive and parts industries. It is not sure, however, that the automobile will concentrate further in this area. One reason is guidance of the Chinese government, which is asking foreign assemblers to locate in other areas. Also the motorization of China has only just started and a rapid growth of demand for automobiles is expected. This may become a trigger for further competition among the regions to attract industries and ﬁrms, with further upheaval taking place. In this context, it is not certain what kind of agglomeration theory will be applicable, or if a new theory will be required to explain and predict it. Regarding Shanghai’s automotive parts manufacturing, we have just pointed out the factors of its growth. Although so far there have been some researches at industry level in the region (see Marukawa 2003, Lee 1997 and Mizuno 2003, for example), it is not clear how the automotive parts industry has been formed. In case of Toyota’s parts suppliers, it is said that indigenous technology in the region such as the spinning and cotton industries became basis of the new industry. In the case of Shanghai, seeking the origin of its parts industry will help in the design and implementation of local industrial policy for other developing economies.

Notes 1. For more detail, see Tsuji 2003. 2. The former name of Toyota City, where most of Toyota factories are located, as mentioned earlier. 3. China Automotive Industry Year Book 2003, pp. 444–6. 4. Marukawa 2003 reports that the average number of automotive assemblers with which parts suppliers make transactions is about 2.32, while that of Japanese parts suppliers is 2.88. He also claims that the area where Chinese parts suppliers are situated is a huge rectangle with sides of 1800–3600-km while Japan’s is 1000 km or less away.

220 Chinese Automotive and Parts Industries 5. The Soviet method implies the large scale of production of a single product and high ratio of domestic production so as to exploit economies of scale. 6. For instance, the Korean government supported only 2 or 3 national automotive companies. 7. It was signed with much ceremony in October 1984, during a visit of West German Chancellor Helmut Kohl. Prime Minister Zhu Rongji also attended the signing, which captured much Chinese attention for such a huge foreigninvested project. 8. Retired German engineers were invited to transfer their skills and know-how. See Lee 1997 for details. 9. The supply chain system is also referred to as the value chain system. In Chapter 4, the latter term is used. 10. This section is based on ﬁeld research conducted in September 2003.The data here were provided by the president of Toyota Tianjin through personal interview, and do not always agree with the ofﬁcial data. 11. This method of delivering parts is referred to as the ‘milk-run’ by Toyota. Although it takes long hours, it is useful to have information on whether particular parts are obtained or not 36 hours prior to arriving in Tianjin. 12. This section is based on ﬁeld research conducted in August 2002.

Bibliography and references Barnett, A. Doak. 2000. Cadres, Bureaucracy, and Political Power in Communist China. New York: Columbia University Press. Coase, Ronald H. 1937. ‘The Nature of the Firm,’ Economica, 4, 386–405. Editorial Committee. 1996. The History of China’s Automotive Industry. Beijing: People’s Transportation Press (in Chinese). Fujita, Masahisa, Paul Krugman and Anthony Venables. 1999. The Spatial Economy: Cities, Regions, and International Trade. Cambridge, MA: MIT Press. Harwit, Eric. 1995. China’s Automobile Industry: Policies, Problems and Prospects. London: Sharpe. Ikeya, Kaichi. 1997. ‘Development of the Automobile Industry in Shanghai’, in Mitsuhiro Seki and Kaichi Ikeya (eds), Japanese Enterprises and Chinese Automotive Industry. Tokyo: Shinhyoron (in Japanese). pp. 136–70. Krugman, Paul. 1991. Geography and Trade. Cambridge, MA: MIT Press. Lardy, Nicholas R. 1987. China’s Entry into the World Economy. Berkeley, CA: University of California Press. Lee, Chunli. 1997. Modern China’s Automotive Industry: The Evolution of Corporation System and Strategy. Tokyo: Shizansha (in Japanese). Mann, Jim. 1997. Beijing Jeep: A Case Study of Western Business in China. Boulder, CO: Westview. Marukawa, Tomoo. 2003. ‘Supplier Network of the Chinese Automobile Industry’, Research in Social Sciences, Vol. (3) 125–52 (in Japanese). Mizuno, Junko. 2003. Automobile, Parts, Mold, and Machine Tools in Asia – Industrial Relation and International Competitiveness. JETRO-IDE. O’Brien, Peter and Yannis, Karmokolias. 1994. Radical Reform in the Automotive Industry: Policies in Emerging Market. Washington, DC: World Bank. Oksenberg, Michel. 1982. ‘Economic Policy-Making in China: Summer, 1981’, China Quarterly, Vol. 90.

Masatsugu Tsuji and Wu Quan 221 Pearson, Margaret. M. 1991. Joint Ventures in the People’s Republic of China. Princeton University Press. Shiomi, Haruo. 2001. China’s Automobile Industry in Shift Term. Tokyo: Nihon Keizai Hyoronsha (in Japanese). Tsuji, Masatsugu. 2003. ‘The Relation between Toyota Motor Corporation and Its Parts Suppliers in the Age of Information and Globalization: Concentration and Dispersion’, in Mitsuhiro Kagami, and Masatsugu Tsuji (eds), Industrial Agglomeration: Facts and Lessons for Developing Countries. Chiba: Institute of Developing Economies. pp. 19–41. Wang, Jixian J. and W. Liu. 2000. ‘Dynamic Relationship between Central Government and Local Authorities in Orienting Foreign Direct Investment: A Case of Automotive Industry’, in S.M. Li and W.S. Tang (eds), Spatial Economic Development in Mainland China. World Motor Vehicle Data, Southﬁeld, MI: Ward’s. Hong Kong: Chinese University Press. Williamson, Oliver. 1989. ‘Transaction Cost Economics’, in R. Schmlensee and R. Willig. (eds), Handbook of Industrial Organization, vol. 1. New York: NorthHolland. Xiao, Wei. 2000. The Management Structure of the Automobile Industry in China. Kyoto: Koyo Shobo (in Japanese). Yahuda, Michael. 1982. China’s Foreign Relations and the Modernization Programme. London: Academic. Yang, Xiaohua. 1995. Globalization of the Automobile Industry: The United States, Japan and the People’s Republic of China. Westport, CT: Praeger.

This page intentionally left blank

Part III Industrial Clusters and IT Industries

This page intentionally left blank

6 Can Malaysia’s Growth Be Sustainable? Yumiko Okamoto

Introduction Since the mid 1980s Malaysia has been successful in attracting a large number of world-class electronics ﬁrms and forming an industrial agglomeration, especially in the electronics industry. As a result, Malaysia has been a leading producer of electronics products in the international market, and the industry has been an important engine of growth. It is often said, however, that since the international competitiveness of Malaysia is being eroded partly owing to the rapid emergence of China as an industrial power, it may not be possible for Malaysia to sustain its current industrial growth. The objective of this chapter is therefore to investigate the validity of the argument. More speciﬁcally, it examines ﬁrst of all, the current ﬂow of foreign investment into Malaysia’s electronics industry and the extent to which it still contributes to the formation of industrial clusters. Second, the chapter observes empirically the international competitiveness of Malaysia, particularly in the electronics industry, using various measures. It also sees whether the Malaysian and Chinese economies are more competitive than complementary or vice versa. Finally it investigates whether the new rise of the ICT industry in Cyberjaya in Malaysia may constitute another industrial cluster to sustain the growth of Malaysia. The chapter is structured as follows. First comes an analysis of the degree to which the electronics industry has been forming an industrial agglomeration in Malaysia. This is performed according to both the international and the domestic context. Second, several statistical measures such as market share in major international markets and the revealed comparative advantage (RCA) index are calculated to examine the industrial competitiveness of Malaysia. It is interesting here to see whether Malaysia is really losing the industrial dynamism in its economy or not. Third, Malaysia and China are compared in both investment and trade structure to see whether China is or is not posing a threat to the Malaysian economy, or whether 225

226 Can Malaysia’s Growth Be Sustainable?

the two countries could be complementary to each other in trade and production. The last section of the chapter provides a preliminary assessment of the Multimedia Super Corridor (MSC), a project expected to create a new industrial cluster consisting mainly of high-tech industries modelled after Silicon Valley in the USA. Although it is too early to provide a comprehensive assessment of the MSC, it is interesting to see whether we can observe the emergence of a new wave of industrial dynamism in Malaysia.

Formation of the electronics industrial agglomeration in Malaysia Importance of the electronics industry in Malaysia Since its establishment in the early 1970s1 the electronics industry has been the driving force behind the rapid transformation of the Malaysian economic structure away from a primary commodity producer to manufacturing. The electronics industry became the biggest export earner in the mid-1980s (Reed Electronics Research 2000:136). Since that time, the industry has continued to be the largest contributor to export earnings, investments, industrial output, value added and employment (MITI Malaysia 2003:91). Entering the twenty-ﬁrst century, the electronics industry continues to occupy a signiﬁcant share of total export revenue (around 68.5 per cent).2 Although the industry was long dominated by component manufacturers, especially the makers of semiconductors, the industrial base has been diversiﬁed and the industrial structure has deepened since the mid 1980s. Malaysia now produces and exports a large volume of both consumer and industrial electronics as well as their parts and components. Besides, the industry continues to account for a large share of total production, value added, employment and investment in Malaysia. In 2000, the industry’s share of total production, value added and employment for the manufacturing sector were 42, 34 and 31 per cent respectively.3 Moreover, the industry accounted for 33.6 per cent of total approved investments in Malaysia in 2002 (MITI Malaysia 2003:91). The formation of industrial agglomeration in an international context Without any doubt, development of the electronics industry in Malaysia has been stimulated by the massive inﬂow of foreign direct investment. As of the end of 1998, the contribution of Japan to the formation of ﬁxed assets in this industry was the largest (30 per cent), followed by the USA (20 per cent) and Singapore (7 per cent).4 Japanese FDI ﬂow into Malaysia increased very rapidly in the latter half of the 1980s owing to push factors in Japan and pull factors in Malaysia (Okamoto and Urata 1997:21). On the other hand, domestic investment accounted for only 17 per cent in the electronics industry.

Yumiko Okamoto 227 Figure 6.1 of 2000

Number of Japanese ﬁrms established in ASEAN countries at the end

Number of establishments

120 100 80

Consumer electronics Industrial electronics

60

Parts and devices 40 20

Vi et na m

In do ne si a

Ph ili p pi ne s

al ay si a M

Th a

ila n

d Si ng ap or e

0

Source: Electronics Industry Association of Japan (2001), List of Overseas Establishments.

Figure 6.1 shows the total number of establishments of Japanese electronics ﬁrms in Malaysia as well as in other ASEAN countries (Indonesia, the Philippines, Singapore and Thailand) as of the end of 2000. It is very clear that Japanese MNCs in the electronics industry, in terms of both assemblers (consumer and industrial electronics) and parts and components suppliers, agglomerated most in Malaysia among ASEAN countries. What distinguished Malaysia from the rest is that not only assemblers but also a large number of the parts and components suppliers were established within a short period of time. Between 1986 and 1995, as many as 83 out of the current total number of 103 establishments started their operations in Malaysia. The industrial agglomeration, particularly in the parts and components sector, laid out a solid basis for further development of the electronics industry. According to MITI Malaysia (2003:75), to date 109 international procurement centres (IPCs) have been approved, with a cumulative estimated turnover of RM40.9bn, and operating expenditure of RM4.1bn. More than half of them (56 IPCs) were approved to Japanese companies largely in the electronics parts and components industries. IPCs are one of the manufacturing-related services identiﬁed for promotion and development in line with the Second Industrial Master Plan strategy for developing the ‘manufacturing plus-plus’ activities of the value chain (MITI Malaysia 2003:73). Geographical dispersion of the electronics industry within Malaysia Although a large number of world-class foreign electronics ﬁrms agglomerated in a small country such as Malaysia within a short period, the industrial

228 Can Malaysia’s Growth Be Sustainable?

activity tends to disperse rather than concentrate in a certain city or state within the country. According to recent studies, there is a trade-off between dispersal and agglomeration, or centrifugal and centripetal forces (Neary 2001:536). Since modern sectors such as the manufacturing sector are often characterized by increasing returns to scale, spatial agglomeration may be a possible outcome, though not an inevitable one. According to Kimura (2003:215), production in the electronics industry is highly concentrated on the coastal area in the case of China. Gini coefﬁcients of the regional production pattern for 2000 show the values of more than 0.7 for three subsectors: electric equipment and machinery, electronic and telecommunication equipment, and ofﬁce machinery. This high concentration of regional production is also observed in Thailand. According to Okamoto and Huang (2003), Gini coefﬁcients of both foreign and domestic regional production show extremely high values. Both exceeded 0.9 in 1996. On the other hand, the industrial activity of the electronics industry tends to be dispersed in Malaysia. Table 6.1 shows the regional distribution of foreign and domestic investments in Malaysia between 1993 and 2002.5 According to the table, investment in ﬁxed assets in the electronics industry tends to be dispersed among such states as Johor, Kedah, Melaka, Negeri Sembilan, Pahang, Perak, Penang, Sarawak, and Selangor. Although there are some states with little investment in the electronics industry, their number Table 6.1 Regional shares of foreign and domestic investment in Malaysia between 1993 and 2002 (per cent) State KL Labuan Selangor Penang Perak Johor N. Sembilan Melaka Kedah Pahang Kelantan Trengganu Perlis Sabah Sarawak Total

Foreign

Domestic

0.2 0.0 15.4 29.0 2.2 12.8 6.0 7.9 22.0 0.8 0.3 0.1 0.1 0.0 3.2 100.0

8.2 0.0 19.6 10.1 13.7 6.4 4.5 6.7 16.9 7.8 0.3 0.0 0.1 0.5 5.2 100.0

Source: Based on data of MIDA. The data of both foreign and domestic investments are those of the approved projects.

Yumiko Okamoto 229

is rather small.6 Gini coefﬁcients of foreign and domestic regional investment patterns are 0.67 and 0.50 respectively for the above period. The values of Malaysia are signiﬁcantly lower than those of China and Thailand mentioned above. Since the regional pattern of industrial activity seems to inﬂuence regional income distribution substantially (Ikemoto 1991:67–8), the geographical dispersion of electronics industrial activities will generate more egalitarian regional development in Malaysia than in other developing countries such as China and Thailand.

International competitiveness of the Malaysian industrial agglomeration It is often said that the industrial competitiveness of Malaysia, especially in the electronics industry, is being eroded rapidly. This section investigates its validity from three different points of view. First of all, the recent inﬂow of FDI to the electronics industry in Malaysia is examined; foreign investor conﬁdence and interest in the electronics industry in Malaysia is one way to measure its international competitiveness. Second, two trade measures are also calculated to investigate the international competitiveness of Malaysia: the share of various Malaysian electronics products in the three major international markets such as Japan, the USA and the EU, and the revealed comparative advantage (RCA) indexes of various Malaysian electronics products. The recent ﬂow of FDI into the electronics industry Figure 6.2 shows the ﬂow of FDI to the electronics industry in Malaysia on an application basis and the number of project applications since 1996. Although the FDI ﬂow showed a sharp drop after the 1997 Asian crisis, it went up enormously again in 2000. Besides, the ﬂow of FDI in terms of the number of applications has not dropped signiﬁcantly, even though the majority of applications are made not for new investment projects but for the expansion and/or diversiﬁcation of existing ones. Moreover, Okabe (2003) observes that although foreign applications for manufacturing projects dropped in 2002, they bounced back again, especially in the electronics industry in mid 2003. Very recent applications seem to concentrate on the production of high value added products such as plasma TVs. This seems to reﬂect the prevalent strong level of conﬁdence among foreign investors in investments undertaken in Malaysia. All the more importantly, Malaysia is attracting new FDI in supporting industries in recent years. For instance, in 2002 the engineering supporting industry encompassing the manufacture of parts and components such as moulds, tools and dies, and the provision of supporting services such as metal surface treatment and/or ﬁnishing and heat treatment registered a total of 56 applications, 72 per cent of which were foreign (MIDA 2003:52).

230 Can Malaysia’s Growth Be Sustainable? Figure 6.2 Foreign investment in the electronics industry of Malaysia on the application basis 5000

250

4500 4000 Number of applications

3500

US$m.

3000

150

2500 100

2000 Value of investment

1500

Number of applications

200

50

1000 500

0

0 1996

1997

1998

1999

2000

2001

2002

Source: Malaysian Industrial Development Authority.

Although most of them are small projects involving investments of less than RM10m., development of such supporting industry is essential for Malaysia to maintain international competitiveness in the manufacturing sector. Development of supporting industry for the electronics industry is also seen in such industries as fabricated metal, chemical and plastic products. As MIDA (2003:68) notes, the increasing investments by foreign companies in the electronic and semiconductor industries to undertake expansion and/or upgrading of technology has encouraged the supporting services companies to make new investments in Malaysia and/or to expand their operations in the country. Despite the close of some establishments and the shifting of their operations to lower-cost production countries such as China and Vietnam, many existing ﬁrms continue to engage in expansion or diversiﬁcation of their investment projects. Besides, in recent years new investments in supporting industries have been seen widely. It seems that Malaysia is ﬁnding a way to survive global competition in the electronics industry. Trade analysis Market shares of electronics products in Japan, the USA and the EU The above ﬁndings are also conﬁrmed by trade analysis as well. Table 6.2 shows the market shares of Malaysia in the major international markets such as Japan, the USA and EU for the years 1996, 1998 and 2000.

Table 6.2

Market shares of Malaysia in major developed countries (%) Japan

SITCR3 751 752 759 761 762 763 764 771 772 773 774 775 776 778

Description Ofﬁce machinery Data processing machinery Parts of 751, 752 TV Radio etc. Sound recorders Telecommunication equipment Electric power machinery Switches etc. Distribuiting electrical equip. Medical equipment Household equipment Transistors, ICs etc. Other parts

USA

EU

1996

1998

2000

1996

1998

2000

1996

1998

2000

12.1 6.5 1.8 29.1 32.5 44.2 9.6 7.4 5.0 1.9 0.0 5.4 5.5 1.3

7.5 10.9 1.3 29.1 32.1 33.0 7.4 5.9 4.2 1.4 0.0 6.3 4.9 2.3

12.6 9.0 7.4 39.7 28.3 28.0 6.2 5.8 4.1 1.3 0.0 4.7 10.3 4.2

3.1 7.0 4.5 11.7 23.5 15.5 6.5 2.8 1.8 0.6 0.0 2.6 13.5 1.5

2.3 6.9 11.6 7.8 15.0 10.8 4.4 2.2 1.8 0.7 0.1 2.0 12.6 1.8

3.9 8.4 10.7 10.9 13.0 10.8 5.1 1.9 1.6 0.4 0.0 1.5 12.9 1.0

1.0 2.9 2.1 1.0 17.1 5.2 1.9 2.0 0.7 0.3 0.1 0.0 6.8 1.1

0.9 2.1 5.2 0.3 12.6 4.0 1.4 1.6 0.8 0.2 0.0 0.0 5.2 0.8

1.0 2.6 2.9 0.2 12.2 5.1 1.8 2.0 0.9 0.4 0.1 0.1 5.5 0.8

Source: Based on trade data from PC-TAS (personal computer trade analysis system), United Nations CD.

231

232 Can Malaysia’s Growth Be Sustainable?

According to it, Malaysia lost signiﬁcant market share in such products as radios (SITCR3 762), sound-reproducing and recording equipment (SITCR3 763). This is consistent with the observation of MIDA (2003:3) that production of these labour-intensive products declined signiﬁcantly owing to the closing or downsizing of some manufacturing facilities, and relocation of manufacturing centres to reduce operational costs. To some extent, this trend continues in Malaysia owing to rising labour costs. On the other hand, Malaysia seems to have gained market share in Japan and the USA, especially in higher value added products such as ofﬁce machines, computers and peripherals, colour TVs, semiconductors and other parts and components. This implies that Malaysia is moving up the technological quality ladder. The recent ﬂow of FDI into higher value added electronics products7 is expected to intensify this shift. RCA index8 The RCA index of Malaysia also conﬁrms a shift in the trade structure of the country (Table 6.3). The RCA index indicates the competitiveness of each commodity within a country compared with the world trade structure. Any value above 1 indicates that Malaysia possesses a comparative advantage in the production of the particular product, any below that it does not. According to Table 6.3, there is no change in that Malaysia continues to specialize in the production of electronics products, since RCA indexes of Table 6.3

RCA index of Malaysia

SITCR3

Description

1996

1997

1998

1999

2000

71

Power generating machinery

0.369

0.294

0.280

0.287

0.268

72

Machinery for special industry

0.202

0.192

0.262

0.194

0.219

73

Metalworking machinery

0.142

0.172

0.190

0.171

0.161

74

General industrial machinery

0.422

0.317

0.319

0.323

0.325

75

Ofﬁce machines

2.138

2.497

2.734

3.309

3.360

76

Telecomm, sound equipment

3.910

3.367

2.945

2.588

2.561

77

Electrical machinery

2.466

2.487

2.623

2.552

2.274

78

Road vehicles

0.067

0.066

0.068

0.052

0.044

79

Other transport equipment

0.677

0.587

0.640

0.331

0.128

See note 8, p. 242, for the method of calculation. Source: Data from PC-TAS.

Yumiko Okamoto 233

SITCR3 75, 76 and 77 remain signiﬁcantly above 1 and the rest below. The magnitude of the relative strength tends however to decline in SITCR3 76 (including such products as radios and sound recorders), but increase in SITCR3 75 (including such products as ofﬁce machines and computers). This means that the industrial competitiveness of Malaysia shifts away from the production of low value added products to that of high value added ones. Analysis of the RCA index also conﬁrms that the overall industrial competitiveness of Malaysia is still maintained at least in the production of relatively high value added products.

Comparison between Malaysia and China in trade and investment structures It is also often argued that the emergence of China as an industrial power is a threat to ASEAN, including Malaysia.9 This section investigates the validity of this argument. The ﬂow of FDI Figures for FDI are often used as a testimony of a country’s ability to attract new investments amid growing competition for global FDI. Figure 6.3 indicates the ﬂow of FDI into both China and ASEAN.10 It is very clear that ASEAN cannot compete with China in terms of the absolute value of FDI inﬂow. Moreover, the gap seems to have become even wider in the recent years as the entry of China into the WTO turned round the declining trend of FDI ﬂow into that country,11 while the ﬂow of FDI into ASEAN remains stagnant. Figure 6.3

Flow of FDI into China and ASEAN

50000 45000 40000

US$m.

35000 30000 25000

China

20000

ASEAN

15000 10000 5000 0 1994

1995

1996

1997

Source: Asian Development Bank, Key Indicators 2001.

1998

1999

234 Can Malaysia’s Growth Be Sustainable?

Although varying across countries, the ﬂow of FDI into ASEAN has not, in general, been remarkable in recent years. In the early 1990s the original ASEAN member countries united themselves and decided to form a freetrade area, mainly to compete with China in attracting FDI. The action was to some extent successful in bringing back the ﬂow of FDI into ASEAN again in the mid 1990s. The loss of dynamism in ASEAN after the 1997 Asian Crisis,12 however, made it difﬁcult for ASEAN to reunite and implement a common policy to compete with China. It is important however to note that the trend of FDI tends to exaggerate the strength of China as a new industrial power since a number of foreign ﬁrms are shifting their production basis because they are attracted by the potentiality of the Chinese domestic market. This means that some are MNCs shifting the site of their production regardless of the competitiveness of ASEAN countries. The ﬂow of FDI into China will become a serious threat to ASEAN if MNCs make the former an important production site for export at the cost of the latter. It is not, however, easy to distinguish between the use of FDI for domestic sales and for exports. Trade analysis RCA index The analysis of the trade structure is also useful to examine whether China may pose a threat to Malaysia. First of all, the RCA indexes are calculated for both countries at the two-digit level of SITC R3. Those indexes are ranked for each country respectively and the Spearman’s rank correlation coefﬁcient between the ranking of two countries is calculated. If they show a high positive correlation, their trade structure is very competitive. That means that the emergence of China as an industrial power may cause a signiﬁcant impact on Malaysia. On the other hand, if the coefﬁcient is near zero or negative, their trade structures are rather dissimilar and they could complement each other to some extent. For comparative purposes, other countries are included in the analysis. Table 6.4 shows the results. First of all, the Spearman’s rank coefﬁcient between Malaysia and China is very low (close to zero) and statistically insigniﬁcant. Moreover, the value of the coefﬁcient not only has become lower since 1996 but also became negative in 2000. This means that the export structures of Malaysia and China are not only dissimilar but have become even more so lately. Table 6.4 also indicates that Malaysia has a rather similar export structure to that of Korea, implying that Malaysia has been upgrading the quality of products and will compete more with Korea rather than with China. Market share analysis Although Malaysia and China are more or less complementary to each other in the overall trade structure, both could be very competitive within

Yumiko Okamoto 235 Table 6.4 Spearman’s rank correlation coefﬁcients of the RCA indexes between Malaysia and some of its major trading partners 1996

1997

1998

1999

2000

China

0.054

0.101

0.082

0.028

Japan

0.124

0.140

0.166

0.132

0.113

–0.192

–0.204

–0.190

–0.154

–0.177

0.178

0.227*

0.228*

0.234*

USA Korea

0.325***

India

–0.081

–0.097

–0.078

–0.065

Australia

–0.244*

–0.282**

–0.333***

–0.315**

–0.028

n.a. –0.273**

n.a.= not available. *= Statistical signiﬁcance at the 10 percent level. **= Statistical signiﬁcance at the 5 percent level. ***= Statistical signiﬁcance at the 1 percent level. RCA indexes of each country were calculated at the two-digit level of SITCR3.

the electronics industry. To examine this, the market shares of China in three major international markets are calculated in the same manner as in Table 6.2. Then the Spearman’s rank correlation coefﬁcients of the rankings of the market shares are calculated between the two countries in each international market (Japan, the USA and the EU). The results (see Table 6.5) show that the coefﬁcients are low in all markets and statistically insigniﬁcant. In 1996 the coefﬁcient was a little bit high (0.453) in the Japanese market, but statistically insigniﬁcant. Moreover, the coefﬁcient got even lower and became close to zero in 2000 (0.134) in the Japanese market as well. Considering the high degree of involvement of MNCs in the electronics industry of both countries, this may indicate that MNCs tend to produce different types of electronic products in China and Malaysia.

Table 6.5 Spearman’s rank correlation coefﬁcients of the market shares between Malaysia and China in electronics products

Japan

1996

2000

0.453

0.134

USA

0.147

0.116

EU

0.262

0.273

1. The market shares in three markets were calculated in the same manner as in Table 6.2. 2. None of the above coefﬁcients show statistical signiﬁcance at the 10 percent level.

236 Can Malaysia’s Growth Be Sustainable?

Intra-industry (IIT) index13 The competitiveness or complementarity of the trade structure between two countries can also be analysed by calculation of the intra-industry index. It takes a value between 0 and 1. The higher the index is, the more the two countries are engaged in intra-industry trade. Table 6.6 shows the IIT indexes at the one- or two-digit level of SITC R3. According to the table, Malaysia and China are highly engaged in intraindustry trade in the machinery sector (SITCR3 7). Besides this, within the Table 6.6

Intra-industry trade (IIT) indexes between Malaysia and China

SITCR3

1996

1997

1998

1999

2000

0

Food and live animals

0.291

0.141

0.094

0.100

0.082

1

Beverages and tobacco

0.107

0.592

0.091

0.205

0.295

2

Crude materials except fuels

n.a.

n.a.

n.a.

n.a.

n.a.

3

Mineral fuels

n.a.

n.a.

n.a.

n.a.

n.a.

4

Animal and vegetable oil

n.a.

n.a.

n.a.

n.a.

n.a.

5

Chemicals

0.304

0.412

0.441

0.474

0.357

6

Basic manufacturers

0.126

0.142

0.467

0.438

0.498

7

Machinery

0.625

0.569

0.705

0.791

0.717

8

Other manufactured goods

0.289

0.272

0.325

0.409

0.425

71

Power-generating machinery

0.348

0.562

0.465

0.415

0.349

72

Machinery for special industry

0.887

0.745

0.754

0.982

0.790

73

Metalworking machinery

0.322

0.422

0.682

0.882

0.601

74

General industrial machinery

0.928

0.538

0.867

0.857

0.519

75

Ofﬁce machines

0.854

0.954

0.974

0.764

0.614

76

Telecom, sound equipment

0.240

0.279

0.375

0.764

0.621

77

Electrical machinery

0.687

0.746

0.943

0.912

0.911

78

Road vehicles

0.583

0.474

0.993

0.853

0.680

79

Other transport equipment

0.368

0.315

0.302

0.048

0.616

n.a. not available. 1. Originally, the IIT index was calculated at the two-digit level of SITC R3. Then, the author aggregated them into the IIT index at the one-digit level using the total value of trade between two countries as a weight. The category of machinery (7) under the classiﬁcation of SITCR3 shows the original IIT indexes at the two-digit level. 2. See Note 13, p. 242, for the method of index calculation. Source: Data from PC-TAS.

Yumiko Okamoto 237

machinery sector the electronics industry in particular shows high values of the IIT index. For instance, both ofﬁce machines (SITCR3 75) and electrical machinery and parts (SITCR3 77) show the high value of the index (close to 1), although that of the former went down in 1999 and 2000. Moreover, the value of the IIT index in telecommunications and sound equipment (SITCR3 76) went up signiﬁcantly, from 0.24 to 0.7, during the latter half of the 1990s. This indicates that Malaysia and China tend to be complementary in trade and production in the electronics industry.

Rise of a new industrial cluster? The Multimedia Super Corridor (MSC) Former Prime Minister Mahathir introduced the concept of the Multimedia Super Corridor (MSC) in 1996.14 The MSC was the government’s ﬁrst major investment in the IT and multimedia sector to attempt to replicate the Silicon Valley model in the USA. The goal of the MSC was to transform Malaysia from a manufacturing-based economy to a knowledge-intensive society by 2020, enabling it to leapfrog into a high-technology industry (Jussawalla 2003a:19). First of all, strong emphasis was placed on the development of an ITrelated infrastructure. The government of Malaysia invested between US$1bn and US$2bn in the installation of a ﬁbre-optic network (Jussawalla 2003a:18) to provide a ﬁrst-class physical and information infrastructure. Despite economic setbacks in 1997 and 1998 due to the Asian Crisis, the government did not postpone investment in the MSC project. Second, in recognition of the enormous gap that still exists between Malaysia and technologically advanced countries, the MSC was designed to attract not only the attention of Malaysian IT and IT-related companies, but also overseas world-class, high-tech companies (Taylor 2003:105). To do so, the government offered a variety of incentives such as exemption from local ownership requirements (or freedom of ownership), allowance of unrestricted employment of local and foreign knowledge workers and generous ﬁnancial incentives, as well as the provision of a world-class physical and information infrastructure. To receive those incentives, or to obtain MSC status from the government of Malaysia, in turn, a ﬁrm must be a primary provider of IT and multimedia services and products, the employer of a large number of professionals and engaged in technology transfer in the case of foreign ﬁrms (Taylor 2003:98–9). Assessment of the MSC Critiques of the MSC project15 As stated previously, one of the important characteristics of the MSC was to provide a world-class digital infrastructure to attract leading-edge, globally well-known companies. To keep its promise, the government of Malaysia has spent billions of dollars so far on the construction of infrastructure. As

238 Can Malaysia’s Growth Be Sustainable?

Alvin Tofﬂer and Bill Gates pointed out to Mahathir, however: ‘The essence of Silicon Valley is not ﬁber optic cables, it is the creative, innovative drive with large numbers of people racing to create new ideas’ (Einhorn 1999: 86).16 Although the role of infrastructure is essential for the development of IT and multimedia industries, something else such as construction of a suitable business environment seems to be equally important. The other critique often heard, although much linked with the ﬁrst one, is that the MCS has failed to attract world-class companies. In 2003, there are 868 MSC-status companies, of which 667 are active.17 Most of the companies are however small and medium-sized ones with paid-up capital of less than RM500,000. Besides, active world-class companies seem to be only a small part of it.18 As Mahathir himself admitted, the MSC has not contributed as much as expected so far (Taylor 2003:113). Emergence of some promising signs It is still too early to give a fair assessment of the MSC. It was only in mid 2000 that major parts of the MSC such as the Cyberjaya, the Japanese NTT R&D Centre and the Central Incubator House were opened formally

R&D expenditure (RMbn)

Export/sales(%)

R&D/sales (%)

Capital expenditure (RMbn)

Employment

Export sales (RMbn)

Economic impacts of the MSC in Malaysia

Total sales (RMbn)

Table 6.7

n.a. 0.258 0.542 0.657

23.3 15.3 17.8 22.8

n.a. 6.6 9.3 8.2

0.048 0.058

53.3 40.5

6.0 5.3

Total 2001 2002 2003 2004

3.000 3.930 5.850 7.980

0.700 0.600 1.040 1.820

n.a. 17,000 17,854 22,398

n.a. 0.620 0.850 n.a.

World-class companies 2002 2003

0.796 1.100

0.424 0.446

n.a. n.a.

n.a. n.a.

n.a.= not available. Figures for the year of 2004 are estimates. Source: Data provided by the Multimedia Development Corporation.

Yumiko Okamoto 239

(Jussawalla 2003b:276).19 Some promising signs have however started to emerge since 2002. Table 6.7 summarizes the economic impact of the MSC, from data provided by the Multimedia Development Corporation (MDC). The MDC, owned and funded by the government of Malaysia, is responsible for promoting the MSC locally and globally, as well as supporting companies which are locating within the MSC. According to it, on entering the twenty ﬁrst century business activities seem to be picking up in the MSC, as both local and export sales values as well as investment and employment are increasing every year. Besides, they sell not only in the domestic market, but also in the international market as well. Moreover, they are also active in R&D activities as the ratio of R&D to total sales value is as high as around 8 or 9 per cent. The business activities of world-class companies also seem to be expanding. Currently, 47 world-class companies are in operation, of which 46 are fully owned by foreign companies. The composition of the nationality of those companies is shown in Figure 6.4. Although the number of US companies is the largest, many European companies as well as Japanese ones have been investing in the MSC. Besides, since 2001 two well-known Indian companies20 have started to operate in the MSC. Moreover, although the investment of world-class companies in R&D is still below the average (Table 6.7),21 their high export ratio indicates that Malaysia seems to possess high cost efﬁciency in the provision of IT and IT-related services. Figure 6.4

Numbers of world-class companies in the MSC by nationality, 2003

16 14 12 10 8 6 4 2

K In Vi di rg a in Is la nd s M al ay si a

Br

it i s

h

H

K an ce Si ng ap or e

U

Fr

U SA er m an N et y he rla nd s Be lg iu m Fi nl Sw a n d i tz er la nd G

Ja pa n

0

Source: Multimedia Development Corporation.

240 Can Malaysia’s Growth Be Sustainable? Table 6.8

Occupational grouping in Malaysia As at end 2002

Management Technical Support staff Total

4,218 10,407 2,375 17,000

25% 61% 14% 100%

Percentage shares of technical staff Software developer/programmer System analyst/designer Business analyst/consultant Systems/hardware engineer Web designer/developer Technical support staff Film director/editor Animator/graphic designer Content developer Other technical staff

32% 8% 8% 9% 3% 21% 2% 2% 2% 12%

Projected 2004 5,168 14,575 2,655 22,398

23% 65% 12% 100%

35% 8% 7% 9% 3% 19% 2% 4% 3% 10%

Source: Data from the Multimedia Development Corporation.

The MSC seems to be beginning to contribute to the Malaysian economy from the point of view of human resource development as well. To mitigate the shortage of skilled workers in this ﬁeld, Multimedia University, a private university owned by Telecom Malaysia, has been established within the MSC itself. It comprises four faculties: creative multimedia, engineering, management, and information technology. The ﬁrst intake of 1300 students to the new university was in 1998 (Taylor 2003:98). According to an interview undertaken by the author at the university in August 2003, almost every new graduate found a job within six months of graduation. Moreover, they found employment not only in Malaysia but also overseas. Since the nationality of both faculty members and students varies a lot, there is a chance that the university will play an important role as a supplier of skilled workers not only for Malaysia but also for other countries within the Asia-Paciﬁc region. The MSC also plays an important role in creating job opportunities for skilled workers. Although the number of employees is still small compared with that of workers in the electronics hardware industry (around 1 to 2 per cent), the percentage of skilled workers in the total number of employees of the MSC is extremely high (86 per cent).22 The occupational grouping of employees in the MSC is provided in Table 6.8. According to the table, the share of skilled workers is not only high but is also expected to increase by the end of 2004. Especially technical staffs such as software designers and developers are in high demand. And the role of the MSC in

Yumiko Okamoto 241

human resource development of new industries seems likely to be very important in the coming decade.

Conclusion Malaysia has been successful in forming an industrial agglomeration in the electronics industry, especially since the mid-1980s and the industry has become very important for the Malaysian economy as a source of value added, employment and export earnings. The uniqueness of Malaysia’s electronics industry is that although Malaysia has been attracting a larger number of foreign ﬁrms, especially in the parts and components industries, than have other countries in ASEAN, those ﬁrms tend to disperse geographically within Malaysia in comparison with other developing countries such as Thailand and China. This may generate a favourable effect on the pattern of regional development. Although the ﬂow of new foreign investment declined substantially in recent years and some of the ﬁrms relocate their production sites outside Malaysia, the majority of existing foreign ﬁrms remain to reinvest in Malaysia. Moreover, Malaysia is climbing up the quality ladder step by step, as is seen in its changing trade and investment structures. The emergence of China as an industrial power and its entry to the WTO is often said to pose a threat to ASEAN countries including Malaysia. However, the continuous reinvestment activities of foreign ﬁrms in the electronics industry of Malaysia seem to suggest that this may not be the case. Moreover, the trade analysis also suggests that Malaysia and China can be complementary in the electronics industry. Although the MSC may not have achieved the outcome as expected, there are now some promising signs. Malaysia has started to create another high-tech hub for both government and private sectors within the Asia–Paciﬁc region. The above analysis seems to suggest that the industrial dynamism of Malaysia is not yet lost. All the more importantly, Malaysia has begun to ﬁnd a new model of growth by combining both hardware and software in the electronics industry. Although it is too early to evaluate the MSC, there are some promising signs.

Notes 1. The development of the electronics industry goes back to investment by Intel in Penang in 1972, a pioneering step in the semiconductor industry in Malaysia. 2. According to Economic Report 2001/2002. 3. According to Census of Manufacturing Industry in Malaysia. 4. This is based on data from the Malaysian Industrial Development Authority (MIDA).

242 Can Malaysia’s Growth Be Sustainable? 5. Since no regional production data for the electronics industry are available, the ﬁgures of regional investment in ﬁxed assets are used instead. However, since the annual ﬂow of regional investment ﬂuctuates considerably, this study uses the regional accumulated stock values of the annual investment in ﬁxed assets between 1993 and 2002. 6. They are Kelantan, Labuan, Perlis, Sabah and Terengganu. 7. See Okabe (2003). 8. The RCA was calculated as follows: RCAij = (Xij/Xj)/(Xiw/Xw), where Xij is the export value of product group i of country j, Xj is the total export value of country j, Xiw is the world export value of product group i, and Xw is the total world export value. In this chapter, country j is Malaysia. 9. Both the challenges and the opportunities the emergence of China poses to other countries are summarized in Okamoto (2005). 10. Here, ASEAN includes original ASEAN members excluding Brunei (Indonesia, Malaysia, Philippines, Singapore, and Thailand) and new ASEAN members (Cambodia, Laos, Myanmar, Vietnam). 11. According to China Statistics Press, China Statistical Yearbook 2002. 12. See Mahani (2002) with respect to the details of the loss of dynamism among ASEAN and its consequences. 13. The IIT index was calculated as follows: IITij = [1–(|Xij–Mij|/(Xij + Mij))], where Xij is the export value of product group i of country j, and Mij is the import value of the same product group i of country j. In this chapter, country j is Malaysia. 14. See Mahathir (1998) for details. 15. See Taylor (2003: 112–14) for details. 16. This sentence was also cited by Jussawalla (2003b:277). 17. According to the MSC Impact Survey downloaded from http://www.mdc.com.my. 18. According to the information obtained from the Multimedia Development Corporation, 47 out of 55 world-class companies are active in 2003. 19. The Cyberjaya along with Putrajaya constitutes an important part of the MSC. 20. Satyam Computer Services Ltd and WIPRO Ltd. 21. According to Table 6.7, the ratio of R&D expenditure to total sales value of world-class companies is below the total average. 22. See Note 17.

References Asian Development Bank. 2001. Key Indicators 2001. Manila: Asian Development Bank. Department of Statistics, Malaysia. 2002. Census of Manufacturing Industries. Kuala Lumpur: Department of Statistics, Malaysia. Electronics Industry Association of Japan (2001). List of Overseas Establishments. Tokyo: Electronics Industry Association of Japan. Einhorn, Bruce. 1999. ‘Mahathir’s High Tech Folly’, Business Week, 29 March, p. 83. Ikemoto, Yukio. 1991. Income Distribution in Thailand. Chiba: Institute of Developing Economies. Jussawalla, Meheroo. 2003a. ‘Bridging the Global Digital Divide’, in Meheroo Jussawalla and Richard D. Taylor (eds), Information Technology Parks of the Asia Paciﬁc: Lessons for the Regional Digital Divide. New York: Sharpe. pp. 3–24.

Yumiko Okamoto 243 Jussawalla, Meheroo. 2003b. ‘The Role of Information Technology Parks in Bridging the Digital Divide’, in Meheroo Jussawalla and Richard D. Taylor (eds), Information Technology Parks of the Asia Paciﬁc: Lessons for the Regional Digital Divide. New York: Sharpe. pp. 254–84. Kimura, Koichiro. 2003. ‘China’s Regional Industrial Disparity from the Viewpoint of Industrial Agglomeration’, in Mitsuhiro Kagami and Masatsugu Tsuji (eds), Industrial Agglomeration: Facts and Lessons for Developing Countries. Chiba: Institute of Developing Economies. pp. 203–34. Mahani, Zainal-Abidin. 2002. ‘ASEAN Integration: At Risk of Going in Different Directions’, The World Economy, 25(9), 1263–77. Mahathir, Mohamad. 1998. Multimedia Super Corridor. Kuala Lumpur: Pelanduk. Malaysian Industrial Development Authority (MIDA). 2003. Report on the Performance of the Manufacturing Sector 2002. Kuala Lumpur: MIDA. MITI Malaysia. 2003. Malaysia International Trade and Industry Report 2002. Kuala Lumpur: Ampang. Ministry of Finance Malaysia. 2001. Economic Report 2001/2002. Kuala Lumpur: Percetakan Nasional Malaysia Berhad. National Bureau of Statistics, People’s Republic of China, 2002. China Statistical Yearbook 2002. Beijing: China Statistics Press. Neary, J. Peter. 2001. ‘Of Hype and Hyperbolas: Introducing the New Economic Geography’, Journal of Economic Literature, 39, No. 2, 536–61. Okabe, Mitsuhiro. 2003. ‘Rapid Recovery of Foreign Direct Investment into Malaysia’, Tsusho Koho, no.15523 (28 August), 1–2. Tokyo: JETRO (in Japanese). Okamoto, Yumiko. 2005. ‘Toward the Formation of an East Asian Regional Arrangement,’ Paciﬁc Asia 2022: Sketching Futures of a Region, edited by Simon S. C. Tay. Tokyo and New York: Japan Centre for International Exchange. Okamoto, Yumiko and Ing-Wei Huang. 2003. ‘Foreign Investment and Regional Development in Developing Countries’, paper presented at the JSPS-NRCT workshop titled ‘Perspectives of Roles of State, Market, Society, and Economic Cooperation in Asia’, Kyoto University, 6–7 November, 2003. Okamoto, Yumiko and Shujiro Urata. 1997. ‘Japan’s Foreign Direct Investment in Malaysia in the 1990s’, in Kazue Sugiyama and Stephen Leong (eds), Revitalisation of Japan’s Economy: Implications for Malaysia. Kuala Lumpur: ISIS. Reed Electronics Research. 2000. Yearbook of World Electronics Data 2000: Market Prospects to 2003. Oxford, UK: Reed Electronics Research. Taylor, Richard D. 2003. ‘The Malaysian Experience: The Multimedia Super Corridor’, in Meheroo Jussawalla and Richard D. Taylor (eds), Information Technology Parks of the Asia Paciﬁc: Lessons for the Regional Digital Divide. New York: Sharpe. pp. 64–118.

7 Bangalore’s Software Cluster Aya Okada*

Introduction Why do some industrial clusters remain robust while others do not? What helps build the competitiveness of such clusters, enabling them to thrive? This chapter empirically examines the conditions under which knowledgebased industrial clusters in developing countries build competitiveness, with particular focus on the dynamics of the local labour market, as an institutional mechanism for promoting innovation and strengthening industrial competitiveness. Various bodies of recent literature – in regional economics, industrial organization, and economic geography – focus on the dynamics of local and regional economic activities in agglomeration, particularly those clusters of small and medium-scale enterprises (SMEs) in technology-based or hightechnology industries, in both developed and developing countries. Prompted by some successful experiences of SME-centered industrial clusters in developed countries, such as Silicon Valley in the USA (Saxenian 1994), Emilia-Romagna in Italy (Piore and Sabel 1984; Pyke et al. 1990) and BadenWürttemberg in Germany (Schmitz and Musyck 1993), many developing countries have recently adopted the model of industrial clusters, as a promising driver of both regional and national economic growth.1 These clusters typically enjoy interﬁrm cooperation among horizontally networked ﬁrms, particularly SMEs working in the same industry or interrelated economic activities, in geographical proximity. Governments of many countries have provided generous support and protection to these ﬁrms in clusters. Yet, while some successful clusters continue to thrive, many more clusters have failed. In particular, regions in developing countries that try to develop knowledge-intensive industrial clusters may face sets of conditions and challenges different from those in developed countries. There is a growing recognition that a region’s economic performance and growth rely on ‘a set of relatively immobile resources – knowledge, skills, institutional and organizational structures’ (Breschi and Malerba 2001: 244

Aya Okada 245

817). Thus, it is crucial to create and develop the local mechanisms to generate and accumulate these resources for developing industrial clusters, particularly in knowledge-intensive industries such as information and communication technology (ICT).2 The most important ‘competitive assets’ (Amsden and Hikino 1994) in these industries are human knowledge and skills, which are often embodied in technical workers and tacit in nature. Thus, it is also crucial to ensure the adequate deployment, mobility and skills development of these workers in order to facilitate the creation and transfer of knowledge and skills within the cluster as well as between clusters. Various strands of the recent literature point to the important role of knowledge spillovers and externalities as a determinant of the growth of regional clusters and of a nation.3 Indeed, the theory of knowledge spillovers suggests that the spatial agglomeration of ﬁrms is most likely to lead to innovative activities in industries where tacit knowledge plays an important role. This is because tacit knowledge, as opposed to information, can be transmitted only informally, and typically demands direct and repeated contacts (Audretsch 1998, as cited in Breschi and Lissoni 2001: 982). However, this argument fails to explain why many knowledge-based industrial clusters are closely associated with distant clusters, often in other countries, and taking advantage of frequent knowledge transfer between clusters, as in the Silicon Valley–Hsinchu connection (Saxenian and Hsu 2001). Moreover, virtually no literature ‘has explored the ways in which knowledge is actually transferred among people located in the same geographical area’ (Breschi and Lissoni 2001: 994). Thus, a more careful examination is needed of the mechanisms that generate and transfer knowledge and skills in and out of a cluster. In particular, the local labour market, as an institutional mechanism, may play a critical role in channelling and mediating the supply of and demand for the knowledge and skills necessary to induce innovation and build competitiveness in clusters. Using a case study of Bangalore, India, this chapter focuses on the dynamics of the local labour market, which has mediated the demand for and supply of a skilled IT workforce and developed cooperative arrangements between ﬁrms and among the state, industry and educational institutions. Bangalore’s IT cluster is of particular interest. As is widely known, in the 1990s India emerged as one of the fastest-growing exporters of software and services in the global economy, and Bangalore became India’s largest IT (especially software and services) cluster. India is the secondlargest producer of IT professionals in the world (NASSCOM 2001), supplying IT engineers not only within India but also to major IT clusters in other countries such as the USA, the UK, Japan and Singapore. With its impressive growth record and global recognition, Bangalore has recently been dubbed ‘the Silicon Valley of India’. Some scholars argue, however, that the Bangalore IT cluster exhibits signiﬁcantly different characteristics from

246 Bangalore’s Software Cluster

Silicon Valley; they caution that the two should not be seen as comparable (Parthasarathy 2000; Saxenian 2001). Drawing on two rounds of ﬁeldwork I conducted in Bangalore in 2001 and 2002, this study speciﬁcally poses two interrelated questions. First, given the ever-accelerating speed of innovation, and the increased mobility of skilled IT workers within the ﬁrm and between ﬁrms, in different locations and often beyond national boundaries, how does the local labour market keep pace with the constantly changing skill needs of ﬁrms in the cluster? Second, what institutional arrangements between the state, industry and education institutions have helped foster the innovation capabilities in the cluster? The aim is to understand the role of the local labour market in shaping the patterns of development of industrial clusters. This chapter is organized as follows. Drawing on the existing literature, the next section considers the role of the local labour market in sustaining the growth of knowledge-based clusters. The third section provides an overview of the evolution and recent development of the software cluster in Bangalore. The fourth section analyses the conditions and recent trends of the labour market for the IT industry in India, and Bangalore in particular. The ﬁfth section examines the roles of local labour market intermediaries, both formal and informal, in the working of the local labour markets. The sixth and ﬁnal section summarizes the ﬁndings and discusses some policy implications.

The local labour market as a determinant of innovation and competitiveness of IT clusters Within the relevant literature, several factors explain the successful growth of regional clusters. Given these factors, how does the local labour market affect the innovation and competitiveness of knowledge-based clusters? Various researchers have pointed to the important role of clustering and industrial districts in promoting regional economic growth. Some point out the critical role of inter-ﬁrm cooperation among horizontally networked ﬁrms in geographical proximity, particularly involving many SMEs, as a driver of regional growth (Piore and Sabel 1984; Pyke et al. 1990; Schmitz and Nadvi 1999). However, to date the empirical case studies of developing countries suggest that the experiences of clusters vary widely (see Schmitz and Nadvi 1999 for a review). Moreover, the mere growth in the number of ﬁrms located in a cluster does not necessarily ensure that the cluster will succeed (Bresnahan et al. 2001). Scholars have found several local factors that explain the successful sustained growth of innovative clusters. First, building on the conventional Marshallian agglomeration theory, many put forth the concept of ‘knowledge spillovers’, or localized external effects arising from the agglomera-

Aya Okada 247

tion of economic activities, as a key factor explaining the clustering of innovative ﬁrms (Krugman 1995; Breschi and Malerba 2001). This occurs because in external economies, the localization of skill, specialized materials and inputs, and technological know-how all help individual ﬁrms reduce their costs and increase the returns to the region as a whole (Saxenian and Hsu 2001; Schmitz and Nadvi 1999). Second, some studies have identiﬁed the so-called regional innovation systems (RIS), a region’s learning capability involving close interactions, networking and learning among various institutions such as ﬁrms, universities, research institutes, science parks and technology transfer centres, as a determinant of regional capabilities to innovate (Cooke 2001; Breschi and Malerba 2001). Building on the literature on embeddedness and networks (Granovetter 1985; Pyke et al. 1990; Powell 1991), these studies stress the importance of local ﬁrms being embedded in a dense network of knowledge sharing, particularly supported by close social interactions, trust, and effective sanctions (Schmitz and Nadvi 1999; Breschi and Malerba 2001). Thus the argument goes, the degree to which individual ﬁrms can utilize localized knowledge and capabilities depends on their ability to establish and accumulate social capital, by maintaining effective social links and connections (Breschi and Malerba 2001). Third, some scholars consider the availability of knowledge inputs, that is, a large pool of specialized and skilled labour, particularly engineers, scientists, entrepreneurs and venture capitalists, as one important condition for the growth of knowledge-intensive clusters (Breschi and Malerba 2001). To meet this condition, universities, large incumbent ﬁrms and foreigneducated people play a crucial role in supplying skills. In addition to availability, the localized mobility of skilled workers, either among ﬁrms or between research institutions including universities and new ﬁrms, may be critically important (Breschi and Malerba 2001: 821; Bresnahan et al. 2001). Fourth, on a related point, some studies suggest the importance of external links that clusters, particularly new ones, may forge, as they may gain access to knowledge, skills, contacts, capital and information about new technological opportunities and new markets (Breschi and Malerba 2001). In this regard, the mobility of a labour force across national boundaries, particularly those returning scientists, engineers and managers educated elsewhere, may play a crucial role in establishing these external links (Breschi and Malerba 2001: 821; Saxenian and Hsu 2001). For example, for their supply of skilled labour, the clusters in Taiwan draw heavily on returning US-educated Chinese engineers who have developed social and economic linkages with Silicon Valley (Bresnahan et al. 2001: 847; Saxenian and Hsu 2001). These authors argue that the emergence of international technical communities among foreign-educated engineers and scientists has facilitated the organization of production and innovation at the local and global levels.

248 Bangalore’s Software Cluster

Fifth, some studies point to the important role of foreign direct investment (FDI) in transferring knowledge and skills between clusters across national boundaries, as they increasingly shift their operations to more geographically distant locations. FDI-related ﬁrms provide managerial and technical training for their employees as well as their suppliers, an important source of knowledge and skill spillovers, playing a key role in fostering the skills upgrading of local suppliers (Okada 2000, 2004; Breschi and Malerba 2001). Finally, some researchers make an interesting conceptual distinction between the forces that account for the formation of a new cluster and those that account for the continued growth of existing clusters (Schmitz and Nadvi 1999; Bresnahan et al. 2001). Bresnahan et al. (2001) found that the determinants of the clusters’ success differ depending on the stage of their development. At the startup stage, ‘old economy’ factors like ﬁrm-building capabilities, managerial skills, and a large supply of skilled labour and connections to markets are crucial, whereas at later stages, agglomeration economies, external effects and spillovers are important to keep them going successfully. Taken together, these factors all point to the important role of the local labour market, as an institutional mechanism to create a skilled labour force, facilitate inter-ﬁrm mobility of labour, and encourage learning through close interactions and networking among ﬁrms and other local institutions. In this regard, the state might play an important role in coordinating the supply and demand at an optimal level to ﬁt with the expected trajectory of economic development, as happened in Singapore and Korea (Amsden 1989; Ashton et al. 1999: 4). Moreover, despite a growing body of literature on knowledge spillovers and spatial externalities, the concept of localized knowledge spillovers is still elusive and the processes that bring them about remain ambiguous, suggesting a need to investigate the speciﬁc mechanisms that link spatial proximity and knowledge diffusion (Breschi and Lissoni 2001). In particular, in the industries where innovation occurs so quickly, like IT, the knowledge they generate tends to quickly become outdated. Thus, ﬁrms must make institutional arrangements to rapidly upgrade their workers’ skills. Therefore, the smooth working of the local labour market, particularly for technical, managerial, and academic employees, seems particularly important (Breschi and Malerba 2001), to facilitate the mobility of skilled workers and thus the transfer of the knowledge they embody. The way a local labour market operates might determine the degree to which the deepening of the knowledge base could occur in the cluster. Despite a large body of empirical literature on industrial clusters, however, the nature and role of the local labour market in industrial clustering remains largely a black box, which is why this study focuses on this subject.

Aya Okada 249

The emergence of Bangalore as a leading software cluster in the global economy In the 1990s, the Indian ICT industry, speciﬁcally its software and services industry, witnessed remarkable growth. The evolution and recent development of Bangalore as India’s largest software cluster will provide some background for the discussions that follow. The rapid growth of the Indian software industry In the 1990s, India emerged as one of the fastest growing global centres for ICT. In particular, as the US market expanded, India’s software and services industry achieved phenomenal growth, over 50 per cent annually during the 1990s, increasing total revenue from US$175m. in 1989/90 to US$5.7bn in 1999/00 (ILO 2001). It rose further to US$10.1bn in 2001/02 (NASSCOM 2002). The IT industry’s share of GDP rose from 0.59 per cent in 1994/95 to 2.87 per cent in 2001/02. Moreover, it earned 76 per cent of its revenue (over US$7.7bn) from exports in 2001/02 (NASSCOM 2002). By 2000/01, India accounted for a signiﬁcant 18.5 per cent of the global market for customized software that is outsourced across borders, and yet only roughly 1.5 per cent of the world software industry, valued at US$500bn (ILO 2001: 170). This rapid growth during the 1990s partly reﬂects India’s policy interventions of the mid-1980s through the 1984 Computer Policy and the 1986 Computer Software Export, Development and Training Policy that promoted the domestic software industry. These policies recognized software as an industry, and provided software ﬁrms with access to the latest technologies and promoted exports.4 Moreover, a drastic transformation of the Indian economy since the early 1990s has greatly inﬂuenced this remarkable growth trend. India has undergone major economic reforms since 1991, when it introduced a New Economic Policy and New Industrial Policy that favoured industrial and trade liberalization.5 Like other emerging ICT centres such as Ireland, Israel and Taiwan, India took advantage of previously untapped market opportunities, easy cross-border transactions with the US market, and the signiﬁcant and growing US demand for ICT products and services during the 1990s (Bresnahan et al. 2001: 843). In the early 1990s, many Indian ﬁrms exported their services by ‘bodyshopping’: sending their software professionals to their clients to deliver low-skill-intensity services ‘on-site’. Thus, they took advantage of the relatively low salaries of Indian software professionals (Parthasarathy 2000; ILO 2001: 170). In recent years, however, Indian software ﬁrms have gradually acquired the technological and managerial capabilities to complete turnkey projects for large ﬁrms, and thus the proportion of on-site exports among India’s software exports has begun to decrease, from 90 per cent in 1988 to

250

Table 7.1

Average annual software labor cost comparisons in selected countries, 1999 (US$)

Project leader Business analyst Systems analyst Systems designer Development programmer Support programmer Network analyst/designer Quality assurance specialist Database data analyst Metrics/process analyst Documentation/training staff Test designer n.a.= Not available. Source: ILO (2001: table 4.6).

Switzerland

USA

Canada

UK

Ireland

Greece

South Africa

India

90,000 89,947 89,947 81,400 68,000 68,000 81,400 86,300 81,400 89,947 71,700 71,700

65,600 46,000 58,300 66,900 49,800 45,000 59,600 60,800 60,800 58,300 43,800 57,100

47,400 43,800 39,000 43,800 35,300 31,600 39,000 34,000 39,000 35,300 31,600 30,400

47,400 45,000 41,300 41,300 35,300 30,400 38,000 40,100 26,700 37,700 25,600 29,200

52,300 43,800 43,800 37,700 25,600 25,600 31,600 35,300 35,300 n.a. n.a. n.a.

35,100 41,000 22,000 22,000 19,000 22,000 22,000 22,000 35,100 22,000 22,000 19,000

34,000 37,000 36,400 25,100 31,700 18,700 26,600 n.a. 26,400 n.a. n.a. n.a.

33,700 31,000 20,500 16,100 11,700 11,700 20,500 20,500 24,900 24,900 11,700 11,700

Aya Okada 251

49 per cent in 1999/00 (ILO 2001: 170). Indeed, offshore delivery has grown much faster than on-site body-shopping, resulting in an almost even distribution by 2001/02 of both types of service in exports: US$3.8bn through offshore and US$3.6bn through on-site services (NASSCOM 2002). As Table 7.1 shows, India’s rapidly growing software trade clearly takes advantage of its comparative advantage – an abundant supply of the lowcost but skilled labour with proﬁciency in English – to tap this global market in IT services. The table shows the differences in labour costs that range from one-fourth to a half between the USA and India, indicating the cost competitiveness of the latter, even though Indian software exports still account for a very small share of the world software trade. The wage differentials between India and developed countries are even greater in lower value-chain services such as remote processing: an Indian medical transcriber’s average annual wage of US$1200 is only 5 per cent of the US$25,000 per annum for an American counterpart (ILO 2001: 136–7). Indeed, more than 185 of the Fortune 500 companies have outsourced their software development and services to India (NASSCOM 2001). The great technological advances in ICT, following the diffusion of the internet on a global scale, have accelerated this trend. India’s software clusters have thus focused on the markets that complemented its main sources of Table 7.2

Indian software ﬁrms’ global market penetration

Area of services

Current market size, 2000 (US$bn)

Indian export revenues, 2000 (US$bn)

India’s market share, 2000 (%)

Custom application development

18

2.5

13.6

Application outsourcing

11

1.7

15.8

Packaged software support and installation

41

0.3

0.7

Systems integration

72

0.15

0.2

IT consulting

19

0.05

0.3

Network infrastructure management

20

0.05

0.3

IS outsourcing

56

0

0

IT training and education

22

0

0

Hardware support and installation

43

0

0

Network consulting and integration

18

0

0

4.75

1.5

Total Source: NASSCOM (2002).

320

252 Bangalore’s Software Cluster

demand (e.g., Silicon Valley) rather than competing with other emerging ICT clusters (Bresnahan et al. 2001: 851). Table 7.2 shows the market niche for India’s software and services sector in the global market. The Indian software industry still suffers from low productivity and remains largely at the lower end of the value chain.6 As the table shows, India’s software exports are concentrated in lower-end services such as low-level custom application development and application outsourcing, involving simple coding, testing and maintenance services, and few well-known products in the global market (Government of India 2001). Indeed, in 2000 India had only a 0.2 per cent share of the world software products segment (NASSCOM 2002). In 2001/02, Indian exports of software services grew by 29 per cent over the previous year, while its IT-enabled services sector grew by 67 per cent (NASSCOM 2002). In the software and IT-enabled services sectors, as developed-country ﬁrms move up the value chain, they increasingly outsource lower value-chain activities, such as remote processing7 and back-ofﬁce functions, to lower-cost ﬁrms in developing countries such as India. As global ﬁrms in such service sectors as airlines, insurance, banking, hospitals and telecommunications increasingly outsource their work, they take advantage of India’s abundant low-cost, semi-skilled labour and time differences. India has become a global hub for a wide range of IT-enabled longdistance customer services, involving at least 310 ﬁrms, according to a recent NASSCOM survey (NASSCOM 2001). India holds a dominant position in call centre and animation services, and new services such as data search and integration services and e-learning are increasingly gaining importance (Government of India 2001). India’s presence in higher-end consulting services such as systems integration and network infrastructure integration is still very small, partly because these are difﬁcult to carry out from an offshore base. But, the industry has recently increased its exports of these high-end consulting services; the increase in R&D spending by Indian software ﬁrms reached around 3.4 per cent of the total revenues in 1999/00 (ILO 2001: 171). Also, leading multinational corporations (MNCs) have recently started outsourcing their R&D to India, most notably to Bangalore. Global players such as IBM, Intel, Sun Microsystems and Texas Instruments (TI) have all greatly expanded their operations in India, many moving their R&D facilities. TI, for example, develops many chips for the world market. Similarly, taking advantage of low labour costs, Sony reduced its costs by between 40 and 50 per cent by moving its internal software services function for its own products to Bangalore from Singapore.8 Such changes have helped some MNCs experience up to 120 per cent annual growth in recent years. Their presence has also made Indian software ﬁrms realize that they must improve the quality of their IT manpower, to raise productivity and move up the value chain, even without a global brand. At the same time, the global

Aya Okada 253

slowdown in the IT industry, beginning in 2002, has led Indian ﬁrms to realize the need to diversify their export markets from the USA to other large markets such as Europe and Japan. The evolution and recent developments of the software cluster in Bangalore Bangalore is India’s ﬁfth-largest city, and the oldest and largest software cluster in India; since the mid-1990s it has emerged as a major global software centre. As of 2003, 1154 software ﬁrms are located in this cluster.9 As Table 7.3 shows, in 2001, of India’s 700 top ﬁrms in the IT software and services sector, 160 (22.8 per cent) were located in Bangalore. Of these, more than 100 ﬁrms were MNCs, many based in the USA and Europe.10 They include major global players such as IBM, Hewlett Packard (HP), Sun Microsystems, Motorola, Ericsson, Oracle and Siemens. Bangalore has also witnessed rapid growth in the number of startup ﬁrms, particularly SMEs. Top 10 ﬁrms account for 50 per cent of exports from Bangalore, while SMEs account for 20 per cent. Of Bangalore’s ﬁrms, 35 per cent have close ties with those in foreign locations through non-resident Indians (NRIs).11 Most large Indian ﬁrms, except Tata Consultancy Services (TCS) and Wipro, grew out of small ﬁrms, without the involvement of large conglomerates such as Tata. On average, these domestic ﬁrms have been in operation for 10 to 12 years.12 Several factors explain the formation of the software cluster in the early 1980s. First, a large and highly skilled technical workforce was traditionally Table 7.3 Locational distribution of headquarters of India’s top 700 IT software and services ﬁrms, 2000 IT cluster Bangalore Mumbai Chennai Delhi Hyderabad Pune Calcutta Gurgaon Noida Thiruvananthapuram Ahemedabad Bhubaneshwar Chandigarh Others Total Source: NASSCOM (2001).

No. of top ﬁrms 160 148 72 62 61 48 32 23 21 14 10 8 7 34 700

254 Bangalore’s Software Cluster

available, because the city has enjoyed industrial development through manufacturing in aerospace, defence electronics and telecommunications since the mid-1950s (Parthasarathy 2000). Second, many higher-education and R&D institutions are concentrated there. In 1909, the Indian Institute of Science (IISc), a top national institution of science, was established in Bangalore. Leading public sector ﬁrms, such as Bharat Electricals, Indian Telephone Industries (ITI), Hindustan Aeronautics and the Indian Space Research Organization have been located in Bangalore, some since the 1950s. The presence of these research-minded public-sector enterprises attracted MNCs such as TI and Motorola in the mid 1980s. Indeed, the formation of the software cluster started when the government invited TI to set up its operation in Bangalore in 1984 as the ﬁrst MNC to operate in the Indian software industry (Sharma 2001). These MNCs’ success in turn encouraged many others to start operations in Bangalore. Third, it also enjoys a relatively comfortable climate all year round, and a cosmopolitan character, partly as a legacy of the British Indian Army station during the colonial period. Today, only 30 per cent of its residents speak the local language, Kannada, while English remains dominant. Fourth, government policies promoted the software industry and its exports in the 1990s. Since 1991, the central government has established the Software Technology Parks of India (STPI) in 35 cities, including Bangalore, Mysore, Manipal and Hubli in Karnataka. STPIs have facilitated a gradual shift from on-site body-shopping to offshore service delivery during the 1990s, supporting new startup ﬁrms in terms of quality infrastructure, telecommunications, a tax holiday for the initial ﬁve years, and dissemination of new technologies (Parthasarathy 2000; Saxenian 2001). In return, these ﬁrms were mandated to export. Together, these factors explain the growth of the software cluster in Bangalore in formative years. They do not sufﬁciently explain, however, its successful continued growth in more recent years, which thus merits a further analysis. Figure 7.1 shows the rapid export growth of Bangalore’s cluster in the software and services industry between 1991/92 and 2000/01. In particular, Bangalore’s IT cluster witnessed a tremendous boom between 1997 and 2001, partly because the Y2K problem led to an urgent demand boost. The number of ﬁrms in the software and services industry in Bangalore increased from 13 in 1991/92 to 728 in 1999/00, and further to 928 in 2000/01. The value of exports increased from US$1.3m. to US$1.6bn during this period. US customers account for 60 per cent to 65 per cent of business in the software and services industry in Bangalore. Leading ﬁrms such as Wipro experienced 30 per cent growth a year during this period. Clearly, as in the case of Hsinchu in Taiwan (Saxenian and Hsu 2001), this close connection with the US market has been a critical factor for the cluster’s growth. The question remains, however: What explains this phenomenal export growth?

Aya Okada 255 Growth of software exports from Bangalore

1000

1800

900

1600

800 700

1400

600 500 400 300 200 100

1000 800 600 400

No. of firms Exports

200 0

19 91 –9 2 19 92 –9 3 19 93 –9 4 19 94 –9 5 19 95 – 19 96 96 –9 7 19 97 –9 8 19 98 –9 9 19 99 –0 0 20 00 –0 1

0

1200

Exports (US$m)

No. of firms

Figure 7.1

Source: STPI (2001).

As Figure 7.1 shows, the agglomeration of ﬁrms as well as the export growth of Bangalore’s software cluster accelerated in 1997, when the state government of Karnataka announced its own IT policy. In 1998, the state government also established its own Department of IT, one year before the central government set up a new Ministry of IT. Indeed, the state government also actively promoted IT industries within the state, and particularly in Bangalore. Such policies include (1) ﬁnancial assistance to IT ﬁrms provided by the State Financial Corporations for technological upgrading; (2) the establishment of the Infrastructure Development Fund; (3) a duty concession to MNCs in the IT sector; (4) entry tax exemption; (5) investment subsidies for new investment; and (6) the 2001 enactment of Karnataka Industries Promotion Bill, which deregulates procedures to speed up all the clearances required to start a new investment project.13 Moreover, in 1998, the central government set up the National IT Task Force which formulated the IT Action Plan a year later.14 Also, at the national level, in 1999, STPIs stepped up their efforts to promote spillovers, by providing a total of US$1m. in seed funds to new startups to become selfsustaining in three years. Further, the state government established the International Technology Park (ITP) in Bangalore, through a partnership with Tata Industries and a consortium of Singapore ﬁrms (Saxenian 2001), where it invited many IT ﬁrms and IIIT to locate. Clearly, such aggressive policy measures for promoting the IT industry in the country and particularly in Karnataka played an instrumental role in inducing the rapid growth of Bangalore’s software cluster. In addition, industrial associations such as the National Association of Software and Service Companies (NASSCOM) have played an important role in linking the industry with computer hardware and manufacturing

256 Bangalore’s Software Cluster

industries, and lobbying the government. For example, with 870 member ﬁrms in 1997, NASSCOM lobbied the government to eliminate all import duties on software and allow software ﬁrms to invest in foreign joint ventures (Saxenian 2001). NASSCOM promotes partnerships between its members and foreign ﬁrms with needed skill sets, among other services. In its early stage, many ﬁrms in Bangalore’s software cluster were engaged in on-site ‘body-shopping’ for consulting ﬁrms. In recent years, it has, however, tried to move up the value chain by attracting global IT players to engage in higher-value software activities. In 2001, in the global software industry, only 40 ﬁrms have achieved SEI-CMM Level 5;15 of these 29 are located in India and 18 in Bangalore (NASSCOM 2001), a clear indication of Bangalore’s rising ability to compete in the global market. Recently, MNCs, such as Ericsson, Lucent and Cisco have rapidly increased their R&D outsourcing. Intel plans to invest US$25m. to set up a new technology development centre in Bangalore in 2003 and expand engineering staff by 50 per cent (NASSCOM 2002). Indian large ﬁrms develop their technological capabilities by collaborating with MNCs in jointly setting up R&D facilities. For example, Sun Microsystems and IBM have created R&D centres in several locations in India. Accordingly, Bangalore has recently witnessed a shift in interest among large ﬁrms in high-end services such as embedded systems, design and manufacturing of chips, and embedded communications. By 2002, however, this demand slowed down, along with the global IT markets. In Bangalore, IT-enabled services have taken off, and also moved to bio-informatics. The state government plans to create a biotechnology park in Bangalore. The cluster is somewhat segmented, because of the clear division of labour among MNCs, leading Indian software ﬁrms, and Indian SMEs. Many SMEs are new startups established after the mid-1990s. While computing and communications technologies, particularly the internet, have widened SMEs’ business opportunities, in some cases even in world markets, most SMEs undertake low-end software services, such as data processing and conversions, and routine application activities, mainly using semi-skilled labour. Or they cater to small local market niches. For example, one small ﬁrm I visited primarily sold a single proprietary software product it had developed to cater to only local SME customers.16 Thus, a hierarchical structure may have emerged within the software and services cluster in Bangalore. Indeed, as illustrated in Figure 7.2, the software and services industry in Bangalore can be divided largely into four tiers, deﬁned by the lines of services, which are in turn deﬁned by the ﬁrms’ skill levels: ﬁrms with similar skill levels carry out similar lines of services. This skill-based segmented structure of the software cluster, and its high reliance on global export markets, together suggest that local ﬁrms within the cluster are not competing so much against each other in captur-

Aya Okada 257 Figure 7.2

Structures of the software and services cluster in Bangalore Highly skilled

R&D/computer hardware engineering Chip design/ embedded software Software product development Network solutions and management Systems development and consulting Skills intensity

Software services: Programming/application development Database development/administration Technical support/customer services IT-enabled services: Call centres; Finance and accounts Data processing: Medical, Insurance, Banking

Semi-skilled

ing export markets. Rather, competition among ﬁrms may exist in the local labour market, in terms of recruiting, retaining and upgrading the right sets of skills necessary to attract foreign client ﬁrms that are interested in outsourcing their operations to Bangalore. The next section will turn to this aspect.

The changing nature of local labour markets in Bangalore’s ICT cluster This section analyses the changing patterns of the local labour market in the IT sector in India, and in Bangalore in particular. Growing demand for IT workers As India’s IT industry has grown, and as computer hardware and software have been developed and diffused both in India and globally, demand for Indian IT professionals has grown fast, both in software development and in IT-related services in other industries. As Table 7.4 shows, both employment opportunities for IT professionals and revenues per employee rose signiﬁcantly in recent years. The number of Indian IT professionals17 increased rapidly from 6800 in 1985/86 to 140,000 in 1995/96, and further to 410,000 by 2000 (see Table 7.4). During the boom years between 1997/98 and 2000/01, it more than doubled. Of 410,000 IT professionals, NASSCOM (2001) estimates that about 140,000 work in the IT software and services export industry. Of the software professionals, 80 per cent have engineering degrees and 12 per cent have diplomas or certiﬁcates from

258 Bangalore’s Software Cluster Table 7.4 Growth of employment and revenue per employee in the Indian software industry

1993/94 1994/95 1995/96 1996/97 1997/98 1998/99 1999/00 2000/01

Employment (persons)

Revenue/employee (US$)

90,000 118,000 140,000 160,000 180,000 280,000 340,000 410,000

6,199 6,998 8,924 11,036 15,000 21,786 35,588 32,927

The employment ﬁgures include professionals engaged in software, IT services, and IT-enabled services including professionals engaged in software development units in other organizations. Sources: ILO (2001); NASSCOM (2001); NASSCOM internal document (2002).

private training institutions (Government of India 2001). The industry projects a further increase in demand for IT professionals in the next several years, estimating that India will require a minimum of 2.2 million knowledge workers in software and related services by 2008 (NASSCOM 2001). Demand for IT professionals is unevenly distributed among regions. In 2000, the most job opportunities for IT professionals were in South India (including Karnataka, Tamil Nadu, and Kerala states), accounting for 43 per cent of all new IT-related recruitment in the country, and the fewest in the Eastern region (including Orissa and West Bengal states), accounting for only 5 per cent (NASSCOM 2001). This reﬂects a rapid growth of IT clusters in South India, such as Bangalore (in the state of Karnataka), Hyderabad (in Andra Pradesh), and Chennai (in Tamil Nadu). The Western region, which includes clusters like Mumbai, Pune and Ahmedabad, accounts for 29 per cent of newly placed jobs. In addition to a growth in domestic demand, the recent serious shortage of IT workers in many developed countries has contributed to a signiﬁcant rise in global demand for Indian skilled IT workers,18 especially as the industry became more engaged in subcontracting to ﬁrms in other countries. The USA has issued more H1-B Visas and countries like Germany, Italy, Japan, Korea, Norway Singapore, and the UK have launched various initiatives concerning immigration to attract Indian IT professionals, which should further boost this demand (NASSCOM 2001). India now faces the challenge of increasing its supply of IT engineering graduates to keep up with the rapid growth in both domestic and global demand, and still retain some of its best talent at home. Reﬂecting the hierarchical and fragmented structure of Bangalore’s software cluster as discussed earlier, the local labour market for IT professionals is fairly segmented (see Figure 7.2). Table 7.5 shows the results of a

Aya Okada 259 Table 7.5

IT-related jobs in demand in India

Types of jobs

Distribution (%)

Programmers/Engineers/Analysts/ Computer scientists

41%

Internet and e-commerce applications Web developers/design

19%

Network specialists/Developers/ Communication engineers

14%

Database administrators/developers

11%

Digital media and technical writing IT-enabled services

5% 10%

Source: NASSCOM (2001).

NASSCOM survey in 2000: among all the IT-related jobs in India, the largest demand (41 per cent of the total) exists in jobs related to software development, such as programmers, software engineers, analysts and computer scientists. With the development of the internet and the growth of e-commerce, professionals with internet-related skills, including network specialists, end-user support staff and internet/intranet developers, have also been in high demand (NASSCOM 2001). Bangalore has recently seen an increase in the number of ﬁrms operating in high-end software development and services such as embedded systems; design and manufacturing of chips; and embedded telecommunication software, as they attempt to move up the value chain. For example, Motorola has moved 90 per cent of R&D activities for mobile phones to Bangalore. Likewise, HP’s R&D, as part of its International Software Operation (ISO) Department in Bangalore, develops embedded software, particularly for its own hardware products, entailing knowledge generation, and requiring indepth knowledge of networking and technologies.19 This has led to higher demand for system software engineers and an urgent need for domain architect skills to develop the highly integrated software required in the consulting business. For example, Sony India focuses on domain knowledge development for its own products, to support its plants in various locations in Asia.20 These ﬁrms require highly specialized technical skills with logical reasoning and design capabilities. Thus, some system software ﬁrms recruit 5 to 10 top graduates from the best institutions. Some MNCs, such as Motorola, TI and HP, constantly look for people with these skills. IT professionals with these skills can receive up to a 35 per cent skills premium, depending on the type of business.21 Project leaders in these high-end services, highly specialized and skilled, do not move to other areas of specialization. On the other hand, IT professionals with more general and thus versatile skills such as C++ and Java move from one project ﬁeld to another.

260 Bangalore’s Software Cluster

MNCs such as HP India, however, are engaged not only in higher-end software development but also in IT-enabled services. For example, HP’s ISO Department, with 3000 IT professionals, puts equal emphasis on highly technical R&D, and software services and solutions. The latter are largely head-count operations for US, European and Japanese customers, involving mainly low-level programming skills.22 While leading Indian ﬁrms such as Infosys, Wipro and Satyam do design software products, they also do more than half of their business on these low-level, routine-type software services and solutions. For these services, few ﬁrms recruit the best and brightest engineers, as they need the basic skills of college graduates with engineering degrees. For software design and development, and software services such as systems solutions, which mainly comprise offshore contracts, ﬁrms employ graduates in computer science. In addition, the industry increasingly needs experts to lead and support IT projects in areas such as customer relationship management (CRM) and e-CRM implementation (including front-ofﬁce automation, and supply chain management), and to provide IT-enabled services such as ﬁnance, accounting and human resources, to ﬁrms in other industries. Thus, in addition to technical skills, they see managerial skills as urgently important, as the industry moves into offshore services on a project basis. For example, Infosys hires many of its graduates from the prestigious Indian Institutes of Management (IIMs) with MBAs. Most of the work in the Indian software industry has however, been relatively non-technical and requires mainly logical and methodical skills and a familiarity with software development tools and languages. Still, the industry prefers engineering graduates to those in other disciplines, because they see them having ‘learnability’, that is, the ability to learn and apply new technical skills.23 For example, workers learn the latest skills such as C++ on their own in the course of their work assignments. Especially, in the software industry, today’s skills set is likely to become obsolete and redundant quickly: the average life of technical skills is only three years!24 Thus, ﬁrms look for more basic skills, such as logical thinking and strong skills in mathematics.25 While college engineering education will equip students with the basic knowledge and skills they need to ﬁnd entry-level jobs in the IT industry, their curricula cannot keep up with the speed of innovation. This suggests the important role of in-ﬁrm skills upgrading, which I discuss further below. Moreover, reﬂecting the recent slowdown in the US market, Indian software ﬁrms started diversifying their markets, such as Europe and Japan. Thus, many Indian IT professionals are seeking language training: private computer training institutes, as well as many leading Indian ﬁrms in Bangalore such as Wipro, are offering their employees Japanese language courses, to expand their opportunities with Japan. On the other hand, the rapid growth of IT-enabled services since 2000 (see pp. 252–7) has increased job opportunities for the semi-skilled work-

Aya Okada 261

force in data entry and conversion, and call centre operations. These jobs, which require no high-quality software skills, attract graduates and diploma holders in many disciplines (NASSCOM 2001); in the USA these jobs attract high-school graduates and dropouts. IT ﬁrms in Bangalore recruit their employees through various methods: (1) on-campus interviews; (2) internal advertisement and employee referrals; (3) stafﬁng agencies; (4) online job portals; and (5) advertisements in national and local newspapers.26 Large ﬁrms such as Wipro typically use campus interviews. For example, various IT ﬁrms come to IISc for a whole month to carry out campus interviews for fresh recruits. During the boom years between 1997 and 2001, as the industry grew tremendously, it experienced a serious shortage of appropriate skills. Thus, ﬁrms needed to attract engineering graduates by offering large skills premiums, comparable to those on the world market. IT professionals’ salaries grew 25 to 40 per cent a year, a rate unmatched in any other Indian industry.27 Intense competition to get the best workers caused ‘compensation spirals’: each company competed to offer better compensation packages than anyone else. Interestingly, this problem led to the evolution of a device for sharing information among local ﬁrms: the IT industry in Bangalore now cooperates almost every year to conduct a survey on compensation and labour market trends. Leading ﬁrms can then adjust compensation levels of IT professionals to create parity; some ﬁrms used the information to offer 20 to 30 per cent higher salaries than other ﬁrms. MNCs pay on average 25 per cent higher salaries to IT professionals than Indian ﬁrms. In such a knowledge-intensive industry, the labour costs account for up to 80 per cent of ﬁrms’ operation costs, a clear indication of how much the ﬁrms’ competitiveness relies on human skills.28 In particular, graduates in computer science from the best academic institutions such as the Indian Institutes of Technology (IITs) and IISc are in high demand and receive up to 40 per cent higher salaries than their counterparts in other disciplines, which makes them least likely to stay with a ﬁrm for very long. Entry-level IIT graduates receive on average US$12,000 per year;29 after a year or two of experience, many move to the USA seeking better jobs. Despite such mobility, the job market at leading ﬁrms such as Wipro and Infosys has become very competitive. Out of more than 300,000 applicants, only 1.6 per cent were offered jobs in 2000; in 2001, only 0.8 per cent.30 MNCs and large leading ﬁrms like Wipro recruit only the best talent from the best institutions such as IITs, IISc and the Regional Engineering Colleges. Other ﬁrms recruited engineering graduates in any subdiscipline, given the acute shortage of graduates in IT-related courses, and paid higher skill premiums to graduates in IT-related courses. As a result, a new class of young ‘techno-elite’ professionals emerged in the local labour market, some even becoming millionaires with houses, cars and stock options from their employers. Firms like Wipro, requiring relatively low skills for their services,

262 Bangalore’s Software Cluster

hired engineering graduates in other subdisciplines such as mechanical and civil engineering. Hiring took place throughout the year, as vacancies are always appearing. Some ﬁrms even had ‘walk-in interviews’; good applicants could be hired on the same day.31 Taking advantage of this skills shortage, many engineers in other industries moved into the IT sector, and some set up their own ﬁrms. As venture capital emerged all over the country, ﬁnance was not a problem.32 Owing to this high labour mobility and growth in demand, leading IT ﬁrms in the Bangalore cluster, such as Infosys and Wipro, have faced serious problems in retaining the best talent among their employees. To reduce attrition, they provide, as a retention strategy, housing loans, stock options, and ﬂexible timing, all new practices for Indian ﬁrms.33 To attract the best and brightest among IT professionals in their mid 20s, they as well as MNCs focus on creating a world-class work environment. For example, Microsoft and HP offered high salaries and fringe beneﬁts and provided work environments at the same levels as in the USA, including a swimming pool! Another retention strategy, among leading ﬁrms like Infosys, is to deliberately recruit engineering graduates in non-IT related courses such as civil and mechanical engineering who are less likely to leave, and train them for a year. With its employee-friendly work environment and such generous compensation and training packages, Infosys keeps its attrition rate below 10 per cent.34 Another retention strategy is to prevent boredom. TCS and HP keep updating their employees with the latest skills and assigning them to challenging high-tech development projects.35 As the US IT market slowed down, so did the Indian software industry, beginning in 2001, which led to a sharp decline in demand for IT professionals. For example, the IT industry absorbed only 30 per cent to 40 per cent of those earning a master’s degree in computer application from Bangalore University, a leading university in the region.36 The starting salary for software engineers was halved in just two years from 2000 to 2002. A large and growing supply base of IT professionals India considers that its comparative advantage lies in its very large pool of low-cost and high-quality scientiﬁc and engineering manpower (Government of India 2001). Undoubtedly, the availability of a large pool of knowledge workers,37 particularly engineers with strong technical skills and ﬂuent English, was a critical factor for not only the formation but also the growth of Bangalore’s software cluster. As of 2001, about 80,000 IT professionals were working in Bangalore, accounting for about 20 per cent of the IT professionals in the whole country.38 In India, higher-educational institutions such as the well-known Indian Institutes of Technology (IITs), universities and engineering colleges are the

Aya Okada 263

main source of newly qualiﬁed IT workers. In 1998, to accelerate its effort to increase the IT workforce, the government established the Indian Institute of Information Technology (IIIT) in seven locations, including Bangalore, as a joint initiative between the government and industry. The IIITs offer not only degree courses but also short courses ranging from 6 weeks to 6 months for IT professionals and industry-sponsored trainees. The IIIT-Bangalore was set up right in the centre of International Technology Park (ITP) located in Bangalore, to improve the quality of training in IT-related courses and train IT professionals to meet industry needs. The institute coordinates closely with the state government’s IT Taskforce. The software industry in the region also provided ﬁnancial support to the IIIT. Its curriculum is uniquely designed to link directly with industry requirements. These IIITs are afﬁliated with schools of leading computer ﬁrms, such as IBM, Microsoft, Oracle and Satyam (NASSCOM 2001). Table 7.6 shows the numbers of formal educational institutions offering IT-related courses in India as a whole and in Karnataka. Among all the states in the country, Karnataka stands at third, after Maharashtra and Tamil Nadu, in terms of the total number of technical institutions in 2001/02 (NITK 2002). As of 2001, there were 106 engineering colleges in Karnataka, 2 government-run and the rest privately funded. Private engineering colleges in Karnataka account for 19 per cent of the 553 all private technical colleges nationwide (Government of India 2001). Since the privatization policy in higher education was introduced in 1993, the entry requirements have eased;39 many more engineering colleges were estabTable 7.6 Higher education institutions offering courses related to computers, all India and Karnataka Type of institutions

National Institutes (6 IITs, 1 IISc and 2 IIITs) Regional engineering colleges Degree colleges at universities including those offering BTech, BSc and BCA

All India (in number) 2000 9

Karnataka (in number) 1999/2000 2 (1 IISc and 1 IIIT)

43

2

860

106

Diploma colleges (polytechnics)

1220

186

Total

2130

297

IIT = Indian Institute of Technology; IISc = Indian Institute of Science; IIITs = Indian Institute of Information Technology; BTech = Bachelor of Technology; BSc = Bachelor of Science; BCA= Bachelor of Computer Application. Sources: NASSCOM, 2001 (for All India ﬁgures); Government of India (2001); The Department of Technical Education, State Government of Karnataka (for Karnataka ﬁgures).

264 Bangalore’s Software Cluster

lished in Karnataka, offering IT-related courses. This greatly increased the IT-skilled workforce for the software industry in the region. With their low capacity, poor facilities and substandard course quality, however, many of these privately funded technical colleges have created serious concerns about quality. In fact, owing to a lack of skilled faculty at local engineering institutes, many students attend classes at non-formal private IT institutes in order to acquire practical skills. Nonetheless, this rapid expansion of technical education institutions in Karnataka helped increase the supply of IT manpower in the 1990s, and particularly during the boom years. To meet a rapid increase in demand, the state government of Karnataka has recently increased the number and size of IT-related courses (computer science, information technology, electronics and communication, and telecommunication), particularly since the mid 1990s (Table 7.7). Today, most engineering colleges have IT-related courses, reﬂecting the increased demand for engineering education in IT-related ﬁelds. Because of the ‘compensation spirals’ discussed earlier, the number of applicants for courses in computer science rose rapidly. In 2000/01, more than 4000 additional spaces were sanctioned in engineering colleges, both public and private, in Karnataka, making an annual total intake of 11,565 students in IT-related courses in engineering colleges. This accounts for 16 per cent of all Indian students in this same category, although Karnataka’s population accounts for only 5 per cent of India’s total population. By 2000, more than onethird of all engineering students in Karnataka were enrolled in IT-related courses. Moreover, the government planned to start another 22 new ITrelated courses in 16 existing diploma colleges (Government of Karnataka 2001). In India as a whole, the output of trained manpower at the degree or diploma level has grown consistently, reaching almost 122,000 per year by 2000 (NASSCOM 2001). Of these, 63 per cent are qualiﬁed workers with degrees in IT-related ﬁelds. However, India’s output of IT professionals is still small compared with that in the USA, where each year 10,000 people earn a master’s degree and 800 a PhD degree in computer science (Government of India 2001). Reﬂecting this tight labour market in the rapidly growing industry, many graduates from other disciplines also ﬁnd jobs in IT-related services. Indeed, more than 90 per cent of non-IT graduates of IITs found jobs in the IT sector (Government of India 2001). In addition, in 2000, more than 70,000 private computer training institutes were estimated to be offering non-formal IT-related courses outside the formal education system in India, although only 5000 of these can be considered of high quality (NASSCOM 2001). Indeed, these training institutes have mushroomed since 1996, diluting the quality of such training. While the government does not recognize certiﬁcates issued by these nonformal training institutes, this type of training has become extremely

Table 7.7

Growth in intake in IT-related courses in Karnataka Diploma colleges

Engineering colleges (BTech.) No. of engineering colleges

Intake in IT-related courses

% in total intake

No. of diploma colleges

Intake in IT-related courses

% in total intake

1980/81 1981/82 1986/87 1991/92

41 42 50 51

0 25 1,240 2,020

0 0.3 8.4 12.1

44 46 150 165

0 0 n.a. 2,364

0 0 n.a. 10.3

1996/97 1997/98 1998/99 1999/00 2000/01

53 70 71 106 109

2,758 4,028 4,190 5,802 11,565

14.1 16.9 17.0 22.0 34.5

178 196 184 186 201

4,379 5,682 5,585 6,120 n.a.

15.2 17.8 17.1 18.0 n.a.

n.a.= Not available. Source: Compiled and calculated from internal documents of the Department of Technical Education, Government of Karnataka, various years.

265

266 Bangalore’s Software Cluster

popular, as trainees who learn to use application software and web-based technologies can take on low-end data entry jobs and thus become absorbed in the IT-enabled services industry (NASSCOM 2001). The software industry relies on a relatively young labour force: the overall median age of the Indian IT professionals was 25.3 years in 2000, with 83.8 per cent of all IT professionals under 35 (NASSCOM 2001).40 IT professionals who are 45 and over account for only 2.8 per cent of the total.41 Of the software professionals, 37 per cent had 4 to 6 years of working experience, while 23 per cent had only 1 to 3 years (NASSCOM 2001). This stands in marked contrast with old-economy industries such as the automotive, where employees are on average in their mid 30s and have longer work experience (Okada 2000). Clearly, the software industry exhibits a different pattern of employment practices from the old-economy manufacturing sector, with more frequent labour turnover and mobility among workers in search of higher salaries and better career advancement opportunities. Both industry and government are aware that they must maintain and enhance India’s comparative advantage in terms of skilled technical IT manpower, by expanding the base of IT skilled manpower and upgrading it with the right skill mix; this will allow individual ﬁrms to meet their skill requirements, and keep the Indian software industry competitive in global markets (NASSCOM 2001). In 2000, the central government set up an HRD Taskforce, to analyse the possible gap between the demand for and supply of the IT workforce in India (Government of India 2001). To accelerate the diffusion of IT skills in Karnataka, the state government also planned to install PCs and start computer-related courses in 2800 government-run secondary schools throughout the state. Leading local academic institutions also keep updating their curricula to respond to the industry’s new skill requirements. For example, IISc’s Computer Science Department has added courses such as computer security, internet technology and biotechnologies in their curriculum in the last few years. Through its Quality Improvement Program, the IISc faculty in turn, in collaboration with some leading ﬁrms in Bangalore, offers shortterm courses in the latest technologies for teachers at engineering colleges in the state so they can keep pace with technological change in the IT industry.42 The large and continued growth in the supply of IT manpower within the state allowed the industry to make even very specialized skills available within the cluster. For example, when a local software ﬁrm specializing in digital signal processing (DSP) advertised for 20 VLSI (very large-scale integration) engineers, most of the 10,000 applicants were from Bangalore. As the CEO of this ﬁrm put it, ‘Nowhere else, with the possible exception of Mumbai, would we have got such a response for a niche specialization’ (Business Today, 23 December, 2001: 64).

Aya Okada 267

As seen above, not only was the availability of a large pool of IT manpower a critical precondition for the formation of the software cluster in Bangalore, but also a rapid increase in the supply of technical skilled manpower in response to the fast growing demand has also contributed to the growth of the cluster. The state government played an important role in supplying the IT manpower required to keep pace with the rising demand and providing various forms of support to the industry. Labour mobility within and between ﬁrms In Bangalore’s software cluster, labour mobility is quite high both internally and across national borders, particularly from and to major IT markets such as the USA, Europe, and Japan. As is widely known, Bangalore has enjoyed very close connections with Silicon Valley, entailing two-way ﬂows of capital, technology and skills. However, unlike the case of Hsinchu (Saxenian and Hsu 2001), a large technical community of US-educated Indian IT engineers with work experience in the USA has not formed in Bangalore, as until recently few such professionals had returned to India (Saxenian 2001). Rather, social and economic links with the USA, through non-resident Indians (NRI), have played an important role in the expansion and growth of startups. Many skilled Indian IT professionals move between nations but within the ﬁrm, through arrangements with partner ﬁrms located in foreign countries, particularly the USA. For example, one ﬁrm I visited has its head ofﬁce in the USA headed by an NRI with 200 Indian workers, while its Indian ofﬁce, established in 1997, served as a back ofﬁce with 150 engineers. The ﬁrm grew mainly by expanding its on-site body-shopping in the USA. At this ﬁrm, Indian IT engineers are typically on assignment for six months in the US ofﬁce on a project basis, where they work under the close supervision of American and Indian managers.43 Most MNCs, such as IBM, HP, Motorola and Lucent, send out their Indian engineers to their locations abroad, sometimes cutting their labour costs by over 50 per cent. Leading Indian ﬁrms such as Infosys and TCS also send their IT professionals abroad for a few weeks to a few months, mainly to provide on-site services. TCS assigns up to 40 per cent of its engineers to its 100 ofﬁces in 50 locations outside India, particularly in the USA and the UK. IT professionals receive higher salaries through on-site assignments abroad, with allowances, which serve as incentives.44 With the increase in offshore services for external customers, however, many Indian workers increasingly remain in India. For example, HP India sends only 5 per cent of Indian engineers to its US ofﬁces, because its human resources deployment is ‘very India-centered’.45 The most popular destination for Indian IT professionals is the USA. Other destinations are less popular among them, partly because of the difﬁculties they face with languages and food, and partly because of high

268 Bangalore’s Software Cluster

living costs in countries such as Japan and Germany, which makes it difﬁcult for them to save their salaries, and thus reduce incentives.46 Mobility is also high within the local labour market. The average IT professional works only 2 to 5 years at one ﬁrm. During the boom years between 1997 and 2001, the average labour turnover rate rose to between 22 and 25 per cent, sometimes even up to 30 per cent, as engineers moved frequently to other ﬁrms offering higher salaries.47 During this period, IT professionals stayed an average of 6 to 8 months at one ﬁrm.48 Even at ﬁrms which took serious retention strategies, such as Infosys, the average length of stay was 2 to 3 years.49 After the IT industry began to slow down in 2001, however, the industry average for labour turnover declined to 8 per cent.50 Skills development Though the supply of IT professionals has grown rapidly through both formal and non-formal education in IT-related courses, clearly the knowledge imparted through such academic training is not adequate for all work in the software industry. Among software ﬁrms in Bangalore, in-ﬁrm training is a principal means to develop IT professionals’ skills. Because technology changes so quickly, ﬁrms need to keep upgrading their skill sets, by retraining their employees in the latest technologies and skills. Indeed, ‘hot skills’ keep changing: the average life of technical skills is only 3 years, as discussed earlier. Thus, leading ﬁrms such as Wipro and TCS closely manage and frequently update the skills proﬁle of their employees through in-ﬁrm training, in technical skills, domain knowledge and communication skills. Wipro allocates 5 per cent of total labour costs to in-ﬁrm training to upskill and reskill its employees, and on average employees spend 10 to 12 per cent of their work time on such training.51 Because the required skill sets differ depending on the project, professionals are trained in speciﬁc skills on a project basis.52 Such training lasts typically for a few days to one week, depending on the nature of skills. Thus, work in software services involves employees’ continuous learning, while also acquiring both technical and managerial skills. In particular, project managers must learn skills to handle customers, manage teams and projects, and formulate business strategies. They learn these skills not only through academic training but more importantly through experience in actual assignments to projects. Such in-ﬁrm training in turn leads ﬁrms to develop their organizational capabilities (Tschang 2001). Unlike those in the old-economy manufacturing sectors such as automotive (Okada 2000), workers in the software and services industry are in a fast career trajectory. In only 4 to 5 years, software engineers reach the top level, becoming project leaders. Software design engineers must have at least 5 years of experience; while software programmers must have at least more than 3 years of experience and junior engineers doing coding need

Aya Okada 269

only one year. A typical IT professional goes through the following career path in 3 to 5 years: (1) trainee (US$168 to US$211 per month); (2) software engineer/system engineer (US$253 per month); (3) senior software engineer/senior system engineer (US$421 per month); (4) team leader/ system analyst (US$632); (5) senior team leader/senior systems analyst (US$737 to US$842 per month); (6) project manager (US$842 per month); (7) head of division. Partly because of this fast-track career development and partly because of high mobility, most leading ﬁrms have performancedriven and merit-based compensation schemes. For example, TCS has a strict appraisal system that reviews each employee’s performance in every quarter.53 Another kind of skills development became popular recently, as the slowdown in the US IT market made India turn to other large IT markets such as Japan, Germany, France and Korea, where Indian IT professionals cannot use their traditional advantage – English proﬁciency. In response, many training institutes and leading Indian ﬁrms such as Wipro started offering courses in the languages of these countries, as well as cultural skills for IT professionals. Spin-offs Spin-offs are quite common among IT ﬁrms in Bangalore. Many Indian workers in large MNCs such as HP and Motorola and in leading Indian ﬁrms leave to set up their own ﬁrms. For example, former Wipro employees have set up more than a hundred ﬁrms with no assistance from the ﬁrm.54 These IT professionals use the knowledge and skills they acquired at large ﬁrms in their new startup ﬁrms. Relatively low capital requirements for such new startups, along with the growth of venture capital across the country, have facilitated this trend. Taking advantage of the rapidly increasing demand in the software industry in the 1990s, many of these startups in Bangalore grew out of small ﬁrms, leading to a large growth of SMEs, including the self-employed in the informal sector. For example, Network Solutions started with 3 employees in 1993: 10 years later, it had 815 employees, and was ranked as the fourth among ﬁrms specializing in network integration in Bangalore. In 2003, of its 815 employees, 650 were engineers located in 7 cities including Bangalore, with 440 engaged in on-site services in 22 cities. At ﬁrst, it was engaged mainly in networking structured cabling (e.g., LAN), data conversion and storage solutions, and then it has gradually diversiﬁed its areas of services to include voice-over IP and network security.55 Spin-offs occur not only from MNCs and leading domestic ﬁrms but also from educational and R&D institutions. For example, IISc has recently liberalized its policy, allowing its faculty members to set up their own ﬁrms. In the mid-1990s, the IISc’s faculty established the Society for Innovation and Development (SID), as an autonomous body, to facilitate consulting

270 Bangalore’s Software Cluster

and commercialize innovation generated from the institute’s laboratories. A well-known example of these innovations is the ‘Simputer’, a simple lowcost mobile computer designed to make the beneﬁts of computers accessible to the uneducated masses. The faculty carries out joint research projects with leading MNCs such as Microsoft, Intel and Nokia, by setting up a lab within the institute.56 As seen above, the local labour market has been dynamic and vibrant, in terms of mediating the demand and supply of IT professionals required in Bangalore’s software cluster. The working of the local labour market was not necessarily automatic, however, as I describe in the next section.

Institutional arrangements for developing and upgrading knowledge and skills in Bangalore’s software cluster Various institutional mechanisms created within Bangalore’s software cluster have inﬂuenced the way in which the local labour market works and facilitated its dynamics. Two such institutional arrangements involve the regional government and labour market intermediaries. University–state–industry linkages in mediating demand and supply of skilled labour Karnataka’s state government closely monitors the capacity and performance of engineering colleges in the state in terms of both intake and output. Each year, it carefully carries out a manpower planning exercise to achieve an optimal ﬁt between the demand for and supply of IT workers in the state. Since 1997, the Manpower Division of the Planning Department, the state’s Department of Higher Education, has worked out annually an elaborate formula based not only on a forecast of the manpower requirements in the region but also on an estimate of the global demand for IT professionals. It forecasts the demand for IT engineers by periodically conducting a sample survey and then adjusts the number of places in IT-related courses offered in engineering colleges in the state.57 This manpower planning helped educational institutions to respond quickly to the enormous growth in demand in Bangalore’s software cluster in the late 1990s, as discussed earlier. Moreover, since 1993, the state government has instituted an inter-departmental taskforce to assess entrance tests in IT-related courses, involving the directorates of information technology, industry and commerce, and technical education. It then introduced a common entrance test (CET), which has helped standardize the process for screening and enrolling students. As observed elsewhere, such proactive initiatives by the state government to closely link its training policies with the promotion of the IT industry manifest a ‘remarkable turnaround in state capacity and learning, in contrast to the over-bureaucratized, undynamic characteristics of the Indian state’

Aya Okada 271

(Dolan and Tewari 2001: 100). Thus, not only had a large pool of skilled labour been available prior to the formation of Bangalore’s software cluster, but the state government has also continued to expand the supply of skilled labour to meet the growing demand for a skilled technical workforce. Also, close interactions between the industry and universities in Karnataka facilitated the processes of skill matching and skill deployment. Leading ﬁrms in the cluster, such as IBM and Infosys, regularly send their key professionals as guest lecturers to engineering colleges, such as Bangalore University, IISc and IIIT.58 These lectures often lead to job interviews for students. Also, these ﬁrms offer scholarships to engineering students at IISc. The role of labour market intermediaries As the software and services industry boomed, private stafﬁng agencies also grew. Bangalore has more than 500 such agencies, doing 90 per cent of their business in the software industry.59 Their three main functions are head-hunting, placing advertisements and maintaining data banks. First, leading MNCs and large domestic IT ﬁrms in Bangalore frequently use local agencies for head-hunting to ﬁll middle and senior positions, or recruiting the best talent for their particular areas. After clearly deﬁning the required skill sets, responsibilities required for given jobs, and compensation levels, they interview and select potential candidates. These stafﬁng agencies receive average fees of 20 per cent of the annual salaries for top level professionals and 12.5 per cent to 16.5 per cent of the total labour cost for other IT professionals.60 For example, Sony India uses several agencies to recruit additional manpower during its peak times. Second, on behalf of their clients, agencies place recruiting advertisements with various media all over the country. Finally, they can maintain over 10,000 IT professionals in their databases and receive 200 resumés a day. This emerging mechanism for skills matching has facilitated the speedy deployment, placement, and mobility of IT professionals with relevant skills mixes within Bangalore’s software cluster. The periodical compensation surveys mentioned earlier also help to mediate the process of setting wages in the local labour market.

Conclusion Bangalore’s software cluster has emerged as the largest in India. This study has focused on explaining the cluster’s rapid and sustained growth, as opposed to its formation. Bangalore’s software and services industry achieved tremendous growth in the 1990s, particularly since 1997, exporting software and IT-enabled services mainly to the US market (see Figure 7.1). Bangalore became a popular location for global players to outsource lower-end software services. In more recent years, leading IT global players started outsourcing their R&D to Bangalore, allowing the cluster to move

272 Bangalore’s Software Cluster

gradually up the value chain by engaging in higher-end services such as embedded systems. At the same time, lower-end IT-enabled services have also grown rapidly. This new trend has led to the emergence of a skilldriven fragmented structure in the software and services cluster in Bangalore (see Figure 7.2). Three main processes have led to the agglomeration of ﬁrms in the cluster: (1) the relocation of global players to Bangalore to gain better access to the local low-cost skilled workforce by outsourcing software development and services; (2) new startups; and (3) spin-offs, that is former employees of large ﬁrms starting their own businesses. Earlier studies cited several factors as explaining the formation of the software cluster in Bangalore. They include the availability of a large pool of low-cost technical skilled labour; the presence of many research and higher-education institutions in the city because of the legacy of manufacturing development since the 1950s; actively engaged government institutions such as the STPI; and a relatively comfortable climate and cosmopolitan ambiance. This study’s closer examination of the local labour market in Bangalore adds more insights into the mechanisms that drove this rapid growth. As the industry relies so much on skilled technical labour – up to 80 per cent of operating costs for some ﬁrms – the functioning of the local labour market proves to be critical for the cluster in terms of ensuring the deployment, mobility and upgrading of the skills required by the different segments of the cluster. During the 1990s, the local labour market in Bangalore was very dynamic and vibrant, creating a virtuous cycle that generated a productive mechanism of learning in the cluster. In the 1990s, Bangalore experienced unusual growth in demand for IT professionals, making the local labour market extremely tight. The shortage of IT-related skills forced MNCs and large ﬁrms to compete ﬁercely to recruit the best local talent, by providing lucrative salaries and world-class work environments, as the industry relies heavily on these skills. At the same time, the fragmented structure within the software and services sector has expanded employment opportunities for a wide range of the workforce, not only engineering graduates in ITrelated ﬁelds but also those trained in other disciplines, serving different segments of the labour market. The large skill premiums paid to engineering graduates in IT-related ﬁelds, however, increased the demand for engineering education in these ﬁelds. Not only was a large and skilled labour pool available before Bangalore’s software cluster formed, but the skilled workforce in IT-related ﬁelds also expanded markedly in the late 1990s. The state government responded quickly to the enormous growth in skills demand, by increasing the number of engineering colleges, IT-related courses in engineering colleges, and the intake of students in IT-related courses. Its manpower planning exercise has been instrumental in adjusting the supply to the rising demand.

Aya Okada 273

High intra-ﬁrm transnational mobility, as well as inter-ﬁrm mobility within the local labour market, has helped ﬁrms to quickly deploy and mobilize the skill sets they require. The emergence of local labour market intermediaries such as stafﬁng agencies and the industry-wide collaboration in carrying out periodic compensation surveys helped facilitate the working of the local labour market, helping ﬁrms in the cluster to coordinate deployment, mobility, information sharing, and wage setting. The fragmented structure of the industry made the local labour market segmented and extremely skill-driven: skills are clearly deﬁned and speciﬁed in trading in the local labour market. This study also revealed that various institutional mechanisms for upgrading the skills of the workforce are built into both the internal labour markets – which are often transnational in MNCs and in leading Indian ﬁrms – and in the local external labour market. Well-coordinated and often project-based in-ﬁrm skills training play a critical role in upgrading the speciﬁc skills of employees. Skills developed through such in-ﬁrm training as well as academic research and training were further utilized through spin-offs, into many startup ﬁrms, particularly among SMEs. The relatively fast-track career path in large ﬁrms might also encourage their employees’ spin-offs. Coordination between the industry and top universities such as IISc and IIIT also helped upgrade the skills industry required. These mechanisms involving close interactions and coordination among ﬁrms, universities and government institutions have in turn helped the cluster better mobilize its skills to meet the global demand and move up the value chain in the software and services industry. Thus, they have promoted learning. The evolution of a very dynamic and vibrant local labour market has facilitated the deployment, mobility and skills upgrading of skilled labour within the cluster. The literature commonly points out that it is ineffective to have ‘public policies attempting to direct the formation of new clusters through top-down interventions, such as technologies, science parks, and ﬁrm incubators’ (Breschi and Malerba, 2001: 821–2). However, in Bangalore’s IT cluster, active interventions by the state government were critical, not only in promoting the IT industry within the state but also in closely linking its technical education policy to the industrial policy for promoting IT. The software and services industry is ultimately a knowledge-intensive one; thus ﬁrms must constantly improve their in-ﬁrm knowledge and skills, not only as a matter of organizational survival but also to move up the value chain. This dynamic and ﬂexible nature of the local labour market has shaped the behaviour of the ﬁrms in terms of recruiting the best talent, continuously offering in-house training, and working hard to retain that talent. It has also facilitated the process of upgrading skills, and the knowledge spillovers and externalities, thereby building competitiveness.

274 Bangalore’s Software Cluster

Notes * I thank the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of the Government of Japan for financial support to this research project under its Grant-in-Aid for Exploratory Research scheme (No. 13873006) for 20001–3. I also thank many people from the software industry, the government of Karnataka and educational institutions in Bangalore who generously shared with me their knowledge, experience and insights. 1. For a review of policies and research on industrial clusters in developing countries, see Nadvi and Schmitz (1994). 2. The ICT industry should be distinguished from the sectors that use ICT. But, in this chapter, following the deﬁnition by NASSCOM (2001), I deﬁne the software industry as consisting of three subsectors: software development, software services and IT-enabled services. 3. For a review of studies on localized knowledge spillovers, see Breschi and Lissoni (2001). 4. For details of these policies, see Saxenian (2001). 5. For details of the 1991 New Economic Policy, see Kaplinsky (1997), Humphrey et al. (1998) and Okada (2000). For implications of India’s economic reforms for the software industry, see Heeks (1996) and Parthasarathy (2000). 6. For discussions on the process of software production, see Heeks (1996). 7. Remote processing work includes a variety of activities such as call centres, data conversion, medical transcription, back-ofﬁce operations, content development, deposition summaries, insurance claims processing, and geographical information systems (ILO 2001: box 4.15). 8. Interview with an HRM manager, Sony, October 2002. 9. Government of Karnataka 2003 (http://www.bangaloreit.com/html/itsckar/ itindustries). 10. STPI, Bangalore, internal documents, 2001. 11. Interview with joint manager, STPI, Bangalore, December 2001. 12. Interview with vice-president, NASSCOM, Delhi, December 2001. 13. Interview with the Directorate of Information Technology, Government of Karnataka, December 2001. 14. For details of the Action Plan, see Saxenian (2001). 15. The Capacity Maturity Model (CMM) is a global quality certiﬁcation framework developed by the Software Engineering Institute (SEI) established in 1984. It describes the key elements of an effective software process and is composed of ﬁve levels; Level 5 is the highest. For details, see NASSCOM (2001). 16. Interview with a small software ﬁrm (Firm M) in Bangalore, October 2002. 17. IT professionals and knowledge workers means professionals engaged in software development, IT services and IT-enabled services including professionals engaged in software development units in user organizations. They include software programmers, project managers, network administrators, networking consultants, system analysts, web programmers, web designers, multimedia developers, EDP managers and IT-enabled services workers (NASSCOM 2001). 18. As of 2001, Japan projected a shortage of close to a million systems engineers in the ﬁrst half of the decade, while European countries have projected a shortage of a total of more than 200,000 IT professionals (NASSCOM 2001: 78). 19. HP India has two main operations: software development and IT-enabled services. Interview with a HRD manager, HP India, October 2002.

Aya Okada 275 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.

38. 39.

40.

41.

42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56.

Interview with a HRM manager, Sony India, October 2002. Interview with Hevitt Consulting, October 2002. Interview with a HRM manager, HP India, October 2002. Interview with a HRD staff, Infosys, October 2002. Rothboeck et al. (2001) also make this observation. Interview with a HRD manager, Wipro, October 2002. Interview with Hevitt Consulting, October 2002. Interview with P. N. Vengapal and Associates, a stafﬁng agency in Bangalore, October 2002. Interview with a HRM manager, TCS, Bangalore, October 2002. Interview with a large IT consulting ﬁrm (Firm I) in Bangalore, October 2002. Interview with vice-president, NASSCOM, December 2001. Interview with a HRD staff, Infosys, October 2002. Interview with a senior manager, Wipro, October 2002. Interview with a HRD manager, Wipro, October 2002. Interview with a faculty member from Bangalore University, October 2002. Interview with a HRD staff, Infosys, October 2002. Interview with a HRM manager, TCS, Bangalore, October 2002. Interview with a faculty member from Bangalore University, October 2002. Knowledge workers refer to ‘those whose work involves not merely the use of knowledge, but who generate ideas and new knowledge,’ being distinguished from those who manipulate information (data workers)(ILO 2001: box 4.2). Interview with STPI, Bangalore, October 2002. To enhance the quality of technical education in the country, in 1997, the central government set up the All India Council for Technical Education (AICTE) as a regulatory body to control the curriculum, norms and standards of technical education, including engineering colleges offering IT-related courses. In Bangalore, the average age of IT professionals was slightly higher. For example, at both Infosys and Sony India, it was 27 years (Interviews with managers, Infosys, Sony India, Bangalore, October 2002). With respect to gender distribution, 79 per cent of IT professionals in software ﬁrms were men, whereas 21 per cent were women in 2000 (NASSCOM 2001). This ﬁgure for female participation is still higher compared with the national average of 13 per cent for all industries. Interview with a faculty member, in the Department of Computer Science, IISc, October 2002. Interview with a managing director, Network Solutions, October 2002. Rothboeck et al. (2001) also make this observation. Interview with a HRM manager, HP India, October 2002. Interview with P. N. Vengapal and Associates, stafﬁng agency in Bangalore, October 2002. Interview with Hevitt Consulting, October 2002. Interview with Hevitt Consulting, October 2002. Interviews with faculty members, at IISc, October 2002. Interview with Hevitt Consulting, October 2002. Interview with a HRD manager, Wipro, October 2002. Interview with a HRM manager, HP India, October 2002. Interview with a HRM manager, TCS, October 2002. Interview with a HRD manager, Wipro, October 2002. Interview with a manager, Network Solutions, October 2002. Interviews with faculty members, at IISc, October 2002.

276 Bangalore’s Software Cluster 57. Interviews with under-secretary, Department of Higher Education, State Government of Karnataka, December 2001. 58. Interview with a professor at Bangalore University, October 2002. 59. Interview with P. N. Vengapal and Associates, a stafﬁng agency in Bangalore, October 2002. 60. Interview with P. N. Vengapal and Associates, a stafﬁng agency in Bangalore, October 2002.

References Amsden, Alice H. 1989. Asia’s Next Giant: South Korea and Late Industrialization. New York: Oxford University Press. Amsden, Alice H. and Takashi Hikino. 1994. ‘Project Execution Capability, Organizational Know-how and Conglomerate Corporate Growth in Late Industrialization’, Industrial and Corporate Change, 3 (1), 111–47. Ashton, David, Francis Green, Donna James, and Johnny Sung. 1999. Education and Training for Development in East Asia: The Political Economy of Skill Formation in East Asian Newly Industrializing Economies. London: Routledge. Breschi, Stefano and Franco Malerba. 2001. ‘The Geography of Innovation and Economic Clustering: Some Introductory Notes’, Industrial and Corporate Change, 10 (4), 817–33. Breschi, Stefano and Francesco Lissoni. 2001. ‘Knowledge Spillovers and Local Innovation Systems: A Critical Survey’, Industrial and Corporate Change, 10 (4), 975–1005. Bresnahan, Timothy, Alfonso Gambardella and AnnaLee Saxenian. 2001. ‘“Old Economy” Inputs for “New Economy” Outcomes: Cluster Formation in the New Silicon Valleys’, Industrial and Corporate Change, 10 (4), 835–60. Cooke, Philip. 2001. ‘Regional Innovation Systems, Clusters, and the Knowledge Economy’, Industrial and Corporate Change, 10 (4), 945–74. Dolan, Catherine S. and Meenu Tewari. 2001. ‘From What We Wear to What We Eat: Upgrading in Global Value Chains’, IDS Bulletin, 32 (3), 94–104. Government of India. 2001. IT Manpower, Challenge and Response: Interim Report of the Task Force on HRD in IT. Delhi: Department of Secondary and Higher Education, Ministry of Human Resource Development. Granovetter, Mark. 1985. ‘Economic Action and Social Structure: The Problem of Embeddedness’, American Journal of Sociology, 91 (3), 481–510. Heeks, Richard. 1996. India’s Software Industry: State Policy, Liberalization, and Industrial Development. Delhi: Sage. Humphrey, John, Avinandan Mukherjee, Mauro Zilbovicius, and Glauco Arbix. 1998. ‘Globalization, FDI and the Restructuring of Supplier Networks: The Motor Industry in Brazil and India’, in Mitsuhiko Kagami, John Humphrey and Michael Piore (eds), Learning, Liberalization and Economic Adjustment. Tokyo: Institute of Developing Economies. ILO. 2001. World Employment Report 2001: Life at Work in the Information Economy. Geneva: International Labour Ofﬁce. Kaplinksy, Raphael. 1997. ‘India’s Industrial Development: An Interpretative Study’, World Development, 25 (5), 681–94. Krugman, Paul. 1995. Development, Geography and Economic Theory. Cambridge, MA: MIT Press. Nadvi, Khalid and Hubert Schmitz. 1994. Industrial Clusters in Less Developed Countries: Review of Experience and Research Agenda, IDS Discussion Paper no. 339. Brighton: Institute of Development Studies, University of Sussex.

Aya Okada 277 NASSCOM. 2001. The IT Software and Services in India: 2001. Delhi: NASSCOM. NASSCOM. 2002. NASSCOM-McKinsey Report 2002: Strategies to Achieve the Indian IT Industry’s Aspiration. Delhi: NASSCOM. National Institute of Technology Karnataka (NITK). 2002. Annual Technical Manpower Review 1999–2003 for Karnataka State Surathkal: NITK. Okada, Aya. 2000. Workers’ Learning through Inter-ﬁrm Linkages in the Process of Globalization: Lessons from the Indian Automobile Industry. PhD dissertation, Massachusetts Institute of Technology. Okada, Aya. 2004. ‘Skills Development and Interfirm Learning Linkages under Globalization: Lessons from the Indian Automobile Industry’, World Development, 32(7), 1265–88. Parthasarathy, Balaji. 2000. Globalization and Agglomeration in Newly Industrializing Countries: The State and the Information Technology Industry in Bangalore, India. PhD dissertation, Berkeley, CA: University of California. Piore, Michael J. and Charles F. Sabel. 1984. The Second Industrial Divide. New York: Basic Books. Porter, Michael. 1990. The Competitive Advantage of Nations. New York: Free Press. Pyke, Frank, Giacomo Becattini and Werner Sengenberger, (eds). 1990. Industrial Districts and Inter-ﬁrm Cooperation in Italy. Geneva: International Institute for Labour Studies/ILO. Saxenian, Annalee. 1994. Regional Advantage: Culture and Competition in Silicon Valley and Route 128. Boston, MA: Harvard University Press. Saxenian, AnnaLee. 2001. Bangalore: The Silicon Valley of Asia? Centre for Research on Economic Development and Policy Reform Working Paper no. 91. Stanford, CA: Stanford University. Saxenian, AnnaLee and Jinn-Yuh Hsu. 2001. ‘The Silicon Valley–Hsinchu Connection: Technical Communities and Industrial Upgrading’, Industrial and Corporate Change, 10 (4), 893–920. Schmitz, Hubert and Bernard Musyck. 1993. Industrial Districts in Europe: Policy Lessons for Developing Countries? IDS Discussion Paper no. 324. Brighton: Institute of Development Studies, University of Sussex. Schmitz, Hubert, and Khalid Nadvi. 1999. ‘Clustering and Industrialization: Introduction’, World Development, 27 (9), 1503–14. Sharma, Ravi. 2001. ‘In a High-Growth Trajectory’, Frontline, December 21, 2001, 65–72. Tschang, Ted. 2001. The Basic Characteristics of Skills and Organizational Capabilities in the Indian Software Industry. ADB Institute Working Paper no. 13. Tokyo: ADB Institute. Rothboeck, S., M. Vijayabask, and V. Gayathri. 2001. Labour in the New Economy: The Case of the Indian Software Labour Market. New Delhi: International Labour Organization.

8 The Local Economy: Lessons from Japan Rika Nakagawa

Introduction Since the Plaza Accords in 1985, many Japanese manufacturing companies have moved production bases from Japan to overseas, such as Southeast Asia and East Asia. One reason for this was that the Plaza Accords caused the Japanese currency to appreciate sharply. This currency appreciation resulted in an increase of production costs at home. In order for Japanese manufacturing ﬁrms to compete with other companies in global markets, they needed to reduce production costs. On the other hand, Southeast and East Asian developing countries need foreign direct investment (FDI) for their economic development. Therefore, governments of those countries implemented several policies in order to invite manufacturing ﬁrms from developed countries. As a result, a large number of manufacturing ﬁrms moved their production facilities from Japan to developing countries in Southeast and East Asia. In the late 1990s, Japanese manufacturing companies began to invest in factories or ofﬁces in China owing to low production costs. These dynamic changes in ﬁrms’ investment strategies, coupled with an economic deterioration in Japan, have had a negative impact on the Japanese local economy. As a consequence, per capita incomes of many prefectures became less than that of Japan’s average. Looking at per capita income in each prefecture throughout the 1990s, the highest, in Tokyo, became twice the lowest in Okinawa. In 2001, owing to the decline of the local economy, the government formulated a plan to activate the local economy by developing industrial clusters. Recently, there has been much discussion of these. Ishikura (2003) explains that industrial clusters can promote product innovations and have a positive impact on the local or regional economy. Matsushima (1998) emphasised that from the viewpoint of small- and medium-sized enterprises (SMEs), industrial clusters promote effective and efﬁcient business activities. 278

Rika Nakagawa 279

Based on the above background, this chapter focuses on policies implemented by central and local governments, because when central and local governments intend to form new industrial clusters in certain areas, policy planning is crucial. The purpose of this chapter is to introduce Japan’s Industrial Cluster Plan of 2001, and to learn how local governments have attracted companies to their prefecture in order to form industrial clusters. As a precedent case to examine how local governments responded to this initiative, this chapter focuses on the case of Okinawa prefecture. The government of Okinawa prefecture has implemented several policies to invite the information, communications and technology (ICT) industry since 1999. This means that Okinawa had already tried to form an industrial cluster before the central government announced the Industrial Cluster Plan. Therefore, Okinawa’s case is a good sample for understanding how a local government has tried to invite private companies. In order for Okinawa prefecture to achieve further development of the ICT industry, this chapter also discusses the relationship between policies implemented by the Okinawa government and those of the central government. As part of the broader discussion, this chapter focuses on whether policies of Okinawa prefecture can be supplemented by the Industrial Cluster Plan and attempts to draw lessons from the case of Japan in order to contribute to the stable industrial development of the local economy. The structure of this chapter is as follows. The next section explains the background of the Industrial Cluster Plan of 2001 and the details of the plan. The third section reviews the case of Okinawa prefecture. Okinawa prefecture has focused on the ICT industry, such as call centres, information services, contents creation companies and software development, and has implemented several policies in order to attract ICT companies. The fourth and ﬁnal section summarizes the chapter and draws some lessons from experiences of Japan.

The Industrial Cluster Plan of 2001 Background of the Industrial Cluster Plan of 2001 The Japanese government formulated a plan to establish industrial clusters in the local economy owing to changes in the international economic situation. There were two reasons why Japan committed to implement policies for forming industrial clusters: one was the agreement of the Plaza Accords, while the other was as a response to the relocation of many Japanese factories to China, the so-called ‘China Shift’. After the Plaza Accords in 1985, the Japanese currency appreciated sharply. This caused Japanese manufacturing companies to move to Asian countries, such as Taiwan, Thailand, Malaysia and so on in order to reduce costs for production. This decision by the business sector resulted in decreasing the number of manufacturing factories. Table 8.1 shows the

Number of manufacturing factories in Japan, 1985–2000 1985

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

280

Table 8.1

1990

Prefecture

Number

Number

Hokkaido Aomori Iwate Miyagi Akita Yamagata Fukushima Ibaraki Tochigi Gunma Saitama Chiba Tokyo Kanagawa Niigata Toyama Ishikawa Fukui Yamanashi Nagano Gifu Shizuoka Aichi Mie Shiga

9,627 2,546 3,408 4,894 3,343 4,843 7,458 9,431 8,711 9,843 23,380 9,681 49,897 17,555 10,659 4,708 6,669 5,018 3,842 10,434 12,925 19,166 35,379 7,517 4,658

9,940 2,701 3,898 5,151 3,875 5,166 7,932 9,887 8,835 10,183 25,042 9,943 42,804 17,822 10,999 4,859 6,637 4,782 3,924 10,286 13,279 19,366 35,458 7,650 4,641

Average growth rate (1985–1990, %) 0.7 1.2 2.9 1.1 3.2 1.3 1.3 1.0 0.3 0.7 1.4 0.5 –2.8 0.3 0.6 0.6 –0.1 –0.9 0.4 –0.3 0.5 0.2 0.0 0.4 –0.1

1995

Number 9,644 2,705 3,716 4,947 3,680 4,719 7,228 9,152 8,037 9,120 21,454 9,361 34,321 15,442 10,058 4,516 5,960 4,449 3,520 9,302 11,735 17,479 31,441 6,895 4,338

Average growth rate (1990–1995, %) –0.6 0.0 –0.9 –0.8 –1.0 –1.7 –1.8 –1.5 –1.8 –2.1 –2.9 –1.2 –4.0 –2.7 –1.7 –1.4 –2.0 –1.4 –2.1 –1.9 –2.3 –1.9 –2.3 –2.0 –1.3

2000

Number 8,817 2,406 3,305 4,493 3,194 4,124 6,335 8,302 7,067 8,154 19,223 8,210 30,096 14,082 8,649 4,198 4,920 3,849 3,082 8,281 10,057 15,736 27,762 6,121 3,985

Average growth rate (1995–2000, %) –1.7 –2.2 –2.2 –1.8 –2.6 –2.5 –2.5 –1.9 –2.4 –2.1 –2.1 –2.5 –2.5 –1.8 –2.8 –1.4 –3.5 –2.7 –2.5 –2.2 –2.9 –2.0 –2.3 –2.2 –1.6

Table 8.1

Number of manufacturing factories in Japan, 1985–2000 1985

26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

continued

1990 Average growth rate (1985–1990, %)

1995

Number

Average growth rate (1990–1995, %)

2000

Number

Average growth rate (1995–2000, %)

Prefecture

Number

Number

Kyoto Osaka Hyogo Nara Wakayama Tottori Shimane Okayama Hiroshima Yamaguchi Tokushima Kagawa Ehime Kochi Fukuoka Saga Nagasaki Kumamoto Oita Miyazaki Kagoshima Okinawa

11,323 45,814 18,798 4,600 4,164 1,816 2,591 7,291 9,495 3,481 3,099 4,250 5,065 2,147 9,661 2,293 3,202 3,714 2,586 2,361 3,745 1,430

10,473 44,585 18,636 4,472 4,087 1,972 2,662 7,120 9,473 3,566 3,110 4,252 5,032 2,093 9,915 2,352 3,193 3,789 2,567 2,506 3,629 1,453

–1.5 –0.5 –0.2 –0.6 –0.4 1.7 0.5 –0.5 0.0 0.5 0.1 0.0 –0.1 –0.5 0.5 0.5 –0.1 0.4 –0.1 1.2 –0.6 0.3

8,920 38,406 15,950 4,019 3,507 1,718 2,346 6,455 8,756 3,161 2,698 3,873 4,568 1,934 9,492 2,346 3,082 3,499 2,503 2,428 3,327 1,519

–3.0 –2.8 –2.9 –2.0 –2.8 –2.6 –2.4 –1.9 –1.5 –2.3 –2.6 –1.8 –1.8 –1.5 –0.9 –0.1 –0.7 –1.5 –0.5 –0.6 –1.7 0.9

7,599 32,557 13,947 3,582 3,012 1,486 2,063 5,435 7,548 2,842 2,290 3,124 3,893 1,686 8,638 2,077 2,853 3,118 2,341 2,132 3,246 1,504

–3.0 –3.0 –2.5 –2.2 –2.8 –2.7 –2.4 –3.2 –2.8 –2.0 –3.0 –3.9 –3.0 –2.6 –1.8 –2.3 –1.5 –2.2 –1.3 –2.4 –0.5 –0.2

438,518

435,997

–0.1

387,726

–2.2

341,421

–2.4

Japan

281

Sources: Ministry of International Trade and Industry (1987, 1992, 1997).

282 The Local Economy: Lessons from Japan

number of manufacturing factories from 1985 to 2000. In 1985, there were 438,518 factories in Japan; however, after that, the number of manufacturing factories decreased. In 1990 and 1995 there were 435,997 and 387,726, respectively. By 2000, the number had decreased to 341,421. From 1985 to 1990, the average growth rate of the number of factories became –0.1 per cent, and from 1990 to 1995 it was –2.2 per cent. From 1995 to 2000, the average growth rate was also negative at –2.4 per cent. Given the above situation, it is clear that Japan has been facing a serious decline in the number of manufacturing ﬁrms. Looking at each prefecture, from 1985 to 1990, Tokyo and Kyoto experienced the most severe situation. In 1985, the number of manufacturing ﬁrms in Tokyo was 49,897, whereas in 1990 it decreased to 42,804. In Kyoto in 1985, there were 11,323 ﬁrms and this fell to 10,473 ﬁrms in 1990. From 1990 to 1995, most prefectures faced a decrease in the number of manufacturing factories. From 1995 to 2000, the situation became more serious as most prefectures experienced a further decline in the number of factories. In addition, the rate of decrease accelerated. In particular, the growth rate in Kagawa, Ishikawa and Okayama became –3.9 per cent, –3.5 per cent and –3.2 per cent, respectively. In the early 1990s, the business situation changed again. Costs for production in Southeast and East Asian countries started to rise. Owing to this, Japanese manufacturing companies shifted their bases from Southeast and East Asia to China. One reason for this relocation was that the Chinese government started to implement policies to attract a great number of manufacturing companies from all over the world. For example, since 1980, the Chinese government has established 206 industrial zones. As shown in Table 8.2, the national government designated 122 zones, whereas local governments set up 84 zones. From 1980 to 1990 the national and local governments set up 32 industrial zones. From 1991 to 1995 the number of industrial zones established by the national and local governments jumped to 146: 77 at the national level and 69 at the local. From 1996 to 2000 it was 28 and 19 at the national level and 9 at the local. Table 8.2

Year 1980–1985 1986–1990 1991–1995 1996–2000 Total

Number of industrial zones in China National level

Local level

Number

Number

Total

9 17 77 19

1 5 69 9

10 22 146 28

122

84

206

Source: Based on JETRO (2001).

Rika Nakagawa 283

In those industrial zones, the government provides beneﬁts to companies in the district. First, companies in the areas can obtain tax deductions. Second, companies can complete customs formalities at a custom house established in the zones. Due to this, export-oriented companies are able to export their goods smoothly. Third, companies in the industrial areas can obtain visas for employees from an ofﬁce on-site, because the government permits administrative ofﬁces in the areas to issue visas. These preferential measures and the low costs of production led manufacturing factories to move to China. This movement, the so-called ‘China shift,’ had a great impact on the Japanese economy. In particular, the local economy has suffered from the China shift. Looking at per capita income in each prefecture as shown in Table 8.3, from 1992 to 2000 the average per capita income in Japan was about ¥3m. From 1992 to 2000, Saitama, Chiba, Tokyo, Kanagawa, Shizuoka, Aichi, Shiga and Osaka reported per capita income above the national average, whereas others fell below. Among all prefectures, Tokyo achieved the highest per capita income and Okinawa the lowest. By comparison, from 1990 to 2000 the per capita income of Tokyo was more than twice that of Okinawa. In the case of growth rates in each prefecture (Table 8.4), in 1992 many prefectures were negative. By 1996 some prefectures had recovered to positive growth rates. However, after 1997 most of them again posted negative growth rates. Thus, the currency appreciation after 1985 caused manufacturing ﬁrms to shift production to developing countries, and the ensuing economic deterioration in Japan damaged the local economy. In particular, small and medium-sized companies (SMEs) faced a serious situation. A number of companies went into bankruptcy. The serious economic situation in the local economy led the national government to take some action. The Industrial Cluster Plan, announced in 2001, was one such response. The government of Japan implemented this plan in order to stimulate the business activities of SMEs and promote innovation by SMEs. The government expects through the implementation of the plan to stimulate recovery in the local economy. The Industrial Cluster Plan of 2001 The Industrial Cluster Plan of 2001 consists of three core programs (Table 8.5). One program is support for companies to form alliances with other companies or universities. This program is aimed at developing networks between the business sector and research institutions. The second program is to help technical innovations in the industrial sector. The third program is to found institutions for the promotion of entrepreneurs. The government appropriated about ¥476bn in funds for the three programs. For the ﬁrst program, ‘network formation’, the central government allocated ¥39bn. There are seven sub-programs in the network formation

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 26 27 28

Per capita income in each Japanese prefecture (thousand yen)

284

Table 8.3

Prefecture

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

Hokkaido Aomori Iwate Miyagi Akita Yamagata Fukushima Ibaraki Tochigi Gunma Saitama Chiba Tokyo Kanagawa Niigata Toyama Ishikawa Fukui Yamanashi Nagano Gifu Shizuoka Aichi Mie Shiga Kyoto Osaka Hyogo

2,503 2,259 2,228 2,487 2,181 2,408 2,430 2,955 2,896 2,759 2,916 3,191 4,198 3,206 2,515 2,925 2,798 2,748 2,698 2,802 2,821 3,047 3,422 2,711 3,050 2,820 3,514 2,750

2,647 2,327 2,312 2,768 2,270 2,584 2,558 3,112 3,039 2,956 3,144 3,319 4,222 3,326 2,707 3,060 2,880 2,875 2,802 2,879 2,904 3,158 3,634 2,885 3,254 2,921 3,696 2,897

2,696 2,377 2,389 2,733 2,249 2,585 2,575 3,019 2,990 2,910 3,193 3,288 4,134 3,323 2,744 3,045 2,884 2,879 2,757 2,881 2,954 3,073 3,508 2,877 3,258 2,891 3,672 2,884

2,726 2,346 2,433 2,675 2,267 2,560 2,528 3,012 3,054 2,932 3,216 3,380 3,851 3,315 2,766 3,071 2,915 2,887 2,752 2,870 2,937 3,048 3,455 2,845 3,194 2,882 3,521 2,890

2,811 2,467 2,574 2,762 2,330 2,619 2,683 3,109 3,000 3,050 3,280 3,348 3,964 3,314 2,836 3,036 2,903 3,033 2,807 2,940 2,897 3,180 3,503 2,878 3,332 2,987 3,447 2,861

2,868 2,471 2,584 2,768 2,383 2,682 2,742 3,071 3,142 3,108 3,252 3,324 4,086 3,383 2,869 3,018 2,926 3,003 2,848 3,039 2,913 3,201 3,612 2,984 3,335 3,054 3,439 3,061

2,901 2,518 2,650 2,881 2,534 2,764 2,907 3,162 3,275 3,138 3,315 3,358 4,270 3,500 3,001 3,225 3,065 3,128 2,940 3,092 3,035 3,297 3,820 3,066 3,599 3,127 3,491 3,162

2,906 2,460 2,625 2,895 2,526 2,752 2,930 3,198 3,299 3,169 3,364 3,390 4,316 3,460 2,982 3,113 3,070 3,147 2,955 3,066 2,998 3,277 3,676 3,046 3,525 3,078 3,475 3,099

2,841 2,420 2,613 2,786 2,462 2,666 2,876 3,058 3,195 3,049 3,289 3,245 4,331 3,341 2,953 2,981 2,992 3,016 2,733 2,945 2,906 3,241 3,571 2,973 3,270 2,981 3,339 2,918

2,866 2,449 2,635 2,792 2,448 2,684 2,804 3,029 3,151 3,056 3,240 3,218 4,440 3,241 2,886 2,904 2,963 2,980 2,825 2,948 2,903 3,161 3,550 2,928 3,240 2,935 3,358 2,810

2,856 2,519 2,673 2,769 2,424 2,685 2,801 3,005 3,213 3,003 3,259 3,205 4,365 3,262 2,892 2,931 2,970 3,030 2,897 3,005 2,886 3,194 3,498 2,972 3,270 3,029 3,303 2,845

Table 8.3

29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

Per capita income in each Japanese prefecture (thousand yen)

continued

Prefecture

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

Nara Wakayama Tottori Shimane Okayama Hiroshima Yamaguchi Tokushima Kagawa Ehime Kochi Fukuoka Saga Nagasaki Kumamoto Oita Miyazaki Kagoshima Okinawa

2,783 2,205 2,447 2,224 2,516 2,876 2,600 2,501 2,513 2,249 2,141 2,573 n.a. 2,055 n.a. 2,387 2,005 2,061 1,928

2,860 2,363 2,560 2,350 2,775 3,013 2,762 2,594 2,656 2,312 2,281 2,638 2,282 2,138 n.a. 2,486 2,110 2,144 1,994

2,800 2,344 2,537 2,368 2,792 3,009 2,858 2,621 2,666 2,350 2,329 2,604 2,403 2,142 2,396 2,495 2,154 2,159 2,024

2,803 2,433 2,569 2,355 2,692 2,983 2,754 2,647 2,710 2,457 2,332 2,584 2,419 2,188 2,403 2,509 2,148 2,121 2,071

2,868 2,403 2,587 2,403 2,733 3,010 2,793 2,688 2,784 2,558 2,455 2,673 2,474 2,270 2,488 2,602 2,299 2,197 2,043

2,927 2,438 2,643 2,441 2,797 3,028 2,809 2,742 2,709 2,558 2,457 2,693 2,530 2,343 2,496 2,638 2,316 2,234 2,052

2,999 2,549 2,708 2,540 2,868 3,095 2,908 2,790 2,893 2,684 2,485 2,751 2,625 2,437 2,652 2,692 2,471 2,296 2,107

3,007 2,522 2,685 2,531 2,773 3,069 2,884 2,800 2,932 2,578 2,438 2,750 2,602 2,427 2,651 2,772 2,456 2,284 2,122

2,959 2,400 2,603 2,501 2,772 3,014 2,793 2,767 2,921 2,517 2,430 2,685 2,574 2,350 2,554 2,699 2,444 2,274 2,124

2,886 2,443 2,598 2,479 2,716 2,962 2,790 2,718 2,793 2,431 2,366 2,635 2,610 2,325 2,544 2,685 2,467 2,289 2,124

2,888 2,462 2,623 2,541 2,649 2,961 2,844 2,705 2,788 2,495 2,404 2,660 2,580 2,345 2,646 2,765 2,440 2,325 2,125

Japan

n.a.

n.a.

3,049

3,012

3,061

3,109

3,211

3,200

3,118

3,100

3,101

Note: Emphasis made by the author means that per capita incomes of these prefectures were more than that of Japan. n.a. = not available

285

Source: Download from the Cabinet Ofﬁce home page. (http://www.esri.cao.go.jp).

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 26 27 28

Growth rates of per capita income in each prefecture (%)

286

Table 8.4

Prefecture

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

Hokkaido Aomori Iwate Miyagi Akita Yamagata Fukushima Ibaraki Tochigi Gunma Saitama Chiba Tokyo Kanagawa Niigata Toyama Ishikawa Fukui Yamanashi Nagano Gifu Shizuoka Aichi Mie Shiga Kyoto Osaka Hyogo

5.8 3.0 3.8 11.3 4.1 7.3 5.3 5.3 4.9 7.1 7.8 4.0 0.6 3.7 7.6 4.6 2.9 4.6 3.9 2.7 2.9 3.6 6.2 6.4 6.7 3.6 5.2 5.3

1.9 2.1 3.3 –1.3 –0.9 0.0 0.7 –3.0 –1.6 –1.6 1.6 –0.9 –2.1 –0.1 1.4 –0.5 0.1 0.1 –1.6 0.1 1.7 –2.7 –3.5 –0.3 0.1 –1.0 –0.6 –0.4

1.1 –1.3 1.8 –2.1 0.8 –1.0 –1.8 –0.2 2.1 0.8 0.7 2.8 –6.8 –0.2 0.8 0.9 1.1 0.3 –0.2 –0.4 –0.6 –0.8 –1.5 –1.1 –2.0 –0.3 –4.1 0.2

3.1 5.2 5.8 3.3 2.8 2.3 6.1 3.2 –1.8 4.0 2.0 –0.9 2.9 0.0 2.5 –1.1 –0.4 5.1 2.0 2.4 –1.4 4.3 1.4 1.2 4.3 3.6 –2.1 –1.0

2.0 0.2 0.4 0.2 2.3 2.4 2.2 –1.2 4.7 1.9 –0.9 –0.7 3.1 2.1 1.2 –0.6 0.8 –1.0 1.5 3.4 0.6 0.7 3.1 3.7 0.1 2.2 –0.2 7.0

1.2 1.9 2.6 4.1 6.3 3.1 6.0 3.0 4.2 1.0 1.9 1.0 4.5 3.5 4.6 6.9 4.8 4.2 3.2 1.7 4.2 3.0 5.8 2.7 7.9 2.4 1.5 3.3

0.2 –2.3 –0.9 0.5 –0.3 –0.4 0.8 1.1 0.7 1.0 1.5 1.0 1.1 –1.1 –0.6 –3.5 0.2 0.6 0.5 –0.8 –1.2 –0.6 –3.8 –0.7 –2.1 –1.6 –0.5 –2.0

–2.2 –1.6 –0.5 –3.8 –2.5 –3.1 –1.8 –4.4 –3.2 –3.8 –2.2 –4.3 0.3 –3.4 –1.0 –4.2 –2.5 –4.2 –7.5 –3.9 –3.1 –1.1 –2.9 –2.4 –7.2 –3.2 –3.9 –5.8

0.9 1.2 0.8 0.2 –0.6 0.7 –2.5 –0.9 –1.4 0.2 –1.5 –0.8 2.5 –3.0 –2.3 –2.6 –1.0 –1.2 3.4 0.1 –0.1 –2.5 –0.6 –1.5 –0.9 –1.5 0.6 –3.7

–0.3 2.9 1.4 –0.8 –1.0 0.0 –0.1 –0.8 2.0 –1.7 0.6 –0.4 –1.7 0.6 0.2 0.9 0.2 1.7 2.5 1.9 –0.6 1.0 –1.5 1.5 0.9 3.2 –1.6 1.2

Table 8.4

29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

Growth rates of per capita income in each prefecture (%)

continued

Prefecture

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

Nara Wakayama Tottori Shimane Okayama Hiroshima Yamaguchi Tokushima Kagawa Ehime Kochi Fukuoka Saga Nagasaki Kumamoto Oita Miyazaki Kagoshima Okinawa

2.8 7.2 4.6 5.7 10.3 4.8 6.2 3.7 5.7 2.8 6.5 2.5 n.a. 4.0 n.a. 4.1 5.2 4.0 3.4

–2.1 –0.8 –0.9 0.8 0.6 –0.1 3.5 1.0 0.4 1.6 2.1 –1.3 5.3 0.2 n.a. 0.4 2.1 0.7 1.5

0.1 3.8 1.3 –0.5 –3.6 –0.9 –3.6 1.0 1.7 4.6 0.1 –0.8 0.7 2.1 0.3 0.6 –0.3 –1.8 2.3

2.3 –1.2 0.7 2.0 1.5 0.9 1.4 1.5 2.7 4.1 5.3 3.4 2.3 3.7 3.5 3.7 7.0 3.6 –1.4

2.1 1.5 2.2 1.6 2.3 0.6 0.6 2.0 –2.7 0.0 0.1 0.7 2.3 3.2 0.3 1.4 0.7 1.7 0.4

2.5 4.6 2.5 4.1 2.5 2.2 3.5 1.8 6.8 4.9 1.1 2.2 3.8 4.0 6.3 2.0 6.7 2.8 2.7

0.3 –1.1 –0.8 –0.4 –3.3 –0.8 –0.8 0.4 1.3 –3.9 –1.9 0.0 –0.9 –0.4 0.0 3.0 –0.6 –0.5 0.7

–1.6 –4.8 –3.1 –1.2 0.0 –1.8 –3.2 –1.2 –0.4 –2.4 –0.3 –2.4 –1.1 –3.2 –3.7 –2.6 –0.5 –0.4 0.1

–2.5 1.8 –0.2 –0.9 –2.0 –1.7 –0.1 –1.8 –4.4 –3.4 –2.6 –1.9 1.4 –1.1 –0.4 –0.5 0.9 0.7 0.0

0.1 0.8 1.0 2.5 –2.5 0.0 1.9 –0.5 –0.2 2.6 1.6 0.9 –1.1 0.9 4.0 3.0 –1.1 1.6 0.0

n.a.

n.a.

–1.2

1.6

1.6

3.3

–0.3

–2.6

–0.6

0.0

Japan Note: n.a. = not available.

Source: Download from the Cabinet Ofﬁce home page.

287

288

Table 8.5

Details of the Japanese industrial cluster plan 2001 (billion yen)

Network formation Formation of networks Training courses for incubator managers

Budget

Research & Development

Budget

Incubator Establishment

6.3

R&D by consortiums

116.1

Establishment of incubators

0.8

R&D by cooperative system between universities and industries*

Promotion of information technology

5.0

R&D by SMEs*

Support for venture businesses

4.1

Coordination for SMEs

3.6

Development and Diffusion of IT-based business model*

Patents fair Support for local governments

2.1 17.1

Support for development of production Technology 39

96.0

Support for local governments* 7.5 108.5

Subsidy for establishment of incubators* Subsidy for purchasing equipments*

26.6 2.4

7.1

Support for development of broad-band Contents

Total

Budget

Total

5.4 67.0 312

Note: * The Japanese government supports up to a half of expenditures by the local governments. Sources: Ministry of Economy, Trade and Industry home page.

Total

125

Rika Nakagawa 289

program. In these sub-programs, the government aims to connect the industrial sector with research institutions to exchange information. In terms of research institutions, the beneﬁt of these sub-programs is understanding what kind of technology is needed in the industrial sector. From the viewpoint of private companies, they can develop production technology with specialists at research institutions. For the second program, ‘research and development (R&D)’, the government allocated more than 65 per cent of the budget, ¥312bn. This program consists of six sub-programs. The purpose of these sub-programs is to enhance research activities in the industrial sector. A remarkable feature of the second program is that some of the sub-programs focus on SMEs. In the IT-based economy, it is hard for SMEs to survive in competitive global markets. In order for SMEs to compete with other manufacturing companies in global markets, the SMEs need to enter a new business ﬁeld or develop their technology. The SMEs, however, have limited resources to keep up with large enterprises in the area of R&D. Therefore, the government has targeted SMEs in this program. For the third program, ‘incubator establishment’, the government assigned ¥125bn. This program consists of three sub-programs. These sub-programs are for establishing incubator facilities and training people to start businesses. Two of the sub-programs are for local governments. Local governments can request funds for the promotion of entrepreneurs and for the establishment of incubators. As a result of these government policies, since 2001 the government has provided information on business alliances to more than 5,000 companies. Moreover, the government has played the role of an agent between companies and universities to foster business cooperation. Local governments have thus far posted some success in attracting private companies and enhancing industrial clusters. As of December 2003, 24 projects have been Table 8.6 Region

Projects of industrial cluster planning in each Japanese region Number of projects

Hokkaido Tohoku Kanto Chubu Kinki Chugoku Shikoku Kyushu Okinawa Total Source: Ministry of Economy, Trade and Industry home page.

2 2 7 3 4 2 1 2 1 24

290 The Local Economy: Lessons from Japan

executed. The Kanto, Kinki and Chubu areas have each executed more than three projects, whereas other areas have executed one or two projects (Table 8.6).

The case of Okinawa Prefecture The current situation of the ICT cluster in Okinawa Since 1999, the government of Okinawa has implemented several policies to attract the ICT industry. The ﬁrst reason why the prefectural government targeted the ICT industry is that the central government appointed Okinawa Prefecture as an ICT promotion zone. The efforts of the Okinawa prefectural government are a precedent case for understanding how the local government has tried to attract private companies. One reason for this, as noted in the second section, is that during the 1990s per capita income in Okinawa was the lowest in the country. Thus, it was crucial for Okinawa to form an industrial cluster and to stimulate economic activities there. In order for local economies to maintain stable economic growth, understanding how and what measures the Okinawa government has taken to invite the ICT industry is essential to drawing some useful lessons for the country as a whole. The ICT industry seems to satisfy requirements, pointed out by Itami (1998), for forming an industrial cluster in Okinawa. According to Itami (1998), there are three minimum requirements for industrial clusters. The ﬁrst requirement is the accumulation of knowledge and technology. Second, companies must be able to coordinate the division of labour at minimal cost. Third, it is important for companies to be able move their premises without any difﬁculty, and to start their operation as soon as they move. The ICT industry does not require a large factory space, as does the automobile industry, for example. ICT employees work on computer terminals and telecommunication networks, which require only ofﬁce space. Because Okinawa is a small island, it is difﬁcult to develop industries that need large factories. In this sense, the strategy for the ICT cluster agrees with the requirements in the case of Okinawa. Figure 8.1 shows the establishment of the ICT industry in Okinawa. As of May 2003, the number of establishments of ICT ﬁrms there was 80; 34 call centres, 17 software companies, 10 information service companies, 13 contents companies and 6 companies in other categories. A common feature among all categories is that a rapid increase in the number of establishments started from 1999, Until 1998, the total number of companies was only 8, but in 1999 it jumped to 21. From 2000 to 2003, the number of ofﬁces doubled from 40 to 80. As Figure 8.1 shows, the establishment of call centres in particular increased rapidly. In 1998, these numbered only 5, but in the next year they increased to 11 and in 2000 to 20; they continued to increase until

Rika Nakagawa 291 Figure 8.1

Total number of ICT companies in Okinawa, May 2003

90 Others

80 70

Contents creation

Number

60 50

Infomation services

40 30

Software development

20

Call centre

10

20 03

20 02

20 01

20 00

19 99

19 98

19 97

19 96

19 94 19 95

0

Source: Okinawa Prefectural Government (2003).

May 2003. In the case of software development, until 1996 there was no establishment, but in 1997, 2 companies came and started business and in 1999, 5 more. After 2000, the total number of software companies continued to increase. In the case of information services, until 1998 no company was established in Okinawa, then in 1999, 2 information services companies started their business there, and in 2000 and 2001, 2 more; in 2002, in total, the number had reached 10 companies. With regard to contents creation companies, the ﬁrst was established in 1997, then no more until a further 5 in 2000; in 2001 and 2002 only a few companies were established, but in 2003, 4 more started business operations in the prefecture. Policies of the Okinawa government As has been noted in the last section, in Okinawa, the industrial cluster of ICT companies has gradually enhanced. Among categories, the number of call centres increased rapidly compared with others. This resulted from the contribution of the policies of the government of Okinawa. This section explains how the Okinawa government has invited the ICT industry to the prefecture, using the ﬁve measures now described to develop the ICT cluster (Table 8.7). Support for reduction of telecommunication costs The prefectural government has provided ﬁnancial support for ICT companies established in Okinawa to reduce telecommunication costs. In order for a company to obtain support, it needs to comply with three conditions.

Table 8.7

Details of support for the ICT industrial cluster in Okinawa Purpose

Project carrier

Terms and conditions for the government’s support

1 Telecommunication cost

To reduce the cost of telecommunications

Tropical Technology Center Ltd

Companies fulﬁlling the following: 1. Hiring another 20 full-time employees within three years 2. Hiring another 10 full-time skilled workers within three years 3. Training people for obtaining special knowledge of telecommunications 4. Contributing to promotion of IT-based industry in the prefecture

2 The ICT cluster

To enhance the ICT cluster in the prefecture

Okinawa Prefectural Government

Companies fulﬁlling the following: 1. Establishing a center for preserving data 2. Hiring new employees

3 Young employees

To increase the employment of young people

Public Employment Security Ofﬁces

Business owners fulﬁlling the following: 1. Expending more than ¥3m. for new investments 2. Hiring employees under 30 years old

4 Human resource development

To train people willing to work for call centers

Operation Department of Employment Activation Organization

People fulﬁlling the following: 1. Residents of Okinawa prefecture

To train engineers to obtain advanced knowledge for IT

Okinawa Prefectural Government

People fulﬁlling the following: 1. IT engineers living in Okinawa

To promote people with advanced skills of contents creating and broadcasting

Frontier Region of Multimedia, Okinawa

People fulﬁlling the following: 1. IT engineers living in Okinawa 2. Hiring new employees

5 Information and communication industry

To provide tax incentives Okinawa to the IT industry Prefectural Government

Source: Okinawa Prefectural Government (2003).

Companies fulﬁlling the following: 1. Operation within the designated areas in Okinawa

292

Target of the project

Rika Nakagawa 293

First, it is required to hire at least 20 employees. Second, it must to hire another 10 skilled employees or engineers as well. In both cases, employees must be full-time, and both these conditions need to be fulﬁlled within three years of obtaining the subsidy. The third condition is that the company trains employees to master advanced knowledge for telecommunications. Based on these conditions, the Okinawa government selects companies which contribute to promoting the ICT industry in Okinawa, and gives them ﬁnancial support (Okinawa Prefectural Government 2003:6–7). Development of the ICT industrial cluster The Okinawa government has allocated subsidies for enhancing ICT clusters because it wishes to make Okinawa a hub for the ICT industry in Japan. Moreover, Okinawa has a big advantage over other prefectures in telecommunication costs, because the Japanese government laid a special telecommunication cable between Tokyo and Okinawa in order to reduce costs for telecommunications. Owing to this, such costs between Okinawa and Tokyo are inexpensive as compared with other prefectures. The government of Okinawa has taken full advantage of this comparative advantage to invite ICT companies to locate in the prefecture. Looking at the terms and conditions set out by the prefectural government, there are two conditions that a company must comply with. First, a company that applies for this subsidy is required to establish a data centre, a facility for preserving information and data. However, it is possible for a company to obtain funds if it transfers its existing data centre from another prefecture to Okinawa. Second, a company must hire new employees. If the government accepts a subsidy request, then a company can obtain funds of up to 50 per cent of the total cost of establishing a data centre or of the costs incurred in transferring an existing facility to Okinawa (Okinawa Prefectural Government 2003:8–9). Enhancement of job opportunities for young people The purpose of this support is to increase job opportunities for people under 30. An entrepreneur who wishes to obtain the subsidy is required to submit an application form to the Public Employment Security Ofﬁce in Okinawa. After the application is received the ofﬁce conﬁrms whether the entrepreneur complies with two conditions. The ﬁrst condition is that the entrepreneur expends more than ¥3m. for new investments. The second condition is that the entrepreneur must hire employees under 30 years old. Moreover, such employees must constitute more than 50 per cent of the total workforce. When the Public Employment Security Ofﬁce ﬁnds the two requirements are fulﬁlled, they supply a subsidy. The maximum amount of the subsidy is ¥100,000 per employee per month. The government provides this subsidy for up to 100 people for 2 years in each company (Okinawa Prefectural Government 2003:10).

294 The Local Economy: Lessons from Japan

Development of human resources The government of Okinawa prefecture, fully committed to the development of human resources, provides four programs to this end. The ﬁrst program is to train people who are willing to work for call centres. The minimum requirement for attending this training course is that the trainee must be a resident of Okinawa. The second program is a training course for engineers to learn advanced knowledge for information technology. This program is for IT engineers resident in Okinawa. It has two forms. In one, engineers can obtain general advanced knowledge for information technology, whereas in the other they can learn advanced skills for broadcasting or creating content. All training programs are free of charge. The last program is a subsidy for companies, useful for when they are planning to train new employees in other prefectures in Japan or overseas for more than one month. A company interested in it submits an application form and a business plan to the Operation Department of Employment Activation Organization, Okinawa, and receives it when these are accepted. This subsidy does not cover all expenses, but it does pay for transportation and housing costs; the maximum amount is three-quarters of total expenses (Okinawa Prefectural Government 2003:11–13). Tax deduction and tax exemption The last policy implemented by the government of Okinawa covers tax incentives and its purpose is to develop the ICT industry in Okinawa. The government exempts or reduces the corporate tax and local taxes payable by an ICT company that is planning to increase its investments for further development. For tax exemption or tax deduction, an ICT company must invest in buildings or facilities in designated areas in Okinawa. The government nominated 11 cities and 13 villages and towns as Information and Communications Industry Promotion Zones.1 The government reduces or exempts from tax for ﬁve years any ICT company that fulﬁls these conditions during the period 1 April 2002 to 31 March 2007 (Okinawa Prefectural Government 2003:16–17). As noted, the government of Okinawa prefecture has implemented many preferential policies in order to invite ICT companies there from other prefectures. Moreover, in the sub-programs it has implemented, particular emphasis has been placed on vocational training. The logic of this emphasis can be clearly understood when we consider that vocational training is essential for the development of industrial clusters, and that human resources are crucial for businesses in the IT-based economy. It is obvious that as a result of the policies outlined above the number of ICT companies locating in Okinawa is increasing, as shown in Figure 8.1, implying that the ICT industry has started to form an industrial cluster there.2

Rika Nakagawa 295

Lessons from policies of the central and local governments This section draws some lessons from policies of the central and local governments which have been executed to improve the local economic situation. According to Takeuchi (2000), ﬁve factors are needed to stimulate the local economy: competitiveness, infrastructure, human resources, business-to-business (BtoB) networks, and markets. First, companies need to be competitive to survive in the market. Second, sufﬁcient infrastructure needs to be in place in the local area for private companies to do business there. Third, well-educated workers and skilled engineers are essential to activate the local economy. Fourth, business-to-business networks among companies to promote business activities are also crucial. Fifth and last, markets where companies sell their goods and services must be big enough. In terms of the above factors, the precedent case of Okinawa offers some lessons. As has been noted, policies implemented by the government of Okinawa have succeeded so far in inviting ICT companies and enhancing the industrial cluster. Two points, however, are required for further development of the ICT cluster. First, local government should stimulate competition and promote cooperation among companies in the cluster. In the case of Okinawa, the prefectural government played an important role in inviting ICT companies to the area, but did not act as a promoter for competition and cooperation in the cluster. To do this, the local government should become a coordinator to connect the business sector with research institutions so as to facilitate the very important exchange of necessary information between them. As a result of smooth ﬂows of information, it is possible for the industry to enhance competition and cooperation. Given this, we can expect business networks among companies in the industrial cluster to expand. Second, the local government should keep a balance between support for vocational training and that for new business. However, the government of Okinawa, while committing itself to support human resource development, has not paid much attention to support for new businesses or venture businesses. To stimulate the local economy, it is also important to help entrepreneurs start new businesses. Thus, local governments need to maintain such a balance. If we looking at policies announced and implemented by the central government, it is clear that emphasis has been placed upon the importance of networks and R&D activities. This implies that these policies supplement disadvantages of policies implemented by the government of Okinawa. If ICT companies in Okinawa compete and cooperate with one another, the ICT cluster in Okinawa will be able to develop further. In the future, this might have a positive impact on the local economy in terms of increased employment, the development of technology for IT-based industry and the development of the local economy.

296 The Local Economy: Lessons from Japan

With regard to the second suggestion, central government has emphasized the need for entrepreneurship and new businesses and has supported incubator facilities and venture businesses. Furthermore, it is possible that the central government’s policies could work to supplement local government policies; as has been noted above, Okinawa has failed to give real support to entrepreneurs to start new businesses. Given the above, it is obvious that the policies promoted by the Okinawa government were well planned; however, they are insufﬁcient for enhancing competition and cooperation. The central government’s Industrial Cluster Plan is expected to supplement insufﬁcient policy factors in the prefectural government’s plan to further develop the ICT cluster in Okinawa.

Summary This chapter has explained the background to the Japanese government’s attempt to support the local economy by introducing the Industrial Cluster Plan. In addition to this, it has shown how the efforts of Okinawa prefecture acted as a precedent for the plan. The Okinawa government started to attract companies to form an industrial cluster in the prefecture before the central government paid attention to the importance of industrial clusters. The reason why Japan focused on industrial clusters and made the Industrial Cluster Plan in 2001 was that Japanese manufacturing companies were transferring their production facilities overseas. This move abroad had resulted from the impact of the appreciation of the Japanese currency and the ‘China shift.’ The government of Japan sought to form industrial clusters in local areas and to let the clusters be the driving force for strong economic growth. Since 2001, the government has implemented several policies in order to achieve this objective. Owing to those efforts, as of 2003, 24 industrial clusters has been formed in Japan. The case of Okinawa prefecture is one of the precedent industrial clusters in Japan. ICT companies have started to move there and establish new ofﬁces. Before 1998, the total number of ICT ﬁrms in Okinawa was in single digits; however, after 1999, it increased rapidly. As of May 2003, 80 companies had established ofﬁces in Okinawa and were in operation. In order to form the ICT cluster, the government of Okinawa prefecture made a large contribution. It executed ﬁve core programs: (1) support for a cost reduction in telecommunications, (2) development of the ICT industrial cluster, (3) enhancement of job opportunities for young people, (4) development of human resources and (5) tax deduction or tax exemption. In this initiative, the government of Okinawa has been eager to promote vocational training. In terms of lessons garnered from experience in Japan, it is important for local governments to make a plan for forming industrial clusters in a stra-

Rika Nakagawa 297

tegic manner. One effective and efﬁcient way to realize this is to cooperate with the central government. As indicated in the precedent case, the policies of the Okinawa government are not sufﬁcient to develop ICT industries there. The Industrial Cluster Plan of 2001 is well planned to supplement the policies introduced by the Okinawa government. The important thing here is that local governments should not duplicate policies implemented by the central government. Moreover, a local government should look for the deﬁciencies of its programs to be made up for by the policies of the central government. If it is impossible for local governments to beneﬁt from the industrial cluster planning of the central government, they need to consider two types of policies: to invite the targeted industry and to stimulate competition and cooperation within the industry in the industrial cluster. Both strategies are critical to form an industrial cluster in the prefecture, and to developing the cluster in the future.

Notes 1. The designated areas are: Naha City, Ishikawa City, Gushikawa City, Ginowan City, Hirara City, Ishigaki City, Urasoe City, Nago City, Itoman City, Okinawa City, Motobu Town, Yanashiro Town, Katsuren Town, Yomitan Village, Kadena Town, Chatan Town, Kitanakagusuku Village, Nakagusuku Town, Nishihara Town, Tomigusuku City, Kochinda Town, Yonabaru Town, Haebaru Town and Ginoza Village. 2. This chapter does not conclude that the ICT industrial cluster in Okinawa will continue to develop in the future, but emphasizes that call centres and other IT businesses established their ofﬁces in Okinawa after the Okinawa government implemented policies which targeted the ICT industry.

References Cabinet Ofﬁce Home page (http://www.esri.cao.go.jp). Ishikura, Yoko. 2003. ‘Naze Ima Kurasuta Nanoka’ [Why are Industrial Clusters Important?], in Ishikura Yoko, Fujita Masahisa, Maeda Noboru, Kanai Kazuyori and Yamazaki Akira (eds), Nihon no Sangyo Kurasuta Senryaku [Strategy for Cluster Initiatives in Japan]. Tokyo: Yuhikaku (in Japanese). Itami, Hiroyuki. 1998. ‘Sangyo Shuseki no Igi to Ronri’ [Signiﬁcance and Logic of Industrial Agglomeration], in Itami Hiroyuki, Matsushima Shigeru and Kikkawa Takeo (eds), Sangyo Shuseki no Honshitu [Essence of Industrial Agglomeration] Tokyo: Yuhikaku. (in Japanese). Japan External Trade and Organization (JETRO). 2001. Chugoku Kaihatsuku Annai [Booklet of China’s Industrial District], Tokyo: JETRO (in Japanese). Matsushima, Shigeru. 1998. ‘Atarashii Chusho Kigyo Ron’ [Industrial Clusters for Small and Medium Sized Enterprises], in Itami Hiroyuki, Matsushima Shigeru and Kikkawa Takeo (eds), Sangyo Shuseki no Honshitu [Essence of Industrial Agglomeration]. Tokyo: Yuhikaku (in Japanese). Ministry of Economy, Trade and Industry. 1987. Kogyo Tokei Hyo [Census of Manufactures]. Tokyo (in Japanese).

298 The Local Economy: Lessons from Japan Ministry of Economy, Trade and Industry. 1992. Kogyo Tokei Hyo [Census of Manufactures]. Tokyo (in Japanese). Ministry of Economy, Trade and Industry. 1997. Kogyo Tokei Hyo [Census of Manufactures]. Tokyo (in Japanese). Ministry of Economy, Trade and Industry home page (http://www.meti.go.jp). Okinawa Prefectural Government. 2003. Information and Communications Industry Location Guide: Support System for the Information and Communications Industry in Okinawa. Naha: Okinawa Prefectural Government. Takeuchi, Sawako. 2000. ‘Chiiki no Koa Kompitansu’ [Core Competence of the Local Economy: from Local to Global], JETRO Sensa, November, vol. 50, No. 600: 11–13 (in Japanese).

9 Japan: Remedies to Activate Local Cities Takuo Imagawa

Introduction The IT1 revolution in Japan has emerged since the second half of the 1990s, and economic activities and lifestyles using information apparatus have been established. According to the ‘2001 edition of Communication Usage Trend Survey’ from the Ministry of Internal Affairs and Communication,2 the number of internet users at the end of 2001 was 56 million, and the population coverage 44 per cent. As those numbers in 1996 were 12 million and 9.2 per cent, respectively, it turns out that the internet has spread rapidly in recent years. The IT industry3 dealing with software, information processing, internet, and so on has received more attention, too; these services have become important in both quality and quantity under the development of communication network. Since the IT industry is easy to enter compared with others, many venture businesses have been born in this ﬁeld. Moreover, it has produced various useful services as well as employment. According to the ‘2001 survey of selected service industries’ from the Ministry of Economy, Trade and Industry, the annual sales of information service businesses4 in that year was about ¥13.6tn, the number of workers 565,000 and the annual sales per worker ¥24m. These numbers increased from 6 years previously by ¥4tn, 110 thousand people and ¥5m., respectively, which had a big inﬂuence on Japan’s economy. Although it is usually thought that decentralization would have been accelerated by development of information and communication technology in recent years, there is, in fact, a tendency towards the accelerated concentration of economic activities and population, which is pointed out as a ‘paradox of accumulation’ (Imagawa 2002). This can be applied also to the IT industry. Approximately 30 per cent of establishments as of March 2002 were located in Tokyo, with a clear tendency towards accumulation in cities. Also, in the USA the central parts of cities that tended to decline were reinvigorated by the location of the IT industry (Konagaya 1999a, 299

300 Japan: Remedies to Activate Local Cities

1999b). There are many local governments in Japan that aim at activating the regional economy by inviting high-growth IT businesses and having them accumulate. However, their efforts have not necessarily reached the stage of realizing the growth of regional economies. In the USA, Silicon Valley, Silicon Alley and Multimedia Gulch are typically cited as examples of the accumulation of the IT industry5 (Yukawa 1999a, 1999b). In Japan, there are Bit Valley in Tokyo and Sapporo Valley (the Development Bank of Japan, Hokkaido branch 2000). In other Asian countries, the Multimedia Super Corridor in Malaysia, Bangalore in India and Zhongguancun in China are well-known clusters of IT businesses.6 There has been academic research that has analysed the location factors of the IT industry focusing on one of those clusters, but no nationwide study that has incorporated all clusters within an entire country. Accordingly, this chapter tries to perform such a nationwide analysis of location factors using well-prepared location data for the IT industry in Japan, on a national scale. In order to realize effective local government policies from now on, it is necessary to analyse and understand how high-tech IT ﬁrms select their location so as to promote growth in regional economies. In this chapter, I try to discern the policy tools that will enhance the growth of the IT industry and also activate local economies by clarifying the location factors of IT businesses in Japan. The remainder of the chapter consists of four sections. The next section introduces the characteristics of the IT industry and predicts the accumulation factors from the existing literature of urban economics. The second section explains the data and the estimation method. The third section performs a formal econometric estimate of the accumulation factors of the IT industry by utilizing the national-scale data of the survey on software-based IT industries. The fourth section concludes and discusses the factors that are important for local governments to consider in order to attract the IT industry.

Accumulation of the IT industry Characteristics of the IT industry The IT industry is a knowledge-intensive industry, that is, it depends strongly on knowledge and information compared with other industries. Since it is hardly dependent on natural physical resources, its location is determined by the use of a highly skilled labour force. With a pooled labour market present, ﬁrms are able to increase their employment of highskilled labour, and the IT industry thereby to realize higher growth (Kolko 2002). For the IT industry, the division of work between companies and the cooperation of various talents are important. When these can be realized within a certain area such as a city, adjustment costs will fall dramatically and the competitiveness of IT ﬁrms will be improved (Yukawa 1999a).

Takuo Imagawa 301

Flexible organization and face-to-face communication and/or collaboration are very important to the IT industry and the specialists within it. They make use of informal networks, in which for example information is passed by word of mouth (Yukawa 1999b). Therefore, it is useful for players in the IT industry to locate close to each other. On the other hand, it may also be possible that the IT industry has no incentive to accumulate, since information technology allows for a dramatic decrease in transaction costs given that these costs are unimportant for this industry whose products are in many cases intangible (Kolko 2002). If transportation costs are very low, ﬁrms are likely to move to suburbs since increasing returns can be pursued owing to the availability of cheaper land prices. There are then fewer incentives for ﬁrms to locate near the centre of demand, and the degree of industrial accumulation should decrease. In particular, in an industry whose products are intangible assets such as service industries, it may be possible to substitute transportation by electronic means, which allows for a dramatic reduction of transportation costs. In addition, thanks to internet technology, we are able to beneﬁt from the externality of knowledge and information spillovers even at a long distance. It is possible for us to communicate very cheaply with many counterparts simultaneously. Even documents or moving pictures can be transmitted. It is also possible to ﬁnd a brand-new communication partner to share unexpected ideas (Kolko 2002). Thus, although inconsistencies may arise with the impact of information and communication technology on industrial accumulation, this is solvable by considering the characteristics of innovation activities in the IT industry. IT industry and innovation For the IT industry, innovation activities are important and the knowledge spillover and sharing of information become indispensable in those activities. In particular, ‘embodied information’ is indispensable for the creation of knowledge (Imagawa 2002). Embodied information such as experience is stockpiled by individuals, and is difﬁcult to transfer. Thus, meeting face to face plays an important role in the sharing of this type of information. An industrial innovation depends on the resources of the fundamental scientiﬁc knowledge of universities, research institutes and so on, and ﬁrms therefore concentrate their own resources of related research into areas where universities and research institutes are easy to access (Baptista and Swann 1998). This geographical proximity to universities and so on allows direct access to individuals who have new and useful information. Geographical accumulation is the most important factor in realizing organizational improvements or technical innovation. By concentrating a variety of specialized knowledge into a speciﬁc area, information exchange will be performed informally, and specialized human capital can be positively

302 Japan: Remedies to Activate Local Cities

affected (Baptista and Swann 1998). In fact, it can be conﬁrmed that computer companies in the early stages in the USA or Europe were located close to universities, which promoted dense communication between researchers and engineers in the semiconductor industry. As a result, information sharing within such areas advanced, accelerating innovation (Swann and Prevezer 1996). Moreover, Kolko (2000) showed that existence of a locally high level of skill and a university combined to create a positive effect on the development the US commercial internet. The IT industry tends to the heavy employment of highly skilled labour, and to obtain beneﬁts from industrial accumulation, which generates a virtuous circle of accumulation (Kolko 2002). The hardware industry (producing for example computers) has a strong tendency to geographical accumulation (Saxenian 1994), which by similar logic may also be applied to the software industry. Although information transaction costs have decreased dramatically and the factor of distance has become far less signiﬁcant, location remains the most crucial strategic factor for ﬁrms’ innovation, production and marketing (Beardsell and Henderson 1999). Feldman and Audretsch (1999) propose that the transmission costs of knowledge are still very high and that this aspect will be the most important factor for the IT industry. Accumulation factors from the existing literature The above considerations imply that the IT industry tends strongly towards accumulation. In this subsection, I therefore review the factors inﬂuencing accumulation that have already been pointed out in the existing academic literature. Kelly (1987) showed that major factors to explain the location of the IT industry7 are: (1) the existence of a link with universities, (2) the availability of a highly skilled labour force, (3) venture capital and (4) government policy. Yukawa (1999a, 1999b) carried out research on the globally famous IT clusters in Silicon Alley in New York and Multimedia Gulch in San Francisco and showed that the major factors of agglomeration are: (1) social amenities for young men, (2) cheap and convenient space or land, (3) the existence of creative workers, (4) related educational facilities and (5) the existence of industries which function as clients of the IT industry (the accumulation effects). Kinukawa and Yukawa (2001) performed research on the Tokyo area and concluded that social amenities, especially fashionable facilities8 which are popular among young workers, are vitally important at the initial stage of accumulation, while the existence of population and businesses (the accumulation effect) is inﬂuential later on. Thus, it can be said that accumulation tends to cause further accumulation. On the other hand, there is no signiﬁcant result for rent, which implies that cost of land or ofﬁce itself is not necessarily a decisive factor in the location of an IT enterprise.

Takuo Imagawa 303

For a systematic survey of the IT industry in Japan, the survey on software-based IT industries by the Ministry of Land, Infrastructure and Transport is available, with ﬁrms’ location one of the main topics in it. According to the questionnaire result in the survey, important factors for location are reasonable rent, good access to a nearby railway station and so on. Moreover, younger companies tend to consider a good communication environment as important. On the other hand, access to special schools, a local government policy targeting of IT ﬁrms and convenient ﬁnancing are not considered as important. While the need for high-skilled workers is very high and those workers are supplied in many cases from a neigbouring area, recruiting by IT firms in special schools (or vocational schools) remains very low. Accumulation factors to be employed in this research From the above, I believe that the most important factor for the location of the IT industry is the availability of a highly skilled labour force. This is consistent with the fact that the IT industry requires knowledge and information for its innovation rather than resources such as physical capital and land. To verify this, I would like to use educational facilities such as universities, colleges, and special schools to see how highly skilled workers are linked with IT ﬁrms’ location. Innovation activities in IT or venture businesses are usually dependent on research and development in academic institutions such as universities, colleges and special schools. In particular, their departments of science, engineering, medicine and so on are the major suppliers of potential workers in this high-tech industry. Therefore, in the estimation framework of the following sections, the existence of educational facilities and the ease of access to those will be incorporated as a proxy of highly skilled labour into a ﬁrm’s selection of a location. In addition to this, I need to include several factors to analyse the accumulation of the IT industry. First of all, population size can be considered to express the demand for IT services by consumers. Therefore, I include population density (normalized by habitable land area) in my estimation framework. High population density implies dense consumer demand, and various IT services will be produced and consumed. Also, under high density, other advantages may be expected such as easier employment or a variety of workers. Second, the proﬁle of the existing industry determines the company demand and the supply chain. Since IT services can be used by both consumers and companies, and since IT ﬁrms also consume products and services of the existing industry as intermediate inputs, the location of the IT industry depends on the size and structure of the existing industry. In addition, there will be cooperation or competition among the industries, mobility of workers in similar industries, regional economic growth and so on,

304 Japan: Remedies to Activate Local Cities

all of which should come into the decision-making process of location selection. Thus, new companies tend to be located where industry has already accumulated (the accumulation effect). In my estimation, this effect of the existing industry has to be controlled. Third, I include social amenities as an important factor for IT ﬁrms’ location. Since the supply of talented high-skilled workers is the key in this high-tech industry, they demand dense information exchange between them, which brings about positive information and knowledge spillovers. As Kinukawa and Yukawa (2001) pointed out, a good quality of social amenities around the workplace may also be related to location as a stimulator of face-to-face communication. Since Kinukawa and Yukawa (2001) checked only the limited area of the centre of Tokyo, it will be meaningful to develop this into a nationwide analysis. Fourth, political or institutional attempts aimed at inviting the IT industry, which many local governments have already enforced, can be included since it would be very interesting to see whether this external intervention is effective or not on a national scale. Finally, it may be reasonable to include rent as one of the explanatory elements if I consider the questionnaire result of the survey on softwarebased IT industries, which tells us that reasonable rent is one of the important factors for the location selection of IT ﬁrms. However, since rent is linked highly with density of population and existing industry, and the appropriate data in the form corresponding to my data set are not available, I excluded rent from my estimation in the following sections. Thus, as explanatory factors for location selection of the IT industry in Japan, I will pick up ‘educational facilities (a proxy for a high-skilled labour force)’, ‘population density’, ‘existing industry’, ‘social amenities’ and ‘policy.’ The next section explains the data set employed for explaining the location of the IT industry and determining the factors of accumulation.

Data and estimation method IT industry As location data for the IT industry, the data from the survey on SoftwareBased IT Industries which the Ministry of Land, Infrastructure and Transport produces every six months are available, and I used the data as of March 2002.9 This data was constructed from a survey on NTT (Nippon Telegraph and Telephone Corporation)’s town-page database. The IT industry is registered10 into one of three types: ‘software business’, ‘information processing’ and ‘internet’. Each type has the following characteristics in general. A ‘software business’ performs the development of software, application, programming, package and so on and also the agent business of those products. ‘Information processing’ includes accepting calculation service, data processing, human support and so on. ‘Internet’ includes internet access and/or connection services such as internet providers.

Takuo Imagawa 305

As of March 2002, taking the IT industry as a whole, there were 35,785 establishments throughout Japan. Of these, 34,248 companies exist in a city or a ward, which means this is a very urban-type industry. Population density Population density is a factor related to the size of consumer demand. Population data were taken from the 2000 population census and used the daytime population as of October 2000. Since there will be a fair amount of population movement between cities if residential and commercial areas are different, it is more appropriate to use daytime population for the analysis of industrial accumulation. Moreover, since mountains or lakes are considered to be non-residential areas, I used the habitable land area to calculate population density. Habitable land areas are taken from Toyo Keizai Regional data 1990, assuming that there is only a small change after 1990. Thus, population density is computed from daytime population and habitable land area. Existing industries Existing industries are considered to increase demand for corporate services from the IT industry, and also to be a measure of economies of scale or industrial diversity as an element of accumulation. To indicate existing industries, I decided to use the number of establishments. Since the data for the IT industry is taken from March 2002, the data of existing industries should be taken from the data before 2002. Thus, the numbers are extracted from those of October 2001, which can be found in the 2001 establishments and enterprise census of the Ministry of Internal Affairs and Communications. Skilled labour force Educational facilities are used as a variable to measure the skilled labour force. Potentially skilled workers are most likely to exist near high-level educational facilities, and existing research also shows that industrial accumulation occurs near universities. Speciﬁcally, ‘educational facilities’ mean the total of universities, graduate schools, junior colleges, technical colleges and special schools. Furthermore, as for universities and graduate schools, I divide them into departments of the arts (law, business, economics, literature, language, policy and so on) and those of the sciences (sciences, engineering, medicine, agriculture and so on).11 It may be possible that students studying at science departments rather than arts departments are in demand by an IT industry that is very technology-oriented. Moreover, there will be collaboration between IT enterprises and academic laboratories in research, development and consulting. In this aspect, it will be very useful to distinguish between arts students and science students. ‘Educational facilities’ are taken from ‘the 2001 national school census’12 and the data extracted were as of May 2000 since these have also to be

306 Japan: Remedies to Activate Local Cities

before 2002. Furthermore, I incorporate the concept of ‘proximity’ into ‘educational facilities’, assuming that the inﬂuence of knowledge or information spillovers will be greatly affected by distance. In this research, when educational facilities exist within the same city, this is expressed as ‘within’. When educational facilities exist in adjacent cities, but not within the same city, this is expressed as ‘adjacent’. When educational facilities exist within a 50-km radius, but not within the same city or in adjacent cities, this is expressed as ‘50 km’. Regarding ‘adjacent,’ the data were constructed from the map of the 2000 edition of city data pack by Toyo Keizai. Regarding ‘50 km’, I extracted both latitude and longitude13 on the basis of the address of each city ofﬁce for every city in Japan, and calculated the distance between every pair of cities. From this database, cities within the sphere of a 50 km radius are extracted, and the number of educational facilities that exist there was added. Social amenities I used the number of restaurants as a proxy for social amenities. According to Kinukawa and Yukawa (2001), fashionable places where a comparatively small number of people with common taste get together have an explanatory power for location of the IT ﬁrms. Restaurants are also representative samples of social gatherings of workers in order to exchange information or perform informal business meetings. Thus, I roughly assume that the number of places for social amenities increases in proportion to the number of restaurants. Since these data also have to come from before 2002, I used the number of general restaurants as of 1 July 1999 extracted from the 1999 Establishment and Enterprise Census. Policy Government policy intervention may also be an important element to attract IT industry enterprises. Using the data from the survey on softwarebased IT industries performed by the Ministry of Land, Infrastructure and Transport, I introduced a dummy variable which takes a value of 0 when there is no information policy14 at the city level, with a value of 1 when information policy is shortly described in general city planning and a value of 2 when speciﬁc plans or measures for information policy have already been introduced. Estimation model Utilizing the above variables,15 I performed econometric estimations using city data. Also, I tried to go further by experimenting to divide the IT industry into three types – software business, information-processing and internet. First of all, I tried to test a hypothesis that the location of the IT industry is inﬂuenced by population density, existing industries, skilled labour force (educational institutions), social amenities and policy (the accumulation

Takuo Imagawa 307

factor analysis). Second, by focusing on the educational facilities that are considered to supply the skilled labour force, I tried to test how and which educational facilities affect the location of the IT industry (the educational facilities analysis). Third, I tried to analyse how much proximity to educational facilities affects the location of the IT industry (the proximity analysis). The estimation model is described as follows. In this framework, we have to be very careful that establishments and restaurants may be endogenous with IT industry enterprises, because accumulation itself is an autonomous and circulative process. Then, in order to control for this endogeneity I relied on an instrumental variable estimation method. As instrumental variables, I used the number of establishments in 1996, and the number of restaurants in 1996, assuming that the accumulation factors in 1996 have only a small power to invite further accumulation, since the IT revolution in Japan seems to have started in the late 1990s. Finally, statistics were White-corrected because there may be heteroscadascity in this estimation framework: IT industry software business information processing internet

= β0 + β1• density + β2• establishments + β3• education + β4• restaurants + β5• policy + ε

Here: IT industry = number of establishments of IT businesses per population of 10,000; software business = number of establishments in software businesses per population of 10,000; information processing = number of establishments in digital informationprocessing services per population of 10,000; internet = number of establishments in internet-related services per population of 10,000; density = population density (daytime population divided by habitable land area); establishments = number of establishments per population of 10,000; education:16 (i) educational facilities = number of educational facilities (per popula tion of 10,000) (ii) univ-art = number of universities and graduate schools with arts departments (per population of 10,000) univ-sci = number of universities and graduate schools with science departments (per population of 10,000) junior = number of junior colleges and technical colleges (per popu lation of 10,000) special = number of special schools (per population of 10,000) (iii) within = number of the corresponding educational facilites within the same city

308 Japan: Remedies to Activate Local Cities

adjacent = number of the corresponding educational facilities in adjacent cities 50 km = number of the corresponding educational facilities within a 50 km radius excluding those within the same city and adjacent cities restaurants = number of restaurants per population of 10,000 policy = dummy variable (2 = information policy in enforcement, 1 = information policy described in general city planning, 0 = no information policy)

Estimation result Basic model Summary statistics of the data employed in this estimation are shown in Table 9.1. First, I tried to analyse whether ﬁve explanatory variables affect the accumulation of the IT industry. Estimation results are shown in Table 9.2. Coefﬁcients of density, establishments and educational facilities are all signiﬁcantly positive. However, those of restaurants are not signiﬁcant. Thus it can be said that the accumulation of the IT industry has a close relationship with existing concentrations of population and business. Table 9.1 IT industry accumulation in Japan: summary statistics (city data) N IT industry software business information processing internet density establishments education al facilities restaurants policy univ-art/within univ-sci/within junior/within special/within univ-art/adjacent univ-sci/adjacent junior/adjacent special/adjacent univ-art/50 km univ-sci/50 km junior/50 km special/50 km

673 673 673 673 673 673 673 673 673 673 673 673 673 673 673 673 673 673 673 673 673

Mean 1.53 0.70 0.48 0.35 2,176.19 515.89 0.35 33.07 0.55 0.05 0.03 0.05 0.22 0.67 0.50 0.43 2.54 5.35 3.59 3.30 17.28

Max. 8.87 5.30 2.44 2.23 18,153.11 970.00 1.66 90.06 2.00 0.98 0.91 0.50 1.11 38.19 30.51 13.93 94.65 48.65 34.65 28.46 159.02

Min. 0.00 0.00 0.00 0.00 96.33 262.34 0.00 3.85 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

S.D. 1.24 0.69 0.44 0.32 2,294.84 104.55 0.29 9.58 0.68 0.11 0.09 0.08 0.21 2.72 2.18 1.07 7.12 8.52 5.97 4.75 25.80

Takuo Imagawa 309 Table 9.2

Estimation results of IT industry accumulation in Japan: basic model (accumulation factor analysis) Dependent variable: number of establishments in the IT industry per capita (1)

(2)

(3)

(4)

density

0.0002 *** (7.013)

0.0002 *** (6.876)

0.0002 *** (6.406)

0.0002 *** (5.958)

establishments

0.002 *** (2.624)

0.002 (3.269)

***

0.002 (2.597)

***

0.001 ** (2.496)

1.342 (7.629)

***

1.342 (7.563)

***

1.170 *** (7.055)

educational facilities restaurants

–0.00005 (–0.009)

policy constant adjusted R2 observations

0.003 (0.53) 0.409 *** (6.165)

0.397 (1.435)

–0.22 (–0.767)

–0.22 (–0.764)

–0.359 (–1.280)

0.162

0.263

0.262

0.309

673

673

673

673

T-statistics are in parentheses, and are all White-adjusted. ***Signiﬁcant by 1%. **Signiﬁcant by 5%. *Signiﬁcant by 10%.

When a policy variable is added, its coefﬁcient is signiﬁcantly positive. A city’s efforts to attract IT ﬁrms are then considered to be effective. However, since it is evaluated only by the existence of information policy at city level, it is not possible to propose what policy it is appropriate to introduce. Education and proximity analysis Second, I tried to divide educational facilities into different levels to be included in the regression. Estimation results are shown in column (5) of Table 9.3. Coefﬁcients of universities and graduate schools are signiﬁcant and large only for science departments (univ-sci/within), and those of special schools (special/within) are also signiﬁcant. On the other hand, junior colleges and technical colleges (junior/within) are not inﬂuential. It is therefore understood that science departments of universities and graduate schools as well as special schools are having a strong inﬂuence on the location of IT industry. Third, I tried to analyse how important proximity to educational facilities is for ﬁrms’ decision-making on location. Estimation results are shown in column (6) of Table 9.3. In this estimation model, I introduced interac-

310 Japan: Remedies to Activate Local Cities Table 9.3

Estimation results of IT industry accumulation in Japan: inﬂuences of education facilities (educational facilities and proximity analysis) Dependent variable: number of establishments in the IT industry per capita (5)

(6)

density

0.0002 (6.804)

***

0.0002 (7.55)

***

establishments

0.001 (2.834)

***

0.001 (1.327)

***

policy

0.387 (5.886)

***

0.312 (4.660)

***

univ-art/within

0.284 (0.637)

univ-sci/within

1.194 (2.184)

**

1.148 (2.192)

**

junior/within

0.639 (1.116)

special/within

1.587 (7.484)

***

1.369 (6.771)

***

univ-sci/adjacent

0.153 (3.808)

***

special/adjacent

–0.063 (–5.215)

***

univ-sci/50 km

0.047 (1.691)

special/50 km

–0.018 (–3.031)

constant adjusted R2 observations

–0.241 (–0.862)

0.427 (1.575)

0.323

0.363

673

673

* ***

T-statistics are in parentheses, and are all White-adjusted. ***Signiﬁcant by 1%. **Signiﬁcant by 5%. *Signiﬁcant by 10%.

tion variables of education facilities and proximity. Here, as education facilities, I use (a) science departments of universities and graduate schools, and (b) special schools, both of which were evaluated as inﬂuential in column (5). As proximity, I use (i) within the same city (within), (ii) in adjacent cities (adjacent), and (iii) within a 50 km radius (50 km). Therefore, six variables in total can be introduced as interaction of education and proximity.

Takuo Imagawa 311

The results tell us that science departments of universities and graduate schools are signiﬁcantly powerful, even for ‘adjacent’ (univ-sci/adjacent) and ‘within 50 km (univ-sci/50 km),’ which implies that positive spillovers from academic institutions are active over city borders. However, coefﬁcients for special schools are signiﬁcantly positive only for ‘within the same city (special/within),’ and those for ‘adjacent’ (special/adjacent) and ‘50km’ (special/50 km) are even negative. Differences among industrial classiﬁcations of IT industry Tables 9.4 and 9.5 show the estimation results when the IT industry is divided into three types. In the accumulation factor analysis of the tables, it is understood that coefﬁcients of population density, establishments, educational facilities and policy are all signiﬁcantly positive, although those of restaurants are not signiﬁcant. However, in the educational facilities analysis, coefﬁcients of science departments (univ-sci/within) are signiﬁcantly positive for software business and information-processing, although those for the internet alone are not signiﬁcant. As for the proximity analysis, the results are very similar for software business and informaTable 9.4

Industrial classiﬁcations of the IT industry in Japan – 1: accumulation factor analysis

Software business

Information processing

(1)

(2)

Internet (3)

density

0.00006 *** (6.320)

0.00006 *** (6.320)

0.00001 ** (2.071)

establishments

0.0004 (1.927)

**

0.0004 (1.927)

**

0.0004 (2.309)

**

educational facilities

0.336 (5.622)

***

0.336 (5.622)

***

0.249 (5.500)

***

restaurants

0.002 (1.170)

policy

0.123 (5.161)

constant

0.002 (1.170) ***

0.123 (5.161)

0.002 (1.427) ***

0.097 (5.206)

–0.121 (–1.211)

–0.121 (–1.211)

–0.097 (–1.167)

adjusted R2

0.225

0.225

0.14

observations

673

673

673

T-statistics are in parentheses, and are all White-adjusted. ***Signiﬁcant by 1%. **Signiﬁcant by 5%. *Signiﬁcant by 10%.

***

312

Table 9.5

Industrial classiﬁcations of the IT industry in Japan – 2: educational facilities and proximity analysis Software business (4)

Information processing

(5)

density

0.0001 (6.286)

***

0.0001 (6.631)

establishmentss

0.0004 (1.662)

*

policy

0.183 (4.797)

***

univ-art/within

0.156 (0.697)

univ-sci/within

0.7 (2.639)

junior/within

0.376 (1.26)

special/within

0.730 (6.681)

(6) ***

Internet

(7)

0.0001 (7.588)

***

0.0001 (8.151)

0.0001 (0.327)

0.0005 (2.462)

**

0.0002 (1.298)

0.149 (3.818)

0.115 (4.896)

***

0.084 (3.572)

(8) ***

***

0.089 (0.537) ***

0.706 (2.869)

***

0.395 (1.824)

0.623 (5.939)

***

univ-sci/adjacent

0.094 (3.796)

special/adjacent univ-sci/50 km

0.491 (6.227)

0.00002 (3.393)

***

0.00003 (5.192)

***

0.0004 (2.982)

***

0.0003 (1.959)

**

0.09 (4.819)

***

0.08 (4.109)

***

0.012 (0.104) *

0.345 (1.672)

*

0.024 (0.119) ***

(9)

0.123 (0.936)

0.128 (0.963) 0.233 (1.279)

***

0.413 (5.429)

***

***

0.042 (2.775)

–0.036 (–4.744)

***

0.029 (1.765)

*

0.367 (5.803)

***

0.33 (5.283)

***

***

0.022 (2.112)

**

–0.017 (–3.753)

***

–0.012 (–3.829)

***

0.024 (2.582)

***

–0.010 (–0.977)

Table 9.5

Industrial classiﬁcations of the IT industry in Japan – 2: educational facilities and proximity analysis Software business (4)

special/50 km constant

Information processing

(5) –0.01 (–2.985)

(6) ***

Internet

(7) –0.009 (–4.104)

continued

(8) ***

(9) 0.058 (0.392)

–0.099 –0.614

0.237 (1.485)

–0.084 –0.821

0.131 (1.309)

–0.063 (–0.773)

0.058 (0.680)

adjusted R2

0.297

0.327

0.239

0.280

0.149

0.164

observations

670

670

672

672

672

672

T-statistics are in parentheses, and are all White-adjusted. ***Signiﬁcant by 1%. **Signiﬁcant by 5%. *Signiﬁcant by 10%.

313

314 Japan: Remedies to Activate Local Cities

tion processing, that is, as with the whole IT industry, the spillover effects of science departments are inﬂuential even for adjacent cities (univ-sci/ adjacent) and within a 50km radius (univ-sci/50 km), but those of special schools are positive only for ‘within the same city (special/within).’ On the other hand, as for the category of the internet, the coefﬁcients of science departments are very different and signiﬁcantly positive only for adjacent cities (univ-sci/adjacent). Thus, difference can be found only for the internet among other types of the IT industry. Implications from estimation results Estimation results for the above three sub-sections (pp. 308 ff.) conﬁrm that population density has a strong explanatory power and that the IT industry is likely to be located in an area with many existing establishments. That is, an accumulation of the IT industry will be observed in an area that has been functioning already as an urban centre of population and businesses. Educational facilities are also signiﬁcantly positive, and especially the science departments of universities and graduate schools, as well as special schools will affect the location of the IT industry. The coefﬁcients of science departments are very large, and even science departments in adjacent cities or within a 50 km radius have explanatory powers, which implies that the scientiﬁc functions (such as natural sciences, engineering, medicine and so on) of universities and graduate schools plays a very important role as an accumulation factor for the IT industry. Although the coefﬁcients of ‘adjacent’ and ‘50 km’ are both signiﬁcantly positive, the magnitude of the coefﬁcient is bigger in the order of ‘within’, ‘adjacent’, and ‘50 km’, which proves that the inﬂuence of science departments depends almost linearly on distance, becoming larger as the distance is shorter. On the other hand, the coefﬁcients of special schools are signiﬁcantly positive for ‘within’, but signiﬁcantly negative for ‘adjacent’ and ‘50 km’. Since these negative impacts are very small, the location effects of special schools are limited to a small area, namely ‘within the same city’, and spillover effects do not cross city borders. Interestingly, arts departments such as law, business, economics, literature and so on of university and graduate schools lack any explanatory power at all for the location of the IT industry, as do junior colleges and technical colleges. Although the existing related literature tells us that educational facilities such as universities tend to have a strong inﬂuence on IT industry location, and to stimulate the innovation and development of IT ﬁrms by being located closer, my research implies that those inﬂuences will come from departments such as natural sciences, engineering and medicine and so on, which have a close relationship with information technology. The questionnaire result of the 2002 edition of the survey on software-based IT industries shows that there is a strong demand for highly talented people in the IT industry.

Takuo Imagawa 315

Almost 70 per cent of the IT ﬁrms answer that they face difﬁculty in employing high-skilled workers, even though they suceed in hiring the necessary number of workers. The estimation results indicate that this demand is in general expected to be for students specializing in scientiﬁc subjects rather in the arts. Regarding restaurants, it turns out that these are not an important factor in the location of the IT industry. In our research, they are a proxy for social amenities. It is possible that social amenities for young workers to have fun or to exchange information are not as important as educational facilities where the highly skilled labour force can easily get knowledge spillovers. Since only restaurants were selected as a proxy for social amenities in this research, it is not appropriate to generalize this result into every kind of social amenity. At any rate, restaurants are not that important contrary to what is said in the earlier literature such as Kinukawa and Yukawa (2001). As for the policy effects, it was proved that a city’s policies relevant to the IT industry that have been introduced already have a positive effect on accumulation. Unfortunately, I cannot state what policy is concretely suitable, since this analysis has been made based on the existence of information policy and not on its type. Future research should include the analysis of those policies by content, and distinguish the better policies from those that are meaningless. When divided into types of IT industry, the estimation results for ‘internet’ differ slightly from the others. The coefﬁcients of the science departments of universities and graduate schools for ‘within’ and ‘50 km’ are not signiﬁcantly positive only for ‘internet’. It is therefore possible that the ‘internet’ industrial type can be located in a place further from the central part of a city than can other types of IT industry. The ‘internet’ type receives positive spillover effects mostly from educational facilities in adjacent cities instead of within the same city. Also, the coefﬁcients of density for ‘internet’ are much smaller than those for other industry types. This is not easy to understand. A possible explanation is that this type of IT industry may avoid being located in the centre of cities. The ‘internet’ type provides a connection for internet services seen typically as internet providers, and needs spacious ofﬁces to hold and manage high-tech machines such as industrial-use servers and so on Therefore, location costs, including land rent, will be reduced by being located in an area somewhat distant from the city centre. If this is true, managerial decision-making should take account of deep concern about trade-offs between proximity to talented young engineers or collaboration with other industries and availability of inexpensive, spacious ofﬁces. Moreover, while other types of software business and information processing incorporate frequent communication with customers in order to develop their marketing and services, the ‘internet’ type may be based on independent and infrequent innovations, which do not

316 Japan: Remedies to Activate Local Cities

rely on daily contacts with the science departments of universities and graduate schools.

Conclusion This chapter has performed a nationwide statistical analysis of location factors of the IT industry in Japan, utilizing well-prepared data on a national scale. Estimation results show clearly that the IT industry tends to accumulate in an area where there are already many people, many business establishments and many educational facilities. Availability of or easy access to these is crucially important. In particular, the effects of educational facilities are more important than other urban functions. While other industries determine their location depending on transaction costs and land rent, the IT industry tends to regard access to educational facilities as an important location factor. This will be a unique feature of knowledgeintensive sectors such as the IT industry, which relies heavily on knowledge workers. Therefore, if local governments wish to attract the IT industry, they should have educational facilities as the minimum requirement. Moreover, as expected or desirable urban elements, they should have universities or graduate schools which include departments of natural science, engineering, medicine and so on. Research institutions or science parks are also important in the same context. Representative merits from these educational functions are access to a pool of high-skilled labour (for example university graduates) and cheaper costs to realize cooperation and share information with prominent researchers (professors and others). Although to policy-makers education and research may seem slow and indirect remedies for supplying well-trained graduates in future, they also incorporate short-term industrial effects that can be a magnet for high-tech ﬁrms. However, it is very important to have education or research facilities such as universities or graduate schools within the city centre, not on some fardistant green hill. Access is a crucial factor in knowledge or information spillover, and industrial policy should reﬂect this. The existing literature in this research ﬁeld including other chapters in this book has focused on Asian IT clusters such as the Multimedia Super Corridor (MSC) in Malaysia, Bangalore in India and Zhongguancun in China. There are in fact universities of science and technology in each cluster: Multimedia University in MSC, the Indian Institute of Science (IISc) in Bangalore and Beijing University and Qinghua University in Zhongguancun. High-level research institutions or science parks are also found in these areas. Although this research is based on Japan’s data, I ﬁrmly believe that similar results will emerge if I apply the same framework to these Asian clusters. The ﬁndings of this chapter can be also applied to Okinawa, which is discussed in Chapter 8. Okinawa lacks top-level education or research institutions in its ‘government-oriented’ cluster and this is pointed out as a

Takuo Imagawa 317

drawback as regards labour supply to the local IT industry. Accordingly, Japan’s central government plans to establish a world-level graduate school of science and technology in Okinawa by 2007, and this will be a driving force to enhance the further accumulation of IT businesses in Okinawa.

Notes 1. As a word to describe digital technology used in information processing and communication and so on, IT (information technology) has been often used in Japan as well as in USA and Korea. However, ICT (information and communication technology), which is used more widely in other countries and international institutions, is now prevailing in Japan as a word to express the importance of communication more appropriately, facing the next generation ‘ubiquitous network societies’. 2. http://www.johotsusintokei.soumu.go.jp/ 3. See p. 304 for deﬁnitions. 4. Information service businesses here include software, information-processing services information-providing services, and others. 5. Yukawa (1999a, 1999b) uses the term ‘contents industry’ instead of IT industry. 6. Discussions of these clusters are found in Chapters 6 and 7. 7. Kelly (1987) uses the term ‘high-tech industry’. 8. These are places where young people get together for drinking, joining a music concert, and so on. Kinukawa and Yukawa (2001) call such a place a ‘sub-culture type amenity’. 9. Local municipal mergers before March, 2002 are taken into account. 10. See Ministry of Land, Infrastructure and Transport (2002) for details. 11. When it is difﬁcult to judge from the name of a department, I checked its web page, and judged the clariﬁcation from its curriculum. 12. The number of schools is counted based on the number of campuses. This can be double-counted when the campus is located in more than one city. The distinction between sciences and arts is also based on the number of campuses and they are double-counted when there are departments of both sciences and arts on a single campus. Therefore, the sum of the number of schools used in this estimation can be bigger than the real number. 13. Latitude and longitude are taken from the ‘address matching service’ of the University of Tokyo (http://fujieda.csis.u-tokyo.ac.jp/cgi-bin/geocode.cgi). 14. This includes ‘informatization policies’ to activate local cities by introducing information-related measures. 15. All variables except ‘density’ and ‘policy’ are divided by population. 16. For the basic model (p. 309, Table 9.2), the number of educational facilities (i) is used. For the educational facilities analysis (p. 310, Table 9.3, column (5)), the number of educational facilities divided into four types (ii) is used. For the proximity analysis (p. 310, Table 9.3, column (6)), interaction variables between 4 types of education facilities (ii) and 3 distance segments (iii) are used.

References Baptista, R. and P. Swann. 1998. ‘Do Firms in Clusters Innovate More?’, Research Policy, 27(5) (00), 525–40. Beardsell, M. and V. Henderson. 1999. ‘Spatial evolution of the computer industry in the USA’, European Economic Review, 43(2), 431–56.

318 Japan: Remedies to Activate Local Cities Development Bank of Japan, Hokkaido Branch. 2000. ‘A Sapporo Valley Core Network’, http://www.hokkaido.dbj.go.jp/ Feldman, M.P. and D.B. Audretsch. 1999. ‘Innovation in Cities: Science-based Diversity, Specialization and Localized Competition’, European Economic Review, 43(2), 409–29. Imagawa, T. 2002. Economic Analysis of Telecommunications, Technology, and Cities in Japan. Tokyo: Taga Press. Kelly, T. 1987. The British Computer Industry: Crisis and Development. London: Croom Helm. Kinukawa, M. and K. Yukawa. 2001. The Location Choice of New Media Companies in Tokyo, FRI (Fujitsu Research Institute Economic Review), 5(2), 28–47. Kolko, J. 2000. ‘The Death of Cities? The Death of Distance? Evidence from the Geography of Commercial Internet Usage’, in I. Vogelsant and B.M. Compaine (eds), The Internet Upheaval. Cambridge, MA: MIT Press. Kolko, J. 2002. ‘Silicon Mountains, Silicon Molehills: Geographic Concentration and Convergence of Internet Industries in the US’, Information Economics and Policy, 14(2), 211–32. Konagaya, K. 1999a. ‘Information Industry Model of US City Revival’, The Quarterly Journal of Economic Studies’, 22(1), 1–19 Konagaya, K. 1999b. ‘City Activation by the Information Industries’, Journal of Municipal Problems, 51(5), 76–91. Ministry of Economy, Trade and Industry. 2001. The 2001 Survey of Selected Service Industries. Ministry of Internal Affairs and Communications. 1999. The 1999 Establishment and Enterprise Census. Ministry of Internal Affairs and Communications. 2000. The 2000 Population Census. Ministry of Internal Affairs and Communications. 2001. Communication Trend Survey in 2001. Ministry of Internal Affairs and Communications. 2001. The 2001 Establishment and Enterprise Census. Ministry of Land, Infrastructure and Transport. 2002. Survey on Software-based IT Industries. National School Data Institute. 2000. The 2001 National School Census. Saxenian, A. 1994. Regional Advantage: Culture and Competition in Silicon Valley and Route 128. Cambridge, MA: Harvard University Press. Swann, P. and M. Prevezer. 1996. ‘A Comparison of the Dynamics of Industrial Clustering in Computing and Biotechnology’, Research Policy, 25(7), 1139–57. Toyo Keizai. 2000. The 2000 Edition of City Data Pack. Toyo Keizai. 1990. Toyko Keizai Regional Data 1990. Yukawa, K. 1999a. Area Dependency of Contents Industries – Multimedia Gulch,’ FRI (Fujitsu Research Institute) Economic Review, 3(2), 68–92. Yukawa, K. 1999b. Development of Contents Industries and Policy Effects – Silicon Alley, FRI Research Report no. 47, Tokyo: Fujitsu Research Institute.

Index

acccumulation effect 23, 304 accumulation of IT industry 300–4 Japan see Japan acquisitions and mergers 173, 174 adaptability 8, 113 Aishin 191 alliances 165, 173, 174 amorphic silicon solar cells 135, 136 anchor ﬁrms 14–15, 170–2, 180–1, 196–8 Toyota’s cluster in Tianjin 187–8, 196–7 Antwerp, Belgium 54 ASEAN 97 ﬂow of FDI into 233–4 Japanese investment in 227 ASEAN Investment Area (AIA) 97 Asian Development Bank 34 automobile components manufacturers 144–5 automobile industry Chinese automotive and parts industries see Chinese automotive and parts industries Tianjin clusters 11–12, 137–49; background to development 138–41; construction direction 146–9; development characteristics 141–6 Vietnam 60, 61, 67, 91 backyard capitalism 44, 45 Baden-Württemberg, Germany 54, 55 Bangalore software cluster 19–21, 54, 58, 244–77, 316 emergence as leading software cluster 249–57; evolution and recent developments 253–7; rapid growth of Indian software industry 249–53 insitutional arrangements for skills development 270–1 local labour markets 19–20, 257–71, 272–3; demand for IT workers

257–62; labour mobility within and between ﬁrms 267–8; skills development 268–9; spin-offs 269–70; supply of IT professionals 262–7 Bat Trang 59 batteries see green battery industry Beijing University 316 Bharat Electricals 254 bicycle enterprise clusters 12–13, 149–63 competitive advantage 157–60 development and structural features 156–7 recommendations 161–3 structure 160–1 Wangqingtuo Town 150–5 Binh Duong province 42 brand marketing 175–7 Brazil 207 business associations 33 incubator facilitators in Japan 283–90 role in facilitating cluster formation and development in Vietnam 92–5 Ca Mau 89 call centres 252, 290–1 Calpers investment fund 105 Canon 170, 198 capacity-building 14, 169, 170, 180–1, 196–8 Toyota’s cluster in Tianjin 183–7, 196–7 capital 154 social 53, 93, 154 career paths 268–9 catﬁsh 94 cement 60, 61, 68, 91 central planning 101 central urban districts 166 ceramics cluster 59 Chemical and Physical Electric Source Institute 130–1 319

320 Index China 7–17 automotive and parts industries see Chinese automotive and parts industries comparison with Malaysia in trade and investment structures 18–19, 233–7 electronic information industry 115–16, 117 ﬂowchart approach see ﬂowchart approach industrial zones 1–2, 282–3 Tianjin area see Tianjin area Vietnam and competitiveness of 99 Vietnam and trade with 35–6 China Automotive Industrial Corporation (CAIC) 215, 216 ‘China Shift’ 279, 282–3 Chinese automotive and parts industries 15–17, 138–41, 200–21 current concentration 201–3 domestic background 140–1 economic and political factors 217 historical background 203–7; early production and policies 1949–65 203–4; reform era 1978–94 205–7; revolution era 1966–76 204–5 international background 138–9 issues of localization 207–9 Japanese manufacturers in China 218–19 Shanghai area 17, 209–16, 219 total output of automobile industry 188–9, 190, 191 cities 29, 165–6 remedies to activate in Japan 22–4, 299–318 coffee 94 commercialization 53 commodity exports 34–5 commodity production industries 65 commune shading maps 63, 64, 79–87 comparative advantage 51, 143–4, 148–9 compensation spirals 261 competition 1–3, 295 competitive advantage bicycle enterprise clusters 157–60 change from comparative advantage in Tianjin automobile industry clusters 143–4, 148–9 value chain management 174–7

concentration policy 207–8 competitiveness current competitiveness index (CCI) 38–9 deﬁning 51 growth competitiveness index (GCI) 38–9 three hierarchies of regional competitive power 114–15 Vietnam 36–9; industrial policy recommendations 99–101 consulting services procedures 185, 186 contents creation companies 290–1 cooperation 1–3, 295 core competence 173, 174, 174–5 core–periphery model 47–9, 200–1 Corruption Perceptions Index 103 costs competitive advantage and 175, 176–7 labour costs 99–100, 250, 251 low cost advantage 124 reduction 178; telecommunication costs in Okinawa prefecture 291–3 transportation costs see transportation costs Vietnam 99–100, 101 credit 31 critical mass 51 cross-dock logistics 14, 15, 188–91, 197 culture bicycle enterprise cluster 154 cluster culture 167 current competitiveness index (CCI) 38–9 customer relationship management (CRM) 260 customer satisfaction 173, 174 customer services, IT-enabled 252 customers with discriminatory ability 8, 114 customs duty 178 Cyberjaya 105, 238

Daihatsu 15, 197 Dapice, D.O. 99 demand for IT workers 257–62 Denso 191, 195–6 dependence 119–21

Index 321 diamond of advantage 50–1, 112 diploma colleges 263, 264, 265 dispersion Chinese automotive and parts industries 202–3, 208 electronics industry in Malaysia 227–9 distribution media 8, 112–13 Dong Nai province 99 donor support 33–4, 101 dot density maps 63, 64, 65–78 dynamic competitive advantage 148–9 dynamic effects of economic integration model 47–9 economic factors Chinese automotive and parts industries 217 competitive advantage of bicycle enterprise clusters 159 economic globalization 111–12 economies of scale see scale economies education comparative educational indicators 41 educational facilities and location of IT in Japan 23–4, 301–2, 303, 305–6, 307–8, 309–15, 316–17 higher education institutions in India 262–7, 270–1 regional economic environment 8, 113 efﬁciency 53 Eighteenth Institute of China Electronic Technological Corporation 128 electronic information industry 9–11, 115–27 characteristics of industrial clusters 116–19 main problems of clusters 119–23 recommendations for development 123–7 electronics Malaysia see Malaysia Vietnam 60, 61, 69, 91 embodied information 301 employees, number of 61 engineering colleges 262–4, 270–1 engineering graduates 259–60, 261–2 enterprise groups 163 enterprises’ competitive power 114 entrepreneurship, environment for 103, 105

European Union (EU) automobile industry 138 Malaysia’s market share 230–2; comparison with China 234–5 existing industries 303–4, 305, 307, 308, 309, 314 exogenous industrial clusters 10, 117–18 export processing zones (EPZs) 1–2, 27 Vietnam 4, 28, 102, 103; role in facilitating cluster formation and development 95, 97–9 see also industrial parks; industrial zones exports bicycles from Tianjin 157–8 Indian software industry 249–52; Bangalore 254–5 Vietnam 34–6, 43; role of business associations 93–4 external economies 194–5 factories, number of in Japan 279–82 ﬁnancial capital 154 ﬁnancial organizations 8, 113 First Automotive Works (FAW) Group 141–2, 142–3, 145, 148, 188, 203–4, 206 First International Exhibition of Bicycles, Assistance-Bicycles and Accessories of North China 158 ﬂexible production 161 ﬂowchart approach 14–15, 169–99 ﬂowchart framework 180–1 Toyota’s cluster in Tianjin 181–96; anchor ﬁrms 187–8, 196–7; capacity-building 183–7, 196–7; industrial zones 182–3, 196–7; related ﬁrms 188–96, 196–7 value chain management 172–80 Ford 174 foreign direct investment (FDI) 1, 248 Malaysia 228–9; ﬂow compared with China 233–4; recent ﬂow into electronics industry 18, 226–7, 229–30 Shanghai area 218 Tianjin area 117–18; promotion 126–7 Vietnam 41–4; IPZs, IPs and EPZs 97–9; overview 95–7 foreign-funded enterprise clusters 164 fruit cluster 58–9

322 Index Fujita, M. 56, 57 Fushida Bicycle Enterprise 157 garments 35, 60, 61, 70, 79, 84, 91, 92, 102 Gates, B. 238 General Motors (GM) 174 Geographic Information Systems (GIS) approach 59–92 cluster analysis 90–1 dataset 60–4 dot density maps 63, 64, 65–78 interpretation of maps 89–90 maps 63, 64 methodology for cluster mapping 64–5 northern and southern region sector maps 63, 64, 79–87 provincial codes for Vietnam 60–1, 105–7 seafood enterprises in Mekong Delta 88–90 Gini coefﬁcients 18, 228, 229 global optimization 177–9 globalization 111–12 government support 1–2, 28, 29 Chinese automobile industry: cultivating dynamic competitive advantage 149; cultivating industrial clusters 145–6 local government see local government Tianjin area 126–7; green battery industry 137; industrial parks 162–3 Vietnam 102–3, 104–5 graduate schools 305, 307, 309–11, 311–14, 314, 316–17 green battery industry 11, 127–37 complete types of products and rudimentary industrial chain 131–2 industrial construction 134–6 market prospects and development potential 132–4 R&D and high ability of production 127–31 recommendations for enterprises 136–7 recommendations for government 137 shortcomings and problems 134

growth 29 long-term 99–101 growth competitiveness index (GCI) 38–9 Hanoi 3, 56 Hanoi–Hai Phong corridor 75, 79–83 heavy industry 43, 65–75 Hewlett Packard (HP) 259, 260, 262, 267, 269 high-end software development 259 high-tech industry 161–2 see also information technology (IT) industry higher education institutions see education; research institutes; universities Hindustan Aeronautics 254 Ho Chi Minh City 43, 89 Ho Chi Minh City-Binh Duong-Dong Nai triangle 75, 84–7 Hollywood 54 Honda 218, 219 Hong Kong 118 Hotelling, H. 5, 45–6 HRD Taskforce 266 Hsinchu Park, Taiwan 54, 55 human capital 97, 154 human resource development (HRD) MSC in Malaysia 240–1 Okinawa prefecture 292, 294 Toyota’s cluster in Tianjin 186–7 human resources regional economic environment 8, 113 Tianjin area 141 see also skilled labour IBM 174, 252, 256, 263, 267, 271 iceberg model 44, 45 imitation 165 in-ﬁrm training 268–9 income per capita 283, 284–5 growth rates 283, 286–7 incubator establishment programme 283–9, 296 incentives investment incentives 98 MSC 237 tax incentives 184, 185, 292, 294 increasing returns to scale 44–5

Index 323 Indian Institute of Information Technology (IIIT) 19, 263 Indian Institute of Science (IISc) 254, 263, 266, 269, 271, 316 Indian Institutes of Management (IIMs) 260, 261 Indian Institutes of Technology 262–3 Indian software industry 245 Bangalore see Bangalore software cluster locational distribution of top ﬁrms 253 rapid growth 249–53 Indian Space Research Organization 254 Indian Telephone Industries (ITI) 254 industrial advantage 123 industrial castle town 3, 27, 102 Industrial Cluster Plan of 2001 (Japan) 21, 279–90, 295–6 background 279–83 programmes 283–90 industrial clusters categories 3, 27, 102 competition and cooperation 1–3 component factors 8, 112–14 evolution of regional industrial clusters 111–12 importance 53 and industrial growth 29 international evidence 52–6 minimal requirements for formation 290 perceived role 103–4 theories of formation 44–51 world-class 53–6 industrial districts/regions 5, 156 industrial parks (IPs) 3–4, 27 Tianjin 134–6, 162–3 Vietnam 28, 102, 103; role in facilitating cluster formation and development 95, 97–9 see also Economic Processing Zones; industrial zones industrial policy China’s automobile industry 141–2; failure of industrial policy 208–9 local industrial policy of Shanghai 211–14 recommendations for Vietnam 99–105

industrial zones (IZs) 3–4, 27 China 1–2, 282–3 ﬂowchart approach 169, 170, 180–1, 182–3, 196–8 Vietnam 28, 102, 103; role in facilitating cluster formation and development 95, 97–9 see also economic processing zones; industrial parks informal institutions 94–5 informal networks 301 information, theory of 200 Information and Communications Industry Promotion Zones 294 information and communications technology (ICT) see information technology (IT) industry information policy 304, 306, 308, 309, 315 information-processing 304, 307, 311–14, 315 information services 290–1 information technology (IT) clusters Bangalore see Bangalore software cluster local labour market as determinant of innovation and competitiveness 246–8 Okinawa prefecture 21–2, 279, 290–4, 295–6, 316–17 Vietnam 58 information technology-enabled services 260 information technology (IT) industry accumulation factors 302–4 accumulation in Japan see Japan characteristics 300–1 and innovation 301–2 information technology (IT) professionals career paths 268–9 demand for 257–62 labour mobility 267–8 supply of 262–7 information technology-related courses 264–6 Infosys 260, 262, 267, 271 infrastructure 2, 27 Malaysia’s investment in 237–41 physical infrastructure in Tianjin 183–5

324 Index infrastructure cont’d. regional economic environment 8, 113 Vietnam 102–3; trade-related 36–8 innovation 53, 165 IT industry and 301–2 institutional framework 2, 28 Bangalore software cluster 270–1, 273 Tianjin area 121–2, 153–4 Vietnam 33, 102–3 institutional reform 185–6 Intel 252, 256 intellectual property regime 40 international procurement centres (IPCs) 227 international supply chain 139, 166–7 International Technology Park (ITP) 255 internet 299, 304, 307, 311–14, 315–16 intra-industry trade (IIT) index 236–7 investment domestic in Malaysia 228–9 foreign see foreign direct investment (FDI) investment incentives 98 Ishikawa, Y. 278 Itami, H. 290 Japan 96, 104 automobile industry 138, 187, 207; manufacturers in China 218–19; see also Toyota FDI: in ASEAN countries 227; in Malaysia 226–7 industrial cluster planning projects 289–90 local economy see local economy location of IT ﬁrms 22–4, 229–318; basic model 308–9; data and estimation method 304–8; education and proximity analysis 309–11; estimation result 308–16; implications 314–16; industrial classiﬁcations of IT industry 311–14 Malaysia’s market share 230–2; comparison with China 234–5 number of manufacturing factories 279–82 per capita income by prefecture 283, 284–5; growth rates 283, 286–7 JETRO study 27–8, 102

Jigang Bicycle Company 157 job opportunities for young people 292, 293 joint ventures 62, 63–4 Karnataka state local government 255, 270–1 supply of IT professionals 263–4, 265, 266 keiretsu 170, 180, 188, 190–1, 192 Kelin Bicycle Company 157 Kelly, T. 302 key ﬁrms 16, 206–7 Kinukawa, M. 302, 304 knowledge clusters 39–41 knowledge-intensive industries 161–2 see also information technology (IT) industry knowledge network, lack of 123 knowledge spillovers 5, 47, 52, 245, 246–7, 248 Korea 104, 207 Krugman, P. 5, 46, 47–9 Kuchiki, A. 169–70, 172 Kyoto prefecture 282 labour costs 99–100 software 250, 251 labour-intensive industrial clusters 166 see also bicycle enterprise cluster; garments labour market intermediaries 271 labour markets 19–20 local see local labour markets thick for specialized skills 5, 47, 52–3 labour mobility 20, 29–30, 247, 267–8 labour turnover 268 language skills 260, 269 large core enterprises 160–1 Las Vegas 54 lead-acid batteries 132, 134 leading enterprises 156–7, 160–1 learnability 260 Li-ion batteries 129, 131, 136 output 133–4, 135 light manufacturing 43–4, 65–75 linkages, backward and forward 5, 47, 52–3 local economy 21–2, 278–98 Industrial Cluster Plan of 2001 21, 279–90, 295–6

Index 325 lessons from policies of central and local governments 295–6 Okinawa prefecture 21–2, 279, 290–4, 295–6 local government Bangalore 255, 270–1 Chinese automotive and parts industries 208; Shanghai 211–14, 217–18 Okinawa prefecture 21–2, 290–4, 295–6 Vietnam 42 local labour markets Bangalore software cluster 19–20, 257–71, 272–3; changing nature of 257–70; institutional arrangements for skills development 270–1 as determinant of innovation and competitiveness of IT clusters 246–8 local product districts 3, 27, 102 local production network 10, 118–19 bicycle enterprises 151–2 not embedded 122 localized R&D system 145 location advantage 124, 152–3 location quotient (LQ) 116–17 London 56 ﬁnancial services cluster 54, 55 long-term growth 99–101 low cost advantage 124 Macao 118 macroeconomic environment 8, 113 Mahathir, M. 237, 238 maintenance-free lead-acid battery 132 Malaysia 17–19, 105, 225–43 comparison with China 233–7; ﬂow of FDI 233–4; trade analysis 234–7 formation of electronics industrial agglomeration 226–9; geographical dispersion 227–9; importance of electronics industry 226; international context 226–7 international competitiveness of electronics agglomeration 229–33; recent ﬂow of FDI 229–30; trade analysis 230–3

Multimedia Super Corridor 105, 237–41, 316 managerial skills 260 manufactured goods, expenditure on 48–9 manufacturing 65–75 market 165–6 bicycle enterprise clusters 153 prospects for green batteries 132–4 market integration 126 market network 166–7 market shares analysis for Malaysia 230–2 comparison with China 234–5 Marshall, A. 5, 44, 47 Matsushima, S. 278 Mekong Delta 88–90, 102 mergers and acquisitions 173, 174 Microsoft 262, 263 Windows XP 40–1 Ministry of Aircraft and Spacecraft (MAS) group (China) 215, 216 Ministry of Land, Infrastructure and Transport (Japan) 303, 304 Ministry of Machinery Industry (MMI) (China) 16, 209, 215–16, 216 Mitsui Bank 174 mobility of labour 20, 29–30, 247, 267–8 modularity 191, 194–5 motorcycles 60, 61, 71, 91 Motorola 118, 124, 164, 259, 267, 269 Multimedia Development Corporation 239 Multimedia Super Corridor (MSC) 105, 237–41, 316 critiques of the project 237–8 economic impact 238, 239 promising signs 238–41 Multimedia University 240, 316 multinational corporations (MNCs) and Bangalore software cluster 252, 254, 256; demand for IT professionals 259–60 FDI see foreign direct investment regional optimization 177–9

326 Index multinational corporations (MNCs) cont’d. Tianjin: as centre of interrelated industrial cooperation system 164; electronic information clusters 117–18, 119, 126–7; promotion of competition among 124 Municipal Automotive Small Group (MASG) 213–14, 218 National Association of Software and Service Companies (NASSCOM) (India) 255–6 national dot density maps 63, 64, 65–78 network formation programme 283–9, 295 Network Solutions 269 networks informal 301 large core enterprises plus leading enterprises plus 160–1 leading enterprises plus 156–7 new businesses 295, 296 new economic geography 5, 46, 47–9, 200 new plants, conditions sufﬁcient for building at a cluster 191–5 Ni-MH batteries 128–9, 131–2, 136 output 133–4, 135 Nissan 174, 218 Nomura industrial zone 99 Okinawa prefecture 21–2, 279, 290–4, 295–6, 316–17 current situation of ICT cluster 290–1 local government policies 291–4; development of ICT cluster 292, 293; human resource development 292, 294; job opportunities for young people 292, 293; reduction of telecommunication costs 291–3; tax incentives 292, 294 oligopoly 138, 140, 142–3, 147 on-site assignments abroad 267 organizational structure 159–60, 161 outsourcing 173, 174 ownership type 31–3, 61–4 paper 60, 61, 72, 80, 91, 92

Philippines 96 Phoenix 210 Phu Quoc Island 89, 90, 102 physical infrastructure 183–5 platform for regional economic development 114–15 Plaza Accords 278, 279 policy advantage 124 political decentralization 208 ‘political struggle’ 204 politics 217 population density 303, 305, 307, 308, 309, 314 Porter, M. 6, 7, 46, 49–51, 52, 111, 112, 169, 176–7 priorities, rational structure of 165–6 private computer training institutes 264–6 private sector Tianjin area 150, 154 Vietnam; 4–5, 31–3, 34, 61–3, 101; role in technological catch-up 104–5 product chain industrial clusters 10, 119, 120, 123 product positioning 141–2 production capacity for green batteries 133–4, 136 Chinese automotive and parts industries 204, 205, 207; failure of industrial policy 208–9; small scale 207–8 ﬂexible 161 production functions 45, 173 production licences 153–4 productivity 53, 99–101 programming engineers 260 proximity analysis 306, 307, 309–11, 314 public sector Bangalore software cluster 254 Vietnam 31, 33, 34, 62, 63 Qinghua University 316 quantities of production 194–6 rational structure of priorities 165–6 RCA index for Malaysia 18, 232–3 comparison with China 18–19, 234, 235

Index 327 recruitment 261–2 labour market intermediaries 271 region of studying type 159–60, 161 regional competitive power, three hierarchies of 114–15 regional economic development, platform for 114–15 regional economic environment 7–8, 112–14 regional innovation environment 149 regional innovation systems (RIS) 159–60, 161, 247 regional optimization 177–9 regional value chain network 14, 179–80 regionalization 97, 111–12 related ﬁrms 14, 171–2, 180–1, 196–8 Toyota’s cluster in Tianjin 188–96, 196–7; conditions sufﬁcient for building a new plant at a cluster 191–5; cross-dock logistics 188–91; Denso 195–6 Renault 174 research and development (R&D) Japan’s Industrial Cluster Plan 283–9, 295 outsourcing to India 252; Bangalore 256 research resources 8, 112–13 Tianjin area: automobile industry 139, 145; green battery industry 127–31 Vietnam 39–40 research institutes 301–2, 316–17 restaurants 306, 308, 309, 315 retention strategies 262 rice 60, 61, 73, 91, 92 Ripley’s K-statistic 64 rooting of industrial clusters 10, 118–19 rural areas 165–6 Samsung 124, 164 Santana 210–11, 213–14 Sanyo Motor Company 131 Saxenian, A.L. 93 scale economies 48–9, 200 ﬂowchart approach 170, 171, 194–6 science departments of universities/graduate schools 305, 307, 309–11, 311–14, 314–15, 316 science parks 316–17

Scotland 54 seafood 60, 61, 74, 81, 85, 91, 92 exports 35 Mekong Delta 88–90, 102 Second Automotive Works (SAW) 204–5, 206 self-strengthening mechanisms 124–5 Shanghai area 17, 209–16, 219 historical background to automobile industry 209–10 local industrial policy 211–14 localization of Santana and parts industry 210–11, 212 production structure 216 success factors 217–18 supply chain network: expansion nationwide 215–16; formation 214–15 Shanghai Automobile Assembly Plant 204, 205, 206, 210 Shanghai Automobile Group 141–2 Shanghai Automotive Electric Motor Factory 210 Shanghai Automotive Industry Company (SAIC) 210, 211–13, 214–15, 216 Shanghai Automotive Transmission Shaft Factory 210 Shanghai Car Plant (SCP) 210 Shanghai Localization Ofﬁce (SLO) 213, 214, 217 Shanghai Municipality Trans-bureau Localization Cooperation Ofﬁce (SMTLCO) 214, 218 Shanghai Santana Commonwealth 213–14 Shanghai Volkswagen (SVW) 210–11, 215 Silicon Valley 53–4, 237, 238, 245–6, 267 Simputer 270 Singapore 54–5, 96 single-core industrial clusters 10, 118, 124 fragility and instability of 122–3 skilled labour 247 Bangalore software cluster 19–20, 257–71, 272–3 FDI in Vietnam and 97 location of IT ﬁrms in Japan 300, 302, 303, 305–6; see also education

328 Index skilled labour cont’d. mobility 20, 29–30, 247, 267–8 MSC and job opportunities 240–1 skills development 268–9 institutional arrangements for 270–1 small and medium-sized enterprises Bangalore software cluster 256 Japan 283, 288, 289 Tianjin 185; automobile industry 147–8; bicycle enterprise cluster 149–63 Vietnam 31, 32–3 Soc Trang 89 social amenities 23, 302, 304, 306, 308, 309, 315 social capital 53, 93, 154 Society for Innovation and Development 269–70 software Bangalore see Bangalore software cluster development 259, 290–1 Vietnam 58, 60, 61, 75, 91 see also information technology (IT) industry software business 304, 307, 311–14, 315 Software Technology Parks of India (STPI) 254, 255 sogoshosha 180 solar cells 135, 136 Sony 174, 252, 259 space 45–6 spatial economics 5, 46, 47–9, 200 spatial regression analysis 64 special schools 305, 307, 309–11, 314 specialization rate of value added 117 specialized skills, thick markets for 5, 47, 52–3 spin-offs 20, 269–70 state see government support; local government state-owned enterprises (SOEs) speeding up reform of in Tianjin 125–6 Vietnam 31, 33, 34, 62, 63 steel 60, 61, 76, 82, 86, 91, 102 Stern, S. 52 subsidies 292, 293, 294 sugar 60, 61, 77, 91 Sumitomo Bank 174 Sun Microsystems 252, 256

supply chain see value chain/supply chain supply of IT professionals 262–7 Sweden 54 Switzerland 54, 56 Takeuchi, S. 295 Tata Consultancy Services (TCS) 253, 262, 267, 268, 269 tax incentives Okinawa prefecture 292, 294 TEDA 184, 185 technological dependence 121 technological resources 8, 112–13 technology bicycle enterprise cluster 153 development of electronic information cluster 126 technology-intensive industries 165 see also automobile industry; electronics; information technology (IT) industry technology overﬂow 123 telecommunications cost reduction in Okinawa prefecture 291–3 Vietnam 36, 37 Texas Instruments (TI) 252, 254 textiles 60, 61, 78, 83, 87, 91, 92, 102 thermoelectric cooling modules 132 Thisse, J.-F. 57 ‘three big, three small’ plan 16, 206 ‘three big, three small, two mini’ plan 16, 206 Tianjin area 7–14, 111–68 automobile industrial clusters 11–12, 137–49; background to development 138–41; construction direction 146–9; development characteristics 141–6 bicycle enterprise clusters 12–13, 149–63; competitive advantage 157–60; development and structural features 156–7; recommendations 161–3; structure of clusters 160–1; Wangqingtuo Town 150–5 cultivating new type local industrial clusters 13–14, 163–7

Index 329 development and spread of industrial clusters 7–9 electronic information industry 9–11, 115–27; characteristics of clusters 116–19; future development of clusters 123–7; main problems of clusters 119–23 green battery industry 11, 127–37; basic situation 127–34; industrial construction 134–6; recommendations 136–7 industrial clusters in 9–13, 115 industrial zones 181, 182–3 Toyota’s cluster see Toyota Tianjin Automobile Group 142, 143, 144, 145, 148, 187, 188 Tianjin Economic and Technological Development Area (TEDA) 182–6 Tianjin Jinneng Battery Technological Company 131 Tianjin Lantian Electric Source Company 130 Tianjin Lantian High-tech Power Sources Joint-stock Company 130 Tianjin Lantian Power Sources Company 130, 131, 132, 133 Tianjin Lantian Sanyo Energy Company 130–1, 133–4 Tianjin Lishen Battery Joint-stock Company 129–30, 131, 133 Tianjin Motorola Company 167 Tianjin New High-Tech Industrial Park 134–6 Tianjin Peace Bay Power Company 128–9, 131–2, 133 Tianjin Tingyi International Food Company 167 Tianjin Toyota Motor Company 145, 167, 187–8 Tofﬂer, A. 238 Tokyo prefecture 282 Toyota 131, 200, 201 in China 201, 218, 219 cluster in Tianjin 15, 142, 145, 148, 164, 170, 179–80, 181–97; anchor ﬁrms 187–8, 196–7; capacitybuilding 183–7, 196–7; FAW Group 143, 188; government role 145–6; industrial zones 182–3, 196–7; related ﬁrms 188–96, 196–7; value chain 146

Toyota China Technology Centre 187 Toyotatsusho Corporation 182, 188 trade 51 inner trade characteristics of automobile industry 139 intra-industry trade (IIT) index 236–7 trade analysis for Malaysia 230–3; comparison with China 234–7 Vietnam 34–6; trade-related infrastructure 36–8 traditional industrial clusters 165–6 traditional industry 161–2 training, in-ﬁrm 268–9 Transparency International 103 transportation costs 45–6, 48–9, 57, 200 ﬂowchart approach 194–5 IT industry 301 turnover 61 tyre production projects 208 United States (USA) automobile industry 138 links with Bangalore software cluster 267 Malaysia’s market share 230–2; comparison with China 234–5 trade with Vietnam 93, 96; catﬁsh 94 universities 301–2, 316–17 India 262–3, 270–1 location of IT ﬁrms in Japan 305, 307, 309–14, 314–15 urban areas 165–6 see also cities urban processing clusters 3, 27, 102 value chain/supply chain 49–51, 172–4 international 139, 166–7 Shanghai automobile industry 214–16; expansion nationwide 215–16; formation 214–15; structure 216 Tianjin area: dynamic competitive advantage 148–9; SMEs 147–8; two interrelated systems for automobile industry 146–7 Vietnam’s position in regional and global value chain 96–7 value chain management (VCM) 172–80 competitive advantage 174–7 global optimization and regional optimization of MNCs 177–9 optimal size of a region of 179–80

330 Index Venables, A. 5, 46, 47–9 Vietnam 3–7, 27–108 economic reforms necessary 7, 100 Enterprise Law 2000 6, 32, 96, 103 environment for business 4–5, 30–44; competitiveness 4–5, 36–9; investment 41–4, 95–7; knowledge clusters 39–41; macro economy 30–4; trade 34–6 evidence for existence of clusters 6, 56–92; GIS approach 59–92; VNCI project 57–9 ﬂowchart approach 197–8 industrial policy recommendations 99–105; competitiveness, productivity and long-term growth 99–101; revisiting conclusions from JETRO study 102–5 JETRO study of 2002 27–8, 102 literature review 5–6, 44–51 provinces, districts and communes 65 provincial codes 60–1, 105–7 role of business associations in facilitating cluster formation and development 92–5 role of IZs and EPZs in facilitating cluster formation and development 95–9 Vietnam Business Forum 7, 100

Vietnam Chamber of Commerce and Industry (VCCI) 33, 92 Vietnam Competitiveness Initiative (VNCI) project 57–9 Volkswagen 210 Wangqingtuo Town bicycle enterprise clusters 150–5 formation and characteristics 150–2 formation factors and development mechanisms 152–4 shortcomings and problems 155 Windows XP 40–1 Wipro 253, 254, 260, 261, 262, 268 world-class clusters 53–6 world-class companies 237, 238, 239 World Bank Poverty Reduction Support Credit (PRSC) 34 World Economic Forum 38–9, 103–4 Xiali cars 143–4, 145 young people, job opportunities for 292, 293 Yukawa, K. 302, 304 Zhongguancun IT cluster, China 316 Zhonghuan Semiconductor Company 125

http://media.us.macmillan.com/jackets/500H/9781403943507.jpg

http://media.us.macmillan.com/jackets/500H/9781403943507.jpg [2/9/2009 7:25:59 PM]