The Dynamics of Urban CG ROWTH in Three Chinese Cities .. F i j R ~ . - I --ii ..R. I OS ,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - * #. .w-. Slialli(i YusuLf Weipin(g Wu A W O R L D B A N K B OO K The Dynamics of Urban Growth _ in Three Chinese Cities The Dynamics A .- of Urban Growth _______ in Three Chinese Cities Shahid Yusuf Weiping Wu PUBLISHED FOR THE WORLD BANK OXFORD UNIVERSITY PRESS Oxford University Press OXFORD NEW YORK TORONTO DELHI BOMBAY CALCUTTA MADRAS KARACHI KUALA LUMPUR SINGAPORE HONG KONG TOKYO NAIROBI DAR ES SALAAM CAPE TOWN MELBOURNE AUCKLAND and associated companies in BERLIN IBADAN © 1997 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. Published by Oxford University Press, Inc. 198 Madison Avenue, New York, N.Y. 10016 Oxford is a registered trademark of Oxford University Press. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of Oxford University Press. Manufactured in the United States of America First printing August 1997 The findings, interpretations, and conclusions expressed in this study are entirely those of the authors and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The boundaries, colors, and other information shown on any map in this volume do not imply on the part of the World Bank any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. Cover: Detail from a Chinese scroll, courtesy of the Freer Gallery of Art, Smithsonian Institution, Washington, D.C. Library of Congress Cataloging-in-Publication Data Yusuf, Shahid, 1949- The dynamics of urban growth in three Chinese cities / Shahid Yusuf and Weiping Wu. p. cm. Published for the World Bank Includes bibliographical references and index. ISBN 0-19-521113-8 1. Urbanization-China-Case studies. 2. Industrialization-China- Case studies 3. Shanghai (China)-Economic conditions. 4. Tientsin (China)-Economic conditions. 5. Canton (China)-Economic conditions. I. Wu, Weiping. II. World Bank. III. Title. HT384.C6Y87 1997 307.76'0951-dc2l 97-2013 CIP Contents Preface ix Abbreviations and Data Notes xi 1. The Dynamics of Urban Growth: Location, Size, Structure, and Reforms 1 Reforms: Background and Framework 1 China's Urbanization 4 Selecting Three Cities 6 Factor Supplies and Their Efficient Use 9 Neighborhood Effects: The Impact of Location 10 Agglomeration: Size, Economies of Scale and Scope, and Extemalities 11 Urban Life Cycle and Costs 13 Outline of the Book 15 Appendix: Definition of Cities and Hinterlands 16 2. China's Changing Urban Geography: The Rise of Three Cities 18 The Urban World of Imperial China 19 The Growth of Coastal Cities 25 Economic and Industrial Policy after 1949 29 Urbanization and Urban Policy 36 Summary 42 v vi The Dynamics of Urban Growth in Three Chinese Cities 3. Shanghai: Renaissance City 44 Industrial Transition: From the 1940s to the 1960s 46 Socialism's Industrial Pillar 47 Preparing the Ground for Reform 48 The Scale of Agglomeration 49 Automobiles and Fashion Garments: Externalities at Work 53 Diseconomies: State-Owned Enterprises and Infrastructure 59 Neighborhood Effects 66 Human Capital and Technology 69 The Infrastructure of Finance 70 Capital Flows and Industry 74 Fiscal Relations and Revenue Generation 77 Summary: Reform and Urban Development 82 4. Tianjin: A Port, Its Neighborhood, and Its Ambition 86 Resource Endowment and Growth 87 Industrial Development and Services 90 Neighborhood Effects and Transport-Induced Development 94 Industrial Prospects 98 The Rise of a High-Technology Corridor: Electronics 100 Colonizing Auto Industry Niches 106 Summary: Reforming for Growth 108 5. Guangzhou: The Pearl in the Delta 112 Urban Infrastructure 113 The Politics of Reform: Why Guangdong Was First 115 Industrial Structure and Agglomeration Economies 117 Neighborhood Effects 123 Transport-Induced Development 132 Labor Mobility 134 The Rise of Producer Services 136 Summary 137 6. Similarities, Contrasts, and Lessons: Three Cities and Others 139 Three Cities Compared 140 Driving Forces for Urban Change in China 153 Urban Prospects: International Experience 165 Summary 174 7. A Map of the Future 175 Strategy in Capsule 176 Municipal Finances and Resource Mobilization 177 Taking the State Sector into the Twenty-first Century 177 Human Capital, Foreign Direct Investment, and Technology Transfer 179 Infrastructure 183 Producer Services 185 Summary 186 Contents vii Notes 188 Bibliography 203 Index 220 Tables 2.1 Population and Rank of Large Metropolitan Centers in China, Selected Years, 1922-93 22 2.2 Higher Education Institutions in China, by Province, Selected Years, 1922-94 32 2.3 Transport Networks in China, by Province, 1994 33 2.4 Gross Value of Industrial Output in China, by Province, 1985 and 1990 (1980 Prices) 37 2.5 Gross Value of Industrial Output in China, by Province, 1990 and 1994 (1990 Prices) 38 2.6 Urban Development in China, by Region, Selected Years, 1949-94 39 3.1 Macro Indicators for Shanghai, Tianjin, and Guangzhou, Selected Years, 1952-93 45 3.2 Major Indicators for Shanghai, Tianjin, and Guangzhou, 1993 50 3.3 Gross Value of Industrial Output in Shanghai, Selected Years, 1985-93 51 3.4 Foreign Trade in Shanghai, Selected Years, 1978-93 58 3.5 Loss-Making Industrial Enterprises in Shanghai and Tianjin, 1987-94 60 3.6 Urban Infrastructure of Shanghai, Tianjin, and Guangzhou, 1990 and 1994 62 3.7 Indicators for Infrastructure in Selected Countries, Various Years 63 3.8 Education of the Labor Force in Selected Sectors in Shanghai, 1990 68 3.9 Foreign Investment in Shanghai, 1986-93 76 3.10 Fiscal Relations in Shanghai, 1980 and 1984-91 79 4.1 Education of the Labor Force in Selected Sectors in Tianjin, 1990 88 4.2 Gross Value of Industrial Output in Tianjin, Selected Years, 1985-93 91 4.3 Foreign Investment in Tianjin, 1986-93 104 5.1 Gross Value of Industrial Output in Guangzhou, Selected Years, 1986-93 118 5.2 Foreign Trade in Guangzhou, 1985 and 1990-93 122 5.3 Foreign Investment in Guangzhou, Selected Years, 1979-93 126 5.4 Macro Indicators for Guangdong Province and Pearl River Delta, 1993 130 6.1 Industrial Comparative Advantages in Shanghai, Tianjin, and Guangzhou, 1988-92 Average 142 6.2 Demographic Profiles of Shanghai, Tianjin, and Guangzhou, 1990-91 149 6.3 Shares of Gross Value of Industrial Output of Selected Subsectors in Shanghai, Tianjin, and Guangzhou and Secondary Cities in the Hinterlands, 1985 and 1991 152 viii The Dynamics of Urban Growth in Three Chinese Cities Figures 1.1 Three Cities and Their Hinterlands, China 7 2.1 China's Railway System, 1994 34 2.2 China's Road System, 1994 35 2.3 China's Urban Population by City Size, Selected Years, 1950-90 41 3.1 Industrial Structure of Shanghai, 1985 and 1993 52 3.2 Shanghai and Its Environs 64 4.1 Industrial Structure of Tianjin, 1985 and 1993 93 4.2 Tianjin and Its Environs 96 5.1 Industrial Structure of Guangzhou, 1986 and 1993 119 5.2 Guangzhou and Its Environs 131 6.1 Textiles Industry in Shanghai, Tianjin, and Guangzhou, 1988-92 143 6.2 Machine-Building Industry in Shanghai, Tianjin, and Guangzhou, 1988-92 144 6.3 Metallurgy Industry in Shanghai, Tianjin, and Guangzhou, 1988-92 145 6.4 Chemicals Industry in Shanghai, Tianjin, and Guangzhou, 1988-92 146 6.5 Transport Equipment Industry in Shanghai, Tianjin, and Guangzhou, 1988-92 147 6.6 Electronics Industry in Shanghai, Tianjin, and Guangzhou, 1988-92 148 6.7 Urban Life Cycle: Population Pyramids for All Cities of China and for Shanghai, Tianjin, and Guangzhou, circa 1990 150 Preface This book would not have been possible without the assistance and sup- port of many individuals and organizations. First, we want to acknowl- edge the contribution of two authors of an earlier World Bank report from which we drew a considerable amount of information: Shekhar Shah (World Bank) and Barry Naughton (University of Califomia at San Diego). Shekhar Shah also made valuable suggestions about the outline of the book. We are also grateful to the anonymous referees, who re- viewed both the previous report and the manuscript of this book, for their constructive criticisms and suggestions. Several individuals and groups deserve special thanks. Gary Jefferson (Brandeis University) and I. J. Singh and Dilip Ratha (World Bank) were extremely generous in providing access to large data bases on China's industrial enterprises. The Consortium for International Earth Science Information Network (CIESIN) provided us with base maps of China at no cost and with valuable technical assistance. Within the World Bank, the Asian Technical Department and the cartography unit helped with geographic mapping. William Easterly, Andrew Hamer, Rajiv Lall, Vikram Nehru, and Robert Taylor provided help and information for the book. We also want to thank those who assisted us technically, in- cluding Larry Austin, Les Barker, Charlene Hsu, Jeffrey N. Lecksell, Julia Li, and Min Zhu of the World Bank, Liu Chuang and Jack Evans of the CIESIN, John Zhao of Brandeis University, and Scott Kennedy of the Brookings Institution. We are very grateful to the Policy Research Department of the World Bank, which provided us with a wonderful working environment. Spe- cial thanks go to the director, Lyn Squire, who supported our effort; to ix x Preface Sushma Rajan, whose invaluable help improved every aspect of the book; and to Robert Akl and Maria Elena Edwards, who offered technical sup- port. We would also like to thank the following persons for invaluable as- sistance with the production and publication of this book: Paola Brezny, James Feather, Kathryn Dahl, Elizabeth Forsyth, Rebecca Glenister, Sherry Holmberg, William 0. Lively, Barbara Malczak, and Jason Peaco. Both of us have survived the sweet, protracted agony of authorship more or less unscathed because of the care and support lavishly pro- vided on the home front. Weiping Wu thanks Michael Crowley for love and encouragement that eased the final and most taxing stages of revi- sion. Behind Shahid Yusuf there stood Natasha, Nadia, and Zain. Their enthusiasm for the project ensured that the manuscript never languished and that the year of writing was lived joyfully. Abbreviations and Data Notes FBIS-CHI Foreign Broadcast Information Services-China FDI Foreign direct investment GDP Gross domestic product GNP Gross national product GVIAO Gross value of industrial and agricultural output GVIO Gross value of industrial output R&D Research and development TEU Twenty-foot equivalent unit Y Yuan Market exchange rates, Chinese yuan/U.S. dollar, 1964-96 (IMF, International Financial Statistics Yearbook, February 1997) 1964-71 2.46 1984 2.80 1972 2.24 1985 3.20 1973 2.02 1986-88 3.72 1974 1.84 1989 4.72 1975 1.97 1990 5.22 1976 1.88 1991 5.43 1977 1.73 1992 5.75 1978 1.58 1993 5.80 1979 1.50 1994 8.45 1980 1.53 1995 8.31 1981 1.75 1996 8.30 1982 1.92 1997 8.30 1983 1.98 xi The Dynamics of Urban Growth: 7 g -- Location, Size, Structure, and Reforms i :-~~~~~f Gazing a few decades into the future, optimists see China moving steadily to the top of the economic league. For example, several sources forecast that China will be the world's second largest economy by 2005 (Austra- lian National University 1995) and the world's largest economy by 2020 (Australia 1997; OECD 1997). Although China's per capita income might still trail behind that of some Western economies, in the aggregate China's domestic product will be the largest. Many skeptics are not persuaded that the recent high growth rates will persist for long. In their view, a variety of systemic forces, some already apparent, will increasingly con- strain China's expansion in the years ahead. Both the optimists and the skeptics base their assessments on an evaluation of China's urban economy, which is the principal driving force. This book should be of interest to both camps. Its theme is the interplay of geography, size, and industrial structure in determining the industrial vigor of cities. Its mes- sage, which is abundantly illustrated by the experience of three Chinese cities, is that turning each of these factors to the city's benefit requires sound policymaking. Unless initiatives are taken to exploit inherited capabilities and to approach comparative advantage in a dynamic frame- work, a strong production base can start to decay, pulling the city into a vicious, downward spiral. Reforms: Background and Framework Reforms began inconspicuously in the mid-1970s. Farmers in a handful of China's poorest agricultural counties demanded to be released from the commune system and returned to the autonomy of household pro- duction. At first the central government resisted, but then, in the face of 1 2 The Dynamics of Urban Growth in Three Chinese Cities agricultural stagnation and rural unrest, it cautiously permitted experi- ments in Anhui, Guangdong, and Sichuan.' By 1978 these stirrings of change had acquired sufficient political backing and economic credibil- ity for the government to announce a reform program at the plenary session of the Eleventh Party Congress, held in December. The congress signaled a break from the past and committed China to pragmatically exploring pathways to modernization (see Fewsmith 1994 and Riskin 1987 for excellent accounts of the circumstances leading up to this deci- sion). It was the beginning of the end of an era in which ideology was an arbiter of institutional acceptability and a touchstone for policy. What was initially a limited mandate for reform has grown with each passing year to cover all facets of economic life in China. The Four Mod- ernizations Program, launched in 1978, was pragmatic and experimen- tal in outlook. Its ambitions for the economy were pitched high, but its scope and the degree of institutional change anticipated initially were rather narrow. They grew wider as reforms proved successful and gen- erated "imbalances," which could only be smoothed away by further changes (see Lin 1994). In the earlier years, agriculture was the principal beneficiary, and expansion in farm output accounted for about half of overall economic growth. Agriculture, which until 1983 included rural industry, grew at an average annual rate of 7 percent between 1978 and 1984 and was responsible for 30 percent of overall growth. Thereafter, growth declined, averaging 4 percent annually for the remainder of the 1980s. After 1983 the spread of reform brought industry to the forefront. Since then, the urban industrial economy has been largely responsible for China's nonpareil industrial growth (which averaged 12 percent a year between 1978 and 1995). This is a considerable departure from trends that emerged shortly after the communist government came to power. Maoist policies were colored by the earlier persecution of communists in cities such as Shanghai. With its antiurban bias, the Maoist vision considered cities to be parasitic and strived to contain "nonproductive," mainly service-related, activities (see Buck 1978; Murphy 1980; Chang 1994). One tenet that was never pursued aggressively was to unite city and countryside. Others that were implemented to varying degrees called for dispersing industry, building regional self-sufficiency so as to en- hance defense capabilities, and curbing the growth of the urban popu- lace. The Third Front policy, whose genesis was the perceived threat to China following U.S. involvement in Vietnam, transferred some heavy industry from the coast to the interior provinces and absorbed between a third and half of all gross investment between 1966 and 1973 (see Naughton 1988; Bramall 1993; Zhao 1996). Urban industrial change in China since the early 1980s has followed a circuitous path, and the next step is still being teased forth (see Shirk 1993; Fewsmith 1994; and the commentary in Yang 1996b). Inevitably, The Dynamics of Urban Growth: Location, Size, Structure, and Reforms 3 the story emerging is complex, but its principal elements are now fall- ing into place. First, China's reforms do not constitute the unfolding of a grand design; rather at each major juncture they have been a hedged and extensively brokered response to emerging contingencies.2 For in- stance, once the government eased controls on agriculture, permitted rural markets, and encouraged collective industry to expand, the result- ing dynamic generated pressure to allow urban industrial enterprises to face the new competition from rural producers. Thus each round of re- forms has carried the seeds for future actions. Second, China's size and its long history of decentralized manage- ment within a unitary framework have meant that reforms are interac- tive (with the initiative shifting between central and local players), de- fined in broad terms, interpreted variously, and implemented unevenly. At the outset of the reform era, the balance of power might have favored the center. But with central leaders having to build coalitions (or fac- tions) composed of party, provincial, and People's Liberation Army lead- ers to shore up their own positions, power has tended to become more diffused (Dittmer and Wu 1995 discuss the importance of personal rela- tionships and factions in Chinese politics). The prosperous coastal prov- inces have acquired a bigger say in the conduct and pace of reforms under the decentralized regime. While retaining the role of key decisionmakers, central government leaders have had to listen, negoti- ate, compromise, and occasionally retreat in their dealings with other national stakeholders. Increasingly they have had to respond to the de- mands of the "selectorate," a term Shirk (1993) uses to describe the key central and provincial players. Third, gradual administrative decentralization since the late 1970s, the rise of rural industry, and price reform have all influenced fiscal de- centralization. The last, in turn, has profoundly increased the incentives to introduce new changes. Administrative decentralization has trans- ferred more autonomy to China's 30 province-level units (prior to 1997, when the city of Chongqing was added as the 31st province-level unit), which have delegated authority to the 2,166 county- and lower-level units. Rural industrialization has provided local governments with ad- ditional tax revenue and off-budgetary sources of income. And indus- trial price reform has drastically eroded the profitability of state enter- prises and the revenues derived from these profits. The upshot of all these is a diminution of central administrative and fiscal capacity, which, in turn, has moderated the central government's ability to manage the course of reforms or to take major initiatives. On balance, the course of fiscal change favored the provinces and very likely induced rapid growth, at least in the coastal region. But tax reforms introduced after 1994 re- flect a cross-provincial agreement that too much decentralization, by fiscally impoverishing the center, would be detrimental to macroeco- nomic stability and national development. 4 The Dynamics of Urban Growth in Three Chinese Cities Fourth, nationwide growth and the double-digit rates at which rural industry has expanded in many provinces have brought out two impor- tant characteristics of the Chinese economy: the extraordinary sense of place and the magnitude of social capital accumulated by China's local communities (social capital, a term coined by James Coleman, refers to community organizations, associations, and networks). Local networks, associations, and kin-based ties help generate resources, the institutional arrangements to minimize entrepreneurial risk, and the capacity to pro- vide effective governance, all of which are necessary for development (see Keating 1995). To an unusual degree, history and tradition have equipped China for decentralized development. Fifth, from the very start, industrial reform has been a carefully circumscribed process, which has kept pace with other reforms. In addi- tion, because ownership of state enterprises is distributed across the cen- tral, provincial, and municipal governments, all parties have had an interest in proceeding cautiously. They have been forced to do so because enterprise reform in China affects well over 100 million work- ers, who constitute the most powerful political bloc in the country. Fur- thermore, a very large proportion of China's industrial capital is tied up in the state sector, and any radical moves to impose market-based solu- tions on the public sector would quickly bankrupt between a third and a half of all state enterprises. This brings us to the final point. China's transition to the market has been successful and prolonged for several reasons. It has succeeded be- cause decentralization mobilized local governments to provide the ad- ministrative infrastructure needed to ignite and sustain the growth of industry and the elaboration of markets. It has been prolonged because all the principal players want to reform the state sector gradually. In this way, they minimize the risks of potential upheaval and the costs for the urban workforce. This does not mean that the social costs have been dissipated, only that they are being shouldered by the dynamic segments of the economy. China's Urbanization If we examine the pattem of urbanization in China from the start of the century to the present day, the first four decades saw rapid urban growth with a concentration of population in port cities and in cities strategi- cally located along major transport arteries. War and dislocation during the 1940s interrupted the process, which resumed once the communist government restored order in the early 1950s. That is, a normal tendency apparent worldwide reemerged, with more and more people gravitat- ing toward cities, especially those in the coastal belt. Urbanization crested in 1960, with about a fifth of the population residing in cities. The Dynamics of Urban Growth: Location, Size, Structutre, and Reforms 5 Thereafter, China diverged sharply from the normal pattern.3 The share of the urban populace actually fell, and the growth of major cities, such as Shanghai, was curbed as some manufacturing plants and workers were shifted to smaller cities in the deep interior. Until the 1970s, severe restraints on mobility, enforced mainly through the hukou (household registration) formally introduced and implemented since 1958, checked the flow of people to cities (see Ma and Fan 1994). The hukou deter- mined access to housing, basic foodstuffs, and urban services such as education and health care. Use of the hukou to limit rural-urban migra- tion and even movement within cities, together with the transfer of sev- eral million city youth to rural areas, helped to maintain the level of urbanization at a nearly constant 19 percent between 1960 and 1980 (Zhou 1991). Step by step, however, these restraints have been scaled back and the antiurban bias erased. The result has been a surge in rural-urban migration (Chang 1994 notes that 117 million people migrated to urban areas during 1979-90, most of them on a temporary basis; also see Goldstein 1990; Gui and Liu 1992). Because development has been most rapid in the coastal provinces, cities along the eastern seaboard have attracted the largest number of migrants. Once a rarity, throngs of visi- tors and migrants are now commonplace in the coastal cities. Their pres- ence, which is especially visible around railway stations, is evident throughout the cities, and they contribute significantly to the workforce on construction sites, in some service activities, and in certain assembly industries. This reversal in urbanization policies has restored the "normal" pat- tern of urban growth. According to estimates by the State Statistical Bureau, China's urban population, which was 134 million in 1980, was 478 million in 1994, or roughly a fifth of the total for developing coun- tries.4 It is growing at close to 6 percent a year, which is among the high- est rates for developing countries. Kojima (1995) estimates that the ad- justed rate of urban population growth during 1982-91 was 6.2 percent a year, which has no precedent in economic history. The number of cit- ies rose from 223 to 622 between 1980 and 1994. In 1994 China had 32 megacities, 42 large cities, 173 medium cities, and 375 small cities (see the appendix to this chapter for definitions). Currently, China's urban economy-industry plus services-is the source of two-thirds of gross domestic product (GDP). If current trends persist, the number of urban dwellers will have doubled by the turn of the century, and all but a fifth of national output will be generated by urban areas. For these reasons, an analysis of China's urban economy that adopts a geographic perspective and is informed by historical trends extending to the beginning of this century can be profitable in three ways: first, attention to geographic and historical angles can enrich the economic analysis of recent development in China; second, concentrating on the 6 The Dynamics of Urban Growth in Three Chinese Cities urban industrial sector, which is the hub of economic activity, makes it easier to apprehend the interplay between reform and growth; third, a well-honed sense of urban dynamics is critical to appreciating China's future economic prospects and to understanding what further reforms will be required to deal with old problems still unsolved and new ones lying in wait. Selecting Three Cities There are several alternative methods of analyzing China's urban economy. It can be done on a macro scale using a broad brush and com- posing a picture from aggregate trends and the pattern of urban devel- opment seen on a national scale. At the other extreme, a single indus- trial subsector can provide insight into certain aspects of urban change. For the armature of our story, we have juxtaposed the macro and the subsectoral experience of three major and strategically situated cities: Guangzhou, Shanghai, and Tianjin (see figure 1.1). These cities were se- lected for several compelling reasons. First, they are among the most important urban centers in China, and two of them enjoy the status of provinces.5 The State Statistical Bureau puts their combined industrial product at 9.9 percent of China's industrial output in 1994 .6 Second, they share certain similarities-each is a major industrial center with a coastal location, a rich hinterland, and a privileged place in China's urban and industrial history. Third, the differences among them are also signifi- cant, and their industrial structure varies considerably-it is most di- versified in Shanghai and narrowest in Guangzhou-but there are signs of convergence, as Shanghai and Tianjin have begun to concentrate on fewer core activities and have joined with Guangzhou in emphasizing the development of producer services. Examining their experience can help to uncover the forces driving this convergence. Fourth, all three cities have been active reformers, but their reform programs have dif- fered in timing and scope and in the determination with which they have been implemented. A comparison across the three cities highlights the efficacy of reforms in overcoming obstacles and in exploiting ad- vantages. Modem industrial development commenced in Shanghai in the late nineteenth century, and the city has remained China's foremost indus- trial center. Shanghai's hinterland, which encompasses the lower Yangtze valley, is the richest in resources and one of the most densely populated regions of China. Tianjin, with its fine deepwater port, initially gained significance from its proximity to Beijing and the industrial cities of Liaoning Province. From being a transport hub, it has grown into a lead- ing industrial center. Guangzhou's industrial presence is of a more re- cent vintage. An old administrative and trading center of the Pearl River delta region, Guangzhou acquired a modest manufacturing base in the 1960s and 1970s. However, the extraordinary acceleration of economic Figure 1.1 Three Cities and Their Hinterlands, China IBRD 28654 120' TIANJIN STUDY CITIES ° SELECTED CITIES HINTERLANDS 0 SELECTED PROVINCIAL CAPITALS NONAGRICULTURAL POPULATION * NATIONAL CAPITAL OF SELECTED CITIES SELECTED PROVINCIAL BOUNDARIES _ ,6.0 MILLION -*- INTERNATIONAL BOUNDARIES 4.5-6.0 MILLION 50' A 1~ 3.5-4.5 MILLION 0-3.5 MILLION 1994 GROSS VALUE OF INDUSTRIAL OUTPUT (BILLION YUAN) CHINA 7,690.9 SHANGHAI 425.5 SHANGHAI + HINTERLANDS 1,991.0 TIANJIN 183.8 TIANJIN + HINTERLANDS 758.3 4 GUANGZHOU 149.2 GUANGZHOU + HINTERLANDS 720.4 Harbin 1994 GROSS DOMESTIC PRODUCT (BILLION US$1 Changchun,--,,\ CHINA 509.6 Kr) HONG KONG 131.9 Or-' '/ .enyo@ ..... ., 5henyng,4 -Japan 40' i -Xian Nanljingg Chengdu , Wuhan . - tna g w <- ~~~~~~~~~~~~~~~~~~~C h-i n; a -30, Chongqing . ( . ~~O0O4O->. - t lil t ~~~~~'. -, -O : 2q 3 4fJO SAES -201- C inn 0 2W. 4k0 600MWTERSE- 20, 2t -e AUGUST 1997 7 8 The Dynamics of Urban Growth in Three Chinese Cities activity in Guangdong Province since the start of reforms has transformed Guangzhou into the principal industrial node for southern China, a re- gion whose economic strength rivals that of the lower Yangtze valley Examining growth and industrialization in these three cities illumi- nates the process of change, underscores its diversity, and enables us to perceive the complex interplay between reforms and economic forces. This approach is valuable at several levels: it demonstrates how a spe- cific geographic locus can affect the course of industrialization; it re- veals how China's cities are reversing decades of neglect and adapting to a new, more competitive environment; and it informs future policy not just in China but also in other countries confronting development imperatives of equivalent urgency. It is the theme of this book that at any point in time the economic prospects of a city are deter-mined by location, historical traditions, and momentum derived from past development and the available produc- tion base, which is composed of infrastructure, physical facilities, human capital, and administrative capabilities. These givens define an initial menu of possibilities, but no city is bound by them. The set of options can be enlarged, comparative advantage can be reshaped, and development paths can be redirected by suitable action, which takes history as a point of departure but then uses policy to augment the re- source base, exploit locational benefits, solve systemic ills, and improve the functioning of the city as a dynamic organism. To define and explicate concepts central to this theme, we weave to- gether strands from economic geography and growth theory as well as issues arising from the "urban life cycle." In particular, economic geog- raphy is the source of two important building blocks that directly affect the resource base and productive efficiency of a city's economy: neigh- borhood effects and agglomeration effects. Growth theory offers a frame- work for incorporating the consequences of externalities arising from industrial structure and of policies affecting spillovers of agglomeration (see Mills and McDonald 1992). At a heuristic level, it also permits treat- ment of conditions that are "growth creating" or "growth diverting" in a manner analogous to the analysis of trade creation and diversion in a customs union framework. Last, bringing in the notion of an urban life cycle, it is possible to treat more explicitly the question of production costs that are associated with characteristics linked to age. Some of these have to do with the labor market and with safety nets for retirees; others relate to the state of infrastructure. Olson (1982), for instance, has drawn attention to strong, entrenched organizations and interest groups, such as the tenured urban labor force and party cadres, that affect the pace of change (see also Mueller 1983). But age is also associated with the depth of institutions of learning, which can provide ideas along with some of the motivation to innovate. The Dynamics of Urban Growthi: Location, Size, Structutre, and Reforms 9 Factor Supplies and Their Efficient Use Industrial growth of cities, much like that of the economy as a whole, is ultimately reducible to flows of capital and labor, the efficiency with which they are allocated across production activities, and the rate of technical change. Although the primary factor inputs, especially capi- tal, have been shown to account for a third to half of aggregate growth, the literature on growth accounting has underscored the contribution of human capital-measured most commonly by education-and a vari- ety of elements that affect the productivity of factor combinations (see Denison 1967,1983; Romer 1989, 1993). The four elements that are most meaningful in the urban context are scale economies, industrial organi- zation, the diffusion of ideas, and services that manage the flow of information and, through it, the efficiency of markets. Thus the larger the urban market, which includes the city as well as its immediate hin- terland, the easier it is for industries to achieve economies of scale. Or- ganizational structures, which maximize learning, adaptive flexibility, innovation, and effective networking among firms, have an important bearing on micro-level productivity and the capacity to exploit econo- mies of scope. The ready diffusion of ideas within the urban system and the ability to attract and harness relevant knowledge from other indus- trial centers will stimulate growth as well as efficiency. Last, producer services that generate and diffuse market information have a large hand in determining the quality of allocative outcomes. (The role of institu- tions, particularly in transition economies, has attracted intense atten- tion, especially the matter of property rights; see Poznanski 1992; Weitzman and Xu 1994.) Each city has production and cost functions that are unique in certain respects. The uniqueness derives from the geography and size of the city, the composition of industry, the use of land, and the state of infra- structure. Aside from geography, none of these givens is immutable; they can all be modified by policies, although an underlying inertia places limits on the rate of change. The interesting point is that specific charac- teristics or legacies associated with institutions (for example, those as- sociated with socialist planning and even with geographic circumstances) will affect the readiness to grasp the reform nettle and, thereby, the speed with which reforms begin to work. Our intention is not to specify and empirically estimate produc- tion functions but to use the "new economic geography" and growth theory to explain urban development, proceeding on the premise that the relevant attributes of a city and policies modifying these features impinge on factor supply, the efficiency of utilization, and the (endogenous) generation of knowledge. We make eclectic use of the standard neoclassical growth theory that stresses factor inputs, 10 The Dynamics of Urban Growth in Three Chinese Cities as well as endogenous growth theories that emphasize scale econo- mies and the generation of operationally usable knowledge through the application of human capital. Neighborhood Effects: The Impact of Location Neighborhood or spillover effects refer to the advantages and draw- backs of a particular location (see Porter 1990; Romer 1990; Ayal 1992; Glaeser and others 1992; Hughes and Holland 1994). These can be either growth creating or growth diverting. Growth-creating effects refer to scale economies obtained when industries in an advantageously located city have access to a wide market in the surrounding region through a multiplicity of forward linkages. The size of the market will be deter- mined by population, income levels, industrial structure, effectiveness of the distribution system, and transport infrastructure. Growth- creating effects also arise from backward linkages to regional sources of factor supply and the provision of knowledge. A rich hinterland with a large population can provide labor, skills, and investable resources for industries within the city. In addition, there can be important knowl- edge spillover effects, some of which might be embodied in capital, while others might be communicated by way of technical assistance. Invest- ment in urban industry originating from other parts of the country or overseas can bring technology with it. Knowledge of production prac- tices, coming in the wake of capital, can further enhance productivity. This can be transferred by word of mouth, but the circulation of manag- ers, technicians, and workers is probably of greater importance. Net- working with industry in the region, which permits subcontracting and shares the fruits of research and greater specialization, can also enhance growth through its influence on organization and allocation. As the tech- nological complexity of industry increases, the development and pro- duction of new products require more team effort between specialized manufacturers, who can collectively harness the skills and research and development (R&D) resources. Hence the significance of networking (see chapter 6). City governments, in collaboration with industry, are in a position to maximize growth-creating effects through policies that develop infrastructure; attract capital, skills, and labor to the city; and cement alliances with producers in surrounding areas. In this manner, the advantages of location can be drawn on to the full. The rich litera- ture on industrial districts refers to the resource-augmenting and capacity-building activities that have been markedly successful in north- ern Italy, parts of Germany, the Republic of Korea, and Taiwan (China) (Brusco 1989; Piore and Sable 1984; Markusen 1995; Pyke and Sengenberger 1992; Schmitz and Musyck 1994). The presence of other strong industrial and financial centers in the vicinity can be a source of fruitful competition, but such contenders can The Dynamics of Urban Growth: Location, Size, Strtctutre, and Reforms 11 also divert growth by acting as alternative poles of attraction for resources (Krugman 1991; Storper and Walker 1989). The presence of strong com- petitors, who might benefit from an early start, have a better developed infrastructure, or use policy incentives more aggressively, can negate some of the neighborhood effects that an individual city might enjoy. Competition between large cities in a region can be a positive sum game, as in the Kanto plain of Japan, but it may not be if a dominant player is determined to enlarge its share. In some developing countries, primate cities such as Bangkok and Seoul have inhibited the growth of alterna- tive foci of industrial activity in the surrounding area because their ac- cumulated advantages exert an irresistible pull on resources. Agglomeration: Size, Economies of Scale and Scope, and Externalities The consequences of agglomeration work through some of the same channels as neighborhood effects, but there are several significant dif- ferences as well. Chief among the benefits are scale economies from a larger market and efficient use of lumpy, urban infrastructure. Large agglomerations gain from externalities associated with industrial diver- sity, faster diffusion of knowledge brought about by face-to-face inter- action, and strong demonstration effects.7 They give rise to concentra- tions of labor skills that lower entry barriers to new firms and permit industrial deepening (see Krugman 1991). Agglomerations support ar- rangements between firms that permit the greatest degree of choice in how production is organized within and between companies. For in- stance, just-in-time production scheduling and tight management of in- ventories are most feasible when all suppliers are located within a short radius of the main firm. In many cases, city size is correlated with the diversity of the industrial base, which allows multiple areas of special- ization and scope for responding quickly to new challenges. Large ag- glomerations are generally more successful in setting up financial mar- kets to mobilize local and regional capital than smaller ones. They can also be more innovative in devising mechanisms of delivery, such as venture capital companies that improve access to finance. In addition, they can enjoy a competitive edge, for reasons linked to history and culture, in producing human capital through a system of universities and associated capabilities in R&D (Luger and Goldstein 1991). Agglomeration economies are advantageous not so much because they exercise direct effects on factor supply but because they increase factor productivity by multiplying the avenues for organizational flexibility and by giving fuller play to scale effects. Scale effects are multiplied by feedback effects, with one round of expansion sowing the seeds for sec- ond and subsequent rounds. However, a negative side to agglomera- tion offsets many of the gains (see Wheaton and Shishido 1981). Although 12 The Dynamics of Urban Growth in Three Chinese Cities urban diseconomies of scale are difficult to test empirically, bigger cities often suffer from congestion, environmental pollution, and problems of governance that raise production costs. For example, governance prob- lems can most directly manifest themselves by way of fiscal imbalances, which in turn make it difficult to supply the desirable volume of essen- tial services. Viewed through the lens of growth theory, agglomeration economies can lower the efficiency of capital because stringent pollu- tion controls might be required, because security concerns may entail the unproductive use of labor, or because congestion, which is a grow- ing problem in Chinese cities, means that more funds are tied up in trans- port equipment. Another dimension of agglomeration diseconomies prominent in China, but also encountered in many other developing countries, is the presence of a large state-owned industrial sector in major cities. When publicly owned firms incur losses, they can impose direct costs not only on the central government's budget but also on the finances of the mu- nicipal economy. Beyond that, their organization and labor policies give rise to deleterious externalities that have challenged reformers and strongly influenced the course of change not only in China but also in the transition economies in East Europe and the former U.S.S.R. Ineffi- cient labor practices and the suboptimal wage structure in state firms can spread to private and collective enterprises, compromising their productive efficiency and inflating their costs. The preferential access to financial resources that state enterprises frequently enjoy can have seri- ous, localized "crowding out" effects, which not only starve private com- panies and, through such misallocation, reduce factor productivity but also saddle publicly owned banks with what can turn out to be nonperforming assets. Furthermore, where state enterprises are verti- cally integrated, as is invariably the case in China, they are likely to produce many items in small lots at an excessive cost. Depending on the size of the public sector, this will inhibit the emergence of a more effi- cient production system based on a hierarchy of specialized manufac- turers linked through subcontracting arrangements. As in the case of neighborhood effects, municipal government poli- cies can enhance the benefits of agglomeration while minimizing the drawbacks. Urban infrastructure is the key to realizing scale economies and also to maximizing gains from the hinterland (see Button and Pearce 1989; Eisner 1991). How well a city fares in the provision of, say, trans- port services depends on proper planning, fiscal policies, and pricing of services. The development of financial markets will depend, to a con- siderable degree, on action at the national level, although local initiative and the ability to create a conducive environment are by no means in- significant. Balance with respect to policies as well as to the provision of municipal services is also critical for an environment in which institu- tions of higher learning can take root and flourish. A city can enhance its The Dynamics of Urban Growvth: Location, Size, Struicture, and Reforms 13 locational advantages by means of policy action. Other cities can strengthen their hand by using policy incentives more aggressively to bring about change, remedy deficiencies, and put the local economy on a firmer footing. Urban Life Cycle and Costs Life cycle effects can take a number of forms, but of most relevance for the purpose here are those that influence production costs. Three ten- dencies associated with a city's life cycle can affect the costs of doing industrial business (see Downs 1979; Branbury, Downs, and Small 1982; Porter 1990). First, older cities in which natural population growth has stagnated and migration has not reshaped the population pyramid are, on balance, likely to have an older workforce and to have a high per- centage of retirees. This phenomenon is particularily noticeable in China because of very low population mobility and an enterprise-level social security system that requires firms to support retirees with housing, pen- sions, and social services. This problem is exacerbated by the relatively low age of retirement (50 for women and 55 for men) for workers. This has several implications: the average wage is greater, labor may be more set in its ways and more accustomed to practices that are dysfunctional, interest groups are likely to be more numerous and obstructive, and as Olson (1982) and others have noted, their questioning of change and protective attitude toward the existing regime of entitlements often make it hard to implement policies that depart from established arrangements. Where these tendencies exist, the burden of social security payments is heavier on both individuals and businesses. Under the socialist system, China's state and collective enterprises were and in large part continue to be responsible for the bulk of services, including pensions. They have served as "total institutions," taking responsibility from the cradle to the grave. Second, population density and patterns of land use are likely to raise the leasing costs and periodic fees for serviced land. This has been starkly apparent in cities such as Tokyo for some time and is a rapidly emerging problem in China (on Japan, see Hill and Fujita 1993). Third, the infrastructure of a city, the distribution of housing, and the space available for commercial use evolve gradually over time. As a city ages, it can lose flexibility. Its structure at any time is likely to be opti- mized for conditions suitable to an earlier pattern of living, technolo- gies, and land use. This does change under the pressure of new forces, but it inevitably does so quite slowly. For instance, many new produc- tion technologies are best accommodated by a single-floor rather than a multifloor plant. The location of factories in downtown areas, as was preva- lent in China, can become singularly inefficient from the perspective of land use or commuting (Gaubatz 1995). In older cities, conveniently lo- cated, large factory sites are difficult to find and expensive to lease. 14 The Dynamics of Urban Growth in Three Chinese Cities All of these handicaps can be overcome, and enough old cities have revitalized their industries to suggest that enlightened policies, success at raising finances, and determined implementation do produce results (Sassen-Koob 1986; Fox-Przeworski, Goddard, and de Jong 1991; Sassen 1991). Although greenfield sites away from an existing center are more attractive for companies establishing new plants and wishing to take advantage of cheaper land and a younger workforce without the bag- gage of industrial traditions, mature industrial centers also have their attractions. Workers may be older on average, but they are likely to be more experienced, which has a counterbalancing effect on costs. In ad- dition, cities that have created an infrastructure of training and research provide an environment hospitable for skill- and knowledge-intensive industries and a rich source of externalities. Migration to the city-en- couraged by housing policies, social services, and a reputation for sound management-can enlarge the pool of workers and lower its average age. Efficient administrative practices can ensure that municipal services are less costly than elsewhere, which would make it easier for employ- ers to accommodate higher average labor costs. Because older cities will have amortized more of the physical plant supplying urban services, prices might be lower than in cities at an earlier stage in the life cycle. The mature, diversified, industrial city can have a head start in accumulating learning, which will have a direct bearing on current pro- ductivity and the potential for future growth. The ability to accumulate learning will depend on the past stages of industrialization and the in- teraction among industrial growth, infrastructure building, and the in- creasing level of skills. Henderson, Kuncoro, and Turner (1995, p. 1069) show that "new high-tech industries are more likely to take root in cities with a history of industrial diversity." But once projects have taken root, they are more likely to thrive in relatively specialized cities where the localization economies are more prominent. If we move from cities to regions, according to the U.S. experience, specialization is on the de- cline, and diversity seems to be the more hospitable industrial environ- ment (see Kim 1995). The cumulative experience that enables a city to graduate from less to more sophisticated industries involves a process of layering. A first round of basic manufacturing activities must be followed by successive rounds of industry building in an upward spiral, which puts in place a hierar- chy of technological capabilities. Firms are more likely to invest in as- cending levels of manufacturing activity if the associated infrastructure is created, as has been attempted fairly successfully in Singapore. Ad- vanced industries are far more demanding in their need for transport and communications. Their skilled workers expect a range of physical amenities. Hence cities that invest continuously in the improvement of urban facilities create a milieu conducive to industrial deepening (creat- ing "smart" and "sustainable" cities is the challenge that municipal ad- The Dynamics of Urban Growth: Location, Size, Structutre, and Reforms 15 ministrators and all other stakeholders must tackle; see Serageldin, Barrett, and Martin-Brown 1995; U.S. Congress, Office of Technology Assessment, 1995). Much of the embedding of skills and the acquisition of tacit knowl- edge of production processes occurs on the job or through training pro- vided by firms themselves. But the learning that promotes transition to higher-order manufacturing activities also rests on training and research that codify and assimilate formal knowledge of new technology. When this formal training is combined with on-the-job experience, the stage is set for the upgrading of industry. Learning, therefore, is related to the methodical pursuit of industrial layering on an ascending curve over an extended period of time. The experience of Singapore's electronics in- dustry aptly illustrates the process. Hobday (1995, p. 1188) notes that firms in Singapore were "engaged in a painstaking and cumulative pro- cess of technological learning [rather than a] leapfrog from one vintage of technology to another." Much of the learning occurred in the preelectronic areas such as mechanical, electromechanical, and preci- sion engineering. Competencies tended to build upon each other incre- mentally. Singapore's progress in electronics required the development of capabilities in plastics, moldings, machinery, assembly, and electro- mechanical interfacing; in acquiring the skills needed to manufacture small precision motors; and in increasing the supply of qualified engi- neers. Indeed the historical evidence testifies to the importance of gradual technological accumulation. In any event, if finances can be found, there is little to prevent a city from radically changing land use and upgrading infrastructure so as to seize a new range of industrial opportunities-to stake its claim to new lines of comparative advantage-whether it is exploiting spillover ef- fects, neutralizing agglomeration diseconomies, or coping with ripe maturity; the readiness to use the full range of policy tools so as to sus- tain industrial advantages comes through very clearly both from expe- rience worldwide and from the particular experience of cities in China. The greater the competition is between cities-and regions-the more urgent it is continually to seek the possibilities inherent in location or size, to anticipate and solve problems even as they appear, and to capi- talize on the advantages of a long industrial tradition rather than fight a rearguard action so as to combat urban sclerosis arising from old age. Outline of the Book The book is divided into seven chapters. Chapter 2 provides the histori- cal backdrop to economic geography in the three cities extending from the early years of the century to the eve of reform. It also describes the course of urban and industrial development in China and situates the three cities within the national context. Chapters 3, 4, and 5 explore the 16 The Dynamics of Urban Growth in Three Chinese Cities salient features of development in Shanghai, Tianjin, and Guangzhou, respectively, employing elements of the framework described above and relating the efforts of each city to enhance its industrial strength through the effective deployment of reforms. Chapter 3, on Shanghai, applies the framework most comprehensively. The approach with respect to the other cities is more selective and emphasizes only those angles that bring out important similarities or contrasts between the cities. In other words, the chapters differ in structure and are tailored so as to test and illus- trate the theme of the book to the fullest. Chapter 6 compares the expe- riences of the three cities and their strategies, relates these to restructur- ing efforts in other metropolitan areas of the world, and looks at the forces driving urban change in China. Finally, chapter 7 reviews the main messages emerging from the juxtaposition of theory and experience in the context of the three cities. It also examines the future strategic choices for these cities with respect to both reform and directions of sectoral development. In doing so, the chapter focuses on the choices for Guangzhou, Shanghai, and Tianjin. But through them it also looks be- yond at the imperatives as well as the options for other major cities in China and the developing world. Appendix: Definition of Cities and Hinterlands The administrative structure of cities (shi) and counties (xian) changed significantly in the late 1970s. Prior to this time, cities, with the excep- tion of those under direct control of central and provincial governments (zhixiashi and shengxiashi), and counties were administered by prefec- tures (diqu). The designation of prefectures was eliminated as a result of the change. Counties now are generally administered by major cities in their vicinity. Thus urban statistics are collected for two levels: the city proper (shiqu), including urban area and suburbs, and the city and coun- ties (diqu), including counties administered by the city. For both levels, total population, which includes both agricultural and nonagricultural population, and nonagricultural population are also accounted for sepa- rately. Because of the urban focus of this book, we use the first level of accounting for the three cities, which excludes counties, unless other- wise specified. A pragmatic concern here is with the availability of sta- tistical data, as industrial subsector-level data are rarely available for counties. For the same reason, village-run enterprises are also excluded in our statistical accounting. One major confusion lies in the accounting of urban population. For the period 1964 to 1982, the official measure of urban population was "city and town" population-the aggregate of all nonagricultural popu- lation in the designated cities and towns. The 1982 census used a differ- ent methodology, which defined urban population as all noncounty population in all districts of cities, irrespective of agricultural or nonag- The Dynamics of Urban Growth: Location, Size, Structure, and Reforms 17 ricultural status. Because of the growing concern with the large propor- tion of agricultural population entering the urban count, the 1990 cen- sus used a more complex system, similar to that used before 1982 in principle (for details, see Kirkby 1994; Kojima 1995). The current classification of city size in China is based on nonagricul- tural population in the city proper (excluding counties): * Megacities have more than 1 million population. * Large cities have between 0.5 million and 1 million population. * Medium cities have between 0.2 million and 0.5 million population. * Small cities have less than 0.2 million population. The counterpart of what is called a metropolitan area in North America would then comprise both the city and counties administered by the city. Metropolitan population would include both nonagricultural and agricultural population in the city and counties. There are two ways of defining the hinterlands of cities. The first, and narrower, one includes only counties that are administered by the cities. As mentioned above, the lack of industrial subsector-level data for counties makes this approach a poor fit. The second one embraces larger regions surrounding the cities, reflecting both eco- nomic and geographic boundaries. In our case, we define the hinter- lands of Shanghai as the lower Yangtze River delta (including Jiangsu and Zhejiang Provinces); of Tianjin as the Tianjin-Beijing-Tangshan region (Jin-Jing-Tang, including Beijing and Hebei Province); and of Guangzhou as the Pearl River delta (Guangdong Province). This definition also allows us to demonstrate the emergence of second- ary cities in these regions and their growing competitiveness rela- tive to the three cities. 2 1 China's Changing W Urban Geography: r The Rise of Three Cities To understand the pattern of urbanization in China we must reach into the past. This will enable us not just to uncover macro tendencies sweep- ing the entire country but also to enlarge our comprehension of how certain cities rose to the upper tiers of the urban hierarchy. Ever since the start of the contemporary era, roughly around the 1840s, coastal cit- ies overtook such ancient inland cities as Kaifen, Luoyang, and Xian, which in earlier times had achieved prominence as dynastic capitals, to become centers of modern industrial and commercial development. A major impetus was the signing of several treaties with Western coun- tries and the designation of a dozen or so coastal cities as treaty ports, which began to function as China's windows to the outside world. hIn particular, Guangzhou, Shanghai, and Tianjin dominated the urban scene for about a century, until the communists took over in 1949. Thereafter, their preeminence waned for close to three decades, only to be restored after 1979 when coastal cities were again recognized as the prime devel- oping areas. The level of urbanization in China at the start of the nineteenth cen- tury was fairly low. Between 3 and 4 percent of the populace resided in cities, or approximately 12 million out of a total of 350 million persons (Feuerwerker 1983; Rozman 1990). This is sirmilar to the rates of urban- ization in the United States and Russia (5 and 6 percent, respectively) but well below those in the United Kingdom and continental Europe as a whole (19 and 11 percent, respectively; see Bairoch 1988).Through the balance of the nineteenth century, the pace of urbanization was also much slower in China than in Europe and the United States. A hundred years later, only 17 million Chinese, or about 4 percent of the population, were living in cities. By comparison, a fifth of all Americans were city dwell- 18 China's Changing Urban Geography: The Rise of Three Cities 19 ers, as were 35 percent of Russians and two-thirds of the British. The slow growth of Chinese cities was largely a consequence of industrial backwardness. To thrive and expand, urban centers must export goods and services to pay for the purchase of food and other materials. In the absence of industry, cities must export administrative and commercial services. For most Chinese cities, the provision of administrative ser- vices was their mainstay (see De Long and Shleifer 1993 on the relation- ship between absolutist power and slow growth of cities). Hence growth was concentrated in the provincial capitals and a few commercial cen- ters strategically located along the arteries through which flowed China's domestic waterbome trade. The Urban World of Imperial China At the start of the early modem era for China, six cities occupied the uppermost rung of the urban hierarchy: Canton (Guangzhou), Hankow (Wuhan), Nanking (Nanjing), Peking (Beijing), Shanghai, and Tianjin.' Peking drew its importance from the administrative functions associ- ated with being a capital city. The others were primarily centers of China's nascent industry and trade, both between regions and overseas. The tug of the coast on urbanization clearly reinforced economic forces in the development of four cities-Canton, Nanking, Shanghai, and Tianjin- while Hankow's presence was a carryover from an earlier time when an insular, inwardly focused China oriented its trading activities along the Yangtze River.2 Hankow's choice location along this economic lifeline made it China's largest city in the fourteenth century. Marco Polo de- scribed Hankow as "without doubt the finest and most splendid in the world." Idovic of Bordenone, who spent three years in Hankow in the 1320s, thought it was the world's greatest city (see Mackerras 1989, pp. 19, 21). Even after economnic activity had begun gravitating toward the coast, Hankow retained, for a time, the distinction of being one of China's largest cities; as a place where tea, salt, rice, and medicinal herbs were actively traded, it was still one of the largest cities in the world in the nineteenth century (see Rowe 1984). But as modern industry and inter- national linkages became the determinants of growth, Hankow quickly fell behind and, by the early twentieth century, was no longer near the top of the city hierarchy. This brings us to the factors responsible for the dominant position that industrial cities along the coast attained in the mid-nineteenth century. Until the end of the eighteenth century, all external threats to China emanated from the steppe frontier extending in an arc along the west- ern and northwestern parts of the country. At intervals, tribal confed- eracies became strong enough to send invincible armies of horsemen sweeping into China to wreak havoc and, on occasion, to topple the dynasty holding power. Thus China's military interaction, its diplomatic 20 The Dynamics of Urban Growth in Three Chinese Cities dealings, and its tributary relationships were principally with its no- madic neighbors.3 The westward orientation of the commercial axis re- inforced the tendency to concentrate on trade links along winding land routes connecting China with Asian countries to the west and south. The absence of a threat from Japan and the limited scope for seaborne commerce meant that there were few coastal cities of significance. Amoy (Xiamen), Canton, Ningbo, and a handful of lesser centers flourished, but they were the exceptions. Until well into the nineteenth century, none of the forces responsible for the rise of coastal cities in the West was strongly evident in China. The occasional need to defend coastal settlements against pirates never compelled the government to construct and maintain a large naval fleet.4 China's self-sufficiency, domestic poli- tics that militated against the maintenance of a large oceangoing fleet after the mid-fifteenth century, and the relative backwardness of its east- ern neighbors reduced the scope for trade. Furthermore, shipbuilding, small-scale manufacturing, and finance, which were the lifeblood of Western cities, remained small in scale. The rapid growth of the great port cities of the Atlantic can be attributed largely to the expansion of shipping, which enabled them to acquire significant manufacturing sec- tors, including the shipbuilding industry (see Konvitz 1994). Officialdom, which rigorously administered the interior of China, es- pecially the cross-provincial trade along the principal waterways and land routes, left the coastal areas largely to their own devices. By turn- ing its back on the coast, the imperial bureaucracy gave towns and cities along China's maritime frontier a measure of administrative indepen- dence not enjoyed by urban centers situated inland. Fairbank (1983, p. 13) explains how the "continent facilitated bureaucratic government." To avoid this involvement, private enterprise flowed into coastal com- merce and overseas trade. Such autonomy induced commercial initia- tive, allowed emigration overseas, and from roughly the eighteenth cen- tury onward facilitated the integration of ports in Fujian and Guangdong into trading networks spanning the South China Sea. The physical dis- tance from the center of imperial power in Peking and the government's disinterest in the economy of China's eastern rim imparted a dynamic similar to the one experienced by European cities that spearheaded the revolution first in commerce and then in industry. Canton Among the cities dotting the coastline, Canton was by far the largest until the 1920s (see table 2.1). Its historical roots stretched back 2,000 years, but the growing importance of trade in silk, tea, and ceramics, starting in the latter part of the eighteenth century, brought a new round of prosperity. Standing at the intersection of two marketing hierarchies, Canton became the principal trading center of southern China.5 Canton China's Changing Urban Geography: The Rise of Three Cities 21 occupied this key position in the marketing system spanning southeast- ern China, but it was also a vertex in the trading network created by migrant Chinese communities who had settled in Indo-China, the Phil- ippines, Thailand, and elsewhere (on the spread of Chinese immigrant communities, see Heidhues 1974; on cultural links between China and Vietnam, see Woodside 1971). Tea grown in Anhui and Fujian was traded, in part, through Canton (Gardella 1994). When porcelain ex- ports to Europe, which also served as ballast for vessels carrying tea and silk, gained in significance, high-quality ceramics came from Jingdezhen 600 miles away in Jiangxi Province. By one estimate, ships of the East India Company carried 24,000 tons of fine porcelain to En- gland between 1684 and 1791 (see Goodwin 1991). The goods were first brought by boat to Nanchang on the Gan River. They then traveled south down the Gan River to a point where they could be offloaded and carried over a mountain range to another river, which flowed to the environs of Canton. An alternative route led from Jingdezhen by river to Nanjing, then down the Yangtze to the coast, and thereafter by junk to the southern port city. The means by which goods were brought to Canton highlights an- other factor of importance in the rise of China's major coastal cities. Dif- ficult terrain and a limited road network made land transport exceed- ingly expensive. Difficult terrain was perhaps more important than poorly developed roads. During the late Han dynasty, China's road net- work, with about 35,400 kilometers of roadway, was comparable to that of the Roman Empire: approximately 8.9 kilometers per 1,000 square kilometers. Cotterell and Kamminga (1990, p. 197) describe the Chinese and the Romans as the most enthusiastic road builders of the ancient world. The road network diminished in size during the Tang dynasty, and more responsibility for construction and upkeep devolved onto lo- cal communities. But as late as the seventeenth century, Westerners ex- pressed admiration for China's road system. From the mid-nineteenth century, decay set in as dynastic capabilities diminished (see Ronan 1995). Long-distance trade, especially of bulky products, moved mostly along inland waterways or depended on the availability of coastal shipping. Thus in the imperial era-and even after some rail services were intro- duced-only those coastal cities located at the terminus of a waterway network could sustain strong and continuous commercial links with market towns in their hinterlands and in adjacent market systems. Can- ton occupied an ideal location crossed by tributaries near the mouth of the Pearl River in the delta region. Not only did the rich agricultural economy of the delta produce a food surplus more than adequate to feed a large urban populace, but its tracery of canals, streams, and wa- tercourses could support a trading center, which exported raw material and finished goods drawn from producers scattered across a radius ex- tending several hundreds of miles inland. Table 2.1 Population and Rank of Large Metropolitan Centers in China, Selected Years, 1922-93 1922 1938 1953 1970 1985 1993 City Population Rank Population Rank Population Rank Population Rank Population Rank Population Rank Chongqing 1,772,000 5 2,400,000 7 6,512,000 4 15,036,500 1 Shanghai 1,500,000 2 3,595,000 1 6,204,000 1 10,000,000 1 11,805,100 1 12,947,400 2 Beijing 1,574,000 2 2,768,000 2 6,999,000 2 9,190,000 2 10,511,800 3 Chengdu 1,250,000 15 3,890,000 13 9,473,000 4 Yangzhou 9,338,900 5 Tianjin 1,223,000 4 2,694,000 3 4,000,000 3 7,779,000 3 8,877,300 6 Xuzhou 8,339,400 7 Shijiazhuang 1,069,620 37 8,273,100 8 Handan 7,817,700 9 Nantong 7,812,000 10 Wuhan 1,242,000 3 1,427,000 7 2,560,000 5 4,179,000 12 6,916,900 11 Qingdao 1,300,000 14 4,255,000 11 6,753,500 12 Tangshan 1,333,000 29 6,717,800 13 Shenyang 2,300,000 4 2,800,000 4 5,142,000 9 6,576,500 14 Changchun 1,200,000 16 5,754,116 5 6,510,400 15 Xian 1,600,000 11 2,940,000 17 6,309,100 16 Guangzhou 1,600,000 1 1,022,000 5 1,599,000 6 2,500,000 6 5,620,000 6 6,236,600 17 Hangzhou 5,280,600 7 5,871,000 18 Luoyang 2,501,475 19 5,836,500 19 Suzhou 5,274,107 8 5,692,800 20 Zhengzhou 1,050,000 20 1,424,000 28 5,677,000 21 Changsha 2,403,500 21 5,554,200 22 Fuzhou 1,650,000 27 5,506,500 23 Jinan 1,100,000 17 3,246,000 16 5,335,300 24 Harbin 1,163,000 8 1,670,000 9 2,550,000 18 5,299,900 25 Dalian 1,650,000 10 4,720,000 10 5,270,900 26 Nanjing 1,092,000 9 1,750,000 8 3,612,000 14 5,147,400 27 Qiqihar 1,320,933 30 4,743,600 28 Jilin 1,071,000 36 4,192,800 29 Hefei 3,359,600 15 3,970,800 30 Yichang 1,215,600 32 3,949,300 31 Zibo 2,234,000 23 3,878,400 32 Nanchang 2,483,185 20 3,827,600 33 Kunming 1,100,000 18 1,990,608 26 3,670,700 34 Anshan 1,050,000 21 1,210,000 33 3,315,200 35 Taiyuan 1,350,000 13 2,220,000 24 2,713,500 36 Lanzhou 1,450,000 12 2,310,000 22 2,612,100 37 w2 Fushun 1,080,000 19 2,059,600 25 2,242,200 38 Huainan 1,036,000 39 1,884,300 39 Baotou 1,042,000 38 1,787,600 40 Guiyang 1,277,000 31 1,602,400 41 Hohhot 1,149,000 34 1,419,900 42 Uramqi 1,076,000 35 1,379,300 43 Note: The population account uses the broader definition, including city-administered counties and agricultural population. Data are listed only for cities where the population exceeds 1 million. Source: Fan 1988, appendix A; State Statistical Bureau, China: Urban Statistical Yearbook, 1993-94. 24 The Dynamics of Urban Growth in Three Chinese Cities Shangnai Shanghai had all of Canton's advantages and then a little more. Located on the coastal periphery, it was spared the imperial bureaucracy's vise- like grip. Built along the Whampou (Huangpu) River in the Yangtze delta, it possessed a sheltered harbor, and the riverine environment eased the transport constraints on the city's development. Its rich hinterland was a source of raw materials for the textile industry. The city was rea- sonably well situated to handle the overseas distribution of tea and silk from provinces to the south and southwest when the Taiping rebellion of the 1850s interrupted the flow of these commodities to Canton. Shanghai entered the pages of history as a fishing village in the tenth century. In 1074 it became a county seat and thereafter skillfully capital- ized on its location and deepened its commercial services.6 Merchant families from nearby Ningbo were instrumental in making Shanghai a part of the coastal trading system. In association with the famous Shanxi merchants, they laid the groundwork of banking services essential for a commercial center that was coming to dominate the regional marketing system (see Rankin 1986). As Shanghai's economic influence grew, Kiangnan (Jiangnan) silk merchants began trading through Shanghai, further bolstering the city's position not just as a center of regional com- merce but also as a gateway of exports from the lower Yangtze valley to other parts of the country and overseas. By 1853 Shanghai had surpassed Canton as China's premier trading city, although the size of its popula- tion lagged behind that of Canton for almost another hundred years. Tianjin Although the Tianjin area was settled in neolithic times, the construc- tion of the Grand Canal allowed an urban center to strike roots. Tianjin started out as a transshipment point at which grain and satin barged up one segment of the canal were transferred to another segment so as to continue their northward journey. The earliest recorded references date back to the mid-Tang period, in the eighth century. By the time of the Song dynasty, a fortified town had been established to protect the region and its valuable trade from raiders. During the Ming period, the city acquired its current name, which commemorates the crossing of the Hai River by the Yongle emperor: hence Tianjin, or Heavenly Ford. The city also consolidated its position as a prime transport node and a vital link in the defensive shield around Peking. Much like Canton and Shanghai, Tianjin benefited both from being located at the point where five rivers draining the northern plains enter the Hai River and from being the most accessible port serving the North China plain. Even though the river network around Tianjin is frozen part of the year and the area is notoriously swampy, the importance of water transport added immeasurably to Tianjin's economic advantages. China's Changing Urban Geography: The Rise of Three Cities 25 Grain storage and distribution activities drew a nucleus of officialdom to Tianjin, and the number of government offices multiplied as the production and distribution of salt grew in importance. The shift of the government-run Changlu salt administration to Tianjin in the early years of the Qing dynasty stimulated commercial development, which in turn led to the growth of long-distance coastal trade. As in the case of Shang- hai, this was financed through the system of remittances established by the Shanxi bankers. In the first half of the nineteenth century, Tianjin acquired yet another commercial function, when it became the entrep6t for opium entering north China. Over an 800-year period, Tianjin had progressed from being a transshipment node on the Grand Canal to be- ing the largest port in north China and the center of the area's market system, combining commercial as well as administrative functions. The Growth of Coastal Cities The next stage in the development of all three cities was closely related to the growing presence of foreigners and their investment in industry, trade, and transport; the spread of manufacturing activities; the build- ing of railways that substantially augmented a system largely depen- dent on waterborne traffic; and the strengthening of commercial links with the outside world. The succession of "unequal treaties" imposed on China by external powers-starting with the Treaty of Nanking of 1849, which ended the Opium War-led to the opening of treaty ports and enclaves along the coast and major waterways. Canton and Shang- hai were among the earliest treaty ports. Tianjin became a treaty port after the Treaty of Peking was signed in 1860. At that time its population numbered some 60,000 (see Hershatter 1986). By 1920 almost 100 of China's most strategically placed cities and towns were part of the treaty network. A form of proto-industrialization was taking place in China for well over a century. This was similar to developments observed in South Asia from the seventeenth century onward and was partly responsible for increasing European involvement in the subcontinent (see Perlin 1983). En China, treaty ports helped to catalyze the next stage of development. Small-scale, household-based manufacturing activities, which dovetailed with agricultural production, had a long tradition in many parts of China. Among these, cloth making, food processing, metalware, and woodwork- ing industries were the most prominent. Proto-industrialization was most advanced in Fujian, Guangdong, Jiangsu, and Zheijiang because of cli- mate, opportunities for trading, or agricultural constraints that forced farming households to seek supplementary sources of income. Thus the situation in some of China's coastal areas and the Yangtze valley region during the eighteenth and nineteenth centuries resembled that of Japan during the same period and Europe in the sixteenth and seventeenth centuries. 26 The Dynamics of Urban Growth in Three Chinese Cities Modem industrialization, which began toward the close of the nine- teenth century in Shanghai and Tianjin plus a few other treaty ports, was firmly grounded in a matrix of small-scale rural manufacturing. For instance, cotton cloth, widely traded in the nineteenth century, was produced by farming households and sold to merchants, who served as conduits for the transfer of goods from rural areas to towns and cities. Although mechanized production in factories displaced a significant amount of cottage industry, urban and rural manufacturing evolved alongside urban production, which used putting-out arrangements to tap skills available in nearby rural households. This reduced the need to invest scarce capital in expensive Western equipment and, so long as the logistics of dispersed production were manageable, gave urban en- trepreneurs flexibility in an uncertain environment. The rural-urban symbiosis allowed China's coastal cities to respond rapidly to the stimu- lus provided by modest doses of foreign investment and the growth of trade. This extensive mode of production, with its low overhead and skillful conjoining of modern industry to rural workshop, was a frugal, suitably labor-intensive, and technologically enlightened outcome. As in other late starters such as Italy, Japan, and Russia, the state took the lead in experimenting with Western production techniques, although the waning administrative capabilities of the imperial government, the innate conservatism of senior mandarins, and grave fiscal shortcomings in the latter half of the nineteenth century prevented it from aggres- sively pursuing industrialization (see Gerschenkron 1968; on the contri- bution of official funding to the start of China's modern industry, see Feuerwerker 1977; on revenue constraints that circumscribed the government's capacity to promote development, see Rankin 1986). Nev- ertheless, the ambition of a new breed of administrator to make China "rich and powerful" resulted in several initiatives: first, to equip Chi- nese armies with weaponry comparable to that of their Western adver- saries and, second, to create a production base in textile and metallurgi- cal industries. Bureaucratic capitalism, albeit hampered by the lack of expertise, began building arsenals and shipyards starting in 1865. Shanghai's Kiangnan Shipyard employed 2,000 workers and was well stocked with machine shops and repair facilities.7 Although its output and efficiency were far below expectations, the shipyard enabled the urban economy to acquire new skills, was the source of demonstration effects, and, through linkages, helped to launch other arsenals. A mix of government and local capital also brought into existence a modern cot- ton textile industry. In 1890 Li Hongzhang, an influential government official, built the Shanghai Machine Weaving Mill, China's first. Other mills, also under government sponsorship, sprang up shortly thereaf- ter, not just in Shanghai but also in Canton, Hankow, and Tianjin.8 As in Japan, government action spurred wealthy local merchants to divert some resources from commerce to modern industry. The extra- China's Changing Urban Geography: The Rise of Three Cities 27 territorial status of the treaty ports meant that industrial property owners could enjoy a measure of protection from the demands of the government and the at times unchecked rapacity of warlords and local officials. These two inducements, together with cost advantages accru- ing from foreign investment in rail transport after 1895, pulled local capi- tal into industry. Much of China's railway system through the mid-1930s was funded by loans from Belgium, Germany, and the United King- dom. Prior to 1911 Chinese engineers and contractors were responsible for only 10 percent of construction, but by the 1920s local capacity had expanded to dominate the construction and operation of railroads (see Chang 1993). On the eve of World War I in 1913, the modern industrial sector included 549 Chinese-owned manufacturing and mining estab- lishments, a nucleus that grew rapidly over the next two decades. Innovations in transport and communication, diffusing in from over- seas, transformed China's economic landscape. Motorized water trans- port and railways significantly increased the volume and patterns of production and trade. For instance, the total volume of goods transported tripled between 1895 and 1936 (Rawski 1989). Improved transport re- duced the costs of moving commodities, opened fresh trading routes, and induced the growth of new transportation hubs (including Chongqing, Hankow, Harbin, Shenyang, and Tianjin). The most drastic changes occurred in and around major urban centers, where virtually every new form of transport could be seen, such as railways, trucks, buses, steamships, and even aircraft. The first two passenger cars- Oldsmobiles from the United States-were registered in Shanghai in 1902 (Harwit 1995). Manchuria, or the northeastern provinces, for example, had close to a quarter of all railways in China (M. Howard 1990). In addition, the construction of the north-south railways reoriented do- mestic trading, which had been based on east-west river arteries. This underscored the growing importance of the eastern region, where many industries were now located. Railways soon became the dominant form of transport in China. The first railway line, 15 kilometers long, was constructed by Jardine Mathison to link Shanghai with Wusong. The story of the first railway line is as follows: soon after the first half of the line opened on July 1, 1876, a Chinese man was hit by a train and died (see Hou 1965). Popular protest and opposition from conservative officials followed, and after tense negotiations Jardine agreed to complete construction, run the line for a year, and then sell it to the Chinese. The line began operating and immediately made a profit early in 1877. In November the line became the property of the Chinese, and on the orders of Shen Baochen, gover- nor of Kiangnan, the tracks were uprooted and the rolling stock was shipped to warehouses in Taiwan (China). Except for the Tangshan- Hsukochwang (Xugezhuang) line, no more railways were built until China's defeat at the hands of Japan almost 20 years later, in 1894-95 28 The Dynamics of Urban Growth in Three Chinese Cities (see Feuerwerker 1983; Hou 1965). Railway investment initially resumed in the northern and northeastern parts of the country but later spread from the eastem coastal cities. Major rail lines were built connecting Beijing to Wuhan in central China and to Guangzhou in the south. The rail system expanded fairly rapidly in the first half of the twentieth century, and by 1949 more than 20,000 kilometers of rail lines were in service (Rawski 1989, table 4.7). Priority was placed on developing a network connecting cities lying in the eastern region. Foreign presence in the treaty ports widened China's trading hori- zons and shifted the economic focus of the country to the coastal re- gions. Foreigners also laid the groundwork of legal institutions that gave private Chinese capital a modicum of property rights, which are widely perceived as the basis of a modem market economy. After the first phase, which lasted until 1913, foreign investment increased in volume and diffused beyond commerce and finance into railways and industry, a shift that accelerated urbanization by adding industrial muscle to the economies of coastal cities. Because food processing was the dominant industry in China's coastal cities during the early period, most foreign investment in manufacturing was channeled into the preparation of tea, mainly in Foochow (Fuzhou). The manufacture of brick tea for the Rus- sian market was the area where foreign technology made early inroads into Chinese production methods. The process, which involved com- pressing tea dust and poor-quality tea leaves, was done most profitably in the treaty ports-Foochow and Hongzhou-where the ingredients and coal were cheap and available. Russian investment in brick-tea fac- tories took place in the 1860s. "The industry grew rapidly ... and was one of the largest objects of foreign investment in nineteenth century China" (Gardella 1994, p. 188). However, Chinese producers could not maintain the quality of tea leaf, and by the 1920s China had lost the market to tea growers in Ceylon and India. Even before this happened, the bulk of foreign investment was moving into the production of cot- ton textiles, cigarettes, pressed oil, pig iron, and machinery (for the im- portance of the tobacco industry in Shanghai's manufacturing sector, see Perry 1993). After 1920 China's exports of tea, silk, and porcelain were overshadowed by exports of bean oil, leather, flour, eggs and egg products, ginned cotton, and hog bristles. The domestic market and the export sector also supported a large cotton textile industry concentrated in Shanghai (which boasted half of all spindles) and Tianjin.9 By the third decade of the twentieth century the foreign share of urban industry was sizable, but Chinese ownership remained dominant. The types of manu- facturing activities that attracted investment were such that a modem industrial system located in the large cities kept intact its linkages with input-supplying rural industries in a hinterland that was continuously being enlarged by the extension of railway lines, the improvement of existing roads, and the introduction of steamers. China's Changing Urban Geography: The Rise of Three Cities 29 Economic and Industrial Policy after 1949 For a 15-year period starting in the mid-1930s, wars in which national- ists fought communists and both fought the Japanese halted and in fact reversed industrial change in China. These pitiless conflicts impover- ished the country and inflicted great damage on China's nascent indus- trial sector. Soviet forces, which occupied parts of Manchuria in 1945, stripped factories of equipment, which they took back with them when they withdrew. Shanghai's capitalists transferred textile factories and many of their skilled technicians to Hong Kong, compounding such losses (see Wong 1988). This hemorrhage, which eroded the industrial base built over decades, finally ended with Mao's declaration, on Octo- ber 1, 1949, that "China has stood up." A period of consolidation and stocktaking followed, and industrial functioning, virtually stilled for several years, resumed. By 1952, with normalcy retuming, Guangzhou, Shanghai, and Tianjin were once again in ascendance, accounting for 28 percent of China's industrial output. For the next several years, these cities, especially Shanghai and Tianjin, were in the vanguard of indus- trialization but in addition were hamessed by the planning system to promote industry in other parts of China. After the mid-1950s, central planning began emphasizing economic self-reliance in order to make the inland regions more self-sufficient in industrial terms and also in the interests of national security. A belt of mountainous and dispersed locations within the inland region was tar- geted for industrial development-the Third Front. This policy trans- ferred industry, in Mao's words, to shen (remote areas), shan (mountains), and dong (caves and tunnels). New military production bases and test- ing stations were also built from scratch. Moreover, to aid military re- search and production, certain departments from major national uni- versities in Beijing were ordered to establish satellite campuses in the area. Much of the investment-one-third of the national total-was con- centrated in east Sichuan, west Hubei, south Gansu, and north Guizhou (see Naughton 1988; Zhao 1996). Under the Third Front policy, Shanghai and Tianjin contributed skilled manpower to other regions of the country, from Sichuan to Xinjiang. Scores of factories and schools were moved to the interior of the coun- try. Following instructions from the central government, Shanghai transferred as many as 1 million skilled workers and technicians, not including rusticated youth with no industrial experience. The largely involuntary resettlement of youth and some professionals from urban to rural areas occurred during the Cultural Revolution, from 1967 to 1972 (see Bernstein 1977). Tianjin sent more than 10,000 skilled workers and technicians from the textile industry alone in order to foster the spread of that industry to other regions. Moreover, when the economy was faced with particular technological challenges, national planners 30 The Dynamics of Urban Growth in Three Chinese Cities sought solutions among enterprises and research institutes located in the coastal cities. For example, Shanghai enterprises aided in the recov- ery of Chinese industry after the withdrawal of Soviet assistance in 1960. They also developed the expertise to permit petroleum drilling to con- tinue after access to Soviet technology was lost. The most important channel for the transfer of technology from coastal cities to the remainder of China was the sale of investment goods, par- ticularly machinery. Shanghai, Tianjin, and to a very limited extent Guangzhou were important sources of industrial machinery for the na- tional economy. Shanghai firms supplied machinery to, among others, the electric power, building materials, shipping, and petroleum indus- tries. Shanghai and Tianjin provided machinery to the textile and other light industrial sectors. In addition, they often served as intermediaries in the importation of advanced technology. Exemplars of advanced machinery were imported to Shanghai, which then had the task of rep- licating the technology and "exporting" it to the remainder of China. Examples include basic oxygen technology in the steel industry and the early development of petrochemicals. Thus, Shanghai and Tianjin, in particular, contributed vitally to the growth of the Chinese economy under conditions of near autarky between 1960 and 1978. An even more important relation between coastal cities and the na- tional economy came through the price and fiscal systems. The cities were manufacturing centers, importing raw materials from other parts of China in exchange for manufactured goods. These transactions were priced according to rules adopted from the U.S.S.R. during the 1950s. Manufactured goods had high relative prices, and agricultural and min- ing products were correspondingly underpriced. As a consequence, manufacturing was exceedingly profitable, and the three cities, which sold their products throughout the country, enjoyed large budgetary revenues. The national government then extracted most of these through the fiscal system. Although the industrialized coastal cities obtained substantially larger incomes than inland municipalities, their tax rates were also much higher. In the mid-1970s development priorities began to favor the coastal regions again, and Shanghai and Tianjin enjoyed a burst of new invest- ment. The shifting development policy during the mid-1970s prefigured the greater changes to come in the 1980s. China began to move away from the xenophobic isolation of the Cultural Revolution and prepared to increase its involvement with the world economy. Because Shanghai and Tianjin produced the highest-quality textiles in the country, they received investment funding and were required to increase exports. Port facilities were also expanded. Perhaps more crucially, the national gov- ernment began to locate large new projects in coastal cities, particularly those that were technologically advanced. Petrochemical complexes were established in Shanghai and Tianjin, and the huge Baoshan Steel Mill was begun in Shanghai in 1978. China's Changing Urban Geography: The Rise of Three Cities 31 On balance, in the period through 1978, the central government neglected the infrastructure of the cities but recognized their superior endowment of human capital and industrial experience. As shown in table 2.2, about half of all higher education institutions were concen- trated along the coast in the eastern region (512 out of 1,080 in 1994). The national government siphoned off resources but regularly turned to Shanghai to realize technological objectives or manufacture high- quality products. The government exploited the human capital and industrial experience accumulated by these large cities, while making inadequate and partial contributions to nurturing their long-run productivity. By 1978 the share of the three cities in national industrial output had shrunk considerably: Shanghai's to 12.1 percent and Tianjin's to 3.7 percent. Guangzhou marginally increased its share of national in- dustrial output to 1.8 percent. This occurred in the context of a slight erosion in the weight of Guangdong Province in national output (from 4.9 to 4.7 percent). Thus, over the entire 1952-78 period, Guangdong's industrial output grew somewhat slower than the national average but became more concentrated in Guangzhou. During the early stage of reform, starting in 1978, the central govern- ment recognized the necessity for balanced industrialization, which gave equal importance to consumer goods and eased the emphasis long given to heavy industry. The coastal cities responded to the signals from the cen- ter by expanding the output of light manufactures. Moreover, China's ex- port surge during 1979-82 relied on the rapid growth of traditional manu- factured consumer goods, primarily from the coastal cities. As the 1980s proceeded, the national government began giving greater economic prior- ity to the coastal regions. The objective of equalizing industrial production throughout the nation was abandoned, and economic development was viewed as gradually spreading inland from coastal poles. Preferential poli- cies were adopted for coastal regions to foster this process, and in 1984, 14 "open cities," including Guangzhou, Shanghai, and Tianjin, were given greater autonomy to manage foreign trade and investment as well as for- eign exchange retention rights over the proceeds from their exports. Transport development, which also favored the eastem region, fur- ther stimulated the economies of coastal cities (see table 2.3). By 1994 China's rail network measured close to 54,000 kilometers, following the completion of several strategic lines extending deep into the southwest (figure 2.1).1o After the start of reforms, the increased demand for trans- port along the eastem corridor induced the government to invest in double tracking and electrification, to upgrade the rolling stock, and to add diesel and electric engines to a fleet composed largely of steam lo- comotives. These additions and technological improvements, although barely keeping pace with demand, permitted the coastal region to double if not treble its growth rates. Rail transport (and more recently road trans- port, see figure 2.2) made it possible for the eastern provinces, which have scarce energy and raw materials, to bring in much of the needed 32 The Dynamics of Urban Growth in Three Chinese Cities Table 2.2 Higher Education Institutions in China, by Province, Selected Years, 1922-94 (numbers, unless otherwise noted) Province 1922 1934 1949 1959 1965 1979 1985 1994 Eastern region Number in region 32 82 121 385 228 301 493 512 Percentage of country total 80.0 74.5 59.0 45.8 52.5 47.6 48.5 47.4 Beijing 7 17 15 50 53 48 62 67 Tianjin 2 - - - - 14 21 22 Liaoning 0 0 8 55 26 34 62 61 Hebei 0 9 11 31 25 27 46 52 Shandong 1 3 7 35 16 34 49 49 Jiangsu 6 11 15 66 30 36 70 67 Shanghai 10 24 37 31 24 27 45 46 Zhejiang 1 4 5 28 13 19 35 37 Fujian 3 4 8 20 10 16 36 33 Guangdong 2 8 12 42 21 29 44 46 Guangxi 0 2 3 27 10 17 23 27 Hainan 0 0 0 0 0 0 0 5 Central region Number in region 7 20 34 307 126 197 328 354 Percentage of country total 17.5 18.2 16.6 36.5 29.0 31.1 32.3 32.8 Shanxi 0 5 1 29 11 16 22 26 Inner Mongolia 0 0 0 18 8 13 18 19 Jilin 0 0 6 35 17 25 44 43 Heilongjiang 0 0 6 39 19 27 40 43 Anhui 0 1 2 32 12 20 36 35 Jiangxi 1 3 5 32 12 17 26 31 Henan 0 3 2 42 12 24 43 50 Hubei 5 6 10 53 23 33 56 60 Hunan 1 2 2 27 12 22 43 47 Western region Number in region 1 8 50 149 80 135 195 214 Percentage of country total 2.5 7.3 24.4 17.7 18.4 21.3 19.2 19.8 Sichuan 1 4 36 55 30 42 56 63 Guizhou 0 0 3 16 6 14 22 22 Yunnan 0 1 3 9 6 15 26 26 Tibet 0 0 0 0 0 4 3 4 Shaanxi 0 1 3 30 21 28 44 47 Gansu 0 1 4 20 7 12 17 17 Qinghai 0 0 0 7 2 6 6 7 Ningxia 0 0 0 3 1 4 6 7 Xinjiang 0 1 1 9 7 10 15 21 Country total 40a 110 205 841 434 633 1,016 1,080 - Not available. a. 40 campuses of 37 institutions. Source: Ministry of Education 1984, 1986; Fairbank and Feuerwerker 1986; State Statistical Bureau, Statistical Yearbook of China, 1995. China's Changing Urban Geography: The Rise of Three Cities 33 Table 2.3 Transport Networks in China, by Province, 1994 (kilometers, unless otherwise noted) Railway Inland Road Region Operating Expanding waterway Total Class 1 Eastern region Kilometers in region 15,882 23,286 58,536 394,843 164,143 Percentage of country total 29.4 32.9 57.0 35.3 46.5 Beijing 1,020 1,606 n.a. 11,532 8,081 Tianjin 502 896 90 4,156 3,624 Liaoning 3,557 4,953 508 42,763 17,155 Hebei 3,230 5,293 75 50,496 33,455 Shandong 2,048 3,060 1,891 50,225 34,781 Jiangsu 747 1,294 23,787 25,891 12,690 Shanghai 246 403 2,100 3,721 2,462 Zhejiang 937 1,355 10,592 33,170 11,602 Fujian 1,024 1,037 3,888 44,608 8,571 Guangdong 682 1,297 10,808 75,716 19,230 Guangxi 219 238 4,521 39,550 10,619 Hainan 1,670 1,854 276 13,015 1,873 Central region Kilometers in region 25,273 33,427 31,572 375,120 108,377 Percentage of country total 46.8 47.2 30.7 33.5 30.7 Shanxi 2,330 3,455 170 32,693 12,828 Inner Mongolia 5,072 5,836 602 44,202 7,223 Jilin 3,487 3,822 1,134 29,581 7,452 Heilongjiang 4,943 6,562 n.a. 48,356 6,453 Anhui 1,753 2,500 5,612 30,876 12,170 Jiangxi 1,581 2,204 4,937 34,556 6,964 Henan 2,133 3,945 1,104 47,704 27,940 Hubei 1,693 2,171 8,003 48,349 14,939 Hunan 2,281 2,932 10,010 58,803 12,408 Western region Kilometers in region 12,838 14,165 12,615 348,308 80,613 Percentage of country total 23.8 20.0 12.3 31.1 22.8 Sichuan 2,881 2,883 7,904 100,002 15,425 Guizhou 1,423 1,480 1,773 32,398 3,558 Yunnan 1,592 1,643 1,324 65,578 13,126 Tibet 21,842 868 Shaanxi 1,826 2,184 998 39,508 12,780 Gansu 2,219 2,940 219 34,984 11,304 Qinghai 1,095 1,098 n.a. 17,061 3,911 Ningxia 459 522 397 8,324 2,753 Xinjiang 1,343 1,415 n.a. 28,611 16,888 Nonprovincial 7,509 Country total 53,993 70,878 102,723 1,118,271 353,133 n.a. Not applicable. Note: Class 1 roads connect major cities and commercial centers. They carry large volumes of traffic and are designed to higher standards. They have a minimum of two traffic lanes of at least 3 meters width and shoulders of a prescribed size. Source: State Statistical Bureau, Statistical Yearbook of China, 1995. Figure 2.1 China's Railway System, 1994 '50° 70° 8b° 90° 100° 11'0° I1 2 13~0 nr~~~ A1 C on Xcldlu) 40~~~~~~~~~~~~~~~~~~~~~~~~~~~0 TI A NJIN S TJDY CITIES -J TooHlmg Dg o NONAGRICULTURAL Y rnIAuY a S KU ~~~~~~POPULATION OFp SELECTED CITIES, 1 99A G... GUANGZHOU: n n e * TIANJN- ON _____ MALLORN ALWY TIN I SECTEDY CITIES Z o SELECTED PROVINCIAL CAPITALS Hou Nannin NATIONAL CAPITAL 2YQ 400MILS - PROVINCIAL BOUNDARIES )bnOn KIOERS' -20' . >-" So INTERNATIONAL BOUNDARIES ~C,o o Figure 2.2 China's Road System, 1994 50° 70° 8b0 90° 10o0 11°o 120o 130° A-'~~~~~~~~~~~~~~~~~~~~~0 \,° Oruqi .~^ * CtCngchun 4O~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~A' 8O> r ~\, ( ;g H ,,-Hohho ''t 80- ' JSp s NONAGRICULTURAL <+.>/,3hichg/.r l POPULATION OF N,.irfhup lTaipuan D CITIES 199 Golmud . GnQon Aa GUANGZHOU: ' / ~ J Xining La zho u n ,TtIow Sea @ TIANJIN: Xia HANGHAI * .1 MILLION _ 3 Ceg 6eeOa SHANGHAILhs uas' TIANJIN STUDY CITIES 'g H gy'fl h MAJOR HIGHWAYS o SELECTED CITIES Ruili / *3 SELECTED PROVINCIAL CAPITALS He ni ®t. * NATIONAL CAPITAL 'iYoyguon 200 400 LES 201 - 0 w-- PROVINCIAL BOUNDARIES 20° S00on 600KILOMETERS C 0 200 400 600 KILOMETERS ~~-20 INTERNATIONAL BOUNDARIES ,5r 1O0_ \ Son lj40 Chin. &no 12Se a. 36 The Dynamics of Urban Growth in Three Chinese Cities inputs and to service the demand for manufactures in the central and western regions of China." The shift in economic strategy spurred industrial development in the coastal provinces, but the large older centers did not immediately par- ticipate in this growth and prosperity for reasons that are related to the manner in which reforms were introduced. During the first half of the 1980s, although some provinces began to forge ahead, fast-growing prov- inces were scattered throughout the country, and no clear geographic pattern emerged. During the second half of the 1980s, growth did in- deed become concentrated in coastal areas, but not in the traditional centers (this description excludes the province of Tibet). Shanghai and Tianjin were the two slowest-growing province-level units. Five coastal provinces emerged as superior performers (see table 2.4): ranked by annual industrial growth rates, these were Guangdong (17.4 percent during 1985-90), Fujian (15.8 percent), Shandong (14.4 percent), Jiangsu (13.0 percent), and Zhejiang (12.4 percent). The south coast became the most dynamic region of the country, with growth radiating northward through the middle coastal provinces. During the early 1990s, however, Shanghai and Tianjin's performance improved significantly and ranked in the midrange for all provincial units (table 2.5). The top five performers remained the same: Guangdong, Jiangsu, Zhejiang, Fujian, and Shandong. Guangzhou, situated in the middle of the most rapidly growing region of all, maintained its growth at the national average during the 1980s. De- spite Guangzhou's respectable overall growth rate, all three coastal cities failed to partake fully of the growth momentum in the 1980s. As a group their share of industrial output fell to 12 percent in 1990 and 10 percent in 1994 (compared with around 18 percent in 1978). To a certain extent, it was inevitable that these early developers would lose market share as China's industrialization spread nationwide. Nevertheless, the relative decline of at least Shanghai and Tianjin at a time when other coastal areas were forg- ing ahead appears to have been overly rapid. It reflects the burden im- posed by a large state sector and past neglect of infrastructure, as well as the central government's unwillingness to give these two cities sufficient latitude to pursue far-reaching reforms. It also reflects the slow pace at which municipal authorities used the measure of autonomy gained by way of decentralization to modernize their economies. Urbanization and Urban Policy Starting in the mid-nineteenth century, industrial and commercial de- velopment promoted urbanization, especially in the larger coastal cit- ies. Among China's macro regions, only three had urbanization rates greater than 4 percent in the 1840s-the lower Yangtze, 5.8 percent; Lingnan, 5.3 percent; and southeast China, 4.2 percent, much of it in Canton, China's most populous city at the time. By the 1890s urbaniza- tion levels had increased substantially in the lower Yangtze and Lingnan China's Changing Urban Geography: The Rise of Three Cities 37 Table 2.4 Gross Value of Industrial Output in China, by Province, 1985 and 1990 (1980 Prices) Province and rank 1985 1990 Average annual by average annual Billions of Share Billions of Share growth rate growth rate yuan (percent) yuan (percen t) (percent) 1 Guangdong 46.1 5.6 102.7 7.9 17.4 2 Fujian 14.1 1.7 29.3 2.2 15.8 3 Shandong 54.0 6.5 105.7 8.1 14.4 4 Jiangsu 86.4 10.4 159.1 12.2 13.0 5 Zhejiang 44.4 5.4 79.5 6.1 12.4 6 Ningxia 2.2 0.3 3.7 0.3 11.0 7 Qinghai 1.9 0.2 3.1 0.2 10.3 8 Anhui 22.2 2.7 35.5 2.7 9.8 9 Shaanxi 16.9 2.0 26.5 2.0 9.4 10 Guangxi 12.4 1.5 19.4 1.5 9.4 11 Jiangxi 15.1 1.8 23.6 1.8 9.3 12 Hebei 33.3 4.0 50.2 3.9 8.6 13 Hunan 26.4 3.2 39.5 3.0 8.4 14 Inner Mongolia 9.5 1.1 14.1 1.1 8.2 15 Shanxi 19.0 2.3 27.8 2.1 7.9 16 Henan 32.1 3.9 46.9 3.6 7.9 17 Yunnan 12.1 1.5 17.6 1.4 7.8 18 Jilin 21.5 2.6 31.2 2.4 7.7 19 Gansu 11.4 1.4 16.5 1.3 7.7 20 Beijing 31.5 3.8 45.4 3.5 7.6 21 Guizhou 8.4 1.0 12.1 0.9 7.6 22 Hubei 42.3 5.1 60.9 4.7 7.6 23 Sichuan 45.4 5.5 64.9 5.0 7.4 24 Xinjiang 7.3 0.9 10.2 0.8 6.9 25 Liaoning 66.4 8.0 87.8 6.7 5.7 26 lianjin 28.6 3.4 37.6 2.9 5.6 27 Heilongjiang 35.2 4.2 45.9 3.5 5.5 28 Shanghai 83.2 10.0 103.1 7.9 4.4 29 Tibet 0.12 0.0 0.14 0.0 3.1 Country totala 829.5 100.0 1,302.4 100.0 9.4 Note: Excludes village-run enterprises. a. Excludes the new province of Hainan. Source: State Statistical Bureau, Statistical Yearbook of China, 1986 and 1991. regions, reaching 8.3 and 6.6 percent, respectively. Urbanization in north- west China had risen to 4 percent, mainly as a result of foreign investment in Manchuria, and the urban population of the southeastern macro region had increased marginally. Compared with Europe or Japan, China had fewer cities overall. This reflected not just the relative dearth of commerce but also the dispersed, rural, household-centered nature of industry. Canton, Shanghai, Tianjin, and Wuhan were the main centers of popu- lation and of industry at the turn of the century, as they were in the 1920s (see table 2.1).12 Thereafter, industrialization was concentrated along the east coast, the northeast, and the Yangtze River valley, mostly in large cities. When the nationalist government made Nanking the capi- tal, the city acquired administrative functions and greater industrial sig- nificance (the capital was moved from Peking to Nanking by Yuan Shikai 38 The Dynamics of Urban Growth in Three Chinese Cities Table 2.5 Gross Value of Industrial Output in China, by Province, 1990 and 1994 (1990 Prices) Province and rank 1990 1994 Average annual by average annual Billions of Share Billions of Share growth rate growth rate yuan (percent) yuan (percent) (percent) 1 Guangdong 138.0 7.4 459.6 10.8 35.1 2 Jiangsu 206.1 11.0 629.5 14.8 32.2 3 Zhejiang 104.8 5.6 316.4 7.4 31.8 4 Fujian 39.4 2.1 108.6 2.6 28.8 5 Shandong 151.7 8.1 394.4 9.3 27.0 6 Guangxi 29.9 1.6 72.4 1.7 24.7 7 Anhui 52.9 2.8 122.5 2.9 23.4 8 Henan 71.3 3.8 158.3 3.7 22.1 9 Jiangxi 34.3 1.8 75.9 1.8 22.0 10 Sichuan 98.1 5.3 216.9 5.1 21.9 11 Hubei 81.6 4.4 174.9 4.1 21.0 12 Beijing 62.6 3.4 125.8 3.0 19.1 13 Shanghai 147.7 7.9 292.3 6.9 18.6 14 Hebei 75.5 4.0 149.2 3.5 18.6 15 Tianjin 52.5 2.8 103.4 2.4 18.5 16 Xinjiang 17.4 0.9 33.9 0.8 18.1 17 Ningxia 5.7 0.3 10.4 0.2 16.2 18 Hunan 58.7 3.1 105.2 2.5 15.7 19 Shaanxi 36.0 1.9 63.3 1.5 15.2 20 Liaoning 130.2 7.0 226.1 5.3 14.8 21 Yunnan 30.8 1.7 53.2 1.2 14.6 22 Inner Mongolia 23.1 1.2 38.7 0.9 13.8 23 Guizhou 18.8 1.0 31.1 0.7 13.4 24 Jilin 47.8 2.6 79.0 1.9 13.4 25 Shanxi 42.4 2.3 68.4 1.6 12.7 26 Gansu 24.8 1.3 39.9 0.9 12.6 27 Qinghai 5.1 0.3 7.9 0.2 11.6 28 Tibet 0.26 0.0 0.34 0.0 6.9 29 Heilongjiang 77.9 4.2 99.9 2.3 6.4 Country total' 1,865.4 100.0 4,257.4 100.0 22.9 Note: Excludes village-run enterprises. a. Excludes the new province of Hainan. Source: State Statistical Bureau, Statistical Yearbook of China, 1991 and 1995. in 1928; see Gray 1990). In 1937, under pressure from invading Japanese forces, the nationalists were forced to move the capital to Hankow and then, in 1938, to Chongqing. This also triggered the relocation of many coastal manufacturing establishments to the interior. Industry con- tinued to be concentrated in a relatively few major cities, but now the mid-Yangtze valley area acquired an industrial status closer to that of Shanghai and Tianjin. By 1953 the census placed Chongqing fifth and Nanjing ninth in the city hierarchy. When the communist government took over in 1949, about 39.5 mil- lion Chinese constituted the urban population, more than two-thirds of whom were living in the eastern region. This region also contained close to half of all cities, including most of the largest ones (see table 2.6). The densely populated central region was predominantly rural, with some China's Changing Urban Geography: The Rise of Three Cities 39 Table 2.6 Urban Development in China, by Region, Selected Years, 1949-94 Indicator and region 1949 1958 1970 1980 1985 1991 1994a Number of cities Eastern region 69 72 68 78 113 191 278 Central region 50 73 75 100 133 194 231 Western region 13 39 34 45 78 94 113 Total 132 184 177 223 324 479 622 Distribution of cities (percent)b Eastern region 52.3 39.1 38.4 35.0 34.9 39.9 44.7 Central region 37.9 39.7 42.4 44.8 41.1 40.5 37.1 Western region 9.9 21.2 19.2 20.3 24.1 19.6 18.2 Total 100.1 100.0 100.0 100.1 100.1 100.0 100.0 Urban population (millions) Eastern region 26.8 51.7 47.8 65.2 99.8 170.3 96.9 Central region 7.9 23.3 30.2 44.0 70.5 121.6 66.2 Western region 4.8 17.2 15.2 24.9 42.1 54.1 28.7 Total 39.5 92.2 93.2 134.1 212.4 346.0 191.8 Distribution of urban population (percent)b Eastern region 67.9 56.1 51.3 48.60 47.0 49.2 50.5 Central region 20.0 25.3 32.4 32.8 33.2 35.1 34.5 Western region 12.1 18.6 16.3 18.6 19.8 15.6 14.9 Total 100.0 100.0 100.0 100.0 100.0 99.9 100.0 Average city size (thousands of persons) Eastern region 390 720 700 840 880 890 350 Central region 160 320 400 440 530 630 290 Western region 370 440 450 550 540 580 250 Total 300 500 530 600 660 720 310 a. Estimated urban nonagricultural population. b. Percentages may not sum to 100 because of rounding. Source: State Statistical Bureau 1990; China: Urban Statistical Yearbook, 1992; Statistical Year- book of China, 1995. 7.9 million urbanites living in 50 small cities. The western region, though it had a handful of medium cities, was almost wholly rural. Once nor- malcy returned, there was an immediate resurgence in the growth of cities. In the course of a decade, the urban population of the eastern region doubled, with large centers such as Shanghai leading the way. The urban populace of central China increased threefold, although from a small base, but the number of cities also rose from 50 to 73, so that the average city size did not climb as much as it did in the coastal zone. Urban policy changed soon after, following a distinctive antiurban, anticoast bias. Urbanization and urban policy since the mid-1950s and before the reforms can be characterized as discouraging the growth of large cities and promoting medium and small cities, as developing in- dustrial and urban centers away from coastal areas, and as using a vari- ety of administrative measures to control the general growth and distri- bution of cities (Pannell and Ma 1983; Yeh and Xu 1990), including the recruitment of city dwellers to work in rural areas, strict controls on rural-urban migration through food rationing and household registra- tion, and the dispersion of millions of young people to work in inland 40 The Dynamics of Urban Growth in Three Chinese Cities and frontier provinces."3 Despite these controls, urban population con- tinued to grow in China, albeit slowly. The government policy of con- taining very large cities appeared to have some effect, and rural-urban migration was far below levels observed in other countries. Between 1953 and 1978 China's rate of urbanization fluctuated between 13 and 16 percent. It reached 20.6 percent in 1982 (see Zhao and Zhang 1995). Unlike many other developing countries, China has never had a full- blown primate city structure dominated by a single large city (Parish 1990). Data for 1953 and 1970 indicate that China's urban system had a very smooth profile and that cities were quite evenly distributed across various size classes (Pannell and Ma 1983). This can be attributed largely to the rapid growth of medium and small cities (figure 2.3). The very largest city, Shanghai, and the very smallest towns lagged behind, al- though the populations of Beijing, Tianjin, and some other large cities were less successfully contained. All the large cities were growing at a steady but unspectacular pace, whereas medium cities witnessed the fastest rate of population growth. The story with the small towns was peculiar. As the economic and administrative functions of small market towns declined, their numbers shrank, and many ceased to qualify as urban places. This was also intended to centralize fiscal control. The net results were a rapid decline in the number and function of cities at the lowest end and a break in the traditional link between city and country- side. On the whole, the state policy of deemphasizing the production of consumer goods and services affected small towns adversely (Chang 1994; Parish 1990). The redistribution policy stimulated the growth and development of interior centers, including Lanzhou, Wuhan, Xian, and Zhengzhou (table 2.1).'4 All cities on the coast, particularly treaty ports, had to demon- strate their contribution to the national economy as industrial producer cities. The underlying rationale was that the treaty ports still suffered from ideological and political "pollution" as a result of their former com- mercial status, foreign participation, scale of service activities, and per- sistent consumer attitudes (Pannell and Ma 1983). The impact of this inland-oriented urban policy was very large. From 1958 to 1970 govern- ment efforts to contain and even reverse urbanization in the eastern re- gion were fairly successful in dampening population growth in cities along the coastal belt. From 1966 until the early 1970s this policy was reinforced by the Third Front policy. In 1949 about two-thirds of the urban population was concentrated in the eastern region, but by 1980 this proportion had fallen to half (table 2.6). The central region experi- enced the most rapid growth of urban population and represented a third of China's urban population in 1980. Much of the industrial in- vestment that took place during these years was in capital goods indus- tries tucked away in small towns and medium cities in Gansu, Hubei, Hunan, Sichuan, and elsewhere. Growth in the absolute numbers of the urban population slowed dramatically. The only growth was in the cen- China's Changing Urban Geography: The Rise of Three Cities 41 Figure 2.3 China's Urban Population by City Size, Selected Years, 1950-90 Percentage of urban population 100 90_ 701 60i - under 0.25 20~~~~~~~~~~~~~~~~~~~~~~~3. - i i .0 401 01. 0 - 2. 1951 1955 1960 1965 1970 1975 1980 1985 1990 City size (millions of people) Note: This figure shows that, for example, the share of the largest cities (city size over 4 million) population in national urban population declined from about 33 percent in 1951 to about 15 percent in 1990. Source: Asian Population and Development Association 1991. tral region, where the populace rose by a fifth. This was also the one region in which the number of cities multiplied. Under the system of planning, cities were assigned quite narrow roles. Although they provided industrial technology and served as revenue collection points, they ceased to be financial, trade, and business centers or the foci of regional economies. Financial and information functions were transferred to Beijing. Central planners and financial officials took over much of the economic coordination that had previously been per- formed in the coastal cities. They ceased to orchestrate the conimercial activities of a region by anchoring a nested hierarchy of markets. Fur- thermore, the vertical integration of urbanized state industry and the insulated nature of communal manufacturing activities severed the long- standing organic links between cities and their hinterlands. China's great cities were forcibly cut adrift from the nourishing rural economy. Most damaging, they also lost their role as traders and contacts with the inter- national economy. A planned economy that sought self-sufficiency had no use for outward-looking commercial centers. With the start of the reform era in the late 1970s, urban policy changed substantially. First of all, the measures to control urban growth by re- 42 The Dynamics of Urban Growth in Three Chinese Cities stricting migration began to lose a great deal of their sharpness, largely as a result of rising rural prosperity and the increasing problems of en- forcement (Kirkby 1985). Second, the government again recognized the advantages of the coastal region and designated various types of growth centers, including five Special Economic Zones and 14 coastal open cit- ies (see Howell 1993). Third, the downward trends among small towns began to reverse as many old market centers sprang back to life. The relaxation of criteria for classifying towns in 1984 also led to a spurt in the level of urbanization (Lee 1989). Last, and most important of all, there was a greater willingness to allow cities and towns to respond to market forces free from government intervention (Yeh and Xu 1990). This can also be attributed to the recognition that the growth of inland cities in the period before reform was achieved at high cost and worked against the economies of scale available to large existing coastal centers. The two-decade transition from central planning has had profound consequences at many levels that are only slowly becoming apparent. The macroeconomic effects are well known and basically positive. China's rate of inflation in the mid-1990s was moderate; the trickle-down effects of high growth on poverty have been weaker than expected since the mid-1980s, and predictably, income equality has declined both within and between sectors. Perhaps more significant for the longer term is the ground swell affecting the urban and industrial geography. An almost 50-year hiatus in the urbanization of China began to come to an end in the 1980s. The push to urbanize, restrained intermittently for decades, began in earnest and promises to move China toward a "normal level" of urbanization in the years ahead. In 1993 the level of urbanization rose to 28 percent, which is close to the average for all low-income develop- ing countries. Attempts to channel urban development toward the cen- tral region and away from the large cities have also slackened. As shown in table 2.6, the share of the eastern region in the number of both cities and urban nonagricultural population has been growing again since the 1980s. It is now up to the cities themselves to contain and manage the migrant streams (see Wong 1994). Summary Starting in the mid-nineteenth century and continuing for 100 years, China entered an early stage of urbanization, drawn at first by increasing com- mercial development and, around the turn of the century, by the emer- gence of modern industry in regions that had long been hospitable to proto- manufacturing. Urbanizing tendencies were strongest in the eastern re- gion and pulled China down a path other countries had trod before. People and industry gravitated toward strategically located agglomerations along the coast and major waterways. A few cities that stood at the intersection of domestic and international markets took the lead. By exploiting the ad- vantages conferred by their hinterlands and the economies associated with China's Changing Urban Geography: The Rise of Three Cities 43 size, these cities established themselves at the top of the urban hierarchy. The central government exercised relatively little direction or control. Re- gional dynamnics were decisive. Growing cities mobilized and effectively employed resources, both domestic and external. This first stage had run its course by the middle of the twentieth cen- tury. The communist victory completely altered the economic as well as the political parameters of urban development. Foreign influence on China's urbanization was at an end. The new government did not view urbanization as either inevitable or necessarily desirable. Cities were seen not as autonomous players but as counters in a larger socialist plan of modernization that emphasized regional self-sufficiency, rural development, and limited mobility of the populace. China's premier cities were harnessed to the plan-cast in the role of handmaidens in the devel- opment of China's interior as users of resources obtained from the rest of the economy and transformed into industrial products using the special advantages conferred by agglomeration. This strategy, implemented with varying degrees of conviction, reached its high-water mark in the early 1970s, but the rules employed were not seriously undermined for almost another 15 years, by which time China was beginning to dismantle the administrative and fiscal checks imposed on its big cities and to give them the capacity to shape their own economic destinies. In the reform era, a new urban policy aimed at using large cities as the loci of regional growth began to take effect. Cities are now expected to serve as growth centers for the surrounding countryside, benefiting the entire region through econo- mies of scale, externalities, and spillover effects (Kuo 1989). The large Chinese cities have come full circle. The decade of the 1980s was a period of profound reorientation, with the economy changing gears and planning being superseded by market autonomy within a few years. First Guangzhou and later Shanghai and Tianjin were challenged by vast new opportunities. An expanding menu of choices replaced the confines of the planned system, in which the city was assigned a delim- ited role. Starting in 1978, reform began transferring responsibility from the center to provinces and municipalities. Cities were given a mandate to modernize and the powers to fulfill it. There was no body of rules on how to proceed, and the changed thinking was assimilated quite slowly, with Guangzhou, as one of the earliest modernizers, leading the way. Its reform initiatives have embraced a wide range of functions, such as building infrastructure and institutions and reviving trade and commer- cial networks. Dominated by large state sectors, Shanghai and Tianjin were much slower to reform. Shanghai's strategic industrial and fiscal position in the national economy constrained the options permitted by the central government. But after the success of reforms in south China, Shanghai was finally set free in the late 1980s, and it has embarked on a path of rapid industrial restructuring. Tianjin, the latecomer among the three in reforms, was still inching its way forward, and significant ef- forts did not commence until the early 1990s. F.~~~~.~~ Shanghai: Renaissance City y~ Shanghai is competing aggressively for the mantle of China's premier metropolis. Already the leading industrial center, it is attempting to in- crease its edge over China's other major cities by augmenting its tech- nological capability in a range of subsectors. By modernizing its port facilities and communications infrastructure, Shanghai is restoring a source of economic dynamism that contributes directly to its industrial strength. Three-quarters of a century ago, modem banking and other producer services began taking root in Shanghai, alongside the growth of manufacturing activities and the expansion of trade. Virtually elimi- nated after 1949, these are being revived once again. Last but not least, early in the century, Shanghai was the crucible in which the cultural activities associated with a modem industrial society made their ap- pearance. By all accounts, the city was the cultural capital of China in the 1920s and 1930s. Together with Beijing, Shanghai was where institu- tions of higher leamning were first established, starting a tradition that has endured and flourished since. Shanghai is again gaining a lead over other cities, drawing on the largest and most diverse pool of talent in China and reserves of cosmopolitanism that survived three decades of cultural drought. These efforts to position Shanghai in the forefront of China's major industrial centers are the result of radical changes in the outlook of ur- ban elites, who are finding a voice after years of enforced passivity thiat extended into the mid-1980s. Shanghai's renaissance also stems from a new equation withi the central government that permits Shanghai to tap into the full range of reform possibilities and to embark on a municipal development strategy that departs markedly from fte strategy of the past three decades. 44 Shanghai: Renaissance City 45 Between 1979 and 1993 Shanghai averaged a GDP growth rate of 7.9 percent a year, well below the 11.3 percent achieved by Guangzhou (see table 3.1). As a result, Shanghai's industrial ranking among China's prov- inces-second place in 1952-fell to fifth place. Since then, growth of municipal GDP has accelerated significantly and has averaged somewhat more than 14 percent a year during 1993-96, raising Shanghai's perfor- mance above the national average, approximately on a par with that of Guangzhou, but well ahead of the performance of Tianjin.' Underlying this recent growth is a series of reform measures and developmental initiatives launched after 1988 that, if sustained, should substantially improve Shanghai's fortunes. This chapter first examines the principal defining characteristics of Shanghai's development strategy through the mid-1980s. It then explores how these have been altered by a series of reforms beginning in the early 1980s but achieving a critical mass only in the 1990s. These reforms have enabled Shanghai to loosen some of the inherited constraints, to begin harnessing more fully resources within as well as outside the municipal economy, to exploit externalities inherent in the diversified municipal economy, and to remedy some of the inadequacies of the economy, es- Table 3.1 Macro Indicators for Shanghai, Tianjin, and Guangzhou, Selected Years, 1952-93 Indicator Shanghai Tianjin Guangzhou China GDP index 1952 100.0 100.0 100.0 - 1978 888.6 619.3 1,065.5 100.0 1991 2,255.4 1,599.7 4,977.5 295.0 1993 - 2,003.0 5,314.6 379.0 Annual growth rate of GDP index (percent) 1952-78 8.8 7.3 9.5 - 1978-91 7.4 7.6 12.6 8.7 1978-93 7.9' 8.1 11.3 9.3 GVIO index 1952 100.0 100.0 100.0 100.0 1978 1,216.6 913.6 1,849.9 1,659.0 1991 3,104.1 3,276.2 9,410.5 7,419.4 1993 - 5,098.6 16,486.3 12,112.3 Annual growth rate of GVIO index (percent) 1952-78 10.1 8.9 11.9 11.4 1978-91 7.5 10.3 13.3 12.2 1978-93 9.1' 12.1 15.7 14.2 - Not available. Note: GDP, gross domestic product; GVIO, gross value of industrial output. a. Annual average rate between 1979 and 1993. Source: State Statistical Bureau, Statistical Yearbook of Shanghai, 1992 and 1994; Statistical Yearbook of Tianjin, 1994; Statistical Yearbook of Guangzhou, 1992 and 1994; Statistical Yearbook of China, 1994. 46 The Dynamics of Urban Growth in Three Chinese Cities pecially in the sphere of producer services and infrastructure (on pro- ducer services, see Healey and Ilbery 1990). Shanghai's size and the complex, multifaceted nature of its modern- ization allow us to examine its development from the angles indicated in chapter 1, including industrial development, producer services, ex- ternal trading and investment links, fiscal relations, and R&D capabili- ties. When we turn to Tianjin and Guangzhou in later chapters, we take a more selective approach that highlights the role of critical sectors and industries. Industrial Transition: From the 1940s to the 1960s In most respects, Shanghai in the late 1940s was typical of large indus- trial cities in a developing country. Textiles, food processing, and a hand- ful of other light industries dominated the manufacturing sector and, by way of backward linkages, supported small-scale engineering and metalworking subsectors. Service industries provided the bulk of em- ployment, whether in formal activities such as banking or in informal ones such as petty retailing (Howe 1981; Sung 1991). Because Shanghai was China's main port, trade was a vital component of the economy. Local merchants managed the city's own commerce and served as inter- mediaries for the external trade of the lower Yangtze region, which was funneled largely through Shanghai. The interactions between hinterland and city, production linkages, and trading ties were numerous and strong. Even when warfare compromised market functioning, interfered with the flow of goods, and caused Shanghai's industrialists to flee to Hong Kong with as much of their equipment as they could ship, Shanghai's basic economic configuration and its role in the regional economy changed very little (Howe 1981). But after 1949, the turn to- ward centralized planning and the placement of increasing emphasis on heavy industry, regional self-sufficiency, and minimal reliance on foreign trade began to transform the character of China's large coastal cities, in particular, Shanghai. One major tenet of Maoist socialism was that cities must be cast in the mold of producers rather than of consumers living off the surplus of the rural economy.2 The practical implications of this were far-reaching and involved a large increase in the share of manufacturing relative to services. Industry was expanded because it was viewed as productive, whereas services, which in the Marxist scheme contributed little to the economy, were reduced in scale. Furthermore, heavy industries were tar- geted for expansion because overriding importance was attached to met- allurgical products, machinery, and petrochemicals. Heavy industry was viewed as the touchstone of economic strength and consequently received the largest allocations of capital through the state plan (see Donnithome 1967; Riskin 1987). Starting in the mid-1950s, Shanghai's industrial center Shanghai: Renaissance City 47 of gravity began to shift away from light manufacturing and toward rma- chine building, a process that continued for the next two decades. The stress on heavy industry was combined with an equally strong impetus toward self-sufficiency in which the widest possible range of subsectors was developed so that the maximum number of input-output relationships could be contained within Shanghai. In fact, an enormous range of industries, spanning 140 of the 146 listed subsectors, was established in order to attain the highest measure of industrial independence. Almost all of the growth took place in the state sector, and the central government rather than the municipality con- trolled state enterprises. Shanghai not only acquired a strong base of heavy industry and a highly diversified industrial structure but also sur- rendered much of the industrial autonomy it once enjoyed, becoming a creature of the planned economy dedicated to achieving narrow pro- duction goals to the virtual exclusion of all else. The pronounced strain of autarky in national policy also meant that earlier trading ties with the international community were largely severed, and the focus of the city turned inward toward the domestic economy. Each of these trends calls for further elucidation. Socialism's Industrial Pillar Shanghai was one among a handful of Chinese cities in the early 1950s that had a functioning modern industrial sector, the supporting infra- structure, and the necessary fund of skills. These cities were the natural candidates to spearhead China's industrialization along socialist lines. Hence the first goal of the central government was to build state-owned heavy industries in Shanghai and to use their production of capital equip- ment as well as intermediate products to initiate industrial growth in other parts of the country. Because machinery alone was insufficient to induce development, thousands of Shanghainese found themselves transferred to exceedingly remote parts of the country, sometimes to train the locals in manufacturing techniques and frequently to run the newly built factories. The accumulation of human capital from Shanghai's universities and nu- merous production establishments was China's key to higher technology, which became all the more valuable when the Soviets withdrew their tech- nicians in 1960 and halted scientific exchanges. Forced to rely on its own relatively meager fund of manufacturing skills and modest scientific base, China began cultivating Shanghai's research potential and using it to ad- vance a number of major industrial projects. Likewise, Shanghai's labora- tories were entrusted with the task of building some of China's ballistic missiles and transport aircraft. In the process, Shanghai added a military wing to its industrial complex, which also contributed to the production of satellite launchers and the development of aircraft (Norris 1994; Aviation Week and Space Technology, March 24, 1997). 48 The Dynamics of Urban Growth in Three Chinese Cities Planned economies committed to maximizing growth using heavy industry as the leading sector have an insatiable hunger for resources. China was no exception. The industrial engine constantly demanded capital, but raising such a large volume of resources in a poor and pre- dominantly rural economy was administratively problematic. The Chi- nese government solved this by using the pricing of industrial products (and also agricultural staples) as the main vehicle for extracting rev- enue. State and collective enterprises manufacturing capital equipment but also consumer durables earned sizable profits by selling their goods at high prices that were fixed by the state, while receiving raw materials at prices that were kept deliberately low. Central authorities then ap- propriated these profits, which constituted three-quarters or more of central government revenue. Shanghai, with its concentration of indus- trial products, became the largest single source of revenue for the state, providing about 25 percent in an average year during the latter part of the 1970s. In fact, nine-tenths of Shanghai's enterprise earnings were not controlled by the municipality, which remained perennially cash- poor and dependent on earmarked funds that central ministries sup- plied for industrial units or research facilities. For example, over the 25- year period from 1953 to 1978, the city devoted Y18.73 billion to capital construction, but central ministries provided more than 70 percent of this for selected state enterprises. For the central authorities, Shanghai served as a pillar of the planned economy. Its manufacturing facilities, among China's most advanced, were crucial for meeting production targets. Shanghai's human resources were instrumental in igniting industrialization elsewhere in China and in making possible a modicum of technical progress. In fulfilling output goals, the city also acted as the state's revenue collector; the pricing sys- tem ensured the profitability of Shanghai's state-owned enterprises, and the central government then exercised its rights to siphon the funds into the budget. Preparing the Ground for Reform Prior to liberation, Shanghai was China's foremost trading city and the largest recipient of foreign investment. Trade and foreign capital intro- duced Chinese entrepreneurs to the industrial technology and producer services essential to a modern economy. The coming of socialism pulled Shanghai into a lower orbit. It ceased to be a "world city" with diverse functions and became instead an industrial workhorse for a plan-bound economy. Its growth rate was constrained first by the limitations on in- vestment in light industry and the supposedly "unproductive" services and next by restraints on migration to the city, as well as the transfer of thousands of engineers and technicians to other parts of China. With industry soaking up the lion's share of capital, little was left for urban infrastructure, and the austere official attitude toward the development Shanghai: Renaissance City 49 of large cities provided little scope for investment in housing, transport, and other urban amenities. Tightly laced socialism turned Shanghai into a somber and congested city, frugally supplied with services and forced to subsist on infrastructure inherited from the early decades of the cen- tury. These lean years were helpful in at least four respects. First, in the industrial sphere, Shanghai acquired industrial diversity and produc- tion experience to an extent that has no parallels in China and few any- where in the world. This was excellent preparation for the reform era and beyond. Second, the embedding of defense-related activities within the municipal industrial system stimulated entry into high-technology activities such as electronics, precision machining, computers, and other relatively advanced subsectors. As was the case with Los Angeles, Sao Paulo, and the western suburbs of London, defense industries provided substantial direct production benefits and knowledge-related externali- ties.3 Moreover, the manufacture of weaponry and aircraft called for research, and this brought into existence a handful of well-equipped laboratories with pilot production facilities and state-of-the-art testing equipment. In a limited but meaningful way, it also established links between research centers and state enterprises along the lines that have proved so fruitful in the United States. Third, the importance of Shang- hai as an industrial center and one of the bulwarks of military research sheltered the city from the turbulence and dislocation that plagued the country during the Cultural Revolution. Fourth, the years of high socialism significantly added to the fund of production, technical, and scientific skills. Undoubtedly Shanghai was taxed each time the central government sought to disperse industry and set up new industrial centers. But the multiplication of industrial enter- prises, and tertiary-level educational institutions and research establish- ments, suggests that the state partially compensated Shanghai for the drain. The city responded sluggishly to the opportunities presented by reforms-and we deal with the reasons for this later-but the speed of diversification in the 1990s and the surge in industrial growth are irre- futable indicators of pent-up economic energy somewhat belatedly re- leased by a well-timed succession of reform policies. The Scale of Agglomeration Whether experienced firsthand or perceived through myriad statistics, the size and sheer density of Shanghai's population are impressive. The scale is economic rather than geographic. The city core covers just 300 square kilometers out of a municipal total of 6,341 square kilometers. In 1980 the metropolis had a population of 11.5 million. This had risen to almost 13 million in 1993,8.9 million of whom resided in the city proper, where the density exceeded 43,000 persons per square kilometer (see table 3.2). None of the other crowded cities in the developing world, 50 The Dynamics of Urban Growth in Three Chinese Cities Table 3.2 Major Indicators for Shanghai, Tianjin, and Guangzhou, 1993 Shanghai Tianjin Guangzhou Indicator Number Percent Number Percent Number Percent Land Land area 6,341 n.a. 11,305 n.a. 7,434 n.a. (square kilometers) Built-up area 300 n.a. 339 n.a. 207 n.a. Population Population (millions) 12.95 n.a. 8.86 n.a. 6.20 n.a. Nonagricultural population 8.93 n.a. 5.00 n.a. 3.75 n.a. Workforce Total workforce (millions) 7.85 100.0 5.03 100.0 3.85 100.0 Staff and workers, 4.90 62.4 2.93 58.3 2.11 54.7 State employees 3.52 44.9 2.07 41.0 1.39 36.0 Primary sector 0.75 9.6 0.89 17.7 0.95 24.8 Secondary sector 4.55 58.0 2.45 48.7 1.46 38.0 Tertiary sector 2.55 32.4 1.69 33.6 1.43 37.2 Average wage (yuan) 5,650 n.a. 4,003 n.a. 6,342 n.a. Economic indicators GDP (billions of yuan) 151.16 100.0 53.61 100.0 74.08 100.0 Primary sector 3.82 2.5 3.54 6.6 4.76 6.4 Secondary sector 90.03 59.6 30.25 56.4 35.17 47.5 Tertiary sector 57.31 37.9 19.82 37.0 34.15 46.1 GDP per capita (yuan) 11,699 n.a. 6,075 n.a. 11,989 n.a. GVIAO (billions of yuan) 341.63 n.a. 147.25 n.a. 121.96 n.a. GVIO (billions of yuan) 1993 286.10 100.0 106.58 100.0 100.25 100.0 State-owned enterprises 155.95 54.5 62.80 58.9 46.53 46.4 Collective-owned enterprises 35.89 12.5 24.13 22.6 16.22 16.2 Others 94.26 32.9 19.66 18.4 37.50 37.4 1994 375.80 100.0 129.38 100.0 - - State-owned enterprises 180.29 48.0 64.81 50.1 - Collective-owned enterprises 48.39 12.9 23.01 17.8 - - Others 147.13 39.2 41.56 32.1 - - n.a. Not applicable. - Not available. Note: GDP, gross domestic product; GVIAO, gross value of industrial and agricultural out- put; GVIO, gross value of industrial output. a. Excludes workforce employed in rural establishments. Source: State Statistical Bureau, Statistical Yearbook of Tianjin, 1994; Statistical Yearbook of Guangzhou, 1994; Statistical Yearbook of China, 1995. such as Calcutta, Dhaka, and Sao Paulo, conveys the extraordinary sense of teeming humanity as does downtown Shanghai during peak hours. Industry has been and remains the lifeblood of the municipality. Shang- hai accounted for 11.6 percent of the national industrial output in 1980. This declined to a still impressive 6.9 percent in 1994, with the bulk of manufacturing facilities crowded into roughly a quarter of the city's built- up area. Shanghai: Renaissance City 51 Table 3.3 Gross Value of Industrial Output in Shanghai, Selected Years, 1985-93 1985 1991 1993 Enterprises Billions of Billions of Billions of in 1993 Indicator yuan- Percent yuanb Percent yuan5 Percent Number Percent City total 82.96 100.0 179.61 100.0 298.98 100.0 13,688 100.0 Light industry 46.15 55.6 89.16 49.6 122.25 40.9 7,937 58.0 Heavy industry 36.81 44.4 90.45 50.4 176.73 59.1 5,762 42.1 State-owned enterprises 64.29 77.5 126.31 70.3 162.97 54.5 4,270 31.2 Collective-owned enterprises 16.01 19.3 25.21 14.0 38.47 12.9 6,762 49.4 Others 2.65 3.2 28.10 15.6 97.55 32.6 2,656 19.4 Top subsectors in 1993 64.31 77.5 136.92 76.2 232.85 77.9 9,228 67.4 Metallurgy, 8.81 10.6 27.36 15.2 61.35 20.5 290 2.1 Machine buildingd 16.12 19.4 30.74 17.1 48.56 16.2 3,066 22.4 Transport equipment 2.80 3.4 11.24 6.3 28.07 9.4 638 4.7 Textiles 14.65 17.7 22.19 12.4 25.29 8.5 1,119 8.2 Chemicals 5.74 7.0 12.91 7.2 17.20 5.8 700 5.1 Electronics 4.58 5.5 8.39 4.7 12.69 4.2 436 3.2 Foode 2.61 3.2 6.37 3.5 10.66 3.6 789 5.8 Metal products 3.02 3.6 6.21 3.5 10.38 3.5 1,205 8.8 Synthetic fiber 3.21 3.9 5.99 3.3 9.42 3.2 51 0.4 Clothing 2.76 3.3 5.53 3.1 9.24 3.1 934 6.8 Herfindahl index 0.077 0.067 0.071 Note: All numbers exclude village-run enterprises. a. 1980 prices. b. Current prices. c. Includes smelting and pressing of ferrous and nonferrous metals. d. Includes machine building and electrical machinery. e. Includes food processing. Source: State Statistical Bureau, Statistical Yearbook of Shanghai, 1992, 1994. Metallurgy, machine building, and transport equipment are Shanghai's biggest subsectors. Together they were responsible for 46 percent of the gross value of industrial output (GVIO) in 1993 (see table 3.3). Behind them, in order of importance, are textiles, chemicals, and electronics. But as noted above, Shanghai's industrial reach is quite extraordinary and extends into virtually all subsectors of any consequence. Compared with the situation in 1985, two changes in the structure of production are noteworthy: textiles had lost ground and were being overtaken by metallurgy and machine building, and transport equipment had almost tripled its share (see figure 3.1). In addition, the degree of concentration of industrial subsectors had decreased somewhat since 1978, as demon- strated by the shrinking Herfindahl index measured in output value (from 0.077 to 0.071). Light and heavy industry were evenly balanced in the late 1970s-each accounting for half of Gvio-and remained so for 52 The Dynamics of Urban Growth in Three Chinese Cities Figure 3.1 Industrial Structure of Shanghai, 1985 and 1993 Clothing Synthetic fiber 1 1993 Metal products 0I 1985 Food Electronics Textiles _ Chemicals Transport equipment Machine building Metallurgy 1 Other l 0 5 1 0 1 5 20 25 30 Percentage of GVIO Note: GVIO, gross value of industrial output. Source: Table 3.3. 13 years, until 1991. However, state-owned enterprises that had claimed more than 90 percent of industrial output saw their share fall to 55 per- cent in 1993 (and 44 in 1995), with the remainder being equally divided between collectives and others, principally joint ventures, which have mushroomed since the mid-1980s. On closer examination, the dynamics of industrial change in Shang- hai reveal the working of significant agglomeration effects and the emer- gence of noteworthy trends. By Chinese standards, Shanghai's state- owned enterprises were among the best equipped and most efficient producers in the country. As reform transformed the economic milieu, they found it relatively easy to introduce new product lines, upgrade their output, and restructure production. Once foreign investors began to explore seriously the possibilities of manufacturing in China, those interested in the assembly of light manufactures concentrated on the southeastern provinces, while large multinationals were more attracted to Shanghai. The wealth of skills is just one factor. Equally important, especially for the production of items such as automobiles and comput- ers, is that a joint venture which is prepared to subcontract, breaking from the long-standing practice of vertical integration, can, with some experimentation and training, source a wide range of components from within the municipality. This took on considerable significance as the government required multinational corporations to aim for a high level of local content. Such diversity would be a plus under any circumstances, but it had special merit in a country where transport and communica- tions facilities were backward in the 1980s and, in spite of feverish in- vestment, remain inadequate in the late 1990s. Shanghai: Renaissance City 53 Industrial depth has been a huge asset for Shanghai, generating exter- nalities and also inducing foreign investment in electronics, chemicals, and transport equipment industries. The entry of foreign companies and the subsequent focus on growth, increased industrial competition from other parts of China, growing exposure to the international economy, and greater scope for nonstate enterprises to expand their operations and compete with state-owned enterprises for market share gradually induced a spiral of growth feeding off positive agglomeration effects that had not yet been fully tapped. The lagged nature of the response is related to the pace of reforms, but once change began, it covered a broad spectrum, drew on Shanghai's advantages, and attempted to cope with its serious handicaps. Automobiles and Fashion Garments: Externalities at Work Over time Shanghai will find it expedient to narrow the compass of its industrial activities and concentrate on fewer leading sectors that will provide the sources of growth well into the next century. Undoubtedly, Shanghai will also diversify into services and in all probability will be the foremost financial center in China. However, the experience of New York suggests that prosperity built on producer services alone is pre- carious and distributionally lopsided (see chapter 6). Abroad economic base including advanced manufacturing and services is the most dy- namic. Among possible industrial leading sectors, Shanghai has the luxury of choosing from among a range of candidates because it is cur- rently well positioned in so many subsectors. Autos, quality apparel, office equipment, telecommunications, pharmaceuticals, and medical equipment are growth industries that have taken root in the municipal- ity. They are all skill- and research-intensive activities peculiarly suited to Shanghai's multifaceted resource endowment. But they by no means exhaust the range of options, which also include heavy machinery, ma- chine tools, and metallurgical industries. If Shanghai perseveres with the development of the auto industry, as is highly probable, metallurgi- cal and machine tool subsectors will be encouraged to invest and to refine their technology. The apparel subsector will build on a textile industry with long experience. Likewise, a robust pharmaceutical in- dustry will promote fine chemicals, and the manufacture of advanced medical equipment will have backward linkages to other chemical industries, optoelectronics, as well as precision engineering. Abrief review of the auto sector and the garment industry offers a taste of how two lead- ing sectors are faring and of the factors that will influence their future. Automobiles At the start of the 1980s, fewer than 2 million autos were on China's roads (output of cars and trucks in 1979 was about 120,000; see Harwit 1995). As a ratio to the population, the number of automobiles was among 54 The Dynamics of Urban Growth in Three Chinese Cities the lowest in the world. The number rose to 4.3 million in 1988 and to more than 10 million in 1995, with passenger cars accounting for close to 3 million of the total. In spite of this steep increase, the number of passenger cars per 1,000 persons was little more than 2.0, compared with 86.2 in Brazil (Maxton 1994; Netv York Times, March 28, 1995). As long as the economy maintains a healthy rate of expansion, the scope for con- tinued growth is substantial. Even if the number of registered cars rises to the projected 4.5 million by 2000, China will have only 3.4 cars per 1,000 persons. China stands on the threshold of automobility. The po- tential demand, looking a decade and more into the future, is huge, and the current scale of the auto sector is small by comparison. Moreover, many of the forward linkages to auto-related services have barely been tapped, suggesting that future direct and indirect employment effects will dwarf the number of jobs created in core manufacturing activities. Although there are more than 160 auto assembly facilities in China, the three front-runners for passenger cars are Changchun (First Auto Works-Audi), Shanghai (Shanghai-Volkswagen), and Wuhan (Magic Dragon-Citroen). In addition, there are three lesser production centers in Beijing (Beijing-Cherokee Co.), Tianjin (Tianjin-Daihatsu Corp.), and Guangzhou (Guangzhou-Peugeot). Total production reached 1.3 million vehicles in 1994, of which about 250,000 were passenger cars. Shanghai has a commanding lead in sedans, with 52 percent of the national out- put of 385,000 sedans in 199 6.4 Volkswagen entered into a 50-50 joint venture with Shanghai Auto and the China Auto Industry Corporation in 1984 with the intention of producing its Brazilian-designed Santana. After a trial assembly from knockdown kits, production commenced in 1985 and built up to 18,000 units in 1990, when 50 percent of the value of the vehicle was being sourced domestically. Production rose to 200,000 in 1996, when 90 per- cent of the value of the Santana was domestically sourced. This consid- erable achievement highlights Shanghai Auto's success in solving a vari- ety of teething problems and serves to identify where the performance of the local manufacturing system falls below international standards. Whenever the leading international automakers are making a fresh start, they prefer a greenfield site and a youthful, well-educated labor force with no experience in auto manufacturing. This ensures that facili- ties incorporate the latest knowledge of plant layout and process tech- nology. An educated but inexperienced supply of workers combines learning skills with the absence of ingrained, possibly dysfunctional habits and internalized organizational routines. In short, the best employ- ees are easily trained and flexible in what they do as well as how they do it. When Volkswagen commenced operations in Shanghai, it had neither of these advantages. Existing facilities had to be adapted or rebuilt, and a sizable part of the workforce was composed of the tenured employees of Shanghai Auto Industry Corp., a state enterprise founded in 1958.5 Shanghai: Renaissance City 55 Volkswagen acquired two assets of great significance in China's ur- ban industrial environment. First, it was linked to a partner with the right bureaucratic and industrial connections. These connections brought access to officialdom, won favors, and greased the wheels of business. They also served as an entry into long-term relations with future suppli- ers. The automobile business depends on efficient coordination between a few assemblers and many suppliers. Ten years ago, most major assem- blers were served by hundreds of suppliers, but experience has shown that the ideal is to have 300 or fewer principal suppliers. Assemblers are now seeking to establish broadly cooperative relations with suppliers but at the same time are anxious to sustain a measure of competition among component manufacturers to keep prices down and maintain quality. Striking a balance between these goals calls for delicate foot- work. The link with Shanghai Auto made it possible for Volkswagen to begin creating a local network of suppliers, starting with the 107 en- terprises related to Shanghai Auto. The Shanghai Santana Common- wealth now embraces not just input suppliers but also universities and research institutes that, with encouragement from municipal authorities, are committed to the progressive deepening of the auto industry in the city. Second, ShanghaiAuto gave the joint venture access to a pool of engi- neering skills, which are invaluable for undertaking a complex manu- facturing operation and for transferring process technology in a short space of time. Volkswagen augmented these skills by sending more than 1,000 Chinese engineers overseas (through 1993) for training. Beyond that, association with a major state-owned enterprise loosened the con- straints on credit from local banking channels and, in the late 1980s, made it easier to bid for foreign exchange. Ten years ago it was difficult to predict how the marriage between an old-line state-owned enterprise and one of the world's premier auto companies would evolve. Urban reform had barely started, and Shang- hai had yet to loosen the state's tight grip on industry. Socialism as prac- ticed by China's state-owned enterprises was far removed from the world of Volkswagen. To fuse such dissimilar cultures and systems into a work- able hybrid posed a challenge for the partners and also for the munici- pal govemment. If the experiment failed, Shanghai stood to lose the investment of the more tentative multinationals waiting in the wings, and Volkswagen would not gain a prized foothold in one of the largest markets of the next century. The municipality's desire to reach a high level of domestic content within a handful of years complicated Volkswagen's task and made it imperative to concentrate from the out- set on the manufacture of components, the bedrock of the auto industry. By dint of trying, the joint venture succeeded in a remarkably short span of time and in so doing underscored the roles of the govemment, the economy of Shanghai, and the approach of the foreign partner. 56 The Dynamics of Urban Growth in Three Chinese Cities By assigning a high priority to the auto industry, the government made it a focus of ministerial interest and municipal commitment. It also pro- vided tariff protection along with a guaranteed market for the vehicles produced during the initial years of operation.6 The combination of bu- reaucratic support, investment funds, and large profits lowered the odds to acceptable levels. Shanghai could not offer a greenfield site, but in the Chinese context it was the ideal place to set up a network of suppliers within a short trucking distance of the assembly plant. Volkswagen dis- covered that there were enterprises in the metropolitan region with the skills to produce almost any item in small lots. With sustained coaching and technical assistance, these firms could learn the design and process skills they needed to participate effectively in a modem automobile pro- duction network. Building such a network with some infusion of for- eign capital into the parts industry took several years, and it is still in- complete! It was a hard slog to win adherence to acceptable standards in the virtual absence of competition, but the industrial strength of the city helped. By 1995, 180 suppliers, many of them joint ventures, were able to produce all but the most technically demanding components for the Santana.' Many are still unable to meet Volkswagen's specifications to the fullest extent, but progress has been steady. Shanghai-Volkswagen was not empowered to recruit an all new workforce, but the city's labor force had the highest level of educa- tion and skills in China. These advantages partially offset other draw- backs. Moreover, the auto industry had a plentiful supply of engi- neers with whom to augment the quality of the assembly-line workforce, troubleshoot, control quality, and repair equipment. The ratio of-engineers to the total workforce is viewed as an indicator of efficiency, and in this regard Shanghai-Volkswagen was operating in one of the best labor markets in China. Such skills are not just an asset on the assembly line; they also underpin R&D capability throughout the supplier network. An early start and a wise selection of product have maximized the gains for Volkswagen. Because it dominates the market for sedans, Shanghai Auto Industries achieved annual sales of $4 billion in 1993, giving it a financial lead over rivals and putting it ahead of other giants such as Daqing Petroleum and Anshan Steel ($3.8 billion and $2.5 billion, re- spectively). Selecting a mature design-the Santana-that had been thor- oughly debugged and customized for a developing-country market smoothed away innumerable problems at the stage of assembly, eased the task of manufacturing components, and simplified the setting up of engine manufacturing in Shanghai (Shaiken 1991). As other producers struggle to launch a first generation of cars, Volkswagen is ready to move into the next generation, having invested $300 million and spent the past decade accumulating hard-won experience with manufactur- ing in China's state sector.9 This firm will face competition on its home Shanghai: Renaissance City 57 turf from General Motors, which in 1997 entered into a $1.57 billion con- tract with Shanghai Auto Industry Corp., to produce Buick sedans. Fashion Garments The car industry is one of Shanghai's new leading sectors that are mak- ing good use of the agglomeration economies at hand. At the other end of the industrial spectrum is a subsector that still has plenty of life but must take some strong medicine if it is to remain competitive. Shanghai once was China's largest textiles producer. With almost 500 mills em- ploying close to 500,000 workers and $2.4 billion in exports in 1992, it is second only to Guangdong."0 But to survive, Shanghai must move up- scale and into high-quality apparel. Modern high-technology production of apparel is research- and information-intensive, requiring, if not proximity to markets, then con- tinuous feedback on demand and shifting tastes. It puts great store in design, type of material, quality, and finish. Sophistication in design has special importance, and the application of computer techniques has raised the technological stakes. Improvements in design must go hand in hand with careful choice of material and access to modem dyeing and finishing facilities if clothing is to appeal to exacting customers in exceedingly competitive markets. Competition is also responsible for the increasing specialization of apparel producers and suppliers of equip- ment. The continuous need to change weave, texture, and appearance calls for an interaction between designers and marketers, makers of cloth- ing, and manufacturers of equipment.'" The business relies on the abil- ity to respond quickly to shifts in the market and to deliver lots of the desired scale with a lead time of a few months. The trend is toward greater capital intensity at virtually every stage, including design and cutting. Expensive, highly specialized machinery is essential for finish, and mechanization also serves to hold down labor costs, although these costs are of less consequence in China, where wages are relatively low. The information system that links retailer to producer is tied to new communications technology. Warehousing and distribu- tion are becoming inseparable from computer networks, and tight in- ventory control is one of the main determinants of competitiveness.12 Recognizing that survival depends on raising value added, Shanghai's producers are shifting to apparel, which now constitutes 80 percent of exports. Local designers have appeared and established themselves through regular fashion shows and advertising.'3 The large metropoli- tan market, China's most sophisticated, is a definite advantage, as is the vertically integrated production system, rich in skills of every sort. Nev- ertheless; only the most efficient and innovative producers will survive. Old facilities, an aging workforce, weak marketing skills, and a fledg- ling infrastructure to cope with design and marketing are all trouble- Table 3.4 Foreign Trade in Shanghai, Selected Years, 1978-93 1978 1980 1985 1990 1991 1993 Millions Millions Millions Millions Millions Millions of U.S. of U.S. of U.S. of U.S. of U.S. of U.S. Indicator dollars Percent dollars Percent dollars Percent dollars Percent dollars Percent dollars Percent Total exports 2,892.6 100.0 4,266.4 100.0 3,360.7 100.0 5,317.3 100.0 5,739.8 100.0 7,381.8 100.0 Light industry 1,636.0 56.6 2,325.0 54.5 1,966.0 58.5 3,388.0 63.7 3,451.0 60.1 - - Heavy industry 355.0 12.3 813.0 19.1 628.0 18.7 1,268.0 23.9 1,561.0 27.2 Agriculture 902.0 31.2 1,128.0 26.4 767.0 22.8 661.0 12.4 728.0 12.7 - Export destinations Asia 1,363.4 47.1 1,914.5 44.9 1,475.4 43.9 2,557.2 48.1 2,955.1 51.5 4,175.4 56.6 Africa 289.6 10.0 326.8 7.7 151.8 4.5 138.2 2.6 163.3 2.8 204.8 2.8 Europe 915.2 31.6 1,361.8 31.9 850.7 25.3 1,142.7 21.5 1,108.7 19.3 1,072.8 14.5 United States 233.5 7.7 481.0 11.3 614.0 18.3 957.1 18.0 1,059.6 18.5 1,542.9 20.9 o Others 101.0 3.5 182.2 4.3 268.8 8.0 522.1 9.8 453.2 7.9 386.0 5.2 Total imports, 132.8 100.0 239.9 100.0 1,812.9 100.0 2,110.5 100.0 2,303.8 100.0 5,349.9 n.a. For production 111.8 84.2 156.9 65.4 1,620.8 89.4 2,027.9 96.1 2,130.0 92.5 - n.a. For consumption 21.0 15.8 83.0 34.6 192.1 10.6 82.6 3.9 173.7 7.5 - n.a. Commodity trade balance 2,759.9 n.a. 4,026.5 n.a. 1,547.8 n.a. 3,206.9 n.a. 3,436.1 n.a. 2,031.9 n.a. n.a. Not applicable. -Not available. a. Includes only imports purchased with municipal foreign exchanges. Source: State Statistical Bureau, Statistical Yearbook of Shanglhai, 1992, 1994. Shanghai: Renaissance City 59 some handicaps. Parts of the industry will not be able to muster the needed capital and skills and will be squeezed out, although the silk industry, for instance, could consolidate its position as a leading sector. China commands 70 percent of the world market for silk and has few competitors in a wide range of products. Diseconomies: State-Owned Enterprises and Infrastructure Even though Shanghai grasped reforms by slow degrees, the city pro- vides rich soil for fresh entrepreneurship. Collective enterprises have enlarged their share of output and provided the local state-owned en- terprises with their first dose of serious competition. Skills and indus- trial traditions, not to mention the pull of a vibrant commercial history, are the basis for the spread of producer services-banking, finance, con- sulting, and information-that are contributing to industrial efficiency and providing an additional impetus to growth of the metropolitan economy. Clearly, agglomeration effects, once released by reforms, can powerfully reinforce the conventional sources of growth, although in Shanghai's case the disabilities of a sprawling state sector and the pov- erty and attendant congestion of the transport infrastructure undermine the advantages of agglomeration. Shanghai's state-owned enterprises include some of the most dynamic industrial entities in the country. Through their own efforts and with the help of imported technology, these enterprises are gradually draw- ing abreast of their competitors in East Asia in terms of productive effi- ciency and quality. Shanghai's exports nearly tripled between 1978 and 1993, and much of the $7.4 billion in goods exported in 1993 was pro- duced by the state sector (see table 3.4) .14 But the average state enterprise is in poor health and struggles with serious disabilities (in 1996, about 30 percent were losing money). Many enterprises have been in existence for two or three decades, their cramped facilities in the central city are frequently run-down, and their equip- ment is technologically obsolescent. The practice of providing the workforce with tenured employment, guaranteed pensions, and health, housing, and other benefits saddles each state-owned enterprise with overheads that are difficult to sustain as the shift to market-based pric- ing, rising interest costs, and the press of competition narrow their profit margins. Overstaffing, which inflates labor costs, and social security for an ever-growing pool of pensioners would be a serious drag on firms in any country, but in China, they compound the problems inherent in outdated management practices, excessive vertical integration,"5 the or- ganization of the manufacturing process, work ethics, and relations with supervisory municipal industry bureaus. Such systemic disabilities are responsible for the financial losses of state-owned enterprises, which amounted to over half of total industrial losses in 1991 (see table 3.5). 60 The Dynamics of Urbarn Growth in Three Chinese Cities Many state-owned enterprises have managed to revamp their opera- tions and achieve financial health, but the persistence of soft budget constraints discourages initiatives, the tax system inhibits enterprises from striving for profitability, and the political economy of labor rela- tions induces firms to honor their tenurial commitments to workers (Rawski 1992; Jefferson and Rawski 1994). The social consequences of large layoffs also dissuade municipal authorities, which own most of the state-owned enterprises, from pressing hard for a reduction of the enterprise workforce. The actual and potential benefits of a many-layered industrial sector are manifold, but they are undermined by the continued preponder- ance of state-owned enterprises that have been slow to respond to the challenge of modernization and have imposed a heavy fiscal burden on the municipality. Enterprise reform, by changing the incentive regime, is forcing state-owned enterprises to become competitive, but it will be many years before the legacy of state-owned enterprise problems will be whittled down to insignificance.16 A partial solution may be found in an experiment launched in 1995 that created a mergers-and-acquisitions market in which assets from state-owned enterprises can be sold to the highest bidder. It is likely that contracts will constrain buyers from shut- Table 3.5 Loss-Making Industrial Enterprises in Shanghai and Tianjin, 1987-94 Loss-making Total losses Losses of state-ozvned enterprises Percentage enterprises Percentage of net value Percentage City of total Billions of industrial Billions of total and year Number enterprises of yuan output of yuan losses Shanghai 1987 674 7.4 0.12 0.4 0.03 25.9 1988 660 6.8 0.23 0.7 0.12 53.1 1989 1,127 11.0 0.57 1.5 0.37 64.5 1990 1,793 17.7 1.47 3.8 1.01 68.3 1991 1,594 15.6 1.35 3.0 0.75 55.6 1992 1,632 19.2 1.15 2.1 - - 1993 2,029 20.1 1.69 - 0.65 38.5 1994 2,818 19.6 3.76 - 1.88 50.0 Tianjin 1987 600 12.3 0.13 1.3 0.08 64.3 1988 551 10.8 0.19 1.6 0.11 57.8 1989 776 14.5 0.69 5.6 0.57 82.0 1990 1,032 19.0 1.11 8.5 0.86 78.0 1991 999 18.0 1.20 8.8 0.85 71.0 1992 1,131 19.1 1.50 9.8 - - 1993 1,990 10.4 2.08 - 1.48 71.2 1994 2,579 22.9 1.79 - 0.88 49.2 - Not available. Source: State Statistical Bureau, Yearbook of Industrial Statistics, various years; Statistical Year- book of China, various years. Shanghai: Renaissance City 61 tering factories and laying off workers. At this stage, the exchange is not open to foreigners, but there are plans to include them in the future. Reinforcing this experience, a program introduced in 1996 aims to con- vert 80 percent of all state-owned enterprises into limited-liability com- panies over a five-year period (Business Week, March 13, 1995; "Shang- hai," Oxford Analytica, June 4, 1996). The state of urban infrastructure, which was starved of capital through- out the Maoist era, makes it hard to realize the full extent of agglomera- tion economies.7 As table 3.6 indicates, per capita paved roads and liv- ing space in Shanghai were lower than the national urban average in 1990. Shanghai was ahead of Tianjin on all but one count, Tianjin having more than twice the area of paved roads as Shanghai and two-thirds more than Guangzhou by 1994. Shanghai also compares favorably with Guangzhou in the availability of public transport. But Guangzhou is better endowed than Shanghai and Tianjin in many respects, particu- larly per capita living space and consumption of water and electricity. While the comparative perspective shows Shanghai in one light, the absolute numbers are much more telling. Living conditions in China's cities are spartan, and both transport and communications facilities fall far short of standards reached by other East Asian countries (see table 3.7). The housing shortage reflects the steady growth of population, which has outpaced the expansion of urban areas and the determined effort to renovate older living quarters, mainly row houses and bunga- lows, and to construct apartment buildings on the outskirts of the city."8 A significant housing reform plan was launched in 1991, introducing a new mechanism of housing finance. It stipulated that the city, employ- ers, and employees would all contribute to a compulsory fund. Other elements of the housing reform scheme included increase in rent, use of housing coupons, discounts for buyers of housing units, and establish- ment of a housing commission. Shanghai's plan was much more com- prehensive than plans in other cities, and it has attracted attention across China (Wang and Murie 1996). There has been some further progress since the start of housing reform during the early 1990s. For instance, a Y3.4 billion savings fund had been raised to finance new housing con- struction, the rebuilding of old endangered housing, and mortgage loans for commercially produced units.19 Shanghai's graceful tree-lined boulevards were designed for an age when the ricksha was the linchpin of urban transport. The system coped when bicycles displaced rickshas after 1949, but problems surfaced when the number of automobiles began increasing after the mid-1980s. In the early 1990s, the tempo of automobility quickened appreciably, and mixed traffic conditions gave rise to severe congestion along the major arteries virtually throughout the day. For instance, it can take up to two hours to cover the short distance from Pudong to the Bund. Currently much of the traffic is composed of bicycles and public transport. The continued Table 3.6 Urban Infrastructure of Shanghai, Tianjin, and Guangzhou, 1990 and 1994 Shanghai Tianjin Guangzhou National urban Indicator 1990 1994 1990 1994 1990 1994 average, 1990 Per capita living space (square meters) 6.6 7.5 6.7 7.1 8.0 8.4 6.9 Per capita paved road (square meters) 2.3 3.5 4.9 7.1 3.0 4.7 2.7 Public transportation per 100 residents 8.5 13.8 3.2 5.6 4.1 10.2 1.8 Telephone installation per 100 residents 9.0 - 4.9 - 9.2 - 2.7 Per capita electricity consumption (1,000 volts per hour) 144.1 - 98.0 - 169.0 - 80.1 Access to faucet water (percent) 100.0' 100.0 77.8a 100.0 87.8X 98.1 81.0b Per capita water consumption (tons) 81.9 - 42.9 - 173.5 - 63.7 Natural gas usage (percent) 46.1 80.8 48.0 90.8 2.3 85.2 8.1 As a percentage of the national urban average in 1990 Per capita living space 95.9 97.2 116.6 Per capita paved road 85.2 181.5 111.1 Public transportation 472.2 177.8 227.8 Telephone installation 332.0 179,4 338.6 Per capita electricity consumption 179.9 122.3 211.0 Access to faucet water, 1987 123.5 96.0 108.4 Per capita water consumption 128.6 67.3 272.4 Natural gas usage 569.1 592.6 28.4 -Not available. a. 1987 data. b; 1985 data. Source: State Statistical Bureau, China: Urban Statistical Yearbook, 1988, 1991; Statistical Yearbook of China, 1995. Table 3.7 Indicators for Infrastructure in Selected Countries, Various Years Hong Kong Rep. of Indicator Brazil China (China) Indonesia Japan Korea Mexico Singapore Tlhailand Telephone installation per 100 residents, 1990 6.3 2.0, 43.4 0.6 44.1 31.0 6.6 38.5 2.4 Access to safe water, urban (percent), 1990 95.0 89.2 99.0 65.0 - 100.0 89.0 100.0 85.0 Households with electricity (percent), 1984 79.4 - - 14.2 - 100.0 74.6 98.3 43.0 Paved road (kilometers per 1 million persons), 1988 703.6 640.0 _ 159.7 6,007.4 236.1 820.2 - 513.4 Roads in good condition (percent), 1988 30.0 - - 29.9 - 69.9 84.9 - 50.0 Fixed investment, housing (percent of GDP), 1985 - - 6.2 - 3.4 4.6 4.4 13.5 3.6 w Fixed investment, transport (percent of GDP), 1985 - 1.3 1.3 - 3.0 3.2 3.0 4.9 1.9 - Not available. Note: GDP, gross domestic product. a. 1993 figure. b. 1990 figure for cities. Source: World Bank, World Development Report, various years; Social Indicators of Development, various years; State Statistical Bureau, Statistical Yearbook of China, 1994. Figure 3.2 Shanghai and Its Environs IBRD 28657 RAILWAYS AONGOLA CENTRAL CITY _ -./ . /'EM. PAOPNEN CITY PROPER (SHIQU) .RE KOREA * ETDZ (ECONOMIC AND . . OREf TECH. DEV'T. ZONE) *P,ANOY - -- - COUNTY BOUNDARIES - - - - PROVINCIAL BOUNDARIES o f . NE - E-AN map Area * ANNUl' ~~~ANGNAI~SHI -32'00' H Ea-t SICHEUAN7 ZIEG N ., , .. hHUNAN.S GUIZHOU FUWIAN. TAJWAN YUNNAN GUANGU. GJANGDOf'E4 I ' '- . 'E 4JPONR- KONG -_ - R # . '> ffi '.-j eVPEENMAM-"'.- ' MACAO, PORT ,' _,THAIEANDj ROEEPM ' > . NAPNiAP Chi.a Se. REP ~~~~~~~~~~~~PHIIPIPINES' . , CHONGMING ' -31°30' ,^ 19 31°30'- JIANGSU PROVINCE Inland '9~~~~~~~~~~~~~~~~~~~jn 5; Pild'Pe ong Ho ng i oigq.oatr/o0 0f A.nt rpotioa H,- gh0 te P;aPrk, -- - 0 :0 0 - . _ - w QINGPU ) ~~~~ONGJIAN ,k ,r . E 5NGJlANGA TDhZai; 0 NANHUI 31'00'- H..un gp. FENGXIAN /,// C j v ) JINSHAN ZHEJIANG Hangzhou Bay PROVINCE 0 5 10 15 20 25 121 00' 121 30 KILOMETERS SEPTEMBER 1997 64 Shanghai: Renaissance City 65 inadequacy of bus transport and the limited use of mass rail transit sys- tems mean that bicycles and walking are still the principal means of transport in China's cities, including the three covered in this study (see Kubuta and Kidokoro 1994). Nevertheless, the number of private ve- hicles is rising steadily, further straining the antiquated road system. Somewhat belatedly, both central and municipal authorities have be- gun attempting to alleviate Shanghai's transport bottlenecks. During 1990 and 1993, Y20 billion was invested in large infrastructure projects, while new investment is running at the same level each year. Projects include two completed bridges across the Huangpu River and a tunnel underneath, an overhead inner-ring road, an outer-ring road, a six-lane north-south expressway along Chengdu Road, a new airport (to the east of Pudong), and new power plants. Shanghai's port is being upgraded so as to accommodate a much larger traffic in containers, and proce- dures are being streamlined to expedite the flow of freight. Under con- struction at great speed is a subway system, which is supported by for- eign capital and technology. This will take some of the pressure off Shanghai's choked streets once sections of it become operational. The first phase of the subway system, which is a single 10-mile-long track with 13 stations, was completed in April 1995. The $680 million project took just five years to complete, a feat unrivaled by other countries. A second 18-mile segment that is expected to become operational by 2000 will connect East Pudong with western Shanghai. Shanghai's future economic prospects rest in part on the successful completion of industrial and infrastructure projects in the old city. They also depend in part on the development of Pudong (see figure 3.2), a zone akin in certain respects to the Shenzhen Special Economic Zone and five-sixths the size of Singapore, and in part on a closer integration of the municipal economy with that of the lower Yangtze region. Con- structed after 1990, Pudong is in fact a new city (see Jacobs and Hong 1994). It has an industrial area of about 18 square kilometers, out of a total area of 500 square kilometers, and encompasses six special zones: Jinqiao (export processing of high-technology products in large-scale operations), Huamu (residential, commercial, financial, and cultural fa- cilities), Liuli (metallurgical and construction industries), Lujazui (finance and trade zone across from the Bund that will be a center of producer services), Waigaoqiao (export processing and petrochemicals and en- ergy), and Zhangjiang (scientific, research, and educational facilities). The development of Pudong, divided into three phases and with $36 billion of government investment, is designed to relieve spatial pres- sures on old Shanghai (the Puxi area) and to provide a framework for infrastructure planning and development, with concessionary policies for foreign investors.20 Perhaps even more ambitious, work commenced on a second international airport south of Pudong in 1995. With both sea and air transport capacity, Pudong can further enhance the city's trading 66 The Dynamics of Urban Growth in Three Chinese Cities capacity. In fact, the airport, which will be the biggest in Asia, and Waigaoqiao port are crucial to the emergence of Shanghai as a center of international com- merce. If the expectations are fully realized, Pudong will become the axis of the city's industry and commerce, particularly high-technology indus- trial and trading activities. By 1994 Pudong's GDP accounted for 15 per- cent of the total for Shanghai's, compared with a mere 8 percent in 1990 (Yeung 1996). Some major industrial and property developments have already taken shape in Pudong. China's biggest steel conglomerate, Baoshan Steel Corporation, has leased 82 acres of land for a giant steel plant to expand its production (East Asian Executive Reports, August 15, 1994). Japan's Yaohan Group, in partnership with Shanghai's No. 1 Yaohan Department Store, has built a $100 million shopping complex, one of the largest in Asia.21 Meanwhile, Carrefour, the French hypermarket chain, is planning a major outlet, having opened one in Beijing. Pudong will also provide space for new industries such as pack- aging, and, in fact, the China National Packaging Corporation is going to construct a modem packaging zone. Since 1995 the Shanghai Securi- ties Exchange and some other financial institutions have occupied new premises in Pudong, and starting in 1996 foreign banks could open branches there. Pudong's growth rate during 1994-95 averaged more than 26 percent a year, and by mid-1996 contracted foreign direct in- vestment (FDI) in 3,900 enterprises had reached $10.3 billion (see Ox- ford Analytica, February 4, 1997). But caution is needed, because even Pudong's success cannot solve the problems facing the metropolis; it can only assist and, in essence, be a catalyst for the city's modernization. Moreover, the price of industrial land is steep, and the quality of services provided needs to be improved.' Many manufacturers prefer other special districts in the city, such as the Minhang District, in the southwest of Shanghai, which is adjacent to the airport and an expatriate community housing estate. As a consequence, much of the industrial and office space in Pudong's Waigaoqiao Free Trade Zone remained unoccupied in mid-1997. Neighborhood Effects Shanghai has the most prosperous economic hinterland of any city in China (Jacobs and Hong 1994). In the pre-1949 period, linkages with the hinterland and the spillover benefits from proto-industrialization in the neighboring provinces, along with the buoyancy of regional commerce, contributed significantly to Shanghai's prosperity. However, in the years following liberation, Shanghai's ties with its hinterland weakened. A number of factors were responsible for this growing isolation. The de- liberate effort to cast Shanghai in the role of "producer" city has already been noted. Shanghai imported industrial materials from across China and in turn provided the country with manufactures and capital Shanghai: Renaissance City 67 equipment. Each province, including the two immediately adjacent to Shanghai (Jiangsu and Zheijiang), attempted to create a relatively self-contained industrial economy and buttressed their attempts to achieve industrial autonomy by minimizing cross-provincial linkages and instituting administrative controls on the flow of trade. Rigidly en- forced greenbelt regulations restricted rural industrialization in the im- mediate environs of Shanghai in an effort to contain the spread of the city into the adjacent farmland. Urbanization in the areas of Jiangsu close to Shanghai was also restricted so that the closest urban settlements, such as Changzhou, Jiaxing, and Suzhou, were arrayed 100 kilometers away. Once reform commenced and the advantages of closer regional inte- gration became apparent, central and municipal governments cobbled together the lower Yangtze economic region in 1982 to coordinate de- velopment and begin dismantling trade barriers between Shanghai and the neighboring provinces. This effort soon foundered, but linkages nev- ertheless multiplied as mutual interests began overriding bureaucratic obstacles and growth-creating effects became manifest. For Jiangsu's flourishing agricultural economy, Shanghai was the preferred source of consumer goods and a range of household durables. Township and vil- lage enterprises that proliferated in southern Jiangsu and nearby Zheijiang turned to Shanghai for skills, equipment, and markets. Many of these were established by workers who were transferred from Shang- hai in the 1970s and not permitted to return because of population con- trols. Retirees from Shanghai's state-owned enterprises returning to home villages in Jiangsu or emigrating from the city because of enticing sala- ries provided much-needed technical skills for setting up factories and supervising production.3 Thus former Shanghainese made it easier for rural-based producers in southern Jiangsu to obtain equipment and tech- nical support from Shanghai. Enterprises in Shanghai quickly discovered the advantages of mov- ing some production facilities out of the city and entering into subcon- tracting arrangements with manufacturers in the hinterland. Such sub- contracting allowed enterprises to rationalize production and cut the expenses of extreme vertical integration. Through much of the 1980s, enterprises wanting to relocate facilities had to leapfrog the protected belt of farmland girding Shanghai and to transplant production in Jiangsu or Zheijiang. But after 1988, the ban on rural industry within the mu- nicipal precincts was relaxed, and the immediate hinterland became part of an industrial continuum extending from the city into the lower Yangtze region. Shanghai's economic neighborhood now embraces 45 of the 100 richest small towns in China, all of which lie in the Sunan area (China Focus, October 1, 1994). Its interconnectedness derives additional force from the multimodal transport network that is being rapidly improved by investment in road, rail, and canal infrastructure. Neighborhood effects have taken close to a decade to flower and are beginning to reshape the regional economy to the advantage of all par- Table 3.8 Education of the Labor Force in Selected Sectors in Shanghai, 1990 Industry Transport Research and development Government Thousands Thousands Thousands Thousands Indicator of employees Percent of employees Percent of employees Percent of employees Percent Total 4,252.3 344.2 98.5 291.4 Level of education University 84.5 2.0 8.5 2.5 32.6 33.1 18.9 6.5 Community college 167.6 3.9 13.5 3.9 16.5 16.7 42.0 14.4 Vocational school 169.0 4.0 19.2 5.6 10.2 10.4 30.3 10.4 Senior high school 1,041.6 24.5 88.1 25.6 18.2 18.5 74.7 25.6 Junior high school 1,997.1 47.0 155.0 45.0 17.8 18.1 92.0 31.6 Elementary school 646.1 15.2 49.4 14.4 2.8 2.8 28.9 9.9 None 146.4 3.4 10.3 3.0 0.5 0.5 4.4 1.5 Nature of work Technical 378.9 8.9 38.5 11.2 53.8 54.6 67.8 23.3 Administration 125.6 3.0 10.3 3.0 6.0 6.1 58.9 20.2 Staff 185.2 4.4 38.2 11.1 8.5 8.6 109.4 37.5 Sales 123.7 2.9 6.4 1.9 1.6 1.6 8.2 2.8 Services 291.3 6.9 30.5 8.9 4.4 4.5 16.0 5.5 Production 3,141.8 73.9 219.3 63.7 23.8 24.1 29.4 10.1 Other 5.8 0.1 0.9 0.3 0.5 0.5 1.6 0.5 Note: Numbers may not sum to totals because of rounding. Source: State Statistical Bureau 1992b. Shanghai: Renaissance City 69 ticipants. Under the planning system, urban centers pursued industrial diversification. This approach is giving way to a pattern of greater spe- cialization and increased trade. To the extent that exit barriers to state- owned enterprises permit, Shanghai is gaining the freedom to expand high-technology and human capital-intensive industries in which it has a comparative advantage. It has also begun concentrating on producer services that supply sophisticated industries with capital, information, and skills. If the trend persists-and the resurgence in growth during the first half of the 1990s suggests that it should-Shanghai and its hin- terland will come to resemble other large industrial regions such as the Kanto plain in Japan and the belt connecting Sao Paulo with Rio de Janeiro in Brazil (see Becker and Egler 1992; Park and Markusen 1995). By 1995 the Yangtze delta region accounted for one-third of China's to- tal industrial output, compared with 10 percent for Guangdong. Human Capital and Technology Human, financial, and fiscal resources provide industrial leverage. Un- til recently, restraints on migration (especially of skilled labor), the vir- tual absence of capital mobility, and the modest extent of fiscal redistri- bution meant that local effort at mobilizing resources was essential for growth to become self-sustaining. Although barriers to the movement of resources remain, they are not as constrictive and will become pro- gressively less meaningful in the second half of the 1990s. Nevertheless, municipal efforts to build skills, raise capital, and attract foreign invest- ment remain profoundly important. Shanghai entered the 1980s with a handsome endowment of human capital and a solid tax base, albeit largely in the grip of the central government. Initial attempts to mobi- lize local resources were diffuse and yielded limited dividends. Since 1988 increased support from the central government and a more deter- mined reform drive have brightened the picture. Compared with those of other industrial cities in China, Shanghai's workforce commands an impressive range of educational skills. Most workers have an elementary education or better, and the vast majority have at least graduated from junior high school. For instance, in indus- try, 85 percent of workers have completed a minimum of eight grades and 6 percent have some college education. The shares are much higher in transport, government, and, of course, R&D (see table 3.8). The state sector employs close to 900,000 technical personnel and more than 40,000 scientific and technological professionals working in state-owned en- terprises, research institutes, and universities. The city hosts at least 266 independent research institutes. In addition, there are nearly 50 uni- versities and colleges, 28 of which are under the jurisdiction of central ministries or commissions.24 With 37 postdoctorate research centers, these universities have become the cradle for young research scientists and produce several hundred research findings every year, but so far a 70 The Dynamics of Urban Growth in Three Chinese Cities shortage of funds has prevented the commercial development of many promising ideas.25 Research institutes are scattered across many different university, min- isterial, agency, and even enterprise jurisdictions. In the majority of cases, they tend to concentrate on their own narrow fields and lack multiple contacts with researchers in neighboring disciplines. Until recently, commercial orientation was weak, and neither the mix of staff nor the incentives in many institutes were conducive to high-quality research, its patenting, or its application. This appears to be changing as stringent budgets force institutes to compete for contracts from the enterprise sector. Although China's economy has grown at double-digit rates since 1993, the demands of infrastructure have absorbed a large volume of invest- ment and constrained the availability of public resources for scientific research. China spent $7.5 billion in 1993 on science and technology, about a third of which went to R&D. Of this, less than 40 percent was devoted to research. China's senior leaders recognize this stagnation and intend to triple the funding to reach 1.5 percent of GDP by 2000. In 1994 the central government spent $275 million on research, a 13 percent in- crease over 1993 ("Leaders Pledge More for a Shrinking Pool," Science 268, June 9, 1995; "The Long March to Topnotch Science," Science 270, November 17, 1995). Nevertheless, the strain on funds means that uni- versity science departments and research institutes must generate part of their own revenue. Shanghai's leaders have enthusiastically endorsed the new policies and are providing a small pot of money that local scien- tists can use to attract larger sums from other sources.26 But the financ- ing of research is likely to remain a problem, even as the research im- perative acquires greater force. The Infrastructure of Finance Human resources can be used fruitfully only if a sufficient supply of financial capital is generated and industrial initiatives are adequately backstopped by producer services. After a lengthy period of gestation in the 1980s, the city is finding its stride, and financial development is proceeding apace, albeit not without difficulties. Indigenous banking activities cornmenced in the mid-nineteenth century, and by the end of the century foreign banks had also arrived on the scene. By 1947 there were 128 government- and privately owned banks, 14 foreign banks, 13 trust companies, and 79 money exchanges, whose activities spanned the entire country (Fung, Yan, and Ning 1992). Following a period of dormancy, financial reforms, introduced in small doses throughout the 1980s, promoted the institutional elaboration of the banking industry and of capital markets. This was inevitably a halt- ing process because banks were inexperienced, financial skills were lack- ing, and three decades of socialism had engendered a high degree of Shanghai: Renaissance City 71 caution. Early attempts to create an interbank market to permit a re- gional pooling of financial resources and to provide both households as well as enterprises with a wider range of instruments had just begun to achieve momentum when these innovations were put on hold. Stabili- zation measures and direct controls were first applied tentatively in late 1988 to arrest inflationary pressures and were then exerted with re- doubled force following the Tiananmen incident in mid-1989 (see World Bank 1990 for an account of China's macropolicies during 1988-90). After 1988 interprovincial credit flows through the banking system slowed to a trickle, and financial development resumed only in 1991. Until the very start of the 1990s financial resource mobilization was treated with utmost passivity. By and large, banks and urban credit co- operatives collected household savings and allocated these in accordance with guidelines handed down by the People's Bank of China or munici- pal supervisory agencies. There was little attempt to stimulate savings, which were admittedly high, by offering attractive rates on a variety of instruments. Nor was allocative efficiency, using market channels, pur- sued systematically and with reference to clear goals. Many ad hoc ex- periments were launched, and these only began to coalesce and acquire the semblance of a coherent program in the early 1990s. The financial initiative with the greatest promise for Shanghai was the setting up of a stock market in December 1990.27 The initial significance of this market has little to do with the mobilization of resources or the effective dispo- sition of capital. Rather its importance lies in its symbolic value and the inducement it gives to institution building.28 An active and growing stock market is tangible evidence of successful transition. It also pushes the enterprise sector to improve its financial performance, pay greater at- tention to auditing and accounting practices, accept the need for legal discipline, and be willing to abide by the letter of the rules. With each passing month the Shanghai Stock Exchange is adding to its fund of financial experience, engendering investor confidence, and staking Shanghai's claim to becoming a world-class player. It is also at- tracting funds that local businesses urgently need to modernize indus- try. By the end of 1995, the Shanghai Stock Exchange had about 7 mil- lion registered investors (IFc 1996). The combined capitalization of the Shanghai and Shenzhen markets is expected to reach $168 billion by the end of 1997 (Financial Times, July 28, 1997). Although government agen- cies still hold a high proportion of the market's capitalized value, indi- viduals and other entities have already demonstrated their willingness to invest. As its listings expand in the future, the stock exchange will pull in many more investors from outside the municipality, initially from other locales in the lower Yangtze region.29 Starting in January 1992, the issue of shares in corporatized state-owned enterprises to overseas in- vestors was a significant step toward broadening the market. Financial reform and the attempt to raise the technological tenor of industrial activity have created a market for a broad range of producer 72 The Dynamics of Urban Growth in Three Chinese Cities services. Shanghai has designated six service industries as the top pri- orities of development. These include finance and insurance, commerce and trade, telecommunications and transportation, real estate, tourism, and information. The first three subsectors have already become the driving force behind Shanghai's service industry, contributing about 73 percent of the sector's value added in 1993. The service sector as a whole accounted for 38 percent of the city's GDP, and employment exceeded 170,000 in 1994 (FBIS-CHI-94167, August 29, 1994, p. 69).3° Since 1990 the city has opened exchanges for securities, metals, coal, farm production materials and equipment, chemicals, grains and edible oils, motor ve- hicles, building materials, and technology. In the coming years, there are plans to expand the money market (already China's largest), as well as the markets for securities, foreign exchange, and insurance (FBIS-CHI- 94107, June 3,1994, p. 43). In the Huangpu, Jingan, and Zhabei districts, clusters of private and cooperative consulting firms provide a wide range of services. The Shanghai Academy of Social Sciences set up a consultancy center as early as 1979 to provide economic, legal, and public opinion services. This network, and the approximately 1,000 computer science graduates from local tertiary institutions, made possible the export of computer software services (primarily to Japan) during the 1980s. It re- flects Shanghai's superior human skills and points toward a type of ac- tivity in which Shanghai may have a real comparative advantage. A rich stock of technical manpower has given life to small businesses offering consulting services and undertaking research contracts.31 Financial services are perhaps the largest and fastest-growing subsector, gradually opening a variety of activities. More than 2,200 bank- ing organizations have branches and offices in the city, including 117 branches of foreign banks at the end of 1995 (in all, 137 foreign financial institutions were in Shanghai by the close of 1995, with assets of $18 billion; FBIS-CHI-94103, May 27,1994, p. 57; "Renminbi Business," Ox- ford Analytica, April 5, 1996). The British Standard Chartered Bank, which dates back to 1858 and retained its presence throughout the post- 1949 period, leads the way. Citibank has shifted the headquarters of its Chinese business from Hong Kong to Shanghai, and many others are expected to follow suit. So far, foreign banks have mainly concentrated their activity in the lucrative area of trade finance and cash settlement of B share issues, but eight foreign banks were allowed to engage in Chi- nese currency business as an experiment in late 1996 (Oxford Analytica, February 4,1997). Loans to Chinese enterprises are still largely arranged in Hong Kong. At the beginning of 1993, the country's first batch of 31 professional brokers to be licensed by the state government for forward trading of treasury bonds started working at the Shanghai Stock Exchange. In November a technology brokerage office was opened to attract domes- tic and overseas brokers and promote technological exchanges (FBIS-CHI- 93225, November 24, 1993, p. 46). By July 1994 Tullet & Tokyo Forex Shanghai: Renaissance City 73 International Ltd., the world's largest transnational currency brokerage company, had set up its first representative office in Shanghai and planned to set up a branch company in the near future to connect with its other four branches in Australia, Hong Kong, Singapore, and Tokyo (see FBIS-CHI-94134, July 13, 1994, p. 39). It was the first overseas cur- rency brokerage institution to receive govemment approval. The mu- nicipal government has also begun rebuilding the Bund, the famous area along the Huangpu River, which housed 113 bank buildings before 1949, as China's Wall Street. Housing management units have established a special institute to help relocate government departments in the former buildings lining the street. Quite a few buildings have already been va- cated and have changed ownership (FBIs-cHI-94016, January 25, 1994, p. 81). The number of persons employed in the securities market has grown rapidly. Shanghai International Securities, the country's largest stock- broker, employed a staff of 120 in 1991; this had grown to 1,500 in 1994. Arthur Anderson, a big transnational accountancy firm, has 120 employ- ees in Shanghai, compared with a mere handful in the early 1990s. Bear Stearns, Goldman Sachs, Merrill Lynch, Morgan Stanley, and Nikko are among the many financial firms now present in the city. However, the local stock market remains too restricted for much foreign investment. Many foreign investors also prefer those Chinese companies listed in Hong Kong and New York. Shares of Chinese firms in Hong Kong are four times more liquid than B shares in Shanghai when measured by tumover per share.32 Since 1992 some drastic reform measures have been introduced in the financial sector. For instance, the Shanghai Foreign Exchange Swap Market, founded in 1986, has removed the ceiling for exchange quota- tion and position to allow individuals to trade foreign currency freely. Another major step in 1994 nudges various state-owned specialized banks into becoming commercial banks through the pursuit of "risk- related management." Credit cooperatives in both urban and rural ar- eas are also being turned into cooperative banks to help spur the growth of the local economy (FBIS-CHI-94004, January 6, 1994, p. 58). To involve local banks in the foreign exchange business, the People's Bank of China set up an unusual U.S. dollar clearing center in Shanghai; this is the first not only in the country but also in the world. In April 1994 China's first interbank currency market opened in Shanghai, replacing the fragmented system of currency swap centers. In August the Shanghai Securities Ex- change was already doing on-line business linked with 15 other securi- ties exchange centers throughout the country. Moreover, the People's Bank of China and the Ministry of Finance selected the exchange as the base on which to establish the issuing, trading, and clearing of treasury bonds. Aside from the producer services that are fairly closely associated with industry, Shanghai's agglomeration economies are a stepping-stone to 74 The Dynamics of Urban Growth in Three Chinese Cities other services that sustain the economies of cities such as New York. These are advertising, book publishing, fashion design, and media con- glomerates active in television, magazines, and newspapers. Although the ease of transferring information by electronic means has lessened the importance of propinquity and hence the need for firms to locate their offices in a central business district, the media industry still thrives on face-to-face encounters and draws its lifeblood from the meetings, lunches, informal encounters, and frequent exchanges of gossip that are most easily conducted within a large, densely populated city. Getting information at the earliest possible stage, maintaining social relations, and making deals require personal interaction and not merely an ex- change of information by way of computers, which nevertheless also greatly facilitate media-related and publishing activities. Many of the industries that once buttressed New York's prosperity, such as textiles and meat packing, have long since deserted the city, which has ceased to be a regional transport hub. But all those activities whose consumma- tion depends on propinquity, such as finance, publishing, and fashion garments, have remained because New York is the ideal urban arena in spite of the drawbacks posed by congestion, crime, and public services of uneven quality (Sassen 1991). The services that now sustain New York's economy are at an embry- onic stage in Shanghai. But they are bound to grow exponentially, and Shanghai more than any other city in China can, with some effort, achieve a commanding lead over other aspirants. It certainly has the advantages of scale, the pool of talent, the industrial diversity, and the supportive traditions from an earlier era. However, the city must enhance its physi- cal and cultural attributes if it is to attract service activities with a high income elasticity. This will require investment, of which much will have to be found locally, but some might be sought abroad. Capital Flows and Industry When China cautiously opened the door to foreign investment in 1978, the trickle of capital first flowed into the Special Economic Zones and the open cities in Guangdong and Fujian. In fact, the pattern of FDI through the first half of the 1980s favored the southeast region (see chap- ter 5). The overwhelming majority of investors were businessmen from Hong Kong and Macao wanting to move some of their assembly-line operations to a familiar part of China. Moreover, other "open cities" such as Tianjin and Shanghai were much less aggressive in bidding for FDI and, to a degree, were constrained by the reform mandates assigned to them by the central government. Until 1985 the cumulative investment utilized in Shanghai amounted to $232 million. A quickening of the efforts of Shanghai authorities to induce FDI in 1987-88 produced a temporary surge in 1987-89, but in- vestor interest waned in 1990-91 as a result of Tiananmen. Deng Shanghai: Renaissance City 75 Xiaoping's visit to Guangdong in January 1992 and his strong endorse- ment of the reform drive galvanized all provincial players, in particular Shanghai, to seek foreign capital. Shanghai's own determined initiatives and a rapid rekindling of interest on the part of overseas investors brought a rush of funds, and utilized foreign investment increased al- most tenfold, reaching $3.2 billion in 1993 (see table 3.9). Contracts with foreigners continued rising over the next two years so that total contrac- tual investment exceeded $50 billion by the end of 1996, with Japan emerging as the leading investor, having committed $5 billion (Financial Times, April 9, 1997). In the second half of the 1980s, much of the FDI was in advanced manufacturing activities, the hotel industry, and affiliated travel services: for example, Volkswagen in automobiles, McDonnell Douglas in aircraft assembly, Pilkington in glass, and Foxboro, IBM, Xerox, and Wang in computers, copiers, and electronic equipment (U.S. Con- gress, Office of Technology Assessment, 1987). Between 1979 and 1993, about half of foreign investment was from Hong Kong, about a fifth from Japan, 12 percent from the United States, and the balance mostly from Europe (roughly two-thirds of all FDI in China originated in Hong Kong and Macao; see Jacobs and Hong 1994). Japanese money went into retailing, and other funds were drawn into real estate, hotels, infrastruc- ture, power, and, of course, manufacturing. The volume of FDI in Shanghai has not been large, although it is now assuming sizable proportions. During 1987-89 it was just about 8 per- cent of annual capital construction outlay. However, it played a catalytic role in several strategic sectors. Foreign capital and expertise were in- strumental in equipping Shanghai with the hotel and other facilities of a caliber to attract long-distance tourists in large numbers. These facilities have also enhanced Shanghai's attractiveness for conventions and con- ferences. FDI helped initiate the modernization of commercial real es- tate, and Hong Kong financiers are responsible for a good proportion of the office buildings that are reshaping Shanghai's skyline. German in- vestment rejuvenated the auto industry and has made Shanghai the most important producer of midsize sedans in China. The aircraft industry has derived much benefit from the local assembly of U.S. MD-82 pas- senger planes and the setting up of repair facilities in collaboration with Swissair. The electronics industry has begun obtaining badly needed capital and technology by way of joint ventures with major international corporations seeking entry into China's market. For instance, Intel is building a $50 million facility to assemble and test flash memory chips and micro controller devices. NEC decided in 1997 to build a $1 billion plant to manufacture state-of-the-art dynamic random access memory chips (DRAM 4s). And in mid-1997 Bell Labs entered into a multimillion- dollar venture with Chinese universities to develop telecommunications products for the Chinese market. Real estate developers, the major hotel chains, and the multinationals have come to Shanghai because it has the potential to join the ranks of Table 3.9 Foreign Investment in Shanghai, 1986-93 (millions of U.S. dollars, unless otherwise noted) Indicator 1986 1987 1988 1989 1990 1991 1992 1993 1979-93 Utilized foreign investment 281.16 575.63 440.47 466.49 321.04 330.25 899.29 3,175.00 10,751.69 Loans 132.26 361.62 207.30 44.37 147.03 185.06 405.68 853.18 5,097.48 Direct investment 148.90 213.66 233.17 422.12 174.01 145.19 493.61 2,317.62 5,132.44 Other - 0.35 - - - - - 4.20 521.77 Contracted foreign direct investment, by source - 338.31 333.28 359.75 374.63 450.01 3,357.25 7,016.14 13,705.29 Hong Kong (China) - 217.89 104.91 113.29 110.91 128.35 - 4,337.76 7,001.96 Percentage of Shanghai's total - 64.4 31.5 31.5 29.6 28.5 - 61.8 51.1 Percentage of Hong Kong's total - 11.7 2.9 3.4 2.7 1.8 - - - Japan - 4.60 22.10 27.71 23.93 131.69 - 370.15 1,190.06 Percentage of Shanghai's total - 1.4 6.6 7.7 6.4 29.3 - 5.3 8.7 Percentage of Japan's total - 2.1 7.4 7.8 11.1 18.3 - - - United States - 55.95 58.21 42.01 88.41 50.07 - 598.58 1,665.52 Percentage of Shanghai's total - 16.5 17.5 11.7 23.6 11.1 - 8.5 12.2 Percentage of U.S. total - 16.5 18.8 8.3 26.4 10.4 - - - Taiwan (China) - - - 7.52 101.81 31.44 - 540.00 916.89 Percentage of Shanghai's total - - 2.1 27.2 7.0 - 7.6 6.7 Percentage of Taiwan's total - - - 1.7 10.3 2.6 - - - Others - 59.87 148.06 169.22 49.57 108.46 - 1,179.65 2,930.86 Percentage of Shanghai's total - 17.7 44.4 47.0 13.2 24.1 - 16.8 21.4 - Not available. Source: State Statistical Bureau, Statistical Yearbook of China, 1987, 1988, 1990, 1992, 1993; data by source for 1987-91, State Statistical Bureau 1 992a; 1993 data from State Statistical Bureau, Statistical Yearbook of Shanghai. Shanghai: Renaissance City 77 world-class cities. A hundred years ago under quite different geopoliti- cal circumstances, foreign capital and technology pushed Shanghai over the threshold into industrial modernity. Capital flowed to the city be- cause its location was a great asset. A deepwater port and proximity to Japan conferred advantages not easily overlooked. Now foreigners are involved in a second industrial transformation for these reasons and fresh ones as well. In this round Shanghai is far better equipped to absorb capital and technology-it is after all China's foremost indus- trial city-and in sheer scale of manufacturing capability Shanghai has no rival in the developing world. The enemies of promise in Shanghai's case are the congestion-related diseconomies of great size and the prob- lems associated with having invested too much in technology that is now outdated and in plant better suited to an earlier industrial era, in embracing organizational practices that are now seen to hobble com- petitiveness, and in adopting workplace rules that severely compromise enterprise flexibility even as reforms steadily enlarge the autonomy of industrial entities. A century ago, FDI helped Shanghai to sink new in- dustrial roots. Now it is helping to stiffen the city's resolve to push ahead with systemic change and has contributed $1.83 billion during 1985-92 toward the renewal of production facilities. More important, it has in- sinuated new skills, new ideas, and new work habits into every major sphere of the municipal economy. Over the long term, that counts for much more. It is influencing grassroots receptivity to changes and pre- paring people in all domains of industry to meet the challenge posed by other cities in China and beyond. Fiscal Relations and Revenue Generation The story of resource mobilization would be incomplete without a fur- ther look at fiscal sources and potential. Under planning and well into the first decade of reform, Shanghai was the single most important ele- ment in the revenue system established by the central government, con- tributing nearly 83 percent of central revenues in 1981 (Jacobs and Hong 1994). However, Beijing controlled and reallocated much of this revenue. A succession of adjustments in tax arrangements between central and city governments have altered the municipality's fiscal circumstances. Between 1949 and 1980, roughly 86 percent of Shanghai's revenue was remitted to the central government. This slowed urban development and placed industrial expansion largely in the hands of the central gov- ernment.33 The only discretionary income under the control of munici- pal authorities was derived from user fees on public services and sur- charges, among which the one levied on the industrial and commercial tax was the most significant. With the start of reforms in 1978, the cen- tral authorities introduced a new decentralizing fiscal regime that visu- alized each provincial entity as a "separate kitchen." Under this arrange- ment, participating provinces were allowed a fixed and an adjustable 78 The Dynamics of Urban Growth in Three Chinese Cities share of revenues. They retained all income collected in excess of these ratios. In exchange for receiving a bigger slice of revenue, provinces were also required to accept responsibility for most expenditures. Because of its strategic role as an industrial center and its revenue- gathering function, Shanghai was largely bypassed during this round of reforms, which conferred most of the advantages on the southeastern provinces. However, the central government did ease its tight grip on expenditure decisions, with the result that municipal spending rose mod- erately between 1980 and 1983 and tripled over the 1983-86 period. In- creased expenditures were not matched by an upward shift in retained income. In fact, Shanghai's overall revenue situation deteriorated as re- forms whittled away at price controls and cut centrally mandated re- source allocation. In 1980 the average profit ratio for Shanghai's enter- prises was 75 percent as against 24 percent for the whole country (Lin 1994). Four years later, the national average had fallen a percentage point, but the profit ratio for Shanghai's state-owned enterprises dropped to 63 percent. Both the central government and Shanghai municipality suf- fered an erosion in tax revenues, with the municipality's earnings de- creasing almost 5 percent between 1980 and 1984. Although Shanghai was allowed to keep 30 percent of the revenues collected above the planned targets and received an additional Y150 million in grants, these were inadequate. In 1980 the city received a fur- ther Y200 million and double this amount in the following year. These ad hoc contributions proved insufficient because growth of industry re- mained sluggish throughout the first half of the 1980s. As Shanghai's fiscal condition worsened, its political assertiveness mounted. It began lobbying intensively for easier treatment. Pressure from Party Secretary Jiang Zemin and Mayor Rui Xingwen, who came to power in 1983, fi- nally carried the day. In 1984 the State Council approved a decree giving Shanghai Y1.5 billion over and above actual expenditures of Y2.26 bil- lion in 1983. This translated into a retention ratio from municipal rev- enues of just under one-quarter (Lin 1994). The fiscal injection permitted the city to take the first few steps to- ward improving urban infrastructure and housing. But with the profit- ability of state-owned enterprises continuing to sink (to 32 percent in 1988), and municipally controlled enterprises being especially hard-hit, Shanghai was compelled to approach the center in 1987, well before the six-year term of the fiscal contract had ended. What forced the municipality's hand was the steep rise in subsidies for enterprises and basic food items. By the mid-1980s Shanghai's industry was losing its lead over other provinces that had been quicker to exploit the opportu- nities posed by reform. Aging industrial plant and an excessive reliance on state-owned enterprises were beginning to tell. Once again Shanghai sought to renegotiate its fiscal contract with Beijing using its political ties with central leaders and citing its crucial significance for China's Shanghai: Renaissance City 79 Table 3.10 Fiscal Relations in Shanghai, 1980 and 1984-91 (billions of yuan, unless otherwise noted) Indicator 1980 1984 1985 1986 1987 1988 1989 1990 1991 Revenue Total revenue 16.92 16.40 18.42 17.95 16.90 16.16 16.69 17.00 17.52 Budgetary income 16.68 - 18.16 - - - - 16.27 16.51 Nonbudgetary income 0.24 - 0.26 - - - - 0.73 1.01 Transport and energy… …… - - 0.30 0.33 funds Self-financing income 0.24 - 0.26 - - - - 0.43 0.68 Revenue sharing Total revenue 16.92 16.40 18.42 17.95 16.90 16.16 16.69 17.00 17.52 Remittances to the central government Amount - 13.69 13.53 14.41 11.99 10.50 - - - Percentage of revenue - 83.5 73.4 80.3 71.0 65.0 - 65.0 - Percentage of GDP - 35.0 29.0 29.5 22.0 16.2 - - - Local spending 1.92 3.03 4.61 5.91 5.39 6.59 7.33 7.56 8.61 Central subsidies Total fiscal subsidies - - 0.67 0.77 1.19 2.46 - - - As a percentage of local expenditures - - 14.6 13.0 22.1 37.3 - - - - Not available. Note: GDP, gross domestic product. Source: State Statistical Bureau, Statistical Yearbook of Shanghai, 1992; Lin 1994; Ho and Tsui 1996. industrial future. The effort paid off, and in 1988 the city entered into a revised arrangement that gave Shanghai terms broadly similar to those enjoyed by Guangdong since the early 1980s. These entailed remitting a fixed annual sum of Y10.5 billion to Beijing during 1988-90 (see table 3.10). In 1991 and 1992 Shanghai was required to remit the base amount and half of all revenues collected in excess of Y16.5 billion. Along with the tax contract, Shanghai was allowed greater leeway in defining fiscal relations with county governments. Notionally, the municipality gained Y11.4 billion in supplementary revenues in 1988 and 1989. In reality, special payments to the central government of Y400 million in 1990 and Y570 million in the next two years ate into Shanghai's enlarged share. The net gain was modest, and it forced the municipal authorities to widen the ambit of resource mobilization. Urged on by the central government, Shanghai launched a wide- ranging program of resource mobilization and expenditure management starting in 1990 using the full discretion allowed under the latest con- tract. First, the municipal authorities enlarged their take from user charges, which were raised in steps for a host of services. Second, fiscal extrabudgetary revenue was increased through fees and other levies on municipally controlled enterprises along the lines of other coastal prov- 80 The Dynamics of Urban Growth in Three Chinese Cities inces. Third, Shanghai began using the lucrative fiscal window opened by urban land reform to raise funds from the leasing of land. This is potentially of the highest significance.34 Shanghai's population density, its relative prosperity, and the nature of commercial development now under way mean that real estate throughout the municipality is extremely valuable and likely to become even more so a few decades hence. In the foreseeable future, Shanghai's land market will rival those of Bombay, London, Seoul, and Tokyo.3s More than 450 plots of land have been leased to foreigners in old central Shanghai on leases that vary from 50 to 80 years ("Shanghai's Amazing Property Boom," Asiamoney, October 1994, pp. 22-25). Although still small, this figure is bound to swell as land use in the central city is rationalized and made more intensive. Factories in the core city are being relocated or closed, and the land is being put to commercial use. More than 2,000 factories producing textiles, machine tools, and aerospace equipment are being relocated from the city center to rural counties to facilitate commercial development in the downtown area ("Factories Forced Out of Shanghai," Wall Street Journal, June 28, 1995), but much investment is needed before the city acquires a modem face. Fourth, the attempt to enhance local property values and draw in capi- tal from the outside is being spearheaded by the building of infrastruc- ture. This is an avenue to resource mobilization in the medium run be- cause it will stimulate local industry and attract funds to Shanghai. Over the long run, these trends will push up land values and, thereby, the tax base of the municipality. Fifth, Shanghai has initiated its own county- level fiscal decentralization so as to generate additional revenue and transfer some of the expenditure responsibilities to lower-level govern- ments. Typically the formula used allows county finance bureaus to re- tain the right to a fixed proportion of revenues collected above a base value. This encourages greater tax effort, some of which contributes to the municipality's budget, while the assignment of expenditures im- proves the allocation of resources. By 1990 the net transfer from local governments to the municipality was on the order of Y80 million. Sixth, the development of Pudong described earlier brought Y6 billion in grants from the center. This was augmented by the flow of FDI into new busi- nesses and infrastructure for the industrial district. The fiscal picture for Shanghai has brightened considerably since 1990. The agreement reached with Beijing in 1988 was a good beginning, and it was usefully buttressed by subsequent actions, direct and indirect, to mobilize funds. Success at accelerating industrial growth beginning in 1992 strengthened revenue performance. It also demonstrated to skep- tics that there is life yet in the old city. Shanghai's recent economic per- formance has allayed some of the fears of outsiders, whose interest was aroused by heightened municipal desire for FDI, by the money being poured into infrastructure, and by the promise of the Pudong project. Shanghai: Renaissance City 81 Fiscal experimentation is bound to continue. The tax package and in- tergovernmental sharing arrangements put into effect in January 1994, as they evolve, will affect Shanghai's finances, but the precise implica- tions are far from clear. The new system is supposed to levy a universal tax rate on all provinces and cities directly under the central govern- ment. The trends regarding revenues are beginning to crystallize. It is likely that the importance of direct taxes and the value added tax will grow at the expense of extrabudgetary sources of revenue. User fees and real estate taxes will also bulk very large as Shanghai modernizes and in the course of doing so adopts the fiscal lineaments of a world- class city. Even the sharing arrangements with Beijing are not immu- table. It is probable that some increment in Shanghai's revenues will have to be transferred to meet the central government's needs. But there is no likelihood of a reversal of decentralization, except on the margin. Municipalities such as Shanghai will continue to be responsible for most local expenditures and be assigned the revenues from local sources to pay for them. The extra revenue effort and expenditure efficiency, dif- ferentiating the dynamic cities from others, will depend on local initia- tive rather than central prompting. Expenditure management is the other half of the fiscal equation. Fis- cal health depends not just on raising revenue but also on keeping a close watch on expenditures. This has been problematic in a city such as Shanghai, with its armada of state enterprises and large tenured labor force. As reform has exposed the state sector to competition, many state- owned enterprises that have been unable to upgrade equipment, skills, and organization are running losses (for the country as a whole, 44 per- cent of state-owned enterprises were running losses in 1995, and more than two-thirds ran losses in 1996; United Press International, February 4,1997). Municipal authorities are unwilling to countenance widespread closure because of the sociopolitical costs of unemployment in the ab- sence of an adequate safety net. Even if unemployment compensation were available, many workers would consider that shutting down the large number of enterprises in distress would constitute breaking a so- cial compact. The Shanghai municipality has shied away from mass clo- sure and preferred to pay the fiscal penalty. A sizable part of the city's limited resources has gone into shoring up weak state enterprises. Ad- ditional funds have been swallowed up by subsidies for food and basic services. These fiscal burdens divert resources away from projects criti- cal for Shanghai's modernization. Clearly, the significance of the state sector has been a handicap in the reform era. Some of the characteristics that made Shanghai an outstanding industrial player in the Maoist years have proved a handicap as liberalization has rapidly diminished the odds favoring state-owned enterprises. Price reform in the early 1990s cut consumer subsidies and decreased fiscal pressure. Enterprise reform and the piecing together of a munici- 82 The Dynamics of Urban Growth in Three Chinese Cities pal social safety net have restored the health of some enterprises and increased the government's readiness to allow bankruptcy. For Shanghai and other cities, this administering of market medicine has extraordinary significance: it allows the municipality to gauge the social acceptability of market-based solutions, it adds to the weight of precedence, and it eases fiscal constraints. The gradual introduction of rational pricing schemes for municipal services beginning in the early 1990s reduced subsidies. Meanwhile, decentralization of spending de- cisions to the county level provided a useful check on the growth of expenditures and enhanced efficiency. Resources have been freed for uses that are more likely to yield the desired economic results. Summary: Reform and Urban Development Looking back over Shanghai's performance in the past 15 years, it is quite evident that nationwide reforms were responsible for creating an environment in which progress was feasible but that municipal policy actions were required to grasp the possibilities. There were extemalities to be realized from industrial depth and neighborhood effects to be ex- ploited, there was increasing scope for domestic resource mobilization, and there were inefficiencies of the state sector that could be remedied. But all these tasks demanded activism on the part of municipal and cen- tral authorities. By sketching the reform context and the unfolding of policy in Shanghai, it is possible to show how policy promoted urban modemization in a transition economy. Six reform events serve to frame the experience of Shanghai. The first brought into existence the household responsibility system in the rural areas and, in stages, curtailed production quotas, state procurement at fixed prices, and constraints on the marketing of agricultural output. The result was a vast increase in agricultural productivity, which raised incomes and widened the markets for industrial products of cities such as Shanghai. Rural counties and townships used their newly gained free- dom to expand industry, which generated employment, catered to rural demand, and gradually branched into a wide range of production ac- tivities. Reform induced rural development in Jiangsu and Zheijiang and enormously increased the prosperity of Shanghai's hinterland. Mar- ket linkages, greatly attenuated by socialist planning and control for more than three decades, resurfaced and provided stimulus and competition for Shanghai's industry, breathing new life into a regional economy that had been marking time throughout the 1970s. A second change, instituted piecemeal, is the reversal of past empha- sis on attaining self-sufficiency and on subdividing the country into a matrix of weakly connected provincial markets (Shue 1988). At first cau- tiously but then more openly, Beijing supported the formation of re- gional markets and their eventual coalescence into a national market. Shanghai: Renaissance City 83 Although initially painful for enterprises accustomed to being sheltered from competition, market integration has been crucial for realizing in- dustrial scale economies (World Bank 1994a). Third, the full benefits of market integration and a rich hinterland have been predicated on the successful implementing of enterprise re- form. One of the reasons why China's urban industry has registered such a dismal record of productivity is its excessive vertical integration. Most large enterprises attempted to produce three-fourths or more of all essential intermediate inputs to minimize reliance on outside suppli- ers. Since the mid-1980s many enterprises have been encouraged to ra- tionalize production and have made full use of a more open environ- ment to subcontract components, specialize, and establish networking arrangements through which to exchange information and share overheads such as marketing or research. Thus enterprise reform, in conjunction with market integration, is critical to fully realizing agglom- eration economies. Fourth, one kind of neighborhood effect arises from the immediate hinterland. A second set is associated with FDi and international trade. The former is an important conduit for technology transfer at many lev- els; the latter enlarges market possibilities and provides the spur of com- petition. Starting with the joint venture law in 1979, which permitted foreign investment in four Special Economic Zones in Fujian and Guangdong, the Chinese government then opened another 14 coastal cities to overseas investors in 1984. Thereafter, a series of decrees elabo- rated the legal framework governing foreign investment and widened provincial discretion in dealing with investors. Apart from affording foreigners the right to invest in China, the government slowly loosened its monopoly over trade, decentralized the trading system, and allowed partial retention of foreign exchange earnings and, eventually, the trad- ing of foreign exchange through special trading centers. Guangdong was the first to obtain these trade privileges, which were extended to Shanghai and Tianjin in 1983. For a variety of reasons, which are discussed in chapter 5, Guangzhou was the quickest to introduce reforms for the purposes of local development. The others followed with a lag. The fifth reform initiative of far-reaching consequence was fiscal de- centralization and the incentives for financial widening touched on above. Decentralization set the stage for intensive resource mobiliza- tion. It also enabled cities to start a virtuous spiral through investments that raised growth directly and via larger agglomeration economies. Higher growth improved revenue prospects and generated more capi- tal, which fed the impulse to grow. The earliest reforms, other than those with a rural focus, were aimed at the southeastern coastal provinces. Other coastal cities began to benefit from the enabling environment first tested in Guangdong only about the middle of the 1980s. 84 The Dynamics of Urban Growth in Three Chinese Cities Last, although Shanghai began asserting at least its fiscal rights after 1983, the city initially showed little determination to define and imple- ment the policies necessary to modernize the economy. Even though Shanghai was an "open city," and foreign investors were clearly inter- ested, the authorities did little to reduce red tape and use the incentive mechanisms at their disposal to bid for overseas capital. When, under prodding from the center and in response to the sluggishness of the municipal economy, the local government began trying in earnest, FDI increased sharply. Similarly, the urgency in dealing with the problems of municipally controlled state enterprises was slow to register. Reform- ing state enterprises posed formidable difficulties, political as well as economic. Tinkering on the margins was clearly insufficient, and a more determined assault would challenge powerful vested interests. It was easier to let matters drift. Only the insistent pressure of fiscal need and the central government's unwillingness to bail out the city induced Shanghai from about the late 1980s to pursue a multipronged restructuring strategy with the assis- tance of international agencies. One prong encompassed vertical disin- tegration, subcontracting with township and village enterprises, and the shifting of production facilities to neighboring provinces, where land and labor are cheaper. Slow to gather momentum, these tendencies are accelerating with amazing speed now that the commitment of the local government is unequivocal. A second prong is the creation of "internal markets" for restructuring. Industrial bureaus are helping to set up in- dustrial corporations, which bring together affiliated producers under a single organizational umbrella so as to precipitate the necessary ratio- nalizing of production and trimming of overhead. The building of a social security system independent of the enter- prise-the third prong-frees firms of onerous obligations, weakens the iron bonds of tenure, and introduces flexibility into the labor market. Fourth, there is bankruptcy, which Shanghai was chary about using so long as its growth performance was weak and employment generation modest. As the economy has gathered speed and job prospects have multiplied in the service sector, resistance to closure of insolvent enter- prise is less troublesome. Shanghai has begun translating the central goverunent's broad mandate for enterprise reform. In many respects this is the most difficult hurdle on the road to modernization, but it is not an obstacle to be sidestepped. Enterprise reform has to be pursued to an industrially and fiscally acceptable conclusion through an incre- mental application of policy. Shanghai's efforts to raise fiscal revenues, the remaining prong, have already been discussed. However, the city's initiatives in mobilizing fi- nancial resources deserve further attention. The central government's financial reforms selectively allowed certain provincial entities the right to issue construction bonds domestically, to raise money in international Shanghai: Renaissance City 85 capital markets, and to permit enterprises to acquire funds by issuing shares to workers or others. Borrowing from the international market commenced on a modest scale in 1986, but Shanghai began lobbying the central authorities for greater access to funds from international agencies toward the very end of the 1980s. Once permission had been granted, Shanghai was able to raise $4.4 billion by 1991. The issuing of construction bonds has grown steadily as the city acquires experience and markets gain in maturity. Less has been done to tap local savings to augment enterprise resources. County-level firms in neighboring provinces are far ahead of the game. However, Shanghai's industrial structure is shifting toward collective and private enterprises, again with support from the authorities. These types of enterprises are far readier to explore a variety of avenues, and the use of new financial instruments sanctioned by the authorities is sure to increase. Shanghai's many advantages are clear: an industrial tradition, a broad manufacturing base, human capital, and a strategic location. It must use these assets, the benefits conferred by agglomeration and by spillovers from surrounding economies, and its ability to mobilize resources to sustain its standing as China's premier industrial metropolis. Through farsighted policies and judicious investments, Shanghai can become a world-class center for producer services. But it must make haste to rem- edy its infrastructure shortcomings, to give sustained attention to the weaker parts of its industrial system, and to meet its social security ob- ligations to an aging workforce in a manner that is socially just while being financially sustainable. Any one of these tasks is well within Shanghai's capabilities. To accomplish all three simultaneously-because that is what the situation demands-calls for creative solutions that squeeze the utmost from each yuan of outlay. A Port, Its Tianjin: Neighborhood, The first impression many have on arriving in Tianjin from Shanghai is a sense of spaciousness, a slow pace of life tied to the rhythms of the bicycle, and a quiet. The quiet is deceptive. Tianjin's industrial reach is almost as great as Shanghai's. Smokestacks with their ubiquitous plumes are everywhere feeding the brown smog that drapes the city. W(hat life remains in the Haihe River fights a losing battle against the ferocious effluent from industries of every stripe: petrochemicals, electronics, metallurgy, dyeing, and textiles. Tianjin lies near the core of China's old industrial heartland. It is a city where state-owned enterprises still domi- nate industrial activity and their influence pervades all economic insti- tutions. Tianjin's story reveals another side of urban development in China. Although its recent economic history closely parallels Shanghai's own experience and Tianjin has many of the same industrial attributes, the difference in economic geography has influenced the dynamics of agglomeration, the pattern of change, and the readiness to challenge socialist practices with the weapons of reform. Because of its location, resource endowment, and proximity to Beijing, Tianjin's options are narrower and its feasible development strategy is better delineated. Tianjin's experience can broaden our understanding of urban devel- opment in a transition economy in four respects. First, it indicates how the expansion of transport facilities can increase the fruitfulness of in- teraction with the hinterland and enable a city to derive additional im- petus from entrep6t trade. Thus transport can offset the industcrial and agricultural weaknesses of the immediate hinterland, promoting trade and the flow of goods from farther afield. Second, Tianjin's deepwater harbor, the presence of mineral resources in neighboring counties, and their availability for industrial purposes have sharpened the choice of 86 Tianjin: A Port, Its Neighborhood, and Its Ambition 87 manufactures. Third, Tianjin's accumulated industrial skills and possi- bility of pooling its human capital with that of Beijing open possibilities in the realm of high-technology industry. Fourth, the large share of state industry within the municipality and the absence of rapidly growing town and village enterprises in the vicinity of the city have slowed the restructuring of the municipal economy. Existing institutions, which af- fect the allocation of labor, cannot be challenged so long as alternative employment prospects are few and industrial workers are the domi- nant political force. Resource Endowment and Growth The Tianjin municipality covers an area of 11,305 square kilometers, and the city proper stands on 222 square kilometers, some 16 percent less than Shanghai. The urban population in 1993 was 5 million, three-fourths of the population of its sister metropolis to the south. The total popula- tion of metropolitan Tianjin was a little under 9 million, compared with 2.69 million in 1953 (table 3.2). Aside from a lower population density, living space per capita is larger than in Shanghai. In the period before liberation, Tianjin was a jumble of unplanned, poorly articulated neigh- borhoods, a legacy of the largely self-contained foreign concessions that appeared after 1860. Piecemeal investment in urban infrastructure gradu- ally imposed a new topography on the city. But the city's urban infra- structure and layout improved significantly when the central govern- ment mounted a drive to repair the damage caused by the Tangshan earthquake of 1976 (which caused extensive damage to structures and killed 25,000 people in Tianjin alone; see Chang, Hu, and Sun 1992). Start- ing in the late 1970s, Tianjin constructed 14 new residential areas, a ma- jor sewage treatment plant, a complete-ring road system, and freeway connections to Tanggu and Beijing. The highway network now integrates intracity routes with intercity trunk lines. Road space per capita, which was well above the level for the two other cities and greater than the national average in 1990, grew from 4.9 square meters to 7.1 square meters by 1994 (table 3.6). Eight hundred years ago under the Yuan dynasty, the town was known as Haijinzhen, or Sea Ford Town. Famous among sailors, the port of Tianjin flourished in spite of problems caused by siltation, the obstacle posed by the Taku Bar at the mouth of the Haihe, and ice buildup in the Bohai Sea during winter. Divided into three parts-Tanggu, Tianjin, and Xingang-Tianjin harbor is the largest in north China. Railway links with Beijing and Shanghai to the west and south and to Harbin in the north enable the harbor to service a number of key economic regions, such as the northeast and northwest. A railway line connecting Tianjin to its satellite port of Tanggu was built in 1888. Later construction established lines to Qinhuangdao and Tangshan. By 1912 railway connections ex- Table 4.1 Education of the Labor Force in Selected Sectors in Tianjin, 1990 Industry Transport Research and development Government Lezvel of education Thousands Thousands Thousands Thousands and type of work of employees Percent of employees Percent of employees Percent of employees Percent Total 1,927.9 207.8 47.3 198.1 Level of education University 31.1 1.6 2.6 1.3 14.2 30.1 12.6 6.4 Community college 69.6 3.6 5.6 2.7 8.2 17.2 31.4 15.9 Vocational school 95.2 4.9 11.1 5.4 5.8 12.3 26.8 13.5 Senior high school 458.4 23.8 46.1 22.2 7.8 16.4 52.2 26.3 Junior high school 914.3 47.4 103.2 49.7 9.3 19.6 53.0 26.7 Elementary school 330.7 17.2 36.6 17.6 1.9 4.0 18.5 9.4 None 28.6 1.5 2.5 1.2 0.1 0.3 3.6 1.8 Type of work Technical 181.3 9.4 20.1 9.7 25.1 53.0 49.1 24.8 Administration 74.4 3.9 5.5 2.7 3.8 7.9 47.4 23.9 Staff 66.7 3.5 14.7 7.1 2.9 6.2 61.2 30.9 Sales 62.7 3.3 4.8 2.3 0.9 1.9 5.6 2.8 Services 124.5 6.5 15.9 7.7 2.1 4.5 12.2 6.2 Production 1,414.2 73.4 146.2 70.4 12.3 26.1 21.6 10.9 Other 4.2 0.2 0.4 0.2 0.2 0.4 0.9 0.5 Note: Numbers may not sum to totals because of rounding. Source: State Statistical Bureau 1992b. Tianjin: A Port, Its Neighborhood, and Its Ambition 89 isted with Shanghai and Wuhan. By the 1930s, 10 to 12 percent of China's trade passed through Tianjin, amounting to cargo throughput of more than 2 million tons. Among the principal exports were wool and raw cotton, while imports serviced the demand for oil, timber, and manu- factures (Todd 1994). In contrast to Guangzhou and Shanghai, Tianjin has a substantial en- dowment of natural resources. The Dagang oil field 60 kilometers away yields 4 million tons of low-sulfur crude a year. Another 1 million tons is produced from nearby offshore fields. Under the suburban counties of Baodi and Jixian, 500 million tons of coal have been verified but so far remain untouched. Salterns in Hangu and Tanggu are the source of Changlu salt for the chemicals industry. And abundant geothermal en- ergy sources are located in the vicinity of the municipality. Whereas scar- city of land for industrial purposes severely constrains the two other cities, plenty of underused land around Tianjin is available for indus- trial purposes. However, unlike Shanghai, Tianjin lacks a rich agricul- tural hinterland comparable to the lower Yangtze basin and is not linked to a powerful, rural-based manufacturing system such as that of south- ern Jiangsu and Zhejiang. However, its proximity to Beijing provides a degree of industrial reinforcement. In the past, Beijing's presence was a mixed blessing, on occasion diverting resources and attention away from Tianjin. More recently, complementarity between the two cities and the scope both for collaborative ventures and for efficient specialization have opened promising vistas in technology. To exploit these, Tianjin will rely on its reserves of human capital as the building blocks for a research network. There are 29 universities and colleges, more than 140 research institutes, and 200 technical schools. Most workers have a junior high school education or better, and roughly 300,000 employees have some technical qualifications (see table 4.1). At the same time, costs per worker, after factoring in the entire suite of benefits, are 10 percent less than the costs for an equivalent employee in Beijing, 35 percent less than in Shanghai, and a full two-thirds less than in Guangzhou ("Heavenly Investment," Business China, July 12, 1993, pp. 8-9). Its resource position notwithstanding, Tianjin lagged behind Guangzhou and Shanghai between 1952 and 1978 in terms of GDP growth (see table 3.1). GVIO also rose more slowly and, in addition, was below the national average. Throughout the period, Tianjin was overshadowed by Beijing, where industrial growth averaged 14.5 percent annually. GDP growth was only marginally better during 1978-93, with much of the improvement taking place in the early 1990s. In this period, the trend in Tianjin's GVIO was superior to that in Shanghai, but only three-fourths the rate in Guangzhou. The sluggish performance relative to the na- tional average through the 1980s draws attention to structural impedi- ments and the cautious implementation of reform, although the spurt in growth since 1991 suggests that progress on both fronts is being made. 90 The Dynamics of Urban Growth in Three Chinese Cities Industrial Development and Services Tianjin, much like Shanghai, started its industrial life as a producer of textiles and light manufactures for the north China economy. In turn, these industries spawned affiliated engineering and metallurgical ac- tivities. That was where matters stood till the mid-1950s, when the cen- tral government promoted a steady widening of the industrial base with emphasis on machine building, chemicals, metals, and manufacturing. By 1985 heavy industry accounted for 49 percent of industrial output, although textiles, together with garments, retained 16 percent of indus- trial production (see table 4.2). During the mid-1980s the 10 largest subsectors were responsible for more than 70 percent of GVIO valued at current prices. In several respects, this industrial structure hampered growth and was disadvantageous for Tianjin. Continued dependence on textiles and garments was a significant handicap. Rising wages and the backwardness of manufacturing, dyeing, and design facilities eroded Tianjin's competitiveness. Furthermore, the weight of these industries in the total was a constraint on productivity. Tianjin's considerable in- dustrial breadth was achieved at some cost. Not only did the city lack a few economic foci to orchestrate development, but its national indus- tries, unlike many of Shanghai's producers, were rarely leaders in their fields (on the comparative advantage of selected subsectors, see chapter 6). Tianjin attained a well-deserved reputation in bicycle manufacturing and was one of the foremost suppliers of television sets, but even in these industries, it was not a force for technological change. In machine tools and metalworking, it was in the front ranks but never dominated either subsector. The metallurgical industry is Tianjin's largest manufacturing subsector. Private sector firms constructed steel rolling mills with two Martin furnaces in 1937. Output in 1952 was 62,000 tons. It rose to 130,000 tons in 1957 and increased further in the early 1960s following the in- stallation of eight Bessemer converters (Etienne 1990). An alloy steel unit began operating in 1969, and there commenced a major expansion in capacity based on imported equipment. After three years of expan- sion, Tianjin became one of the seven major production bases for iron and steel in China. In 1993 it produced more than 2.2 million tons of steel and 1.5 million tons of pig iron, adequately meeting current local demand. Tianjin is an established medium-size steel production center, and in 1993 its output ranked eighth in the country. It has developed a complete chain of production, embracing iron foundries, steelmaking, rolling mills, and final products. Steelmaking and other metallurgical industries dovetail effectively with downstream producers of machin- ery and transport equipment, creating a vertically integrated system of considerable competitive potency. Much like similar industrial cities in Japan and Korea, Tianjin can be a key player in several strata of heavy industry. Tianjin: A Port, Its Neighborhood, and Its Ambition 91 Table 4.2 Gross Value of Industrial Output in Tianjin, Selected Years, 1985-93 1985 1991 1993 Enterprises Billions of Billions of Billions of in 1993 Indicator yuana Percent yuan' Percent yuan' Percent Number Percent City total 29.10 100.0 59.93 100.0 106.81 100.0 12,454 100.0 Light industry 14.79 50.8 27.47 45.8 35.48 33.2 6,813 54.7 Heavy industry 14.31 49.2 32.47 54.2 71.33 66.8 5,641 45.3 State-owned enterprises 23.23 79.8 43.64 72.8 62.88 58.9 2,548 20.5 Collective-owned enterprises 5.34 18.4 10.94 18.3 24.28 22.7 7,874 63.2 Others 0.53 1.8 5.35 8.9 19.66 18.4 2,032 16.3 Top subsectors in 1993 20.50 70.4 41.76 69.7 81.92 76.7 - - Metallurgyb 3.02 10.4 7.60 12.7 17.73 16.6 - - Machine building' 4.97 17.1 8.38 14.0 15.13 14.2 - - Transport equipment 1.11 3.8 3.49 5.8 12.38 11.6 - - Chemicals 2.51 8.6 6.46 10.8 8.81 8.3 - - Textiles 3.77 12.9 5.44 9.1 6.56 6.1 - - Metal products 1.44 4.9 2.86 4.8 6.37 6.0 - - Electronics 1.31 4.5 2.91 4.8 5.34 5.0 - - Petroleum processing 0.87 3.0 1.36 2.3 4.38 4.1 - - Clothing 0.76 2.6 1.84 3.1 2.74 2.6 - - Building materials 0.75 2.6 1.42 2.4 2.50 2.3 - - Herfindahl index 0.062 0.057 0.064 - Not available. Note: All numbers exclude village-run enterprises. a. Current prices. b. Includes smelting and processing of ferrous and nonferrous metals. c. Includes machine building and electrical machinery. Source: State Statistical Bureau, Statistical Yearbook of Tianjin, 1991, 1992, 1994. Tianjin was the first city in China to establish a production line for black and white television sets (in 1959), and over the years it capital- ized on this base of research and manufacturing to maintain its role as one of China's leading producers (Hussain, Lanjouw, and Li 1990). Re- search was fed by defense considerations, as in Westem countries, and the city acquired a complex of research institutes and factories that steadily added to scientific knowledge as well as production skills in the field of electronics. Once government policies began favoring con- sumer goods in the early 1980s, Tianjin quickly expanded the output of consumer electronics, especially television sets. To achieve this, the city imported assembly lines and technology from overseas and assimilated these within the existing industrial structure. By 1990 production of items such as black-and-white televisions, color televisions, and radios had risen manifold. Tianjin's Great Wall Television Company sold 70 per- cent of its black-and-white televisions in northern China, but it also ex- 92 The Dynamics of Urban Growth in Three Chinese Cities ported to Europe, the United States, and the Middle East, mainly for industrial use. Tianjin was also developing backward linkages into the high value added components that are at the core not just of televisions but of tele- communications equipment, industrial robots, and modem automobiles. Motorola's decision in 1992 to set up a plant to manufacture pagers was a breakthrough. It is introducing advanced, large-scale production meth- ods and giving incentives for the expansion of research. Because the electronics industry is intrinsic to progress in so many other subsectors, its development has large spillover effects. Starting in the second half of the 1980s, Tianjin embarked on a pro- gram of restructuring and specialization: textiles production was reduced in scale; some facilities were transferred to Handan, Shijiazhuang, and Xingtai; others moved to Xinjiang in the northwest; and large resources were devoted to upgrading the textiles sector.' After reappraising its competitive standing in other areas, the municipality narrowed its sights and opted for greater specialization in a more limited group of indus- tries: metallurgy, machine building, transport equipment, electronics, and chemicals. To finance this strategy, the city redirected its own con- siderable resources and, in addition, requested funding from the central government, from international agencies such as the World Bank, and from foreign investors, who are also the principal suppliers of technology. By the early 1990s the strategy had begun to yield results, albeit spar- ingly. Industrial growth accelerated after 1991. The broad sectoral ag- gregates had shifted, with the share of light industry falling to just over a third as Tianjin scaled down production in those subsectors where rural industry in other parts of China had gained a decisive lead. At the level of subsectors, much greater concentration was evident. The top four subsectors now accounted for 50 percent of output, and the share of textiles had fallen from close to 13 percent to a shade over 6 percent (see figure 4.1). In a handful of years, Tianjin had restricted the compass of its activities and carved out areas of core competence in certain choice heavy industries. In the face of a continuing integration of the national market and increasing competition from producers in other cities, the strategy is clearly a rational one. But the full benefits of agglomeration have yet to be exploited for the purposes of maximizing productivity and reforming enterprises. Like all other Chinese cities, Tianjin was lightly equipped with producer services. State-owned banks provided financial services, foreign trade cor- porations and other government organs handled all forms of commerce, and research centers embedded within the domain of an industrial bureau supplied consulting services. To the extent possible, enterprises struggled to be self-sufficient, doing in-house what firms in market economies have been subcontracting for decades. Without the adequate availability of pro- ducer services, industrial change is seriously hobbled. Tianjin: A Port, Its Neighborhood, and Its Ambition 93 Figure 4.1 Industrial Structure of Tianjin, 1985 and 1993 Building materials Clothing Petroprocessing Electronics 319 3 Metal products _1985 Textiles _ Chemicals 1 Transport equipment Machine building _ _ _- Metallurgy - - T Other 0 5 10 15 20 25 30 Percentage of GVIO Note: GvIo, gross value of industrial output. Source: Table 4.2. Early in the 1990s Tianjin finally began taking steps to deepen service activities. This was almost a decade behind Guangzhou and several years after Shanghai. As a consequence, industrial strategy was guided far more closely by bureau directives, with less support and feedback from the market. Tianjin's enterprises had to grope their way forward in an environment that is still quite poorly furnished with institutions and service providers relative to Shanghai and Guangzhou. Thus the pace of industrial reorganization has been more tentative, and the outcomes have been uneven. Nevertheless, Tianjin has been remarkably effective in lessening the preponderance of state-owned enterprises. In 1985 state- owned enterprises accounted for 80 percent of GVIO and joint ventures for a bare 2 percent. By 1993 the share of state-owned enterprises had declined to 59 percent. Collective-owned enterprises had increased their share by 3 percentage points to about 23 percent, but the output of joint ventures exceeded 18 percent (see table 4.2).2 Hence the proportion of more innovative enterprises, free from some of the constraints binding state-owned enterprises, was starting to leaven the industrial environ- ment, weakening the resistance to new technology and to changes in labor relations, flexible work practices, and new managerial techniques. However, compared with those of both Shanghai and Guangzhou, Tianjin's state sector remains dominant. The market power of state en- terprises can still deter the entrance of small private firms and put pres- sure on the ones already operating. Tianjin's large, vertically integrated firms have been under less pres- sure to restructure and to subcontract than their counterparts in Shang- hai. One reason is that they are still relatively insulated by transport costs from aggressive competitors in the coastal provinces to the south 94 The Dynamics of Urban Growth in Three Chinese Cities and are not sufficiently challenged by producers from the old industrial heartland of Manchuria. Another reason is the sparseness of rural in- dustry, which has been the source of economic buoyancy in other coastal provinces. Rural enterprises in the counties encircling Shanghai and Guangzhou have profoundly influenced the transformation of industry in the cities through production and market linkages. Production link- ages are composed of subcontracting deals that permit urban state-owned enterprises to focus on the principal operations and allow affiliated town and village enterprises to supply components. Market linkages are a combination of product competition and market demand. Although the effects of these on urban firms are difficult to quantify, casual empiri- cism suggests that the interaction between a city and its surrounding rural industry is of vital significance. Such symbiosis has yet to materialize in Tianjin.3 Possibly because the agricultural economy lying beyond the city is poor and the commercial tradition of Guangdong or Zhejiang is absent, rural industry of compa- rable vigor has not emerged. Tianjin's enterprises have therefore not yet managed to arrive at a mutually fruitful division of labor with enter- prises in the hinterland.4 Moreover, Tianjin's policymakers do not have a large enough nonstate sector to absorb the pain of industrial restruc- turing by sustaining growth momentum and absorbing displaced work- ers. For this reason, the political and social costs of enterprise reform have rendered change a perilous venture. If failing enterprises are al- lowed to go under, it is harder to place the unemployed, and the city is not sufficiently prosperous to absorb the cost of extensive layoffs. The voice of labor in Tianjin is far stronger than in Guangzhou and stronger, perhaps, than in Shanghai, where exit is less problematic and a limited social security system is easier to finance. The full benefits of agglom- eration are beginning to accrue, but slowly, because reformers must bal- ance political and economic concerns much more finely. Neighborhood Effects and Transport-Induced Development The neighborhood effects on which Tianjin is attempting to capitalize are different from those that are central to the strategies of Shanghai and Guangzhou. Exploring Tianjin's approach unveils the workings of ge- ography along different pathways. The two prime determinants of spillover benefits for Tianjin are its potential as a transport hub and its nearness to Beijing. A glance along China's northeastern coast and across the northern one-third of the country reveals, first, that, besides Dalian, no other sheltered, deepwater port is a match for Tianjin.5 Much of the developing trade of this highly industrialized region is being funneled through Tianjin harbor. Once China's northeastern provinces become more outward-oriented and jointly develop the region's surface trans- port network, the significance of Tianjin as an entrep6t can only increase. Second, the city is near the eastern apex of the triangular north China Tianjin: A Port, Its Neighborhood, and Its Ambition 95 plain, where it intersects with the northeast region and Inner Mongolia. The shape of the Bohai peninsula is such that Tianjin lies closer to the inland provinces of northwestern China-as well as Mongolia-than any other port. In fact, an expansive delineation of Tianjin port's hinter- land covers 1 million square kilometers embracing the municipalities of Beijing and Tianjin, five contiguous provinces (Gansu, Hebei, Shaanxi, Shanxi, and Qinghai), and three autonomous regions (Inner Mongolia, Ningxia, and Xinjiang). The current configuration of Tianjin port dates back to the early 1890s, when the Japanese colonial administration shifted the main port from the head of the river closer to the sea. Through the 1950s and 1960s traf- fic was divided between the upriver Tianjin port and the outpost at Tanggu Xingang. Gradually the former was relegated to the status of a river port, being finally eclipsed by the destruction wrought by the 1976 earthquake. From the early 1970s, container handling facilities were built at Tanggu Xingang, and Tianjin became part of the containerized traffic network that tied together several of China's coastal cities. By 1990 the port was equipped to handle 400,000 TEUS (twenty-foot equivalent units), although actual throughput was 287,000 TEUS. In 1993 the port had a throughput of 37 million tons and functioned both as a seaport (Tanggu Xingang) and as a riverport (Tianjin proper). By 1996 handling capacity had risen to 700,000 TEUs, which exceeds that of the major nearby ports such as Dalian, Lianyungang, and Ningbo; by early in the next century, Tianjin port expects to have a handling capacity of 100 million tons as against 60 million in 1997 ("China: The Bohai Sea Rim," Oxford Analytica, May 29, 1997; Ports and Harbors, July-August 1992, p. 32). The role of a busy entrepot serving multiple, rapidly developing re- gions can be exceedingly advantageous, as the experience of Singapore clearly underscores. By expanding its port facilities, systematizing and computerizing information requirements, and mechanizing all cargo handling, Singapore has made itself the busiest transport hub in Asia.6 Port operations and associated warehousing, trade-related services, and repair facilities have become a major source of income for the island republic and a pillar of Singapore's future prosperity. Tianjin is proceed- ing down a similar path, although some elements of its strategy are less well specified than in the case of Singapore: there is a strong emphasis on port development, which is under the control of the municipal au- thorities, whereas plans for pursuing transport projects that would speed the flow of cargo to and from the hinterland are much less certain. In 1992 the city began a major effort to turn the existing, largely com- mercial harbor with its 31 cargo and 4 passenger berths into an interna- tional port with multiple functions and modern facilities. It began con- structing new wharves and berths, improving container transportation, and laying special railway lines to allow easier access. The port now has the country's first 10,000-ton, three-sided freight berth. An expanded wharf is capable of handling 3.5 million tons of grain. In the southern Figure 4.2 Tianjin and Its Environs I4 11 6' SELECTED CITIES Cf ngd (SYMBOL SIZE PROPORTIONAL TO POPULATION) Zhangjiakou SALTPAN s--- RAILWAYS -.-- PROPOSED RAILWAYS EXPRESS ROADS Datong ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~PORTS eiling Qinhuangdao~~~~~~~~ X/0 50 100 KILOMETERS Lanfang Tagshan C~nI i~1'v SH SHNOO el OINGHAI - ' SI RencjIU~~~~~~ Daiclkeng AS HENIANJ JI t 0 Hunhu±RN~PN x 51~~~~~~~~~~~~~~~~~~RCHUAN ZEI > 111.11AN )I3~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~A c: 'VAN~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~GX AGDN -, ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ f~~~~~~~~~~s~~~'-,j ~VIETNAM> ',-- ' MCOPR 114' ~~~~~~~~~~~~~~~~ - DEMI HI I '0 N~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~0 Tianjin: A Port, Its Neighborhood, and Its Ambition 97 part of the port, berths have been developed for crude oil, coal, timber, and chemicals. The neighboring Qinhuangdao port ranks as the largest coal-export terminal in China. About $4.6 million has been invested in Tianjin to build China's largest unloading facility for liquid chemicals. The port is the first in China to have autonomy over collecting funds and expanding cooperation with overseas companies. It has a bonded free trade zone of 5 square kilometers, which is the only free trade zone in China for international transit trade (to Mongolia and the Central Asian Republics) and is similar to Hong Kong (FBIs-cHI-92045, March 6, 1992, p. 68). By the end of 1993 more than 2,300 enterprises had been established in the free trade zone, with total investment of $1.76 billion; goods valued at $1.3 billion were shipped into and out of the zone dur- ing 1991-93. However, the difficult fluvial environment of the port im- poses some constraints. Although Tanggu Xingang can accommodate container ships with 3,000 TEU capacity, it cannot yet berth the newer 4,000 TEU vessels, unlike its competitor, Dalian (Todd 1994). The attractiveness of port facilities depends also on their accessibility to users in the surrounding regions. Around Tianjin, the road system is of good quality and is bolstered by the network of waterways that per- mit delivery by barge. Before the advent of railways, the five rivers linked to the Haihe River-Beiyun, Daqung, Nanyun, Yongding, and Ziya- provided the avenues for distributing goods entering through Tianjin widely across the hinterland. Railway connections to the coastal areas and principal cities are also adequate. Tianjin is connected to the major industrial city of Lining and to provinces to the north: Jilin and Heilongjiang. It is also connected via Beijing to Shanghai and points south. Construction of highways, which will become increasingly imn- portant for the flow of containerized cargo, is a priority of the govern- ment. Seven highways of national standard link Tianjin with neighbor- ing cities, and there is a 142-kilometer expressway to Beijing (see figure 4.2). The city spent Y800 million on transport development in 1994 and has earmarked sizable funds for this purpose in the coming years. Tianjin airport has been upgraded to handle international flights. Currently it accommodates flights diverted from Beijing, but the expansion program initiated in 1994-95, once completed, will make it the mainland's largest airport.7 In spite of all this investment, the intermodal infrastructure that would permit Tianjin to serve its hinterland, especially the interior areas, re- mains inadequate. Neither road nor rail facilities leading to the western provinces are developed sufficiently, and containerized cargo still can- not penetrate much beyond Beijing and the neighboring parts of Hebei Province. Thus the advantages of the capabilities built up at Tanggu Xingang and at Tianjin airport have yet to be exploited. Building this infrastructure, initially to accommodate freight but later to facilitate high- speed multimodal passenger traffic, requires a cross-provincial effort 98 The Dynamics of Urban Growth in Three Chinese Cities spearheaded by the central government and involving, possibly, the in- fusion of foreign capital. By modeling itself as a regional transport hub, Tianjin has opened the door to growth of a variety of commercial services and transport-intensive industries. Warehousing is but one of the services linked with port and airport development. Of equal significance are storage, trucking, forward- ing, brokerage, insurance, and trade financing services. Together these gen- erate thousands of well-paid jobs, some of which are open to workers dis- placed by the restructuring of state-owned enterprises. Relative to open- ings in many other service occupations, such jobs call for a higher level of skills, and there is more scope for increments in productivity. The building of warehousing, container handling, and refrigerated storage facilities around the airport, together with the scheduling of air cargo services, could catalyze a response across several industries. It would reinforce the trans- port-led strategy and allow Tianjin to use neighborhood effects to a much greater extent. Industrial Prospects Tianjin's broad industrial base and the presence of an international air- port open up possibilities of a radically different kind. Over the past two decades there have appeared on the international stage light manu- facturing industries and some horticultural activities that depend on close interaction with their final markets and the ability to airship goods to destinations in high-income countries. Two products that loom large in this trade are of special interest for Tianjin: designer clothing and fash- ion watches.8 Each one of these could build on the success of existing production, and their export would involve acquiring and combining many new skills. Sophisticated air transport facilities would bring in foreign expertise and encourage Tianjin's most dynamic enterprises to view the world as their oyster. Garments and textiles have been a mainstay of Tianjin's industrial sector since early in the century. Now, as rising costs diminish the city's competitiveness in these fields, Tianjin is being forced to close some fac- tories, move upscale, modemize dyeing and finishing facilities, and search for niches, which the remnants of its industry can occupy. As in the case of Shanghai, fashion garments made in small lots for select Western markets are an obvious choice. But to establish itself, Tianjin's producers must forge connections with retailers and distributors over- seas. Beyond this, there is a need to improve design standards, quality control, and capacity to manufacture small lots on quite short cycles and to ship the clothing on schedule, usually by air, to their destina- tions. Many of the ingredients of such restructuring are coming together as the city confronts a market environment in which survival demands competitiveness. In some subsectors-including textiles-marketing, de- Tianjin: A Port, Its Neighborhood, and Its Ambition 99 sign, and a reputation for reliability as a supplier are being recognized as critical. But weaknesses in quality standards, flexibility of response, and speed of delivery will be overcome only through a concerted effort of industry as well as government. This will entail radical changes in managerial practices and attitudes of the workforce and parallel changes in technology as well as work organization. These changes will be slow to materialize without a credible hardening of the budget constraint sig- naled by closure of several major money-losing enterprises.9 Tianjin's mechanical watch industry is among China's largest indus- tries. It offers the prospects of entering the lower end of the automatic (and mechanical windup) watch market and, in conjunction with the electronics subsector, the mass market for quartz fashion watches. Here again (as the success of the Indian company Titan has shown), the keys to success are innovative design, marketing, production technology, and the ability to supply retailers across the world with efficiency and speed. The hurdles are higher in this area than in garments. Neither municipal enterprises nor private entrepreneurs have moved rapidly to develop producer services or to seek foreign partners who could provide capital together with skills. This is a serious drawback. The possibilities exist, but firms have not actively sought spillover effects to reorient and mod- ernize this industry. Shipping facilities of a high order make possible a class of industries, some already in existence and others that can be established. Among the industries that are present and could be expanded are petrochemi- cals, oil refining, and metallurgy. The first two use feedstocks from Dagang and salt from counties within the municipality. In the early 1970s three petrochemical plants and an oil refinery with a throughput of 2.5 million tons were built in the suburbs. Recently more petroleum and marine chemical projects have been completed, including a 140,000-ton ethylene project, a 200,000-ton polyester plant, and a 2.5-million-ton re- finery. Tianjin's efforts to build its petrochemicals subsector are being aided by the desire of European producers to shift capacity to China, which could constitute a $100 billion market by 2000.10 The Bohai Chemi- cal Group, the country's largest salt-chemical industrial group, was listed in 1994 on the Hong Kong Stock Exchange in an effort to raise addi- tional capital. It is embarking on joint ventures with Rhone-Poulenc to manufacture methionine, with Showa Denko of Japan to produce epoxy resins, and with Lucky Chemicals of Korea to establish a plant for poly- vinyl chloride. Fine chemicals are also a priority. The combined share of chemicals and petroleum refining in the municipal economy was more than 12 percent in 1993, making it the third largest subsector. Transport development is crystallizing neighborhood effects that are setting the stage for industrial progress in another area. The proximity to Beijing has had its drawbacks. In the distant past, Beijing's impor- tance as the administrative capital greatly enhanced Tianjin's role as a 100 The Dynamics of Urban Growth in Three Chinese Cities port and transshipment point. More recently, Beijing has competed against Tianjin for resources and industry. This might explain the slow pace of urban modernization from 1949 through the 1970s. Competition between the two cities has continued during the reform era. But as the spirit of industrial rationalization has taken hold, each city has begun to pursue a measure of specialization and to seek cross-metropolitan syn- ergy. The expressway connecting Tianjin with the capital is a first step, but a long one, toward building an industrial corridor that might eventually mimic Route 128 outside Boston. Whether or not such an industrial axis eventually exists will depend on the future of several industries, with microelectronics as a leading candidate. In fact, the elec- tronics sector can serve as a useful prism for understanding how Tianjin's industry might neutralize some of its constraints and join forces with industry around Beijing so as to create a concentration of high- technology manufacturing in north China. The Rise of a High-Technology Corridor: Electronics Industrial expansion in Beijing and trends in applied research have, to some extent, paralleled those occurring in Tianjin. In 1986 Beijing mu- nicipality designated a 100-square-kilometer zone surrounding Zhongguancun in Haidian District as a high-technology industry de- velopment zone offering tax holidays and 45 percent tax rebates on profits equivalent to those given by Special Economic Zones. An enterprise seek- ing certification to produce in the zone must ensure that a third of its employees have university-level education and that 3 percent of all rev- enue from sales is plowed into research." Within eight years, the zone attracted more than 2,000 enterprises and became the most successful of the country's 51 economic and technological development zones (Seki 1994). Some of the most advanced segments of Beijing's manufacturing es- tablishment are producing electronic equipment of all kinds but have staked out a position in personal computers, peripherals, components, and software. Suppliers of aeronautical equipment and other defense products have used their expertise to acquire a foothold in medical elec- tronics, an immense and fast-growing field. A major new electronics product that looms on the horizon is the wireless personal communica- tion system for short-distance use. Such devices, which are being per- fected in Japan, are ideal for China's geographically compact and densely populated cities, where a small number of cells can cover a large num- ber of people. Production of phones and cell stations could generate lucrative manufacturing opportunities far into the future.'2 China's market for computers grew 65 percent annually during 1990- 94, and this trend is likely to persist through the balance of the decade, with annual sales of personal computers projected to reach 6 million by Tianjin: A Port, Its Neighborhood, and Its Ambition 101 2000. Local and foreign producers are competing aggressively for a share of sales, which reached 1.15 million in 1995, with foreign producers such as AST, Compaq, Digital, and IBM accounting for 80 percent."3 Among Chinese producers, Legend had the largest slice of the market in 1995, followed by Great Wall, Long Chao, and several small producers. Both Great Wall and Legend have capitalized on the pool of skills in the Beijing area. In the case of Legend, scientists from Academia Sinica teamed up with a Hong Kong firm to produce their own products and serve as intermediaries for Apple, AST, IBM, and Sun Microsystems. This strategy has adroitly brought together Chinese talent, foreign technology, and Hong Kong capital. There is considerable scope for more collaboration along these lines and for the development of indigenous research ca- pacity and technology. Compaq has invested $1 million in a research center in Beijing, and other multinational corporations will find it ad- vantageous to invest in China, as they have in India and Israel. Many small start-ups are engaged in the specialist software and applications business. In particular, "hanzification" has opened new avenues, as soft- ware engineers refine the ability of computers to read and write Chi- nese characters.'4 With many other foreign manufacturers such as NEC, DEC, and Casio prepared to license technology or set up facilities in China, the Beijing-Tianjin Silicon Axis is moving from a vision to a stirring of action in the two cities and finally to actual agglomeration of industry that could one day create a long industrial bridge. In both Tianjin and Beijing, the rise of electronics is being supported by backward linkages with engineering, transport, and telecommunica- tions industries, whose own advance is keyed to the successful incorpo- ration of microchip technology. Another linkage that has begun acquir- ing significance as China integrates with the international economy is the flow of contracts for software and data processing from abroad. This is creating openings for businesses that are the handmaidens of a suc- cessful, full-service electronics sector. Thus at the two ends of the new expressway, many "Silicon Shoots" are in evidence, as the two cities begin mobilizing their reserves of human capital and nudging industry along new pathways. Whether a "Silicon Corridor" will emerge and whether Tianjin will be able to contribute to its building will depend on whether policies are woven to minimize constraints on high-technology industries that are rich in spillovers. The semiconductor industry, which constitutes the axis of the elec- tronics subsector, illustrates both the nature of the constraints and the policies to deal with them. Semiconductor production is one of the pil- lars of the industry and a technology driver."5 It is also the entry point for countries seeking to establish a strong electronics complex. The tech- nology is relatively accessible, and competitiveness is a function of pro- duction scale, clean room practices, and process skills. However, as Korea's experience shows, entry costs are high because a plant of opti- 102 The Dynamics of Urban Growth in Three Chinese Cities mal size for producing a mass-produced item such as DRAMs or SRAMS involves a large outlay of capital; several years of learning are needed to produce defect-free chips that are on a par with industry standards; and innovation or the manufacture of higher value added items, such as ASiCs, requires heavy expenditure on R&D. Abimodal industrial structure seems to have several advantages. Frequently, small new entrants introduce new technology, experimental products, and fresh ideas; produce cus- tomized chips; and design application-specific software. However, only large firms with deep pockets can undertake mass production, which greatly reduces unit costs but is exceedingly capital-intensive. Role of State-Owned Enterprises Tianjin has an abundance of big state enterprises that are controlled by both the central government and the municipality. The municipal au- thorities have adopted policies that accommodate collectives and joint ventures and have increased the number of entrants, so the emerging industrial structure is achieving a good balance between state and nonstate enterprises. But problems remain. First, the established state- owned enterprises, while long on experience, are short on process skills of the sophistication needed to produce advanced electronics. Quality control, clean room practices, and in-plant research capability are par- ticularly backward. Second, red tape still hampers the entry of new firms and the product diversification of existing large state-owned enterprises. The planning and control mentality, although in retreat, continues to interfere with industrial functioning. Third, specialization among firms and networking are crucial in the field of electronics. Without such in- teraction, firms cannot internalize spillovers, shorten the product cycles, or keep costs in check. Spillovers also depend on labor mobility between firms, which acts as a transmission belt for ideas and production prac- tices. Since the late 1980s, Tianjin municipality, together with the central government, has accelerated the tempo of industrial restructuring, which includes networking and loose conglomerate arrangements. It has pushed labor contracting so as to introduce flexibility into the job mar- ket. Both initiatives have begun to influence behavior, but ingrained habits and beliefs do not change overnight. Much like other Chinese cities that are implementing reforms, such as Shanghai, Shenyang, and Wuhan, the first few steps are the hardest. Once some momentum is attained, progress can be much more swift. A "Silicon Corridor" will emerge and flourish over the long term if technological advancement is a prime and consistently pursued goal across an interrelated range of fields encompassing components and pro- duction equipment."6 Such simultaneous effort in several areas lessens the dependence of firms on outside suppliers and eventually reinforces technological momentum. To achieve minor but cumulatively signifi- Tianjin: A Port, Its Neighborhood, and Its Ambition 103 cant innovations necessarily calls for networking among many small niche-targeting companies and involving large firms capable of devel- oping and mass producing a promising innovation. Networking princi- pally means the sharing of technology, joint product development, and subcontracting, which allows for specialization, the pooling of skills, and the internalizing of many links in the production chain. Technical Manpower and Foreign Direct Investment Tianjin and Beijing together have the largest slice of research manpower in China and several of the most prestigious universities. There are many government-owned research laboratories in the area, some of which are well equipped. In any case, acquiring equipment is not a vast expense. Furthermore, the government-through state-owned enterprises-is a reliable source of contracts, especially for R&D on electronics, which of- fers growth prospects for large state-owned enterprises and smaller nonstate enterprises alike. The Tianjin municipal authorities, taking their cue from other open cities and neighboring East Asian economies, have created an economic and technology development zone (ETDZ). Cover- ing an area of 24.8 square kilometers, it is Tianjin's equivalent of the technology development zone in Beijing. By the end of 1996, 2,800 foreign-funded enterprises had been es- tablished, with foreign investment amounting to $8.0 billion, as had a small number of local start-up and joint ventures ("China: Bohai Sea Rim," Oxford Analytica, May 29, 1997). Some of these enterprises produced high-technology products. With a GVIO of 1.45 billion yuan, the Tianjin ETDZ was the largest and most profitable in China. Among investors in the Tianjin area were a number of multinational corpo- rations, from Japan, the United States, and Korea, whose willing- ness to remain in Tianjin and expand their operations will certainly boost the prospects of a technology corridor. Mention has already been made of Motorola's investment. The company's $300 million state-of-the-art factory produced $1.5 billion worth of pagers and cellular telephones in 1994. Even though the venture has yet to achieve profitability, Motorola is planning to invest another $1 bil- lion in production facilities, which includes a $500 million semicon- ductor wafer plant, also in Tianjin. Others include Otis from the United States, NEC from Japan, Henkel from Germany, Zanussi from Italy, and Karry from Hong Kong. More FDI targeted toward selected areas of the electronics industry could provide a much needed spillover for local producers, linkages for input suppliers, and a demonstration effect to stimulate the entry of home-grown producers. Tianjin can learn from China's southeastern provinces how to entice FDI in sufficient volume into the desired indus- tries (see table 4.3). Two other policies thus far neglected have the po- tential to stimulate the growth of electronics. One is the provision of Table 4.3 Foreign Investment in Tianjin, 1986-93 (millions of U.S. dollars, unless otherwise noted) Indicator 1986 1987 1988 1989 1990 1991 1992 1993 1979-93 Utilized foreign investment 134.8 223.4 242.7 167.8 98.6 260.9 831.2 935.0 4,016.5 Loans 83.4 90.3 181.5 136.3 61.6 128.3 599.2 391.1 2,739.8 Direct investment 51.4 127.4 31.9 28.0 34.9 132.2 231.4 541.2 1,218.0 Other - 5.7 29.3 3.4 2.0 0.5 0.6 2.7 58.8 Contracted foreign direct investment, by source - 14.4 91.7 84.6 163.7 196.6 1,219.3 2,255.7 4,270.2 Hong Kong (China) - 6.7 28.2 42.3 87.0 92.8 629.2 996.5 2,002.7 Percentage of Tianjin's total - 46.8 30.8 50.0 53.2 47.2 51.6 44.2 46.9 Percentage of Hong Kong's total - 0.4 0.8 1.3 2.1 1.3 1.6 - - Japan - 2.1 18.6 25.1 16.3 6.6 90.1 113.2 307.2 Percentage of Tianjin's total - 14.9 20.3 29.7 10.0 3.4 7.4 5.0 7.2 CD Percentage of Japan's total - 1.0 6.2 7.1 7.6 0.9 4.2 - - United States - 0.9 4.9 4.7 19.5 32.2 237.8 337.9 687.8 Percentage of Tianjin's total - 6.1 5.3 5.6 11.9 16.4 19.5 15.0 16.1 Percentage of U.S. total - 0.3 1.6 0.9 5.8 6.7 7.6 - - Taiwan (China) - - - 4.8 13.7 18.6 90.9 288.1 411.7 Percentage of Tianjin's total - - - 5.7 8.4 9.5 7.5 12.8 9.6 Percentage of Taiwan's total - - - 1.1 1.4 1.5 1.6 - - Korea, Rep. of - - - - - - 19.9 172.0 211.4 Percentage of Tianjin's total - - - - - - 1.6 7.6 5.0 Percentage of Korea's total - - - - - - 4.8 - - Others - 4.6 40.0 7.7 27.2 46.4 151.5 348.0 649.3 Percentage of Tianjin's total - 32.2 43.7 9.1 16.6 23.6 12.4 15.4 15.2 -Not available. Source: State Statistical Bureau, Statistical Yearbook of China, 1987; 1988; 1990; 1992; and 1993; data by source for 1987-91, State Statistical Bureau 1992a; 1992 and 1993 data and cumulative figures from State Statistical Bureau, Statistical Yearbook of Tianjin, 1994. Tianjin: A Port, Its Neighborhood, and Its Ambition 105 venture capital to launch products developed jointly by Tianjin's uni- versities and entrepreneurial researchers or other sponsors. State-owned enterprises are prepared to sponsor initiatives, and county authorities are ready to put up risk capital. At this juncture, institutional arrange- ments that enable the research establishments of Tianjin and also Beijing to make commercial use of research results could truncate what might otherwise be a slow maturation of university-business partnerships. The Massachusetts Institute of Technology has shown that the use of ven- ture capital can be unusually fruitful in pushing university-based re- search down the path of commercial realization (Saxenian 1994; Kargon and Leslie 1994). Harvard has taken similar steps to commercialize re- search. In 1988 it created Medical Science Partners Inc., a $36 million venture capital fund managed by Harvard Medical School to back mar- ketable medical discoveries.17 Research Acquiring effective R&D skills is time-consuming. It demands experi- ence, project management capability, knowledge of markets, and de- tailed understanding of not just the frontier of a field but also the direc- tion in which it is expanding. There are no shortcuts, although there are ways of tapping into Western research networks and thereby telescop- ing the acquisition of knowledge and the development of products."8 FDI is, of course, one obvious route, as are alliances with foreign companies. A second approach is to increase incentives for R&D by the large state- owned enterprises and to encourage small collective-owned enterprises, private firms, and joint ventures to set up research facilities that cater to the needs of many small producers. Even in the mid-1990s, big state- owned enterprises have made minimal outlays on research and have minimal in-house capability. Testing, quality control, troubleshooting, new product development on a limited scale, and occasional reverse engineering define the range of enterprise activities. Few product and process innovations emerge from all this. Even the giant state-owned enterprises in Tianjin lack the strategy, managerial aptitude, or quality of the workforce needed to participate in the process of technological change, which is one of the prime determinants of a competitive strat- egy to attain international competitiveness. Licensing arrangements offer other possibilities that strengthen medium-term prospects and build longer-term potential. Many enter- prises in China are already exploring the full range of possibilities, and competitors in nearby Beijing are beginning to prod Tianjin's enterprises into action. To take one well-known example, Beijing's Stone (Si Tong) Group has emerged as a leader in a number of fields, including com- puter software, office automation (especially typewriters and cash reg- isters), and electronic components. Stone Group has accomplished this by employing deft entrepreneurship and mobilizing scientific manpower: 106 The Dynamics of Urban Growth in Three Chinese Cities nearly 70 percent of its employees are university graduates. In addition, it has entered into a number of fruitful joint ventures with Matsushita, Mitsubishi, and Mitsui to produce word processors, printers, and floppy disks (Financial Times, April 30, 1996). Stone Group's strategy of using abundant human capital, mainly from Beijing, to develop competence in a few technology-intensive fields is something that enterprises in Tianjin are beginning to emulate but must pursue more aggressively. Colonizing Auto Industry Niches Other industries already established in Tianjin could capitalize on ag- glomeration economies and the benefits from externalities, annex a large share of the expanding market in China, and be able to compete in the export sector. Among the most noteworthy is the automobile industry: Tianjin ranked sixth among the seven principal Chinese manufacturers of autos in 1994. This is a sub sector with many linkages backward as well as forward. Tianjin's diversified industrial structure already encompasses many of the industries that support automobile production. For instance, steel, glass, petrochemicals, machine tools, electronics, and component manufactures are well established and could support an expanded auto industry. Some years ago, Tianjin, which was already assembling autos, made a big decision with respect to future output mix: it entered into a licensing arrangement with Daihatsu to assemble a minivan and a minicar, the Charade. In the process, Tianjin laid claim to two of the fastest-growing segments of the automobile market in China. The minivan is a practical and cost- effective vehicle for businesses wanting to transport passengers and cargo. The Charade is relatively simple to produce, ideally suited to China's narrow urban streets, and low priced. This puts it within reach of China's emerging middle class, which is composed of two-earner families engaged in either business or the professions. New, flexible process technologies and lean production methods make it possible for medium-scale producers to achieve unit cost levels comparable to those of large multinational corporations. The experience of Daihatsu (and also of Renault) has clearly shown that manufacturing for selected niches can be a viable strategy for medium-size producers.19 The Tianjin Auto Industry Corporation entered into a licensing agree- ment with Daihatsu in 1986. In 1988,2,873 cars and 9,329 minivans were assembled. The number rose in 1996 to almost 88,000 cars. Production of both types of vehicles is projected at 150,000 in 1997 (Maxton 1994; Harwit 1995; "Toyota Targets Chinese Markets," Financial Times, Novem- ber 16, 1995; "Long March to Mass Market," Financial Times, June 25, 1997). Both vehicles have been in heavy demand throughout China, and trends point to steadily rising sales. Minivans are also the preferred mode of transport in rural areas, where the roads and repair facilities are not suited for more sophisticated vehicles. Tianjin: A Port, Its Neighborhood, and Its Ambition 107 By 1996 sales of autos in China had risen to 1.6 million, of which 382,000 were cars, most of them manufactured domestically. By some estimates, the market for cars alone could double by 2000. Demand for minivans is also likely to increase rapidly. Undoubtedly the competition for such vehicles will be sharp because they are particularly suited to the profiles of likely buyers, whether individuals, enterprises, or taxi companies. However, Tianjin has the benefits that accrue to an early entrant, and it has a local manufacturing base that few cities in China can rival. Hence Tianjin could evolve into a major center for China's auto industry, and in certain respects it is offering a more attractive mix of products. There are hurdles to be crossed, and these bring us back to weak ag- glomeration effects. First, Tianjin's partner in the automobile venture is one of the smaller Japanese producers. Daihatsu is partially owned by Toyota, which only recently began to assign substantial resources to widening the demand salient created by its subsidiary. Furthermore, the licensing arrangement with Tianjin Auto was quite frugal in transfer- ring skills, process technology, design capability, and all the other fac- tors that determine manufacturing strength. This will change with the increase in Tianjin's market share and Toyota's decision to enter the market.20 The Japanese company set up a technical assistance center to explore opportunities for manufacturing components in the early 1990s. In 1996 it entered into a $250 million joint venture with the Tianjin Auto Industry Group to manufacture car engines, a prelude possibly to full- scale production of automobiles. Parts production is the second hitherto missing element of agglom- eration. Unlike Volkswagen, which has been instrumental in persuad- ing its German suppliers to begin producing certain parts in the Shang- hai area, Tianjin received little assistance from Daihatsu in attracting parts producers to the city. All 48 component producers are affiliates of Tianjin Auto, and because FDI in the production of auto parts has been negligible, less than 50 percent of the Charade is from locally sourced items. High value added items such as transmissions, electronics, and engines are still being imported. In fact, Tianjin Auto and the local in- dustry bureaus have lagged significantly behind other major auto pro- ducers in China, both in the transfer of technology and in the domestic production of parts. This too is set to change because Tianjin Auto has entered into a dozen joint ventures to manufacture components such as oil filters, moldings, and mirrors. Toyota also appears ready to create a network of suppliers, much as Volkswagen has largely done in Shang- hai. The small size of engines used (0.98 liter) in both the Charade and the minivan, together with the poor quality of parts and materials required for automobile manufacturing, is a third hurdle that must be crossed. Production of 1.3-liter and even larger-capacity engines in collaboration with Daihatsu and Toyota will remove one of the deficiencies. But the problem of parts may take longer to solve. Although Tianjin has invested 108 The Dynamics of Urban Growth in Three Chinese Cities heavily in basic industries over the past four decades, improvements in quality have lagged behind growth in output, and this is immediately apparent from a casual inspection of the vehicles assembled by Tianjin Automobile in the first half of the 1990s. Because China's petrochemical industry is still not technologically on a par with those of more advanced countries, various products such as paints and sealants have poor res- ins. Quality problems also beset all rubber products. Locally produced glass is still deficient in certain respects, and the raw material continues to be largely imported. Cold rolled steel is now available from the modemized steel plant, but there is a dearth of zinc-coated, corrosion- resistant grades of the desired width. The same problem applies to electronic components, from connectors to more sophisticated items. Some components are produced locally but fall far short of international standards. Such shortcomings stem in part from inevitable delays in assimilating technology, in establishing stan- dards, and in creating a market environment, which induce firms to com- pete on grounds of technology and quality. In part also they stem from weaknesses in the production process that arise from the incentive struc- ture embedded in the internal labor market of state-owned enterprises, work discipline and supervision, quality control procedures, and the availability of testing equipment, as well as the skill to put it to effective use. Of all these, work incentives and work organization on the factory floor are arguably the most important. Chronic overstaffing is a long- standing feature of state-owned enterprises in Tianjin, as it is in other cities. Combined with payment practices that do not recognize effort and skill and with lifetime tenure of the core workforce, they under- mine motivation. Employees of state-owned enterprises function in a factory ambience where evidence of slack is widespread and rules of discipline can be fairly relaxed. Financial incentives to work harder or to upgrade skills are relatively modest. And the majority of employees are shielded from layoffs. Those state-owned enterprises that are now operating in buyers' markets have begun responding to competitive pressures by increasing the proportion of contract labor and reorganiz- ing the workplace. But thus far the auto industry still faces excess de- mand, and many dysfunctional practices linger, slowing the ascent of Tianjin's auto industry along the learning curve. Productivity gains have not been fully realized, and vehicles still leave the assembly line with many defects. Summary: Reforming for Growth This brief examination of two industrial subsectors that could spear- head the future growth of Tianjin's economy indicates the breadth of choices at hand and draws attention to elements of industrial backward- ness. Socialist development strategy was responsible for the wide in- Tianjin: A Port, Its Neighborhood, and Its Ambition 109 dustrial base and the experience accumulated by Tianjin. In this respect, it certainly enlarged the scope for agglomeration benefits and potential externalities. It also endowed Tianjin with a number of leading sectors, both actual and potential. Electronics and automobiles are among the two most promising. Together with industries that produce basic materi- als and are engaged in transport-related activities, they could give Tianjin the lead over other cities in north China, with the exception of Beijing. Socialist development strategy isolated Tianjin from the technologi- cal spillover effects emanating from other countries. The autarkic, inward-oriented approach insulated the economy from fresh ideas, new techniques, and competition.2" In this milieu, the system of planning, as well as the vertically integrated organizational structure of state-owned enterprises, dampened industrial dynamism. Furthermore, research with a commercial slant did not flourish, nor was the application of experi- mental findings given priority. Although scientific manpower was fairly plentiful, it was not used efficiently to advance the industrial fortunes of the city. The primitive state of the services sector meant that consult- ing, engineering, extension, and information services contributed little to agglomeration effects. The compartmentalization of the workforce as a result of lifetime ten- ure in state-owned enterprises had macro- as well as micro-level conse- quences. At the macro level, it affected labor mobility and hence the flexibility of resource use. At the micro level, state-owned enterprises, faced with having to employ more workers than they needed, used la- bor inefficiently, made little effort to exert stringent work discipline, and in the interests of trouble-free relations, opted for minimal wage differ- entiation. The transition from plan to market has enhanced Tianjin's autonomy from the central government, improved its growth prospects, and brought many possibilities for innovation within reach. However, new organizational structures are needed that efficiently use physical and human resources, particularly the latter. As Landes (1991, p. 317) ob- served with regard to the industrial revolution in Germany, "reorgani- zation of work entailed reorganization of labor: the relationships of men to one another and to their employers were implicit in the mode of pro- duction: technology and social pattern reinforced each other." Germany succeeded by creating organizations hospitable to emerging technolo- gies. By contrast, the United Kingdom fell behind because organizational disabilities emerged: initially the social power of craft workers prevented factory owners from modernizing their facilities; later labor's ability to exercise its political voice prevented the dismantling of the work prac- tices hindering innovation. There are lessons in this for China: the social organization in state-owned enterprises can be a major stumbling block, and micro-level restructuring and the political power of the urban, in- dustrial workforce influence the content and pace of reforms overall. 110 The Dynamics of Urban Growth in Three Chinese Cities Tianjin has been a late starter at serious micro-level reforms. By 1992 it had secured a fiscal arrangement with the central government that allowed the municipality to retain a larger proportion of revenues and shared most sales taxes with the central government. This system will increase Tianjin's share of total revenues.22 Toward the end of the 1980s, industrial strategy had been sharpened and refocused on a core set of competencies that included the transport sector, certain basic industries, electronics, and autos. In order to build capacity in high-technology ar- eas, Tianjin has created the development zone and is actively seeking FDI. Beyond that, the idea of a "Silicon Corridor" linking Tianjin with Beijing is arousing interest among the politicians and entrepreneurs who will see it to fruition. A multimodal transport network now links Tianjin with the immediate hinterland, but connections with more distant areas are poor. Hence the full benefits of transport investment remain to be realized. One phase of enterprise reform, which encouraged the establishment of nonstate enterprises, has been pursued since the mid-1980s. But the reform of large state-owned enterprises gathered momentum only in the early 1990s. So far, the reform of labor contracting, the merger of enterprises, the setting up of umbrella corporations, and the closure of ailing state enterprises have yet to be molded into an effective strategy. Although some of the smaller loss-making state-owned enterprises are being closed, the tendency is to rely on mergers, which in the Chinese context entails transferring the staff and assets of one enterprise to an- other, usually in the same bureau.23 State-owned enterprises are still absorbing the largest amount of capitaL which otherwise could be used to finance private enterprises, and are responsible for more than 70 per- cent of all industrial losses in the city (see table 3.5). However, Tianjin has attracted foreign investment despite its disad- vantageous position compared with both Guangzhou and Shanghai. Unlike Guangzhou, whose proximity to and historical affinity with Hong Kong have proved to be a magnet for foreign investment, Tianjin was not particularly favored by any major source of investment.24 And un- like Shanghai, Tianjin did not possess a wide range of industrial activi- ties, a tradition of financial and producer services, a large pool of skilled labor, or a strong financial investment from the central government in recent years. In fact, Tianjin has to compete with its close neighbor, Beijing, which, as the nation's capital, can use its political muscle to poach potential investors. To overcome these handicaps, Tianjin has lured FDI with a combination of concessions and incentive policies. It has intro- duced measures to cut capital costs, reduce bureaucratic procedures drastically, and improve the quality of the infrastructure in the Tianjin Economic Development Area; provided postapproval services such as worker training and social insurance; and augmented fiscal incentives.25 Tianjin: A Port, Its Neighborhood, and Its Ambition 111 Tianjin's experience, which is similar to that of other Chinese cities as well as East Asian economies, shows that a good incentives package, sound management, and high-quality infrastructure are key attractions for foreign investment. Incentives are particularly important for high- technology industries, which may be more dependent on them than other industries. City authorities could also provide direct financial assistance to multinationals involved in R&D activities. Other elements of a tech- nology policy include forming a network of research institutions and helping establish linkages between suppliers and customers and between foreign and local firms. Tianjin has yet to define a comprehensive program to reform these state- owned enterprises and their losses. But the city has a real advantage that Shanghai and Guangzhou can only envy: the availability of a large tract of undeveloped land around the city. Tianjin's infrastructure is also in relatively good shape because the 1976 earthquake forced the city to re- build itself. Once the expansion of both the port and the airport is com- pleted, Tianjin will have the makings of a formidable transport hub for north China. If it is able to catalyze multimodal transport development in the hinterland, the city can become the center of a range of transport- induced activities, much like Chicago in the early decades of the century. MM~~~~~~~~~~~ Guangzhou. f The Pearl in - the Delta , Ohh ceG g eiht ov r historic Of t ththree cities, Guangzhou easily has the most venerable historical record. It occupies the place where, legend has it, five genies riding on goats brought the first cereals to the inhabitants of the area. In memory of this event, Guangzhou is often referred to as Yang Cheng, or Goat Town, and a commemorative statue depicting a group of five goats (or rams; see Ikels 1996) was emplaced on a hill overlooking the city earlier in this century. The name Guangzhou dates back to the Wu dynasty, which ruled the area during the time of the Three Kingdoms (roughly around the third century). From ancient o tie toc he present, Guangzhou's history is densely eventful. Long stretches of prosperity were interspersed with reversals, when war or dynastic change wreaked destruction or weakened the local economy. Guangzhou's location in the Pearl River delta region is uniquely ad- vantageous. The city is surrounded by some of the richest agricultural land in China, which has supported the cultivation of rice, sugarcane, vegetables, and the mulberry trees that sustain sericulture. Shanghai's location is no less advantageous, but it achieved prominence only in modem times, when the Yangtze delta region opened outward and be- gan trading with the world economy. Guangzhou, in contrast, has been a regional market center and a port since the Han dynasty in the first century B.C. The coming together of international trade and local com- merce made Guangzhou the largest city in China during the nineteenth century, but the inability to build an industrial base caused it to fall be- hind Shanghai. The relative backwardness of the economies ringing the South China Sea, through the first half of the twentieth century, also slowed Guangzhou's development. As the capital of Guangdong, a stra- tegically vulnerable province near China's sensitive border with Viet- 112 Gutangzhou: The Pearl in the Delta 113 nam, Guangzhou suffered neglect throughout the Maoist era. Choice industrial projects were almost never located in the southeastern coastal provinces.' Nor did the central government provide funds for infrastruc- ture. Under these conditions, growth suffered, and by 1977 Guangdong ranked sixth in net material product per capita.2 Guangzhou's locational advantages became more prominent with the ending of the Vietnam conflict and the rising prosperity of Southeast Asia. Guangzhou's experience since 1978 is interesting, both because similarities with Shanghai and Tianjin deepen our understanding of cer- tain fundamental forces driving urban development and because the contrasts draw attention to the richness as well as the complexity of the forces at work. This chapter analyzes seven factors that have a bearing on Guangzhou's performance: urban infrastructure; the politics of re- form and its influence on resource availability; industrial structure and associated agglomeration economies and diseconomies; neighborhood effects emanating from buoyant economies along the rim of the South China Sea, kinship networks that bind Guangzhou to the economy of Southeast Asia, and the emergence of strong industrial growth poles in the Pearl River delta area; the role of transport; the labor market of the Pearl River delta region; and the rise of producer services in Guangzhou's economy. Urban Infrastructure Guangzhou was the sixth most populous city in China until the late 1980s and, in the early 1990s, the third largest urban economy in the country, surpassing Tianjin and trailing only Shanghai and Beijing. In 1993 Guangzhou also had the highest GDP per capita among all major cities in China, even ahead of Shanghai (table 3.2).3 Guangzhou mani- fests the same problems facing many large cities in China, such as over- crowding, population density ranging from 27,000 to 55,000 persons per square kilometer, housing shortages, traffic congestion, and conflicts in land use (see Yeung, Deng, and Chen 1992). In Guangzhou, industrial land occupies the largest share of serviced urban real estate (more than 30 percent), and the concentration of factories is very high: more than half of all industrial enterprises in the municipality are located in the urban area. Guangzhou has enormous potential for upgrading the area bordering the Pearl River that is now cluttered with factories and ware- houses. Within the central city, differences in the living environment and social status of districts are largely reflected in housing.4 A scarcity of housing has persisted for decades. The average living space per person was only 4.5 square meters in 1945. A trickle of invest- ment raised it to a still meager 4.9 square meters in 1976. Investment in housing construction rose substantially in the late 1980s and in a hand- ful of years pushed per capita living space to 8.5 square meters in 1993 114 The Dynamics of Urban Growth in Three Chinese Cities and to almost 10 square meters in 1995. Annual investment of Y3 billion since 1993 has added a total floor space of more than 3 million square meters (FBIs-CHI-93158, August 18, 1993, p. 44). But the city is facing a new problem in the wake of a growing influx of migrant workers (see Ikels 1996).5 More than any other district, the Huangpu cluster in the central city has the highest concentration of factory workers. The city will have to expand the market for rental housing to accommodate the demand for affordable housing and prevent social disorder associated with the unchecked influx of migrants. Low housing prices will be a major factor in holding down the increase in wages, as was the case in Hong Kong during the 1960s and 1970s. Housing reforms already under way in Guangzhou are designed to deal with the problem of unmet demand and poor maintenance. A ma- jor goal of the reforms is to increase the private ownership of housing. The city has created a public housing fund, based on the Singapore Cen- tral Provident Fund model of matched employer-employee contribu- tions (currently 5 percent of wages from each). It is also attempting to delink housing from work units. But the ultimate goal of treating hous- ing as a commercial commodity to be produced, exchanged, and man- aged in response to market signals requires much institution building.6 Traffic congestion is also a problem. Although Guangzhou is still at the low end of per capita road coverage and total road length, it has the highest ratio of motor vehicles to population in China (Yeung, Deng, and Chen 1992). By the end of 1992, the city's 6 million residents owned some 320,000 motor vehicles, a fivefold increase from 1985. Together with more than 3 million bicycles and 2,000 public buses, Guangzhou daily struggles with incipient traffic strangulation. In this riverine city, water transport could play a vital role, but unfortunately it has been allowed to decline since the 1960s and now only accounts for 10 and 28 percent of passenger and freight traffic, respectively. This decline was mainly a result of limited mechanization of Huangpu harbor, but lack of funds and space for expanding the old harbor and ferry terminals has also contributed. The city has responded to the increasingly severe traffic congestion by embarking on the construction of ring roads. It is tackling the grave shortage of public transport by building an 18.4-kilometer subway to be completed by 1998. The city will invest about Y35 billion by 2000 in road and bridge projects, including the subway. An expressway around the city and four new expressways in the downtown area will be constructed. Moreover, Guangzhou is ahead of other municipalities in its innovative approach to financing: transferring ownership rights to private compa- nies of certain bridges, tunnels, and waterworks helps to ease the finan- cial burden of the municipal government and provides new sources of funding for public utility construction. Guiangzhou: The Pearl in the Delta 115 The Politics of Reform: Why Guangdong Was First As indicated in chapter 1, China's reform strategy as it was tentatively formulated in the late 1970s had two separate dimensions. First, it sought to respond to the dissatisfaction of farmers with planning and controls imposed through the communes and to release the full productive ca- pabilities of the agricultural economy. Second, it sought to respond to the demonstration effect of increasing prosperity among China's imme- diate neighbors along the eastern rim in a manner that was ideologi- cally defensible, that would be politically tolerated, that might generate large economic returns if it were to proceed according to expectations, and that could be contained and defused at acceptable cost were the experiment to fail. Several factors decreed that Guangdong and Fujian would be at the forefront of China's attempt to modernize by carefully managing the opening of its economy. First, the two provinces were situated at a safe distance from the economic heartland and, although populous, played a marginal role in both agriculture and industry. A small dose of "open- ing and reform" could only improve matters and, with sufficient over- sight, would not be explosive. Second, to succeed, any action had to win the backing of a neighboring economy with capital, technology, and in- ternational trading connections to share. This made the choice of a coastal province inevitable. Third, China was keen to enlist the support of econo- mies and groups with which it could bargain on relatively equal terms. Hong Kong and Macao were obvious candidates because they were small, capital-rich, and already linked to China through trade and geo- political circumstances. If Hong Kong and, through the island economy, countries in Southeast Asia were to serve as China's partners, then Guangdong and Fujian were the obvious candidates. Fourth, the two provinces possessed certain attributes that reformers could not over- look. The density of contacts with overseas Chinese was one consider- ation. If China were to attract FDI, only kinship ties and deep-rooted commitments to the homeland would negate the risks of investing. Fur- thermore, the Pearl River delta area was agriculturally rich and well endowed with labor that could be engaged in light industry. And then, there were the vestiges of a commercial tradition that had made the re- gion wealthy in the past and, if revived, could do so again. All these considerations tilted the balance toward the southeastern coastal provinces. However, politics had a large hand in the initial deci- sion to experiment with reforms and in subsequent decisions to ensure that Guangdong benefited from the decentralization that governed the local availability of fiscal resources and from the freedom to interpret and implement central reform directives. Spillover effects from setting up the special economic zones in 1978 and opening up the Pearl River 116 The Dynamics of Urban Growth in Three Chinese Cities delta area benefited Guangzhou economically. In addition, the city gained, directly and indirectly, from the influence its leaders had with the central authorities. In the late 1970s and the first half of the 1980s, Guangzhou enjoyed a close relationship with Beijing because of several important personal connections. It was granted a high degree of administrative autonomy well before other major cities in China. Some observers attribute this to the military and civilian ties of the late Marshal Ye Jianying with Guangdong. For a long time, Marshal Ye was the only major representa- tive from Guangdong on the national scene. Bom in Meixian, a small town in Guangdong, he served briefly as the first mayor of Guangzhou and later as govemor of Guangdong before moving to a senior position in the central govemment. He emerged as the fourth-ranked member of the leadership hierarchy in the early 1970s and subsequently became one of the most powerful elder statesmen in the reform period (Sigel 1993). His son Ye Xuanping later became Guangzhou's mayor, deputy secretary of the Guangdong Provincial Party Committee, and then gov- emor of Guangdong.7 Guangdong also served as the power base for several other leaders who later achieved national prominence, includ- ing Tao Zhu, Zhao Ziyang, Wei Guoqing, Xi Zhongxun, and Yang Shangkun (see Vogel 1989; Cheung 1994b). With such contacts, Guangdong became one of the first provinces designated as a priority area for engaging in foreign trade and promoting foreign investment. Arguably the most important incentive that the central government provided to Guangdong and Fujian was a special fiscal arrangement fixing the amount of revenue that each province had to transfer to the central government. Negotiated in 1979, this fiscal contract gave the two provinces the widest possible budgetary autonomy (Guangdong trans- ferred 31 percent of its revenue to Beijing in 1979 and only 3 percent in 1984; see Sigel 1993; Cheung 1994a; Goodman and Feng 1994). The dif- ference between the province's anticipated revenues and expenditures determined the sum to be transferred. Under the terms of the contract, central ministries ceased to issue mandatory targets and gave Guangdong full control over local spending. Ye Xuanping was able to extend the initial five-year fixed payment, remitting only Y1.2 billion to the central government each year, even though Guangdong's annual revenues topped Y1O billion after 1988. Thus Guangdong, and through it Guangzhou, stole a lead over other provinces whose fiscal contracts were commensurably less favorable. The fiscal contract was combined with measures that allowed Guangdong autonomy in formulating its own development plans, managing labor, and introducing price reforms. Other actions enlarged Guangdong's financial decisionmaking pow- ers, including the authority to establish its own financial institutions, engage in foreign exchange account business, and issue bonds overseas. After the Tiananmen event in 1989, hard-liners in Beijing proposed to Guangzhou: The Pearl in the Delta 117 take back the economic and fiscal autonomy enjoyed by Guangdong and a few other provinces.' But Ye Xuanping prevented such an out- come. Even after his formal departure from Guangdong in 1991, he con- tinued to safeguard the province's interests. His capacity to do so was further enhanced by the important position of his brother-in-law Zou Jiahua, who served as minister in charge of the State Planning Commis- sion and deputy premier.9 The fiscal contracts negotiated in 1988 brought other provinces on a par with Guangdong, but the earlier political inter- ventions put the province on the path to rapid development on which it continues at an unabated pace. Industrial Structure and Agglomeration Economies Guangzhou has already gained a foothold in a number of major subsectors, including clothing, textiles, consumer electronics, household durables, bicycles, motorbikes, and food processing. It has also begun exploiting backward linkages to metallurgical and petrochemical indus- tries. There is ample scope for expansion and upgrading in all of these fields. In particular, garments, household electrical products, leather, and sporting goods industries might provide avenues for growth. Similarly, there are vast opportunities in food and cosmetics industries: the de- mand for custom-designed convenience foods will grow, requiring in- dustrial enzymes, flavors, enhancers, and colors and encouraging the emergence of a biotechnology subsector. Among the three cities, Guangzhou's industrial structure shows the lowest degree of specialization, as demonstrated by a lower level of Herfindahl indices than both Shanghai and Tianjin. The top 10 indus- trial subsectors accounted for 57 percent of Gvio in 1986 and 62 per- cent in 1993 (see table 5.1). In the mid-1980s the four leading subsectors in order of size were machine building, textiles, pharmaceuticals, and metallurgy. Machine building was by far the largest, with a fifth of output, followed by textiles, with 8 percent. Both of these have lost ground, and their shares were 14 and 5 percent, respectively, in 1993. The large gains in shares were registered by transport equipment, chemicals, leather products, and building materials. This represents Guangzhou's attempt to diversify into motor vehicles and motorbikes, chemical feedstocks, fibers for the textiles industry, and a wide array of light manufactures. The development of leather goods and clothing has strengthened export-oriented light industry. At the same time, the emphasis on transport, chemicals, and building materials increased the salience of heavy industry. Overall, heavy industry has gained in share, from a third in 1986 to 43 percent in 1993. Although the shift away from machine building represents a realistic assessment of Guangzhou's competitive strength relative to producers in other parts of China as well as overseas, the investment in autos-as distinct from 118 The Dynamics of Urban Growth in Three Chinese Cities motorbikes-and in chemicals suggests that the municipality remains uncertain about its areas of core competence for the long haul. Com- pared with Shanghai and Tianjin, Guangzhou has diffuse goals for in- dustrial composition, which calls into question its otherwise impres- sive performance. Several subsectors are gaining a stronger presence in Guangzhou's in- dustrial economy, including petroleum processing and chemicals, trans- port equipment, and miscellaneous light manufacturing (see figure 5.1). The petroleum-processing and chemicals industries grew rapidly during recent years. Together they accounted for 11 percent of total GVIO in 1993, up from 8.6 percent in 1986. Chemicals are now the third largest industry in terms of output. The Guangzhou General Petrochemical Plant, with ca- pacity to refine 10 million tons of imported sulfur-bearing crude oil, is being substantially expanded. If completed as planned by the end of Table 5.1 Gross Value of Industrial Output in Guangzhou, Selected Years, 1986-93 1986 2991 1993 Enterprises Billions Billions Billions in 1993 Indicator of yuan' Percent of yuan' Percent of yuana Percent Number Percent City total 17.80 100.0 53.68 100.0 100.25 100.0 5,261 100.0 Light industry 11.86 66.6 32.77 61.1 56.24 56.1 3,564 67.7 Heavy industry 5.94 33.4 20.91 39.0 44.01 43.9 1,697 32.3 State-owned enterprises 11.67 65.6 31.80 59.2 46.53 46.4 1,006 19.1 Collective-owned enterprises 4.53 25.5 8.54 15.9 16.22 16.2 3,081 58.6 Others 1.60 9.0 13.35 24.9 37.50 37.4 1,174 22.3 Top subsectors in 1993 10.11 56.8 32.42 60.4 61.71 61.6 2,657 50.5 Machine buildingb 3.63 20.4 8.20 15.3 14.11 14.1 746 14.2 Transport equipment 0.73 4.1 4.17 7.8 9.59 9.6 247 4.7 Chemicals 0.86 4.8 3.62 6.7 6.03 6.0 225 4.3 Metallurgy' 0.89 5.0 2.69 5.0 5.74 5.7 76 1.4 Textiles 1.42 8.0 3.07 5.7 5.03 5.0 264 5.0 Petroleum processing 0.67 3.8 2.97 5.5 5.00 5.0 3 0.1 Leather products 0.14 0.8 1.83 3.4 4.37 4.4 198 3.8 Building materials 0 40 2.3 1.45 2.7 4.19 4.2 258 4.9 Clothing 0.43 2.4 1.77 3.3 3.94 3.9 569 10.8 Pharmaceuticals 0.94 5.3 2.64 4.9 3.72 3.7 71 1.4 Herfindahl index 0.059 0.056 0.044 Note: All numbers exclude village-run enterprises. a. Current prices. b. Includes machine building and electrical machinery. c. Includes smelting and pressing of ferrous and nonferrous metals. Source: State Statistical Bureau, Statistical Yearbook of Guangzhou, 1987, 1992, 1994. Gutangzhou: The Pearl in the Delta 119 Figure 5.1 Industrial Structure of Guangzhou, 1986 and 1993 Pharmaceuticals Clothing Building materials 1 993 Leather products l3 1 986 Petroprocessing Textiles Metallurgy Chemicals Transport equipment Machine building I . I Other H - r 0 5 10 15 20 25 30 35 40 45 Percentage of GVIO Note: GVIO, gross value of industrial output. Source: Table 5.1. this decade, the plant will be able to produce high-quality lead-free gaso- line, diesel oil, and kerosene for use in aviation. In addition to domestic sales, it will export its products to Singapore and other countries in South- east Asia (FBIs-cHI-94111, June 9, 1994, p. 50). Guangzhou's transport equipment industry got an early boost from a joint venture with Peugeot to produce passenger cars. But Peugeot's decision in 1995 to halt production of its 505 sedan suggests that the market for midsize cars is already too crowded and that smaller players face an uphill battle.10 Motorbikes mightbe better suited to Guangzhou's capabilities, and the market prospects are brighter (penetration rates are 3 to 5 percent) than for midsize sedans, although competition from large producers in Chongqing, Jinan, Shanghai, and Luoyang is strong. China produced 7.8 million motorbikes, mostly in the 100-cubic-centimeter range, in 1995, and capacity distributed across 140 producers is nearly 10 million units. By the end of the century, estimated demand could be as high as 12 million, a huge market for assemblers and component mak- ers alike." The collaborative venture with Honda that began in 1984 gives Guangzhou an opportunity to become a major player. However, only through aggressive marketing, improved technology, and expanded capacity can Guangzhou ensure that it is one of the top five dominant producers five years from now. Guangzhou's approach is similar to that of Japan. Earlier in the cen- tury, Japan relied on its well-articulated and efficient bicycle industry to gain entry into the manufacture of motorcycles. Guangzhou's bicycle industry, led by the Five Rams Company, had already established a net- work of suppliers with the sophistication required to produce parts for 120 The Dynamics of Urban Growth in Three Chinese Cities the motorbike subsector (certain engine electronics and instrumentation were still being imported in 1995 and may still need to be imported). This company has also shown initiative in linking up with defense sup- pliers to arrange for technology transfer in the area of metallurgy and carbon-fiber composites. No doubt they will need a few years to reach acceptable standards of quality, but the experience of Taiwan (China), and even Pakistan, indicates that once the skills are present, movement along the learning curve can be rapid. Differences in the structure of output are also evident in miscellaneous light manufacturing, which is considerably larger in Guangzhou than in Shanghai or Tianjin (and even larger in Guangdong as a whole). Here is the strongest evidence that Guangzhou is following the export-oriented path pioneered by Hong Kong and Taiwan (China). Guangzhou's light industry is labor-intensive, has low skill requirements, and emphasizes design and product diversity so as to facilitate the pen- etration of international markets. Examples of this type of manufactur- ing include toys, shoes, and sporting equipment. Guangdong has avoided the problem of low income elasticity for these goods by dis- placing Hong Kong and, to a lesser extent, Taiwan (China) with the help of suppliers who have consciously relocated production from these economies to southern Guangdong. Guangzhou, and more generally Guangdong, has followed the standard "stepladder" or "flying geese" pattern of export development (Yamazawa 1990, p. 13). Entering at the bottom of the technology spectrum, they have taken over production of the simplest goods with the largest labor component. Guangdong has only just begun the transition to higher-quality, more skill-intensive goods and still has a long way to go (Bateman and Mody 1991). Growth of miscellaneous light manufactures also characterized Shanghai dur- ing the past decade. Rapid increases in garments, shoes, printing, sports goods, and handicrafts have been responsible for a portion of Shanghai's growth. However, in both Shanghai and Tianjin this process is less ad- vanced than in Guangzhou. Food processing may well become a profitable pursuit for Guangzhou. The region around Guangzhou yields an assortment of produce, only a small percentage of which is processed, compared with 70 percent in Brazil and 83 percent in Malaysia. As a result, waste is still substantial. Processing of fruits and vegetables is likely to be an important growth industry, which will spur investment in storage facilities and packag- ing. The market for nonalcoholic drinks throughout China is enormous and still largely untapped. Compared with the annual per capita con- sumption of 16 liters in Hong Kong, the average Chinese consumes just 4 liters of nonalcoholic drinks. Firms such as Vitasoy from Hong Kong have already established joint ventures in Shenzhen, and beverage pro- duction in Guangzhou, with or without foreign participation, could cer- tainly expand and find markets throughout China as well as overseas. Guiangzhout: The Pearl in the Delta 121 But the tasks of enlarging market share and moving upscale into items with a higher value added call for investment in services to raise skills, enhance quality, improve the access of small businesses to financing, and increase the level of design and marketing capacity. For instance, training can augment the productivity of garment workers and their stitching skills. Garment manufacturers with support from municipal authorities could add training facilities, as the Thai Garment Manufac- turers Association has done. Likewise, raising the standards of university courses and establishing design institutes sponsored by the industry or the government, as is the case in Hong Kong and Japan, can address deficiencies in design. Quality is also a function of equipment, and small enterprises often have difficulty obtaining sufficient credit to purchase the latest machinery. Although informal arrangements and urban credit cooperatives satisfy much of the need, the infrastructure of financing is still backward. Guangzhou's state sector is rather different from that of Shanghai in that small-scale state enterprises have played a much more important role in production, providing more than half of the city's GVIO. The city is restructuring municipally owned state enterprises by transforming property rights and attracting foreign investment. In early 1994 it se- lected 50 loss-making state-owned enterprises for a two-year trial re- structuring. Under this program, firms will move out of prime sections of the city to make way for real estate development. Proceeds from land leasing will be used to pay the debts of these enterprises.'2 Other enter- prises owned by the municipality are being offered to foreign investors, who can purchase them outright or enter into joint ventures with the Guangzhou authorities. In this way, it is hoped that problem enterprises will be put back on their feet, their technology improved, and their pro- duction facilities modernized. Guangzhou has also been the most active in cultivating the nonstate sector, especially joint ventures and cooperatives. Ease of entry (and exit) by smaller nonstate enterprises offers the best chances to build robust networks of subcontractors, absorb new technology, and enhance the competitiveness as well as the flexibility of the local economy. As they have done in Korea and Taiwan (China), state enterprises can continue to dominate the upper reaches of industry, where capital intensity and scale economies are significant. Elsewhere, the flexibility of nonstate enterprises can be a tremendous asset. All three cities need to reverse the vertical integration that characterizes large state enterprises, but Guangzhou took an early lead in building a base of potential suppliers. The changing structure of ownership is heartening and indicates that the role of the market is on the rise. In the mid-1980s, state-owned enter- prises generated two-thirds of Guangzhou's industrial output, a large proportion even though it is below the ratios registered by Tianjin and Shanghai. By 1993 this had diminished to 46 percent, but so, interestingly, Table 5.2 Foreign Trade in Guangzhou, 1985 and 1990-93 1985 1990 1991 1992 1993 Millions Millions Millions Millions Millions of U.S. of U.S. of U.S. of U.S. of U.S. Indicator dollars Percent dollars Percent dollars Percent dollars Percent dollars Percent Total exports 413.3 100.0 1,442.9 100.0 1,840.7 100.0 2,456.8 100.0 3,482.2 100.0 Trading companies 314.5 76.1 543.3 37.7 621.1 33.7 607.4 24.7 474.4 13.6 Light industrial products 116.2 n.a. 182.6 n.a. 193.8 n.a. 190.9 n.a. 137.6 n.a. Crafts 37.0 n.a. 53.8 n.a. 48.0 n.a. 46.0 n.a. 35.1 n.a. Textiles 54.1 n.a. 75.8 n.a. 88.7 n.a. 75.7 n.a. 61.9 n.a. Chemicals 7.7 n.a. 15.7 n.a. 13.9 n.a. 13.0 n.a 13.3 n.a. Machinery 1.6 n.a. 9.0 n.a. 11.6 - 15.2 n.a. 14.8 n.a. Equipment 2.0 n.a. 18.3 n.a. 47.0 - 50.0 n.a. 36.8 n.a. Industrial enterprises 36.9 8.9 440.2 30.5 523.8 28.5 634.4 25.8 909.7 26.1 9 Foreign-invested enterprises 11.6 2.8 349.2 24.2 559.7 30.4 1,073.3 43.7 1,416.3 40.7 Processing/assembly plants 50.3 12.2 110.2 7.6 136.2 7.4 141.7 5.8 542.6 15.6 Others 0 n.a. 0 n.a. 0 n.a. 0 n.a. 139.2 4.0 Total imports - - 671.6 100.0 1,178.9 100.0 1,746.9 100.0 2,485.1 100.0 Trading companies - - 114.1 17.0 153.9 13.1 187.8 10.8 238.5 9.6 Industrial enterprises - - 226.1 33.7 265.5 22.5 361.0 20.7 413.7 16.7 Foreign-invested enterprises - - 328.8 49.0 756.1 64.1 1,191.8 68.2 1,473.3 59.3 Processing/assembly plants - - 2.7 0.4 3.5 0.3 6.2 0.4. 359.6 14.5 Commodity trade balance - n.a. 771.3 na. 661.8 n.a. 709.9 n.a. 997.1 n.a. -Not available. n.a. not applicable. Source: State Statistical Bureau, Statistical Yearbook of Guangzlhou, 1992, 1994. Guangzhout: The Pearl in the Delta 123 had the share of collective-owned enterprises (table 5.1). The big gainers were joint ventures and private enterprises, whose share increased four- fold, from 9 to 37 percent, underscoring the sheer volume of FDI in Guangzhou and its dramatic impact on the industrial system. Prospects for continuing reform of state-owned enterprises may depend on the maturing of the still embryonic social security system. So long as jobs are abundant and few state employees are being laid off, a state-supported social security system is not essential. However, an accelerated closure of state-owned enterprises and the concomitant loss of welfare services could engender labor unrest and in turn rising labor costs. In this regard, Shang- hai has acted faster than Guangzhou in devising a framework for provid- ing social welfare outside of state-owned enterprises."3 Guangzhou's growth performance puts it ahead of the two other cit- ies in some respects. Its efforts to restructure, which have been pushed ahead more forcefully by reforms, have substantially enlarged the share of nonstate enterprises, as well as service industries. Guangzhou's hos- pitality toward foreign investors has provided funding for industry and promoted export activities, enabling the city to achieve a high ratio of trade to municipal output. Where Guangzhou is weakest is with regard to industrial focus. In the light manufacturing sector it has areas of specialization, such as leather goods, but overall the impression is less positive. The city is be- ing pulled in several different directions, spreading resources thinly across subsectors, such as autos and chemicals, where its comparative advantage is uncertain."4 Some industries should either be pursued com- prehensively, with a long-term vision and the willingness to commit a large volume of capital, or be left alone. The robustness of Guangzhou's economy and its competitive strength are also apparent from export trends. Exports were valued at $413 mil- lion in 1985. By 1993 they were almost $3.5 billion, a more than eightfold increase yielding a net surplus of approximately $1 billion on the trade account (see table 5.2). More than half of the exports were light manu- factures produced by joint ventures and enterprises processing or as- sembling items on contract. Such activities have multiplied all across Guangdong and Fujian. Guangzhou, because of its size, industrial base, port facilities, and services, has emerged as a manufacturing base that foreign investors find especially attractive. Neighborhood Effects The overwhelming importance of neighborhood effects is arguably the most remarkable feature of Guangzhou's development. These derive from three separate sources: FDI primarily from Southeast Asia, proxim- ity to Hong Kong, and growth of the Pearl River delta economy. The roots of FDI in Guangzhou lie in the emigration of Chinese to Hong Kong 124 The Dynamics of Urban Growth in Three Chinese Cities and to other countries in the region. A trickle before the mid-nineteenth century, emigration mounted rapidly after the Qing authorities relaxed the ban in 1890. A hundred years later, in 1990, roughly 37 million Chi- nese were living overseas in 136 countries and economies, with two- thirds residing in Indonesia (20 percent), Thailand (16 percent), Hong Kong (15 percent), and Malaysia (15 percent). Including Singapore and the United States raises the ratio to more than 75 percent (Poston, Mao, and Yu 1996). A majority of the emigrants are from southeastern China, and many of them have retained their ties with the cities and villages of their birth. Even during the Maoist period, when contacts were difficult to sustain and travel to China was problematic, overseas Chinese kept alive their relations with kinfolk whenever possible by sending money and packages. This large, interconnected, entrepreneurial, and wealthy Chinese diaspora is the conduit for much of the FDI in China and the force behind neighborhood effects (Financial Times, August 16, 1995). The turning point for Guangdong and for its capital city was Central Document No. 50, issued in 1979, which not only formalized the gener- ous new fiscal contract but also enabled the province to add operational content to its latent ties with Hong Kong and other economies of South- east Asia. Under the terms of the agreement with the central government, Guangdong retained 30 percent of foreign exchange in excess of a fixed amount of $20 million from export processing, compensation, trade, and joint ventures. This was 5 percent over the level allowed to other open regions. The province was granted greater authority to conduct foreign trade and to create investment entities so as to bring in foreign capital. Two later documents, No. 41 (mid-1980) and No. 27 (1981), encouraged Guangdong to pursue enterprise, labor, and price reforms so as to maxi- mize benefits from the "opening to the outside world." Financial mea- sures permitting, the ability to issue stocks and bonds so as to absorb foreign capital strengthened Guangdong's hand yet more (Cheung 1994b). With the blessing of the central government and the latitude to ma- nipulate policies so as to induce FDI, Guangdong moved quickly to breathe life into cross-border relationships. Businessmen in Hong Kong, seeking opportunity and some relief from rising land and labor costs, were quick to reciprocate. For all the initial uncertainty, the attractions of shifting some production facilities into Guangdong were exceedingly powerful. The absence of language barriers, the ease of dealing with kinfolk, and the convenience of being a short truck drive from Hong Kong outweighed the headaches caused by red tape, the avarice of local authorities, and the problems of operating in a country that had adopted a posture of studied aloofness for 30 years. In an amazingly short span of time, economic considerations tri- umphed convincingly over political reservations, over ideological resis- tance to the seeping in of capitalism, and over the many day-to-day an- noyances of doing business in the China of the early 1980s. Starting with Guiangzhou: The Pearl in the Delta 125 the four Special Economic Zones, FDI diffused into Guangdong. Soon it spread throughout the delta and especially to Guangzhou, which cre- ated an Economic and Technological Development District in 1984, ini- tially covering an area of 7 square kilometers that by 1996 had expanded to 50 square kilometers. Between 1979 and 1993 Guangzhou actually received $3.0 billion in FDI, three-quarters of it from Hong Kong (see table 5.3; about 64 percent of all foreign investment in Guangdong Prov- ince was in Guangzhou and Shenzhen). By the end of 1993, there were nearly 5,000 joint ventures in Guangzhou, with a gross output of Y30 billion and exports of $1.4 billion (FBIs-cHi-94104, May 31, 1994, p. 49). The effect of this great wave of capital was far-reaching for Guangzhou and the entire Pearl River delta."5 Guangzhou and the surrounding re- gion acquired a modern industrial base of light industry composed mainly of joint ventures, collectives, and private enterprises.'6 By the close of the 1980s, 7,000 joint ventures were operating in the Pearl River delta, while another 20,000 local factories were working with Hong Kong businessmen in various kinds of subcontracting relationships. Nearly 6 million workers, mostly in Guangdong, were employed in these busi- nesses by 1994, compared with about 615,000 manufacturing workers in Hong Kong itself. Moreover, the number of manufacturing workers in Hong Kong has been declining in absolute and relative terms, from 42 percent of total employment in 1980 to 20 percent in 1991 (Wong 1991; Sung 1991). Thus, the export of jobs from Hong Kong has made possible the rapid growth of manufacturing employment in Guangdong. Hong Kong itself has already made the transition from being a significant manufacturing center to serving as the financial and commercial focus of a manufacturing network that spans the Pearl River delta. In 1978, 75 percent of Hong Kong's exports were manufactured in the colony; by 1989 the figure was only 39 percent. The share of employment in com- mercial, financial, and real estate services increased from 25 percent in 1980 to 33 percent in 1990. Despite rising land and labor costs, foreign investors continue to enter Guangdong in droves. Provincial planners expect high costs in the open coastal areas to push processing/assembly and other labor-intensive industrial activities gradually out of the Pearl River delta and into more remote areas. In 1992 all major cities in Guangdong were declared "open" to promote foreign investment. But weak transporta- tion and distribution links to the mountainous areas have inhibited most firms from moving out of the delta. Foreign-invested enterprises in Guangdong still rely heavily on Hong Kong affiliates for marketing, re- pair, and other support services, which are rarely available domestically or at least are not available cheaply (see Brudvig 1993). The spillovers from Hong Kong and later from Taiwan (China) as well as other economies have covered a wide spectrum. Most immediately, FDI was the source of capital manufacturing technology and market outlets. Table 5.3 Foreign Investment in Guangzhou, Selected Years, 1979-93 (millions of U.S. dollars) Indicator 1979 1980 1985 1990 1992 1993 1979-93 Contracted foreign investment 23.58 249.05 701.75 554.26 4,710.80 7,047.64 16,320.92 Loans 0 0 159.82 50.67 155.71 175.66 938.80 Direct investment 19.70 247.94 515.75 471.83 4,496.54 6,836.34 14,963.90 Equity joint ventures 0 0 70.73 107.06 526.67 444.83 1,918.73 Contractual joint ventures 19.70 247.94 445.02 242.27 3,661.13 6,034.09 12,017.34 Wholly foreign-owned 0 0 0 122.50 308.74 357.42 1,027.83 Other 3.88 1.11 26.18 31.76 58.55 35.64 418.22 Leasing 0 0 0 14.55 46.01 25.09 132.16 Compensation trade 3.88 1.11 . 6.09 4.79 2.40 2.22 107.87 Assembly/processing 0 0 20.09 12.42 10.14 8.33 178.19 Utilized foreign investment 9.84 30.13 157.82 267.37 728.80 1,463.92 4,151.06 s Loans 0 0 39.37 66.43 155.71 175.66 907.88 Direct investment 1.65 12.87 103.89 180.87 554.20 1,278.28 3,019.50 Equity joint ventures 0 0 29.64 85.05 244.84 332.80 991.20 Contractual joint ventures 1.65 12.87 74.25 60.31 210.67 796.92 1,681.34 Wholly foreign-owned 0 0 0 35.51 98.69 148.56 346.96 Other 8.19 17.26 14.56 20.07 18.89 9.98 223.68 Leasing 0 0 0 7.65 15.42 3.61 39.36 Compensation trade 0.56 3.63 7.15 3.77 0.52 0.50 75.25 Assembly/processing 7.63 13.63 7.41 8.65 2.95 5.87 109.07 Source: State Statistical Bureau, Statistical Yearbook of Guangzhou, 1994. Guangzhou: The Pearl in the Delta 127 In less than a decade, Guangdong had acquired a large and dynamic in- dustrial sector, which was responsible for 9.5 percent of China's manufac- tured exports in 1993.17 A province not known for its industrial perfor- mance and whose growth rate barely reached the average was transformed into an industrial powerhouse and became one of the three fastest-grow- ing provinces in the country. The longer-run effects of Hong Kong's participation and the FDI en- tering the province are equally significant. First, industrialization and contact with foreign engineers, technicians, factory supervisors, and oth- ers have sharply raised labor skills. Second, many economic institu- tions-legal, regulatory, for propagating information, and for selling stan- dards-have been absorbed into the life of the province, squeezing out earlier socialist practices. This has made Guangdong a leader among China's modernizing provinces. Third, neighborhood effects have given rise to an irresistible reform dynamic. Greater integration with the inter- national economy by way of Hong Kong has created pressure for more reform. Each step forward gives rise to demands for more change, which wears down the residual political resistance (from supervisory bureaus, state enterprises, and more conservative elements in the party) to the extension of market institutions. Fourth, Hong Kong is the source not only for the bulk of FDI, but also for a full suite of producer services critical to the effective functioning of joint ventures and the export of products assembled in Guangdong (see Enright, Scott, and Dodwell 1997). Although financial services are at the top of the list, industry and trade would be helpless without marketing insurance, brokerage, and trans- port services that see goods from the factory gate to their overseas cus- tomers, most commonly via the port in Hong Kong (increasingly, seabome freight is being channeled through other ports, such as Shekou, Yantian, and Zhuhai; see Chu 1994). China's long isolation from the world of international trade in manufactures and its relative industrial back- wardness made it necessary for Hong Kong businesses to provide train- ing in quality control, to supervise production closely, to train key fac- tory personnel, to arrange for professional inspection of items before they are shipped, and to send skilled technicians to repair equipment (three Hong Kong firms do the bulk of inspections in the Pearl River delta area). As the scale of cross-border activities mounted in the 1980s, the supply of such services rapidly grew in volume. Toward the end of the decade, more than 50,000 people from Hong Kong were crossing into Guangdong each day, mostly to service industrial activities. Hong Kong has been the main gateway for capital, skills, informa- tion, and services into Guangzhou (Sung 1991). Other economies, such as Taiwan (China), Thailand, Singapore, and Indonesia, have also in- vested in the province. Although their shares are smaller, these rapidly growing economies have generated other kinds of spillover benefits. Together they constitute a large and highly competitive market for Guangzhou's industry. Collectively they had a gross national product 128 The Dynamics of Urban Growth in Three Chinese Cities (GNP) of $675.0 billion in 1992, and their total trade, exports plus im- ports, amounted to $685.2 billion, which has increased 10.5 percent an- nually since 1985.18 Because Chinese businesses dominate the industrial scene even in countries where the majority of the populace is not Chi- nese, there exist ready-made channels through which enterprises from Guangdong can acquire market contacts, gather market intelligence, and strike up alliances. Redding (1992, p. 21) has mapped the spread of these alliances and the diffusion of Chinese capital throughout Southeast Asia. He notes that: Chinese have been holding various international meetings in recent years. These gatherings are significant in that they tran- scend the traditional family, geographical, and business ties of overseas Chinese.... They are aimed at building a global net- work of expatriate Chinese and Chinese enterprises. Further- more, the recent modernization and diversification of Chinese businesses are combined. It seems that the international confer- ence is the place where Chinese enterprises create the new net- work for becoming an international company by making the best use of the traditional networks which have been created in the past. It is easy to think, therefore, that recent international conferences are harbingers of moves to create a strategically more broadly based worldwide network that would replace the tra- ditional network which has lost some of its vitality and resil- ience because of generational change. If that is the case, the pres- ence of Chinese networks in the world economy, not just in the regional economy, will become a factor of decisive importance. The size of the Southeast Asian market means that there are scale economies to be exploited in certain export industries. Economies that have access to an abundant supply of low-wage labor and are able to invest in large-scale modern facilities can gain the upper hand. Guangdong Province can marshal the labor, and it commands a sub- stantial investable surplus. This permits a rapid supply response and the achievement of scale economies. As service industries clustered around Guangzhou mature, Guangdong's overall productivity and com- petitiveness will increase even more, adding to its power in the region. Thus the size of the potential market has feedback effects that the economy around Guangzhou can use to gain in strength. The region lying beyond China's southern periphery is exerting a pro- found influence on Guangzhou, shaping its development, and regulat- ing the pace of growth. But the Pearl River delta economy is of consider- able consequence as well, and its evolution is also leaving an imprint on Guangzhou. The so-called open region of the Pearl River delta has an area of 45,005 square kilometers and contained 17.2 million persons in Guangzhout: The Pearl in the Delta 129 1993. The delta as a whole was responsible for close to 59 percent of Guangdong's gross product, with industry accounting for much of its preponderance (see table 5.4). Factories in the Pearl River delta produced almost three-quarters of Guangdong's industrial output and an equiva- lent percentage of provincial exports. Within the delta, Guangzhou is the largest city in terms of population and industrial output, but manu- facturing centers have grown around it in the counties of Dongguan, Foshan, Zengcheng, and Zhongshan, not to mention the thriving spe- cial economic zones of Shenzhen and Zhuhai (see figure 5.2). Each pro- duces a range of manufactures, though in most instances there is some specialization. For instance, Foshan is especially noted for the assembly of consumer electronics, textiles, and toys, and Dongguan has a reputa- tion in ceramics and clothing. In a little more than a decade, 1:DI and the opening of the economy to trade have transformed largely rural coun- ties into thriving industrial districts with clusters of networked firms whose competitive strength equals if not surpasses that of other pro- ducers in Southeast Asia. The drive for wealth in the Pearl River delta is most striking in Dongguan, a metropolitan area of 2 million persons that includes 29 surrounding townships. The local economy has grown about 20 percent annually since 1979. To lure bigger enterprises, Dongguan has put up factory buildings, which it rents to manufacturers from Hong Kong and Taiwan (China). It now has more than 7,000 processing/ assembly manu- facturing and "compensation trade" enterprises in which foreigners have invested (migrants from Dongguan account for a sizable share of Hong Kong's population-650,000-and Dongguan was the first to attract in- vestment from Hong Kong; see Fitzgerald 1996).'9 In 1993 production of household appliances, textiles, foodstuffs, pharmaceuticals, building materials, and machinery pushed GVIO to Y21.5 billion. Dongguan is improving its infrastructure as well as human resources. Since 1979 it has invested Y6 billion in infrastructure, including the rapid conver- sion of the port into a container terminal. It is building a Y1 billion, 186-megawatt power plant. Dongguan will benefit tremendously from the new rail connection between Guangzhou and Kowloon because the double-track line runs through eight towns in the county. It is also forg- ing direct relationships with universities and other educational institu- tions in other parts of China. Since the early 1980s Foshan has emerged as an important industrial area focusing on consumer durables, textiles, aluminum products, and electronics (Foshan's position as a significant commercial town and ad- ministrative unit dates back to Ming times; see Faure 1990). Some of China's best-known brands are produced there, and GVIO in 1993 was Y76.7 billion. It also manufactures electronic equipment, textiles, and construction materials, in part in the Foshan high-technology industrial development zone. Transportation links have fueled the growth of Table 5.4 Macro Indicators for Guangdong Province and Pearl River Delta, 1993 Pearl River delta region Guangdong Open econotmtic Guangzhou city Shenzhen Special Zhuhai Special As a percentage of Indicator Province region, proper Economic Zone Economic Zone Total provincial total Population (millions) 65.82 17.20 3.73 0.52 0.31 21.75 33.0 Area (square kilometers) 180,000 45,005 1,444 328 121 46,897 26.0 Gross value of agricultural and industrial output (billions of yuan) 613.64 230.44 68.26 43.69 17.12 359.51 58.6 w Gross value of industrial output (billions of yuan) 523.74 211.56 66.59 43.62 16.81 338.57 64.6 Exports (billions of U.S. dollars) 27.03 8.23 0.81 5.77 0.95 15.76 58.3 Utilized foreign investment (billions of U.S. dollars) 9.65 3.92 0.28 1.11 0.52 5.84 60.5 a. Composed of 28 counties and municipalities, excluding the city proper of Guangzhou and the Special Economic Zones of Shenzhen and Zhuhai but including the counties administered by Guangzhou, Shenzhen, and Zhuhai. Source: State Statistical Bureau, Statistical Yearbook of Guangdong, 1994; Statistical Yearbook of Guangzhou, 1994. Figure 5.2 Guangzhou and Its Environs IBRD 28659 114~ MONGOLIA ( o - JN / O' " 4FAANGSWE ----------- COUNTY BOUNDARIES 25' SHUAN EI - -'A -t CITY BOUNDARY S I ICHUJAN --, ZHEXAA3G RJIANGXI ,,E C t EO PROVINCIAL BOUNDARY H'Ž EUNAN1 GU.,,. , ,U, Are F .-U . IAN- INTERNATIONAL BOUNDARIES ? (l ,---~~~ .Area ef-mn-p ' TA>Yww T.> , GUANGXI ,G PN 1 o* 6~~~Huizkou aoin .. i,, _Ce . Heshan - / 1- .liangmenq( J ~ - Boai Xinlui, Kaiping~,,- oTaish,e*one~o "<¾'- -22,~~~~~~~~< 2 >22<'~~~~~~~~~~~~~~~~~~~~~~~~2- 0 20 40 60 so KILOMETERS 7 1 i3 l 114 AUGUST 1997 131 132 The Dynamics of Urban Growth in Three Chinese Cities Foshan, and new highways to Zhuhai and Guangzhou are strengthen- ing its links with the rest of the delta area. A daily direct train offers service from Hong Kong to the new railway station in Foshan. The adja- cent container port at Lanshi, just 9 kilometers from Foshan railway sta- tion, has an annual capacity of 2 million tons. The economy of Guangzhou complements and competes against the manufacturing establishments sprouting in nearby counties.20 Guangzhou's lead in the sphere of producer services allows it to sup- port the industrial push in the nearby cities. Because it is the main port of the region, it has the best sea, air, and rail links to other parts of China as well as the outside world. Whereas cities such as Wuhan have found that their interests as open and autonomous cities conflict with those of the provincial authorities (Hubei), Guangzhou remains very much the administrative center of Guangdong, as it was in the past. It is not just the largest industrial city in the region, but one with several roles that have taken on a greater importance as the Pearl River delta has become industrialized. The rise of Dongguan, for instance, has had spillovers for Guangzhou: demand has grown for transport and other services available in the city, local manufacturers have increased their clientele in the neighboring countries, and provincial administration has become much more significant. The element of industrial competition and some collaboration is also noteworthy. Enterprises in Guangzhou, especially state-owned enter- prises, are under far more competitive pressure from producers in the area. State-owned enterprises are being driven to restructure their op- erations and to use their reserves of skilled and trained staff to compete, if not in terms of price, then on the basis of quality and innovation. Other enterprises must offset the higher land and labor costs of operating from Guangzhou by improving their productivity. In any event, competition from the hinterland has beneficial macro and micro consequences. At the macro level it helps drive the tempo of reform in Guangzhou. Al- ready among the front-runners with respect to reforms, Guangzhou has been compelled by intraprovincial circumstances to strive even harder to create an efficient economy keyed largely to market signals and to develop financial instruments, technical skills, and producer services so as to enhance the competitive position of local firms. Transport-Induced Development Continuing economic prosperity will require the province to expand transport capacity, including that of the airport, seaport, connecting road, and water system, because transport-dependent light industry and trade are the lifeblood of the delta area's economy.21 Guangzhou harbor, the third largest seaport in the country, is undergoing expansion in a bid to attract a greater volume of international marine traffic. With more than Guangzhou: The Pearl in the Delta 133 70 berths, 34 of which are able to accommodate 10,000-deadweight-ton ships, it handled 65 million tons of goods in 1993. Starting in 1996, the port authority began investing Y750 million over a five-year period to dredge the mouth of the harbor and add 9 million tons of cargo-handling capacity annually. Moreover, an entirely new port zone is to be built before 2005 to raise the port's annual handling volume another 40 percent (FIS-CHL-94171, September 2,1994, p. 36). China's first container route, between Guangzhou and Shanghai, was inaugurated in November 1993. A ship capable of car- rying 235 containers is plying between the two cities five or six times a month. This has relieved pressure on the overloaded rail system. More- over, the new port at Yantian near Shenzhen and Hong Kong is emerging as a major transshipment terminal for coal and containers. Guangzhou airport is the third largest in the country and one of the busiest.22 The city has proposed building a new facility instead of ex- panding the existing Baiyun airport, which is located 40 kilometers away from the city center. The local authority is allowed greater flexibility to choose the location and size of the airport and to share management responsibility and profits. Guangzhou will also permit foreign funds to be invested in the new international airport in Huadu. Foreign sources may provide as much as 60 to 70 percent of the $1.15 billion that the airport is estimated to require. A company will be created to construct the airport, which will be listed on either the Shanghai or the Shenzhen Stock Exchange. In the Pearl River delta, infrastructure facilities, such as highways, deepwater ports, and airports, are constructed at a much higher density than in other regions in China (Nanyang Commercial Bank, Ltd. 1992). Within 100 kilometers of the Pearl River estuary, there is a high concen- tration of ports, including Guangzhou port, the four ports of Shenzhen, the two ports of Zhuhai, and six others. Three more ports are under planning or construction. In addition to passenger transport, these ports are mainly used to ship goods to Hong Kong for reexport. There are two international airports in the delta, Guangzhou and Shenzhen, and four domestic ones in Foshan, Huizhou, Jiangmen, and Zhuhai. With comple- tion of the Guangzhou-Shenzhen-Zhuhai and Huizhou-Shenzhen free- ways, most of these airports will become part of an effective multimodal network.23 However, the surface transport network is not as well developed, in- hibiting door-to-door delivery, which is advantageous for light industries. Although the density of road networks and inland waterways in Guangdong is higher than the national average, there is still a severe short- age of transport capacity, which is compounded by outmoded technol- ogy and high costs. Roads are heavily congested in the fast-growing areas because road capacity has not expanded in line with the growth of traffic in recent years. Growth rates of total road traffic in the province have averaged an estimated 20 to 30 percent a year since 1980. hi particular, 134 The Dynamics of Urban Growth in Three Chinese Cities there is a lack of class 1 and 2 roads, as well as expressways. The mixed use of roads by motor vehicles, tractors, bicycles, and pedestrians, with- out separation, exacerbates the situation. Congestion and road conditions greatly reduce speeds and consequently increase the costs of operating a vehicle. Moreover, many trucks are of 1950s vintage, and the introduction of new equipment is poorly planned. For instance, the use of modem heavy trucks is causing rapid deterioration of pavements because few roads were constructed to handle such loads. To deal with these problems, the Guangdong authorities have launched two initiatives. One is an intensive program to upgrade the road system to class 1 or class 2 standards. Another is the heavy emphasis on invest- ment in highway construction using both domestic and foreign sources (Luk 1994). The Shenzhen-Guangzhou highway, by Hong Kong's Gor- don Wu, is the most notable instance of foreign investment in infrastruc- ture. Highways will play an increasingly large role in the future because the fast-growing industries in the province will need the speed, flexibil- ity, and door-to-door delivery that only road transport can provide. Although roads are receiving attention, construction and improve- ment of inland waterways continue to suffer neglect. The condition of water channels, which ranges from poor to good, limits long-haul ship- ping. Despite having several thousand kilometers of navigable inland rivers, the Pearl River delta area has not increased its use of water trans- port. In addition, many river ports are not equipped with modem han- dling facilities, and loading and unloading are often done manually. Because of outmoded technology and limited port capacity, state-owned ships achieve a utilization rate of only about 20 percent (see Harral, Cook, and Holland 1992). Recently, larger barges, barge trains, and container barges have increased their services, and efficient handling equipment has been installed in several locations. Outside the delta, inland river transport tends to specialize in bulk goods such as construction materi- als, cement, and fertilizer. Channel dredging could allow larger vessels upstream and extend the navigation season. But thus far, inland water- way transport is not a priority for investment in the eyes of provincial authorities, and opportunities to alleviate a critical bottleneck are being allowed to slip.24 Labor Mobility Guangzhou, as a pioneer of reforms among the three cities, is also an appropriate place for observing the state of labor markets. From early on, the city recognized the need to enhance labor mobility and to ex- pand the authority of enterprises over labor market decisions. A series of measures have been introduced, particularly after 1986, to revamp the "iron rice bowl" system and to improve labor productivity, includ- ing labor contracting and wage reform. Two of the most successful changes in matching people with jobs have been the circulation of infor- Gutangzhou: The Pearl in the Delta 135 mation about job openings and the policy that applicants must take exams (Vogel 1989). A municipal labor service company was set up to facilitate recruitment and serve as a training center (see chapter 6). To accommodate migrant workers, a system of temporary registration was also established to regulate entry into the municipality and to help en- force a system of annually renewed employment contracts. Although local labor is plentiful, it is unlikely that the expansion of industry and construction could have proceeded at the rate actually at- tained without voluntary and induced migration from poorer parts of the province as well as farther afield.25 Rural people from Hunan, Guangxi, Sichuan, and deeper in the interior have flocked to Guangdong, whose prosperity is a byword in China. Many of them find their way to Guangzhou to work in factories, in construction sites, and in the myriad formal and informal service industries. Factory managers, taking a prac- tical approach, frequently send supervisors to recruit young workers from distant villages once a few migrants from these places have estab- lished a good reputation. Given the nature of factory work, fresh blood sustains productivity and fills the gaps left by locals who graduate to less demanding jobs or begin to press for higher wages. The existence of unskilled and semiskilled migrant labor has sustained elastic labor supply conditions that have allowed Guangzhou to retain its industrial competitiveness. But the city and its sister counties have also relied on the experience of technical staff, who are drawn by attrac- tive salaries from Shanghai, Tianjin, and Liaoning Provinces. These people have made good Guangzhou's deficiencies in factory operation, plant engineering, repair work, and a host of engineering tasks associ- ated with rapid industrialization. According to a 1987 population sur- vey, technical personnel accounted for close to one-third of male mi- grant workers (see Li and Siu 1994). In fact, the flow in skills was on such a large scale that Shanghai authorities attempted to curtail it start- ing in the early 1990s. Migrant labor is also more highly educated than Guangdong's population as a whole: 62 percent have a junior high school education or above compared with only 36 percent of the general popu- lation in Guangdong (Fan 1996). The magnitude of intraprovincial migration is remarkable, in fact larger than interprovincial movement, and was estimated at around 2.5 mil- lion in 1990. Close to two-thirds of this movement is from rural areas, particularly in the mountainous periphery, to cities and towns. As may be expected, the majority of migrant workers end up in the Pearl River delta region. Major destinations include Shenzhen, Guangzhou, Dongguan, Foshan, and Zhuhai (in that order). The eastern part of the delta-Shenzhen and Dongguan-attracted the most migrants during the 1980s. Both cities capitalized on their proximity to Hong Kong and concentrated on developing processing/assembly industries and opera- tions, which are highly labor-intensive by nature. Estimates of how many migrants, from both within and outside the province, are present in the 136 The Dynamics of Urban Growth in Three Chinese Cities Pearl River delta area tend to be imprecise, but by 1990 they were close to 2 million, with Guangzhou accounting for half that number (see Sawada 1992). Since then, the number has risen, possibly to 2.5 million or 3 million, and is likely to continue inching upward as the attractions of a large prosperous metropolis exert their inexorable pull. The Rise of Producer Services Another noteworthy feature of structural change in Guangzhou relates to services. Possibly because the central government allowed Guangzhou greater leeway than either Shanghai or Tianjin, the percentage of ser- vices in municipal output was greater than in the other two cities. In fact, Guangzhou is the only large city in China where the tertiary sector dominates. The increasing extent of market-directed development has brought about a further growth in share, with the result that Guangzhou's economy is structurally balanced and in a somewhat better position to realize productivity gains. Retailing, fast food, and tertiary business are set to continue their strong growth. In 1992 Guangzhou ranked third in retail sales after Beijing and Shanghai; it also recorded the largest through- put of foreign visitors-2 million-of any city (Business China, Novem- ber 1, 1993, pp. 8-9). From the second half of the 1980s, there was a steady increase in the import of consulting services from Hong Kong as Chinese firms started taking more initiative in the areas of design, product innovation, research, and streamlining of the production process. Real estate development and the building of better urban infrastructure also generated its own needs for architectural, engineering, and consulting services. Within less than a decade, enterprise managers, party bosses, city mayors, urban planners, and officials in the myriad bureaus realized the necessity of developing a producer-services sector to sustain manufacturing indus- try. Initially, there was the ingrained ideological resistance to services, which were viewed as unproductive. Where possible, the tendency was to spend money on hardware but to skimp on services. Experience even- tually drove home the irrationality of this attitude, and the use of im- ported producer services increased. In due course, domestic suppliers learned the tricks of the trade, and import substitution began in earnest, with Guangzhou emerging as the focus for these activities. Guangzhou's edge over other candidates in the region derives from its being the pro- vincial administrative center, a major port, and the second largest re- cipient of FDI in the delta region (where Shenzhen is the largest recipi- ent). But municipal authorities also were quick to see the possibilities and to call for the acquisition of skills that would allow Chinese profes- sionals to enter service industries. Although Hong Kong is likely to re- tain its dominance during the balance of the 1990s, the intensive devel- opment of this sector in Guangzhou is certain to erode Hong Kong's role as a provider of services to firms in Guangdong. Guangzhout: The Pearl in the Delta 137 One example of a service activity largely neglected in China's so- cialist economy is packaging. Chinese enterprises, unassailable in seller's markets, were oblivious to the need for attractive packaging and unwilling to sort and pack items to suit consumer demands. How- ever, in the competitive international market, slipshod packaging was a serious drawback, and Hong Kong expertise proved essential. Sung (1991, p. 133) observed that "packaging is not a particular[ly] capital- or skill-intensive activity. . . and repackaging Chinese goods in Hong Kong [went against the grain of entrenched practices]. A good example of reexports arising out of sorting [was] the labor-intensive sorting of Christmas decorations into variety packages." This is now being shifted to China, with Hong Kong firms providing the expertise and the Chi- nese investing in "imported packaging machinery to remedy past neg- ligence." Universities in Guangzhou were among the first in the country to seek commercial outlets for their academic research. This may be related to the deep-seated business culture and the early introduction of market mechanisms into the local economy. For instance, the South China Engi- neering University developed a robot production line for assembling ceiling fan motors, the first of its kind ever designed and manufactured completely by China itself. The university was also one of the first units to start research into robots. Summary Whether it is growth or reforms, Guangzhou is the front-runner of the three cities. It continues to benefit from spillover effects from the delta area and abroad. The network of overseas Chinese has extended its po- tential economic reach deep into Southeast Asia, giving the city access to capital and making it part of the trading network that embraces one of the most dynamic economic regions in the world. Infrastructure de- velopment in Guangdong and in Guangzhou-with the help of foreign capital-is strengthening the all-important communication links with the hinterland, other parts of China, and external markets. At the same time, special rules governing the fiscal and foreign exchange regimes have been terminated. Guangzhou is on a level playing field with the other cities, but it has a head start and no longer needs an extra push. Starting with fewer resources and less industrial capability than the other two cities, Guangzhou has used the advantages of neighborhood to strengthen its economic position. These neighborhood effects are partly an accident of geography-Hong Kong is close by, and it is an economic powerhouse-and partly an outcome of historical conditions that drove the Cantonese to emigrate. Partly also they derive from the sequencing of China's reforms and the political guanxi (connections) that were in- strumental in giving Guangdong the autonomy to pursue policies deemed too radical initially for the rest of the country. 138 The Dynamics of Urban Growth in Three Chinese Cities Reviewing Guangzhou's performance, and juxtaposing it with that of the other cities, highlights two contrasts. First, Guangzhou's agglom- eration economies are weaker, it has fewer industrial options, and its longer-term growth prospects are less clearly defined. Second, the neigh- borhood effects, although mostly positive thus far, could turn against the city. On the first point, Guangzhou's narrower industrial base, smaller stock of human capital, and more limited education infrastructure re- strict its opportunities. Moreover, the city has not yet adopted a coher- ent development strategy that will keep it near the front of the pack well into the next century. Although it has tackled the problems of the state sector more forcefully than the other two cities, Guangzhou has not yet closed this difficult chapter from the past. The many loss-making, inef- ficient state-owned enterprises exert a serious drag on the municipal economy, and their partially reformed circumstances have undoubtedly shaved perhaps 1 to 2 percentage points off Guangzhou's annual growth rate during the last 10 years. Shanghai, and, to a lesser extent, Tianjin dominate their immediate neighborhoods. Shanghai is China's leading industrial and financial cen- ter, and Tianjin is an industrial power and the greatest port of the re- gion. Guangzhou also is a major port, but it competes against two other world-class ports-Hong Kong and Yantian-that share the business generated in the region. Similarly, as a middleman, a financial sector, and a supplier of producer services, Hong Kong is far ahead of Guangzhou and will probably retain that position after 1997. Thus, Guangzhou might be constrained by the presence of such a redoubtable contender in a sector where scale economies, an early start, and a repu- tation are important. The industrial field in southeast China is crowded, and several cities are approaching Guangzhou's level of industrialization. As industrial- ization spreads inland, Guangzhou will be exposed to competition from centers where labor and land are cheaper and the costs of urban over- head are much lighter. With so many cities vying for a place, Guangzhou will have to play its industrial cards with some dexterity and be on the lookout for promising niches. Agglomeration and neighborhood effects will still be a factor in the future, but their significance will be less. The major reforms will also have had their effect by the end of the decade. In the future, the gains in productivity, mobilization of resources, innova- tions, and improvements in the urban environment will depend on the effective combination of many micro-level policies. Similarities,6 Contrasts, and Lessons: _________Three Cities and Others China's urban development is both very old and relatively recent. Many of China's cities, including the three that occupy the center stage in this study, were established centuries ago and were imnportant players in their corner of the world. What might be described as modern urban development in China is a phenomenon of this century and one that initially embraced a relatively small number of cities, the majority lying on the eastern seaboard. This first stage lasted through the mid-1930s, when internal turmoil halted modernization. The urbanization that resumed after 1952 was both limited and in certain respects one-dimensional: cities were viewed as a necessary evil for promoting industry. By the mid-1980s, as reform and external contacts chipped away at the socialist imTage of the city, these preconceptions were seen as out- dated, and China's cities braced themselves for rapid expansion, a di- versification of functions, and a fundamental change in spirit and in autonomy. The past 18 years have provided only a taste of what is to come. But already the direction of change is becoming apparent, and Chinese cities are both learning about the problems of rapid urbaniza- tion and devising instruments to cope with them, using the advantages of size, of hinterland, of location, and of acquired industrial capability. The three preceding chapters situate Shanghai, Tianjin, arid Guangzhou withidn a specific historical-geographic context and a framework of devel- opmental possibilities. They seek to provide a sense of place and to convey a perspective on how these cities have begun adjusting to an unforgivingly competitive world, where the rewards are a function of ceaseless effort and are no longer simrply assigned by central planners in Beijing. It is a world in which budget constraints are hardening and municipalities them- selves must bear a large share of the costs of bad policies. 139 140 The Dynamics of Urban Growth in Three Chinese Cities This chapter operates along two planes: the first is comparative, and the second is concerned with models of municipal development synthe- sized from a broad range of experience. In earlier chapters we frequently juxtapose and contrast the performance of the three cities in particular areas. Here we look more systematically at the economic record and perceived strategies. We then examine forces that will drive urban change in China during the balance of the 1990s and into the next century. The concluding section explores the relevance of and lessons from the inter- national experience for the three cities. Three Cities Compared Shanghai, Tianjin, and Guangzhou have much in conmon, including certain trends in development. All three are, first and foremost, indus- trial cities to varying degrees. Their economic center of gravity is in the manufacturing sector. Second, each city has a distinctive geographic lo- cation within the regional context: they are on or near the coast, they have large and busy ports, and they are well served by the hinterlands around them. Third, as was typical for major industrial centers in China, the three cities have concentrations of state-owned industry, which are slowly being whittled down as enterprise reform continues to make headway. Fourth, because of the large share of state-owned industry, which sustained a workforce enjoying lifetime tenure, and tight restric- tions on labor mobility, the labor market in these cities was subject to many rigidities. These are diminishing but still interpose barriers to change. Fifth, all three belong to the select group of "open" cities that have spearheaded China's "opening to the outside world" and have taken the lead in introducing reform. Sixth and final, the three cities have re- ceived a substantial volume of FDI and have benefited from spillover ef- fects emanating from neighboring economies. In these respects, they are comparable and relatively well matched. They are also, to a considerable extent, representative of other large industrial cities of China such as Chongqing, Fuzhou, Hangzhou, Ningbo, Qingdao, and Wuhan. In other respects, the three cities diverge from one another in ways that draw attention to size, industrial composition, location, share of state industry, and reform initiatives. Of the three, Guangzhou has con- sistently grown the fastest (see table 3.1). During 1952-78, its GDP in- creased at an annual rate of 9.5 percent, with Shanghai in second place at 8.8 percent and Tianjin bringing up the rear with a growth rate aver- aging 7.3 percent. In the 15-year period after the start of reforms (1978- 93), Guangzhou maintained the lead with a growth rate of 11.3 percent, which was somewhat above the national average, while Shanghai and Tianjin grew at a roughly similar pace of 8 percent, more than 1 percent- age point below the national average. Industrial performance was broadly similar, with Guangzhou once again leading the group during 1952-78, albeit not by a wide margin. But after the start of reforms, Similarities, Contrasts, and Lessons: Three Cities and Others 141 Guangzhou's lead widened substantially, with GVIO rising annually 15.7 percent compared with 9.1 percent for Shanghai, although part of this might be traceable to biases in measurement. Guangzhou's superior performance in the prereform era was the out- come, most probably, of greater economic autonomy and the smaller share of centrally managed state industry. Unlike the other two cities, Guangzhou was less subject to central direction, less bound by plan tar- gets, and not forced to serve as a vehicle for generating state revenues. It did not receive major state-funded investment projects, but by the same token neither were its development and fiscal policies tightly regulated by Beijing. Although both Shanghai and Tianjin started out in the mid- 1950s with a commanding lead over Guangzhou in per capita GDP, this gap no longer exists. During the reform period, Guangzhou also went further in dissolv- ing large, inefficient state enterprises. By 1993 state-owned enterprises accounted for less than half of the city's GVIO and only 20 percent of all industrial enterprises (see table 5.1). Its trial enterprise restructuring pro- gram made headway by transferring property rights of more than 50 loss-making state-owned enterprises to shareholding or foreign owner- ship. As a result, joint ventures and cooperatives multiplied and by 1993 produced close to 40 percent of municipal GVIO. By contrast, Tianjin was the slowest to reform its state-owned enterprises, and by 1993 the state sector still accounted for close to 60 percent of production. In particular, Tianjin lagged behind in devising a workable program of state-owned enterprise reform, whereas Shanghai recently moved forward with an acquisition and merger program. There are other differences as well, which are only briefly summa- rized here. Both Shanghai and Tianjin have an industrial base that is broader and stronger than that of Guangzhou, with Shanghai decisively in the front. These cities enjoy more generous agglomeration economies deriving from the breadth of their industrial base, although during the 1980s they suffered relative to Guangzhou with respect to the scale of the services sector. When various aspects of subsectoral performance are compared, Shanghai dominates Tianjin and Guangzhou in profit- ability and productivity across six major industries (see table 6.1). Inter- estingly, Guangzhou is ranked second in the majority of industries, sug- gesting that its relatively small share of the state sector, its more com- petitive market environment, its greater FDI, and better developed pro- ducer services outweigh Tianjin's industrial reach and agglomeration economies. Shanghai's indicators of costs are by and large the lowest, with Guangzhou again in second place and Tianjin at the bottom of the list. It is only with respect to the competitiveness of wage rates that Tianjin has an edge over the others, a reflection of greater control over prices and the relative slowness of the reform process. When all the indicators are pulled together, the index of overall performance favors Shanghai in each 142 The Dynamics of Urban Growth in Three Chinese Cities of the sampled industries except one-transport equipment-where Tianjin seems to have an advantage. In other subsectors, Tianjin ranks above Guangzhou in metallurgy and electronics, which seems very plausible from the evidence presented in earlier chapters. Guangzhou is ahead in textiles, a testimony, no doubt, to the heavy FDI in that subsector. It also has an advantage in chemicals, again possibly because of investment in new plant. The higher ranking in machine building may reflect Guangzhou's greater specialization and market orientation, which offset steeper wage costs and relative lack of accumulated experience in the sector (see figures 6.1-6.6). Two of the cities, Shanghai and Tianjin, are starting to confront the problem of an aging workforce. Compared with the national urban av- erage, Shanghai's population in 1990 was the most mature (see figure 6.7). Persons above 50 years of age, both male and female, accounted for close to 20 percent of the municipal population, against the benchmark (Text continues on page 149.) Table 6.1 Industrial Comparative Advantages in Shanghai, Tianjin, and Guangzhou, 1988-92 Average Cost Profit Productivity Wage rate Industry (PT/GVIO) (GVIO/OVFA) TPC/GVIO (thousands of yuan) Index' Textiles Shanghai 0.12 2.86 0.80 2.92 0.19 Tianjin 0.01 1.93 0.92 2.27 0.13 Guangzhou 0.03 2.60 0.82 3.60 0.15 Machine building Shanghai 0.17 1.89 0.82 3.28 0.17 Tianjin 0.06 1.47 0.89 2.55 0.13 Guangzhou 0.07 1.96 0.94 3.98 0.15 Metallurgy Shanghai 0.12 5.17 0.84 3.58 0.27 Tianjin 0.10 4.88 0.86 3.26 0.25 Guangzhou 0.11 3.28 0.76 4.05 0.20 Basic chemicals Shanghai 0.16 2.40 0.79 3.95 0.19 Tianjin 0.12 1.62 0.86 2.63 0.15 Guangzhou 0.18 1.79 0.75 4.96 0.17 Transport equipment Shanghai 0.18 2.91 0.83 3.62 0.21 Tianjin 0.10 4.34 0.88 3.11 0.24 Guangzhou 0.10 3.39 0.89 4.72 0.21 Electronics Shanghai 0.08 2.65 0.91 2.99 0.18 Tianjin 0.06 1.71 1.04 2.43 0.15 Guangzhou -0.10 2.69 0.91 3.35 0.12 Note: PT, profits and taxes; GVIO, gross value of industrial output; OVFA, original value of fixed assets; TPC, total production costs. a. Index equals {[(PT / GVIO) + (GVIO/OVFA) / 101 + [(TPC / GVIO) / 5]1 / 3. Source: Survey of Industrial Enterprises, conducted by the Policy Research Department, World Bank, and the Department of Economics, Brandeis University. Figure 6.1 Textiles Industry in Shanghai, Tianjin, and Guangzhou, 1988-92 Profits and taxes/GVIO Total production costs/GVIO Ratio Ratio 0.15 120 -2 O'.10~~~~~~~~~~~~~~~01 010; ; 080 0.60 0.00 // - |b: --/ Guangzhou 040 / Guangzhou -0 05 1' / /Tianjin 0.20 l Tianjin -010 1988 1989 / hnhi00 ~/Sanghai 1990 19 1921988 1989 1990 1989 1990 1991 199Z Gvtoloriginal value of fixed assets Thousands Wage rates Ratio of yuan 4.00 5.00 3.00 4.00 3.00 2.00 1 00 G~~~~~uangzhou 1 Guangzhou ,/ Tianjin 100 ' Tianjin 0.00 Sh~~~~~~anghai 0.0- Shanghai 1988 1989 1990 19 19/ 1988 1989 1990 1 1992 1988 1989 1990 1991 1992o19n0o1tk, Dt nd i Source: Survey of Industrial Eniterpr-ises, conducted by the Policy Resear-ch Department, World Banik, and the Departnment of Economics, Brandeis University. Figure 6.2 Machine-Building Industry in Shanghai, Tianjin, and Guangzhou, 1988-92 Profits and taxes/GVIO Total production costs/GVIO Ratio Ratio 0.30 02020 1.no 0.10 0.00 Guangzhou 0.50 Guangzhou 0.00 // ' , . , / / ~Tianjin /Tianjin -0.10 88 ' / Shanghai Shanghai 1988 1989 1990 1988 1989 1990 Gvio/original value of fixed assets Thousands Wage rates Ratio of yuan 2.50 21.o5o0 4.00 |/-/ /// / Guangzhou 2.00 Guangzhou 0.50 //Tianjin a ianjin 0.00- < Sanghai ° °° /Shanghai 1989 1990 1991 1992 S1989 1990 9 Source: Survey of Industrial Enterprises, conducted by the Policy Research Department, World Bank, and the Department of Economics, Brandeis University. Figure 6.3 Metallurgy Industry in Shanghai, Tianjin, and Guangzhou, 1988-92 Profits and taxes/GVIO Total production costsl/Cvio Ratio Ratio 0 25, ~~~~~~~~~~~~~~~1.00 0.20 0.80 2 ~~~~~~0.40Ganho 0.10 Guangzhou Guangzhou 0.20 ( Tianjin 0.00- 0.0 hanghai 8 19990 1991 S a 18 1 1990 1991 1992 911992 Gvio/original value of fixed assets Th d Wage rates Ratio -- of yuan 10.0)0< , , i _ ] 6.00 8 0(0'--,--S- - 5-0010- j 8.00) 6.00 -~~~~~~~~~~~~~~~30 4.00 20 Guangzhou 2.00- u 2.00 Tianjin 1.00 Tianjin 09800> ~-- Shanghai 0.00 / Shanghai 1988 1989 19 I991 1992 1988 1989 1990 by t P 1990 1991 1992 1990 ~~~~~~1991 1992 Source: Survey of Induistrial Enterprises, conduuced by the Policy Research Department, World Bank, and the Department of Economnics, Brandeis University. Figure 6.4 Chemicals Industry in Shanghai, Tianjin, and Guangzhou, 1988-92 Profits and taxes/GVIO Total production costs/GVIO Ratio L 00 |Ratio 0 -- ; 0.25 - 1.00-- 0.80 0.20 , 1 0.40 55-59 50-54' Male &>* 45-49 Female Eggss sg g.gg w-40-44' 35-39 < E a gii ~~~~25-299................. > >n g, 20-24 15-19 a 5-9 . at> , , w~~r1-~W'&& Igr9 @ r. 0-4 . } >»> g aj. >s 1-3 1; 9 7 5 3 i ) j 3 5 76 9 t1 13 Percent Percent Shanghai, 1990a Age group S .....................b.<;b55-59 == _saa gagamw>.50 54 YRgi a _ 45 4945_4_ Male 40 44 g >-& >l Female & 35-39 *30-34 .9a'7711= ggygaa9ySg < ix- ia af9s@> 25-29 agigg'ggrYg 2 5 19 .10-14 agi 99.9 ....- 0-4 13 11 9 7 5 3 1 0 0 1 3 5 7 9 11 13 Percent Percent Similarities, Contrasts, and Lessons: Three Cities and Others 151 Tianjin,1 991 Age group 70+ 65-69 77-. ] 60-64 7 55-59 50-54 45-49 Male --.- 40-44 . . . Female 35-39 _____________ =___________ 30-34 25-29 I . ' ',20-24 15-19 10-14 - . 7 .5-9 i . 0-4 13 11 9 7 5 3 1 0 0 1 3 5 7 9 11 13 Percent Percent Guangzhou,1 990 Age group 70+ 65-69 60-64 55-59 50-54 45-49 Male -- ...40-44 _ M ale . .- - - ...... 35-39 7 Female 30-34 25-29 20v24 E---_-- ___*_, _,-g-_,; _, 15-19 l = s , ~~10-14 __*_______;-,______:- __-___.,._. _; _________... __ 5 -9 ___ _ . ___ _ : - . - __ |~~~ ~~~ Ti: 5--t:9-1i 1- ^°4 1 1 ~~~ 1 0-4 13 11 9 7 5 3 1 0 0 1 3 5 7 9 11 13 Percent Percent Note: Urban population includes city-administered counties and agricultural population. a. Based on 1990 population census. Source: State Statistical Bureau 1992b; Statistical Yearbook of Shanghai, 1992; Statistical Year- book of Tianjin, 1992. 152 The Dynamics of Urban Growth in Three Chinese Cities tive shares. By 1993 they accounted for close to 55 percent of GVIO in Shanghai and 51 percent in Tianjin. Moreover, each of the three cities is striving to develop service industries, particularly financial and producer services. Shanghai has the potential to become a world-class financial center, whereas Guangzhou might pursue trade-related services, and Tianjin is well positioned to develop transport-induced activities. All three cities, Tianjin in particular, are facing increasing competition from secondary cities in their hinterlands. For instance, during the pe- riod from 1985 to 1991, Shijiazhuang became a serious contender with Tianjin in both textiles and clothing industries (see table 6.3). Foshan is challenging Guangzhou in textiles and electronics; in fact, its electronics production surpassed that of Guangzhou in 1991. Dongguan is another contender in both clothing and electronics industries. Of course, all three cities are still clinging to their advantages in such large industries as machine building, chemicals, metallurgy, and transport equipment. However, only Shanghai is likely to remain competitive across these subsectors over the longer term. The 1980s were a period of ferment. The retreat from planning and centralized management induced each of the three cities to redefine its relationship with the center and to begin chalking out a modernization strategy, which would be implemented largely by itself. All three cities are starting from a broadly similar situation with regard to urban infra- structure: urban amenities are poor, transport and communication links are quite weak, housing is scarce, and the environment is heavily pol- Table 6.3 Shares of Gross Value of Industrial Output of Selected Subsectors in Shanghai, Tianjin, and Guangzhou and Secondary Cities in the Hinter- lands, 1985 and 1991 (percent) Textiles Clothing Machinery Electronics Chemicals City 1985 1991 1985 1991 1985 1991 1985 1991 1985 1991 Shanghai 14.4 9.3 15.6 8.2 12.6 10.8 16.1 10.3 11.9 8.2 Tianjin 4.1 2.8 5.0 2.5 3.9 3.5 5.4 4.0 5.0 5.2 Guangzhou 1.3 1.4 2.0 2.2 2.0 2.2 3.1 1.4 1.7 2.7 Hinterland Shijiazhuang 1.7 1.8 1.6 1.5 0.7 1.0 0.6 0.8 0.7 0.7 Tangshan 0.5 0.4 0.6 0.4 0.5 0.4 0.0 0.0 0.4 0.2 Qinhuangdao 0.1 0.1 0.2 0.1 0.1 0.1 0.0 0.0 0.1 0.0 Wuxi 3.6 2.3 1.3 0.9 2.4 1.5 4.5 2.2 1.4 0.7 Suzhou 4.8 1.5 2.2 0.5 1.9 0.8 3.1 1.5 2.1 1.0 Ningbo 1.7 0.9 1.8 1.0 1.0 0.5 1.1 0.5 0.4 0.4 Foshan 1.4 0.9 1.4 0.6 0.4 0.3 1.8 1.7 0.3 0.4 Zhanjiang 0.1 0.1 0.1 0.4 0.1 0.2 0.1 0.4 0.1 0.2 Dongguan 0.1 0.3 0.3 0.8 0.1 0.3 0.0 0.9 0.1 0.1 Source: State Statistical Bureau, China: Urban Statistical Yearbook, 1986 and 1992; Statistical Yearbook of China, 1987. Similarities, Contrasts, and Lessons: Three Cities and Others 153 luted. The funds to remedy these problems are almost equally scarce, but each city is relatively well placed to mobilize fiscal and financial resources and to attract foreign capital. The three cities have unexploited agglomeration and neighborhood economies that they can use, and all are beginning to do so, with Guangzhou displaying the most energy. Likewise, much momentum remains to be derived from enterprise and urban land reform. Guangzhou was the first to initiate reforms, and this is reflected in its growth rates and productivity. Now Shanghai and Tianjin are attempting to catch up. Developing producer as well as other services and modernizing the infrastructure are clearly a priority for Guangzhou and Shanghai and, to a lesser extent, for Tianjin, whose fu- ture growth is dependent to a greater degree on its role as a regional transport hub. Above all, each city has begun tackling the matter of in- dustrial strategy in a piecemeal fashion, but the time is approaching when a comprehensive reappraisal is called for. Driving Forces for Urban Change in China Urban modernization in China is a river that is fed by numerous streams of greater and lesser significance. It is neither necessary nor feasible to give each its due because attending to the major ones is adequate to depict the big picture. Beyond that, analyzing urban modernization from several vantage points can impart a sense of China's urban dynamics in the final years of the twentieth century, a sense, moreover, that is fil- tered through the accumulated experiences of cities in other countries. Because this study emphasizes the industrial facet of urban develop- ment, we highlight three factors that are currently driving urban change in China and will remain pivotal for at least the next decade and more. They are industrial organization and urban industrial networks, urban finances, and the labor market. Industrial Organization and Networks Two forces are propelling the growth of industry in urban China: the emergence of many new firms, which are generally quite small, and the creative restructuring of large state-owned enterprises and collective- owned enterprises.' The emergence of new firms is the outcome of buoy- ant entrepreneurship and the fortuitous coming together of many rela- tively specialized enterprises into networks that distantly mimic similar constellations in industrial countries. The creative restructuring of state- and collective-owned enterprises is expressed in several ways: * Vertical disintegration whereby a big industrial concem is divided into a core unit and affiliated subsidiaries, which provide inputs and subassemblies to the mother enterprise but also to other producers 154 The Dynamics of Urban Growth in Three Chinese Cities • The formation of corporate groupings that bring together firms into formal networking arrangements in which overall financial control and strategic functions are centralized in a parent holding company * Loose alliances that permit rationalization of production, joint financ- ing of activities subject to externalities, and agreement over rules of market behavior along with cooperation among firms. Industrial networking has its intellectual antecedents in Alfred Marshall's view that the matrix of production is a region or an indus- trial district and not a firm. Networking is a means of pooling informa- tion, capital, expertise, and infrastructure and is a mode of insurance. It lowers entry barriers by providing new start-ups with a bundle of es- sential ingredients.2 Thus a dynamic, networked, industrial community has many layers. At one level, there is the layer composed of social or kin-based overlapping relations, which entangle the individual with the ties of reputation, family, and friendship. Local banks, credit societies, and informal savings societies embedded in the networks depend criti- cally on the flow of information. While most small start-ups in China and elsewhere draw mainly on personal savings of the entrepreneur or resources obtained from the community through various relational mechanisms, the growth of small enterprises is frequently a function of the availability of credit from local institutions of finance. At another level of the network, there are circuits that transmit R&D. Powell and Smith-Doerr (1994, p. 387) note that "bonds of professional membership greatly expedite the formation of collaborative R&D net- works. The sense of common association with a technological, intellec- tual, or scientific community is a glue that thickens cooperation. This membership in scientific or industrial associations is ongoing, occur- ring outside of commercial relationships; thus members monitor indi- viduals' behavior and reputation." Interfirm movement of individuals helps knowledge circulate and stimulates a trade in information, which is the key to an innovative milieu. "When [an] R&D network brings firms together, the sharing of different competencies can generate new ideas." Services constitute still another layer. These cover a broad spectrum. At one end there are technical colleges and vocational institutes that infuse networks with skills. At the other there are testing facilities and organizations that generate information on technical standards enforced in foreign markets and the translation of tenders advertised overseas (World Bank 1994a). In sum, the layered nature of networks and the involvement of enterprises at many levels-resource exchange, infor- mation, board of directors-account for the effectiveness of the ones that work best. Networking arrangements are particularly suited to certain kinds of industries where production on a small scale, specialization, and collaboration between geographically clustered firms not only is efficient but also promotes innovation. Similarities, Contrasts, and Lessons: Three Cities and Others 155 Industrial modemization and restructuring in the three cities require forms of organization that will ensure gains in efficiency and technol- ogy sufficient to put China's firms on a par with those of its more ad- vanced East Asian neighbors. How industrial organization evolves will depend on a host of factors, in particular administrative, legal, and in- stitutional initiatives pursued in each city. International experience, al- though remaining short on industrial specifics for individual countries, is increasingly unanimous on the broad guidelines. Major elements include industrial groupings, mergers, joint ventures, and subcontracting. Industrial groups in which a major enterprise exerts strong financial and strategic control over the members have certain advantages. Unlike alliances, groups need not materialize through some lengthy organic process but can be created rapidly, with the precise details regarding management and financial control determined over time. Given current administrative arrangements in China, the deliberate creation of groups (or corporations, which include a smaller number of enterprises) through the joint action of municipal or provincial agencies saves time and per- mits a linking together of enterprises drawn from different jurisdictions. Groups can have a definite subsectoral focus, in order to cultivate a basic competence and streamline the production system. The experience of au- tomobile assemblers in Japan and the initial results achieved by the group headed by Second Auto Works (Changchun) suggest that when the core enterprise can provide able leadership and other members are fully com- mitted to its goals, industrial groups can tap deep reserves of dynamism. The use of groups as a vehicle of industrial policy confronts two reali- ties. Large-scale restructuring and modernization of industrial subsectors by means of centralized planning are not feasible. Likewise, the market cannot efficiently guide macro changes in the structure of industry or major investment decisions with a long-term strategic purpose. An in- termediate organizational form is needed that is capable of making and implementing strategic choices and coordinating the actions of multiple participants. But the organization must partake fully of the market mi- lieu and reserve some independent decisionmaking role for the various enterprises subsumed within the group. One of the most ambitious efforts in this sphere, which embraces en- terprises from all cities, is the Chinatron Corporation. This electronics giant, which brings together such enterprises as Great Wall Computers (Tianjin) and Panda Electronics, is attempting to concentrate R&D, train- ing, electronics, and machinery engineering within a single corporate structure.3 Chinatron represents an initial effort to create a viable corpo- rate grouping, and it mirrors the organizational weaknesses of other corporations created in the three cities. By and large, differentiating be- tween administrative units and corporations is still difficult. Often only the name is changed, with the structure and functions of the old bureau remaining intact. Even when the bureau and corporation are separate and distinct, the former retains control over major divisions and per- 156 The Dynamics of Urban Growth in Three Chinese Cities sonnel management. The idea of setting up autonomous corporations to restructure enterprise clusters and stimulate market competition is a good one, but so far evidence of true autonomy is sparse and few corpo- rations have gone far in rationalizing plant or closing obsolete facilities. Among other corporations created in the three cities, a typical example is the Wanbao Corporation of Guangzhou, founded in mid-1988, which produces refrigerators, fans, and washing machines.4 Wanbao is made up of 26 enterprises, each consisting of an independent accounting unit with its own contract. During the early 1990s, it began to remake itself into a holding company and to work on the precise legal, structural, and managerial implications of such a transformation. To be successful over the medium term, the corporation must achieve autonomy from direct administrative supervision, while remaining accountable for its perfor- mance to shareholders and stakeholders. Wanbao, much like other state enterprises in transition, is attempting to define the nature of ownership and to ascertain the hardness of budget constraints in the environient of reform. Much depends on the degree of competition. Only if competitive pressures become more severe will the corporation be induced to use au- tonomy to ensure that its various operations satisfy the test of long-run profitability. Clearly, measures that increase market competition must pro- ceed hand in hand with attempts to change industrial organization. Mergers offer yet another path to rationalizing production capacity within more or less optimally sized units and to lessening market un- certainty by increasing market concentration. Free market advocates see mergers as a way to bring productive assets under the control of the most efficient managers and to internalize certain transactions within a larger whole that cannot be conducted efficiently through market deal- ings. The common tendency is to merge a failing enterprise with a stron- ger one. In a number of instances, this is a surrogate for bankruptcy. It is a means of salvaging usable physical and human capital. To the extent that mergers achieve this purpose, they prune obsolete or inefficient capacity and maximize the returns from industrial resources. But using mergers to maintain inefficient and loss-making firms in production not only perpetuates waste and maintains obsolescent production capacity but also compromises the health of the successful enterprise, which has to take on new managerial and production burdens, operate multiple facilities, sacrifice funds that might otherwise be used to upgrade its own plant, and bear the cost of assimilating the culture of a company in decline. For rea- sons of administrative and financial convenience, mergers have, thus far, been the preferred route to restructuring, especially in Shanghai and Tianjin. Empirical research conducted in industrialized economies on the re- sults of horizontal mergers is equivocal at best, while research on con- glomerate mergers is negative. Even where mergers are the market- driven outcomes of voluntary and well-informed decisions, half or more pull down the profitability of the merged entity from the level achieved by each individual firm prior to the union (George 1991). Changes in Similarities, Contrasts, and Lessons: Three Cities and Others 157 ownership that reduce organizational slack can bring the performance of moderately inefficient enterprises close to the average, but this re- quires two to three years at least, and performance is likely to deterio- rate initially (Lichtenberg 1992). Another problematic side to horizontal mergers is their effect on in- dustrial concentration. Views on this have tended to change, with much stress being placed on scale economies, the minimum efficient scale of R&D, and the investment outlay required to sustain competitiveness in industries with short product cycles and high tooling costs. Neverthe- less, the dangers of excessive oligopolistic power loom large. The im- portance of scale economies can also be overplayed. In a wide range of industries, medium-size firms are as competitive and innovative as large companies, if not more so (Freeman 1982; R. Howard 1990; Pratten 1991; for a historical perspective, see Mowery and Rosenberg 1989). In Ger- many and Italy, the "backbone" enterprises are midsize specialist firms (Rommel and others 1995). Restructuring in the sense of technological modernization can also be sought by way of joint ventures with domestic or foreign partners. These have become a commonplace in industrial economies and are increas- ingly widespread in China as well. Four concerns motivate firms to en- ter into joint ventures. First are the advantages of sharing product or process technology. Second is the ability to spread risks in the develop- ment of complex new products involving large outlays. Third is the pos- sibility of market access. Last is the ability of firms to pool complemen- tary strengths. One partner might have labor and a choice production site, while the other might have the capital and marketing apparatus to permit full use of the existing plant. The success of joint ventures depends on balancing the benefits and reconciling differing operational philosophies. It calls for perseverance by both parties, a fairly high degree of organizational flexibility, and a willingness to weather the inevitable frictions. Over the past 10 years or so, foreign investors have entered into many joint ventures with Chi- nese partners. They have infused capital, technology, organizational change, managerial skills, and marketing expertise into a handful of important manufacturing subsectors: automobiles, clothing, and elec- tronics. One striking instance is the Xerox Corporation of Shanghai, which has followed an effective strategy of training 60 suppliers in production and materials handling so that they can meet quality and delivery stan- dards. Xerox has expended several million dollars in this effort, with the Shanghai Economic Conmmission providing supplementary funding. Most joint ventures have encountered start-up problems having to do with style of management, complexity of labor relations, low degree of formalization in operating procedures, and quality of infrastructure. In- dustrial enterprises in China also have to cope with the thickets of bu- reaucracy that surround an industrial enterprise in China (China-Euro- pean Commission Management Institute 1990; World Bank 1991c; 158 The Dynamics of Urban Growth in Three Chinese Cities Pearson 1991). Foreign investors entering into a joint venture with a state enterprise often have difficulty coping with the overhead and complexi- ties of personnel management that arise from the far-ranging social re- sponsibilities of the enterprise. Overstaffing and permanent employment pose one set of problems. But the enterprise must also take care of fam- ily records, health care, education of dependents (frequently also the employment of wives and children), housing, and public security. Usu- ally the Chinese partner attempts to keep a tight grip on personnel man- agement and finances, which tends to frustrate the foreign company's efforts to improve labor use and financial accountability. The majority of joint ventures are eventually able to overcome the teething difficulties and turn a profit after a few years, but not without a struggle. Joint ven- tures yield promising results only when the foreign partner is committed and the Chinese partner is fully prepared to be flexible; to adapt technol- ogy, soft as well as hard, to local circumstances; and to modify work prac- tices so that they begin to approximate those found in market economies. The activity of agglomerating, creating alliances, establishing groups, merging, and setting up joint ventures can introduce organizational changes, some positive and others less obviously beneficial. Among the organizational changes, one of the most far-reaching is the role assigned to subcontracting. The utility of subcontracting within certain manufac- turing subsectors has undergone a fundamental reappraisal. Earlier in the century, when the industrialized countries were pushing toward maturity, major corporations preferred a high degree of vertical integra- tion within giant facilities (Chandler 1990). This was the era of manage- rial centralization and scale economies. It not only shaped the corporate sector in Europe and the United States but also proved attractive to the centrally planned socialist economies. Eastern Europe and the former U.S.S.R. embraced giant, vertically integrated production complexes that promised managerial and scale economies. They exported this organi- zational strategy to China, which diluted it by emphasizing decentral- ized development and retained vertical integration while tempering very large-scale production and tight control from the central government to allow for Chinese realities. There is a long tradition in China's enter- prises of setting up subsidiaries so as to employ relatives and children of staff. In recent years, the tax system has reinforced the incentive to do this. To minimize tax obligations, a state enterprise often finds it advan- tageous to create a largely tax-exempt collective to produce a profitable new product rather than to manufacture the item on its own premises. The larger state enterprises in the three cities buy components from subcontractors, most of which are located within the municipal bound- aries, although firms in Shanghai have dealings with affiliates in nearby provinces as well. Nevertheless, the scale of subcontracting is fairly small by Japanese, Taiwanese, and now even U.S. standards, averaging about 20 to 30 percent among the large state enterprises. It is conspicuously low among major machinery producers, who stand to gain the most Similarities, Contrasts, and Lessons: Three Cities and Others 159 from subcontracting. Enterprises and bureaus confront a number of prob- lems. Smaller subcontractors, unless they happen to be located nearby and are closely supervised, cannot meet delivery and quality standards. Because of China's transport constraints, regular delivery from distant suppliers can be difficult to arrange. And when a supplier in another prov- ince or municipality receives a contract, the local finance bureau loses tax revenue. These problems can be resolved, but they require initiatives on the part of both the bigger firms and the supervising bureaus. China's major municipal centers already possess the institutional and cultural essentials of successful subcontracting. They have a tradition of networking between enterprises and a strong belief in the binding sig- nificance of informal long-term relations. Furthermore, as revealed by the experience of coastal provinces, the Chinese extended family can be the enterprising nucleus for dynamic small-scale enterprises (On the role of the family in building businesses in Hong Kong, see Redding 1990.) People do possess the business acumen and skills to set up small firms. The family provides the organizational glue and a mechanism for mobi- lizing funds. Interfamily and community networks help to generate business opportunities, thereby lessening the degree of market risks. Enterprise reform that encourages the vertical disintegration of larger firms, the full exploitation of new information technology, and the build- ing of networks would help local industry to exploit other strengths and draw closer to international competitors. Urban Finances Urban industrial restructuring is costly. Both international experience and the lessons from China's own efforts over the past decade are unequivocal on this score. To achieve their dual objectives of urban mod- ernization and industrial restructuring, all three cities must devise suit- able policies to ensure that an adequate volume of financing is forthcoming for these activities. Making provisions for the necessary funding is a precondition for the successful implementation of an urban-industrial strategy. This is especially critical at the current junc- ture: cities are still facing an uncertain tax base; revenue transfers to the central government, although lower than in the past, remain sizable; mounting competition is squeezing the profits of enterprises; and in- vestable surpluses are diminishing. Even as traditional sources of funding decline, new opportunities are appearing in both fiscal and financial spheres. New taxes and user charges can restore the revenue base. Moreover, the progressive development of capital markets can mobilize a larger volume of local savings as well as resources from elsewhere in the economy. Because fiscal-financial reforms are still at an early stage, the full resource potential of the cities and the neighboring regional economies has yet to be utilized. Once that is done, the problems of restructuring will appear far less formidable. 160 The Dynamics of Urban Growth in Three Chinese Cities Each of the municipalities operates under a slightly different set of fiscal rules, but two principles apply to all. First, none of the munici- palities can legally establish its own taxes or other formal sources of revenue. The central government imposes taxes, sets rates, and then assigns revenues to the local authorities. Second, within each of the municipalities, budgetary revenues accrue separately to central and municipal authorities (and, in the case of Guangzhou, to provincial authorities as well). In general, central government revenues come from enterprises or entities under central government jurisdiction: for ex- ample, railroads, the banking system, and a few industrial subsectors. Municipal governments have no direct control over these revenues. Although local fiscal authorities have no formal authority to create new avenues of financing, they have usually been able to find new types of financial resources that they can control more effectively. For instance, in the early 1990s fiscal authorities in each of these cities created a "re- volving fund" that was used for small-scale investments, predominantly in industry. Ordinary fiscal grants for industrial investment were trans- formed into repayable loans, and both repayment of principal and in- terest payments were channeled into the revolving fund, gradually build- ing available resources. Loans were made at interest rates equal to, or up to a point below, bank interest rates, typically for periods of around one year. Such funds are operated in a legal gray area, in the face of opposition from officials in the banking system (perhaps as a result, sta- tistics on the funds were uniformly imprecise in the three cities). Municipal fiscal authorities are clearly in a difficult position. They are caught between two important economic trends. On the one hand, tra- ditional sources of revenue are declining as the monopoly profits of the state industrial sector are competed away. On the other hand, the grow- ing need for municipal infrastructure, education, urban amenities, and social security is placing new demands on local budgets. Thus far, mu- nicipal authorities have resolved this conflict by reducing the share of revenues turned over to the central government and by drawing on new financial sources to fund industrial investment. Nevertheless, they con- tinue to be squeezed between conflicting demands on their resources. Unable to provide large amounts of direct finance, municipal fiscal au- thorities have fostered the restructuring process through a variety of tax breaks and concessionary policies. Concessionary tax policies are typically coordinated by the municipal economic commission. They include exemption from product and value added taxes for up to two years for items that are new to the municipal- ity, subsidization of interest on loans, permission to repay loans from before-tax income, remission of import duties on imported materials and equipment, accelerated depreciation (30 percent additional) for projects in accord with national industrial policy, and generous expense provisions for small investments associated with restructuring. These concessionary policies can be an important aid to the restructuring pro- Similarities, Contrasts, and Lessons: Three Cities and Others 161 cess, but municipal authorities find it difficult to restrict their use to those cases where tax relief is genuinely required. In Guangzhou, 80 percent of tax rebates were given to the city's 20 centrally controlled enterprises. Tax concessions frequently threaten to become a general form of subsidy, rather than targeted restructuring assistance. More signifi- cant, in financial terms, is the rebate of the value added tax on exports, which is automatic rather than discretionary. Financial constraints are not the sole, or even the primary, obstacle to industrial restructuring. In many cases, fiscal authorities are willing to con- duct restructuring programs but face obstacles from industrial bureaus or concerns about employment. All three cities face difficult tradeoffs caused by the relatively high tax burden that state-owned factories currently bear. Because Guangzhou's economy and fiscal revenues are growing overall, fiscal authorities in Guangzhou are willing to bear substantially more risk, lowering tax burdens temporarily on individual firms in the knowledge that the money could be raised subsequently from vibrant sectors of the municipal economy. By contrast, growth and structural change in Shang- hai and Tianjin were slower, which made it difficult, until recently, for fis- cal authorities to bear that kind of risk, because they had difficulty identi- fying the alternative sources of tax revenue. Although Guangzhou has succeeded in defining a more flexible policy for itself, the dilemma facing the three cities underscores the obstacles that high tax burdens impose on restructuring. It may also illuminate a key feature of restructuring: the more rapidly it progresses, the more resources will be available to carry it out. Growing economies can undertake restructuring more easily. Bank credit has emerged as a major source of financing. Over the past decade, China's industrial financing shifted from budgetary grants and retained funds (internal financing) to bank credits. China increasingly resembles market economies in which surpluses generated in the house- hold sector are loaned to the enterprise sector to finance a large portion of investment. However, in the current partially reformed state of China's economy, a large proportion of loan financing is channeled through the state banking system, which is still subject to a high degree of govern- ment control, exercised in part by the earmarking of loanable funds for specific purposes. One such purpose is industrial restructuring. Accord- ing to one estimate, the government earmarks two-thirds or more of the total amount of bank fixed investment lending (both capital construc- tion and renovation investment), mainly for state-owned enterprises. Thus, although bank loans are a significant financial resource, they are allocated administratively by various levels of government, which lim- its the efficiency with which they can be used. The current system of allocating bank loans retains some of the practices embedded in the old system of administrative allocation of investment, which is responsible for the problems of industrial restructuring that are so pressing today. The banking system is now capable of moving substantial funds across provincial boundaries. As a result, the cities have also attracted a certain 162 The Dynamics of Urban Growth in Three Chinese Cities amount of portfolio investment, at least in the form of bank deposits. Nonresidents of the municipality hold more than 10 percent of total household savings deposits, in the case of Shanghai, reflecting the supe- rior convenience and confidentiality of holding savings in a major ur- ban center. Similar shares are reported for the other cities. The superior productivity of coastal regions and the greater flexibility of economic policy outside the major cities enhance the attractiveness of these areas. So the development of new financial instruments sets the pace at which coastal cities are able to attract capital from other regions. Stocks and, especially, bonds have already demonstrated their attractiveness to Chi- nese households. Household saving is extremely high, but the range of assets available to households is still quite narrow (overwhelmingly concentrated in savings deposits). As the range of available assets ex- pands, the coastal cities should be able to tap these sources. Another new option may lie in the use of venture capital.5 Venture capital is different from other forms of capital in that institutionalized venture capitalists provide equity rather than debt financing and fre- quently take an active role in managing the enterprises they finance. In the United States, for example, many new companies and technologies, including Apple Computer, Fairchild, Intel, and Sun Microsystems, would not have been launched nor would they have attained commer- cial success so quickly without such capital.6 The major centers of ven- ture capital include Silicon Valley, New York City, and the Route 128 corridor outside Boston, where such capital has been used mainly to establish high-technology enterprises. In Silicon Valley, the archetypal start-up was formed by a group of friends and colleagues with an inno- vative idea that they could not realize in their current workplace. They initiated a business plan and sought funding from local venture capital- ists. The expanding local circle of university researchers and consult- ants provided additional assistance (see Saxenian 1994). Venture capital would be conducive to the development of new tech- nologies and enterprises because venture capitalists often bring techni- cal skills, production experience, and networks of contacts, in addition to their equity investment. Geographic proximity and the existence of a critical mass of researchers can be a further advantage. An expanding network of special suppliers and service providers also facilitates the start-up process. Both Shanghai and Tianjin could use much more ven- ture capital to finance technological innovations and enterprises. Labor Market Labor market flexibility, by enhancing productivity, can impart greater momentum to industrial change. Restructuring becomes a positive sum game where the winners compensate the losers and thereby retain their support. China's major cities are appropriate laboratories for observing the state of labor markets and for devising policy measures that will Similarities, Contrasts, and Lessons: Three Cities and Others 163 assist not just industrial restructuring but also the larger purpose of self- sustaining urban development. All three cities recognize the need to enhance labor market mobility and to expand the authority of enterprises over labor market decisions. The "iron rice bowl" not only has tied the hands of employers but also has sharply circumscribed the ability of workers to quit and has bound them in a dependent relationship in which employers are responsible for almost all their living needs (Walder 1991).7 Although local govern- ments are no longer obligated to find jobs for all graduates and enter- prises are not required to accept those being placed, labor authorities and work units are still subject to employment imperatives and recruit- ment plans. Labor reforms in a city such as Shanghai have enlarged a graduate's choice of jobs: an individual is no longer assigned to a unit. Likewise, an enterprise need not accept a candidate selected by the la- bor bureau but can recruit directly and only later notify the labor bu- reau. However, the greater degree of choice has increased frictional un- employment among school leavers (World Bank 1992b). Managing the transition in labor markets and relocating workers displaced by the merger or, in rarer instances, the closure of enterprises pose serious prob- lems (middle-age female textile workers in Shanghai and Tianjin are especially hard to place in new jobs). They constitute one of the major challenges facing each city. This recognition has motivated a range of reforms intended to make the labor market more flexible. The purpose of the labor market reforms, introduced in stages since 1986, has been to improve labor productivity and industrial efficiency by gradually dislodging permanent employment and state allocation of jobs. The introduction of labor contracting has been accompanied by somewhat greater scope for employee discharge, usually for disciplin- ary reasons, and by "waiting-for-employment" or unemployment in- surance. Second, although fixed-term contracts should increase labor mobility, they may be insufficient to address the larger goal of enhanc- ing labor market flexibility in the three cities. Chinese enterprise man- agers indicate that the hiring of contract workers has not fundamen- tally altered labor-management relations, despite the large changes, at least on paper, in job security. Even if worker contracts were much more prevalent, the degree of job insecurity desirable from the stand- point of labor productivity is hard to define against the historical background and expectations of permanent tenure. Too much job secu- rity, as well as too little, can damage worker productivity. Therefore, the mere existence of a contract may not be a sufficient incentive to raise productivity. Anecdotal and circumstantial evidence points to the automatic renewal of long-term contract workers, though the extent of this phenomenon is not known (Liu 1989; Davis 1988). Rigid job defini- tions and work rules, which are rarely observed and therefore keep multiplying in a bewildering array, are essentially untouched by the contract system. 164 The Dynamics of Urban Growth in Three Chinese Cities Broadening job assignments and enhancing worker mobility within a given enterprise is another approach to enhancing labor flexibility. Nar- rowly delineated career paths severely restrict the ability of enterprises to respond quickly to changing technologies, production processes, new competition, and shifting demand. Highly circumscribed worker and managerial assignments make sense primarily in mass production sys- tems and where trainability of the workforce is a concern. This is why traditional job security and wage structures are based on narrow work rules in the United States. But the resulting work patterns generate nei- ther the skills nor the incentives to seek flexibility. Generalized skills cannot develop where the adaptation and innovation such skills make possible do not have ready outlets. They develop when there is an exist- ing skill base and when workers have opportunities to upgrade their skills in specific directions that are determined within the enterprise. Labor use in Chinese enterprises is highly compartmentalized; work- ers are trained in a narrow specialization. In this context, the presence of a labor overhang in China's urban sector further complicates efficient labor allocation within and between firms. Authorities in the three cit- ies, though conscious of the rigidities in labor use, are faced with the tradeoff between underemployment and open unemployment. Over the medium term, they have opted to absorb surplus labor in service opera- tions (such as catering, retailing, repair and maintenance works, grounds maintenance, guest house operation, and cleaning services) within the firm. This service employment is being generated either from existing operations, which are now run as stand-alone subunits within the en- terprise and are required to be fiscally independent, or from new activi- ties that are being started especially for this purpose. Industrial restructuring involves the breakup, regrouping, birth, and demise of individual industries and industrial sectors. Almost no part of this process is devoid of implications and consequences for educa- tion and training. The quality and quantity of the initial training of new entrants to the labor force influence turnover costs in enterprises. Recur- rent vocational training, either on the job, within the enterprise, or on leave from the enterprise, has important effects for incentives and technology absorption. The retraining of workers, either to respond to shifts in pro- duction within the enterprise or to facilitate employment after the enter- prise closes, influences internal and external labor mobility and the pace of industrial restructuring. Finally, the retraining of older workers affects de- cisions about retirement and the cost of pension benefits for society. The vast state-sponsored education and training establishment in China is of a socialist pedigree. Whether it is at the level of the central, provincial, or municipal government, of the industrial bureau, or of the enterprise, the sheer number of layers and the volume of educational services imparted are impressive (for details on China's educational and vocational education system, see World Bank 1986, 1987). At one level, with respect to training policies and programs, the three cities must con- Similarities, Contrasts, and Lessons: Three Cities and Others 165 sider whether they simply need to do more of what they are doing and perhaps do it better. At another level, and in the context of industrial restructuring, they need to consider more fundamental questions that do not yet command much attention. These are questions about the in- herent uncertainty of the demand for skills during the process of indus- trial transition, about the potential for mismatching skills unless flex- ibility is built into decisions regarding training, about the relative value of vocational and on-the-job training, and about the incentives for train- ing under different labor market conditions. Several factors will increase the uncertainty about the composition of demand for skills in the three cities: demographic trends, the opening up of China's markets to international competition, the likely narrow- ing of industrial specialization, and technological change. This greater uncertainty will make planning for the requisite skills all the more diffi- cult. Very likely, greater emphasis will have to be placed on devising flexible training arrangements and on-the-job training, with the respon- sibility for this devolving more on enterprises than on planning agen- cies, and on determining and anticipating the required mix of skills.8 Urban Prospects: International Experience This section paints a broad canvas but examines it from a particular angle, thereby filling out and extending the earlier discussion. It deals with the economies of cities worldwide but is concerned with those few charac- teristics that must economically explain success, failure, or resurgence. It uses a few cities as examples of these qualities and, from their experi- ence, extracts a handful of lessons germane to the three Chinese cities. These refer to cities in ascendance, troubled cities able to sustain a level of performance, and cities attempting comebacks. Lesson I The clustering of industries in and around a particular city often occurs simply through a historical accident, thereby creating new urban cen- ters (Krugman 1990; Scott and Angel 1991). Examples are the carpet in- dustry in Dalton, Georgia, the shoe industry in Massachusetts, and the tire industry in Akron, Ohio. These cases are all drawn from the United States, but similar ones abound in Europe. Aggressive entrepreneurship by individual communities, states, or provinces seeking to woo indus- tries has also influenced the geographic distribution of industry. Local politicians have frequently been decisive in steering industry to their city by tirelessly extolling its virtues and by using a variety of instru- ments; tax relief, quality of infrastructure, and creation of science parks are among the most common. The science park adjacent to Stanford University-the genesis of Silicon Valley-is now famous (Scott and An- gel 1991). The first company to locate there was Fairchild Semiconduc- 166 The Dynamics of Urban Growth in Three Chinese Cities tor, which subsequently spawned almost 50 independent spin-off firms. These became the core of a vertically desegregated production complex, which includes every aspect of microchip production from basic research, production, and testing equipment through bonding material, silicon wafers, and photomask, all the way to metal plating deposition and etch- ing services. The agglomeration economies from such clustering have been very significant. North Carolina's Research Triangle, which is a highly successful research park, is built on the research and institutional foundations provided by the universities of Raleigh, Durham, and Chapel Hill, the standard-bearers of the second most lavishly funded state univer- sity system in the United States. The growth of an industry producing packaging and boxing machines around Bologna, Italy, followed a similar pattem. The first two or three factories were set up between 1920 and 1950. These became the source of technical and entrepreneurial skills that have spawned nearly 300 small and medium-size specialized firms occupying a variety of niches within the packaging industry (Capecchi 1989). Across the Pacific, Tsukuba Science City, near Tokyo, represents the Japanese government's attempt to establish a major center for research that would provide a nucleus for high-technology manufacturing. Ap- proved by the cabinet in 1963, construction was begun in the early 1970s, and a substantial research infrastructure was in place by the late 1980s at great public cost. Tsukuba has certainly emerged as a hub of scientific research composed largely of public institutions. However, the verti- cally integrated structure of laboratories and the isolation of city from industry-both factors present in China-have limited the spillover ben- efits. Tsukuba might be finding its stride in the mid-1990s, but the gesta- tion period was exceedingly prolonged and the return on public invest- ment questionable (Castells and Hall 1994). Another arguably more successful venture initiated by the Japanese government is located in Kyushu. Politicians closely connected with powerful central government ministries made rural Kyushu into Japan's Silicon Island (Calder 1988). Governor Hiramatsu Morihiku, formerly a key figure in the Ministry of International Trade and Industry, set in motion forces that contributed to industrial success: he mobilized local support and funding and persuaded the central government to defray the start-up costs of high-technology industrialization by financing in- frastructure and much of the cost of retraining the local workforce. Once the government had been won over, its own championing of Kyushu prompted businesses to act. Finally, the need for human capital was met through the creation of a network of training institutes, universities, and research centers, such as the Kyushu Institute of Technology and the Fuzzy Logic System Institute, both in the former coal mining town of lisuka.9 Two of the newly industrializing economies have also set up their own versions. Some 90 kilometers from Taipei, Hsinchu Industrial Park stands next to two major universities and the government's Industrial Similarities, Contrasts, and Lessons: Three Cities and Others 167 Technology Research Institute. And Korea's Daedok Science Town, 160 kilometers from Seoul, is home to 55 research institutes, whose findings are enticing industry to the area. Lesson 2 It is becoming increasingly evident that reputable universities with ex- tensive science and engineering programs are necessary for industrial health, especially with regard to technologically advanced activities. However, synergistic interaction with important government or corpo- rate research establishments provides an additional spur. The aircraft and electronics industries around Los Angeles and along the M4 auto route near London could hardly have arisen without the research infra- structure built up by government funding. A significant part of this fund- ing was earmarked for defense purposes but generated substantial spillovers into civilian sectors, primarily through the movement of per- sonnel (see Block 1988; Castells and Hall 1994). Thus, in several notable instances, government funding has determined the industrial future of cities not only through the money channeled into physical infrastruc- ture but also through the sums poured into research, whether in univer- sities or in public sector establishments. Even where the direct transfer of technology from defense-related research has been meager, civilian industry has profited indirectly from the training imparted to a vast body of professionals, the experience engendered, and the pure knowledge propagated. The potency of these externalities is evident all along Route 128 out of Boston. Successful cities have capitalized on these trends in production, in organization, and in services. To stay ahead, they have also achieved workable tradeoffs in two other areas: public finances and regulation. Cities compete with one another for industry by exempting companies from certain local taxes. However, if nontax revenues do not offset a narrower tax base, infrastructure and services-important attractions for industry-suffer. A balance has to be struck, and cities that offer the winning combination are unusually efficient with their expenditures and ingenious in persuading the private sector to fund a variety of capital projects. Regulations can contribute to a city's livability, for instance, by sustaining high environmental standards, as in parts of California. They also can saddle companies with expensive red tape. In market econo- mies where capital is mobile, institutional arbitrage ensures, within lim- its, that industry moves to those cities where the regulatory framework is most advantageous. Such competition is an important check on the scope of regulation and discourages the unwanted interference of local authorities (see Siebert 1991). Lessons 1 and 2 are extremely relevant for the three Chinese cities. They underscore the importance of good urban infrastructure and envi- ronmental conditions. They draw attention to the role of premier uni- 168 The Dynamics of Urban Growth in Three Chinese Cities versities. And they suggest that local as well as central governments should at least prime the pump by adequately funding research facili- ties and encouraging the commercialization of research. Both lessons also highlight the significance of increasing private sector involvement in supporting research and tertiary education, in stimulating interac- tion with researchers, and in promoting entrepreneurship that builds high-technology industry. Municipal authorities in Shanghai could pour money into public research institutes, but spillovers into industry re- quire initiative from the business sector. Lesson 3 A city is most vulnerable when its economy is dominated by a single firm with a relatively narrow specialization. Should the company's busi- ness decline because of a secular trend in demand or falling competi- tiveness, this very quickly affects the urban economy. There are many examples of this in western Europe and North America, most conspicu- ously with respect to cities dependent on producers of basic metals, ships, transport equipment, and textiles. For example, centers of steel produc- tion all across northern Europe and Japan have had to contend with a steady ebbing of business activity since the mid-1970s (see Houseman 1991). The practice followed by German steel manufacturers in the nine- teenth and early twentieth centuries of purchasing large tracts of land and controlling local politics so as to prevent other industries from chal- lenging their monopoly over local labor markets became a severe handi- cap for the Saar region when the retreat of steel increased the demand for alternative employment. Old industries such as steel, textiles, and shipbuilding, which have a long history and have employed successive generations of workers in a particular location, exercise a profound influence. The labor force be- comes inbred and wedded to the company through multiple points of contact. Social networks and local traditions breed intense loyalties, which discourage workers from seeking jobs farther afield. The need to support increasing numbers of unemployed is an additional drain on the economy, which compounds the effects of soft demand for the prin- cipal manufactured product.'0 Cities where a single firm is the main economic axis represent an ex- treme case. More common is the concentration of multiple producers, all operating within a narrow industrial segment, such as the hard-hit centers of the textile industry in the British Midlands. Battered by for- eign competition, a once-proud manufacturing establishment has now been reduced to a small handful of vertically integrated corporations occupying some of the more capital-intensive segments. The eclipse of textiles was a blow for cities whose economic life revolved around this industry. It was only a little less serious for areas in which textiles shared the spotlight with another major industry, such as machine tools. Similarities, Contrasts, and Lessons: Three Cities and Others 169 The disquieting aspect of this experience is that in many instances the failure of one industry has not been counterbalanced by the rise of an- other. Furthermore, even when an industry, for example steel, has been effectively downscaled, the prolonged orchestration of local policies and substantial grants from central governments are required to salvage the affected urban economies (Rodwin 1991). Hill and Fujita (1993) describe how Japanese steel companies have ameliorated the effect of closure in Kamaishi and Kitakyushu without being able to offset the drastic re- duction in jobs in the local communities. Financing from the central gov- ernment has eased the pain of severance and unemployment, but re- gional policies have rarely succeeded in reversing the decline. Shanghai and Tianjin have already lost jobs in the textile and light consumer in- dustries. Restructuring of state-owned enterprises and competition from elsewhere in China will undoubtedly force some other subsectors to shrink. The experience of other countries shows that resisting such pres- sure is costly. Instead, local authorities should work with the enterprises to cushion the shock and relocate displaced workers to other enterprises. Thus the emphasis must be on encouraging industrial diversity and putting in place labor market mechanisms that enable displaced work- ers to find alternative employment. Lesson 4 Old and troubled industrial cities are populated by firms that survive by cutting costs and postponing capital expenditure. Most of them are resigned to being technological laggards and are disinclined to open or exploit new niches aggressively. By and large, the labor force is com- posed of older or female semiskilled workers whose geographic and occupational mobility is often limited. Because of unionization and con- ditions peculiar to local labor markets, average wages are often high in relation to the level of skills. Furthermore, the capacity to augment skills either within companies or through urban training institutions might be poorly developed. Aging infrastructure can be another drawback. Whether in education, telecommunications, housing, or recreation, older cities frequently lag behind. All this plus an unpleasant environment discourage the inflow of capital, entrepreneurship, scientific talent, and young workers. These conditions are not by any means irreversible, but once the com- petitiveness of the foremost industrial players begins to wane, their dif- ficulties infect the finances of the city and shrink the fiscal base. This in turn adversely affects the supply of services and investment in social overhead, unless the city is in a position to tap other sources of funds. Ultimately, the factor preventing cities from staging a comeback is the paucity of capital with which to invest in infrastructure and clean the environment. The early stages of decline are caused by the erosion of industrial capability. In today's world, this is principally a technological 170 The Dynamics of Urban Growth in Three Chinese Cities failing, with the term being interpreted broadly to cover not just prod- uct and process but also the organizational innovations conducive to efficiency. The future of steel and machine tools is as closely tied to tech- nological increments as is the future of electronics.'" Both industries, es- pecially machine tools, need to forge links with producers in neighbor- ing subsectors so that a concerted effort can yield product diversity as well as productivity gains. Lesson 5 Some cities have managed to restructure their industrial bases and to sustain steady growth. This is most evident in the newly industrializing economies, and, in most cases, the outcome has been encouraging. Take the example of Hong Kong. Its early success as a center of low-end manu- facturing exports is quite unusual in that it chose to specialize in only a few stages of the production process. Because of the small size of the capital goods sector, which is limited by the dearth of land and natural resources, and the government's noninterventionist industrial policy, Hong Kong lagged behind other newly industrializing economies in tech- nological innovations and R&D capacity. By the late 1970s and early 1980s, the price competition from some lower-cost countries in the region was seriously challenging Hong Kong's advantage. With the opening up of China, the city restored the competitive advantage of its labor-intensive light manufacturing by relocating production facilities to the mainland. By the early 1990s, between 80 and 90 percent of Hong Kong's electron- ics establishments had moved across the Shenzhen River. Hong Kong's process of "deindustrialization" is often regarded as positive because it facilitated the expansion of financing and trading activity, which in turn sustained full employment. The share of the in- dustrial sector in employment (including manufacturing and construc- tion) fell from 50 percent in 1980 to a mere 36 percent in 1991, while that of the service sector rose from 48 to 63 percent. Every branch of the ser- vice sector, including financial services, trading services, transportation, and communication, experienced steady growth in employment (Ho and Kueh 1993; Ash and Kueh 1993). Real wage growth in the service sector was also much higher than in the manufacturing sector. Hong Kong now has perhaps the most service-oriented economy in the world. Lesson 6 The city-state of Singapore took a related but contrasting route. In the 1950s and 1960s Singapore relied on its cheap labor to produce for export markets, particularly in the electronics industry. Although knowledge-based industries such as electronics require a large contin- gent of technically qualified personnel to design the product, integrate Similarities, Contrasts, and Lessons: Three Cities and Others 171 production equipment, and maintain and supervise the production pro- cess, the bulk of the workforce is composed of unskilled and semiskilled labor. The elastic supply of such labor has buttressed manufacturing capability in and around the science parks. It also explains the rise of the electronics industry in Hong Kong and Singapore, cities not especially well endowed with scientific manpower but with an abundance of semi- skilled workers and factory-level technicians. As its labor advantage diminished in the 1970s, the Singapore gov- ernment took drastic measures to restructure the economy. One impor- tant element was the technology policy, which emphasized basic education for the entire population, strong fiscal incentives for foreign investment in R&D, disincentives for low-end processing, reduction of capital costs of foreign investment, and organization of local research institutions.'2 These policies enabled Singapore to hold on to the large multinationals in the electronics industry and to become a center of high- technology production. Singapore is now the world's largest exporter of computer hard-disk drives, with a 60 percent share, and is the fifth or sixth largest exporter of computer products overall. The technology policy, which pushed Singapore up the learning curve, has successfully maintained the city-state's competitiveness and its attractions for multi- national corporations. Another element of its success consists of regional cooperation, with its hinterlands, in industrial development. Such efforts include a sectoral and technical division across Singapore, Johor in Malaysia, and the Riau Islands in Indonesia (K. C. Ho 1994; Lim 1993). Land and labor con- straints require the low value added, labor-intensive stages of the pro- duction process to be subcontracted or relocated to Johor or Riau. Singapore also serves as a regional headquarters for multinationals with production facilities in the region, offering transport and communication infrastructure, a range of financial services, and a good urban environ- ment. Singapore's approach finds an echo in the strategy of Korea's capital city, Seoul, which is being challenged by new industrial cities such as Ansan, Kumi, Pusan, and Ulsan, despite its historical dominance in a number of trade-oriented industries (Park and Markusen 1995). Even within the capital region, Seoul was losing to the suburban areas, as demonstrated by a relatively large reduction in industrial jobs in the former and gain in the latter (Park 1993). To meet the challenge, Seoul in the late 1970s launched a concerted effort to attract high-technology in- dustries. Within 10 years, four-fifths of all such activities were located in the vicinity of Seoul (as against 58 percent of total manufacturing) in spite of a government-sponsored effort to disperse R&D centers through- out the country. Side by side with the spread of high-technology manu- facturing, the rapid growth of consulting, technical, and computer ser- vices sustained Seoul's growth. 172 The Dynamics of Urban Growth in Three Chinese Cities Lesson 7 Urban-industrial decay is by no means a one-way street. City govern- ments have been able to work with corporate management to revive old industries and attract new ones. In most cases, it has been a multipronged effort, but there are instances where a city has retained a modicum of dy- namism by building on its base as an international supplier of services. Earlier in this century, New York's economy was dependent on the manufacture of clothing, food processing, the printing industry, and its role as a transport hub (Glazer 1991). These activities financed the de- velopment of an efficient infrastructure, the growth of urban services, and the founding of a university system that remains one of the best in the country. Between 1950 and 1994, New York lost over 700,000 manu- facturing jobs (Moss 1997). By the 1960s the city's economic focus had shifted to services, and it slowly consolidated its claim to being the pre- mier financial center. Vestiges of past manufacturing activities remain, most notably in fashion garments, but for the past three decades, New York has derived its prosperity from services-financial, professional, tourism-related, and cultural-supported, in turn, by communications facilities of high quality. Between 1979 and 1989 the city also benefited from the surge in defense spending, both directly and as a result of height- ened industrial activity in Connecticut (Netzer 1997). New York has been able to live off services for several reasons (Piore and Sable 1983). First, it stands at the apex in the world of finance (albeit with increasing difficulty). As a result, the city was a logical site for cor- porate headquarters that, in turn, have pulled in a range of legal, consultancy, and real estate services."3 Second, New York is well endowed with skills. Its universities not only nurture local talent but also draw in some of the best minds from around the world. Third, the city's cultural resources are exceedingly rich, keeping alive its attractiveness in spite of endemic urban blight. Fourth, its principal competitors in the 1990s are equally high-cost cities in Europe and East Asia, some of which lack New York's other attributes. The experience of cities as diverse as New York, Baltimore, London, Amsterdam, and Dortmund suggests that sustaining dynamism or find- ing a second wind are critically dependent on three factors. First, the political leaders and opinion-makers of the city must crystallize and endorse a vision of the future that is acceptable to businessmen and com- mon taxpayers alike. For such a consensus to form, the tenor and par- ticipatory nature of local politics and trustful relations between key po- litical figures are important assets (Richardson 1991). Second, public and private sectors must collaborate so as to spread the burden of cost and risk. By drawing in the business community, a city can be sure that the money spent to attract industry is going to the right places. Industrial- ists have quite specific concerns that might be related to labor or physi- cal infrastructure or quality of the environment, as indicated by the ex- Similarities, Contrasts, and Lessons: Three Cities and Others 173 perience of Pittsburgh, Rochester, Research Triangle in North Carolina, Utah Research Park, and Silicon Valley. Access to university faculty, the ease of recruiting professionals, the degree to which the university is responsible for administering the research park, and the provision of amenities appear to be the critical factors. The more directly these con- cerns are addressed, the greater is the impact. Third, the actual implementation of projects has in some notable instances been most efficacious when it was made the responsibility of quasi-public bodies (de Jong 1991). These can join public interest with private initiative. Quangos, as they are called, can rely more fully on the economic calculus to neutralize political concerns that dilute the effects of public schemes. The justly acclaimed Inner Harbor development in Baltimore is an example of deft political inception and efficient realization by a quasi-public body, which put forward a project credible enough to draw federal funding. Lesson 8 The above lessons for three Chinese cities are based mainly on the expe- rience of large cities in the industrialized economies. However, Shang- hai, Tianjin, and Guangzhou also have much to learn from major urban centers in lower-income and newly industrializing countries. Cities such as Bangkok, Bombay, Karachi, Manila, and Seoul do not have to cope with decline or recovery. They are struggling with explosive population growth that is placing immense burdens on an infrastructure unable to keep up with demand. In many of these cities, municipal politics are murky, administration is weak, crime is rampant, and local revenues fall far short of requirements (Asian Development Bank 1997). Although a city such as Seoul has enjoyed respectable growth, led first by industry and more recently by a combination of industry and producer services, other large cities in lower-income countries have turned in a less robust economic performance. This weakness stems from the inability to increase rapidly the number of relatively well paid jobs in manufacturing and producer services. One constraint is the quality of the infrastructure (energy and water supplies as well as transporta- tion and housing services). Another is the transaction costs of doing business. There is too much red tape, taxes and fees are often excessive and unpredictably chaotic, local officials are extortionate, and criminal mafias thrive largely unchecked by the authorities. A third constraint that derives from the first two is that the quality of the physical and institutional environment deters FDI, negates agglomeration economies, and interferes with the functioning of dynamic industrial networks that strive to compete and give due attention to technological gains. The Bangkoks and the Bombays are expanding cities, but their growth is seemingly involuntary, propelled more by population than by a strat- egy that seeks to promote a mutually reinforcing mix of industry and services under environmentally sustainable conditions. 174 The Dynamics of Urban Growth in Three Chinese Cities The lesson here relates to the quality of municipal leadership, adminis- tration, development strategy, and governance. Megacities in developing countries can thrive only if they retain the initiative and have clear strate- gic goals and the political determination to implement them. Cities that allow themselves to be engulfed by myriad problems and then struggle to survive have a bleak, crowded, possibly unsustainable future. Summary An important difference between the three Chinese cities and declining cities in developed market economies is their degree of industrial spe- cialization. As noted above, cities in market economies are particularly vulnerable when they depend on a single manufacturing sector. By con- trast, the three cities possess extremely diversified manufacturing sec- tors. Although this has hindered specialization and the focused use of resources, it protects the municipal economies from the decline of an individual subsector and allows them to explore different industrial options. Thus, the task of industrial restructuring takes on a different slant. Rather than trying to move out of a single declining industry, the three cities might allow a range of noncompetitive industries to shrink in relative size. The difficult but nonetheless crucial task is to identify the growth industries and producer services that will serve as the basis of re- structuring and will absorb labor and capital from declining industries. The experience of reviving cities in industrial countries shows that no industrial sector is intrinsically obsolete, even at high wage levels. Some textile and garment producers flourish in Germany, Italy, Japan, and the United States. Thus, it is impossible for planners to determine the shape of future restructuring from first principles, even when those principles are firmly grounded in economic logic. It is impossible to say that Shanghai's textile industry will not survive and prosper simply because Shanghai's wage and skill levels are higher than those in the interior of China. Nor can it be said that coastal regions will necessarily succeed in a particular high-technology sector because they share the same factor endowments. Municipal planners succeed best when they work in con- cert with markets. Urban modernization, or revival, takes place when city planners pro- vide effective support for industrial restructuring by developing mu- nicipal infrastructure and improving labor and capital markets, while allowing market demand to pick the winners and losers in the process. This is particularly important in the coastal cities. Some declining in- dustries will effectively "downsize" into competitive and flexible "niche" producers. Some will disappear completely. And some apparently de- clining industries will surprise everyone and turn into engines of dy- namic growth for the future urban economy. The objective of the three cities should be to allow these surprising trends to emerge without ob- struction. A Map of7 the Future Aristotle believed that the ideal city contained just 5,000 people and that a settlement with more than 100,000 inhabitants could no longer be con- sidered a city. Timies and opinions have changed. In 1995 the largest city in the world-Tokyo--had a population of 27 million, and 21 megacities had populations in excess of 8 million. There were only two megacities in 1950. Urbanization has slowed in industrial countries to 1 percent a year or less but is projected to continue at rates of 3.5 percent and more in developing countries. It is fastest in Africa and Asia, where urbaniza- tion ratios are between 30 and 35 percent. By 2015 there are likely to be 33 megacities, 27 of them in the developing world. In addition, the num- ber of cities with populations ranging from 1 million to 10 million will have risen from 270 in 1990 to 516 in 2015 (World Resources Institute 1996). Recent trends suggest that China's rates of urbanization will be among the highest in the world, possibly in the range of 4 percent a year, and that between one-half and two-thirds of China's populace will be living in towns and cities within a couple of decades (see Kojima 1995). Already in 1995, both Shanghai and Tianjin were classified as megacities, Shanghai occupying sixth place and Tianjin occupying thir- teenth. Guangzhou will officially catch up in the near future, although in reality it may already be a megacity. Thus one of the biggest chal- lenges for the three cities will be to absorb the flood of migrants with- out experiencing a paralysis of services and a sharp deterioration in living conditions. But each city can only do so much in the face of relentless migration. Only a nationally coordinated strategy for urban development can en- sure a degree of regional balance and ease the pressure on the main 175 176 The Dynamics of Urban Growth in Three Chinese Cities coastal centers. For this reason as well as many others, China will need to have a strong central government and efficient mechanisms for fiscal sharing so that poorer but densely populated interior provinces benefit from progress elsewhere (on the range of problems to be solved in the interests of sustainable development and the necessity for local-central cooperation, see Drakaki