OGob 1156 CHIS Long-Terin Development Issues and Options A WORLD BANK COUNTRY ECONOMIC REPORT FILE,XC . rI N I F I" U I m C:HINA AWorld Bank Country Economic Report This report is supplemented by six annex volumes published by the World Bank: 1 China: Issues and Prospects in Education 2 China: Agriculture to the Year 2000 3 China: The Energy Sector 4 China: Economic Model and Projections 5 China: Economic Structure in International Perspective 6 China: The Transport Sector In addition, nine background papers, numbered here as they are referred to in the text, have been prepared in connection with the report. They are available as World Bank staff working papers. I The Asian Experience in Rural Nonagricultural Development and Its Relevance for China (Staff Working Paper 7 57) 2 International Experience in Urbanization and Its Relevance for China (Staff Working Paper 758) 3 Alternative International Economic Strategies and Their Relevance for China (Staff Working Paper 759) 4 International Experience in Budgetary Trends during Economic Development and Their Relevance for China (Staff Working Paper 760) 5 Productivity Growth and Technological Change in Chinese Industry (Staff Working Paper 761) 6 Issues in the Technological Development of China's Electronics Sector (Staff Working Paper 762) 7 The Environment for Technological Change in Centrally Planned Economies (Staff Working Paper 718) 8 Managing Technological Development: Lessons from the Newly Industrializing Countries (Staff Working Paper 717) 9 Growth and Structural Change in Large Low-Income Countries (Staff Working Paper 763) These supplementary works may be ordered, by annex or working paper number, from local distributors of World Bank publications and from the World Bank Publications Sales Unit, 1818 H Street, N.W, Washington, D.C. 20433, US.A. > CH[INA Long-Term Development Issues and Options Tbe report of a mission sent to China by The World Bank Edwin Lim Adrian Wood Cbief of mission Deputy chief of mission Contributing autboars Ian Porter Robert P. Taylor William Byrd Gene Tidrick Timothy King Wouter Tims Gerhard Pohl Publisbedfor The World Bank The Johns Hopkins University Press Baltimore and London Copyright ' 1985 by The International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W, Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First hardcover and paperback printing October 1985 The Johns Hopkins University Press Baltimore, Maryland 21218, U.S.A. The World Bank does not accept responsibility for the views expressed herein, which are those of the authors and should not be attributed to the World Bank or to its affiliated organizations. The designations employed, the presentation of material, and any maps used in this document are solely for the convenience of the reader and do not imply the expression of any opinion whatsoever on the part of the World Bank or its affiliates concerning the legal status of any country, territory, city, area, or of its authorities, or concerning the delimitation of its boundaries or national affiliation. Library of Congress Cataloging in Publication Data Main entry under title: China, long-term development issues and options. "A World Bank country economic report." 1. China-Economic policy-1976- . 2. China- Social policy. l. Lim, Edwin. II. International Bank for Reconstruction and Development. HC427.92.C46452 1985 338.51 85-22777 ISBN 0-8018-3313-2 ISBN 0-8018-3314-0(pbk.) Cover photograph by Curt S. Carnemark: Family outing at Tiananmen Square, Beijing Foreword The idea for this study originated during my visit to China in 19 8 3. In Beij- ing, I had the opportunity to meet with Chairman Deng Xiaoping and Pre- mier Zhao Ziyang. Through (discussion with them and with other distin- guished Chinese leaders, I learned a great deal about the Government's perspective on development ond the long-term goals for the country. We talked about the national aim to raise China's annual per capita income to US$800 by the year 2000, and we contemplated some of the exciting possi- bilities and complex questions that China would encounter in the process of reform and modernization. Out of these discussions in Beijing came the agree- ment that the World Bank would undertake a study of some of the key devel- opment issues that China might face in the next twenty years. In particular, this study would examine, in light of international experience, some of the op- tions for addressing these issues. This report summarizes the conclusions of that study. It is based primarily on the findings of an economic mission led by Edwin Lim and Adrian Wood, which visited China in early 1984. It was presented to the Board of Exccutive Directors, which represents the 148 member governments of the World Bank, in May 1985. This report, in my view, is a good example of one of the many ways that the World Bank can assist its member countries. Indeed, China's ambitious at- tempt to modernize and reform its economy will shape the future of not only the Chinese people but also people throughout the world. We hope that this report will be a useful contribution to the debate in China about some of the important issues of economic and social development. A. W. Clausen President The World Bank September 6, 1985 I Contents SUMMARY AND CONCLUSIONS . ...................................................... I Pattern of Growth .................................... ... ................... I Managing the Economy ............................................................ 8 Social Issues and Policies ............................................................ 16 Overview .... . .......................1............9.......... ... 19 1 GROWTH AND CHANGE ...................................... 21 Objectives ...................................................................... 21 Macroeconomic Chang ............................................................ 22 Microeconomic Change ............................. ............................... 26 Inequality .................................................................. .... 29 2 ILLUSTRATIVE PROJECTIONS .................. .................................. 32 Description of the Model ......................... .................................. 3 2 Three Alternative Projections ..................... ..... 3 3 Growth and Demand .............................................................. 3 6 Production and Investment .................................... 3 8 Employment and Income .............................. ........................... 40 Projections and Predictions .................................... 43 3 AGRICULTURAL PROSPECTS AND POLICIES ................... ............... 44 Production Possibilities ............................................................. 44 Potential Demand-Supply Imbalances .................................................. 45 Changing Food Demand ............................................................ 50 Managing Household Agriculture ..................................................... 5 3 Agricultural Incomes Policy .......................................................... 5 5 4 ENERGY DEVELOPMENT ....................................................... 58 Demand and Supply: Trends and Balances ............................................... 58 Coal Production, Transport, and Utilization ............................................. 60 Oil and Gas . .................................................................... 65 Electric Power . .................................................................... 67 Rural Energy and Biomass ........................................................... 68 Investment, Planning, and Prices ............................................... ...... 69 5 SPATIAL ISSUES . ................................................................... 7 3 Specialization and Trade ........................................... ................ 7 3 Transport, Commerce, and Other Services ............................................ 82 Urban and Rural Development ....................................................... 86 Rural Poverty and Regional Inequality .........9..................................... 91 vii 6 INTERNATIONAL ECONOMIC STRATEGY .......................................... 97 China in the World Economy ........................................................ 97 Foreign Trade Prospects and Policies .............................................. .... 101 External Balance and Finance ..................................... 107 Regional Issues .................. ................................................. 109 7 MANAGING INDUSTRIAL TECHNOLOGY .......................................... 110 Technology in Perspective ................................................... 110 Planning for Technological Development ................................................ 112 The Make or Buy Decision ....... ..................... ... ................. 116 Incentives for Innovation and Diffusion .... ............................................. 118 Obsolescence and Scrapping .. .................................. 121 8 HUMAN DEVELOPMENT .......................................................... 123 Education and Training Issues .. ...................................................... 123 Employment and Productivity ..................... ................... ... 127 Allocating and Motivating Workers ............................. ...... ...... 130 Population Issues beyond the Year 2000 . ............................................... 136 9 MOBILIZING FINANCIAL RESOURCES .............................................. 143 Alternative Sources of Saving ....................................................... 143 Subsidies and Transfers .......................................... .. ............ - - 148 Revenue Needs and Sources .................. 152 Tax Reform ......1........................... 156 10 DEVELOPMENT MANAGEMENT ............................ ............. 160 The State and the Individual ......................................................... 160 The State and Enterprises ........................................................... 164 Investment and Price Reform ........................................................ 171 Economic Planning ................................................................ 177 Overview ............................................................ 181 Tables 1.1 Size Distribution of Industrial Enterprises in Selected Countries ............................... 28 1.2 Rural Income Distribution, 1979-82 .................................,,,,... 29 2.1 Average Annual Growth of National Income, Alternative Projections, 1981-2000 ..... 36 2.2 Composition of Final Demand, Alternative Projections, 1981-2000 .... 36 2. 3 Average Annual Growth of Sectoral Gross Output, Alternative Projections, 1981-2000 ... 38 2.4 Composition of Capital Stock, Alternative Projections, 1981-2000 .39 2.5 Sectoral Composition of the Labor Force, Alternative Projections, 1981-2000 .................... 40 2.6 Sectoral Composition of National Income, Alternative Projections, 1981-2000 ................... 42 4. 1 Energy Supply and Demand, Alternative Projections, 1981-2000 . ............................ . 60 4.2 Share of Solid Fuels in Commercial Energy Use in Selected Countries, 1960, 1980, and 2000 .. . 62 4.3 Estimated Investment Requirements in the Energy Sector, 1986-2000 .......................... 70 5.1 Dispersion of Production Activities among and within Provinces, 1982 .......................... 73 5.2 Gross Output and Profits of State-Owned Industry, by Province and Subsector, 1982 .74 5.3 Industrial Concentration, by Region and Key City, 1982 .................................... 74 5.4 Extent and Composition of Rural Nonagricultural Activities in Selected Countries ................. 79 5.5 Urbanization and Economic Development in Selected Countries .......................... . 87 5.6 Per Capita Income in Urban and Rural Areas, 1982 ..... ......................... .. .... 87 viii 5.7 Distribution of Employment and Population, Alternative Projections, 1981-2000 ................. 89 5.8 Average Annual Growth of Urban and Rural Income, Altern ative Projections, 1981-2000 .... ...... 90 5.9 Urban-Rural Income Disparities, Alternative Projections, 1931-2000 ........................... 90 5.10 Rural and Provincial Poverty, 1982 ..................................................... 92 5.11 Rural Development in Dingxi County (Gansu) and Wuxi County (iangsu), 1983 .................. 92 5.12 Wage Levels, by Province and Type of Unit, 1982 ...... .................................. 94 5.13 State Budget Revenues and Expenditures in Wuxi and Dingxi Counties, 1983 .................... 96 6.1 Total Foreign Trade Ratios, Alternative Projections, 1981-2 300 ............................... 102 6.2 Share of Manufactures in Total Merchandise Exports, 1960 and 1981 ........................... 103 6.3 Sectoral Foreign Trade Ratios, Alternative Projections, 1981-2000 ............................. 104 6.4 Average Annual Growth of Foreign Trade, Alternative Projcctions, 1981-2000 ................... 104 6.5 Structure of Foreign Trade, Alternative Projections, 1981-2000 ............................... 105 6.6 Foreign Trade Balance and Debt, Alternative Projections, 1 5'8 1-2000 ........................... 107 7.1 Index of Total Factor Productivity in State-Owned Industry, 1952-82 ......................... 111 7.2 Sources of Successful Innovations in European Firms ........................................ 118 8.1 Educational Attainment of the Population, by Age and Sex, in Selected Countries ............ .... 124 8.2 Population Growth, Alternative Projections, 1980-2 100 .................................... 137 8.3 Population of Working Age, Alternative Projections, 1980- 2100 ............................. 139 8.4 Population Sixty-Five Years Old or More, Alternative Projeztions, 1980-2100 .................... 139 9.1 Gross Domestic Savings in Selected Countries, 1976-80 .................................... 145 9.2 Subsidies, 1981 ..................................................................... 148 9.3 Subsidies and Transfers as Percentages of Government Revenue in Selected Countries ..... ......... 149 9.4 Subsidies and Transfers, Alternative Projections, 1981-2000 .................................. 151 9.5 Government Expenditures, Alternative Projections, 1981-2000 ............................... 153 9.6 Illustrative Projections for Figure 9.2 .................................................... 155 9.7 Illustrative Projections for Figure 9.3 .................................................... 157 Figures 1.1 Composition of Aggregate Demand in China and Other Developing Countries, 1981 ..... .......... 22 1.2 Structure of Production in China and Other Developing Countries, 1981 ........................ 2 3 1.3 Structure of Employment in China and Other Developing Countries, 1981 ...................... 24 1.4 Enrollment Ratios in Formal Education in China, 1965-83 and Targets for 2000, and in Other Developing Countries, 1965-80 ....................... ............................. 31 2.1 Share of Agriculture in Total Employment in China, Alternative Projections for 1981-2000, and in Other Selected Countries, 1960-82 .............................................. 41 3.1 Crop Exports and Imports, Alternative Projections, 1981-2000 ............................... 48 3.2 Per Capita Food Intake in China, 1981-2000, and in Other Selected Economies, 1960-80 .... ...... 51 3.3 Share of Animal Products in Total Food Intake in China, 1980-2000, and in Other Selected Economies, 1960-80 .............................................................. 52 5.1 Variations in Per Capita Gross Agricultural and Industrial Output, 1982 .. . ....................... 76 5.2 Distribution of Employment, Capital, and Gross Output, Alternative Projections, 1981-2000 .... .... 88 6.1 Exports of Goods and Nonfactor Services in Selected Countries, 1978-82 ....................... 98 6.2 Trade in Goods and Nonfactor Services, 1978-83 ................... ...................... 102 6.3 Direction of External Trade, 1983 ...................................................... 105 6.4 External Balance, 1978-84 .. . .......... .......... 107 7.1 Output and Cost of Production ofthejA 1-1 Sewing Machine, by Enterprise, 1980 ................ I11 8.1 Population, Alternative Projections, 1980-2100 .................... ...................... 137 9.1 Sources and Uses of Funds, 1981 ...... ............. ...... 144 9.2 Income Flows and Taxation, Alternative Projections, 1981-2000 ............................ 154 9.3 Sources and Composition of Saving, Alternative Projections, 1981-2000 ........................ 156 ix Boxes 1.1 Size and Growth of Enterprises in Market Economies . ....................................... 27 2.1 Wages, Profits, and Prices ................................................ 42 3.1 World Grain Trade, Past and Future ................................................ S0 4.1 Cogeneration for Industry and District Heating . ........................................... 6 3 5. I The Spatial Distribution of Dairy Production ............................................. 77 5.2 Locational Issues in the Iron and Steel Industry . ........................................... 80 5.3 Locational Issues in the Bicycle Industry ................................................ 81 5.4 Trade, Pricing, and Transport of Cement ................................................ 84 5.5 Agriculture on the Loess Plateau ....................................................... 93 5.6 The Benefits and Costs of Migration to Communities of Origin ......... ...................... 94 6.1 Contrasting Types of International Economic Strategy .............. ........................ 99 7.1 Continuous Casting of Steel . ................................................. 113 7.2 Strategic Options in Power Station Boiler Technology . ....................................... 114 7.3 A Successful Strategy in Electronics ............................................ .... 115 8.1 The Second Health Care Revolution ................................................ 140 10.1 Dispersed Social Ownership in the Federal Republic of Germany ........ ...................... 167 10.2 Monetary Control . ..................................................... 174 Maps 5.1 Per Capita Net Material Product, by Province, 1982 ....................................... 75 5.2 Provincial Budget Expenditures and Revenues, 1982 ........................................ 95 Relief and Administrative Divisions ..................................................... 18 3 % Preface This report was prepared principally by members of an tant), and Fred Bentley (consultant on arid agriculture). economic mission that visited China twice in 1984, for * Energy. Roberto Bentjerodt (senior economist, coal four weeks in February-March and for five weeks in projects), Weigong Cao (power engineer), Abdel El- April-May. In addition to visiting Beijing, the mission Mekkawy (engineer, petroleum projects), Robert Taylor went to three provinces: one coastal and relatively high- (energy economist), and Darrel Fallen-Bailey (consultant); income (Jiangsu); one inland and average-income (Hu- D. C. Rao (assistant director, Energy Department) led the bei); and one interior and low-income (Gansu). It re- team in the field. ceived a lot of information, as well as numerous valuable * 1I4dustrial technology. Gene Tidrick (team leader), comments and suggestions, from officials and others in Anupam Khanna (industrial economist), Reza Amin (in- these provinces, as well as from those in many central dustrial specialist), and Josephine Woo (research assis- agencies and institutions, including: the State Planning tant). and State Economic Commissions; the Ministries of Fi- * Location and trade. Ian Porter (team leader), Vernon nance, Agriculture, Coal, Communications, Education, Henderson (consultant on urbanization), John Sheahan Foreign Economic Relations and Trade, Labor and Per- (consultant on industrial location and trade), and Samuel sonnel, Petroleum, Railways, Urban and Rural Construc- Ho (consultant on rural nonfarm activities). tion, and Water Resources and Electric Power; the State The following also contributed to the preparation of Statistical Bureau; and various universities and research the rLport: Wlodzimierz Brus (consultant on socialist institutes of the Chinese Academy of Social Sciences. A economies); Gerhard Pohl (energy and transport); Robert series of seminars was organized by the Technical-Eco- Drysdale (Annex 1); Helena Ribe and Nikhil Desai (An- nomic Research Center under the State Council. The nex 3); Shujiro Urata (Annexes 4 and 5); and Lily Uy generous and thoughtful assistance of all these people in (Annex 6). Larry Westphal, Carl Dahlman, and Bruce China contributed greatly to this report. Ross-Larson organized background work on technology. The Bank mission was led by Edwin Lim (mission Behrouz Guerami-N, Tejaswi Raparla, and Kong-Yam chief) and Adrian Wood (deputy mission chief) and also Tan helped with the multisectoral model, the input-out- consisted of William Byrd (economist), Mats Hukin (se- put table, and data for international comparisons. Ann nior education adviser), Erh-Cheng Hwa (senior econo- Orr, 'Kenneth Hill, Moshe Syrquin, J. V. S. Sarma, Ken- mist), Timothy King (senior economist), Jacques Yenny neth Cochran, Chang Hsin, Liu Ying, and Cai Jinyong (senior transport economist), Umnuay Sae-Hau (research unde took research. Linda Mitchell, Terrice Bassler, and assistant), Betty Ting (interpreter), Luc De Wulf (senior Caro Cole Rosen edited the report; Helen Kung assisted economist, International Monetary Fund), Benjamin in irs processing; and Catherine Ann Kocak prepared King (consultant on statistics), Wouter Tims (consultant graphics for the text. on planning and agriculture); and the following teams. The report also benefited from comments of a review * Agriculture. J. Goering (team leader, April-May), pane] consisting of Anne 0. Krueger, Luis de Azcarate, Tom Wiens (team leader, February-March), Lang-Seng Kemal Dervis, Janos Kornai (consultant), and managers Tay (irrigation specialist), Lo-Chai Chen (fisheries consul- of the East Asia and Pacific Regional Office. xi I I Technical Notes Currency equivalents. The Chinese currency is called renminbi (RMB). It is de- nominated in yuan (Y): I yuan = 10 jiao = 100 fen. In early 1984 the official exchange rate of the yuan to the U.S. dollar was around Y 2 = US$1. The in- ternal settlement rate (ISR) of Y 2.8 = US$1, however, was used in most mer- chandise transactions. The official exchange rate is now about Y 2.8 = US$1. OnJanuary 1, 1985, the Government abolished the ISR. Weigbts and measures. Chinese statistics are usually in metric units; in addition, mu and jin are often used: I mu = 0.1647 acres = 0.0667 hectares; I jin = 0.5 kilograms. Fiscalyear. The fiscal year is January I to December 31. Transliteration. The Pinyin system is used in this report. Terminology. With the recent reorganization of rural administration, "com- munes" have been replaced by "townships," and "production brigades and teams" by "villages." This report retains the former terminology in one respect, however, by referring to "commune and brigade enterprises." The term "national income" is used in this report to encompass both the Chinese measure (net material product) and the Western measure (gross na- tional product). Where the context makes the distinction between these two measures important, they are more precisely identified. Tables. Individual items may not sum exactly to totals because of rounding er- rors. The following symbols are used: . . indicates data are not available; (.) indi- cates a negligible amount; n.a. indicates not applicable. xiii I Summary and Conclusions China's ultimate economic objective is to catch up with highei than in China and will increase in the future by at the developed countries, while maintaining a socialist sys- least 2 percent and perhaps 3 percent per year. To catch tem in which the benefits of prosperity are widely shared. up by 2050, China's per capita income would have to Major, though uneven, progress toward this goal was increase at an average rate of at least 5.5 percent and made in the past three decades. During the next two perhaps 6.5 percent per year. Such rapid progress has decades, there could be substantial further progress. The been tare elsewhere. In 1960-82, excluding small and oil- foundations for rapid and equitable growth in the dominated economies, only two developing countries twenty-first century could also be laid. But this will re- achieved annual per capita income growth above 5 per- quire steering a difficult course, in both development cent: South Korea (6.6 percent) and Greece (5.2 percent). strategy and system reform. More generally, only one country-Japan-has indisput- This report looks at some of the key issues for China in ably caught up with the developed nations from a posi- the next twenty years, and at some of the options for tion of economic backwardness. addressing them, especially in light of international expe- China's past record, on the other hand, is quite encour- rience. It covers many specific topics, including agricul- aging. From 1952 to 1982, despite relatively rapid popu- ture, energy, technology, transport, industrial location, lation growth and periods of acute economic mismanage- internal and external trade, population, education, em- ment, per capita national income grew at an average ployment, and finance (health was addressed in another annua.l rate of 4.0 percent, with phases of significantly recent report'). The first section of this summary assem- faster growth. Of particular relevance is China's perfor- bles some of its conclusions in the context of a more mance during the past few years of policy and system general discussion of the pattern of economic growth. reforrs: from 1979 to 1984, per capita national income Common themes regarding the system of economic man- grew at 6.8 percent per year. agement and planning are pulled together in the second Although the basic objective is to improve living stan- section. A third section covers some related aspects of dards and eliminate poverty, the Government has set social policy. At the end of the summary there is a brief targets of quadrupling the gross value of industrial and overview. agricultural output (GVIAO) between 1980 and 2000 and The most careful forecasts are often confounded. The increasing per capita national income from about $ 300 to lessons of international experience are often ambiguous $800 (about 5 percent per year). If China's future invest- and controversial. In any event, they are hard to apply to ment efficiency were similar to the average of the past China, a country that in important respects differs from three decades, but allowing for increased investment in all others and is not easy for outsiders to understand. The economic and social infrastructure, quadrupling GVIAO projections of this report are thus not predictions, and its would require an investment rate of about 30 percent of suggestions are tentative. They are no more than an at- natio:zial income. This is comparable to the rates in other tempt to contribute to the debate in China about the fast-growing East Asian economies, as well as East Euro- difficult questions that must be confronted during the pean countries, and in line with China's past investment country s unique socialist modernization. rate. With reasonable improvements in energy and mate- Pattern of Growth Per capita national income in the industrial countries is 1 Dean T Jamison and others, China: The Health Sector (Wash- (in terms of purchasing power) approximately ten times ington, DC.: World Bank, 1984). 1 rials use, and provided that the population in 2000 is sector. The pattern in China, however, is characteristic fairly close to the official target of 1.2 billion people, also of the Soviet Union and Eastern Europe. Continued quadrupling GVIAO would allow the Government's per emphasis on increasing the physical volume of produc- capita income target to be attained. tion (for example, the target of quadrupling GVIAO) The target growth rate of per capita income is far might cause this pattern to persist. But China could also above the 2.6-3.6 percent range projected by the World shift to a different pattern of growth, with accelerated Bank for middle-income countries in 1985-95. It seems development of the service sectors. feasible because China's projected population growth is Rapid service sector expansion and reform of the sys- unusually slow (about 1 percent per year, compared with tem of economic management are closely related options. an expected 2.2 percent in middle-income countries) and Expansion of commerce would go hand in hand with because China's investment rate has consistently been expansion of market regulation and increased specializa- unusually high (the 1982 average for middle-income tion of production units and localities. Expansion of en- countries was 24 percent). But its attainment depends terprise support services, including finance, accounting, crucially on the efficiency with which resources are used. and law, would likewise be a corollary of increased enter- For example, if the overall efficiency of investment were prise autonomy and specialization. And expansion of per- similar not to China's 1952-82 average but to the 1957- sonal services would reflect a change in the relative 77 average-roughly comparable to that of the Soviet weights attached to planners' priorities and people's pref- Union in 1950-75 and India in 1960-82-the growth erences. Moreover, rapid growth of services probably rate of per capita income would be about one percentage could not be accomplished through administrative direc- point lower. tives and centralized resource allocation, but would have If agricultural or energy production (discussed below) to be pulled by demand. Individual and collective enter- increased unexpectedly slowly, or if domestic or interna- prises may also be better suited than state enterprises to tional misfortunes adversely affected China's efficiency or providing many types of services. investment rate, growth could be slower still. In any This association with system reform makes it likely event, experience in China and other socialist countries that the benefits of faster service sector growth would strongly suggests the desirability of cautious planning. more than offset its costs (in terms of investment, skilled Unrealistically high growth targets cause fluctuations, labor, and-to a lesser extent-materials). In addition to shortages, and inefficiency, while aiming too low has few the direct benefit to households from more convenient adverse consequences. Indeed, continuation of China's retail commerce and from more restaurants, tailors, and recent practice of setting modest annual and medium- so on, there would be indirect contributions to the effi- term targets could greatly facilitate reform of the system ciency of industry and agriculture. A bigger, better- of economic management. equipped, and more responsive commercial sector could reduce requirements for both circulating capital and-by Sectoral Balance permitting greater specialization and realization of econo- Experience elsewhere also confirms the importance of mies of scale-fixed capital. Enterprise support services striking an appropriate balance among the different broad also make fuller use of specialized equipment and person- sectors of the economy. In particular, favoring industry at nel, while financial institutions can contribute to better the expense of other sectors does not seem to generate investment decisions. Larger commerce and enterprise rapid and efficient growth. In China, incentives for agri- support service sectors may also contribute to reducing cultural production have recently been substantially in- material input requirements, especially by accelerating creased, but additional measures to support agriculture improvements in product quality, which make the cost of (discussed below) will be needed. Moreover, emphasis on materials a smaller fraction of the price. quadrupling GVIAO should not imply neglect of infra- An alternative development path, involving both structure and services, which are vital for industrial and greater emphasis on services and more efficient resource agricultural efficiency. The adverse consequences of past use, is therefore analyzed in the report. With a level of underinvestment in electricity and transport are well rec- investment efficiency similar to the average in Japan in ognized in China, but there are still difficulties in divert- 1950-80 and in all middle-income countries in 1960-82, ing sufficient funds from directly productive investment it could attain the Government's target national income into infrastructure. And the importance of commerce and growth rate with an investment rate of only 26 percent, other services (including finance and enterprise support and hence faster growth of consumption. (The growth services) has only recently come to be appreciated. rate of industry would be 7 percent per year rather than The share of the service sectors-including com- 8 percent, however, and hence GVIAO would less than merce-in output and employment in China is at present quadruple.) On this path, China would also have the strikingly small by comparison with other developing option of maintaining the investment rate at 30 percent, countries, despite an untisually large education and health which would cause national income to grow one percent- 2 age point faster than the target rate (and GVIAO to slightly target -or 2000. Given the growth rate of nonagricultural more than quadruple). output implied by this target, and the present very large agricultural employment share, such a large decline in Human Resources agricultural employment could be achieved only if nonag- Faster service sector expansion could also make it easier ricultural labor productivity grew slowly. This might well for China to absorb growth in the labor force, averaging be incompatible with efficient industrialization. In the about 10 million workers per year in 1980-2000. In the early (lecades of the twenty-first century, however, agri- usual pattern of development-even in rapidly industrial- cultural employment is likely to shrink rapidly. izing countries-the main intersectoral employment shift Especially because many of the children of today's is from agriculture into services, with a more moderate farmers will move into nonagricultural employment at increase of employment in industry, which is less labor some point during their working lives, it will be impor- intensive. In China, however, employment has grown tant for China to continue to emphasize the wide diffu- faster in industry than in services. Continued emphasis on sion of basic education and to improve its quality. Inter- increasing the physical volume of production could per- national experience strongly suggests that this could petuate this pattern. This is especially likely because the contribute not only to faster economic growth, but also prospective decline in the primary school age group will to a less unequal distribution of its benefits. In recent cause the number of teachers-now approaching a quar- years, however, there have been problems in maintaining ter of all service sector employment in China-to increase the primary school enrollment ratio (although it remains only modestly. Service sector employment in 2000 might well above the developing country average); and the sec- thus remain as low as 70 percent of industrial employ- ondary school enrollment ratio has dropped below the ment (similar to the ratio in the Soviet Union in 1959). develcping country average. Wide variations among lo- On the alternative development path, by contrast, ser- calities in enrollment ratios and school quality also per- vice sector employment in 2000 could be as much as 150 sist. These problems are being aggravated by increased percent of industrial employment (similar to the ratio in local financial self-reliance in basic education, as well as Japan in the early 1950s). Such a change in the composi- by key schools (recently modified in name, but less in tion of nonagricultural employment could ease some substance) at primary and secondary levels. Policy problems in China's urban areas, which now suffer from chang2s, and greater financial support from central and a combination of surpluses of some sorts of labor and provincial authorities, will be needed if the important shortages of other sorts of labor (including competent, target of making nine years of basic education universal able-bodied workers even at relatively low skill levels). A by 2000 is to be attained. much higher ratio of service sector to industrial employ- As regards advanced education and training, the need ment would help to reduce this structural mismatch be- for rapid progrcss to make up for the damage inflicted by tween labor demand and supply, since service sector jobs the Cultural Revolution is well recognized in China, but tend to be disproportionately filled by the categories of some changes in emphasis seem advisable. University- labor that are now in surplus in urban China-particu- level courses are too specialized, partly because many larly women. educational institutions are subordinate to sectoral minis- Perhaps more important, faster service sector expan- tries and other agencies. The Government's present pol- sion, by raising the overall level of nonagricultural em- icy oi vocationalizing secondary education could suffer ployment, could reduce the proportion of China's labor from s;imilar problems. The case for sound general educa- force that must remain in agriculture, where the earnings tion, supplemented by subsequent, briefer training in and contribution to production of most workers are highly specialized skills, is particularly strong in China likely to remain unsatisfactorily low for the next two or over the next few decades. Rapid technical advance and three decades. On the material-production-oriented quad- structural change will require constant alterations in the rupling path, the proportion of the labor force in agricul- skill ccmposition of the labor force. The slow prospective ture-excluding brigade industries-might decline by growth of the labor force in the twenty-first century will 2000 to only slightly under 60 percent. On the alterna- reduce the scope for achieving such alterations through tive development path, though, with the same growth changes in the pattern of training of n-ew labor force rate of national income, it could decline to nearer 50 entrants. It will correspondingly increase the required percent, a difference of perhaps 40 million workers. amount of retraining, which is more difficult for people Even the latter percentage, which implies an absolute whose original education was highly specialized. increase in agricultural employment of 13 million, is much higher than most Chinese projections, which envis- Food ind Agniculture age a decline in the agricultural employment share to Despite its long history of food problems, China's per under 40 percent. This would be normal for a country capita consumption of calories and protein is currently with a per capita income roughly double the Chinese similar to the average for middle-income countries. The 3 share of animal products in the diet, however, is that of a Feedgrain imports could make up any conceivable short- low-income country, and a major issue is how far and fast age, probably without an appreciable increase in world this share should increase. This issue is hard to address prices, but it would be possible to export only a part of because of uncertainty about China's agricultural produc- the potential rice surplus at economically attractive tion potential: depending on whether the recent remark- prices. Even if overall crop production targets can be able surge in production wanes or persists, the future attained, it may thus be necessary to switch some rice trend growth rate of agricultural output could lie any- land either into coarse grains or into other crops that where between 2- 3 percent and 5-6 percent per year. could be more easily exported. In addition to the successful measures already imple- The future pattern of agricultural production and for- mented in recent years, especially introduction of the eign trade thus depends quite heavily on the rate of in- production responsibility system, the Government could crease in consumption of animal products. The possibility take a number of steps to enhance agricultural growth of a rapid increase cannot be ruled out, but there are prospects. These include improvements in agricultural re- some significant problems and uncertainties, which could search, education, and extension services; irrigation and be greater in the twenty-first century. (Even with a favor- drainage projects; better nutrient balance in fertilizer sup- able feed conversion ratio, an increase of ten percentage plies, as well as changes in the pricing and distribution points in the share of animal products in the average diet system so that fertilizer is allocated among crops and is equivalent to the disappearance of 3S-40 percent of localities more in accordance with its potential contribu- China's cultivable land, or a 35-40 percent increase in tion to production; increased availability of agricultural population.) Particularly undesirable would be an in- credit; and improvements in rural transport, storage, and crease in per capita consumption of animal products to a marketing facilities. level that could not be sustained-because of a slowdown To permit consumption of animal products to rise from in crop production growth, poor feed-meat conversion 6 percent of total caloric intake in the early 1980s to 1 5 ratios, or unwillingness or inability to sustain large feed- percent by the end of the century, gross agricultural out- grain imports-and therefore had to be reduced. put would have to increase at about 4.5 percent per year, Although there appears to be little danger of a decline with crop production growing at 3.6 percent per year in China's present very satisfactory food intake level, (less than the 1965-8 3 average) and animal husbandry at caution is accordingly needed in managing food demand. 7.5 percent per year (well above past rates). This agricul- It seems especially important that consumers should pay tural growth rate could simultaneously satisfy the other the full economic cost of animal products, without subsi- demands of economic growth at the government's target dies or administrative restrictions on prices. Japanese ex- rate, with agricultural exports and imports more or less in perience suggests that high prices can bring consumer balance in 2000. But it is very high by international demand for animal products into line with limited sup- standards-well above the 3.0 percent average for mid- plies in a socially acceptable manner. By contrast, the dle-income countries in 1960-80 and surpassed only by experience of countries which have subsidized and ra- Thailand and the Philippines, which started with more tioned consumption of animal products has been ex- uncultivated land and lower yields. tremely unfavorable. To achieve such rapid growth of animal production would also not be straightforward. China's pastures are Energy Production and Use IIow seriously overgrazed, which makes it unlikely that Because China's initially high consumption of energy per present targets for beef and mutton production growth unit of output offers great scope for conservation, it can be achieved. Much higher priority than at present would be possible to attain the Government's target of would need to be given to improving techniques for the quadrupling GVIAO with a much smaller proportionate production of poultry, a potentially efficient converter of increase in energy production-though probably not as feed into animal protein. The most important issue, how- small as the originally envisaged doubling. Electricity pro- ever, appears to be availability of grain and protein-based duction probably has to quadruple by 2000 to meet de- feeds, on which an increasing proportion of animal hus- mand, which will require a huge amount of investment bandry will have to be based. This is a matter partly of (averaging about 2 percent of national income), although the backwardness of China's feed-processing industry, it is technically feasible and in line with current plans. partly of achieving the required level and pattern of crop Crude oil production in 2000 is hard to predict, since it production. depends heavily on success in replacing the output of With rapid growth of animal husbandry, the Govern- existing fields with that of new discoveries, but the 200 ment's present targets for increasing production of spe- million ton target (twice the present level) could be more cific crops would tend to create a large shortage of the than enough to meet the likely level of domestic demand. coarse grains needed for animal feed and a large surplus of It is coal, however, that will bear most of the burden of rice, whose consumption is likely to grow only slowly. bringing future energy production into line with demand. 4 How much coal will be needed could vary widely, de- in exis:ing large-scale uses such as steel and cement pro- pending on the rate and sectoral pattern of growth, as duction-and to invest heavily in suitable production and well as on the degree of energy conservation achieved distribution facilities. Careful studies will be needed to within individual sectors. With growth of national in- select the least costly of the various alternatives (for exam- come at the Government's target rate, the material-pro- ple, coal gasification versus cogeneration and district heat- duction-oriented quadrupling path would require 1,400 ing) in each case; and their implementation and operation million tons of coal to be produced in 2000 on the most will elntail much closer cooperation than at present optimistic assumptions about progress in conservation, among a variety of agencies and organizations. and more probably 1,600 million tons (as compared with 770 million tons in 1984). At the same national income TransptPnionties growth rate, the alternative-more service-oriented and There is concern in China that economic growth over the efficient-development path mentioned earlier could re- next decade or two might be held back by transport duce coal demand in 2000 to 1,200 million tons on the shortages, especially given the length of time needed to most optimistic conservation assumptions, and 1,400 construct new railways and roads. But the volume of million tons on intermediate assumptions. freight transport in China is now so high (per unit of Especially because less optimistic, but still plausible, outpu:) by international standards-though lower than in assumptions about progress in conservation could in- the Soviet Union-that it is hard to predict how rapidly it crease the potential demand for coal in 2000 to over will nced to increase in the future, given any particular 1,800 million tons, it wiU be vital to promote economical trend growth rate of national income. This will depend use of all forms of energy. The mixture of administrative on the sectoral pattern of growth (for example, industry controls and financial incentives applied in recent years is more transport-intensive than services) and on its spa- has led to impressive energy savings. But a quota-based tial pattern (which influences average transport distance). energy allocation system is fundamentally unsuited to It will also be affected by system reform, which will on achieving large and economically rational reductions in the one hand tend to increase transport requirements energy intensity over the longer term. It gives enterprises because of increased specialization and exchange, but an incentive to exaggerate their needs and does not take could on the other hand greatly reduce wasteful use of enough account of the widely varying economic costs of transport facilities, for example, through less cross-haul- reducing energy consumption in individual enterprises. It ing, nmore preliminary processing of materials, and more is very probable that more energy could be saved, with rational location decisions. less waste of other resources, if enterprises were more If national income were to grow at the Government's strongly motivated to reduce all costs and faced with target rate, with continuing strong emphasis on material prices that better reflected the economic values of all their production, and with agriculture and industry using inputs and outputs. transport only moderately more efficiently than in the It should be possible to mine the required quantities of past, freight volume could reach 3,200 billion ton-kilo- coal, given China's ample reserves and the straightfor- meters by 2000 (approaching four times its 1980 level). ward production technology, although this would require But with the same growth of national income achieved some advance planning, substantial investment, and con- through faster expansion of services and slower expansion tinued support for small collective mines. Much more of of industry, and with somewhat greater efficiency in a challenge will be to transport the coal (as discussed transport use, freight volume in 2000 would be 2,400 below) and to use it cleanly and efficiently. The share of billion ton-kilometers. Even in the latter case, China's coal in total commercial energy use is likely to remain at projected freight intensity would remain high in compari- 70-75 percent (compared with 30-35 percent world- son with other countries, which suggests that the possibil- wide), which will require it to be put to an unusually ity of even slower growth of freight volume cannot be wide range of uses, including many where backward ruled out. technology and small scale in China today result in waste- Althaough the overall volume of transport requirements ful and dirty combustion. in the longer term is highly uncertain, there will certainly To prevent further increases in already unacceptably have :o be significant changes in the composition of trans- high levels of urban air pollution, as well as to economize port investments, as well as in organization and manage- on fuel, will often require the replacement of decentral- ment Agricultural and industrial specialization will in- ized and uncontrolled burning of coal in households and volve much movement in fairly small lots of a great enterprises with centralized, large-scale, environmentally variety of goods, often perishable or fragile, over short to controlled combustion to produce cleaner forms of en- medium distances, to and from dispersed origins and des- ergy (gas, electricity, steam, hot water) for distribution to tinations. This is generally only feasible by road. The final users. This in turn will require China to import and demand for road passenger transport will also increase disseminate new technology-which will also be critical rapid y. However, China now has a rural road network 5 half the size of India's and fewer trucks per rural person sufficiency. This appears to result partly from transport than the impoverished Sahelian countries of West Africa. problems, but also from shortcomings in commerce and Major intercity highways, the vehicle fleet, and vehicle material supply, the nature of the fiscal system, distorted fuel production and distribution are also at present strik- prices, and insufficient competition. It could have a seri- ingly backward. ous adverse effect on China's overall investment effi- That road transport should play a larger role in the ciency. future is widely recognized in China. Yet there seems to To prevent continued proliferation of industrial plants be no strategic plan to bring this about. Underutilized of suboptimal scale and product quality, there is thus an agricultural labor could, as in the past, be mobilized to urgent need for measures to reduce the present conflict build up an adequate rural road network. But higher between what is rational for a particular locality and what levels of government (central and provincial) will need- is rational for China's whole economy. Of particular im- as in other countries-to play a much more active role in portance would be better transport, fiscal reform, and planning, coordinating, and financing the road develop- changes in the system of economic management (dis- ment schemes implemented at lower levels. In addition, it cussed below), regarding enterprise motivation, competi- will be essential to produce a much larger number and tion, and prices. In addition, the central government wider range of more fuel-efficient trucks, to improve the might consider legal measures to back up its existing gene- utilization of vehicle fleets, to upgrade the quality of ral prohibition of barriers to interlocality trade, including motor fuels, and to distribute and price these fuels more possibly the establishment of a special regulatory institu- rationally. All this will require improved coordination tion, with the power to levy large fines. among higher-level government ministries and agencies. Suboptimal plant size is not China's only current indus- Railways will remain the most efficient means of trans- trial location problem. There is also too much concentra- porting raw materials and heavy industrial products over tion of industry in large and medium-size cities. Research long distances. There will be a large increase in demand and experience in other countries suggest that economic for coal transport, even with more coal washing and efficiency requires weight-reducing heavy industry (for minemouth power generation. But there is apparently example, iron and steel production) and agroprocessing much scope for economizing on rail transport of all com- to be located near raw material sources, with other space- modities through stronger incentives for cost reduction intensive or standardized industrial activities in special- and less administrative intervention in materials alloca- ized small cities. Industries such as petrochemicals and tion and pricing. It also seems physically and financially bicycles benefit from large-scale operation, but not from possible to accommodate the overall increase in rail trans- proximity to other industries-or at any rate not enough port demand. This could be accomplished largely by to justify the much higher economic cost of land and switching from steam to diesel and electric traction, al- labor in large cities. By contrast, industries whose tech- though substantial double tracking and some new lines nology is evolving rapidly or which must respond to will also be necessary. constantly changing demand, as well as many specialized Increased use of water transport could appreciably re- service activities, benefit substantially from mutual prox- duce the load on the railways, expecially if future heavy imity and clustering in large urban areas. Development of industrial investments were appropriately located. And both large and small cities, as well as smaller towns pro- China's entire transport system could benefit from better viding goods and services to rural areas, could thus con- intermodal coordination: this is partly a matter of im- tribute to rapid and efficient growth in China; but plan- proving technology and infrastructure (for example, con- ners and prices (discussed below) will need to give better tainers and their handling facilities), but it also depends guidance on what activities should go where. on rationalizing tariffs, increasing competition, and en- How much the overall level of urbanization should couraging new specialized entities to provide an appropri- increase in the decades ahead is a difficult issue, partly ate variety of transport, leasing, storage, and transfer ser- because the distinction between urban and rural is itself vices. blurred in many densely populated parts of China. In any event, substantially increased rural-urban migration Specialization and Urbanization seems probable and desirable. Indeed, without some ru- China's size and past emphasis on local self-sufficiency ral-urban migration, China's overall urbanization rate (as offer opportunities for enormous economic gains through officially measured) would decline from 20 percent in increased specialization and trade among localities. In re- 1981-quite normal for a low-income country-to 19 cent years, agricultural specialization has proceeded rap- percent by 2000, simply because the natural population idly. In industry, by contrast, specialization does not seem growth rate in existing urban areas over the next two to be increasing, and indeed local planning bureaus are decades is likely to be much lower than in rural areas. still promoting the development of a wide range of indus- This could cause the general labor shortages that have trial activities, with continuing emphasis on local self- already appeared in some Chinese cities to become wide- 6 spread within a few years. Increasing urbanization would income country to run foreign trade surpluses, as China probably be a more economically efficient response to has done in recent years, has some obvious disadvantages, such shortages than accelerated automation of urban pro- as compared with a net inflow of foreign capital to sup- duction. An urbanization rate in 2000 of 30 percent plemeni: domestic savings. But experience in other coun- (implying that more than a third of the urban population tries, including socialist ones, has shown how easy it is to would have come from rural areas) would still involve an slide into excessive or unproductive borrowing, especially unusually large amount of rural nonagricultural employ- when domestic enterprises and planners are not suffi- ment. Lower-middle-income countries today (at a similar ciently conscious of investment costs or responsible for income level) have an average of 34 percent of their bad decisions. population in urban areas. However, China also seems wise to encourage direct Whatever the overall rate of urbanization, efficient de- foreign investment, less for the foreign capital or ad- velopment of towns and cities will require a very large vanced technology it brings than for the demonstration increase in urban infrastructure and services. The admin- effect of modern management techniques. The example istrative responsibilities for their provision, and potential of, and competition from, well-run foreign companies financing and cost recovery mechanisms, need to be care- can help domestic firms identify weak links in manage- fully reviewed. In many countries, efficient and equitable ment, product design, material supply, and so on, and development of urban services has been seriously con- spur them to make changes they might otherwise never strained either by fragmented and unclear allocation of consider. But such examples of the way modern industry responsibility among different administrative levels or by operates in more advanced countries are effective only if inadequate growth of urban revenue sources. In China, contrasted and competing with locally managed firms. moreover, special attention should be given to the poten- This requires foreign and joint ventures to be spread-as tial impact of economic and social reforms on the level is increasingly the case in China-among a wide range of and composition of local government revenues and ex- localities and activities, rather than confined to special penditures. zones cr particular sectors. International experience suggests that decentralization Foreign Trade and Capital of foreign trade to the enterprise level can also make an China's recent resumption and expansion of external eco- important contribution to realizing the potential benefits nomic contacts have raised exports to 9-10 percent of of increased external economic contacts. Direct exposure national income, which is well within the normal range of exporting firms to foreign buyers and competitors has for large countries. However, inflows of foreign loans proved elsewhere to be an extremely effective way not and direct investment have remained small, partly be- just of learning in the abstract about new and better cause of foreign trade surpluses. For the future, there products and processes, but also of learning how (and appear to be two large and related issues. The first is how being put under pressure) to introduce them in practice. much, and in what directions, to further increase external Similarly, allowing enterprises-including commercial en- contacts. The second is how best to manage such con- terprises-and consumers to choose directly between im- tacts, and in particular whether to continue the present ported and domestically produced goods could greatly insulation of the domestic economy from the world out- increase competitive pressure on Chinese producers to side by administrative intermediation and separate price introduce new, better, and cheaper products. systems. Such decentralization of foreign trade could have some Over the next two decades, China will need to change potentially serious disadvantages, too, including damage the composition of its manufactured exports. Exports of to promising but newly established industries in China, as textiles and clothing, in particular, are likely to grow well as greater exposure to world price instability and fairly slowly. China will therefore need to make a major unpreclictable fluctuations in foreign markets. But skillful effort to expand exports of machinery and metal prod- use of policy instruments-including the exchange rate, ucts. Rapid growth in such exports would permit rapid tariffs, and selective administrative interventions-could growth of intermediate and capital goods imports, as well minimize the costs while securing many of the benefits. as consumer goods imports (which can stimulate impor- China will also always be a large and diverse economy, tant improvements in domestic production). Exports of with a relatively small foreign trade sector, which greatly manufactured goods could also be used to pay for pur- reduces the possible damage that unexpectedly unfavor- chases of licenses and technological assistance, which in able external developments could inflict, even though many advanced as well as developing countries have particular industries and localities might experience quite proved to be very effective ways of upgrading domestic sharp swings in prosperity. technology. Especially following the recent unification of foreign As regards foreign borrowing, the Government's pol- exchange arrangements-and given the current account icy of cautious progress seems well founded. For a low- surpluses of recent years, large reserves, and the ready 7 availability of external capital-China appears to be well not only in bringing together the several elements of placed to move toward a decentralized and mainly indi- market regulation, but also in combining them with an rectly regulated foreign trade system, without the prob- appropriately modified system of planning. lems (of external insolvency or internal inflation or defla- tion) that many other countries attempting a similar Enterprise Motivation move have simultaneously had to contend with. To do The unsatisfactory results of direct administrative control this immediately, or at one stroke, would probably be of enterprises-whether by central ministries or local gov- unwise, since most domestic enterprises do not yet have ernments-lead irresistibly to the principle of greater en- appropriate motivation or independence, and since some terprise independence. Yet establishing independent and prior changes of domestic prices would ease the adjust- appropriately motivated enterprises may be the hardest ments required. But it could proceed in parallel with single aspect of reform in a socialist system-and perhaps internal economic reform. also the most fundamental, since an unmotivated enter- prise cannot be expected to motivate its workers, or to Managing the Economy respond to price and tax signals, or to compete. Appropriate enterprise motivation should include a How far China will have risen into the middle-income strong desire to increase profits. Avoidance of losses is an range by 2000 will depend crucially on how successfully important aspect of this, but is insufficient, since China the economy is managed. This is a matter partly of mobi- needs enterprises that not merely pursue a passive strat- lizing financial and human resources, and of importing egy of staying out of trouble, but actively seek to increase and developing modern technology, partly of putting production and sales and to cut costs of all kinds. With these things to effective use. China's past record of eco- rational prices and competition, the best single measure nomic management, for all its unevenness, is superior to of enterprise performance in this regard is usually me- that of other low-income countries, both in promoting dium-term profits (especially after deduction of the cost- growth and in reducing poverty. But maintaining the past depreciation and interest-of the capital employed). The pace of economic progress is likely to require greater active desire of peasant households to increase sales and efficiency. cut costs, because their standard of living depends on it, Efficiency in the relevant sense means ceaseless, intense lies at the heart of China's recent agricultural successes. efforts by producers to reduce costs, to increase produc- Outside agriculture, the same motivation is a natural fea- tivity, to improve product quality, to introduce new ture of individual and family enterprises, and of small products, and to seek out and respond to changing needs enterprises owned and operated collectively by their and opportunities. It involves bold yet frugal investment workers. In medium-size and large enterprises, however, decisions, increasing specialization, and the constant dis- things are not so simple. placement of more expensive or inferior products and processes by cheaper and better ones. STATE ENTERPRISES. Some important enterprises These conditions for rapid, sustained growth are at should clearly remain subject to direct government con- present far from fully met in China, except in agriculture, trol and should not primarily pursue profit. At a mini- where outstanding progress has been made in recent mum, this applies to public utilities (for example, electric- years with the introduction of the production responsibil- ity, railways, and telecommunications), large-scale ity system. Recognizing this, the Central Committee of mineral exploitation, and many defense-related indus- the Chinese Communist Party in October 1984 an- tries. Direct control of other key enterprises might also nounced a program of urban economic reforms, with help to achieve specific development objectives, including invigoration of state enterprises as its central theme, to be the technological upgrading of particular sectors. implemented over a period of about five years. Much of What is less clear is the best way to organize and the discussion in the present report is premised on the motivate the remainder-probably the great majority-of Central Committee's decision. state enterprises, which would be relatively independent. To be efficient, enterprises must be motivated to im- Once a suitable economic environment is created through prove their economic performance; they must have some price reform and increased competition, and provided freedom of maneuver; they must be faced with economi- that indirect levers such as taxes and credits are skillfully cally rational prices; and they must be subjected to com- applied, pursuit of profit should lead most state enter- petition. None of these elements is individually easy to prises in economically appropriate directions. It would establish, and the absence of any one of them reduces or thus no longer be necessary for particular individuals nullifies the benefits of the others. In addition, the state within enterprises to be charged specifically with repre- must retain the ability to direct the overall pace and senting the interests of the state. But each of the alterna- pattern of development. The essence-and the diffi- tive possible internal management arrangements has culty-of a successful and comprehensive reform thus lies strengths and weaknesses. 8 Giving direct control of state-owned enterprises to are banks, pension funds, and insurance companies.) their workers is a socially attractive possibility. Japanese Such a system of socialist joint stock ownership could experience also shows the economic advantages of strong perhaps be created initially by suitable dispersion of the worker commitment to the enterprise. But there are also ownership capital of existing state enterprises. Over time, disadvantages, especially because state enterprises should it could be reinforced by a more diversified pattern of operate for the benefit of the whole society and not only investment finance (discussed below), with a variety of of those who work in them. Experience in Yugoslavia and state institutions acquiring financial interests in existing elsewhere suggests that worker control could result in and new enterprises. excessive wages and worker benefits, inadequate labor discipline and effort, restriction of employment, and COLLECTIVE AND INDIVIDUAL ENTERPRISES. China's comparative indifference to profitability. policy is to encourage the coexistence of state and other Experience in industrial capitalist economies suggests sorts of enterprises. In small-scale economic activities, that giving control of enterprises, instead, to their man- whose importance will increase with economic develop- agers would increase their propensity to expand, im- ment and reform, individual and collective enterprises are prove, and innovate. But completely independent man- already making a valuable contribution. They could be agers sometimes choose a quiet life or, more commonly, even more productive if given better access to materials, pursue personal power through expansion with insuffi- credit, skilled labor, and technology. Their potential role cient attention to profitability. In socialist countries, in medium- and large-scale activities, however, remains to moreover, managerial control tends to differ only slightly be determined. from worker control, since managers find it hard to resist Experience elsewhere suggests that genuine collectives worker demands for greater benefits or to insist on the flourish only in certain lines of activity and rarely above a often unwelcome changes in work practices that are certain size. Many of the most successful examples else- needed for innovation and increased efficiency. This is where are in commerce and are marketing or purchasing already a problem in China: studies of experimentally cooperatives. There are far fewer examples of medium- reformed enterprises controlled mainly by their managers size gcnuine worker collectives or of any sort of collec- reveal large increases in worker benefits, but small in- tives in manufacturing. China's numerous township and creases in economic efficiency. village enterprises may appear an exception; however, An alternative approach, common in nonsocialist these are not collectives in the ordinary sense, but com- countries (and now being tried in Hungary), is to give munity enterprises. They have been successful partly be- strategic decisionmaking authority in each enterprise to a cause of restrictions on personal mobility among commu- board of directors. Such a board could contain some nities, which has given local residents a strong common representatives of society at large, as well as of the work- interest in their establishment and profitability. ers. But to impart the necessary motivation, the board Collectives and community enterprises could in the would mainly have to represent institutions with a strong future constitute a significant proportion of China's me- interest in the enterprise's profits. It could insist that the dium-size enterprises. Many individual enterprises could enterprise managers behaved accordingly, partly by ap- also giow rapidly to medium size, if allowed to do so, and pointing and dismissing them, partly by linking their re- could be an important dynamic force in the economy. As muneration to profitability. The managers might in prac- an alt-rnative to the present restrictions on the size of tice make most decisions, but would be greatly influenced individual enterprises, these enterprises might be obliged by their ultimate accountability to the directors. to sell (at a fair price) majority ownership to one or more Such boards should be free from direct government state institutions once they reached a certain size. The intervention. But, precisely because these are state enter- then rniinority owners could continue as managers, super- prises, this may be difficult to achieve. In China, for vised by a board of directors representing all the owners. instance, even if the boards were to consist of representa- tives of the Ministry of Finance or a new Ministry of ACC:OUNTING AND LAW. The best way of motivating State Property, rather than of sectoral ministries or the independent state enterprises, and the best balance governments of the localities in which the enterprises are among state, collective, individual, and mixed enterprises, situated, informal connections and pressures for direct can only emerge from experience and experiment over a administrative control could persist. A possible solution protracted period. But certain necessary conditions for might be to spread the oNvnership of each state enterprise success can already be identified. among several different institutions, each in some way One is improvement of enterprise accounting and au- representing the whole people, but with an interest diting, which at present leave much to be desired. With- mainly in the enterprise's profits rather than directly in its out accurate accounts, subject to thorough, compulsory, output, purchases, or employment. (Examples of such and independent audits, with severe penalties for non- institutions, in addition to central and local governments, compliance, it is hard to see how any system of indepen- 9 dent enterprises-state or nonstate-could function prop- Another important step toward free entry could be to erly. allow individual and collective enterprises to operate in Moreover, a decentralized economy cannot function more lines of activity, including technologically dynamic properly without a comprehensive system of commercial sectors. Studies in other countries have shown that small and contract law and institutions and personnel to imple- enterprises are just as innovative as large ones. (Chinese ment it. Laws are also needed to regulate the economic visitors to foreign countries are often given a misleading activities of independent enterprises. Among other impression in this regard, because they are usually taken things, laws or decrees should prohibit specified types of only to well-known, large enterprises.) Particularly be- monopolistic, anticompetitive, or exploitative behavior, cause of the strong incentive that small enterprises have with legal institutions to interpret and implement these to innovate, the overall pace of technological advance prohibitions. At present in China, the situation is unsatis- could very probably be accelerated by letting them com- factorily vague, with enterprises and local officials free to pete for some of China's skilled manpower, technology place their own interpretations on general guidelines imports, and research support. from the center. Lastly, effective competition has to involve the elimina- tion of obsolete or unwanted products and processes. Competition Subsidies or protection for backward technology and in- As the Central Committee's 1984 decision clearly states, efficient enterprises-other than on a temporary basis to competition among enterprises is crucial to efficient permit reorganization-hold back the growth of efficient growth-without it, customer requirements are ne- enterprises and remove a powerful negative stimulus to glected, innovations diffuse slowly, and market-regulated innovation and improved efficiency. At the same time, prices are distorted for monopolistic gain. One key ingre- this aspect of competition can obviously cause hardship, dient of competition is allowing purchasers to buy from especially for displaced workers. Social policies and insti- the best and cheapest source. China has already taken tutions (discussed below), and government retraining some significant steps in this direction, but further schemes, can alleviate this hardship. But the conflict be- progress is needed. It could probably be assisted by dis- tween the gains of the majority from greater competition mantling the annual production planning and allocation (and faster technological change) and the losses of particu- system, either in one step (as in Hungary) or more gradu- lar individuals and groups cannot be completely elimi- ally, by reducing each year in each enterprise the propor- nated. tion of material requirements covered by the state alloca- tion and, in parallel, the obligatory state production and Pnices procurement quota. The Government is well aware of the need to establish It would also be assisted by dismantling administrative economically rational prices, without which the decisions barriers to internal trade, which would be easier if fewer of independent, profit-oriented enterprises would often enterprises were directly controlled by government or- be inefficient for the whole economy. The Government gans anxious to avoid competition within their "families" also recognizes that this must be accomplished by chang- of enterprises. Consolidation of enterprises into large sec- ing the price-setting system to give market supply and toral corporations, by contrast, would normally tend to demand forces a greater role, and not only by adminis- reduce competition, although this could be offset by in- tered changes in prices set by the state, which tend to lack creased exposure to foreign competition. Equally impor- the flexibility, complexity, and precision needed in a mod- tant would be measures to enlarge, diversify, and ern economy. The prices of many minor items, and of strengthen commerce and related services. Individual and some transactions in more major items, have already collective enterprises might be given a greater role in been successfully decontrolled. wholesaling and material supply and better access to Further progress in this direction is impeded by chronic transport facilities. shortages of many goods. It is feared that these shortages, Another key ingredient of competition is free entry of if prices were no longer subject to direct state control, new producers into particular markets. One step in this would cause general price inflation, as well as obstructing direction for China could be to allow larger enterprises to the allocation of materials to key projects. For this rea- diversify more freely out of their existing lines of busi- son, it is often argued in China that further price decon- ness. Provided that prices are rational, such diversifica- trol should be postponed until rising production has elim- tion can transfer profits earned in one line of activity into inated most of the shortages. other lines where the economic returns to investment are Soviet and East European experience suggests, how- greater. And it can permit enterprises to transfer workers ever, that chronic shortages are not the temporary result who can no longer be profitably employed in one line to of inadequate production capacity, but an enduring fea- other lines, without having either to discharge them or to ture of administrative economic management, which can keep them idle. be eliminated only by systemic reforms, including price 10 decontrol. This view is supported by China's experience ternative would be to have wholly market-determined in the past few years, especially in agriculture, where prices, but with a price stabilization scheme-Govern- relaxation of direct controls has turned long-standing ment intervention to keep fluctuations at any given time shortages into abundance. The same could happen in within a predetermined range by purchasing for addition industry. Increases in specific prices could eliminate spe- to stocks at the lower end and selling from stocks at the cific shortages by stimulating supply and reducing de- upper end. mand. These price increases, moreover, would tend to A third instance concerns the relationship between pro- reduce purchasing power over (and hence the prices of) ducer and retail prices, which needs to be altered not only other goods, provided that the government kept strict to reduce the presently high level of subsidies, but also to control of the budget balance and credit. encourage consumers to buy less of things (such as energy Nonetheless, there is substance to the fear of immedi- or animal products) in short supply or whose production ate and general decontrol of prices in China. The down- costs are increasing and more of things in abundant sup- ward inflexibility of some prices, and of wages, could ply ancl whose production costs are falling. This could be obstruct the smooth adjustment outlined above. The full achieved by establishing relatively rigid margins between response of demand and supply to price changes may producer and retail prices, sufficient to cover the costs take years. China lacks experience in indirect fiscal and and normal profits of commerce as well as indirect taxes monetary regulation of the general price level, and other (whose rates could vary from commodity to commodity). countries with much greater experience still suffer from In the near term, what is mainly needed are some rising prices. The elimination of shortages, moreover, substantial increases in the retail prices of staple food and would also require increased competition and stronger coal and in rents. These should be as fully and accurately enterprise motivation to hold down costs, including in- compensated as possible, by wage increases and income vestment costs, which cannot be brought about over- supplements to households with high dependency ratios night. For these reasons, price decontrol probably has to (discussed later). Special supplementary interest payments be gradual though steady-perhaps in conjunction with on saving deposits which would otherwise lose part of the gradual dismantling of annual production planning their real value might also be needed (financed perhaps by and allocation mentioned earlier. a corresponding special levy on borrowers, whose loan Some steps, however, could be taken more quickly. repayment burden would otherwise decrease in real Large increases, either administratively imposed or mar- terms). A large-scale program of selling state- and enter- ket-determined, in the prices of most forms of energy- prise-owned housing to its present tenants, who would and some other materials-are a case in point. Without otherwise face increased rents, could soak up a large frac- them, there would be little chance of achieving the essen- tion of existing saving deposits and hence reduce the tial reductions in energy use discussed earlier. There is possible scale of panic buying in anticipation of price concern in China that large energy price increases would increases. have ripple effects, since energy-intensive industries and There is room for disagreement as to whether such a enterprises could not absorb all the increased cost major realignment of retail prices, incomes, and assets through conservation and profit reduction and would should be done in one step, or more gradually. The social therefore have to raise their prices. But most such ripple problems that have sometimes followed large retail price effects should in fact be welcomed (and cushioning increases in other countries seem to have occurred mainly through reduction of taxes and profit remittances because of the absence of compensation-often deliber- avoided): experience elsewhere suggests that increases in ate, because of the need to reduce real consumption in the prices of energy-intensive products, by discouraging the face of economic difficulties. In China, however, their use, can make a vital contribution to raising the there is no need for a cut in household consumption, and economy's overall energy efficiency. full compensation could be provided (although this Agricultural prices are another instance. With the new would mean no net improvement in the state budget system of household agriculture, if production is to be balance). A carefully prepared and well-explained one- efficient and surpluses and shortages of particular com- step adjustment could thus be quite acceptable and would modities avoided, prices must be allowed to respond flex- avoid the protracted uncertainty and delays to other nec- ibly to supply and demand trends. At the same time, it essary reforms associated with gradual adjustment. But it would be desirable to avoid the large short-term price would also obviously increase the cost of errors in prepa- fluctuations-often because of weather-that characterize ration (especially calculating the required changes in unregulated markets for agricultural commodities. The prices wages, and other forms of compensation) or im- Government's present strategy is to decontrol completely plementation. the prices of minor items and, for major items, to have Following this initial phase of adjustment, subsequent two-tier pricing (flexible prices for amounts in excess of rises and falls in producer prices could be reflected in official, fixed-price procurement contracts). A possible al- retail prices. Experience in other countries suggests, how- 11 ever, that it would be inadvisable to provide continuing constraining China's growth in the next two decades, automatic compensation for retail price increases in the measures to improve their allocation and motivation form of general wage or income indexation, since this seem worth considering. Among the more important tends to aggravate inflation. Even selective indexation (of possibilities would be freedom of job choice, for both state income supplements to poor households, for exam- new graduates and experienced staff, allowing employers ple, or for pensions) should be approached cautiously. to recruit and release freely and competitively, and higher A fourth instance concerns prices that influence the and more flexible salaries. Provided that they were ac- spatial location of economic activities. One is transport: companied by reform of enterprise motivation, competi- prices for each mode should more accurately reflect long- tion, and prices, these measures could help move highly run marginal economic costs, without cross-subsidies skilled labor to where its economic contribution was among different length hauls, to provide incentives to greatest, speed the diffusion of new technology, improve reduce the present waste of rail facilities, to encourage motivation (with better matching of jobs to personal pref- greater use of roads, and to enable proper assessment of erences and skills just as important as more money), and the costs and benefits of locating particular activities in send much clearer signals of needs and priorities to educa- particular places. The other is urban land: for social as tion and training institutions. well as economic reasons, enterprises and planners should More generally, it will be essential to make better use be made to feel the dramatically varying economic useful- of all categories of labor. This is because long-term ness (or opportunity cost) of different sites-higher in growth of average per capita income will be almost en- coastal cities than in more remote regions, higher in large tirely determined by growth of labor productivity, most than in small cities, higher near the center than in the of which must come not from movement of labor out of outskirts (in a city the size of Shanghai, experience in agriculture into other sectors, but from higher productiv- other countries suggests that a central site is worth ap- ity within each sector. Unskilled labor is currently in proximately 150 times as much as one in the suburbs). surplus, but this will change in the twenty-first century; A differentiated urban land tax reflecting these varia- experience elsewhere shows how hard it is at a late stage tions could be introduced. Alternatively, a competitive to break out of an established pattern of low productivity rental market might be established-which would in prin- and indifference to labor costs. ciple be more efficient and quite consistent with public It might therefore be desirable to give enterprises pro- ownership of all land. (Either or both could provide a gressively more discretion in deciding how many, and useful supplement to municipal revenues.) This would which, less-skilled workers to employ. This should in encourage better use of existing sites and more economi- principle encompass not only the right to dismiss idle or cally efficient location decisions. The presently huge in- negligent workers, but also to release or reject workers centives for rural-urban migration would likewise be re- who are simply not needed for production. Lifetime em- duced to more economically rational proportions, ployment would not necessarily disappear (Japanese expe- especially if differing land values were reflected in house rience confirms its potential advantages in training and rents and wages, and hence in the costs of living and motivation), but would be likely to be limited to larger employing workers in large urban areas. enterprises, which would be highly selective recruiters. Finally, further strengthening of the linkages between Some open unemployment would be an inevitable con- internal and world prices could be beneficial, though sequence of these changes. In addition to its adverse hu- there would need to be some exceptions-for example, man consequences (which could be mitigated by changes rice, where an increase to the present world price could in social policies and institutions, discussed below), this aggravate incipient overproduction. For most goods, es- would involve waste of human resources. The waste of pecially manufactures, enterprises should eventually di- unemployment, however, has to be weighed against the rectly feel the relationship between their value (for ex- waste of human resources associated with the present ports) or cost (for imports) to China in world markets employment system, which transfers unemployment and their domestic prices. Such linkages (which are now from the streets into the factories and discourages man- standard in Hungary and have begun to be introduced in agers from realizing the full productive potential of their China) are essential for the efficient functioning of a de- workers. It is also important to bear in mind the tremen- centralized foreign trade system. They would also con- dous potential for creation of new jobs in the service and tribute to a much-needed widening of price differentials small enterprise sectors. Nonetheless, movement should between low- and high-quality products, without which be gradual, perhaps with an early end to the obligation of enterprise incentives to innovate and improve quality will enterprises to accept unwanted new workers (usually as remain small. part of a package also containing some useful recruits), and subsequent release of unneeded workers each year up Labor and Wages to a specified percentage of an enterprise's labor force. To reduce the risk of shortages of highly skilled workers Greater flexibility of less-skilled wages might smooth 12 this transition, as might stronger linkages between indi- Experience suggests that this would encourage bold and vidual pay and performance. But these changes, too, innovative thinking combined with thorough analysis, could be only gradually introduced, to avoid a sharp drop hard-headed calculation, and the avoidance of waste. in unskilled wages. It will also be necessary for some time Delegating more responsibility for investment decisions to retain administrative control of total and average to independent enterprises might also help in striking a wages (and other benefits) in state enterprises, to prevent better overall balance between infrastructural and other excessive increases in response to pressure from workers. investment. China's local governments have proved dedi- In this regard, the Government is contemplating formal cated and resourceful investors, but are probably unduly linkages between the remuneration of workers and the biased :,n favor of directly productive industrial invest- performance of their enterprises. But experience else- ment. Making enterprises independent, with more invest- where suggests that this could generate inequities and ment responsibility, could thus give local governments a misrepresentations that might undermine aggregate wage stronger incentive-as in other countries-to improve in- control. It might be preferable to have wage guidelines frastructure in order to attract industrial investment. that were more uniform across enterprises, coupled with China's recent mixed experience with devolution of greater managerial discretion in the distribution of wages investment decisions to enterprises, however, confirms among workers within particular enterprises. that without other systemic changes such an arrangement Gradual wage adjustments could also help remote and can have undesirable results. Inappropriate managerial backward localities. At present, nationally uniform state incentives, irrational prices, lack of competition, and sector wages, and unofficial pressure to tie collective sec- shortcomings in commerce and material supply can lead tor wages to state wages, make it harder for enterprises in enterprises to undertake investments of little economic these localities to compete effectively against enterprises merit, while neglecting other investments of much in more advanced regions or in international trade. Non- greater use to the economy. agricultural employment in backward areas might there- Reforms in enterprise motivation, competition, and fore increase faster if unskilled wages in these areas de- prices could eventually overcome many of these prob- clined in relative terms. At the same time, though, these lems, making it possible to consider devolving a much areas probably need to offer skilled workers higher wages greater proportion of investment decisions to enterprises than elsewhere. and, relatedly, allowing state enterprises to retain a much larger proportion of their profits. More mobility of in- Investment Decisions and Finance vestment funds, through more varied channels, could also In energy, transport, education, health, and defense, the increase the efficiency of decentralized investment deci- Government will undoubtedly retain direct control over sions. most investment decisions, mobilizing the necessary fi- Investment flows in China in the past have been largely nance through the budget. The same will apply to major vertical, with most savings being mobilized upward land and water development projects. Budgetary saving through the budget and dispensed downward through to finance other sorts of investment may also have to sectoraL ministries and state banks. In addition, there has remain substantial, if the high aggregate saving rate neces- been some compartmentalized reinvestment of savings sary for rapid growth is to be attained. But households generated within particular sectors and localities. In the and enterprises could contribute a much larger share of future, although vertical flows and compartmentalized aggregate saving than in the past and could make an reinves ment will remain important, they could advanta- increasing proportion of investment decisions. This pat- geously be supplemented and replaced to an increasing tern is already established in agriculture and other small- extent by horizontal investment fund flows of three main scale activities. In larger-scale industry and services, how- types. The first, mentioned earlier, would be allowing ever, the degree to which investment decisions should be existing enterprises to diversify into new lines of activity. devolved to enterprises remains an important issue. The second, which overlaps with the first, is direct invest- Such devolution could have a number of advantages. It ment linkages among economic units: enterprises could would give greater meaning to enterprise independence, increasingly be permitted to invest in other enterprises, especially since the economic performance of an enter- establish new enterprises, or participate in various sorts of prise at any given time depends heavily on previous in- joint ventures. The third sort of horizontal flow is vestment decisions. It would increase the medium and through socially owned financial institutions, acting as long-term responsiveness of production to evolving needs intermediaries between the suppliers and users of re- and demands. It would make fuller use of the detailed sources. knowledge of enterprise managers. It might also provide Such financial institutions, offering interest rates that a way of strengthening incentives for sound investment reflected the scarcity of capital, could mobilize funds decisions, in whose outcome enterprise managers could from households and institutions, including enterprises be given a substantial direct personal financial interest. with lmited internal investment opportunities. They 13 could make these funds available to potential investors, INDIRECT CONTROL. In the remaining areas of the especially enterprises of all kinds, perhaps mainly in the economy, where decisions will increasingly be made by form of loans, but also in the form of ownership capital- households, farmers, and independent enterprises, there partly because many worthwhile projects are too risky to will be a shift from direct to indirect control. This, how- be largely loan-financed, partly because directly sharing in ever, need not diminish-and indeed might well in- the risks and benefits would motivate financial institu- crease-the Government's ability to manage the econ- tions to provide more assistance in project design and omy. For up to now in China, what has not been directly implementation, including information on markets and controlled has often not been properly planned. The technology. These institutions could in effect create a broader reach of indirect controls, which shape the envi- socialist market for investment funds. ronment in which all economic units operate, could more For such a market to function well, the financial insti- than compensate for the associated loss of direct control tutions should be numerous and diverse, acting in most over particular economic units. respects as independent state or collective enterprises. Instruments of indirect control fall into two main cate- Nonetheless, government regulation of their activities gories-those connected with credit and those connected would be essential. At a minimum, as in all other coun- with taxes or subsidies. A third category is prices. But tries, rules to protect depositors would need to be de- indirect control through administrative determination of signed and enforced, with other rules and central bank prices has the drawback of being able for any particular intervention in the financial market to regulate the over- commodity to govern either demand or supply, but not all supply of money and credit. In addition, the govern- in general both, and hence having to be supplemented ment could-as for many years in Japan and South Ko- with administrative quotas either on demand (in the case rea-influence the allocation of funds in accordance with of shortages caused by a low price) or on supply (in the strategic economic objectives. This could be done in varn- case of surpluses caused by a high price). Moreover, re- ous ways, including channeling some budgetary savings duction in the scope of administrative price determina- through these institutions and subsidizing or taxing loans tion will gradually change prices-apart from their tax or for particular purposes, as well as by selective direct con- subsidy element-from instruments of government con- trols on interest rates and loan allocation. trol into independent indicators of value, cost, and scar- For China, continued experimentation and exploration city. These both influence and are influenced by the in the area of investment decisionmaking and financing actions of producers and consumers, usually in economi- will be necessary, especially since the experience of other cally appropriate directions. Government intervention, countries provides no precedent for a socialist financial however, may still be needed in some cases to dampen market. The relative shares and importance of the various fluctuations, as well as to set one very basic price, namely possible elements discussed above-vertical and compart- the exchange rate, which strongly affects exports and mentalized, as well as horizontal-should be allowed to imports in a decentralized economy. evolve with experience and the development of institu- Control of the overall level and growth of credit, which tions. In general terms, however, it seems likely and desir- depends not only on caution in financing budget deficits able that diversified investment arrangements will emerge by note issue or borrowing from banks, but also on as both consequence and cause of a diversified pattern of regulating the whole banking system (according to well- enterprise ownership. established rules), is a fundamental element of indirect macroeconomic management. And, as mentioned earlier, preferential interest rates and repayment terms in certain Planning sectors or regions, as well as direct government interven- Reform of the system of economic management along tion in the allocation of credit, are feasible and effective the lines of the Central Committee's 1984 decision will ways of influencing the composition of investment. But reduce the Government's direct involvement in produc- experience elsewhere suggests that they should be used tion, commerce, prices, employment, and so on, but cautiously, to avoid undesired side effects. Low interest should not diminish its responsibility or capacity to pro- rates on rural credit, for example, can lead to uneconomic mote and steer China's development. The preceding dis- mechanization in areas with surplus agricultural labor. cussion has touched repeatedly on areas in which contin- Personal income taxation, including progressive rates ued, and often enlarged, direct government involvement and selective relief, can influence the proportion of is needed, especially in establishing the physical, educa- household income saved and the composition of con- tional, scientific, institutional, and legal infrastructure es- sumer demand. Above-average indirect taxes on particu- sential for rapid and efficient growth. The Government lar goods (for example, luxuries) can reduce the amounts will also continue to intervene directly in some other key purchased and supplied by driving a larger wedge be- sectors and enterprises, though more in investment deci- tween the price paid by the customer and that received sions than in current operations, by the producer; the reverse is true for below-average 14 indirect taxes or subsidies (as on food, children's clothing, government agencies and between the center and local books and the arts in many countries). Taxes on enter- governments (who would still have their own, comple- prise profits, including selective concessions by product mentary, medium-term plans, at least at the provincial or locality, can be designed to regulate the total amount level). Other sorts of planning could also play an impor- of funds available to enterprises, the proportion of profits tant role. reinvested, and the amount of external finance used and One is what is called in other countries "policy plan- can stimulate investment in certain activities or regions ning." This is the design of packages of policies to achieve while discouraging it in others. Similarly, import tariffs specified medium-term objectives-for example, improv- can be used to promote increased production in particu- ing road transport, or increasing machinery exports, or lar sectors through import substitution (although without raising primary school enrollment. It is of particular value complementary export incentives they produce an ineffi- when indirect policy instruments are used, or when sev- cient bias toward the domestic market), and in certain eral dif'ferent government organizations are involved, or cases it is rational to tax exports. when different objectives and policy instruments are in- This range of instruments can enable planning and mar- terrelated. For this reason, policy planning is generally kets to coexist reasonably harmoniously, though it in- the responsibility of a high-level government agency, with volves greater emphasis than hitherto in China on manag- strong analytical capabilities, which can evaluate alterna- ing demand rather than supply. Direct control of supply tive approaches to an objective, make recommendations will, as mentioned earlier, continue in some important to the political leadership, and monitor implementation. sectors. But in most cases, rather than dictating what Anoi:her is location planning. Even with more econom- should be produced, the Government could use taxes and ically rational pricing of transport and urban land, most credits to guide the changing composition of demand and countries have found that zoning plans and regulations to influence the relative profitability of different sorts of are necessary to preserve urban amenities. Moreover, in production and investment, allowing supply to respond industries where a single optimal-size plant would supply through the decentralized decisions of enterprises and a high proportion of the national market (for example, farmers. automobiles or electronic chips), but many alternative Experience in other countries suggests that this indirect sites would be equally attractive in economic terms, the approach can be as effective as direct controls in shaping Government needs to arbitrate among competing locali- the pattern of development and often can be more eco- ties to prevent wasteful duplication. In other industries nomically efficient. But the relationship between objec- (cement and fertilizer, for example), government inter- tives and indirect instruments is much less precise and vention is required because the location and scale of all predictable, especially in the absence (as in China) of the enm:erprises need to be simultaneously determined to much systematic study of the behavior and responses of minimize transport plus production costs. economic units. This element of unpredictability can be In addition, Japanese experience suggests the potential reduced by research and by flexible and responsive imple- usefulriess of long-term sector planning, especially where mentation, with adjustment of policies to achieve desired major technological changes are expected or desired. (In objectives over the medium term. But it can never be China, such plans would provide a particularly useful eliminated, and many countries (including Hungary in framework for establishing industry-specific priorities for recent years) have experienced practical difficulties in in- modernization-a very difficult task because of the com- directly regulating their economies. The increased uncer- prehersiveness as well as the backwardness of the whole tainty associated with greater reliance on indirect controls industuial sector.) However, Japanese experience also sug- thus has to be weighed against the inefficiency of direct gests that such plans should not be designed and imposed controls. from above and must have competitiveness in interna- tional narkets as their not-too-distant objective. Through SCOPE OF PLANS. To the extent that direct control of close contact with the enterprises in an industry, sectoral production and allocation was reduced as part of the planners need to create a consensual 'vision" of future reforms, annual plans would become less significant in development, which increases the consistency (and re- themselves, although they (and the budget, with which duces :he risks) of investment decisions. This vision has to they should be linked) would remain important as a be backed up by, for example, government financial par- means of monitoring and adjusting medium-term plans. ticipat on in-or loans to-key investments, including pi- The medium-term plan, by contrast, would become the lot plants, or by temporary restrictions on imports and core of the entire planning process. Although the instru- assistance to exports. But enterprises that wish to pursue ments for its implementation could be different, the gene- a different strategy should often be allowed to do so: ral form of the plan document might be much the same as though this may seem wasteful, the flexibility it provides at present. Its preparation could serve even more than has pioved valuable, since neither planners nor enter- now as a focus for the exchange of views among different prises can pick technological winners consistently. 15 Long-term (ten- to twenty-year) plans in individual sec- trends, and major changes in the pattern of disease, will tors need not, and probably cannot, be coordinated in also pose new problems. Some changes and innovations detail. But they could benefit from a projected long-term in social policies and institutions may thus be necessary. framework for the whole economy-a broad and fairly aggregated "vision" of future growth and structural Income Dispanties change, worked out and regularly revised by central gov- Improving the allocation and motivation of skilled labor ernment planners. Such a plan could also provide guid- could involve some substantial increases in relative wages. ance to enterprises in industries without specific sectoral Even higher incomes could be earned-and in some in- plans, as well as for the planning of direct state invest- stances already are being earned-through entrepreneur- ment in infrastructure and natural resource development, ship. Although perhaps not yet fully accepted in China, and as a context for the medium-term plans. entrepreneurship can make an important economic con- In all these areas, although the common sense and good tribution. Entrepreneurs perceive and supply hitherto un- judgment of planners will remain fundamental, China met needs, initially at a high profit, thus luring others could make some significant technical advances in plan- into the same activities, which increases supply and drives ning methods, including greater use of quantitative profits down. If allowed to, entrepreneurs in China could models of various kinds. This could be assisted by im- also-as in other countries-play a major role in techno- provements in economic training (for both planners and logical innovation, whose risks may require the stimulus other relevant people). Increased use of computers could of high rewards. permit better and more flexible use of the large amounts In such cases, there is unavoidably some conflict be- of data now collected at lower levels. These, and other tween considerations of economic efficiency and the de- sorts of economywide and sectoral information, should sire for minimal income inequality. But this conflict can increasingly be disseminated to lower-level decision- be eased by progressive income taxation (tax rates that makers, including enterprises-who at present receive far rise with income level) and by high indirect taxes on less information from the Government than in many luxury goods. China already has a progressive personal other countries. income tax, aimed mainly at resident foreigners, which could be extended to cover, say, the top quarter of per- Social .ssue andPoliciessonal incomes. Although this would have administrative Social Issues and Pollcies D costs, it would probably make it easier for the central China has been outstandingly successful in reducing ex- government to prevent the informal income redistribu- treme poverty. Although there is substantial income in- tion that is now reportedly occurring in some localities. equality between urban and rural areas and among differ- International experience also suggests the desirability of ent rural localities, the hunger, disease, high birth and not attempting to conceal economically necessary income infant death rates, general illiteracy, and constant fear of disparities. In some countries, for example, bribery has destitution and starvation that haunt very poor people in been aggravated by keeping the wages of public officials other countries have been more or less banished. It will low for political reasons. In other countries, formal in- be of the utmost importance in the future to preserve and comes are fairly equal, but for important people there is build on this achievement. But will the policies and insti- access to shops providing goods not available elsewhere, tutions that have worked so well in the past meet the and sometimes even unrecorded but regular supplements social objectives of a middle-income China? And are they to official salaries. Since these and similar practices compatible with changes in the system of economic man- quickly become well known, their social benefits are tem- agement? porary, while they seem to have a high long-term cost in To a large extent, the answer to both questions is "yes." terms of corruption and the evolution of a society based Social ownership of land and other productive assets, on special privileges. widely dispersed access to basic education and health care, and a guaranteed minimum food intake should all Social Security contribute to maintaining much less inequality and pov- At the other end of the income spectrum, a social secu- erty than in other developing countries. Continued en- rity system consistent with a reformed economy is ur- couragement of agriculture and of other small-scale and gently needed. It will be increasingly difficult to reconcile labor-intensive activities should also-in light of experi- the need for greater enterprise independence and effi- ence both in China and elsewhere-play a vital role. ciency with the present role of enterprises as providers of However, economic reforms and greater technological social welfare benefits and lifetime employment. Prices dynamism could increase income disparities, while eco- and wages will be increasingly governed by economic nomically desirable changes in the price, wage, and em- considerations, which will make it much harder to use ployment systems could make it more difficult to use them to guarantee minimum urban living standards. these things as instruments of social policy. Demographic Faster structural change will involve contraction and clo- 16 sure of some enterprises. Some skills will become obso- including the self-employed as well as state and collective lete, and some workers-often through no fault of their employees. It could be hard, however, to extend the own-may become at least temporarily unemployed. same scheme to rural areas, because incomes are much In devising new institutions and policies to tackle these lower and vary widely among localities. A compromise problems, China might draw on the experience of the also always has to be struck among the desire to alleviate developed Western countries, all of which over the past existing poverty, the need to give people an incentive to few decades have established social security systems avoid poverty, and the wish to minimize budgetary and aimed at providing an adequate income for all citizens administrative costs. For example, setting unemployment and at eliminating poverty due to old age, illness, disabil- benefits high, though it reduces hardship, also reduces the ity, and unemployment. These systems have not been incentive to seek (or remain at) work, particularly if the free of problems, but have effectively attained their main alternat.ve to unemployment is a monotonous or ardu- objectives. All but a few developing countries, however, ous job at a low wage. have been deterred from introducing such comprehensive Especially if noncontributory income supplements social security systems by their financial cost and adminis- were chosen in preference to state-run social insurance, trative complexity. The challenge for China will thus be the statc should encourage collective and local develop- to devise a social security system that is effective, afford- ment of contributory pension and insurance schemes. able, and workable for a lower-middle-income country. Some of these pension and insurance schemes might, as In industrial countries, the social security system con- in other countries, be organized by enterprises and other sists mainly of a state-run insurance scheme, financed employers. But they should be financially independent primarily by compulsory wage-related contributions (or "funded"), so that, for example, a scheme would be from workers and their employers, which provides old- able to continue paying pensions even if the enterprise age pensions, unemployment benefits, sickness and ma- that hac. originally organized it were closed. Such pension ternity benefits, and so on to those who have contrib- and insurance schemes, which would have substantial uted. In addition, some noncontributory income funds tc invest, might play an important role in a socialist supplements are usually provided to households whose financial market. per capita income (including social insurance benefits) Becau.se many complex issues are involved, and be- would otherwise fall below some minimum level. This cause improved social security arrangements could be es- would be one possible approach for China, which could sential to a successful reform of the economic system, a be implemented by consolidation and extension of the commission could perhaps be established soon to formu- present enterprise-based labor insurance and welfare ar- late proposals. These might be implemented in stages, rangements. and should probably initially be of a modest kind. But the An alternative approach might be for the state in China long-term objective should be to maintain China's social to concentrate its resources on noncontributory income achievements, as in the past, well above those of other supplements to low-income households. These could per- countries at comparable income levels. haps advantageously be financed partly from the taxes on high incomes mentioned above, although some addi- Housing and Social Services tional contribution from general revenues might well be China's large and medium-size enterprises are now required. The total fiscal cost of these supplements could unique in the degree to which they provide housing and be substantial, but, because they would be targeted on education and health services to workers and their fami- the poor, would be a fraction of the cost of equivalent lies. As with social security arrangements, it would be price subsidies, which benefit rich and poor alike. Their much easier for enterprises to function as independent, administrative cost could be minimized, and their effec- efficient economic units if these obligations were re- tiveness maximized, by making use of existing institu- moved from them. This would enable managers to con- tions. In urban areas, there is the household registration centrate on economic activities, facilitate labor mobility, system used to administer grain rations. This might be a reduce the social problems caused by the release of un- particularly useful basis for providing income supple- needed workers and closure of inefficient enterprises, and ments to households with high dependency ratios that diminish the present reluctance of many workers to con- would otherwise on balance lose from the upward adjust- sider employment in small urban enterprises. ment of retail prices and wages mentioned earlier. In rural The social services now provided by enterprises could areas, the scheme could be implemented by the adminis- be taken over by local governments. So might some of trative successors to communes, brigades, and teams. the hotsing. But experience in China and elsewhere (in- In administering either state-run social insurance or an cluding Eastern Europe) suggests the advantages of giving income supplement scheme, China, like other countries, households as much responsibility as possible for their would face some difficult choices concerning coverage own homes, through individual ownership or housing and benefits. There might be a uniform urban scheme, cooperalives. This can provide a powerful stimulus to 17 personal saving, release government resources for other force, with shortages especially of younger, more recently purposes, and produce a higher quality and better main- trained, more adaptable workers. In addition, maintain- tained housing stock. The Government could usefully ing very low fertility could cause the proportion of el- assist this, as mentioned earlier, by selling more existing derly dependents (those age sixty-five or more) to bulge urban housing to tenants, as well as by "sites and ser- temporarily above its ultimate long-term level. vices" projects (planning and providing basic utilities for The increase in the proportion of elderly people will be new individually constructed housing), coupled with a serious social issue in the twenty-first century even with technical assistance and limited subsidies or tax conces- a more moderate reduction in fertility. It will increase sions to housing cooperatives. At the same time, univer- from about 5 percent today eventually to about 20 per- sal entitlement to a minimum standard of accommoda- cent. Though there will be an offsetting decline in the tion could be maintained, through both income proportion of young dependents, providing financial sup- supplements and direct provision. port to elderly people is more costly and more difficult to Increased central and provincial government support of organize. China's existing institutions, including the fam- rural social services seems to be needed, especially in poor ily, for supporting elderly people are unlikely to be able and backward localities. The introduction of the produc- to cope with their increased numbers. This strengthens tion responsibility system adversely affected social set- the arguments mentioned above for starting now to plan vices in such places, with a sharp decline in cooperative long-term reform of the social security system. The early health programs, as well as in primary and secondary establishment of wider and better pension schemes in school enrollment rates, especially among girls. This is rural areas would, among other things, make it easier to unfortunate, especially since powerful evidence from reduce fertility. other countries confirms that improvement of basic edu- cation and health is one of the most effective ways of Health helping poor people, as well as contributing to economic The aging of the population will also contribute to the growth. Thus although self-reliance may remain an ap- fundamental changes that are occurring in the nature of propriate policy in more prosperous rural areas, the state China's healh problems. These have already moved should consider financing a larger proportion of the costs from the pattern typical of low-income countries to one of social services in poor areas, for example, through where the leading causes of death are heart disease, matching grants. strokes, and cancer, as they are in high-income countries. These chronic diseases constitute a fundamental challenge Population for China, which other countries have had difficulty in A further reason for taking financial and other steps to meeting despite vast financial outlays. maintain and improve basic social services is that studies One essential ingredient of the required "second health in other countries have shown that both education- care revolution" will be prevention. Some steps-particu- especially of females-and health are major contributors larly control of salt intake and tobacco consumption- to voluntary fertility reduction. As noted earlier, the ex- could be taken straightaway. But much additional re- pected slow rate of population increase, due also to suc- search is needed to identify and field test appropriate cessful birth planning policies, is an important ingredient preventive strategies, which are much less straightforward of China's relatively favorable economic growth pros- than with communicable diseases. Another essential in- pects. Given, in addition, the limited amount of cultiva- gredient will be strategies for dealing with the large num- ble land in China, the achievement of replacement level ber of cases of chronic disease that will, inevitably, occur. fertility (2. 1 children per woman) is clearly desirable. These strategies should include capacity for treatment It might not be so desirable, however, to hold fertility where treatment holds promise of results and can be well below the replacemcnt level for a long time (which afforded; they should pay careful attention to affordable would be necessary, for example, to attain the ultimate plans for rehabilitation of incapacitated individuals; and population of 700 million suggested by some Chinese they should be concerned with humane care for the ter- demographers). This is because the gains from having a minally ill (an area in which there have been major and smaller population in the late twenty-first century could quite affordable advances in Western medical practice). be more than offset by the costs of transitional changes in A third ingredient of the second health care revolution the population's age structure. Of particular concern will be designing an insurance and financing structure would be the long-term impact of very low fertility on that encourages prevention and discourages current ten- the labor force, which could decline rather rapidly in the dencies toward overuse of facilities and introduction of middle decades of the twenty-first century. Negative la- high-cost procedures. Assembling these three ingredients bor force growth might have some economic advantages, will, inevitably, be difficult. But, to the extent that success but would inevitably cause problems of adjustment. is achieved, China could become a world leader in the There would also be a rise in the average age of the labor effective and humane handling of the burden of chronic 18 disease without succumbing, as other countries have, to ally require packages of complementary measures, whose endlessly costly investments in medical technologies of composition would vary according to the particular cir- limited efficacy. cumstances of the locality and other government priori- ties. Improved education and health services, in addition Rural Poverty to meel:ing immediate social needs, build up capacity to Increases in agricultural labor productivity should permit work and innovate. Focused research and extension ser- agricultural incomes to grow quite rapidly over the next vices can release untapped agricultural potential. Agricul- two decades. It is not easy, however, to find a plausible tural specialization and industrial development can be set of assumptions under which agricultural productivity made possible by better roads and other infrastructure, would grow fast enough to narrow the large agriculture- and as5isted by temporary subsidies for local industry. nonagriculture earnings gap (even in proportionate terms) Allowing part of the population to move elsewhere, thus until the twenty-first century. Because members of farm increasing per capita land and water availability, can also families will increasingly be engaged in nonagricultural be an efective and economical way of reducing poverty. activities, their total incomes will rise faster than their earnings from agriculture, but could still remain unsatis- Overview factorily low. Additional measures to raise them are worth considering. To sustain rapid growth for several more decades will be a Higher agricultural product prices and lower prices for hard and complicated task for China. Better access to agricultural inputs could conflict with other objectives of advanced foreign technology than in the past will make it agricultural pricing policy, including avoidance of surplus easier, but will not be sufficient: the cumulative economic production and encouragement of efficient input use. effect of individually inconspicuous improvements in The present two-tier pricing system, though, would in processes and products has been shown to be larger than principle make it possible to boost the incomes of farmers that of radical innovations, and the size of the technologi- by increasing offical procurement prices, while allowing cal gap between the best and the average enterprise in an market-determined above-contract prices to balance sup- industry to be as important as the technological level of ply and demand for the various agricultural commodities. the best enterprise. Nor will availability of energy, land, But this would be quite complicated, especially if unac- and other natural resources-despite their importance- ceptable inequities among farmers in different places and be the deciding factor. China's economic prospects will situations were to be avoided. Direct income supple- depend, rather, on success in mobilizing and effectively ments to broad categories of farm households might pose using al available resources-especially people. fewer administrative problems. This in turn will depend largely on success in reforming Increased government expenditure on things that the system of economic management, including coordi- mainly benefit the agricultural population, financed by nated progress on three fronts. First is greater use of taxation of other sectors, could provide another means of market regulation to stimulate innovation and efficiency. transferring nonagricultural productivity gains to those Second is stronger planning, combining indirect with di- who must remain farmers. Development of agricultural rect economic control. Third is modification and exten- and other rural infrastructure, improvements in agricul- sion oi' social institutions and policies to maintain the tural research and extension services, and support to rural fairness in distribution that is fundamental to socialism, social services are obvious examples. Another possibility despite the greater inequality and instability that market is subsidies to rural participants in social security schemes. regulation and indirect controls would tend to cause. Though generally low agricultural incomes are an im- On cach of these fronts, there are promising ways for- portant problem, the most serious rural poverty in China ward, but also problems and hard choices to be faced. In is concentrated in specific localities, where incomes are addition, it is hard to overstate the importance and diffi- far below the average. (For example, in Dongye town- culty of striking a correct balance among the three. Very ship, in Dingxi county, per capita income in 198 3-the few countries have combined state and market regulation township's best-ever year for agricultural production- in such a way as to produce rapid and efficient growth, was about one-sixth of the national rural average.) Many and fewer still have also managed to avoid intolerable such localities will benefit absolutely, and some relatively, poverty among substantial segments of their populations, from reform of the economic system-this appears to be On the contrary, there are far more countries in whichi the reason why the production responsibility system has unhappy combinations of plan, market, and social institu- not yet generated the anticipated increase in overall rural tions have produced neither rapid growth nor efficiency income inequality. But others will tend to remain very nor poverty reduction. poor and could even become worse off, especially if they There is thus a vital need to guard against losing the are remote or have meager land and water resources. strengths of the existing system-its capacity to mobilize To solve the problems of poor localities would gener- resources, as well as to help the poor-in the course of 19 overcoming its weaknesses. This will surely not deter progress in the urban economy. Though in many ways China-a successful pioneer in many areas-from moving more complicated and troublesome than rural reform, ahead. But it argues for a gradual advance, with experi- urban economic reform probably has the advantage of mentation and evaluation at each step, even though a not needing to be so uniform. Other countries have suc- one-stroke change would in principle involve fewer inter- cessfully applied different management methods in differ- nal inconsistencies. Experience in Eastern Europe also ent sectors and enterprises, and China should be able to suggests the importance of moving steadily and of trying do the same, while constantly seeking to refine and im- to avoid ill-judged steps in the direction of market regula- prove the mixture. tion that subsequently have to be reversed or administra- In system reform, and in the many other areas covered tively tampered with, thus creating needless uncertainty. by this report, both the potential for progress and the Not all the steps need be small, though. In some cases, problems involved are so large, and there is so much that despite the greater risk of error, it may be best to intro- is without historical precedent, that an even-handed and duce substantial packages of simultaneous reforms. Nor credible conclusion may be impossible. At a minimum, need progress be slow. What has been accomplished in though, China's long-term development objectives seem China's rural areas in the past few years has provided not attainable in principle, and if recent experience is any only an example, but also an excellent opportunity and guide, there is a good chance that they will be attained in indeed a vital need for complementary and similarly rapid practice. 20 Growtb and Change Over the next two decades, China-now a low-income percent: South Korea (6.6 percent) and Greece (5.2 per- country-will become a middle-income country. This cent). chapter explores possible implications for China's devel- More generally, only one country-Japan-has indis- opment strategy of international experience during the putably caught up with the developed nations from a low- to middle-income transition. Its three main sections position of economic backwardness. Several Latin Ameri- deal with macroeconomic change, microeconomic can countries, including Brazil and Chile, entered the change, and inequality. But first are some objectives- modern growth phase at least as early as Japan and have both for China and for this report. subsequently experienced periods of rapid growth, but still reriain far behind.3 Even the Soviet Union, with an Objectives exceptiDnally high rate of investment in physical and hu- man resources over many decades, has raised its per cap- China's ultimate economic objective is to catch up with ita GN]P in real terms to only about half the U.S. (and the developed countries, and as quickly as possible. The two-thlirds of the West European) level.4 impressive progress of the first three decades of the Peo- China's past record, nonetheless, is quite encouraging. ple's Republic, described in an earlier World Bank report From .952 to 1982, despite relatively rapid population (China: Socialist Economic Development, 1983) represents the first step toward that goal. The next step envisaged by Chinese leaders is to give most Chinese people a relatively 1. "A relatively comfortable living standard" is a rough trans- comfortable living standard and eliminate the worst man- lation cf the Chinese term xiaokang, which originated from a ifestations of poverty. ' Toward this objective, quadru- Chinese classic Li Ji. In that context the term was contrasted with pling of agricultural and industrial output between 1980 datong, xhich refers to a very rich society. Some people in China now cItssify the various stages of development in terms of im- and 2000 has been set as a target, with per capita GNP to proveme-nt in living standards, as follows: jihan (poverty), wenhao increase from about $300 to $800 (in 1980 dollars). (basic needs satisfied, which is considered applicable to the present Beyond this, the path is as yet uncharted. But the direc- situation in China), xiaokang (which is to be achieved by the end of tion of development during the next two decades will this century), and fuyu (rich). See Liu Guoguang, ed , Zbongguo powerfully and perhaps irreversibly influence China's Jingi F,zhban Zhanlue Wenti Yanjiu (Issues of China's economic economic prospects in the twenty-first century. development strategy) (Shanghai: Shanghai Renmin Chubanshe, The magnitude of the task ahead is readily illustrated 1984), p. 98. Inhreal (purcasning power) taskher is, G percapit inathed 2. At actual prices and exchange rates, the gap is much larger- In real (purchasing power) terms, GNP per capita in the a factor of thirty rather than ten. But the huge nominal income gap industrial countries is approximately ten times greater between rich and poor countries is apparently due more to differ- than in China2 and will increase in the future by at least 2 ences in prices than to differences in purchasing power, large percent and perhaps 3 percent per year. To catch up by though the latter are. See 1. B. Kravis and others, World Product 2050, China's GNP per capita would have to increase at and Income (Washington, D.C.: Vorld Bank, 1982), and China: an average rate of at least 5.5 percent and perhaps 6.5 Socialist Economic Development (Washington, D.C.: World Bank, percent per year (the Government's 2000 target implies a 1983), Mvlain Report, para. 3.32. rate of 5.0 percent). Such rapid progress has been rare 3. See, for example. Lloyd G. Reynolds, T09he Spread of Eco- nomic Growth to the Third WVorld: 1850-1980,; Journal of Eco- elsewhere. In 1960-82, excluding small countries and nomicLmterature, vol. 21 (September 1983), pp. 941-80. economies dominated by oil, only two developing coun- 4 This is the conclusion of several Western studies. Soviet tries achieved annual per capita income growth above 5 calculations suggest smaller gaps. 21 Figure 1.1 Composition of Aggregate Demand The future is inherently unpredictable, and the most in China and Other Developing Countries, 1981 careful forecasts are often confounded. The lessons of (shareof GDP) international experience are commonly either obvious (though frequently ignored) or ambiguous and controver- Foreign sial and in any event are hard to apply to China, a ,.z-borrowing country that in important respects differs from all others and is not easy for outsiders to understand. Thus, al- though this report is about planning in the broadest sense, its projections and calculations are not predictions, _ ,,-- ---____ 4 1 ! l l l l 1ll and its suggestions are no more than an attempt to con- , ,-___ | | | | 0 X tribute to the debate in China about the difficult ques- tions and problems that must be confronted during the country s unique socialist modernization. Macroeconomic Change China's future economic growth will require and cause changes in economic structure. At the macroeconomic level, the pattern of structural change associated with rising income has been broadly similar in most countries, ____----- and many of its features are likely to recur in China. But China Typical Typical because individual countries have deviated significantly large large from the average pattern, it is also important to consider low-income middle-income country' country' China's particular starting point and circumstances. Figures 1.1 to 1.3 summarize some essential results of Investment available research.6 Each figure juxtaposes three things: China in 198 1; a typical large low-income country; and a a. Data are from the period 1980-81. typical large middle-income country. The first figure deals Sources Annex 5, Table 3. 1. with the sectoral composition of domestic demand, the second with the sectoral composition of production, and the third with the sectoral composition of employment. growth and periods of acute economic mismanagement, Cbina's Present Economic Structure per capita national income grew at an average annual rate in International Perpective of 4.0 percent, with phases of significantly faster growth.' The figures show that China is in essence a low-income Of particular relevance is China's performance during the country with an unusual pattern of domestic demand and period of policy and system reforms that began in 1979 production. Output per agricultural worker is small and (described in more detail in World Bank Report 4072- agriculture's share of total employment is large-similar CHA, March 1983). From 1979 to 1984, largely as a to the typical low-income country. Output per worker in result of much accelerated agricultural growth, per capita manufacturing is higher than in the typical low-income national income grew at 6.8 percent per year. country, because of massive investment in heavy indus- The objectives of the present report are to identify try; with nonagricultural employment only 30 percent of some of the key issues and problems that China may face in the next twenty years and to review, in tight of inter- in the next twetyseasadtrvie,nlght5. State Statistical Bureau, Statistical Yearbook of Cbina, 1983 national experience, some of the options for addressing (Hong Kong: Economic Information and Agency, 1983), p. 6. them, with particular emphasis on decisions that might 6 The discussion in this section is based on Annex 5, which need to be made soon. As well as considering issues compares the present structure of China's economy with those of within specific sectors, such as agriculture and energy, other countries, and on Background Paper 9, which presents the and intersectoral linkages-the ways in which prospects results of empirical research on the structural changes normally and problems in each sector depend on circumstances and associated with the low- to middle-income transition in large coun- decisions in other sectors-the report ves special atten- tries. Figures 1.1-1.3 are derived from Tables 3.1, 3.6, 3.8, and n o t r e 3.9 of Annex 5. In the figures, per capita GNP is taken to be $ 300 tion to cross-sectoral or economywide Issues, including and $1,500 (1981 dollars) for the typical low-income and middle- human and financial resources, spatial location of eco- income country, respectively. In the Annex tables, the predicted nomic activity, and reform of the system of economic values for large lower-middle- and upper-middle-income countries management. are presented separately. 22 the total, however, the higher output raises China's per Figure 1.2. Structure of Production in China capita income only modestly.7 and Other Developing Countries, 1981 The most unusual aspect of demand in China (Figure (sbare oj GDP) 1.1) is the high share of saving-unique among low-in- come countries, well above the average for middle-in- come countries, and matched only byJapan and the East Mani-u European socialist countries. However, although house- faauilng hold consumption composes only half of China's low per _ - - capita income, basic living standards-as measured for struc:ure -' example by calorie intake, literacy, and life expectancy- are similar to those of middle-income countries. This has been achieved partly through a share of public spending (particularly on education and health) well above-and N better focused on basic needs than-that of the typical N low-income country, partly through imposing an austere but cost-effective pattern of household consumption. Al- though the share of food in household consumption ex- penditure is quite normal for a low-income country, sta- ples such as grain constitute an unusually high proportion China Typical Typical of food consumption. large large Especially because exports and imports are small in low-income middle-income relation to total output (though not unusually so for a country' country' very large country) and almost balance in both agricul- ture and manufacturing, China's unusual pattern of de- Note Inirastructure includes electricity, transport, and construction; services mand contributes to an unusual pattern of production includes housing. a. Data ire from the period 1 980-8 1. (Figure 1.2). The high saving share has its counterpart, Source Annex 5, Tables 3.8 and 3.9; World Bank, World Tables, 3d ed. because of the need to produce a large volume of invest- (Washington, D.C., 1984), vol. 2. ment goods, in a manufacturing share far larger than that of the typical low-income country and similar to that of the typical middle-income country. The share of agricul- ture in total production, by contrast, resembles that of in China is similar to that of the typical low-income the typical low-income country, simply because of the country, the composition of nonagricultural employment need to allocate a substantial proportion of limited re- is rather different (Figure 1.3). The share of manufactur- sources to feeding the population. The share of infra- ing in employment is significantly higher than in the structure (electricity, transport, and construction) is also typical low-income country, but by less than its share in similar to that of other low-income countries, despite output, since output per worker in manufacturing in China's large manufacturing share. China is unusually high (which in turn mainly reflects the The share of services in output is thus much smaller unusually low share of labor-intensive consumer goods (about 17 percent of GDP) than in the typical low-income within the manufacturing sector). The share of services in country (35 percent) or middle-income country (40 per- emplcyment in China is correspondingly lower than in cent). The share of education and health services is (as the typical low-income country. already mentioned) significantly above average. The low International comparisons of economic efficiency are overall service share thus mainly reflects the small shares difficult. But the available evidence (reviewed in Annex of commerce (including restaurants), banking and fi- 5) gives considerable support to the view that China's nance, and miscellaneous business and personal services. economy is relatively inefficient. Usage of intermediate This in turn reflects the small share of household con- material inputs-most conspicuously energy, but also sumption in national income and the high proportion of other industrial and agricultural products-seems high. food consumed directly by its producers, which tend to (Indeed, inefficient use of industrial intermediate inputs reduce trade (especially retail), as well as past official ne- appears just as important as the high saving rate in ex- glect of personal services. In addition, China's present system of economic management, and the high degree of enterprise and local self-sufficiency, reduce the provision 7. As explained in Annex 5, China's per capita GNP in 1981 is of (especially wholesale) trade, and financial and other $300 at Chinese prices and official exchange rates and is tentatively business services. estimat ed to be $ 3 50 at prices comparable to those in other devel- Although the share of agriculture in total employment oping countries. 23 Figure 1.3 Structure of Employment in China country levels of output per worker require about five and Other Developing Countries, 1981 times as much capital per worker as in a low-income (share of the labor force) country; to accumulate this capital rapidly requires a high investment rate, as for example in other fast-growing East Asian economies. structure s> China has, though, the option of supplementing do- -------- ~~~~~~~~mestic savings with external finance. In recent years, as Figure 1. 1 makes clear, China's balance of payments sur- _ _ ------ plus has added to the need for domestic savings-reduc- S. _ l l l oMining ">ing consumption below the already low level required to finance domestic investment. By contrast, the typical large developing country (low or middle income) runs an external deficit, with domestic investment exceeding do- mestic savings, thus easing the conflict between consump- tion and growth. Recent international experience, how- ever, has underscored the need-discussed further in Chapter 6-for caution and good judgment in external .--------- --------- borrowing, and the essential role of matching increases in China Typical Typical export earnings. large large low-income middle-income EFFICIENCY A high rate of investment (in education couor> ~~~countrvY country' country' and training, as well as physical capital), though neces- sary, will not be sufficient to attain China's growth tar- Note. Infrastructure includes electricity, transport, and construction; services gets. It must be accompanied by increased efficiency and includes housing. improved technology, which Wester economists cal a. Data are from the period 1980-81. prdc tivt ec Sources Annex 5, Table 3.6. growth of total factor productivity (TFP)-the difference between the rate of growth of aggregate output and that of some aggregate of inputs. Though different studies have often yielded signifi- plaining the large share of industry in China's total pro- cantly different estimates for particular countries and duction.) The data do not permit a firm conclusion re- time periods, most studies suggest that TFP growth has garding efficiency in the use of fixed capital, but the typically contributed about one-third of aggregate net amount of circulating capital used appears unusually output growth in middle-income developing countries, high. and nearly one-half in industrial market economies. In the Soviet Union, TFP growth has apparently been slower Possible Implicationsfor China's Future Development than in most industrial market economies and has con- Partly because China now differs in important respects tributed a smaller proportion of output growth. Rough from any other developing country, low or middle in- calculations for China suggest that TFP growth during come, the macroeconomic structural changes that occur 1952-82 was on average low by comparison with other during the typical low- to middle-income transition (sum- developing countries, though better in some sectors and marized in Figures 1 .1-1.3) are of only limited relevance subperiods than in others.8 to China's future development. Nonetheless, some impli- Research on the causes of variations in TFP growth cations-explored further in subsequent chapters-can be across countries and sectors and over time has had limited drawn from individual elements of this transition, the success, partly because of difficulty in quantifying poten- experiences of some other atypical countries, and China's tial explanatory variables (such as improved technology unusual starting point. and knowledge, increased division of labor, and incen- tives to economize on materials and capital and to use SAVING. China's saving rate is already so high-about workers more effectively). However, differences in TFP 30 percent of national income-by international stan- growth among sectors are apparently less systematic and dards that the usual low- to middle-income trend of a pronounced than differences among countries: in a fast- rising saving rate (Figure 1. 1) is unlikely to apply. In growing country, TFP growth in all sectors tends to be principle, indeed, China could reduce saving to a level- higher than in other countries. This-like China's experi- perhaps 20 percent of national income-similar to other developing countries. But this would be incompatible with China's long-tcrm growth targets. Middle-income 8. Annex 5, Table 3.16; Table 7.1 herein. 24 ence since 1978-confirms that the general orientation of those connected with housing, travel, and recreation, and policies and the system of economic management are with a disproportionate increase in expenditure on rele- crucial determinants of the pace of economic advance. vant services (as distinct from goods). The same finding-and much other evidence, including However, with any given level and distribution of in- China's own experience-confirms the need for a bal- come, the composition of household consumption varies anced allocation of resources among sectors. In particular, significantly among countries. This appears to be caused favoring industry at the expense of other sectors does not by variations not so much in national tastes (which influ- seem to generate rapid and efficient growth, even though ence only the details) as in the relative prices and availabi- industry is usually the fastest-growing sector and the shift lities of different items, which in turn vary partly because of employment from agriculture to industry contributes of natural resource endowments, partly because of gov- significantly to overall productivity growth. Countries ernment policies. For example, people in Japan, as com- that have stimulated agricultural production and incomes pared with those in the United States, eat less meat and have generally experienced faster, rather than slower, in- dairy products and are less well housed, mainly because dustrial growth. Similarly, even in countries where manu- land is scarcer, but also because imports of agricultural facturing production has grown very rapidly, infrastruc- products are restricted to protect Japanese farmers. In the ture and services-which are vital for industrial (and Soviet: Union and Eastern Europe, retail trade and other agricultural) efficiency-have absorbed a much larger services are a much smaller share of household consump- share of resources. For example, in South Korea in tion tnan in Western countries, mainly because planners 1965-74, manufacturing accounted for only 27 percent have allocated few resources to these activities. of total annual fixed investment, excluding housing, and China's future pattern of household consumption, in Japan in 1951-65, 30 percent. In China, the corres- though following certain universal trends, will thus to ponding share in the period 1950-80 seems to have been some degree accommodate itself to physical constraints, close to 40 percent (Annex 5, Table 3.11). as well as to government policies affecting the supply and prices of particular goods and services. Some of the Gov- CONSUMPTION. Like the domestic saving rate, the ernment's policy options (especially as regards consump- share of public consumption in national income in China tion of food, energy, and services) are examined in later is already above the normal level for a middle-income chapters. Naturally, at a given level of aggregate real con- country and will thus not necessarily follow the rising sumprion, if Chinese people are induced-by high prices trend observed in other countries (Figure 1.1). Indeed, it or restricted quantities-to consume less of one thing, could even fall, given the large weight of education in they must also be induced or allowed to consume more government expenditure, because of the large and highly of something else. But planners often find it difficult to atypical prospective decline in primary school enrollment identify items (other than basic necessities) whose con- in China over the next two decades, as a result of low sumpl:ion should be disproportionately increased. In most birth rates (see Annex 1). Increases in enrollment at market economies, by contrast, profit-seeking enterprises higher levels, improvements in educational quality, and seek continuously to produce consumer goods and ser- pressure for greater public expenditure on health, admin- vices whose consumption can be expected to increase istration, defense, and other public services will, on the unustLally rapidly (sometimes with the assistance of adver- other hand, tend to increase public consumption. Clearly, tising). Changes in China's system of economic manage- the Government has many options, whose budgetary ment (Chapter 10) may thus be needed, among other implications can only be assessed in relation to possible reasons, to enable household consumption to grow at the trends in subsidies, transfer payments, and public invest- same high rate as national income. ment-discussed in Chapter 9. But public consumption (much of which is in fact essential investment in human FOPEIGN TRADE. Choices connected with the level skills and knowledge) in the narrow sense of current pur- and pattern of foreign trade are less significant for large chases of goods and services seems unlikely to rise faster countries than for small ones, since the former generally than national income in the next twenty years. have rnore balanced resource endowments, higher inter- Since household consumption in China is probably nal transport costs, and domestic markets large enough to more concentrated on basic necessities than in the typical realize economies of scale in basic industrial activities. low-income country, it will almost certainly change, as China appears to be an exception to this generalization per capita income rises, in the same directions as it has only in respect of resource endowment: its 0. I hectares changed in other countries. These include a decline in the of cultivable land per person is one of the lowest ratios in share of expenditure on food, with an increase in the the world. The long-term possibility of net agricultural relative weight of nonstaple foods, especially meat; a imports paid for with manufactured exports-the pattern slight rise in the share of expenditure on clothing; and a observed in South Korea and Japan-thus requires careful sharp rise in expenditure on other items, most notably investigation. 25 But China's foreign trade options (Chapter 6) are sig- dence and specialization. And expansion of personal ser- nificant mainly because of the relationship elsewhere be- vices would reflect a shift in the balance between plan- tween foreign trade and economic performance. Devel- ners' priorities and people's preferences as criteria for oping countries-including Japan-that have become allocating resources. More generally, rapid and efficient sucessful exporters, especially of manufactures, have gen- expansion of the service sectors probably could not be erally also experienced faster growth. The causation runs accomplished through administrative directives and cen- in both directions, but exposure to trade seems to be a tralized resource allocation. International experience necessary ingredient of rapid and efficient growth. Thus, strongly suggests that growth of services should be pulled although the ratio of exports (and imports) to national by demand, with government in a permissive rather than product is typically no higher in large middle-income an active role. Individual and collective enterprises may countries than in large low-income countries, the possibil- also be better suited than state enterprises to providing ity of raising China's foreign trade ratio seems worth high-quality, flexible, and customer-oriented services. considering. SERVICE SECTORS. A declne in agriculture's share of Microeconomic Change production and employment is the most universal feature The changes in macroeconomic structure illustrated in of economic development, and one that will undoubtedly Figures 1.1-1. 3, and discussed in the preceding section, continue in China. The two issues are, rather, how fast are slow-at most, a few percentage points per decade. this decline will occur and what the relative sizes of the Moreover, these broad structural shifts are more a conse- various nonagricultural sectors will be. Especially as re- quence than a cause of growth. By contrast, microeco- gards employment, the two issues are intimately related, nomic structural change (affecting individual products, since, in the usual pattern of development, the main shift processes, enterprises, and workers) is typically much is from agriculture into services, with a more moderate faster, and its speed is a crucial determinant of the aggre- increase in employment in industry, which is less labor gate growth rate. intensive. The conspicuous smallness of the service sec- Microeconomic change contributes to growth mainly tors in China at present raises questions about their fu- by enabling cheaper and better products and processes- ture contribution: will they continue to be unusually in agriculture and services, as well as in industry-to small, or will they-as in the past few years-expand displace more expensive or inferior ones (Chapter 7). unusually fast? This involves introducing new technology-better seeds, China's small commerce, finance, personal and enter- faster machine tools, computerized inventory control- prise support service sectors, partly offset by a large edu- and abandoning obsolete products and equipment. It in- cation and health sector, constitute a pattern that also volves increasing specialization, which realizes technical emerges when the Soviet Union is compared with West- economies of scale and allows people to become expert in ern countries at comparable income levels. Neither in the particular sorts of work. It involves responding to needs Soviet Union nor in China can this pattern be explained and opportunities created by alterations in economic cir- by the "concealment" of service activities within other cumstances-energy and materials prices, the pattern of sectors.9 As in other countries, the share of employment demand, transport facilities, and so on. It involves innu- in services in the Soviet Union has risen with income, but merable small modifications to existing products, proc- has always been at an unusually low level, with a corre- esses and practices-in procurement and marketing, as spondingly above-normal share of employment in agricul- well as production-which cumulatively can reduce costs ture. There are variations also among developing coun- and improve the quality of products and services by strik- tries. But the striking contrast between the Soviet Union ing amounts. (and Eastern Europe) and virtually all other countries Much microeconomic change takes place within exist- (including Japan, which has always had a large service ing production units-farms and enterprises-sometimes sector) is more indicative of the range of options China involving radical transformation of the product mix. But has. This contrast also strongly suggests that rapid service sector expansion (discussed further in Chapter 2) and 9. Gur Ofer, The Service Sector in Soviet Economic Growth (Cam- reform of China's system of economic management are bridge, Mass.: Harvard University Press, 1973). See also Annex 5 closely interrelated options. Expansion of commerce of this report. Nor can this pattern be explained by the classifica- would go hand in hand with expansion of the role of tion in China and the Soviet Union of many services as not market regulation and increased specialization of produc- materially productive (and hence excluded from national income): the shortfall is in fact more conspicuous for material services (com- tnon units and localities. Expansion of enterprise support merce and enterprise support services) than for nonmaterial ser- services, including finance, accounting, and law, would vices (where the gaps in finance and personal services are offset by likewise be a corollary of increased enterprise indepen- education and health). 26 Box 1.1 Size and Growth of Enterprises in Market Economies The seemingly disorderly process of enterprise births and between profitability and growth; enterprises that are deaths and widely differing enterprise growth rates in most more profitable also grow faster. This is largely because market economies has a strong underlying element of evo- they depend mainly on retained profits to finance invest- lutionary natural selection: enterprises that provide ment needed for expansion, but also because higher profits cheaper and better products or services, and market them make it easier to borrow and issue new shares. Since more effectively, expand at the expense of their less effi- greater profitability generally reflects greater efficiency, this cient rivals, which are forced to improve or to disappear.' financial linkage between profitability and growth plays a Studies-mainly of U.S. and British enterprises-have also crucial role in enabling efficient enterprises to expand at established that this process exhibits certain statistical regu- the ex pense of less efficient ones. Success and failure, larities.2 Small enterprises are a large and stable proportion moreover, are persistent: enterprises with above-average of the total (whose overall size distribution is approxi- growth and profitability in one (say, five-year) period are marely lognormal). On average, small enterprises grow just likely also to have above-average performance in the next as fast, and are just as profitable, as large enterprises, but period. But the evolutionary process takes time: in every small enterprises display much more diversity of growth sector, there is always a wide spectrum of enterprise profit and profitability, which means that the most profitable and and lo6;s rates. fastest growing enterprises, as well as the least profitable and slowest growing (or contracting) enterprises, are mainly the smaller ones. As a result, a significant propor- 1R Several relevant studiei are Srh and Productivity Differenceso tion of initially very small enterprises are rapidly propelled Journal of Ec literature, vol 19 (September 198 1), pp. 1029- into the large enterprise category, while an equally signifi- 64. cant number collapse altogether. 2. Many studies are surveyed in Robin Marris and Adrian Wood. There is also a strong association, independent of size, eds., T1se Corporatm Economy (London, 1971), Appendixes A-C. microeconomic change is also associated with the birth of 1.1). Individual enterprises have expanded at widely vary- new enterprises, marked changes in the relative sizes of ing rates; some have merged, diversified, and changed existing enterprises, and the disappearance-through clo- their product mix, such as one Shanghai engineering en- sure or merger-of particular production units (Box 1.1). terprise, whose expansion from a repair shop to a manu- For individual workers, change may be less rapid, partly facturer-first of gear lathes, then of welding equipment, because workers with obsolete skills retire and are re- then of electrical cable equipment, then of pumps-re- placed by new entrants with new skills, partly because semb,es the history of many engineering firms in other processes change less than products, and partly because countries. Some enterprises, including many small ciga- changes in enterprise product mix reduce the need for rette and machinery manufacturers, have been closed workers to move among enterprises. But microeconomic dowr, especially in recent years. change inevitably entails considerable relearning of skills But: there have also been some basic differences be- and movement of workers among jobs and-in most tween China and most other countries in the area of countries-enterprises (Chapter 8). It may also affect the micrceconomic change. Product and process innovations relative prosperity of different localities-sometimes in a by existing enterprises appear to have been less frequent; persistent and even cumulative way-and thus create the trend of increasing specialization has been much more pressure for geographical movement of people and capital muteA; and producers have rarely sought new customers (Chapter 5). and markets (for example, smelters in China, unlike those The ingredients of microeconomic change described in the United States, normally produce only pure metals, above can be discerned, in some form or to some degree, with users having to make their own alloys). Changes of in China over the past three decades. Many new products product mix, diversification, merger, and spinoff of new and new processes have been introduced. New enter- enterprises seem to have been less common. Enterprise prises have been born at high rates, especially since 1970 closu,e rates have been low. Workers have generally re- with the expansion of commune and brigade enterprises mained employed by the same enterprise throughout and (more recently) individual enterprises. Largely as a their career. result, the proportion of small enterprises in China is Th- competitive process by which better and cheaper quite similar to that in a typical developing or industrial products drive out worse and more expensive products country, and far higher than in Eastern Europe (Table likewise seems to have been slow in China. Especially in 27 Table 1.1 Size Distribution of Industrial Enterprises in Selected Countries (ercent) United United Soutb Size of China, Kingdom, States, Korea, Japan, India, Yugoslavia, Hungary, enterprise 1982 1979 1977 1981 1972 1976-77 1981 1981 5-33 employees 59.2 65.2 56.4 70.6 80.2 51.7 6.6 2.2 33-75 employees 19.5 15.7 20.3 14.4 10.7 35.3 15.8 4.8 75-189 employees 12.2 10.8 12.4 9.2 6.1 7.8 32.1 18.7 189-243 employees 8.5 1.4 3.8 1.5 0.8 0.8 12.0 9.2 Morethan 243 employees 0.6 6.9 7.1 4.3 2.2 4.4 33.5 65.1 Note Data are percentages of the number of enterprises with five or more employees in the country. Sources For China, estimates based on data in the State Statistical Bureau, Statistical Yearbook of Cbina, 1983 (Hong Kong: Economic Information and Agency, 1983); for other countries, various national sources. the state sector, backward enterprises have been pro- In all countries, the potential for "extensive growth" tected, and progressive enterprises held back-sometimes through duplication of existing production is limited, no deliberately in order to maintain profits, capacity utiliza- matter how much capital is accumulated, by shortages of tion and employment in less efficient enterprises. In re- natural resources and (ultimately) of labor and by the cent years, for example, despite general agreement that unwillingness of consumers to buy ever-increasing there are too many bicycle producers, the output of en- amounts of the same goods. To attain a high level of terprises producing good-quality bicycles has been re- economic development thus requires "intensive growth" stricted by the Ministry of Light Industry in order to based on innovation, cost reduction, and product im- protect the producers of inferior bicycles. Similarly, a provement; this in turn requires much faster microeco- cotton mill in one province was recently obliged by the nomic change. With close to 70 percent of the labor provincial authorities to stop sending its cloth elsewhere force still employed in agriculture, extensive growth for dyeing and printing, because the province's own sub- alone could probably carry China up into the middle- standard printing and dyeing enterprise was underuti- income range over the next two decades or so, albeit at lized. Subsidies to keep inefficient plants and old equip- rather high cost. But, over the longer term, intensive ment in operation are common, partly because growth will be essential to China's catching up with the enterprises play a key role in providing housing and social industrial nations, and the experience of other countries services. such as the Soviet Union has shown the difficulty of Chronic shortages, strong emphasis on production tar- switching to intensive growth at a late stage of develop- gets, small price differentials between high- and low-qual- ment. ity products, and inadequate incentives have made prod- Faster microeconomic change in China will require- uct quality improvement, innovation, and cost reduction above all-changes in the system of economic manage- occasional, rather than everyday, concerns of Chinese ment, many of which have already been implemented or enterprises. The former policy of local self-sufficiency and proposed.10 System reform is thus a pervasive theme of deficiencies in the material supply system have led local this report: the general treatment in Chapter 10 builds on governments to construct many plants of suboptimal the discussion of specific aspects of system reform in scale and inferior product quality to supply local needs other chapters, especially those on technology, spatial is- for commodities such as steel. (This may explain why, by sues, and human and financial resources. In all chapters, comparison with all the other countries in Table 1.1, the need for difficult choices will emerge; this is partly China has so few enterprises in the largest size category.) because, beyond a certain point, the objective of faster Deliberate and incidental obstacles to microeconomic growth, with which this chapter has so far been solely change of course exist in other countries. Declining indus- concerned, conflicts with the objective of equity in in- trial sectors are protected by trade barriers; large firms in come distribution. difficulty are helped out with loan guarantees and subsi- dies. Monopolistic practices, distorted prices, corrupt or incompetent bureaucracies, and inadequate retraining fa- cilities, too, retard the speed and efficiency of develop- ment elsewhere. Nonetheless, the obstacles to microeco- mnom clsewha . ioneChinasee toh o bs e un usuall 10. In addition to the major reforms already carried out in rural nom vc change in China seem to have been unusually areas, and numerous experiments in urban areas, a general strategy substantial and comprehensive and could-if allowed to for reform of the urban economy is contained in the Communique persist-prove incompatible with China's objective of of the Third Plenary Session of the Twelfth Central Committee of sustained rapid growth. the Chinese Communist Party, October 20, 1984. 28 Inequality rates, general illiteracy, and constant fear of destitution The distribution of income-and of lving standards more and starvation that haunt very poor people in other coun- generally-is of critical importance fr a stries have been more or less banished from China. Life generally-IS of critical importance for a socialist country epc:ny hs eedneo ayohreooi suc as Chin , whs ceta ecnoi obetv is no expectancy, whose dependence on many other economic such as China, whose central economic objective IS not and social variables makes it probably the best single merely rapid growth, but growth whose benefits are indicator of the extent of real poverty, is on average in widely spread. Considerations of equity will inevitably China outstandingly high (sixty-seven years in 198011) play a central role in shaping China's future development ri statg an sytmreomor a low-income country; even in the p)oorest province, strategy and system reform. life espectancy is not far below the average for middle- Inequality in Cbina in International Penpective income countries. China's past strategy and present system have created, on 'NEW INFORMTION. Since the first World Bank eco- the whole, an extraordinarily equal society. The first World Bank economic report on China dealt with this at normc report on China was written, substantial addi- tional information has become available and further stud- length (Main Report, paras 3.46-3.102); its conclusions, ies made. The preliminary results (from a 10 percent which are essentially unaffected by more recent informa- sampl.) of the 1982 census have been published, as have tion and developments (to be described subsequently), more household survey results. Health nutrition and merit summary recapitulation. Inequality of urban incomes is uniquely low in China, populaton Issues were exammed by the World Bank m with virtually no extreme poverty in urban areas. There China The Healtb Sector (1984). In general, this material is, however, a large gap-comparable to that in other has reinforced the conclusions of the first report. But a developing countries-between average urban and rural few changes of emphasis are required. incomes,which in te rms of living standards is magnifie, The distribution of income in rural areas is less unequal incomes, which in terms of living standards is magnifed, than the World Bank's first estimate (based on fragmen- as in some other countries, by concentration of social services on urban residents. In addition, there is substan- tary ata) had suggested. Rural household survey dat a, . . ' . . ' ...................summ arized in Table 1 .2, show the Gini coefficient in tial inequality of rural incomes in China, largely because 199t ave Tble 0.26, rher than0 Thiseans of wide differences in the quantity and quality of agricul- that rural income inequality in China is significantly less tural land per person, coupled with tight restrictions on than ln other South Asian countries (with Gini coeffi- geographical mobility.tavnohrSuhAincut's(ihGn ofi geogr apcal moi it y cients of 0. 30-0.3 5), rather than in the same range. Thus because China is a low-income country, and be- The same data, however, underscore the importance of cause of substantial rural income inequality, a large mi- geographical differences in rural incomes. Table 1.2 al- nority of the population have extremely low incomes. These people, however, have a much higher standard of living than their counterparts in most other developing 11. World Bank estimate based on the 1982 census and other countries. Agricultural collectivization has prevented the recently released demographic data: see Kenneth Hill, "China: An emergence of a class of impoverished landless laborers; Evaluation of Demographic Trends-1950-82,' Technical Note the state guarantees.ainimumfoodspply r DEM 4 (Washington, D.C.: World Bank, Population, Health, and thrsat uu fd s y pri y Nutriton Department, 1984), which is Supplementary Paper I of school enrollment is high; and basic medical care and Dean r. Jamison and others, Cbina: The Health Sector (Washing- family planning services are available to most people. As a ton, D.C.: World Bank, 1984). The estimate for 1979 in the first result, the hunger, disease, high birth and infant death economic report (sixty-four years) was too low. Table 1.2 Rural Income Distribution, 1979-82 (percentage ofpopulation) Income National average Gansu, Jiangsu, per capita 1979 1980 1981 1982 1982 1982 Less thanY 100 19.6 10.1 4.9 2.9 21.1 Y 100-200 54.0 53.2 39.5 25.6 45.2 16.8 Y 200-300 20.7 26.0 36.2 39.3 22.5 36.2 Y 300-400 4.2 7.7 13.5 20.2 8.2 27.6 Y 400-500 1.1 2.2 4.0 7.4 2.3 11.4 More than Y 500 0.3 0.9 1.9 4.6 0.7 8.0 Gini coefficient 0.257 0.237 0.231 0.225 Source: Estimates based on household survey data (Annex 5, Appendix 1). 29 lows comparison of the rural income distribution for the income inequality in rural areas. The statistics in Table whole of China with the distributions in a relatively rich 1.2, although they may have various biases (Annex 2, province (Jiangsu) and in a poor province (Gansu); the para. 1 .34), suggest that so far this has not happened and differences are extremely striking, with four-fifths of the that if anything rural inequality has declined. This could people in Jiangsu in the Y 200-500 or more range, and be due to increases in inequality within localities having two-thirds of the people in Gansu in the range below Y initially been more than offset by reductions in income 200. Moreover, because Chinese provinces are large and differentials among localities: the reforms spread fastest in internally diverse, these provincial statistics do not ade- poor and backward areas, some of which had also been quately convey the range of incomes among localities. especially disadvantaged by the former policy of concen- For example, Gansu's average rural per capita income in trating on grain production even when other activities 1983 was Y 228 (as compared with the national average would have been more economic. In any event, the lack of Y 310); but in Gansu's poorest county, Dingxi, the of deterioration in the distribution of rural incomes has average was only Y 108; and in one of Dingxi's poorest meant that the sharp rise in average rural incomes has townships, Dongye, it was only Y 55-and this was the greatly reduced the number of households with very low township's best-ever year for agricultural production. incomes. On the basis of the data in Table 1.2, and a Census data on educational attainment by sex and age poverty line based on food intake requirements of 2,185 group (Table 8.1) confirm that China's performance in kilocalories per day, it is estimated that the proportion of basic education surpasses that of most low-income coun- the rural population in poverty declined from 31 percent tries. Moreover, they clearly reveal the adverse effect of in 1979 to 13 percent in 1982 (Annex 5, Table 1.7). the Cultural Revolution on the proportion of younger Though the latter figure still means 100 million people, people with postsecondary education. The same data and though people in places like Dongye township have also show a persistent (though narrowing) gap between gained little, the speed and scale of the improvement is male and female educational attainment. This, together probably unprecedented in human history. with the census data on occupation by sex, suggests that In other respects, however, the rural reforms have had the World Bank's first report may have overstated the some unfortunate consequences for inequality. One has degree to which China has reduced inequality between been a sharp decline in rural cooperative medical insur- men and women, though it is still less than in most ance, financed from collective welfare funds, to which developing countries. individual households are reluctant to contribute. This has been particularly pronounced in poor and backward RECENT DEVELOPMENTS. The data recently released areas (in richer provinces, the swelling profits of com- also allow an examination of trends over the past five mune and brigade enterprises have by contrast led to years-a period of important policy and systemic improvements in collective medical and other social ser- changes, including above all the replacement of collective vices). The drop in the proportion of production brigades agriculture by household farming. This, in conjunction covered by cooperative health insurance from 85 percent with major increases in procurement prices and rapid in 1975 to 58 percent in 1981 (with further declines since development of rural nonagricultural activities, has then) thus represents a considerable step backward, al- caused an astounding surge in agricultural production though the Government is making efforts to upgrade the (more than 7 percent per year since 1978) and rural quality of primary medical care by, for example, giving incomes-up in real per capita terms between 1979 and barefoot doctors more training. 1983 by about 70 percent (Annex 2, para. 1 .33). Urban Similar problems have arisen in basic education, mainly wages and employment have also increased and urban because the household responsibility system has induced staple food prices have been held constant, which has more rural parents-as in many other countries-to keep raised real urban per capita incomes by about 40 percent their children out of school to work on the family farm. over the same period, but the urban-rural income gap has As shown in Figure 1.4, the proportion of primary still narrowed significantly. Rising agricultural production school age children enrolled (the net enrollment ratio) has also caused a sharp increase in food consumption, dropped in 1980-82, although subsequent government which, at about 2,700 kilocalories per person per day in efforts seem to have reversed the decline; the gross enroll- 1982 (Annex 2, para. 2.03), compares favorably with ment ratio at the secondary level has dropped from about levels in most middle-income countries. Nonetheless, 46 percent to about 30 percent, below the developing- food consumption in many localities still falls far short of country average of 39 percent (Annex 1, paras. 2.04- the average, and the composition of the Chinese diet is 2.11). The declines have been particularly marked in still that of a low-income country. poorer areas, where enrollment ratios were already unsat- It was generally anticipated that the rural reforms, by isfactorily low (and actual attendance rates lower still); in making the incomes of individual households more de- richer areas, such as Jiangsu, rural parents see clearly that pendent on their effort, skill, and luck, would increase their children will not be able to join the rapid exodus 30 into nonagricultural employment without a good educa- making diligent workers and successful individual busi- tion. The declines in enrollment have also been particu- ness proprietors richer, and the idle or unlucky poorer. larly marked among girls. Increa,ed interlocality specialization and trade will surely benefit some localities more than others. Possible Future Trends in Inequality In addition, China may be affected by the forces that China may not in the future be able to achieve such a low have t:ended to increase inequality during the low- to degree of inequality as in the past-and indeed, the Gov- middle-income transition in other countries. The incomes ernment's position is that past strategy was excessively of the poorest 40 percent of the population have usually egalitarian. The rural responsibility system, though thus increa,ed more slowly than the average; the gap between far it has increased inequality only through its impact on agricultural and nonagricultural earnings has widened; education and health care, will probably also start to and interregional income disparities have increased.'2 At increase income inequality before too long. Systemic higher income levels, though, all these trends have usually changes in urban areas could also increase inequality, by been reversed, as the modern sectors have come to domi- nate the economy and people have moved in search of higher incomes. And even in middle-income countries with an unusually high degree of inequality, most people have a much higher standard of living than in a low- Figure 1.4 Enrollment Ratios in Formal Education in income country. China, 1965-8 3 and Targets for 2000, and in Other China will also have to deal with the losses that micro- Developing Countries, 1965-80 economic change may inflict on certain families, occupa- tions, enterprises, and localities-even though these may Enrollment ratio (logarithmic scale) not bL visible in aggregate measures of inequality and 100._ _ _ _ even though the gains of other groups may hugely out- 70. __ l___ Prrmaryyeducation- weigh these losses. An example would be a backward enterprise suffering from competition from cheaper or better goods. Governments in all countries, especially 3 Secondary ed lucation socialist ones, feel impelled to reduce the social and politi- 30/0 cal co,ts involved. Yet protecting potential losers, or rou- 20.0 ( 7 tinely compensating them, also has a high cost, because protection and compensation can slow microeconomic lo o change and reduce growth. Ways of easing this dilemma .____ _____ ___'_ will be discussed in subsequent chapters, but some com- 7.0 _ _ -e d_ _ promise between conflicting objectives is unavoidable. 4_0°_. _ H _ Despite all this, if appropriate policies are followed .~O - -7(experience in other countries suggests the importance of 3.0- support for agriculture, labor-intensive industry and ser- 2_0 ___ __-___vices, and small-scale economic activities) and if impor- /University education tant features of the past system are preserved (including social ownership of land, widely dispersed access to basic 1._ t_7_education and health care, and a guaranteed minimum 0.7 \ / food Lntake), China should be able to maintain a much o.s__ 7lower degree of inequality than in the typical developing 0.4 t /country. In addition, the new dynamism of the rural 0.3 \ / _ sector could further narrow the urban-rural income gap. 0 2 \ / And in the longer term, regardless of the precise pattern of distribution, the great majority of China's people will unquestionably gain substantially from rapid growth- although special attention will have to be given to people 1965 1970 1975 1980 1985 1990 1995 2000 and localities excluded by isolation or inadequate re- ---China sourcos from the general process of development. Other developing countries Notes The enrollment ratio expresses the number of students in a level of 12. See, for example, World Bank, World Development Report education as a percentage of the relevant age group. 1980 (New York: Oxford University Press, 1980), Chapter 4; and a. Targets for nonuniversity higher education in 2000 are not availableBakondPpr. Source: Unesco, Statistical Year Book; World Bank data. Backg ound Paper 3. 31 Illustrative Projections To explore further China's long-term prospects for simultaneously and in an internally consistent way, with growth and structural change, discussed in the previous allowance for the many accounting linkages among them. chapter, and to provide a consistent numerical frame- To make a projection requires many assumptions, both work for the discussion of other topics in subsequent about the future path of the numerous variables that are chapters, projections to the year 2000 have been made in the model, but not determined by it (examples of such with an economic model. This chapter provides a simple "exogenous" variables being population and energy con- description of the model, explains the main assumptions servation rates in particular sectors), and about the nu- underlying three alternative projections, and summarizes merous causal-as distinct from accounting-linkages their results. (A fuller, technical account of the model and governing the variables that are determined by the projections is in Annex 4. The underlying data, including model. By comparison with some recent models, the an estimated input-output table, are in Annex 5.) assumed causal linkages in this model are simple and mechanical and not based on any particular theory of Description of the Model household, farmer, or enterprise behavior. This makes them less precise and complete, but also easier to grasp The model is a simplified representation of China's whole and less controversial. economy, broken down into twenty sectors. There are two agricultural sectors (crops-including forestry-and Growth animal husbandry-including fishing), four energy sectors Growth of national income in the model depends mainly (electricity, coal, oil extraction, and oil refining) and seven on the level and efficiency of investment. The level of other industrial sectors (metallurgy, chemicals, machin- investment is determined by the proportion of national ery, building materials, food processing, textiles, and income saved and (to a much lesser extent) the amount of other manufacturing), as well as construction, transport, foreign borrowing, both of which are treated as govern- commerce, education and health, public administration ment policy decisions. Investment is added to the existing and defense, miscellaneous services, and housing. The last capital stock, which increases production capacity in ac- four sectors are not regarded as "materially productive" cordance with assumed linkages in each sector between in China; they are included partly to permit calculation of capital stock and gross output (sectoral capital-output ra- Western-style national accounts, but mainly to make it tios). Because these ratios are not the same in every sector easier to handle some important components of invest- (a unit of capital adds more, for example, to the gross ment, public expenditure, and household consumption. output of textiles than of metallurgy), the impact of in- Chinese-style national accounts are also generated by the vestment on growth depends partly on its allocation model, however, and in most respects the sectoral classifi- among sectors. It also depends on the ratio of net output cation corresponds with that used in Chinese statistics.' (that is, contribution to national income) to gross output For each of the twenty sectors, the model has produc- in each sector, which in turn depends on efficiency in the tion, use of materials, employment, investment, and use of materials, energy, and other intermediate inputs. prices. And for each sector, several sources of demand are distinguished: intermediate purchases by other sectors, household consumption, public consumption, exports I. Except that brigade enterprises are included in the relevant (less imports), and use for investment. The model projects industrial sectors rather than in agriculture and that transport in- all these variables-about 600 in total-into the future, cludes both passenger and freight. 32 Increased production in the model requires not only As regards final demand, the division of national in- more capital and more materials, but also more labor. come between saving (the main determinant of demand Sectoral labor requirements, and hence employment lev- for investment goods), public consumption, and house- els, are governed by assumed growth rates of labor pro- hold consumption is treated in the model as a direct ductivity. But the overall growth rate of the economy is government policy decision, as is the composition of pu- not constrained by availability of labor. This is partly blic consumption (mainly for education, health, public because skilled labor is not distinguished from unskilled administration, and defense). Household consumption is labor, partly because there is assumed to be (over the next spread among sectors by assumed "income elasticities" two or three decades) a pool of surplus labor in agricul- that cause the share of food to decline and the shares of ture, which can be drawn into employment in other other goods and services-including housing-to increase sectors with no reduction in agricultural output. at varving rates as income rises. The Government can, of The determinants of growth in the model are a great coursc, influence these elasticities by taxes and subsidies simplification of reality. In particular, the sectoral capital- or by rationing or other administrative means, but these output and net-gross output ratios summarize and con- policy instruments are not included in the model. ceal a host of important underlying factors-product and Exports (less imports) are also a significant component process innovations, the quality of investment decisions, of final demand in several sectors. They are determined the training and effective use of skilled labor, and pres- over tie longer term in the model mainly by assumptions sures on enterprises to improve product quality and to regarding the openness of each sector (the ratio of exports use less materials, labor, and fixed and circulating capital. or imports to sectoral production), the degree of self- Growth in the model is also smooth. In particular, it is sufficimncy in each sector (the sectoral trade deficit), and never held back by bottlenecks or shortages in individual the composition of manufactured exports-all of which sectors (such as metallurgy or transport), partly because are treated as government policy decisions, although the imports and increased use of existing capacity are as- instrunents necessary to implement these decisions are sumed to provide short-term flexibility, partly because not explicitly included. Sectoral exports and imports are capacity shortfalls in specific sectors are assumed to be also influenced in the model by the need to equate the remedied through increased investment within a year or overa]l outflow and inflow of foreign exchange, given the two. These assumptions, though technically convenient, Government's chosen amount of foreign borrowing. and reasonable for analyzing growth over a twenty-year About half of total demand is not final, but intermedi- period, make the model unsuitable for analyzing me- ate. For any given sector, growth of intermediate demand dium-term (say, five- to seven-year) growth prospects. depends partly on growth of production in the other sectors that use its output as an input (for example. inter- Structure mediate demand in metallurgy is governed largely by Except for agriculture and energy, where natural resource machinery production), partly on changes in the pattern constraints are important, the structure of production- of intermediate use within those sectors (for example, the relative sizes of the different sectors and the way in reduc.ions in the amount of metal per machine). The which these change as the economy grows-is deter- compmsition of intermediate demand in the model thus mined in the model by the structure of demand. This is depends on all the factors that govern the structure of because the allocation of investment among sectors in the final demand, plus the assumed pattern of constancy or model is governed by the expected growth of demand for change in input-output coefficients, which may be influ- their output: in any given year, demand and production enced in various ways (not explicitly modeled) by govern- capacity in particular sectors may not balance,2 but sec- ment policies. tors where there are shortages, and where the trend growth of demand is more rapid, receive larger shares of total investment, and hence experience faster growth of Three Alternative Projections production. Since a large number of assumptions, both about the The model's basic assumption that the long-term struc- future values of exogenous variables and about specific ture of production is determined mainly by the structure of demand means only that the Government's produc- tion policies will be based mainly on assessment of needs. 2. In sectors where there is foreign trade, temporary demand- These production policies may be implemented directly, produ.-rion imbalances are accommodated in the model by adjust- by quotas and administrative decisions, or indirectly, by ment of exports or imports, in other sectors by temporary changes using economic levers to regulate the pattern of demand in capacity utilization (reflected in temporary deviations of sectoral capital-output ratios from their assumed long-term values). De- and allowing farmers and enterprises to respond. The mand-production imbalances do not affect prices These assump- model is compatible with both methods of implementa- tions are made more for simplicity than for accuracy, since the tion. model is not intended for short-term analysis. 33 causal linkages among variables, are required to make a duced use of metals (especially in machine building). Total projection, there is a colossal number of possible alterna- nonenergy intermediate requirements per unit of gross tive sets of assumptions (and hence different projections), output in individual sectors (other than agriculture) were even if each individual assumption is varied only within a assumed to remain approximately constant, as in most limited, plausible range. After many experiments, three other countries. have been selected for discussion (some variants will also The division of consumption between public and be mentioned). household consumption was assumed in QUADRUPLE to These three projections-named QUADRUPLE, MODER- stay the same as in 1981, as was the proportionate alloca- ATE, and BALANCE-have no monopoly of merit, but are tion of public consumption among sectors. The share of useful for illustrative purposes. Although all of them lie household consumption allotted to food declines, despite within the range of international (including Chinese) ex- a steep increase in expenditure on animal products (see perience, they span a fairly wide range of possibilities, Chapter 3 and Annex 2). The share of clothing in house- both for policy and for factors beyond government con- hold consumption was assumed to rise slightly, and the trol. They also reveal a number of specific questions and shares of other manufactured goods, electricity, transport, problems, which are addressed later in this report. The miscellaneous services, and housing to increase signifi- present section outlines the assumptions of each of the cantly-broadly in line with average international experi- three projections. The rest of the chapter describes their ence. Household consumption of fuel, though, was as- results. sumed to increase more slowly than income, because of improvements in thermal efficiency (see Annex 3). QUADR1JPLE Other assumptions of QUADRUPLE include growth of The QUADRUPLE projection attains the Government's crop production at an average annual rate of 3.6 percent, target of quadrupling the gross value of industrial and doubling of crude oil production (in accordance with the agricultural output (GVIAO) between 1980 and 2000. official target), attainment of the Government's targets Specifically, given a reasonable set of assumptions-de- for hydro and nuclear power, and coal production in scribed below-about the efficiency of investment and 2000 of 1,400 million tons. The present orientation of the forces shaping the structure of the economy, the ag- foreign trade policy is assumed to be broadly maintained, gregate saving rate was varied until the rate that quadru- but the composition of manufactured exports shifts away pled GVIAO was found. from textiles toward machinery and other manufactures, As regards the efficiency of investment, sectoral capital- and there is somewhat greater dependence (to save en- output ratios in agriculture and energy were direct ergy) on imports of metals and chemicals. Foreign bor- (though approximate) estimates of the likely investment rowing increases gradually toward a level consistent with cost of increasing gross output in these sectors, drawing a 15 percent ratio of debt service (interest and repayment) on World Bank project experience in China and else- to exports. where. In other industrial sectors and in construction, it was assumed that capital-output ratios would be equal to MODERATE the average of the past thirty years in China (based on The MODERATE projection maintains most of the as- estimates of sectoral capital stocks in 198 1), with reforms sumptions made in QUADRUPLE, including the same ag- checking their past upward tendency. In transport, the gregate saving rate, but takes a less optimistic view of the capital-output ratio for new investment was set about 25 future efficiency of China's economy. Sectoral capital- percent above its past level to correct for past underin- output ratios in (nonenergy) industry and construction vestment. In commerce, it was assumed that greater dis- are assumed to increase gradually, as in the Soviet Union cretion in purchasing would cut circulating capital re- and less efficient developing countries. Specifically, the quirements by 20 percent, and hence that the sector's amount of investment required per additional unit of future capital-output ratio would be lower than in the industrial output is assumed to be 40 percent greater than past. in QUADRUPLE. In addition, crop production is assumed Energy conservation rates in individual sectors in to grow more slowly (an average annual rate of 2.9 per- QUADRUPLE (and also in MODERATE and BALANCE) are cent), because of smaller increases in the efficiency with those regarded as most likely-the consequences of more which investment and industrial inputs are used in agri- optimistic and more pessimistic assumptions are discussed culture. Lower efficiency in converting animal feed into in Chapter 4. Intermediate use of other materials and meat is also assumed. Coal production in 2000 is 1,200 goods was assumed to change in accordance with average million tons. international experience. Specifically, the assumed pat- Lower efficiency was also assumed to be associated tern of change in input-output coefficients includes some with slower growth of labor productivity in nonagricul- substitution of manufactured for agricultural materials, tural sectors. In QUADRUPLE, on the basis of past experi- increased use of chemicals (especially plastics), and re- ence in China and other countries, gross output per 34 worker was assumed to increase at an annual rate of 5 these sectors are indispensable to reducing costs and im- percent in heavy industry (3 percent in coal mining), 4 proving quality in material production through increased percent in light industry, 3 percent in construction, trans- reliance on markets, more specialization and competi- port and commerce, and 2 percent in miscellaneous ser- tion, and greater orientation of production toward cus- vices. In MODERATE all these rates were reduced by one tomer requirements. percentage point. BALANCE CAP[TAL. In this broader context, faster expansion of the service sectors could reduce investment requirements The BALANCE projection represents an alternative way of in several ways. A larger, better-equipped, and more re- attaining the same growth rate of per capita national in- sponsive commercial system could require-in light of come as in QUADRUPLE, giving greater weight to the ser- experience elsewhere-smaller amounts of circulating vice sectors-specifically, to commerce and miscellaneous capital, especially within the sector, but also in other business and personal services. This shifts the future struc- sectors, which would be less likely to produce unsalable ture of China's economy away from the Soviet pattern and goods and would have less need to stockpile materials and toward the pattern of Japan and most other countries at semifinished goods (of which it is now difficuk for Chi- comparable stages of development (Chapter 1). nese enterprises to get regular supplies of suitable quality In the model, the shift is accomplished partly by chang- and specifications). A larger and better commercial sys- ing the pattern of household consumption. Relative to tem would also permit greater specialization in produc- household incomc, consumption of miscellaneous ser- tion, and hencc largcr-scale enterprises with lower capital vices and of commerce (retail trade is treated as part of costs, and a greater payoff to agricultural investment. consumption in the accounting framework used) grows Bus ness service enterprises can also make fuller use of faster in BALANCE than in QUADRUPLE. To compensate, specialized equipment (and personnel). This is obviously the share of household income spent on food declines true cf equipment leasing and rental enterprises. But it somewhat faster-although this is partly offset by in- also applies to professional and technical services-such as creased food consumption in restaurants, which are also advertising and market research, law, accounting, design, part of commerce-and the share spent on manufactured engineering, repair and maintenance, and data pro- goods increases somewhat more slowly. There is also cessing-which can often be provided at lower cost and at faster growth of intermediate demand for services. Rela- a higher standard by specialized entities than by an enter- tive to gross output, the volume of intermediate com- prise's own staff. Even mundane business services such as merce and material supply activities is assumed in BAL- catering, cleaning, and trash removal can sometimes be ANCE to increase by 3 percent per year (it remains con- undertaken at lower cost by specialized enterprises. Fi- stant in QUADRUPLE); expenditure on miscellaneous busi- nally, banks and other financial institutions can contrib- ness services is assumed to increase gradually in every ute to better investment decisions (Chapter 10). sector from virtually nothing in 1981 to about 3 percent The potential for reducing investment requirements in of gross output in 2000-comparable to other countries. these ways cannot be accurately calculated or appor- Faster expansion of the service sectors involves costs- tioned between fixed and circulating capital. However, investment and use of energy, materials, and labor in on thr basis of rough international comparisons, it was those sectors. In BALANCE, thesc costs arc deliberately assumed for illustrative purposes in BALANCE that faster augmented by making the fixed capital required per unit expansion of the service sectors would enable circulating of gross output in commerce three times its QUADRUPLE capital requirements to be reduced (below their QUADRU- level, to reflect improvements in the size and quality of PLE levels) by 30 percent in nonenergy industry, con- warehouses, shops, vehicles, and other equipment, which struction, and miscellaneous services and by about 70 at present in China lag far behind those in other coun- percent in commerce. tries. In addition, to allow for improved staffing of large- scale commercial facilities, a larger share of more labor- VA-TERIALS. It was also assumed in BALANCE that the intensive smaller facilities, and a shift in the composition increased intermediate purchases of commerce and busi- of miscellaneous services toward more labor-intensive ac- ness services would be offset by reduced purchases of tivities, the growth rates of labor productivity in com- manufactured intermediate goods (and hence indirectly merce and miscellaneous services are reduced in BAL- of agricultural materials and energy, as well as other in- ANCE (to I percent per year in both sectors). dustrial materials). This is the pattern observed in the few Faster expansion of commerce and miscellaneous ser- studies that have been made in other countries-increas- vices would also have economic benefits-in addition to ing service-intensity of production in individual sectors, the gains that consumers would realize directly-espe- but ceclining materials-intensity, with little systematic cially in association with comprehensive reform of Chi- change (outside agriculture) in total intermediate pur- na's system of economic management (Chapter 1). For chases per unit of gross output. Part of the explanation 35 apparently lies in the transfer of particular activities from Table 2.2 Composition of Final Demand, manufacturing enterprises to more specialized service en- Alternative Projections, 1981-2000 terprises. But a more important factor appears to be con- (percent) stant improvements in product quality, with the cost of 2000 materials becoming a smaller part of the product price, Cowmponent 1981 Quadruple Moderate Balance and the value added by processing (including indirect Shares ofnational income processing, design, packaging, and marketing) a greater (expenditure)a part. Investment 28 (29) 29 (29) 29 (28) 26 (26) Of course, many materials-saving improvements in Public consumption 15(10) 15 ( 9) 15(10) 16(10) product quality are unrelated to service sector expansion. Household consumption 56 (61) 56 (62) 56 (62) 59 (64) Independent technological advances, including methods Shares of household of weight reduction and miniaturization, as well as gen- consumptipnb eral enterprise incentives to reduce waste and improve Food (including 48 49 44 products, are crucial. But the service sectors stimulate and Manufactures 24 29 28 25 facilitate these improvements. Design agencies and tech- Services (including nical consultants provide essential information, while a commerce) 18 20 20 27 good commercial sector, supplemented by advertising Fuel, electricity, and and market research enterprises, increases the flow of transport 3 4 4 4 information between customers and producers and a. Figures without parentheses are Western measures; those within paren- sharpens competition, making it both easier and more theses are Chinese measures. The Western measures may not sum to 100 sharpens cmpetition,making itboth easie and more because of external trade imbalances. necessary for enterprises to improve their products. b. Calculated on a Western basis, at 1981 producer prices, with all com- mercial margins included in services. Source: World Bank projections (Annex 4). Growth and Demand For each of the three projections, using both Western and Chinese measures, Table 2.1 presents aggregate growth percent, although the investment rate is closer to 30 per- rates over the period 1981-2000, while Table 2.2 shows cent for most of the period because of foreign borrowing. the composition of both total final demand and house- This is higher than China's average investment rate in hold consumption. (More detailed breakdowns are avail- 1952-82 (28 percent), but lower than the 1970-82 aver- able in Annex 4. The projections start in 1981 because age (32 percent). It is also quite close to Chinese projec- this is the latest year for which all the necessary data exist; vestment rate needed to quadruple GVIAO the projected values for 1982-84 do not conform closely in8-00(-2 rent), to ug sewhAt J in~~~~~~~~~ 1980-2000 (26-29 percent), though somewhat to actual developments in those years.) higher, perhaps because QUADRUPLE assumes more of an In QUADRUPLE, as mentioned earlier, the saving rate increase-over China's past levels-in capital require- was deliberately chosen so as to attain the target of quad- ments in agriculture, energy, and transport.3 rupling GVIAO in 1980-2000 (annual average growth of rntinalricome- anresl of thanypoth e 7.2 percent). The required domestic saving rate is 29 Nas tions imae, raer tha d antehoe-aso assumptions made, rather than by deliberate choice-also grows in QUADRUPLE at a rate consistent with the Gov- ernment's targets. Per capita GDP (which in China differs Table 2.1 Average Annual Growth of National only trivially from GNP) grows at 5. 5 percent per year- Income, Alternative Projections, 1981-2000 from $ 300 in 1981 to $ 8 30 in 2000-which is the mini- (Percent, at 1981 prices) mum rate that China would need to sustain to catch up Measure Quadruple Moderate Balance with the industrial countries by the middle of the twenty- first century (Chapter 1). The Chinese measure of per National income GDP 6.6 5.4 6,6 capita national income grows slightly more slowly-5.2 NMP 6.3 5.1 6.2 percent-because the relative size of the nonmaterial ser- National income per capita vice sectors increases. GDP 5.5 4.3 5.5 In MODERATE, with the same investment rate, the as- NMP 5.2 4.1 5.1 sumption of lower efficiency causes significantly slower Gross value of industrial and agricultural output 7.2 6.0 6.4 Note, A Western measure of national income, GDP (gross domestic product) 3. The Chinese projections are from Liu Guoguang, ed., Zbong- is the net output of all sectors, including all services, plus depreciation. The guo Jingji Fazban Zhanlue Weti Yanjiu (Issues of China's eco- Chinese measure, NMP (net material product) is the net output of the materially productive sectors. nomic development strategy), (Shanghai: Shanghai Renmin Chu- Source. World Bank projections (Annex 4). banshe, 1983), pp. 405-406. 36 growth. The growth rates of national income and of is the average amount of investment required to produce GVIAO are reduced by rather more than one percentage an additional unit of output. It can be calculated by point, and per capita GDP rises to only $670 by 2000. To dividing the average share of investment in national in- attain the same aggregate growth rates as in QUADRUPLE, come in a given period by the growth rate of national the investment rate would have to be increased to about income. 36 percent. (All three projections, incidentally, assume The projected ICOR in MODERATE (calculated using that the Government's target of a population of 1.2 bil- Western national income measures) is about 5.5, much as lion by 2000 is attained. A higher population would in China in the two decades before 1978, and compara- cause a commensurately lower level of per capita in- ble also to the Soviet Union in 1950-75 and India in come.) 1960-32. In QUADRUPLE, the projected ICOR is substan- In BALANCE, with greater weight given to the service tially lower-about 4.5, which is similar to the 1952-82 sectors, the saving rate was deliberately chosen to attain average in China. In BALANCE, it is lower still-about the same growth rate of per capita GDP as in QUADRU- 4.0, which is comparable to the average in Japan during PLE, in line with the Government's long-term objectives. 1950-SO, but also to the average for all middle-income The Chinese measure of national income in BALANCE develo?ing countries in 1960-82. The assumptions on grows only slightly more slowly than in QUADRUPLE, investrnent efficiency underlying the three projections because most of the increase is in materially productive thus seem to span a plausible range of international expe- services (commerce and business services). GVIAO, how- rience.5 ever, grows significantly more slowly, because the greater It should be emphasized, though, that China's experi- share of services means smaller shares for agriculture ence in the next two or three decades-in terms of and-especially-industry. Partly as a result, but also be- growth rate and investment requirements-will not nec- cause of the assumed reductions in use of circulating essarily lie within this range. The possibility of faster capital and materials within individual sectors, the same growth cannot be ruled out. To illustrate this, a variant of growth rate of national income as in QUADRUPLE is at- BALAM CE was constructed, with the saving rate increased tained in BALANCE with less investment-26 percent of to the same level as in QUADRUPLE. Per capita GDP in this national income. The increase in consumption that this projection increases at an annual average rate of 6.5 per- makes possible is divided proportionately between public cent (one percentage point higher than in QUADRUPLE consumption and household consumption (which in real and BALANCE, though still slightly below the 6.8 percent per capita terms in 2000 is 9 percent higher than in of 1979-84) and reaches $990 by the year 2000. GVIAO QUADRUPLE). in this variant of BALANCE more than quadruples (growth By most international standards, the growth rates pro- of 7.4 percent per year), with a commensurate increase jected for China are quite high. Even in MODERATE, per also in energy and materials use. capita GDP growth is slightly above the 4.1 percent Equally, if internal or external misfortunes were ad- 1960-82 average for upper-middle-income developing versely to affect China's saving rate or investment effi- countries (3.2 percent for lower-middle-income coun- ciency, growth could be slower than in MODERATE. In tries), and similar to Western estimates of Soviet growth addition, it is possible that all three projections may have in 1950-75. It is well above the 2.6-3.6 percent range underestimated the backlog of past investment needs- projected by the World Bank for middle-income coun- especially in transport and in housing (the estimates for tries in 1985-95.4 Nonetheless, the projected MODERATE electricity are more reliable)-in which case their saving growth rate of national income per capita is almost ex- and investment rates would be too low, or their growth actly what China achieved in 1952-82 (4.0 percent by rates too high, or some combination of the two. In any the Chinese measure), and the rates in QUADRUPLE and BALANCE are no higher than those attained by China in periods of good economic management. These rates in 4 orld Bank, World Development Report 1984 (New York: turn have been surpassed by a few other countries, in- Oxford University Press, 1984), Table 3.2. Most other developing cluding Japan in 19 5 3 - 7 3 (8. 5 percent). countr'/ statistics in this chapter are from the Indicators at the back China's unusually rapid projected per capita income of this World Development Report. growth, as compared with most other developing coun- 5. T3tal factor productivity (IFP) for the whole economy grows tries, is due mainly to an unusually high investment rate at 1 .6 percent per year in QUADRUPI.E (contributing 24 percent of (the 1982 average for middle-income countries was 24 output growth), 0.8 percent in MODERAIE (14 percent), and 1.9 percent) and unusually slow population growth (I per- perceni: in BALANCE (29 percent). The average IFP growth rate in a cent per year, as compared with a projected 2.2 percent sample of mainly middle-income developing countries between 1950 and 1980 was 2 percent and contributed 31 percent of in middle-income countries), rather than to unusually output growth (Annex 5, Table 3.16). In these TFP calculations, efficient investment. This can be seen by looking at the output is measured by GDP; for China, labor force and capital stock aggregate ICOR (incremental capital-output ratio), which growtr are given equal weight. 37 event, experience in China and elsewhere (including East- reflects both lower efficiency in crop production and ern Europe) strongly suggests the need for cautious plan- slower overall growth, which reduces the demand for ning: the fluctuations and inefficiency caused by setting meat. (In these projections, the output of nonagricultural unrealistically high growth targets are usually much more brigade enterprises is included in other sectors, rather serious than the problems caused by unrealistically low than-as in present Chinese statistics-in agriculture.) growth targets, which can be gradually adjusted upward. The projected growth rate of total industrial output in The projected changes in the composition of household QUADRUPLE (about 8 percent) is in line with most Chi- consumption (Table 2.2) are broadly consistent with ex- nese projections, as is the slightly faster growth of heavy perience elsewhere. In all three projections the share of than of light industry.6 Within heavy industry, the fastest- food declines: in QUADRUPLE and MODERATE the reduc- growing sector (9.1 percent) is machinery, pulled along tion is unusually small, mainly because of rapidly increas- mainly by investment demand and by household and ing consumption of animal products; BALANCE is more public consumption, but also by foreign demand (the normal by international standards (Chapter 3). In QUAD- share of machinery output exported rises over the period RUPLE and MODERATE, most additional nonfood con- from 4 percent to 11 percent). Chemicals production- sumption is manufactures, although expenditure on hous- pulled mainly by intermediate demand-grows almost as ing and miscellaneous personal services also grows quite fast (8.8 percent), but metallurgy production grows more rapidly. In BALANCE, consumption of services (including slowly (7.0 percent), partly because of increased effi- retail commerce) grows abnormally fast, to bring China ciency of metal use in the machinery sector, partly be- more into line with the usual pattern. The share of house- cause of increased net imports. Coal and petroleum are hold income spent on manufactures thus increases slowly, the slowest-growing heavy industrial sectors (on average, but because-as mentioned earlier-aggregate consump- 4.8 percent). Within light industry, the fastest-growing tion increases faster, real per capita consumption of man- sector is food processing (8.7 percent), followed by ufactures in BALANCE in 2000 is only 5 percent less than wood, paper, and miscellaneous manufacturing (8.4 per- in QUADRUPLE. cent). Textiles and clothing, the biggest light industrial sector, grows more slowly (7.1 percent), partly because Production and Investment consumer demand grows only slightly faster than income, partly because textile exports grow relatively slowly (but Table 2.3 shows gross output growth in the three projec- still by 6.0 percent per year in real terms). tions, broken down among five broad sectors (the In MODERATE, both heavy and light industry grow twenty-sector breakdown is in Annex 4). In all three more slowly than in QUADRUPLE (because the whole cases, agriculture-discussed further in Chapter 3-is the economy grows less fast), but the difference between slowest growing sector, although its growth rate in their growth rates is somewhat more pronounced, largely QUADRUPLE and BALANCE is very high by international because, by assumption, coal production is not much standards, partly because animal husbandry production is lower and petroleum production remains the same. Gross assumed to keep up with rapidly increasing consumer output in other industrial sectors in 2000 in MODERATE demand. Slower agricultural growth in MODERATE thus is about 80 percent of its level in QUADRUPLE-the range is from 84 percent in chemicals to 77 percent in food processing. Table 2.3 Average Annual Growth of Sectoral Gross In BALANCE, the assumed greater efficiency of invest- Output, Alternative Projections, 1981-2000 ment and materials use-and the consequently lower in- (pereent, at 1981 prices) vestment rate-enables heavy industry to grow slightly more slowly than light industry and significantly more so than in QUADRUPLE (even though national income grows Agriculture 4.5 3.7 4.6 just as fast). Machinery production in 2000 in BALANCE Heavy industry' 8.1 6.9 7.0 is only 80 percent of the QUADRUPLE level; metallurgy, Light rdustrye 7.9 6.5 7.1 74 percent; building materials, 85 percent; and chemicals, Servicesd 7.2 6.0 10.5 82 percent. Light industry in BALANCE also grows more slowly than in QUADRUPLE, partly because some con- Note: These broad sectoral definitions correspond only approximately to sumer dean is sitDRom manufaureso ces, (hose used in China. ~~~~sumer demand IS shifted from manufactures to services, those used in1 China. a. Metallurgy, coal and petroleum (including extraction), chemicals, and building materials and machinery; excludes electricity. b. Food processing, textiles and clothing, wood, paper, and other manufac- turing. 6. Most Chinese projections envisage this relationship for the d. Commerce, miscellaneous serics a uanspirte(freughteand passenger)d whole period 1980-2000, but with faster growth of light than of health, public administration, and defense. heavy industry in the 1980s (see, for example, Liu, Zbongguo Jingji Source World Bank projections (Annex 4). Fazban Zbanlue Wenti Yanjiu, pp. 146-49). 38 but mainly because of smaller stock building, more effi- Table 2.4 Composition of Capital Stock, Alternative cient intermediate use, and smaller net exports (because Projections, 1981-2000 of less need to generate foreign exchange to pay for im- (percent) ports of energy, chemicals, and metals). 2000 The infrastructure sectors (electricity, transport, and Sector and type 1981 Quadruple Moderate Balance construction) in all three projections grow nearly as fast as industry, the main user of their outputs and supplier of Sector as outputs ~~~~~~~~~~ ~ ~~~~~~~AgriCUJLnre 9.6 8.8 8.4 9.8 their inputs. The link between infrastructure and indus- Heavy iredustry 29.5 33.4 37.3 28.4 try is particularly clear in QUADRUPLE and MODERATE- Light Inclustry' 7.9 7.7 8.5 6.4 with slower industrial growth in the latter projection Infrastructurea 14.1 18.1 16.4 18.4 causing an equiproportionate reduction in infrastructure Servicesb 38.9 32.0 29.4 37.0 growth. In BALANCE, relative to QUADRUPLE, the reduc- Type of capital stock tion in infrastructure growth is much less pronounced Fixed (nt of than the reduction in industrial growth, because of in- depreciation) 65.3 67.5 66.9 74.8 creased use of infrastructure by the service sectors. Elec- Circulatiig 34.7 32.5 33.1 25.2 tricity projections are discussed further in Chapter 4 (and a. For sectoral definitions, see notes to Table 2.3. Annex 3), transport projections in Chapter 5 (and Annex b. Includo s education, health, public administration, and defense, as well as Annex 3, tranport poj'ectons inChapte 5 (an Annex commerce, housing, and miscellaneous services. 6); in both sectors, the plausible range of outcomes in the Source WAorld Bank projections (Annex 4). year 2000 in each of the three projections is quite wide (depending largely on the degree of success in reducing wasteful use of electricity and transport). The construc- ciency--reflected in the table in the lower ratio of circu- tion industry's output growth is determined mainly by lating to fixed capital. investment demand: it therefore grows more slowly in In QUADRUPLE, industry's share of the capital stock MODERATE (6.1 percent) than in QUADRUPLE (7.2 per- increases over the period-implying that its share of total cent), because of slower overall growth; it also grows investment (about 42 percent) is greater than in the past somewhat more slowly in BALANCE (7.0 percent), be- (about 37 percent).7 This is the result of rapid output cause the investment rate is lower. growth in sectors such as machinery and chemicals Excluding education and health and public administra- (which between them absorb around 20 percent of an- tion and defense, where gross output is a particularly nual investment), coupled with rising capital-output ra- elusive concept (employment in these sectors is discussed tios in ,lower-growing coal and petroleum. Light indus- below), the service sectors in QUADRUPLE and MODER- try's small share of the capital stock does not change ATE grow at about the same rate as the infrastructure much. In MODERATE, lower industrial investment effi- sectors and somewhat more slowly than industry. Com- ciency causes the capital stock shares of both heavy and merce grows at almost the same speed as GVIAO, while light industry to be even higher than in QUADRUPLE. In miscellaneous services and housing (both of which are BALANCE, these shares decline slightly, because of slower propelled by above-average growth of consumer demand) industrial output growth and greater industrial invest- grow rather faster. In BALANCE, commerce grows faster ment efficiency.8 Industry's share of total investment (36 (9.1 percent) than GVIAO (6.4 percent), with an increased percent) in BALANCE is still large, however-excluding amount of trade relative to production, while consumer mining and extraction, it is about the same as in Japan in and business demand for miscellaneous services grows 1951-65. rapidly (15.3 percent). The service sectors thus grow In all three projections, the infrastructure sectors in- faster than industry. crease their shares of the economy's capital stock. In Table 2.4 shows broad sectoral shares of the economy's QUADRUPLE, transport absorbs about 10 percent of an- total capital stock (fixed and circulating) in the three pro- jections. These depend, of course, on the sectoral alloca- tion of investment, which in turn is related both to sec- 7. These percentages include all mining (including oil extrac- toral gross output growth and to sectoral capital-output tion), biut exclude electricity (included with infrastructure). The ratios (investment requirements per unit of output). In all past estinate is based on the 1981 capital stock data in Annex 5, three projections, agriculture's share of the capital stock which, is with the projected estimate, covers all investment in does not change much over time, since the sector's be- fixed and circulating capital, including nonstate and nonproductive low-average output growth is largely offset-as in other investment. countries-by a rising capital-output ratio. Its share in 8. The share of light industry declines more than that of heavy industry partly because coal and petroleum capital requirements BALANCE is higher than in QUADRUPLE simply because are not reduced, partly because greater investment efficiency is the total capital stock is 10 percent smaller, because of the assumed to be reflected in lower circulating capital requirements lower aggregate investment rate permitted by greater effi- (which are more important in light than in heavy industry). 39 Table 2.5 Sectoral Composition of the Labor Force, cause of faster growth of public and household consump- Alternative Projections, 1981-2000 tion. Miscellaneous services absorb three times as much (percent, unless otberwise noted) investment as in QUADRUPLE, but stllH less than 2 percent Sector 1981 Qyadruple Moderate Balance of the total. Agriculture 70 59 61 52 Employment and Income Heavy industry 9 11 11 9 Light industry 6 9 8 7 Table 2.5 shows projected sectoral employment shares Infrastructure 5 7 7 7 (the twenty-sector breakdown is in Annex 4). These de- Services 10 14 13 25 pend in the model on sectoral output growth rates and Labor force (millions) 452 631 631 631 assumptions about sectoral labor productivity growth Population (millions) 990 1,196 1,196 1,196 (mentioned earlier)-except in agriculture, where em- Note. For sectoral definitions, see notes to Table 2.4. ployment is a residual, and in education, health, public Source World Bank projections (Annex 4). administration, and defense, where it is determined by public consumption. In QUADRUPLE, nonagricultural employment increases at 3.5 percent per year. This is somewhat slower than nual investment (excluding purchases of vehicles by over the past thirty years in China-in 1952-82, the rate farmers, households, and nontransport enterprises). was 4.5 percent-because of slower gross nonagricultural Transport and electricity together account for a higher output growth (7.8 percent in QUADRUPLE, excluding proportion (19 percent) of total investment than in the nonmaterial services, versus 8.9 percent in 1952-82). past (12 percent), but about the same proportion as in Average nonagricultural labor productivity in QUADRU- Japan in 1951-65. In MODERATE, these sectors' shares of PLE grows at the same rate as in 1952-82 (about 4 per- the capital stock increase less, mainly because of greater cent), though significantly faster than in 1957-77 (about absorption of investment by industry. In BALANCE, their 3 percent). Labor productivity is assumed to grow fastest capital stock shares are much the same as in QUADRUPLE, in heavy industry, with the result that this sector's em- although in absolute terms somewhat lower (because de- ployment share increases only modestly, despite rapid mand for the output of these sectors grows more slowly). output growth. Conversely, the share of employment in At the outset, the service sectors have a large share (39 services increases more, despite slower output growth, percent) of the economy's capital stock-of which about because labor productivity does not rise so much. one-half is commerce (mainly circulating capital), one- Although faster growth of productivity in industry third housing, and one-sixth education, health, public than in services is the normal international pattern, the administration, and defense. In QUADRUPLE, this share share of employment in services in QUADRUPLE not only declines, partly because circulating capital requirements remains unusually low, but also increases more slowly in commerce are assumed to be somewhat lower than in than normal (even by comparison with the low service the past, partly because capital-output ratios in other sec- share countries such as the Soviet Union). This is because tors are rising (they rise faster in MODERATE, which of the prospective decline in the primary school age causes the service sectors' capital stock share to decline group, which will cause the number of teachers-now even more). Nonetheless, these sectors absorb substantial approaching a quarter of all service sector employment in amounts of investment. In QUADRUPLE, about 12 per- China-to increase only modestly, even given higher en- cent of annual investment goes into housing and 11 per- rollment rates (Annex 1). Faster growth of employment cent into commerce (still mainly circulating capital), while in health, research, culture, and so on will to some extent education, health, public administration, and defense be- offset this. But the projected increase of 15 million, or 80 tween them absorb about 5 percent (the underlying as- percent, in education and health employment during sumption is a doubling of capital per worker over the 1981-2000 in QUADRUPLE-MODERATE and BALANCE period in these social sectors). Miscellaneous services ab- are similar-may in fact be an overestimate, as may the sorb only 0.5 percent of investment. projected increase of 5 million, or 50 percent, in employ- In BALANCE, the capital stock share of the service sec- ment in public administration and defense (which is cur- tors also declines slightly, but is significantly higher than rently approaching a quarter of total service sector in QUADRUPLE, partly because the total capital stock is employment). smaller. In commerce, faster output growth and higher Nonagricultural employment in QUADRUPLE increases fixed capital requirements are offset by lower circulating between 1981 and 2000 by rather more than 120 mit- capital requirements, and the sector's capital stock in lion. The total labor force-assuming a slight decline in 2000 is 5 percent smaller than in QUADRUPLE (its share the participation rate of the adult population-increases of annual investment is virtually the same). The social by nearly 180 million. Employment in agriculture thus sectors and housing absorb slightly more investment be- increases by 56 million, even though agriculture's share 40 of employment drops from 70 percent to 59 percent. Figure;!.1 ShareofAgricultureinTotalEmployment This pattern is typical of the low- to middle-income tran- in China, Alternative Projections for 1981-2000, sition elsewhere, with an absolute decline in the agricul- and in Other Selected Countries, 1960-82 tural labor force occurring only at a later stage of develop- ment. But the projected decline in the agricultural Percentage of employment employment share in QUADRUPLE is rather small by in- - ternational standards (Figure 2.1). Thailand It is also much smaller than in most Chinese projec- Thalan tions, which envisage a decline in the agricultural employ- 80 Turkey ment share by 2000 to less than 40 percent.9 This would India Indonesia be normal for a country with a per capita income roughly 70 St Ko\r double that projected for China in 2000 (Annex 5, Table South Kor 3.8). Given the rates of gross output growth in QUADRU- Palnstan PLE, it could be achieved in China only if nonagricultural 60 Pakistan labor productivity increased at less than 2 percent per Egypt year-half the past rate and probably incompatible with 50 \ Malaysia efficient industrialization. In 1960-82, industrial labor Ph productivity increased on average in all middle-income 40 Philippines countries at rather more than 2 percent per year; but in ga the faster-growing upper-middle-income group, at 3 per- Yugoslavia cent; in South Korea, at 6 percent; and in Japan in the 30 Brazil 1950s and 1960s, at more than 8 percent. (Productivity and surplus rural labor are discussed in Chapter 8.) 20 I i Sectoral employment shares in MODERATE are similar 12:: 250 500 1,000 2,000 4,000 to those in QUADRUPLE. This is because the assumed Per capita income (1982 U.S. dollars, logarithmic scale) lower efficiency that causes slower output growth is also * *QUADRtUPLE projection assQciated with slower growth of nonagricultural labor - B---- -ALANCE projection productivity (on average about 3 percent). But the nonag- ricultural labor force grows slightly less than in QUADRU- Source V,orld Bank, World Development Report 1984 (New York: Oxford PLE, and hence the share of employment in agriculture Universit; Press, 1984), table 21; World Bank, World Tables. 3d ed. (WVash- does not decline quite so much. ingtonn D C., 1984), vol. 2. In BALANCE, employment in services grows much faster than in QUADRUPLE-partly because the demand ment is 4.3 percent). Agricultural employment in for commerce and miscellaneous services expands faster, BALANCE thus increases between 1981 and 2000 by only partly because improved quality and the changing com- 13 million, and declines to 52 percent of the labor force. position of activities within these sectors are assumed to This would be quite normal for a large lower-middle- entail slower labor productivity growth. By 2000, the income country (Figure 2.1 and Annex 5, Table 3.8). service sectors, which employ 14 percent of the labor Sectoral employment shares largely determine sectoral force in QUADRUPLE, employ 25 percent in BALANCE- shares af national income, although these also depend on almost exactly the 1980 average for lower-middle-income the amount of capital per worker, on wage rates, and on developing countries, with an average per capita income prices--especially in relation to costs. The projections of very similar to that projected for China in 2000. Because sectoral national income shares in Table 2.6 (both West- industrial output grows more slowly in BALANCE, and ern and Chinese measures are presented) assume that industrial labor productivity grows just as fast, industrial sectoral prices change in parallel with sectoral costs, employment does not increase so much as in QUADRU- which means that the relative profitability of different PLE-and indeed increases very little as a proportion of sectors remains constant (Box 2.1). Different assumptions the total labor force (although in absolute terms it rises by about prices-for example, that they do not change or 35 million workers). Employment in the infrastructure that there is movement toward equalization of sectoral sectors, however, increases only slightly less than in QUADRUPLE. The greater increase in service sector employment in BALANCE much more than offsets the smaller increase in tural employment in 2000 is 225 million-SO percent of a rural industrial employment, so that nonagricultural employ- labor force of 450 million. The projection cited in Annex 2, Table ment increases by over 40 million more than in QUADRU- 5.1, is somewhat higher-250 million, or 39.6 percent of the PLE (the annual growth rate of nonagricultural employ- World Bank's total labor force projection of 6 3 1 million. 41 Table 2.6 Sectoral Composition of National Income, profit rates-would somewhat alter the projected shares Alternative Projections, 1981-2000 of national income. (percent) In QUADRUPLE, the share of agriculture in national 2000 income drops by ten percentage points between 1981 Sector 1981 Quadruple Moderate Balance and 2000, with increases in the shares of other sectors, especially heavy industry and infrastructure. In MODER- Heavyindustry 24(26) 28 (30) 29(31) 23(28) ATE, the movements are similar, although the share of Light industry 15(17) 15(18) 15(18) 12(17) agriculture declines less. In BALANCE, there is a much Infrastructure 11(10) 14 (14) 13 (13) 13 (14) greater increase in the share of services, especially using Services 16 ( 7) 17 ( 7) 16 ( 7) 28(13) the Western measure (partly because some of the increase Note Figures without parentheses are Western measures (shares of GDP); is in nonmaterial services, partly because the material those within parentheses are Chinese measures (shares of NMP). In the latter contribution of miscellaneous services is in the Chinese case, infrastructure excludes passenger transport and only commerce is in measure included in the output of the material produc- services; the contribution of business services to material production is included in the net output of other sectors. Other sectoral definitions are as tion sectors rather than of services). The share of industry gven in Table 2.4. Figure 1.2 differs from the present table mainly because in national income in BALANCE increases less by the Chi- of price adjustments (see Annex 5). Source: World Bank projections (Annex 4). nese measure and declines significantly by the Western measure (the difference is again due mainly to the treat- ment of business services). Agriculture's share of national income declines more in BALANCE than in QUADRUPLE, Box 2.1 Wages, Profits, and Prices Sectoral relative prices-which are producer (ex factory or mediate inputs (such as agriculture), or where labor pro- farmgate) prices-are determined in the model by sectoral ductivity is rising more slowly than average (such as ser- costs, including intermediate input costs, wage costs, and vices), or where the capital-output ratio is rising faster than capital costs. Intermediate input costs depend on usage average (such as energy), relative prices rise. rates-input-output coefficients-and on input prices. Conversely, in sectors where intermediate-including Wage costs in each sector depend on labor productivity energy-use is reduced, or with above-average labor pro- and on wage rates, which in turn depend on the rate at ductivity growth or below-average capital-output ratio in- which wages in general are rising and on the relationships creases (many industrial sectors), relative prices fall. Some among wage rates in different sectors. In the model, gene- sectors are subject to conflicting cost pressures, which, ral wage increases are tied to household consumption in- moreover, may vary among projections-for example, ag- creases; in these projections, sectoral relative wage rates ricultural prices rise less in BALANCE then in QUADRUPLE (agricultural carnings are included in wages) are assumed to because agricultural labor productivity rises more, as a remain constant in proportionate terms-except in educa- result of greater absorption of labor by other sectors. tion and health, where the initially low wage rises gradu- The assumptions of the present trio of projections thus ally to the same level as in public administration. allow in a simplified way for the most important long-run Capital costs in each sector are governed in the model by tendencies in sectoral relative prices, but not for any modi- the capital-output ratio and the sectoral profit rate (gross of fication of the present relationships in China among wages depreciation), which in turn depends on the economy- in different sectors or among sectoral profit rates. Such wide average profit rate and on the relationships among modifications could easily be incorporated (for example, to profit rates in different sectors. The average rate of profit approximate the results of specified types of wage and (including tax) is determined in the model by the need to price reform), although they would make it harder to generate a large enough surplus to finance the planned interpret the projected structural trends with which this level of saving (net of household saving) and government chapter is mainly concerned. It would also be possible-if consumption. In these projections, it was assumed that complicated-to extend the model to distinguish retail relative sectoral profit rates would remain constant in pro- prices from producer prices, enterprises from government, portionate terms. and urban households from rural households. The deter- In short, prices in these projections adjust over time so as minants of the absolute price level and the rate of inflation, to maintain constant the relative wages of the workers in however, probably could not be handled satisfactorily in each sector and the relative rate of profit on capital in each this model. sector. As a result, in sectors with increasing use of inter- 42 although the difference is more pronounced by the West- culture, energy, transport, urbanization, foreign trade, em measure. and public finance. It is important, however, to interpret The share of agriculture in national income in 2000 in all these projections appropriately. They are not predic- BALANCE would be normal for a large lower-middle-in- tions, but simply complicated calculations based on many come country (Annex 5, Table 3.6). The share of ser- assumptions, whose results could be altered considerably vices, however, is almost ten percentage points lower, by different, but still defensible, assumptions. As such, notwithstanding the similarity of the service sector em- they should be taken seriously-as an indication of some ployment share, and the projected income share of indus- future Fossibilities-but not literally. try and infrastructure is correspondingly higher than in The choice of terminal year (2000) is also quite arbi- these countries-indeed, it exceeds that of almost all trary; the purpose of the projections is to illustrate possi- other countries, including Japan. These discrepancies be- ble long-term trends, most of which would continue into tween employment and income shares partly reflect real the twenty-first century, rather than to assess China's differences in the composition of industry (more heavy positior. in any particular year. And for planning pur- industry in China) and services (a lower proportion of poses, the process of making projections is often as help- social services elsewhere). But they also reflect the cur- ful as the particular model and numbers selected for pre- rently unusual structure of prices and wages in China, sentation. Cumulatively, the innumerable "unsuccessful" most features of which in these projections are assumed- experiments with different assumptions and model speci- perhaps incorrectly-to persist (Box 2.1). The differences fication, are very illuminating, especially in practical judg- between the projected sectoral income shares for China ment of particular plans and policies. in 2000 and the normal shares in large lower-middle- For these reasons, and because there is much scope for income countries are of course even greater in QUADRU- improvement of the model and the underlying data (espe- PLE and MODERATE than in BALANCE. cially the input-output table and sectoral capital require- ments), the projections presented in this report should be viewed as no more than a preliminary contribution. Fur- ther wcrk, and better statistics, would be needed to turn Subsequent chapters will further discuss the implications them into a reliable basis for planning. of these three (and some additional) projections for agri- 43 Agricultural Prospects and Policies During the next two or three decades, agriculture will and 1983. Giving adequate incentives to individual remain one of the largest and most important sectors of farmers-the essence of the reforms-will continue to be China's economy.1 Even in 2000, food will account for a vital ingredient in expansion of agricultural production about half of household consumption, and about half the over the longer term. But their effectiveness will depend labor force will be engaged in agricultural activities. The on the availability of key inputs such as agricultural land economic linkages between agriculture and other sectors and irrigation, fertilizer, and improved seed. Also impor- will become ever closer. The pace of agricultural progress tant, though less tangible, will be improvements in sup- will depend increasingly on growth of nonagricultural port services, such as research and extension, and trans- demand for food, materials, and labor and on the avail- port and marketing. ability of nonagricultural goods and services. And the Future availability of land will depend heavily on the pace of nonagricultural progress will continue to depend extent to which reclamation can compensate for land lost on growth of agricultural purchasing power and on agri- to nonagricultural uses. Taking account of land quality culture's capacity to supply the right amount and mixture and use by herdsmen, perhaps another 3 million to 5 of produce for industrial processing and nonagricultural million hectares of "wasteland" are suitable for develop- consumption. System reform and government policies, ment in the medium term for sustained production of moreover, will greatly influence these mutually reinforc- annual crops. Losses of land to nonagricultural uses in ing interactions between agriculture and other sectors. 1959-78 were about 1 million hectares annually; the This chapter, which draws heavily on Annex 2, first current surge of rural housing construction suggests that reviews China's long-term agricultural production poten- these losses will continue to be substantial and are un- tial. It then examines the long-term balance between the likely to be fully offset by increases in the amount of land demand for and supply of agricultural products, with irrigated, which has not increased significantly in recent special emphasis on China's future options regarding the years. level and pattern of food consumption. Finally, the chap- Although organic fertilizer will remain an important ter discusses the choice of policy instruments for manag- source of nutrients (particularly phosphorus, potassium, ing agricultural production under the recently imple- and some trace elements), it is unlikely to meet the Gov- mented responsibility system, and possible ways of ernment's target of half of total nutrient offtake by 2000, narrowing the large gap between agricultural and nonag- if China's fertilizer use patterns follow those of other ricultural incomes. (Rural nonagricultural activities, inter- countries. However, China now has the capacity to de- regional agricultural specialization and trade, and the sign and construct efficient, large-scale plants to produce problems of poor localities are discussed in Chapter 5.) nitrogenous chemical fertilizers; if raw materials are made available, the rather modest future production targets for Production Possibilities nitrogen seem realizable. But achieving the 2000 targets for both phosphorus and potassium would require an The rural reforms of 1979-80 have stimulated exception- early commitment of substantial resources to develop lo- ally rapid growth of agricultural production, and of rural cal raw materials and introduce new production technol- incomes and food consumption (Chapter 1), based on rapidly rising yields. In grain production, for example, average yields-which were already high by most interna- 1. "Agriculture" is defined throughout to exclude brigade and tional standards-increased by 22 percent between 1979 team industry. 44 ogies. Even so, sizable amounts of phosphorus and potas- the problems of poor seed quality, low fertilizer applica- sium would still need to be imported. Equally important tion rates, and less advanced cultural practices than those will be improvements in the fertilizer distribution system, applied to wheat and rice. to ensure the timely availability of supplies in economi- Among the industrial crops, prospects seem good for cally optimal quantities (discussed later). For seed, how- further yield and production gains in cotton. Increased ever, the basic mechanisms for production and distribu- product:ion of soybeans is essential for meeting livestock tion are already in place, but processing facilities and producc targets; there is also a need to improve processing quality assurance procedures need to be improved. of oilseeds to increase supplies of protein meals suitable Experience in other countries suggests that building an for livestock feeding. For industrial crops such as sugar effective national agricultural research capability requires cane and beets, emphasis should be on increasing produc- an annual investment of up to 2 percent of the value of tive efficiency through better cultural practices, improved agricultural output-considerably higher than current ex- planting materials, and careful choice of planting loca- penditure in China. Despite some very notable achieve- tion. Elevelopment of the state forestry sector is ham- ments, work at many research institutes is hampered by pered by limited investment and low administered prices, poor facilities and inadequately qualified staff. Good although price and marketing restrictions have recently progress has, however, been made in reorganizing the been abolished for individual and collective forestry. extension service in accordance with the rural reforms, so that it can deliver technical advice to many small pro- Prospec*sfor Livestock ducers. Close links need to be maintained between re- The Government expects rapid growth of the livestock search and extension. Moreover, agricultural education sector (tentatively projected to increase its share of gross needs to be strengthened to support these programs: re- agricultural output-excluding brigade industry-from cent studies indicate shortages of appropriately trained about .8 percent in 1982 to over 30 percent in 2000). personnel (at both professional and technical levels), as But achievement of the livestock growth targets could be well as a need to upgrade the qualifications of those constrained by the supplies of grass and feed, the weak- already employed. nesses of the systems of transport, processing, and distri- It will be essential to improve rural transport, process- bution, and the availability of investment funds. Many of ing, storage, and distribution, whose limited capacitv is China's grasslands are overgrazed and probably cannot already an important constraint on the restructuring of provide additional meat and wool in the short to medium China's agricultural sector and increased specialization. term without further degradation. An urgent require- Efforts are needed to both expand the rural marketing ment is therefore to match livestock numbers with the system and increase its efficiency, with simplification of land's carrying capacity, through herd reduction and im- procedures and reduction of administrative barriers to proved range management. The prospects for pigs and internal trade. The Government is currently taking fur- poultry are more favorable, but depend on increased stup- ther important steps in this direction, by switching from plies ol high-quality energy and protein feeds, as well as procurement quotas for major crops to contracts, by in- on the price relationship between feed concentrates and creasing price flexibility, and by allowing collective and meat products. Because poultry provide Chinese con- individual enterprises a greater role in marketing. In the sumers wvith a preferred meat and are, moreover, efficient provision of support services, as in agriculture itself, the conver.ers of feed to high-quality protein, China's poul- economic environment will thus need to be conducive to try industry should be a priority development area. individual and collective investment. Especially because The prospects for dairy development over the long state investment in agriculture is not expected to increase term will depend largely on the availability of low-cost, significantly, it will be critical to put much of the growing high-quality forage and feed protein and on the establish- volume of rural savings to productive use. ment of efficient milk processing and distribution sys- tems. MAilk supplies for some urban consumers might also Prospectsfor Major Crops be increased at lower economic cost by importing pow- The Government's long-term growth target for grain dered milk. Aquatic production will continue to be a production is about 2 percent per year-to be achieved valuable supplementary source of high-quality protein, by raising yields, with the cropped area declining. Rice and good potential exists for increasing the output of production is projected to grow at 1.8 percent per year- fresh anid marine fish. a target that appears readily attainable. The projected 2.6 percent annual growth in wheat yields would require Potential Demand-Supply Imbalances continued attention to wheat diseases, good water con- trol in irrigated areas, and improved nutrient balance in Projecting China's long-term agricultural demand-supply fertilizer use. Similarly, the projected growth in corn balance is peculiarly difficult. This is partly because of the yields (2.8 percent per year) would require attention to dramatic difference between the sector's performance in 45 the past five years (gross output up by over 7 percent per of 3.6 percent. This particular growth rate was chosen year) and in the preceding twenty years (average under because it approximately balances agricultural demand 2.5 percent). Depending on what past period is regarded and production in 2000-net crop imports equal to 0.5 as representative, and what other countries as relevant percent of crop production. Though well above the long- comparators, the future trend growth rate could be pre- term trend increase in China prior to 1979, this projected dicted at anything between 2-3 percent and 5-6 percent crop production growth rate is close to the 1965-83 per year-which, over twenty years, is the difference average (3.7 percent per year). between less than doubling and tripling. Moreover, even Given the underlying assumptions about agricultural given a particular view of potential production growth, efficiency, this crop production growth rate requires use most projected imbalances between agricultural demand of chemical and other industrial inputs to increase at and supply will not actually materialize-or at least not to about 6 percent per year, and investment in fixed and the extent indicated-simply because producers or con- circulating capital averaging about Y 6 billion (or 2-3 sumers or the Government will automatically react in percent of total investment) per year. Production of ani- ways that reduce them. The following analysis (which is mal products-responding to the assumed steep increase based partly on the economywide model and projections in consumer demand-grows at 7.5 percent per year, presented in the previous chapter, partly on the more which requires, in addition to rapid growth of animal detailed agricultural projections in Annex 2) is thus not feed consumption, investment averaging about Y 15 bil- an exercise in prediction, but an exploration of possible lion (or 5-6 percent of total investment) per year. In tendencies and problems-some of which could be total, gross agricultural output grows at an average annual avoided if anticipated sufficiently early.2 rate of 4.5 percent, and the sector absorbs 8 percent of A starting point for the analysis was the assumption- total annual investment. based partly on Chinese planning targets-that over the This agricultural growth rate lies within the fairly wide next two decades there would be a rapid increase in the range (under 4 percent to over 5 percent) contemplated share of meat, poultry, fish, eggs, and dairy products in by most Chinese economists.3 Though well below the the average Chinese diet, with a corresponding decrease growth rate of the past five years, it is higher than any in the share of direct grain consumption. Specifically, it past long-term trend rate in China (in 1965-83, the aver- was assumed that, if the Government's targets for na- age was 4.1 percent). This is because of the high pro- tional income and population growth were attained, per jected growth rate of animal husbandry production (in capita consumption of animal products would approxi- 1965-8 3, growth of livestock and fishery production av- mately quadruple (by weight), raising their share of a eraged 5.4 percent and 3.6 percent per year, respec- slightly increased total caloric intake from 6 percent in tively). By international standards, it is a very high long- the early 1980s to about 15 percent by the end of the term growth rate-double the 2.3 percent average for century. (In the next section, some possible alternative low-income countries in 1960-80, well above the 3.0 food consumption patterns are considered.) The economywide model also assumes that investment will be allocated-and hence production will respond to 2. The strength of the economywide model lies in relating demand-in such a way as to maintain approximate na- agricultural production, input use, and employment, as well as tional self-sufficiency in animal products and processed demand for food and agricultural materials and agricultural exports and imports, to corresponding developments in other sectors (in- food. Agricultural demand-supply imbalances thus show cluding food processing) and in macroeconomic aggregates. Its up mainly in projected net crop exports or imports weakness is that-to avoid unmanageable complexity-it distin- (though these are powerfully influenced by the rate of guishes only two agricultural subsectors, crops (including forestry growth of animal husbandry production, since animal and sidelines) and animal husbandry (including fishing). Annex 2, feed is an important component of crop demand). The conversely, contains projections for more detailed subsectors and potential size of these imbalances, given the overall pace some specific agricutural commodities, but without full consider- ation of all the linkages with other sectors. These two sets of and pattern of economic growth, can accordingly be in- projections, though they cannot be precisely reconciled, were vestigated by varying the assumed future rate of growth made in a coordinated way and are fundamentally consistent in of crop production (which, as already mentioned, cannot their assumptions and implications. be predicted with any accuracy). 3. Some Chinese estimates of future agricultural growth ap- The QUADRUPLE projection introduced in the previous pear significantly higher because they include rapidly expanding chapter is a convenient point of reference. It incorporates brigade and team industries. For agriculture proper, the Ministry the assumption of a rapidly changing diet, as well as percent (Annex 2p Table 5l3n aHghets imply a growth rate of 3 8 atiigthe government's other main long-term target, pret(ne ,Tb et 5.3. Hihretmtsaegvni i attaining , , , ,, , Guoguang, ed., Zhongguo Jingii Fazban Zhanlue Wenti Yanjiu (Is- but allows for some substitution of industrial for agricul- sues of China's economic development strategy), (Shanghai: tural raw materials in manufacturing. Crop production is Shanghai Renmin Chubanshe, 1983), p. 155 (4.7-5.7 percent) assumed to grow in 1981-2000 at an average annual rate and p. 404 (4.4-4.5 percent, perhaps 5 percent). 46 percent average for middle-income countries, and sur- If aggregate income growth were slower still, the pro- passed only by Thailand and the Philippines, which jected crop deficit could disappear, even with crop pro- started with more uncultivated land and lower yields. duction growth at 2.9 percent. This is so in the MODER- ATE projection (introduced in the previous chapter), in Alternative Outcomes in 2000 which lower efficiency in industrial investment causes the China's strict birth planning policies are motivated partly whole economy to grow at an annual rate about one by concern about food availability. A projection was thus percentage point lower than in QUADRUPLE, while lower made with the economywide model that maintained the agricultiral efficiency manifests itself in slower growth of assumptions of QUADRUPLE, except that the population crop production. Though crop production in 2000 in rises to 1.3 billion in 2000 (rather than 1.2 billion). This MODERNTE is 12 percent lower than in QUADRUPLE, causes a larger foreign trade deficit in crops. But the there is a significant trade surplus in crops (about 2 per- projected deficit is surprisingly small-about 2 percent of cent of domestic crop demand), because slower growth of total crop production in 2000, which is similar to that in incomes; reduces demand by even more. Direct human the early 1980s. This is because the higher population consumption of crops is only 6 percent less, mainly be- causes per capita income to be lower, which, given the cause grain remains a larger proportion of the diet at the assumptions linking consumer demand to income, causes lower per capita income level (indeed, direct grain con- lower per capita consumption of crops and of commodi- sumpticn might be higher in absolute terms if incomes ties whose production directly and indirectly requires were lower; see Annex 2, Table 2.1). But indirect con- crops.4 If per capita food consumption were unchanged, sumption-mainly animal feed-and other uses of crops the impact of a larger population would be much together are 17 percent lower. greater-an increase in direct and indirect grain require- ments in 2000 of about 40 million tons, or 8 percent of Cbangirg Imbalances over Time projected requirements (Annex 2, Table 2.3). Moreover, In considering the long-term agricultural demand-supply the possible range of alternative population sizes for balance, it may be misleading to look simply at projected China becomes far wider in the twenty-first century (see values foDr 2000. Although the present model is not de- Chapter 8). signed to analyze medium-term movements, the projec- Projections were also made under somewhat more pes- tions suggest some changes in the pattern of imbalances simistic assumptions about agricultural efficiency, both in during 1981-2000, with possible implications for policy crop production (given the same resource inputs, gross in the rest of this century and beyond. output was assumed to grow about 0.7 percent per year Figure 3.1 a illustrates this for the QUADRUPLE projec- more slowly) and in animal husbandry (feed requirements tion, by showing what happens to crop exports and im- per unit of product in 2000 up by about 15 percent).5 ports. Ulp to the mid-1990s, domestic demand for crops Retaining all the other assumptions of QUADRUPLE, falls short of production, causing a projected foreign trade lower agricultural efficiency pulls down the aggregate surplus and-not shown in the figure-abnormally high growth rate (GDP grows at about 6.4 percent per year, accumulation of stocks because of an assumed upper limit rather than 6.6 percent). Its impact on the agricultural on crop exports imposed by world market conditions demand-supply balance depends, however, on whether it (discussed further below). This represents a continuation is assumed to cause more resources to flow into agricul- of the actual situation of abundance in the early 1 980s as ture (to maintain production) or to cause slower agricul- a result of the surge in production induced by the rural tural production growth. reforms This initial surge is assumed eventually to dimin- In the former case, with growth of crop production ish, with somewhat slower growth of crop production in maintained at 3.6 percent per year, exports and imports the 1990s (3.4 percent per year) than in the 1980s (3.8 of crops are almost unaffected. This is because lower percent per year). In consequence, and despite an as- efficiency in converting feed to livestock products, which sumed reduction also in the rate of increase of food con- would tend to increase the crop deficit, is almost exactly sumption, the demand-supply balance begins to change in cancelled out by slower growth of total and per capita income, which reduces crop consumption both directly and indirectly (through lower consumption of meat and 4. In the model, faster population growth has almost no effect other commodities). In the latter case, with growth in on total national income in 2000, partly because it would not crop production reduced to 2.9 percent per year, lower significandly alter the size of the labor force until later, partly agricultural efficiency causes a substantial crop deficit by because of the assumption of surplus agricultural labor. 2000'net imports eqalton5. Tl e assumed increase in animal feed requirments is substan- 2000-net imports equal to nearly 9 percent of domestic tial. But the difference between low and high efficiency in conver- demand for crops (and nearly 20 percent of total im- sion of animal feed to meat could in fact add as much as 50 ports). Were it not for the slower overall growth of in- percent, Dr 60 million tons, to feedgrain requirments in 2000 come and demand, the deficit would be even larger. (Annex 2, Table 2.2). 47 Figure 3.1 Crop Exports and Imports, Alternative Projections, 1981-2000 Billions of 1981 yuan Billions of 1981 yuan 14 14 a QUADRUPLE projection b. QUADRUPLE projection, t higher population variant 12 - 12 - /~~~~~~~~~~ 10 - t1 - 8 /8 6 6 4 4 2 2 1981 1985 1990 1995 2000 1981 1985 1990 1995 2000 Billions of 1981 yuan Billions of 1981 yuan 14 14 c QUADRUPLE projection, d. MODERAFE projection low crop production variant / 12 - 12 - 10 = 10 I x 8 8 6 6 48 2 2 1981 1985 1990 1995 2000 1981 1985 1990 1995 2000 - - -Upper and lower bounds on exports * Trade surplus M =ImportsTrddeit X = ExportsTrddeit Source: World Bank projections. 48 the early 1990s, with an absolute decline in crop exports the substantial projected imbalances within the crop sec- and, in the last years of the century, the emergence of a tor would have to be resolved through international trade crop trade deficit. Though the deficit in 2000 is very or by attering production patterns, both of which would small, it is growing: if the projected demand and produc- be feasible up to a point, but not without problems. tion trends of the late 1990s were to continue, it would increase to 5 percent of domestic demand by 2020. INTERNATIONAL TRADE. It would generally be more Most of the other projections discussed above follow a economic to import feedgrains rather than an equivalent similar time path, with crop surpluses in the earlier part amount of animal products, simply because wages in Chi- of the period diminishing or disappearing in the later na's agriculture are likely to remain far below those of the part, although the speed of the transition varies (Figures main animal product-exporting countries. It also appears 3.1b and 3.1c). In the variant on QUADRUPLE with that China could purchase a large amount of feedgrains slower crop production growth, for example, the surplus on the world market without much effect on prices, becomes a deficit before 1990 and thus becomes large by mainly because of the large unexploited production po- 2000. MODERATE, however, is an exception (Figure tential in grain-supplying countries. World trade in coarse 3. Id): despite slower crop production growth, the slow grains, on the basis of past trends, could increase by 7 overall growth of the economy causes projected crop million to 8 million tons per year, and prices could con- exports to exceed imports throughout the period (though tinue to decline in real terms (Box 3. 1, and Annex 2, abnormal stock accumulation ceases in the 1990s); even paras. '.17-5.19). Even if the whole of the large feed- in the last few years of the century, there is no tendency grain shortfall identified above were to be met through for a deficit to emerge. imports, these would not exceed 15-20 percent of world market growth, and by the end of the century China's Product Mix witbin Agriculture imports would be unlikely to exceed 10 percent of world Expansion of animal husbandry at the high rate (7.5 trade in grain. Diversification of suppliers in recent years percent per year) entailed by demand growth in the has also reduced the ability of any single large supplier to QUADRUPLE projection might be technically feasible, but use grain embargoes or restrictions for political ends. would require attention to some curent problems and Nonetheless, it might be unacceptably risky for a country weaknesses in the livestock sector (as discussed earlier and of China's vast size to become heavily import-dependent in Annex 2, paras. 4.41-4.88 and 5.28-5.38). It would (60 million tons would be about a third of China's pro- also require substantial diversion of crop output to animal jected feedgrain requirements in 2000). feed, which could cause problems of demand-supply im- There is also the necessity of earning sufficient foreign balance within the crop sector. exchang,e to pay for these imports. This need not be Even with an overall crop balance, as in QUADRUPLE earned by agricultural exports: indeed, in some of the and in a more detailed alternative projection with lower projections mentioned above, in which there is an overall animal husbandry production (see Annex 2, Table 5.2), agricultural deficit, some of it would have to be covered the Government's present targets for increasing produc- by exports of manufactures and services (conversely, ain tion of specific crops could well lead to sizable surpluses overall agricultural surplus could be a net source of fi- of rice (possibly 30 million tons, unprocessed, by 2000), nance for nonagricultural imports). But the projected co- and of tubers and pulses, with a large shortage of feed- existence of shortages of feedgrains and surpluses of other grains (perhaps 60 million tons, or 12 percent of total crops o3viously suggests the possibility of exporting the grain production, by 2000). Though the magnitudes latter. vary, a broadly similar pattern of imbalances arises in all Here a major question arises. The surpluses are mainly the other projections described above. It occurs partly in rice, and world trade in rice is comparatively small. because direct consumption of fine grains by humans is Indeed, the potential rice surplus for 2000 mentioned unlikely to increase on a per capita basis if income growth above (some 20 million tons on a milled basis) is about as and dietary change proceed according to Government large as most estimates of total world rice trade at that targets (the projected surplus of rice would be even time (the total is now about 12 million tons and is ex- greater if China were assumed to experience the shift of pected co grow at about 2.5 percent per year), which consumption from rice to wheat that has occurred in would i nply a doubling of total trade and an increase in other Asian countries in response to higher incomes- China's share from the present 10 percent to around 50 Annex 2, para. 2.11 and Table 2 3). These imbalances percent. It is unlikely that world trade in rice could be arise also because of the projected rapid growth of de- doubled without depressing the world market price of mand and production in animal husbandry and the need rice to uneconomic levels. However, the present relative for such production to be increasingly based on grain and level of the world rice price-2.5-3.0 times that of protein-based feeds, rather than pasture or forage. wheat-has little foundation in nutritional content. And Taking the projected pattern of food demand as given, in China, the cost of rice production is probably lower 49 ::Box 3.10 WoldGranTrd, Pat ad Futuret00000000000000000 ft 00000 000 0 Over the pastwenty years, andindesincethe 1930s, morethanhalftheirgraiimports foranimal feed. world tae in cereals-wheat, rice, and coarse ns-has Most projections (see Annex 2, paras. A.17-5.20) sug- increased in volme by about 4 percent per year. Gain gest sloer growh of world grain trade in the 1 9810s. This production has grown more slowly,so that tradehas in- isX because of slower demand growth, due to increased 0 crased its jshare:i of output fiErom aboutW 400 percent0t in the 0 productiont in jIndia andf China, and -conomic difficulties 930s to 010 percent aroundf 1960 andmore than 14 per- for4some otherlarge gran importers. As a result, and 00f i; cent tat; present. Trade in coarrsegrains has: grownafthe ifast- 0 because workld production potential will still greatly exceed est, fat5percent tperf year, and riethe slowest,; at2. 5 demand (in theUnited Staetes alone, 50 million hectares of percent per year. rAltough prices have flucuated, their unurilie land could prouce an, additional 200 million trend has en dowwadin ralterms (&fo eat and tlonsper year), grain prices are expected to continue to comr a decline e World Bank projections 1981 for rieL a decline of more thani 25percent over the (made before the analysis of demand and production past twenty years) Developing counties have acounted : trends in China discussed in this chapter) suggest reduc- for two-thirds ofthe increase in world grain imports over tions of 1w1opercent,O13 percent, and7 percent in theprices the past fif to twenty years; in particar, teir coarse of rice,wheat, andr corn rpectively by 1995,:as compared grain imports have increased tenfold, mainly because of with their 1976-82 averages. dietary changes in middle-ncome countries, which use than that of wheat production (Annex 2, Table 1.3). It feedgrains are grown in the North, while most pork and could therefore be in China's economic interests to ex- poultry are produced and consumed in the South. In the pand rice exports substantially-but gradually, for future, there will therefore have to be either more North- smooth adjustment-and to reduce the world rice price South movement of feedgrains or an increase in livestock to a level much nearer that of wheat.6 production in the North for transportation to the South (which would involve a smaller volume of transport, but PRODUC'I'ION PANl-rERN. In addition, or as an alterna- more expensive refrigerated equipment). tive, to using international trade to match demand and supply for particular crops, there is scope for altering the Changing Food Demand domestic production mix. This would involve switching The preceding discussion of potential agricultural de- rice land either into coarse grain production or into the mand-supply imbalances has assumed a rapid increase in production of other crops that could be more easily ex- the share of animal products in a gradually rising total ported to pay for coarse grain imports. The extent to food intake. Because animal products are relatively ex- which either of these switches would be economic, let pensive, this implies an income elasticity of demand for alone the economically optimal combination, cannot be food (0.85) that is rather high by international standards determined on the basis of presently available informa- (0.70 would be normal for a country at about China's tion on China's soil and water resources. Much of the income level).7 Clearly, however, adjustment of the level presently irrigated rice land would not be as well suited to other crops, but in rainfed upland, rice yields are much lower in relation to those of corn and oilseeds. Major 6. To maximize China's economic gains, the price should be expansion of some ofChna'stradtioreduced to the point where the increase in foreign exchange earn- expand mus ooms r exams difficultea ings from additional rice exports-taking account of the loss of and mushrooms, for example) would be as difficult as revenue on existing exports caused by the price reduction, as well expanding rice exports; major expansion of alternative- as the increase in volume at the new price-is approximately equal and potentially promising-export crops would require to the cost of additional rice production, measured in foreign tremendous improvements in storage, transport, and exchange. The price level in question could be roughly identified in packaging facilities. advance on the basis of information on the price elasticity of world Since these changes in the production pattern would demand, but its attainment would require gradual adjustment and careful monitoring of demand and cost development. require certain areas to become more specialized in pro- 7. The income elasticity of demand for food is defined as the duction of animal feed or export commodities, the Gov- growth rate of real per capita expenditure on food divided by the ernment would have to further relax local food self-suffi- growth rate of real per capita household income. The figure of ciency requirements (see Chapter 5). In addition, the 0.85, for example, implies that each 10 percent increase in income present regional pattern of production is such that most leads to an 8.5 percent increase in food expenditure. 50 and pattern of food consumption is an additional possible Figure }.2 Per Capita Food Intake in China, response to these imbalances. 1981-2000, and in Other Selected Economies, The present food situation in China is very unusual, in 1960-80 two senses. First, by comparison with other countries: total food intake, in terms of protein as well as calories, is Food intaice (kilocalories per day) above the average for middle-income countries; yet the 3,600 Germany, Fed. Rep. proportion of animal products in the Chinese diet is simi- F lar to the low-income country average. This limits the 3,400 France relevance of international experience, since a "normal" diet at the Government's target income level for 2000 3200 Spain would involve significantly less food (though more ani- mal products) consumption than at present. Second, by Turkey comparison with the past: for much of China's history, 3,000 Egypt/ Japan feeding the population at a level-at best-just above subsistence has been a major problem; yet in just five 2,800 Hong Kong years a remarkable surge of agricultural production has transformed the situation. Even looking a long way 20 ahead, agricultural production seems likely to grow at China least twice as fast as population, which implies that there ~ ~ razil should be no difficulty in maintaining or improving on 2,400 - Indonesia Nigeria China's present satisfactory average per capita calorie and Pakistan Thailand protein intake. 2,200 - The projections discussed above also allow the possibil- ity of an increase in animal product consumption over India the next two decades as rapid as present targets imply. 2.000 But this would require crop production to grow at a high rate, international trade or changes in production pat- 1,800 , terns to balance supply and demand for particular crops, 125 250 500 1,000 2,000 4,000 8,000 16,000 and feed-meat conversion ratios to improve substantially. Per capita income (1982 U.S. dollars, logarithmic scale) There is no guarantee that any or aLL of these conditions * Relationiship in about 1960' will be fulfilled, especially over the longer term. There are * Relationiship in about 19801 some difficult technical and economic obstacles to over- * Q(UADRtD Pl.E projection, 2000 come, as well as the possible risks of dependence on feedgrain imports. The meager scale of China's cultivable a. Based cn 1960 per capita income and the 1961-65 average for food land resources and the possibility of substantial popula- intake. b. Based en 1982 per capita income and the 1978-80 average for food tion growth in the twenty-first century are also reasons intake. for caution. Even with a favorable feed conversion ratio, Source W'crld Bank data. an increase of ten percentage points in the share of animal products in the average diet is equivalent (other things being equal) to the disappearance of 35-40 percent of Chn' cutval ln, ora3.0prennraei figurc shows that it would lead to roughly the food con- China's cultivable land, or a 35-40 percent ncrease m sumption level ofJapan (under 3,000 kilocalories), rather China's population.f than to tie substantially higher levels (around 3,500 kilo- One possibility of particular concern would be an in- calories) of Western Europe and the United States. (In crease in per capita consumptio,n of animal products to a the U.S.S.R., the corresponding figure is 3,300 kilocalo- level that could not be sustained, because of a future rie iES te Europ, ing from 3,200 toc3,80 slowdown in crop production growth, inefficient meat rie;nastr n es.) production, or inability or unwillingness to import feed- B . .ocaorr C ' t grains on a sufficient scale. International experience sug- animal products, relative to the target increase in per gests that the cutback in meat consumption that would capita income, is high by international standards (Figure then be required could cause acute social problems. Alternative Food Consumption Patterns 8. This is because more than four kilograms of grain are re- China's target increase in total food intake is modest by quired to produce a kilogram of meat, which contains onlv three- international standards (Figurc 3.2). If thc 1981-2000 quarters a! many calories as one kilogram of grain consumed di- trend were continued to a much higher income level, the rectly. 51 3.3). Over similar low- to lower-middle-income ranges, Japan's food consumption pattern is of special rele- the share of animal products in the diet in most other vance to China's future development, largely because the countries either increased only slightly (Egypt and Nige- two countries have in common a very small amount of na) or declined (Indonesia, Pakistan, Thailand, and Tur- cultivable land per person (although China's is more than key). Moreover, if this Chinese trend were continued, it double Japan's). Japan's pattern differs from that of other would lead to West European levels, rather than to the industrial market economies not only in its substantially much lower level of Japan. Indeed, China at an income lower caloric intake and far lower animal products intake level of $800 would have a share of animal products in (although the latter increased from 300 kilocalories per food consumption similar to Brazil at a comparable in- day in 1961-65 to 550 in 1977), but also in the composi- come level, but 50 percent higher than Japan had at an tion of the animal products intake, with an unusually income level of $2,500, and only 25 percent lower than high share of aquatic products and unusually low shares at Japan's present income level of $ 10,000. of meat-especially beef-and dairy products. This di- etary pattern, moreover, has been achieved through rela- tively high food prices, with the result that expenditure Figure 3.3 Share of Animal Products in Total Food on food is actually a higher proportion of consumer ex- Intake in China, 1980-2000, and in Other Selected penditure in Japan (around 30 percent) than in Western Economies, 1960-80 Europe and the United States (around 20 percent). Prices for Japanese farmers that are well above world market levels-enforced by restrictions on most agricultural im- Percentage of food intake ports-have been passed on to consumers. Germany, Fed Rep. Variants on the QUADRUPLE projection were con- structed to illustrate the impact that changes in the food 36 consumption pattern in 2000 could have on agricultural France demand-supply imbalances. For example, reducing con- 32 / sumption of animal products by about 6 percent, but 3 2 increasing direct crop consumption by a roughly equiva- lent number of calories, transforms the slight overall crop 28 Hong Kong deficit into a small surplus. The impact of this change is Spain greater for individual crop subsectors: the potential rice 24 /pain surplus is reduced from about 20 million tons to 15 million tons, and the coarse grain deficit from about 60 /apan million tons to 40 million tons. 20 Another variant assumed a lower income elasticity of demand for food (0.70, as mentioned earlier), with signif- 16 /icant reductions in both meat consumption (about 12 6 ~ Pakistan China razil percent in 2000) and direct crop consumption (about 8 percent). This transforms a projected crop deficit of 0.5 12 percent of production into a 2 percent crop surplus. The change would be larger, except that the reduction in 8 / Turkey consumer expenditure on food is matched by an increase Thailand / , in expenditure on other things, especially light manufac- India Egypt tures, whose production requires considerable amounts 4 India --,, _iei of cotton and other industrial crops. The BALANCE pro- ilgeriall g 1 * .jection introduced in the previous chapter also assumes o 2 Indonesia I an income elasticity of demand for food (0.80) lower 125 2S0 500 1,000 2,000 4,000 8,000 16,000 than in QUADRUPLE, but attains almost the same level of Per capita income (1982 U.S. dollars, logarithmic scale) food consumption in 2000 because of faster growth of total household income and consumption. However, do- * Relationship in about 1960a mestic demand for crops in 2000 is 2 percent lower * Relationship in about 1980 because of the increase in the weight of services in the * QU'ADRUPLE projection, 2000 economy relative to that of industry. a. Based on 1960 per capita income and the 1961-65 average for food Managing Food Demand intake. b. Based on 1982 per capita income and the 1978-80 average for food As the preceding discussion suggests, the long-term capac- intake, Source: World Bank data. ty of China's agriculture to provide its people with, in 52 particular, a higher level of consumption of animal prod- adjusts to the changing pattern of demand, thus minimiz- ucts cannot be accurately predicted. The uncertainties are ing the osts of the potential imbalances discussed earlier. especially great in the twenty-first century-although past These issues all raise the question of how prices and other long-term predictions of food production potential (in economic levers should be used in managing household other countries and worldwide) have almost always been agriculture and of the appropriate roles of the state and unduly pessimistic, because they have underestimated the collective in the new system. technological progress in agriculture. Recent changes in government policies will substan- These uncertainties, and the adverse social conse- tially increase the regulatory role of the market. From quences of possible future reductions in per capita con- 1985, compulsory procurement and production quotas sumption of animal products (or food in general), suggest for grain and cotton have been abolished. This is to be the need for caution in managing food demand in China. extended gradually to all agricultural products, with an One essential policy objective will be to continue to guar- accompanying loosening of price controls. However, the antee a minimum supply of food to everyone, and espe- state will continue to fix the price at which it procures cially to people in poor rural areas. Indeed, some of the agricultural products under contracts that will account for present grain surpluses could be used to increase the guar- a large proportion of the total marketed by farmers. Im- anteed minimum. This would not only reduce the gap portant inputs to agriculture are still administratively allo- between the minimum and the average level of food cated. The state also remains significantly involved in intake, which has probably increased in the past five agricultural production-including the state farm sys- years, but it could also help to eliminate the stunting due tem-and directly productive investment, and it is appar- tomild chronic malnutrition that persists amonga signifi- ently envisaged that large-scale state-owned units will cant proportion of China's rural children (Annex 2, para. play an important role in the rapid growth of meat and 2.05). dairy production. At the other end of the social spectrum, considerations of efficiency and equity suggest that consumers should Output Pricing and Marketing pay the full economic cost of animal products (including The discussion of imbalances in earlier sections, and addi- the opportunity cost of diverting scarce agricultural re- tional analysis in Annex 2 (paras. 3.30-3.39), suggest sources and foreign exchange from other uses). A long- some general directions in which agricultural producer term food trade strategy, especially with regard to tariffs prices may need to move during the next few years. or quotas on imports of animal products and feedgrains, Potential shortages of animal feed and surpluses of rice has to be established in advance. Given this, it would could be alleviated by increases in the price of coarse probably be inadvisable to subsidize or administratively grains (and possibly soybeans, the most useful source of restrict either the retail prices of animal products or the animal feed protein-Annex 2, para. 4.75) relative to the prices of inputs into animal husbandry and aquaculture, price of rice. Pork and fish prices will need to be higher particularly feedgrain. The experience of Japan suggests than in the past to make it financially attractive for pro- that high prices can bring consumer demand for animal ducers to invest in the more feed-intensive systems neces- products into line with limited supplies in a socially ac- sary for rapid output expansion-and of course the re- ceptable manner. By contrast, the experience of other quired animal product price increases will be made countries with subsidized and rationed consumption of greater by the increases in animal feed prices. Higher animal products has been extremely unfavorable, both timber prices also seem necessary to stimulate produc- economically and politically, notwithstanding its short- tion, although some other tree crop products (such as tea term social appeal. and rubber) at present appear overpriced. More important than these possible specific price changes, however, is the issue of whether necessary price Managing Household Agriculture adjustments should be administratively imposed or gener- The production responsibility system has been extraordi- ated by the market. With the new system of household narily successful. Yet there may be substantial scopc for agriculture, and with increasing reform of the nonagricul- further improvement of agricultural management, in light tural economy, it will not be possible, let alone efficient, of China's own experience with the new system, as well to have an administered structure of agricultural prices as experience in the many other developing countries that ignores or fundamentally contradicts the forces of where households are the basic agricultural production demand and supply. Yet it would also be impossible to units. The major issues include: how best to stimulate calculate and apply an economically efficient set of ad- growth of household agricultural production; how to im- ministered agricultural prices, especially because they prove its economic efficiency, which is essential to in- would constantly be altering. Even at a very aggregated creases in both national and rural incomes; and how to level, the projections discussed earlier suggest the possibil- ensure that the composition of agricultural production ity of changes in agricultural imbalances over time. The 53 point is strengthened when specific categories and quality the price limits could be as significant as those of altering subdivisions of agricultural products are considered: the purely administered prices. Nonetheless, it could have detailed pattern of demand is bound to change in innu- certain advantages over the two-tier (contract and above- merable ways, with the added complication that demand contract) pricing system now being implemented in for particular products is affected by the prices and avail- China, especially in terms of administrative simplicity and ability of other products; the detailed pattern of produc- responsiveness to changing circumstances. tion costs will also change, again with the complication Though a price stabilization scheme could not be used that the willingness of farmers to supply particular crops directly to redistribute income, as is possible with two- at particular prices is affected by the prices and produc- tier pricing (discussed later), it could contribute indirectly tion potential of the alternative crops they might grow. to raising rural incomes by enabling the state to reduce its Nonetheless, experience in other countries and interna- present heavy direct involvement in agricultural com- tional markets suggests that market forces alone may pro- merce. Indeed, the Government would need to purchase duce undesirably unstable agricultural prices. Actual or and sell agricultural products only as required for stabili- anticipated changes in production-often weather-re- zation purposes and to secure emergency stocks and sup- lated-and demand can cause large short-term price fluc- plies for poor areas. Scarce state resources could thus be tuations that inflict needless hardship on producers or released from commerce for other, more vital, agricul- consumers, increase the cost of adjustment to fundamen- tural support activities. And allowing collective and indi- tal changes in economic circumstances, and create risks vidual enterprises a large share of (at least) the increment that can discourage farmers from specialized production in agricultural commerce over the next two decades could for the market rather than subsistence. For these reasons, make an important contribution to boosting rural non- there are few countries in which governments do not in farm incomes-as experience in other countries confirms. some way intervene in agricultural price formation. Simi- The same applies to food processing, whose volume is larly, the Chinese government intends to limit the influ- projected to increase nearly fivefold between 1981 and ence of market forces on agricultural prices: although the 2000, and to the processing of other agricultural prod- prices of minor products will be allowed tO float freely, ucts. for major products market-determined prices will apply only to above-contract sales (that is, sales in excess of Input Pricing and Allocation contracted state procurement) and even then will be sub- Under the production responsibility system, farmers have ject to a state-set floor price. a strong incentive to make more efficient use of industrial A possible alternative approach for major agricultural and other inputs to agricultural production, applying products would be for the Government to let prices be them just up to the point where their contribution to determined by demand and supply, but to intervene in increased output value is balanced by their cost. How- the market to prevent excessive fluctuations. At any ever, to ensure that farmers use inputs in a way that is given time, prices would be allowed to move only within optimal for the economy as a whole, it is necessary that a range established by the Government's willingness to output prices and input prices be economically rational, purchase for addition to its stocks at the lower end and to and that farmers be able to obtain sufficient quantities of sell from its stocks at the upper end. The range could be inputs at these priccs. At present in China these condi- narrow, which would require frequent interventions at tions are not always fulfilled. both ends and comparatively large stocks. Alternatively, Chemical fertilizer is a good example (Annex 2, paras. as in most schemes of this sort elsewhere, the range could 3.17-3.25, 3.35-3.36). Although its price seems quite be wider, with price volatility reduced by starting inter- rational by international standards, especially in relation vention on a modest scale before prices reach the estab- to the price of wheat, it is allocated to farmers through a lished limits. Over time, if a price persistently remains at cumbersome administrative procedure. The allocation its upper or lower limit, with the Government's stocks criteria are not at all clear and are not always followed, becoming very small or very large, the Government grad- although they appear to include priority for food crops ually alters the two limits, shifting the whole range in the over feedgrains and compensation for state procurement direction required by demand and supply trends. at low prices. In any event, the present system does not To operate such a price stabilization scheme would not allocate fertilizer among farmers or localities in a way that be simple. It would be technically quite complicated, maximizes its contribution to production (which would especially given the likely need for coordination with involve equalizing the marginal output response to a kilo- export and import policies and for interregional price gram of fertilizer in all uses). variations within China. It would involve financial costs, Nor, more importantly, could any administrative sys- especially because it is possible to be wrong about market tem of fertilizer allocation in practice achieve this desir- trends (and hence to adjust price limits too soon or too able result: the yield response to fertilizer varies not only late); and the political and social repercussions of altering among crops and localities, but also among individual 54 plots of land within a given locality. Individual farmers needed to improve rural transport, storage, and market- are usually better informed about such variations than ing ficilities (Annex 2, paras. 3.61-3.70). At present, government officials and could use their knowledge in China's rural road network is less than half the size of deciding how much fertilizer to purchase in a competitive India s; there are fewer trucks per capita in rural China market. China should thus consider moving toward mar- than in the impoverished Sahelian countries of West Af- ket allocation of fertilizer, although this would make it rica; and there is less cold storage capacity in the whole of even more necessary to maintain economically rational rural China than in northwest India (with a population of agricultural product prices. The immediate impact of 50 nrillion). Improvements in these aspects of China's market regulation on the general level of fertilizer prices agricultural infrastructure would almost certainly be a cannot be predicted, but could be cushioned by suitable better use of Government resources than, for example, changes in imports. The level of imports, in conjunction direct state investment in animal husbandry, which in with economic calculations of the relative cost of imports most countries is inefficient compared with household and domestic production, would then provide a guide to production (Annex 2, para. 4.70). investment needs in the domestic fertilizer industry. Expansion of collective activities in certain areas-espe- Similar arguments for moving, in the context of a sys- cially marketing and input supply-could also make an tem of household agriculture, away from administrative important indirect contribution to the development of measures and toward market allocation of inputs (with household agriculture. The original collective principles market-regulated prices) can be made for pesticides, ani- of ClHna's supply and marketing cooperatives are being mal feed, farm machinery, fuel, and so on. They also revived, but they could be given a much larger role in the apply to credit, which, together with the farmer's own commercial system. Both in perishable crops and in ani- saving, is in household agriculture the main source of mal husbandry, including dairying, cooperative market- finance for fixed investment and of circulating capital. ing facilities in other countries have very effectively stim- Extensive experience with rural credit in other developing ulatec and facilitated production, as well as increasing countries has made clear the disadvantages of subsidizing farmers' incomes. interest rates and of the administrative rationing conse- quently required by excess demand for available funds. China's farmers, including the poorer farmers, will proba- Agricultural Incomes Policy bly be able to get more credit, and hence to increase their Increased state and collective activity in these various production and incomes more, if the interest rates areas, by helping farmers to increase production, to re- charged reflect the scarcity of credit and repayment is duce costs, and to market their output, could make a vital strictly enforced. These conditions would also make it contribution to raising agricultural incomes over the next harder for China's rural credit institutions to discriminate two dacades. So could more economically rational alloca- against loans to households and in favor of loans to col- tion a'sd pricing of agricultural inputs. But are there other lective or state-sponsored activities (Annex 2, paras. measures that could help to close the large gap between 3.45-3.46). And they would discourage uneconomic agricultural and nonagricultural incomes? (Policies to- mechanization.' ward very low and very high incomes within agriculture are considered in Chapters 5 and 9, respectively.) Economic forces are, by themselves, unlikely to lead to The preceding discussion has identified a number of ar- an early reduction in this gap. Indeed, experience from eas--pricing, marketing, input allocation-in which less other countries suggests that the gap is more likely to extensive or direct government involvement in managing widen-even in proportional terms-during the low- to China's household agricultural system could be advanta- middle-income transition. The basic problem-in China geous. At the same time, there are a number of areas as in cther countries at a similar stage of development-is (identified in Annex 2) where increased government ac- that agricultural labor productivity does not increase fast tivity would be highly desirable. Given overall limitations enough to close the gap, even when agricultural output on government resources for agriculture, moreover, it grows quite rapidly, unless or until agricultural employ- may in fact be essential to curtail some activities in order ment starts to shrink absolutely. In the QUADRUPLE pro- to expand others with larger economic benefits. Among the areas for increased state activity identified in Annex 2 are agricultural research, education, and ex- 9. ]Especially in poorer localities, where it would aggravate the tension services (paras. 3.48-3.58), range management labor surplus, with (as in other countries) potentially untortunate 4.48-4.51), fertilizer utilization studies (para. social and economic consequences. In richer localities, shrinking (paras. 4.48-4.SI), fertlllzer utlllzatlon studles (para. agricultural labor surpluses will make mechanization more attrac- 3.25), and seed testing and certification (para. 3.27). tive to farmers, but the alternative of permitting more movement Equally important, state support-both directly and by of labor from poorer to richer areas may be beneficial to the encouraging collective and individual provision-is national economy (this issue is discussed in Chapter 5). 55 jection, for example, although gross agricultural output products, which-as emphasized earlier-will be very im- increases at 4.5 percent per year, agricultural employ- portant in managing China's new system of household ment also rises (at 0.9 percent per year). Thus, although agriculture. Demand and supply forces, coupled with for- average gross output per worker in agriculture almost eign trade policies, may require agricultural price in- doubles over the period 1981-2000, it increases more creases-as for example in Japan, where high agricultural slowly than in other sectors. Even in the BALANCE pro- prices are a concomitant of the Government's policy of jection, where faster growth of employment in services limiting agricultural imports. But this will not necessarily causes slower growth of agricultural employment (0.2 be so, or not to a sufficiently great extent. And the percent per year), agricultural labor productivity-net of experience of Western Europe, where artificially high ag- increased outlays on inputs such as fertilizer-increases ricultural prices are maintained specifically to boost farm slightly more slowly than in other sectors. incomes, has not been favorable, because of the financial Agriculural employment tends to continue to expand cost of stocking the resulting surpluses or exporting them in the early phases of development fundamentally be- at a loss. cause it constitutes such a large initial share of total em- In China, however, additional room for maneuver is ployment. The absolute number of farmers can therefore provided by the two-tier agricultural pricing system. Even be reduced only by extraordinarily rapid growth of non- following the recent decision to unify the state procure- agricultural output (as in South Korea in 1960-80, with ment prices for grain, cotton, and oilseeds (based on industry growing at around 15 percent per year) or by weighted averages of the former quota and above-quota unacceptably slow growth of nonagricultural labor pro- prices), procurement prices will generally differ from the ductivity (Chapter 2). But in the twenty-first century, the market prices for above-contract sales and purchases. situation in China will become more favorable, mainly These market prices will strongly influence farm house- because agriculture's share of the total labor force will hold decisions to increase or decrease production of par- have shrunk. In 2000-2020, for example, if the total ticular commodities, and hence are the prices relevant to labor force were to grow at I percent per year,10 and balancing demand and supply. But the state procurement nonagricultural employment were to continue to grow at prices will powerfully affect the average price that farmers the same rate as in QUADRUPLE (3.5 percent per year, receive and hence their incomes. In the past, state pro- from a share of 41 percent in 2000), the agricultural labor curement prices have generally been below market force would decline from 3 70 million to 2 55 million, or prices-the difference being an implicit tax on agriculture. 3 3 percent of the total labor force. The corresponding But the Government apparently envisages that the new extrapolation for BALANCE, with faster growth of nonag- unified procurement prices for major crops will be above ricultural employment, would be from 3 30 million to 85 their market prices (otherwise it would not be possible to million, or 11 percent of the labor force (similar to Japan replace compulsory procurement quotas with voluntary in 1980). contracts). For the remainder of this century, however, the Gov- The immediate impact of the latest price and procure- ernment may need to intervene actively to prevent the ment reforms on farm incomes will depend on whether gap between agricultural and nonagricultural incomes the market prices for the crops concerned turn out to be from becoming (or remaining) unacceptably wide. Price above or below their unified procurement prices. But in adjustments are one obvious mechanism for transferring the longer term, regardless of whether market prices are the benefits of nonagricultural productivity growth (in initially above or below procurement prices, the Govern- both rural and urban areas) to agricultural workers, ment could in principle boost the incomes of farmers by rather than giving them mainly to nonagricultural work- increasing procurement prices, while allowing market- ers in the form of higher wages. To illustrate this, the determined above-contract prices to equate supply and macroeconomic projections specifically assume that the demand for the various agricultural commodities. To fi- relative price of agricultural output adjusts so that agricul- nance its increased expenditures on procurement, the tural earnings per worker remain a constant proportion- government would of course have to raise additional rev- about one-half-of nonagricultural wages. In QUADRU- enues (or reduce other expenditures or subsidies), largely PLE, for example, the relative price of crops has to at the expense of the nonagricultural population. But the increase by 17 percent over the period 1981-2000 to same would be true of any redistributive scheme to nar- prevent the earnings gap from widening. In BALANCE the row the agriculture-nonagriculture income gap by a given required relative price increase for crops is 6 percent. amount. The key issues are rather (a) how much redis- Larger increases in agricultural prices could in principle tribution from the nonagricultural to the agricultural reduce the gap in earnings between agriculture and non- agriculture. But they might conflict with other objectives of China's agricultural price policy, including the need to 10. This is similar to the fastest of the three labor force projec- balance demand and supply for particular agricultural tionsin Table 8.3. 56 population is desired; and (b) whether increasing procure- farm households-an administratively costly process, ment prices within a two-tier system is the best of the with wide scope for abuse, especially since considerable various alternative redistributive schemes available, discretion would have to be given to local officials. Direct Although the first of these issues is fundamentally a incorne supplements to broad categories of farm house- political matter, it is complicated by the fact that the holds would be an alternative means of redistribution incomes of suburban farmers-who are in close proxim- that might pose fewer administrative problems, because ity to the urban population-are generally substantially of its greater simplicity and transparency. higher than those of the majority of farmers in more Another means of redistributing nonagricultural pro- distant rural areas. Redistributive schemes that raise the ductivity gains to farmers that would probably pose even incomes of all farmers across the board are thus more fewer administrative problems is increased government likely to provoke an unfavorable response from the non- expenditure on development of agricultural infrastructure agricultural population than schemes whose benefits are and support services. Greater budget expenditure on ru- concentrated on farmers with average and below-average ral social services, particularly education and health, incomes. This could be a disadvantage of redistribution would constitute yet another approach, as would subsi- through two-tier pricing-although it could be minimized dies to a rural social insurance scheme that provided pen- (as appears to be the Government's intention) by exclud- sions and other welfare benefits (discussed in Chapters 8 ing the most important suburban agricultural products, and 10). All these approaches naturally have some limita- such as meat and vegetables. tions, especially as to the amount of money that could Redistribution through two-tier pricing would also be usefully be spent and the possibility of spreading it equita- technically quite complicated, especially if it were to be bly within the agricultural population. But, in conjunc- done on a large scale and if serious inequities among tion with two-tier pricing or otherwise, they offer sub- farmers in different places and situations were to be stantial scope for economically efficient government avoided. This is because its impact depends crucially on action to raise agricultural incomes. the allocation of procurement contracts among individual 57 Energy Development Energy, though a much smaller sector than agriculture, is most important energy source (more than 25 percent of just as critical to China's future growth and subject to primary consumption). Oil-all domestically produced- similar uncertainties. There is uncertainty about energy accounts for approximately one-fifth, and natural gas for supply prospects, particularly for oil and gas, but also for less than 3 percent, of primary commercial energy con- coal and electricity, whose successful development will sumption. Electricity in 1980 accounted for just 18 per- require much investment and planning. Uncertainty also cent of final commercial energy consumption, a far lower surrounds both the magnitude of potential economies in share than in most countries.' energy use (by reducing waste, modernizing technology, Industry-especially metallurgy, chemicals, and build- and changing industrial structure), and how much of this ing materials-accounts for at least half of the final con- potential China will be able to realize without compro- sumption of coal, oil, and gas, and three-quarters of elec- mising standards of living. Equally important, coal is tricity consumption. Households and commerce together likely to remain in China-unlike almost all other coun- accounted for only 20 percent of final commercial energy tries-the overwhelmingly predominant source of energy. consumption in 1980-but, including biomass, for ap- To harness it cleanly and efficiently in an unusually wide proximately 43 percent of final energy use (the same range of uses will be a major challenge. share as for industry). The transport sector's share of final This chapter, which is based largely on Annex 3, first commercial energy consumption (8 percent) is low com- reviews overall energy demand and supply prospects and pared with other large developing countries, because of targets to the year 2000. It then looks at issues and the small role of road transport. policies in specific energy subsectors, with special refer- ence to the production, transport, and use of coal. Fi- Energy Intensity nally, it reviews possible investment requirements in the Cross-country comparisons suggest exceptionally high sector and the future role of planning and demand man- consumption of both commercial and total primary en- agement. Two common themes-which will echo ergy per unit of GDP in China. This can be explained through subsequent chapters-are the importance of partly by China's high share of industrial output in GDP more general improvements in the efficiency of resource and by space heating requirements. But energy consump- use in China and the need for economic as well as techni- tion per unit of gross output value in Chinese industry is cal analysis in choosing among options. Both will require also exceptionally high compared with most other coun- the strengthening of economic planning and intersectoral tries. Although the share of energy-intensive subsectors in coordination as well as increased reliance on market regu- industry is not significantly different from other coun- lation. tries, Chinese industrial output is weighted toward en- ergy-intensive goods. High energy consumption in indus- Demand and Supply: Trends and Balances trial production is also explained by relatively backward Coal already dominates energy consumption in China. In 1980, it accounted for about three-quarters of primary 1. Final ener consumption excludes transformation losses in commercial energy consumption and one-half of total I power electric power and other energy sectors (about 1 5 percent of pri- primary energy consumption (including biomass), the mary commercial energy use in 1980); commercial energy excludes highest share for coal in any major country. Biomass biomass fuels (about 27 percent of total primary energy use in fuels-used mainly in rural households-are the second 1980). 58 technology, the small scale of industrial plant, and the resource use), energy demand grows significantly more type of fuel used (particularly coal). Low energy prices slow .y-by 3.8-5.0 percent per year from 1980 to 2000. and insufficient cost consciousness among enterprises and In the MODERATE scenario (with GDP growing at 5.4 planners also appear to have contributed substantially to percent per year, as compared with 6.6 percent in QUAD- the low efficiency of energy use in China. RUPI.E and BALANCE, and with the same emphasis on In recent years, steps have been taken to reduce both expansion of manufacturing output as in QUADRUPLE), the growth of overall energy consumption and energy use energy demand grows more slowly still-by 3.4-4.6 per- per unit of output. These steps have so far consisted cent per year. mainly of regulations concerning the amount of energy Tli e Government's original tentative target was to use per unit of output, reductions in energy supply quo- quadruple GVIAO between 1980 and 2000 while only tas, and bonuses for energy use below established quotas. doulling the production and use of energy (to somewhat Conservation centers have been established to provide less than 1,300 million tons of coal equivalent of primary technical assistance. A larger share of investment funds commercial energy). This was to involve doubling the has recently been allocated for technical transformation production of coal, oil, and natural gas (output levels of of existing enterprises and in particular energy conserva- 1,201) million tons for coal, 200 million tons for oil, and tion measures. Financial incentives are now receiving 25 billion cubic meters for gas by 2000). Preliminary greater emphasis, and there has been some movement forecasts by the Ministry of Water Resources and Electric toward rationalizing the energy price structure. Pow.-r (MWREP) suggested that electricity generation As a result, during the past five years an impressive would increase from 301 terawatt-hours in 1980 to reduction in energy use in relation to economic activity 1,000-1,200 terawatt-hours by 2000-with some 230- has taken place. Energy consumption per unit of gross 280 terawatt-hours from hydro and nuclear power. value of industrial and agricultural output (GVIAO) was The energy demand projections discussed above, and reduced by 7 percent per year during 1979-81 and by 3 more recent government estimates, suggest that it will percent per year during 1982-83. Preliminary data indi- probably be necessary to increase the production of en- cate a further reduction of almost 7 percent in 1984. ergy somewhat faster than originally envisaged (Table Technical and operational improvements are estimated to 4.1). For example, even under optimistic assumptions have accounted for about 40 percent of these energy abou. improvements in energy efficiency, the estimated savings. The other 60 percent came initially from a de- growth in energy demand resulting from quadrupling dline in the relative importance of heavy industry and the GVIAO (4.2-5.5 percent per year) would be more than closure of some inefficient small-scale plants, but since could be accommodated by doubling energy production 1981 mainly from structural changes within major indus- (which implies a growth rate of about 3.5 percent per trial subsectors. year) Increasing oil and gas production substantially above the original targets is a possibility that cannot be Demand and Supply Prospects relied on, and it would be extremely difficult to increase To investigate the future growth of demand for various primary electricity production much faster than already types of energy in China, the projections made with the planned. For these reasons, and because the need for multisectoral economic model (discussed in Chapter 2) more than the originally anticipated amount of energy were combined with a plausible range of assumptions production is concentrated on fuel, the burden of bring- about changes in unit energy consumption in particular ing energy production into line with demand is likely to sectors. The resulting estimated annual average growth fall nr ainly on coal (oil distillate and total electricity pro- rates of primary commercial energy demand between duction plans conform reasonably well with the demand 1980 and 2000 range from 3.4 percent to 5.5 percent. projections). Projected elasticities of growth in energy demand relative In the QUADRUPLE and BALANCE projections discussed to growth in GDP range from 0.6 to 0.85, which is far in Chapter 2, it is assumed that coal production in 2000 lower than the past trend in China (1.5 in 1965-78), will r.2ach 1,400 million tons-a figure regarded as feasi- though higher than the elasticities realized during the past ble by relevant agencies in China, even allowing for the few years (0.2 in 1978-81 and 0.5 in 1981-84). need to develop additional transport facilities and other In the QUADRUPLE scenario (which involves the quad- infrastructure. How much coal will in fact need to be rupling of GVIAO), the projections indicate that total pri- prodLced, however, will depend heavily on the rate of mary commercial energy demand could be expected to econemic growth, on the sectoral structure of the econ- grow at 4.2-5.5 percent per year, depending on the de- omy, and on the degree of success in reducing energy gree of success in reducing unit energy consumption. In intensity within individual sectors. the BALANCE scenario (which has the same rate of In QUADRUPLE, production of 1,400 million tons growth of GDP, but with slower growth in manufacturing would meet coal demand only on the most optimistic output, faster growth of services, and greater efficiency in assumptions about reductions in energy consumption per 59 Table 4.1 Energy Supply and Demand, Alternative Projections, 1981-2000 Totalprimary Fuel Oil distillates Electrity commeraal energy Component (millions of TCE)a (millions of tons) (terawatt-bours) (millions of TCE) Projections of demand QUADRUPLE 1,135-1,455 100-140 1,060-1,285 1,385-1,765 MODERATE 955-1,225 85-115 885-1,070 1,180-1,500 BALANCE 1,030-1,315 90-130 955-1,150 1,270-1,610 Production levels in line with original targets 1,020 112 1,000-1,200 1,275-1,295 a. Fuel includes coal, fuel oil, and natural gas for power generation, for other energy industries, and for final consumption. Frc indicates tons of coal equivalents. Source: Annex 3, Chapter 2. unit of output in each sector. On the least optimistic (but together contain only 10 percent of the economically not implausible) assumptions about unit energy consump- recoverable reserves, but by international standards are tion, coal demand in QUADRUPLE in 2000 would exceed comparatively well endowed; recoverable reserves in 1,800 million tons. Even in MODERATE, with signifi- these two regions are sufficient to sustain 1980 produc- cantly slower economic growth, coal demand in 2000 tion levels (210 million tons) for more than 300 years. would on these less optimistic assumptions approach Production costs are low by international standards. 1,500 million tons. In BALANCE, with slower growth of Operating costs are typically less than Y 20 per ton, while industry compensated by faster growth of services, and capital costs average about Y 110 per ton of capacity.2 higher overall economic efficiency, projected coal de- But geological conditions and mining costs vary widely mand is substantially lower than in QUADRUPLE-1,400 among different coal producing areas. For example, costs3 million tons on intermediate assumptions about unit en- are as low as Y 30 per ton for extracting the few deposits ergy consumption, and 1,200 million tons on the most amenable to open-cast mining in remote areas of Nei optimistic assumptions. Monggol and Shanxi and are approximately Y 40 per ton for the huge underground reserves at moderate depth in Coal Production, Transport, and Utilization Shanxi province. But smaller and less favorable deposits Coal close to major consumption centers in the South and East In any event, the share of coal in commercial energy use might have production costs of up to twice those in the by 2000 is likely to remain exceptionally large-some North. 70-75 percent (compared with 30-35 percent world- Between 1965 and 1979, coal production grew at an wide). Even if oil production reaches 200 million tons by average rate of almost 30 million tons per year. Coal 2000, it would still account for less than one-quarter of production fell by 15 million tons in 1980, however, and commercial energy production. If the target of doubling remained at the same level in 1981, because of insuffi- gas production by 2000 is achieved, natural gas would cient attention to mine development during the 1970s continue to account for only 2-3 percent of primary and the economic adjustment program of 1980 which energy consumption. Even if the gas production target sharply reduced investment allocations. Production has were exceeded, which is quite possible, it is extremely since picked up, increasing by about 45 million tons per unlikely that its share of energy consumption in 2000 year during 1982-84. would exceed 5 percent. (By contrast, natural gas pro- An average increase of about 40 million tons per year vides close to 20 percent of primary energy in most major would be required to reach an annual production level of industrial countries, and 30 percent or more in the 1,400 million tons by the year 2000. To achieve this will United States and the U.S.S.R.) China thus inevitably require further strengthening of production incentives, a faces some unusually substantial challenges in coal pro- stepped up program of project preparation (including duction, transport, and utilization. Coal Mining 2. Y 140-150 per ton for large-scale central mines and Y 75 per China has abundant reserves of good-quality coal that can ton for small and less mechanized local mines (excluding social be mined at relatively low cost. Economically recoverable infrastructure). 3. Including an appropriate charge for capital recovery, for ex- reserves are concentrated in the North, which accounts ample, 15 percent of invested capital per year (based on an interest for two-thirds (Shanxi province alone for one-third) of rate of 12 percent, reflecting the opportunity cost of capital and an the total. The populous southern and eastern regions economic life of fifteen years). 60 mine infrastructure), and coordination between the Min- transfers (for example, rail to ship to rail to road) are istry of Coal Industry (MoCI) and the transportation min- required to transport coal, say, from Shanxi to an inland istries to increase coal transportation capacity more city xn the South, total transport costs could be Y 40-60 quickly than originally planned. Organizational changes per :on, justifying coal production in the South even for central mines could also help to speed project imple- under unfavorable conditions and at high costs (up to Y mentation. State-imposed design norms leave little flexi- 80-100 per ton-or 2 to 2.5 times the production costs bility for taking account of geological conditions in the at th~ more favorable deposits in Shanxi province). design of specific coal mining projects, which can lower Nonetheless, the concentration of China's coal reserves the cost effectiveness of some projects. Time required for in the North requires increasing large-scale, interregional mine construction is currently longer than in other coun- coal :ransportation. At present, railway transport is prob- tries for similar work (construction time is more than six ably more of a constraint on energy supplies than coal years for central mines) and could be shortened by better mine development, because of insufficient railway invest- coordination among the many government entities in- ments in the past and the slow replacement of inefficient volved in project execution. steam locomotives by electric or diesel traction. It will Equally important is the balanced development of cen- take several years to complete the railway projects that tral and local mines. Central mines, primarily operated by were recently initiated to alleviate critical bottlenecks be- MOCI, and local state and collective mines each accounted twee:i major mining areas and the coast. If coal produc- in 1984 for about one-half of production. Because local tion were to reach 1,400 million tons in the year 2000, as mines are smaller and less mechanized they require only much as 450 million tons might need to be transported about half as much capital spending per unit of output as out of Shanxi and adjacent areas in Nei Monggol and large-scale central mines. And because they are more Shaanxi, even if current plans for developing mine-mouth widely scattered across the country they put less pressure power generation were implemented. To handle this traf- on transport infrastructure than the large-scale mines in fic, it would probably be necessary to construct an addi- the major coal basins (one-third of the coal produced by tional new double-tracked electrified line for heavy unit local mines comes from south of the Chang Jiang, as train (similar to the new Datong-Qinhuangdao line cur- compared with only 3 percent of coal produced by MOCI rently planned) in a southeastern direction and to in- mines). In recent years, the existence of a market for coal crease capacity on existing lines through double tracking supplies beyond plan allocations has contributed to the and electrification. Slurry pipelines might ease the pres- development of local mines, which accounted for most of sure on the railways, but water shortages in Shanxi could the overall increase in coal production in 1979-84. If, in constrain their widespread use. addition, access to investment funds and transport facili- Since a considerable part of incremental coal produc- ties were improved, local mines could continue to play an tion (an additional 2 50 million to 350 million tons) will important role in meeting energy demand, especially in be for electric power generation, coal-fired power genera- areas with less favorable coal deposits and in supplying tion n mining areas in conjunction with long-distance local needs. There is also considerable scope for improv- transmission of electric power to load centers represents ing mining techniques in local mines, with selective intro- one alternative to transporting large quantities of coal. duction of mechanization and greater attention to mine Current preliminary plans are to locate roughly half of safety. However, large coal deposits are often more suit- thermal power capacity additions during 1986-2000 in able to large-scale development with higher mechaniza- coal mining areas. The relative economics of electricity tion, lower costs, and better safety and resource recovery. transmnission compared with rail transportation of coal are project-specific, as costs vary substantially. However, Coal Transport in most cases, rail transport of coal would be less expen- Transport costs greatly affect the optimal spatial pattern sive than long-distance transmission of electric power (see of coal production and utilization. Just as coal mining AnneK 3, Appendix E). The optimal mix of mine-mouth costs vary considerably in China, depending on the type power generation and load-center-based power plants of deposit, coal transport costs also vary widely depend- will therefore be determined mainly by other factors such ing on factors such as the available transport mode (rail, as the availability of low-quality coals or middlings from ship, road), degree of capacity utilization, and topogra- coal preparation plants (discussed later), the availability of phy. For example, long-run marginal costs (LRMC)4 for water for mine-mouth power plants, and the potential for transport over a 1,000 kilometer distance by double- cogeneration of electric power and heat or steam for track railway may be some Y 20 per ton, but about Y 30 per ton on a single-track railway line. Coastal shipping over several thousand kilometers may be Y 20-30 per 4 Total costs including capital construction, operating, and ton, but road transport costs might be as high for a maintenance costs, calculated by using an appropriate factor to distance of only 200-300 kilometers. If several modal reflect :he opportunity cost of capital. 61 most industrial countries coal accounts today for only Table 4.2 Share of Solid Fuels in Commercial Energy 20-40 percent of energy consumption (Table 4.2), and Use in Selected Countries, 1960, 1980, and 2000 even less in most developing countries. Coal use in other trcenlt) countries is thus by and large limited to large-scale uses Countoy 1960 1980 2000 such as electric power generation, the steel and cement United States 24 23 .. ndustries, and large industrial boilers, where environ- Japan 54 19 . mental control measures can be more easily implemented Germany, Fed. Rep. 78 33 and are less costly than in small-scale uses in industry or South Korea 79 36 . the residential and commercial sectors. U.S.S.R. 64 33 In some respects, energy utilization patterns will have China 96 77 70-75 to remain quite different in China from those prevailing in most other countries today. But a number of industrial Source United Nations, Energy Statistics Yearbook (New York, 1982); An- A nex 3. countries used to have a similarly high share of coal use twenty to thirty years ago and are now again increasing the role of coal, albeit not to the high levels that are likely to prevail in China. The main issue in China will be how industrial or space heating uses in the case of power to use coal in environmentally acceptable and affordable plants located near load centers. These economic choices ways, especially in urban areas where air pollution has in need to be thoroughly studied before long-term plans can many cases already reached very high levels. Making coal be finalized. use environmentally acceptable will require many things, While railway transport of coal is likely to remain a but will often involve replacing decentralized and uncon- constraint on energy supplies for some years to come, it trolled combustion of coal with combustion in large facil- nevertheless remains one of the most economical means ities with better environmental controls, and distribution of energy transport in China, as other technologies (water of clean energies (for example, steam and hot water) from transport, slurry pipelines, power transmission) are either these facilities to final consumers. However, these distri- not as widely applicable or more costly. Lead times for bution networks are usually very costly, and careful stud- railway investments need not be longer than for other ies need to be carried out to minimize costs and to avoid alternatives. But long-term investment plans for railways duplication (for example, between district heating and gas and other transport facilities need to be closely integrated distribution). Particularly uncertain is the extent to which with coal mining development, and sufficient resources coal conversion (coal gasification and liquefaction) tech- will have to be allocated, especially to ease critical bottle- nologies can be justified at present, given the cost and necks. An optimal coal development and transportation availability of alternative clean fuels or alternative ways to strategy must also incorporate measures to reduce trans- make direct coal combustion cleaner and more efficient port requirements, for example, through balanced devel- (for example, cogeneration and district heating). opment of large-scale coal mines in major coal basins High priority will have to be given to the transfer, remote from demand centers and mining under less favo- development, and introduction of coal utilization tech- rable conditions and at higher costs in smaller deposits nologies including coal preparation and cleaning, im- closer to consumption centers. Other means to reduce proved boiler systems, heat recovery equipment, process transport requirements include optimal location of raw controls, cogeneration systems, district heating technol- material processing industries and more extensive use of ogy, coal briquetting technology, and emission control coal preparation to reduce transport of inert material, devices such as electrostatic precipitators and bag filters. possibly in conjunction with mine-mouth power genera- Improved coal utilization also will require greater coordi- tion. Realignment of coal and transport prices to reflect nation throughout the entire coal production, beneficia- economic costs and greater enterprise autonomy in ar- tion, distribution, and consumption chain, calling for ranging supply could greatly facilitate development of greater interaction between the many agents involved least-cost solutions for coal production and transport. (for example, coal producers, transportation ministries, the power sector, commercial departments, municipal Coal Utilization governments, and industrial enterprises). Many of the Improving the efficiency of coal use could much reduce options for improving the utilization of coal (for example, the amount of coal that will need to be mined and trans- cogeneration, district heating, coal gasification) cut across ported. Even so, projected increases in coal consumption the responsibilities of many different institutions. In- will have serious implications for environmental pollu- creased cogeneration could result in major energy savings tion-and hence high welfare and health costs-unless (see Box 4.1), but would require institutional arrange- improved environmental protection measures are ments for power generation to be more flexible. adopted. This is particularly important in China because Coal beneficiation through washing has been slow to coal is expected to remain the dominant fuel, whereas in develop in China (about 18 percent of China's coal was 62 Box 4.1 Cogeneration for Industry and District Heating Cogeneration, or combined heat and power (CHP), is the ably more attractive, especially if environmental benefits combined production of two forms of energy-electricity are considered, and could justify construction of large CHP or mechanical power plus useful thermal energy-in one plants in industrial estates to supply steam within a radius technological process. The total amount of fuel needed to of, say, five kilometers. produce both electricity and thermal energy in a cogenera- District Heating. District heating has recently received tion plant is less than would be needed to produce the renewed interest in view of sharply increasing prices of same amount of electric and thermal energy separately. clean fuels such as distillate oil or natural gas. District Cogeneration is used intensively in Western and Eastern heating is also a way to increase the use of coal and other Europe, accounting for up to 30 percent of installed elec- lower-grade fuels to meet residential and commercial en- tric power capacity, and is about evenly split between ergy requirements in an environmentally acceptable way. industrial and district heating uses, but its use for district Althougi network costs for district heating are high (de- heating has recently expanded. In West Germany, for ex- pending mostly on urban density). the superior efficiency ample, cogeneration for district heating doubled over the of heat extraction from large CHP plants usually more than past ten years while industrial cogeneration stagnated, as compensates. Improved materials and network construc- industrial restructuring and improved energy conservation tion tec,niques have reduced the cost of heat distribution. reduced industrial heat requirements. In the United States, At the kyw temperatures required for district heating (70- the role of cogeneration is small and limited to industry. In 140°C), cogeneration is particularly efficient, with incre- China, the role of cogeneration is relatively small (Box mental fuel requirements for heat extraction equivalent to Table 4.1A), and its technology not very advanced by only about one-fifth of conventional boiler fuel require- international standards. ments. In Europe, a large share of new coal-fired electric power plants are now being built as cHP plants. New CHP plant de!;igns have been aimed at increasing efficiency of Box Table 4. AA. Estimates of Installed Cogeneration heat extraction and maximizing flexibility between electric Capacities in Selected Countries power and heat dispatch. Percentage of eectn pwer Market penetration of district heating has reached very Total generaing caacity high levels in Eastern Europe and Scandinavia (up to 50 Total Distria Country (gigawatts) Total Industrial heating percent of total space heating requirements). Apart from climate, market penetration of district heating is related to United States 14.9 2.3 2.3 8 availability and costs of aternative clean fuels, particularly United King- 9natural gas. Countries that rely mostly on market mecha- dom 2.5 3.6 3.6 O nisms in energy supply have found that the most impor- Finland 3.2 29.4 16.5 12.8 tant obstacles to wider use of district heating are existing U.S.S.R. 76.6 27.3 .. .. natural gas distribution networks that make market pene- China 4.9 7.1 tration for a new and costly distribution system very diffi- cuk. But district heating is often the lower cost alternative Source, Union Internationale des Distributeurs de Chaleur (Zurich), Inter- if natural gas is scarce and gas distribution networks do not national District Heating Conference, Kiev, 1982; World Bank staff esti- ye eit, an If urban desitia c hetigrequire- mates. yet exist, and if urban densities and space heating require- ments are sufficiently high. These conditions appear to be fulfilled in northern Chinese cities. Industrial Cogeneration. The pulp and paper, steel, and Cogen.ration and district heating require close coordi- chemical industries typically account for a very large share nation between electric power generation, urban planning, (up to 80 percent) of industrial cogeneration. In the pulp and industrial location decisions. Transfer pricing mecha- and paper industry, large amounts of burnable wastes are nisms between electric power utilities, cogenerators, and generated and used to fuel combined heat and power district heating utilities are crucial and should reflect the plants, as are blast furnace gases in the steel industry. In costs and benefits of cogeneration as accurately as possible. other industries, a lack of waste fuels has limited cogenera- Otherwise, potential industrial cogenerators will not find tion. With increased fuel prices, many industries have cogenera:ion a profitable investment. In market econo- found it more profitable to invest in energy-saving mea- mies, large CHP plants are often owned and operated sures, rather than in cogeneration schemes. However, if jointly by, the electric power and district heating utilities to clean fuels are not available (as in China) cogeneration of facilitate coordination and render transfer pricing arrange- electricity and steam for industrial uses might be consider- ments more transparent. 63 washed in 1983) and needs to be expanded. Increased some cities, air pollution has already reached a multiple coal washing can substantially improve end-use efficien- of internationally accepted standards. The largest and cies in combustion and reduce the amount of inert mate- most noticeable source of street-level pollution is the di- rial that must be transported. The optimal role of coal rect burning of coal in household stoves and numerous washing must be carefully evaluated, however, because industrial and commercial boilers without even simple capital costs are high and energy losses are significant, environmental control equipment. In Beijing, for exam- depending on coal characteristics. In the United States, ple, average dust (total suspended particulate) levels were though, some electric utilities have found that the cost of reported at 0.5 milligrams per cubic meter, or about coal preparation and coal cleaning is compensated by seven times greater than U.S. air quality standards. De- higher utilization rates and lower maintenance costs of spite large reserves of low-sulfur coal in North China, electric power plants. Whether this would also apply to sulfur dioxide levels in Beijing are also very high, as coals China will depend among other things on the combus- of unusually high sulfur content are used by households tion characteristics of the particular coal being used and for space-heating and cooking. would need to be studied further. The development of There are several options for making the use of coal in local uses for washery tailings from coal preparation the residential and commercial sectors cleaner and more plants is of particular importance, to minimize energy efficient. Over the long run, these include greater devel- waste and hence reduce the cost of coal beneficiation. In opment of district heating, electric cooking, and possibly countries where large proportions of coal output are coal gasification. Over the short- and medium-term, pol- washed (for example, West Germany), washery develop- lution problems associated with small-scale combustion of ment is closely linked with thermal power development coal in urban areas could be alleviated by greater use of and production of coal briquettes. China's planned ther- higher-quality coals, in particular anthracite, and dissemi- mal power development in coal mining areas such as nation of improved coal stoves. Under the present coal Shanxi should be well integrated with washery develop- allocation system, some of the least desirable coals (for ment-washery throughputs of 300 million to 400 mil- example, briquettes from coal fines with high volatile lion tons in 2000 would produce tailings that could fuel matter and high sulfur content) end up in urban areas. 20-30 gigawatts of power generating capacity. This could be remedied by allocating high-quality coals Further development of fluidized bed combustion (FBC) (anthracite) to urban areas and restricting the use of less is of particular importance if coal mining and coal wash- desirable grades of coal. ery wastes are to be used in mine-mouth power plants. Urban air pollution could also be alleviated by locating While China has a larger number of FBC boilers operating energy-intensive heavy industries outside densely popu- than any other country, most are small industrial FBC lated areas. The construction of large cogeneration plants boilers using low grade coals, and only a few mine-mouth can be justified in urban areas, however, provided ade- power plants with somewhat larger FBC boilers (25 mega- quate environmental protection measures (dust and noise watts, electric) are in operation, using coal mining wastes. control) are adopted. As low-sulfur coal is available in One of the reasons that larger units are not used more substantial quantities, expensive flue-gas desulfurization widely is that transfer prices for electric power are not equipment may continue to be unnecessary, but much sufficiently attractive for sales to the power grid and limit more attention will need to be given to installing im- power generation by coal mines to internal use. Again, proved particulate removal systems. Installation of such institutional changes for electric power generation, and in equipment in new coal-fired plants is now standard prac- particular appropriate transfer pricing arrangements, tice in most countries. Fly ash collected from power would be required to improve the efficiency of coal utili- plants and industrial enterprises can be used profitably as zation. Mine-mouth power plants using coal wastes could an input in cement manufacturing (up to 40 percent of be operated by the the power grids, the coal mines, or the raw material required) or as a substitute binder for jointly. Appropriate economic incentives will be required building materials, as is already done to some extent in to make cogeneration attractive. China. In a number of industrial countries that have experi- Environmental Issues enced environmental damage from acid depositions China fortunately has very substantial resources of rela- ("acid rain") it is now believed that these effects are tively high-quality (low-sulfur) coals,5 and total emissions probably due to a complex interaction of different pollu- of sulfur dioxide and nitrogen oxides are still low, compa- tants, involving not only sulfur and nitrogen oxides but, rable to, for example, the United States around 1920-30 (or one-half and one-quarter, respectively, of present U.S. levels). Nonetheless, air pollution from coal burning, es- 5 Most Chinese coals have a sulfur content of 0.4-2 percent, pecially particulates (dust and soot) is already a serious probably averaging less than I percent, but a few deposits have problem in urban China, particularly in the North. In high sulfur content (3-5 percent). 64 in a key role, ozone and other photochemical oxidants mine the most promising areas for immediate explora- formed by reactive hydrocarbon compounds, emitted tion, establish priorities for future exploration, and for- primarily by petroleum refineries and other chemical mulate policies on the use of foreign technology and plants. Although there are still large scientific uncertain- capital. Moreover, although exploration, development, ties about the mechanisms and effects of acid depositions, and production practices have been improved consider- these deserve attention in the choice of energy utilization ably in recent years, China's techniques and practices will technologies. Large-scale coal gasification is of particular need to be brought closer to international standards. In concern because pollutant emissions from coal gasifica- the recent past, petroleum exploration and development tion plants are similar to those from refineries. Some of have been characterized by too much attention to short- the most serious environmental effects of air pollution in term production targets. Reservoir management has Eastern Europe are believed to be related to large-scale sometimes also been less than optimal, with many fields coal gasification. producing at rates that reduce ultimate total recovery of The experience of other countries indicates that enter- oil frorn the reservoirs. prises are often unwilling to make substantial investments To meet its petroleum production targets, China must in pollution control equipment without direct govern- upgrade equipment, technology, and management at all ment intervention in the form of financial incentives or stages of exploration and development. Current opera- administrative measures, such as a system of enforced tion and maintenance practices, particularly for field penalties for exceeding specific industry-by-industry stan- equipment, need to be improved substantially. Future dards. In China, the development of adequate pollution discoveries of the magnitude envisaged probably can only regulations and enforcement provisions is necessary and be located by drilling in the deeper parts of sedimentary will become even more important as state enterprises gain basins, in more complex geological structures, and in more autonomy and independence. Lack of attention to more remote locations where access is difficult. All this pollution control at the time of installing new plants may will recquire, in addition to imports or joint venture man- ultimately result in higher costs. Investments required to ufacturing of equipment, more intensive personnel train- rectify the situation could be higher than if adequate ing, both on the job through service contracts and abroad measures had been taken at the time of plant construc- in accredited institutions and with operating companies. tion. In view of rapidly increasing energy use and remain- International oil companies (locs) are already involved in ing uncertainties about causes and effects of environmen- China's offshore oil activities and recently the Govern- tal damage, it would also be highly desirable to sharply ment decided to expand Ioc activities to limited onshore increase the extent and quality of environmental moni- areas tfhrough joint venture contracts. Appropriate legisla- toring. This could be instrumental in determining cost- tion needs to be prepared before contracts can be final- effective environmental pollution control measures in the ized. Such involvement would allow the Government to future. share the risks of development and exploration with for- eign investors and might facilitate technology transfer. Oil and Gas The transportation of crude oil, petroleum products, and nal ural gas does not appear to face major bottlenecks Oil Production at present, but the options and costs of future transport China's largest oil fields have already passed their peak systems should be analyzed. Increased use of pipelines productivity. The extent of their natural rate of decline is could help to reduce the burden on the railways and being reduced somewhat by a program of infill drilling, improve refinery utilization. Transport will be an impor- well stimulation, and secondary recovery. China's success tant factor in developing the hydrocarbon reserves that in meeting its goals for petroleum production in 2000, may be discovered in the western basins, as transporta- however, will be largely determined by the extent of new tion costs will be critical in determining the economic discoveries and the speed with which they can be devel- viabilitv of such efforts. oped. By 2000, most of the fields producing today will be twenty-five to forty-five years old, and it is unlikely that Natural Gas they will provide more than a small share of the produc- Natural gas production is targeted to increase to 25 bil- tion target of 200 million tons. A large share of total lion cubic meters by the year 2000-roughly double the production by the year 2000 will have to come from new current level. A significant part of this increase is expected reserve additions during the next eight to ten years. to be achieved by doubling production from the Sichuan China's prospective oil areas onshore and offshore are Basin to 10 billion cubic meters and through develop- extensive, geologically complex, and largely unexplored. ment cf new onshore and offshore finds such as the To achieve the 200 million tons per year production recent Zhongyuan and Hainan Island discoveries. Nonas- target by the year 2000, China will have to develop a sociatecl gas production by 2000 could be as much as four long-term strategy for exploration, which could deter- times greater than the current level of 5 billion cubic 65 meters. If, in addition, the oil production target is tween 30 million and S0 million tons in 2000, even with reached, total gas production could reach 3 5 billion to 40 major improvements in energy efficiency. Transport fuel billion cubic meters, or three times current levels. If ex- requirements thus would account for one-fourth of tar- ploration efforts for nonassociated natural gas are stepped geted oil production, or more if oil production falls short up, production could be even higher, and some estimates of the target. Policies to restrain the growth of road indicate that it could reach as much as 5 percent of total transport could of course reduce growth in distillate de- commercial energy requirements. There is also an urgent mand, but at high costs in terms of the efficiency and need to increase gas exploration in older producing areas dynamism of economic development, especially outside where production declines are leading to shortages of gas large cities (see Chapter 5). for existing plants. While the projected high share of transport fuels in In many feedstock and fuel applications, natural gas has total petroleum product demand is not unusual by inter- a relatively high value per unit of energy compared with national standards, the bulk of Chinese crude oil is rela- other energy sources. Evaluating its optimal use, how- tively heavy with, on average, more than 70 percent ever, is highly complex and needs to be pursued on a site- residuals in primary distillation. Chinese crude oils are by-site basis, taking into account regional variations in particularly deficient in very light fractions (naphtha and demand, in the opportunity cost of alternative fuels or gasoline). Consequently, about 35 million tons per year feedstocks, and in the costs of gas transmission and distri- of secondary conversion capacity has been installed, pri- bution. Given the characteristics of natural gas, it is nec- marily catalytic cracking (24 million tons per year) and essary to make a comprehensive plan that includes re- thermal cracking. Refinery complexity in China (second- quirements for exploration and development as well as ary conversion equivalent to 35 percent of primary distil- transport and use. To accommodate uncertainties about lation capacity) is thus already high in comparison to gas reserves, production rates, and the size of the market, most other countries, except the United States (48 per- this plan should be flexible, with staged construction to cent). Refinery throughput is about 80 million tons per permit continuous revisions as new information becomes year (80 percent capacity utilization), and refinery losses available. The value and use of natural gas is likely to be have been reduced from 8 percent in 1980 to about 6 very location specific. Given the projected small share of percent in 1984. Exports of crude oil have increased to natural gas in commercial energy production and use, over 20 million tons in 1984, and product exports have natural gas is likely to be used primarily for high-value averaged some 5 million tons in recent years. industrial uses such as feedstocks. Nonetheless, some uses The major issue for China's refinery industry in the of natural gas which may be regarded as "low value"- future is how to match domestic demand, geared toward electric power generation, for example-should not be transport fuels and other distillate products, with China's excluded as they could be instrumental as initial uses to domestic resources of heavy, high-wax (but low-sulfur) develop gas production and pipeline networks. House- crude oils. While some degree of flexibility regarding hold use for cooking and heating could be rational in domestic burning of residual fuel oil appears justified, some areas. given the high costs of converting plants originally de- signed to burn oil to use coal or the high costs of coal Petroleum Refining handling equipment for small boilers, there is little doubt The demand for oil distillates6 is projected to grow from that it will be more economical in most cases to increase 34 million tons in 1980 to between 90 million and 140 coal supplies to meet boiler fuel demand and to either million tons in 2000. The attainment of both oil produc- export or further refine much of China's residual fuel oil. tion and economic growth targets might result in a bal- While it is technically feasible to convert Chinese crude ance between oil distillate supply and demand around the oils to transport fuels and other distillate products by midpoint of the projected demand range with approxi- secondary conversion, yielding virtually no residual fuel mately the same refinery complexity as today (Table 4. 1). oil, these processes are inherently very capital intensive If crude oil production levels were to reach only 150 and expensive. Considerable savings from the present million tons in 2000, either major imports would be strategy of self-sufficiency in petroleum products for do- required to satisfy domestic demand for distillates or re- mestic consumption could be achieved by greater integra- finery distillate yields would have to be increased to their tion into the international oil market. Other countries in technical maximum by installing more sophisticated (and the region and elsewhere have a different demand struc- expensive) secondary conversion facilities. ture, and other crude oils are available with very different If road freight and passenger transportation both grow at close to 10 percent per year during 1980-2000 (imply- ing significant increases in the share of road transporta- 6. Defined to incLude gasoLine, diesel, kerosene, lubricants, light tion in total transport), the demand for transport fuels petrochemical feedstocks, and liquefied petroleum gas (I.Pc). See would grow from about 10 million tons in 1980 to be- Annex 3 66 characteristics. Refineries elsewhere are meeting product 84, ;nd the power sector has been unable to meet de- demand not only by installing conversion facilities, but mand fully. also by optimizing the mix of crude oil feedstock and by trading some refined products in international markets. Def2nd Projections Some of the options that should be considered before Electricity demand growth could average 6.0- 7.5 percent investments in further secondary conversion facilities are per rear during 1980-2000, with the projected range finalized include: (a) importing foreign crude oils in ex- depending to a large extent on the rate of growth of change for increased exports of Chinese crude oils to manlfacturing industry (Table 4.1). Generation require- increase the yield and quality of products (for example, menis in 2000 would be 960-1,290 terawatt-hours, transport fuels or asphalt) in which Chinese crude oils are which is quite consistent with MWREP's preliminary esti- deficient; (b) importing (or exporting) deficient (or sur- mates of 1,000-1,200 terawatt-hours. The share of elec- plus) products to meet domestic demand; and (c) con- tricity in final commercial energy consumption is pro- tracting for the processing of a marginal share of China's jectedl to continue to rise from 18 percent in 1980 to crude oil in international refining centers with excess ca- 26-27 percent in 2000 (excluding cogeneration in indus- pacity. try), but the elasticity of growth in electricity demand Institutional changes would be necessary for more ra- relative to growth in GDP is expected to decline to 0.9- tional investment and production decisionmaking in the 1.15. Electricity use in transportation and households is refining industry. While distillate prices are more or less expected to grow particularly fast (14-15 percent per in line with international prices, crude oil prices are low, year and 11-1 3 percent per year, respectively, from 1980 making oil refining a very profitable industry. This has led to 2C00); though this is from a low base in both sectors, to duplication of investments in the past, as different it may result in increased peak load generating require- localities and organizations tried to maximize their reve- ment,. The manufacturing sector is projected to continue nue from oil refining. The consolidation of virtually all to account for at least three-quarters of final power con- refining and basic petrochemical production in one state sumption, although electricity use per unit of gross manu- corporation (SINOPEC) in 1983 has helped to rationalize facturing output value is likely to fall slightly, because of the refining and petrochemical sector and has made trans- changes in the structure of manufacturing output. fer pricing of petroleum products and petrochemical feed- stocks less cumbersome and controversial. However, Investment Planning some problems remain. For example, the present prefer- No official long-term development program for electric- ence for product rather than crude oil exports is appar- ity development for the remainder of the century has yet ently mainly related to the allocation mechanism for for- been prepared. The present mix of generating capacity, eign exchange. Less than optimal export patterns (and however, is not expected to change much over the next resulting revenue losses) could be avoided if domestic two decades. By the end of the century, nuclear power prices for crude oil and products were set at international would supply less than 4 percent of total generation, prices and if foreign exchange allocation mechanisms while hydropower would stabilize its share of supply at were improved. Decentralization of the petroleum indus- 18-19 percent (or 22-23 percent if small plants are in- try would become possible only after price reform. cluded). Thermal power will account for the balance, and under current plans all new capacity will be based on Electric Power coal. The bulk of additional hydropower is expected to come from four large-scale river basin development Electric power development in China has proceeded rap- schemes, which have already been the subject of substan- idly during the past three decades. By the end of 1983, tial study. total installed capacity reached 76,000 megawatts, com- China's exploitable hydropower resources are among pared with less than 2000 megawatts in 1949. Approxi- the largest in the world, but the bulk of undeveloped mately one-third of generating capacity in 1983 was hy- poteniial is in the Southwest and Northwest, where dro, while the remaining two-thirds was thermal. No large-scale development would require transmission of nuclear power plants have been commissioned. There electricity over distances of 1,200-1,500 kilometers to were thirteen power grids with capacities of more than major industrial load centers. Gestation periods for the 1,000 megawatts each; four of these had capacities ex- large-scale projects are relatively long (eight to ten years ceeding 10,000 megawatts. The combined installed ca- from project approval even under ideal conditions) and pacity of the thirteen largest grids accounted for more have been further extended in some recent projects be- than 80 percent of the national total. Between 1965 and cause of unexpected geotechnical problems, shortages of 1979, total generation increased at an average rate of funds, and other unanticipated problems. Higher infra- 10.7 percent per year, but has slowed down in recent structure costs and rising costs due to inundation are also years, increasing by 5.8 percent per year during 1979- increasing the costs of large-scale hydro projects, and 67 careful economic analyses will need to be carried out in and power transmission, environmental implications, and selecting sites and timing of large-scale hydro projects, other external factors such as increased use of cogenera- taking into account economic effects on agricultural pro- tion. duction, inland water transport, and other users of water The demand for power currently exceeds supply in all resources. of China's major grids, and the cost of interconnecting MWREP's preliminary power production profiles for most regional grids, which requires transmission lines of 1985-2000 suggest a commissioning of 5,000-8,000 1,000 kilometers or longer, may not be justified by re- megawatts of nuclear capacity by the end of the century. ductions in reserve requirements and load diversification Because of relatively low-cost coal and competing de- alone. However, with large-scale hydropower develop- mands for financial resources, it is unlikely that a nuclear ment and construction of some mine-mouth power power program beyond MWREP's targets could be justi- plants, greater interconnection will become more com- fied, at least in the medium term. The high capital costs, pelling and eventually imperative, because large amounts long lead times, and uneven performance of nuclear of power will have to be transmitted from West to East. plants in some countries underscore the importance of In addition, integration will allow wider use of modern, evaluating the parameters that determine the economics large-scale generating equipment with improved technical of nuclear generation in relation to the costs of other and economic results. Currently, China has adopted a alternatives (particularly coal, but possibly also natural step-by-step approach to grid expansion and integration, gas). Commissioning of a few nuclear units in China with plans to establish a national integrated power grid could be justified in certain circumstances, for example, by the end of the century. in locations far from low-cost coal mines and on the basis Increased efforts to optimize investment planning by of the continuing need to develop technical expertise and focusing on long-term, least-cost development programs operating skills. The competitiveness of large-scale nu- will be essential. Of particular importance is the need to clear power development could probably be enhanced by consider the full economic costs of large-scale projects the experience gained from a modest-size program and by with long lead times (for example, nuclear or large hydro careful planning-foreign experience shows that a well- projects). Such projects clearly can provide major bene- designed and well-paced program could reduce capital fits, but the opportunity costs of money invested and the costs. costs of sacrificing flexibility need to be more carefully Nuclear power requires a major long-term commit- evaluated. The adoption of up-to-date system planning ment to developing manpower and infrastructure. It in- techniques also will help in the evaluation of broad strate- volves technically complex plants, severe economic conse- gic decisions about the generating plant mix, power plant quences in case of operational failures, and strict safety location, and grid architecture. Modern computer requirements, all of which are unique to this type of models allow planners to look beyond the unit cost of power. Major investments of effort, time, and resources energy in determining least-cost sequences; they take into are needed to develop the technology and standards for account such factors as demand patterns, variation in equipment manufacturing in the early stages of the pro- hydrology, and random outages of generation and trans- gram, as well as autonomous safety and regulatory insti- mission facilities. These models are not without their tutions to provide essential oversight functions. Particu- limitations, however, and much training is required to larly important over the long term is that plants be based use them appropriately. on a standard design that has been carefully selected and developed. Rural Energy and Biomass Under current plans, conventional thermal power gen- eration based on coal will continue to provide at least According to Chinese estimates, the consumption of bio- three-quarters of total generation at the end of the cen- mass fuels (that is, fuelwood, crop byproducts, and dung) tury. Probably the most important issues are those of by rural households was about 220 million tons of coal plant location and transmission network planning. For equivalent in 1980, thus accounting for more than 25 example, the location of many plants will have to be percent of total primary energy use. Crop byproducts determined by the relative advantage of development at and fuelwood each account for roughly one-half of con- coal mine sites-requiring high-voltage transmission to sumption, while dung contributes an almost negligible major load centers-versus development near load cen- share. Biomass fuels provide some 85 percent of the total ters and port areas where large quantities of coal would energy consumed by rural households. In many areas, be transported by rail or water. Decisionmaking should biomass fuel supplies fall short of demand, but supply include consideration of factors such as the regional distri- even at current levels no longer can be sustained by the bution of coal resources, availability of low-grade fuels agricultural system and local environment through tradi- from coal preparation plants, regional differences in coal tional means without serious adverse consequences. development costs, relative costs of coal transportation While the production of crop byproducts has probably 68 doubled since the first half of this century, the proportion Given the importance of local conditions, central poli- used as fuel has increased from about 50 percent during cies should be highly flexible and should emphasize con- the early 1930s to 60-80 percent today, inhibiting their tinued strengthening of local capabilities for both techni- use as animal fodder, organic input for soil improvement, cal development and economic evaluation of alternatives. and construction material. Increased demand on local All options should be considered, including increased reli- nonagricultural land to produce other types of biomass ance on supplying energy to households from outside a fuel has exacerbated long-standing water and soil conser- given locality. vation problems, as traditional fuel collection methods have often gone unchecked and immediate fuel needs Investment, Planning, and Prices have impeded efforts to reforest local areas. Hence, China faces major challenges in developing alternatives to Investment Requirements traditional supply and use practices. Continued reliance The energy sector is one of the most capital-intensive on traditional patterns of biomass supply and use to meet sectois in any country, and efficient use of existing and rural household fuel needs will result not only in greater new plants is of utmost importance for the efficiency of disparity between supply and demand, but also in greater the sector and the economy at large. The average annual disruption of local agricultural and ecological systems. investment requirements for developing the coal, petro- Current rural energy policies emphasize the develop- leum, and electric power industries are estimated on the ment of local energy resources for local needs and im- basis of current government goals to increase from about provements in the efficiency of energy use. In 1980, an Y 11 billion in the Sixth Five-Year Plan (1981-85) to Y initiative was launched to promote the production of 30 bi'lion to 40 billion during 1986-2000. If petroleum fuelwood in private woodlots in areas that have suitable production targets are to be achieved, a significant in- uncultivated land. By 1984, small plots had been allo- crease in investment for exploration and production de- cated to almost one-half of China's rural households for veloprnent will be required. Investments in the power this purpose. The development of local coal resources sectot will need to be sharply increased if electric generat- continues, for use by local households and industries and ing capacity is to be installed to meet forecast demand. for sale to consumers in other areas. Solar cookers have While the power sector accounted for less than 40 per- been popularized in a few areas and, where conditions for cent of original energy sector investment allocations for small-scale hydropower development are particularly fa- the Sixth Five-Year Plan, power will need to account for vorable, the use of electricity for cooking is beginning to two-thirds of energy investments according to Western be promoted. Despite past setbacks in the biogas pro- definitions, or 60 percent according to Chinese defini- gram,7 the promotion of family-size digestors continues, tions of capital construction (Table 4.3). Actual spending with greater emphasis on quality construction. In some for el ctric power during the Sixth Plan has been consid- areas, larger, community-size digestors are now being erably higher than originally planned. promoted to supply gas for household use. A major pro- gram to develop and disseminate more efficient rural PETIROLEUM. On the basis of knowledge of some Chi- cooking stoves was started recently, and by early 1984 nese Fields and using some international yardsticks, the had put into use more than 40 million improved stoves. minirmum investment required to discover and develop The potential and the economic viability of the differ- the additional reserves required to meet China's produc- ent options vary dramatically from area to area, depend- tion targets can be estimated at Y 30 billion and Y 70 ing on local conditions. Moreover, even the implementa- billion for exploration and development, respectively, tion of seemingly simple solutions can be quite complex. over the next twelve to fifteen years. Considering the In the case of biogas generation, for example, the eco- risks mnd uncertainties associated with petroleum invest- nomic viability of the higher-quality but more expensive ments, this estimate simply provides a benchmark. Major digestors currently being promoted hinges on a wide adjustments in investments targeted for the sector may be range of local direct and indirect factors (for example, required as risk assessments and estimates of development local slaughtering rates, winter temperatures, sanitation needs change. Petroleum sector investment targeted un- problems). Experience in other developing countries sug- der the Sixth Five-Year Plan amounted to Y 15.5 billion gests that, while the benefit of energy conservation de- (an atinual average of Y 3.1 billion), 70 percent of which rived from improved stoves appears attractive relative to was for petroleum development and 30 percent for explo- the costs involved, dissemination can be difficult. For ration. Thus a future average investment figure of Y 6 example, acceptance can be hindered because the new stoves may compromise convenience in cooking or may prove inappropriate for preparing certain traditional 7. Of the 5 million family-size biogas digestors constructed in foods. In China, household heating needs, which are of- Sichuan during the 1 970s, only about one-third are reported to be ten met by traditional stoves, must also be considered. in operation. 69 Table 4.3 Estimated Investment Requirements been about I percent of GDP or less, mainly because of in the Energy Sector, 1986-2000 lower growth rates of electric power demand.) Amount Percentage of (billions energy sector COAL. The average yearly financial requirements for Component ofyuan) investment capital expenditures to produce 1,400 million tons by Coal mining' 50-70 12-13 2000 (also in 1983 prices) are expected to rise to about Y Petroleum 80-120 20-21 4 billion to 5 billion for mine development plus Y 2 Exploration 20-40 5-7 billion to 3 billion for social infrastructure. This may be Development 60-80 14-15 compared with, for example, the 1983 state budget for Electric power 2 80- 370 66-68* Eletric power 280-3270 68-69 capital expenditures in coal for MOCI mines (Y 3.6 billion Generation 200-2 70 48-49 Other 80-100 18-20 including social infrastructure, or about Y 2.2 billion ex- Total 410-560 100 cluding social infrastructure). The Government's pro- gram for coal price increases should provide more incen- Energy-related investments tives and financial resources to increase coal production in industry and transport Coal transport 40-50 - particularly in mines operated by local governments and Refineries 20-40 - collectives. Resources made available to MOCI are being increased under the Seventh Five-Year Plan (1986-90) to a. Assuming coal production of 1.2 billion to 1.4 billion tons bv 2000. more than Y 6 billion a year social infrastruc- Includes only direct production-related investment. Associated social infra- I (icludig structure (housing, schools, etc.), which is included in Chinese estimates, ture). To supplement domestic funds, efforts are also be- would add at least Y 20 billion to 30 billion. ing made to mobilize foreign investment resources in the Source: Data provided to the economic mission, form of foreign loans as well as direct foreign investment. Planning and Coordination billion to 8 billion a year represents about two or three To achieve optimal energy production and efficient use of times the recent level of expenditure. The foreign ex- energy, decentralization of decisionmaking and greater change requirements for substantial imports of sophisti- use of prices and other economic levers will be essential. cated tools, material, and equipment, along with the pro- At the same time, however, long-term planning and inter- posed involvement of foreign contractors, would likely sectoral coordination will need to be improved. In partic- amount to US$ 15 billion to 20 billion over the next ular, economic evaluation of specific development alter- twelve to fifteen years. natives to reflect all relevant costs and benefits could help to avoid serious economic distortions. For example, eco- ELECTRIC POWER. Preliminary estimates of the invest- nomic evaluation of a coal gasification scheme for urban ment required for capacity and grid expansion to gener- cooking should take into consideration all economic ate 1,000-1,200 terawatt-hours in the year 2000 have costs, including gasification, gas distribution and con- been made by MWREP. Given the increasing unit costs sumer hook-ups, and the relative costs and benefits of expected in the sector, a dramatic increase in average other alternatives, for which different institutions may be annual investments will be required if capacity is to be responsible. expanded not only to relieve current supply constraints, Beyond specific project analysis, long-term energy' plan- but also to provide the additional power required to meet ning should consider explicitly the impact of broad as- the needs for rapid economic growth. Under the Sixth pects of economic development strategy on energy de- Five-Year Plan, the average annual investment for the mand, as well as long-term programs for energy power sector has so far been about Y 4.2 billion. Average production and distribution and improvement in energy annual investment requirements are estimated to grow to efficiencv. Potential tradeoffs between the achievement some Y 13 billion to 15 billion during 1986-90 and Y 22 of short-term gains and attainment of long-term goals billion to 29 billion during 1991-2000. This implies a may exist, for example, in cases where emphasis on in- major increase in the share of the power sector in total creasing short-term energy production compromises ul- domestic investment compared with recent years. timate reserve recovery or overlooks preparation for The share of power sector investment has been lower long-term needs. Tradeoffs may also arise where the de- in China than in many other developing countries. The ployment of outdated technology in new industrial ca- increased share estimated for the future (about 2 percent pacity provides an easier and faster way to meet pressing of CI)P) is not unusual by international standards. The current demands, but ultimately may compromise goals World Bank estimates that power sector investment re- to improve energy efficienc'. quirements in all developing countries together will repre- Because effective long-term planning for energy pro- sent about 2 percent of GDP from 1982 to 1992. (Electric duction, distribution, and utilization involves so many power investments of industrial countries have generally interrelated and intersectoral issues, an integrated ap- 70 proach to energy planning is essential. The development Demand Management and Pricing of markets in various sectors and regions must be consid- Energy in China has been-and still is, despite recent ered along with the development of energy production, reforms-largely allocated administratively, with market transportation, and distribution. Major imbalances may forces playing only a minor role. Financial constraints be avoided through review and coordination of long-term have not been a critical consideration for most enter- plans for specific projects in the various economic sectors. prises, whose overriding objective has been to meet phys- Better energy planning also requires an improvement in ical output quotas. Energy prices in China likewise have the quality and quantity of statistics relating to energy been primarily determined by administrative decisions, production and consumption. some made decades ago. As a result, most energy prices Intersectoral coordination has been difficult in the past, today reflect neither scarcity nor production costs-pe- for example, in relation to the location of large coal mines troleum refining is very profitable, for example, while and power plants and to the distribution of natural gas official prices for coal barely cover operating costs. Some and the development of its market. To meet even greater improvements have recently been made by permitting needs for such coordination in the future, institutional market- -elated prices for above-quota supplies. This two- mechanisms at both central and regional levels must be tier pricing system has spurred production (for example, strengthened. Prices that fully reflect production costs from strlall coal mines under local control) and may also and scarcity are the most important tool for ensuring have he]ped to reduce energy consumption. intersectoral coordination, but market research by pro- Overall, administered energy prices are low in relation ducers and interagency coordination at all levels are also to economic costs, with the exception of some petroleum important. product, and electricity for residential consumers. Com- pared with the structure of international prices, the prices Technology Transfer and Development of coal, crude oil, and fuel oil are particularly low. Aver- The rapid transfer and development of energy-efficient age electricity tariffs are much lower for nonresidential technologies will also be crucial because the power plants than for residential consumers. Centrally allocated coal is and energy-using industrial equipment that are now being priced at about 60 percent of long-run marginal costs, built will last for several decades and will determine the and heavy fuel oil at only one-third of international efficiency and environmental effects of energy use into prices. There is little information on the level and struc- the next century. With rapid economic growth, the de- ture of natural gas prices, although they appear to be low gree of energy efficiency of new plants will be more in relation to the economic costs of competitive fuels. important for energy conservation over the long term Although recent measures (mentioned earlier) have than energy savings in existing plants. Moreover, the been effective in promoting energy savings, a quota-based most energy-efficient technologies available abroad are energy allocation system is fundamentally unsuited to not necessarily more costly, particularly if attributes such achieving large and economically rational reductions in as better reliability, product quality, and environmental energy intensity over the longer term. Enterprises are controls are considered. discouraged from reducing energy consumption because Technology transfer is critical for thermal power plants, quotas are related to past consumption levels, and unnec- combined heat and power plants and district heating, essarily large quotas are sought for security or stockpiling, small- and medium-scale industrial boilers, heat recovery or for sale or exchange for other commodities. The in- equipment, industrial process controls, transport equip- flexibility of the administrative allocation system also ment, and a long list of other industrial equipment. At makes enterprises reluctant to introduce new products least as important is the transfer of disembodied technol- and prccesses, including energy-saving measures, where ogy, for example, modern management and organiza- these would involve changes in the required level or pat- tional techniques (see also Chapter 7). Importation of tern of energy supply. Enforced reductions in energy quo- disembodied technology through manufacturing licensing tas, for lack of information, have to be imposed too agreements, service contracts, joint ventures, and other uniformly on all enterprises, with insufficient regard to procedures is one of the most economical means of trans- the widely varying economic costs of reducing energy ferring technology, as is job training in conjunction with consumption. technology service imports. An important example of the The two-tier pricing system now in operation for some need to import disembodied technology through service forms of energy, and under consideration for others, is a contracts in the energy sector is the transfer of petroleum considerable improvement on the pure quota system, exploration, development, and reservoir management since the market-determined prices of above-quota sup- techniques that could considerably enhance the efficiency plies give more appropriate signals to producers and users of China's petroleum sector (Annex 3, Chapter 5)-a of energy. But for the two-tier system to operate effec- critical element if petroleum production targets are to be tively, quotas need to be set and regularly revised so that reached. all prod icers and users are obliged to make significant use 71 of the above-quota market. This is difficult, especially ally tradable fuels, such as petroleum, prices should be set because production potential and energy consumption at world prices plus or minus transport margins, unless requirements change constantly and often cannot be di- there are limits to imports or exports (as, possibly, in the rectly and reliably ascertained. Indeed, the two-tier pric- case of coal), in which case domestic prices should float to ing system itself increases the incentives for energy pro- clear the market, especially if there are many producers ducers and users to supply misleading information to the and markets that could be competitive. For electric authorities responsible for setting quotas. power, tariffs should on average at least cover long-run There would thus be clear advantages to phasing out marginal costs that fully reflect investment costs and the the whole quota system gradually, with more autonomy opportunity cost of capital. Lower tariffs may be set at for individual enterprises, and the establishment of a uni- off-peak periods, and higher tariffs are needed to bring fied energy price structure that reflects the relative scar- demand efficiently into line with supply in periods when city of competing fuels (see Chapter 10). Two key advan- generating capacity is insufficient. An appropriate electric tages of higher energy prices as compared with power tariff would make physical rationing of electric administrative regulation are: (a) that they are passed on power unnecessary, as users would restrict their demand in higher product prices and hence reduce consumption if the costs of electricity consumption exceeded the bene- of energy-intensive products; and (b) that reductions in fits, particularly in peak-load periods. An appropriate energy use occur automatically where the economic cost electric power tariff would also promote development of reductions is least, provided that other prices are ra- of efficient cogeneration schemes and mine-mouth power tional and enterprises are profit sensitive. For internation- generation. 72 Spatial Issues Spatial aspects of development will play a critical role in Table 5. 1 Dispersion of Production Activities how efficiently and equitably China makes the transition among .nd within Provinces, 1982 from a low- to a middle-income country. First, China's Number Number ofprefectures or municipalities sheer size and past emphasis on local self-sufficiency offer of provinces involved in activity opportunities for large gains in national economic effi- involved in * . ,. . , I Pronduaon activity Jiangsu Hubei Gamu clency through increased specialization and trade among activity (out of 29) (out of 14) (out of 14) (out of 1) regions and between urban and rural areas. Second, re- maining poverty and inequities have a major spatial di- Foodgrains 29 14 14 1 3 mension-personal incomes are much higher in urban Coal 27 9 7 .. than in rural areas, and poverty is concentrated in specific Cement 29 14 14 13 rural localities. Pig iron 2 7 9 3 2 To increase internal specialization and trade and to fur- Steel 28 13 8 4 ther reduce rural poverty will require a lot of invest- Steel products 28 14 7 Fertilizers 2 8 14 12 6 ment-and some basic changes of orientation-in trans- Machine tools 28 13 9 2 port and commerce. It will also require changes in the Cloth 28 14 14 4 system of economic management and in the roles of dif- Bicycles 26 12 6 1 ferent levels of government. For example, to ensure that Sewing machines 24 8 0 the spatial decisions of households, farmers, and (increas- Watches 24 1 1 .. 0 ingly) autonomous enterprises are efficient, special atten- Source, State Statistical Bureau, Statistical Yearbook of Cbina. 1983 (Hong tion will need to be given to economic levers such as Kong: Economic Information and Agency, 1983); data provided to the economic mission. transport tariffs and land use charges, as well as to remov- ing barriers to competition and trade among localities. Direct government intervention in some spatial deci- sions-and of course in infrastructural development-will Specialization and Trade continue to be necessary. But the appropriate degree and China's past emphasis on local self-sufficiency rather than form of direct intervention, and the level (or levels) of specialization has resulted in all twenty-nine provinces, government that should intervene, will need to be reas- and huidreds of prefectures, being involved in a wide sessed. Thisess apter-whichednavoidably.deals more with ge- range cf production activities (see Table S.1). In fact, T is chpe-hc unaoidbl almost all provinces and many areas within provinces neral principles than with geographical specifics-looks prost all basic s and ma thin suces first at policies and instruments to guide specialization produce not just basic foodstuffs and materials such as and trade in agriculture and industry. It then considers cement--whose production is dispersed in most coun- associated changes in the volume, pattern, and efficiency tries-but also iron and steel products and consumer do- of transport, and in commerce and related services. Next, it discusses some implications of structural change and 1. This chapter draws on the fuller discussion of transport in specialization for the pace and pattern of urban and rural Annex 6 and on Background Papers 1-3, which cover rural nonag- development. It concludes with a review of measures to ricultural development, urbanization, and international economic reduce rural poverty and regional inequality. strategy. 73 rables, whose production in other countries tends to be provinces (see Map 5.1 and Figure 5.1). Both per capita much more concentrated because of large economies of agricultural and industrial output tend to be much higher scale. In many areas of China, production is small scale, on the coast than in the interior.2 The difference, for high cost, and frequently of such low quality that com- instance, between Jiangsu, a prosperous coastal province, mercial bureaus can sell the output only by restricting and Gansu, a poor province in the interior, is substantial: imports from other localities. Partly as a result, there is per capita gross agricultural output is Y 390 in Jiangsu wide variation among provinces in the profitability of and Y 160 in Gansu; per capita gross industrial output is enterprises in particular industrial subsectors (Table 5.2). Y 820 inJiangsu and Y 410 in Gansu; and per capita net The diversity of production in most localities in China does not imply an even distribution of output. In fact, industrial output is quite concentrated, especially in 2. The interior is defined to include the provinces of Shaanxi, Shanghai (Table 5.3). And there are large differences in Gansu, Qinghai, Ningxia, Xinjiang, Sichuan, Guizhou, Yunnan, per capita output between, and also within, regions and and Xizang. Table 5.2 Gross Output and Profits of State-Owned Industry, by Province and Subsector, 1982 (percent) Gross output Profit rate' National National Subsector Jiangsu Hubei Gansu average Jiangsu Hubei Gansu average Metallurgy 3 17 16 11 22 17 10 23 Power 6 5 9 5 38 19 24 23 Coal 2 . 3 4 -4 -10 1 3 Petroleum 4 6 17 7 137 36 51 74 Chemicals 15 10 13 12 37 20 18 38 Machine building 18 22 20 19 34 14 14 21 Building materials 3 3 3 3 33 20 17 21 Forest industry 1 1 2 48 36 15 22 Foodstuffs 1 s1 . 1 3 88 101 . 93 Textiles 28 19 5 16 100 58 54 81 Paper making 1 I 1 50 17 7 33 Other 5 5 13 7 49 29 25 37 Total 100 100 100 100 45 23 19 32 a. Defined as profits plus taxes divided by net value of fixed assets. Source Data provided to the economic mission. Table 5.3 Industrial Concentration, by Region and Key City, 1982 (percent) Gross industrial Population Ratio of output output shares shares to population Region Total Heavy Ligbt Total Urban Total Urban and city (1) (2) (3) (4) (5) (1):(4) (1)(5) Three municipalitiesa 19.3 17.8 20.8 2.8 8.8 6.9' 2.2c Beijing1 3.9 43 3.4 0.5 2.7 7.8 1.4 Tianjin' 3.7 3.2 4.2 0.5 2.5 7.4 1.5 Shanghai' 9.0 8.0 10.0 0.6 3.0 15.0 3.0 Other provinces 80.7 82.2 79.2 97.2 91.2 0.8 0.9 Otherkeycitiesc 15.9 17.0 14.7 3.1 15.1 5.1 1.1 Total, all provinces 100.0 100.0 100.0 100.0 100.0 1.0 1.0 Total, all key cities 32.5 32.5 32.3 4.7 23.3 6.9 1.4 a. Includes some rural counties surrounding the cities of Beijing, Tianjin, and Shanghai. b. Theequivalent ratios for some majorcities in othercountries are Chicago 2.6 in 1914 and 1.4 in 1977; Bangkok 3.7 in 1978 c. The equivalent ratios for some major cities in other countries are: Chicago 1.2 in 1914 and 1.1 in 1977; Bangkok 0.6 in 1978. d. Excluding the rural counties. e. Includes Shenyang, Dalian, Changchun, and Harbin in the Northeast; Taiyuan in the North; Jinan, Qingdao, and Nanjing in the East; Wuhan and Guangzhou in the Central South; Xian and Lanzhou in the Northwest; and Chengdu and Chongqing in the Southwest. Source State Statistical Bureau, Statistical Yearbook of China, 1983. 74 Map 5.1 Per Capita Net Material Product, by Province, 1982 U. S. S. R. MON GO L IA JAPAN ; - s 2- r . . . s \ u i i Sea otjaps ""t - ; itC -s ;- ' sv R g __ EOFKOPL RE ,P/L \ Despite fiscal and other redistributive measures, these 030 % \ 41% inequalities in per capita output result in substantial, Regional though smaller, interregional variations in personal in- Y 900 Y 516 comes. The gap in personal incomes between Jiangsu and Coast Interior Gansu, for example, is 1.4 to 1; between Wuxi County, Jiangsu, and Dingxi County, Gansu, it is 2.8 to 1; and between Qianzhou Township, Wuxi, and Dongye Town- 32% 28% ship, Dingxi, it is over 10 to 1.4 The challenge will thus Provincial be to increase efficiency through greater internal speciali- y 571 zation and trade, while simultaneously reducing the pov- Y 1,211 / erty associated with these substantial income inequalities. Gansu Province ineq JIaguPoic Agnicultural Specialization + l ~~~~~~~~~~In recent years, rapidly increasing agricultural specializa- tion in China has contributed to both growth and pov- 41% _j erty reduction. The specialization has mainly involved Prefectural Y 199 53% changes in the location of nongrain agriculture (with con- tinuing, though diminishing, emphasis on local self-suffi- Y 1,548\ Dingxi Prefecture, ciency in foodgrains). In the case of industrial crops, pro- Gansu vincial specialization according to comparative advantage Suzhou Prefecture, Jiangsu j may now be more characteristic of China than of India.5 Further specialization both within and between provinces I | will, however, be critical for future agricultural growth. I l An efficient spatial pattern of agricultural development 44% i will require increasing concentration of production in County Y 169 7 3% areas where agroclimatic conditions are most conducive to high yields. But land and transport costs will also play Y 2,00 Dingxi County, an important role. For example, areas close to cities and Dingxi towns will usually have higher economic land values than Wuxi County, Suzhou areas farther away, because there are much greater possi- I 2/5% 3. The gap in per capita NMP between the richest province (Shanghai) and the poorest province (Guizhou) is 11. I to l; if the I three municipalties are excluded the gap falls to 3.2 to I between Township _00% Liaoning and Guizhou. These gaps compare with a 3.3 to I gap in Y w TU.S. state per capita domestic product in 1950 (between Delaware Dongye'rownship, and Mississippi); a 1.8 to I gap in U.S.S.R. national per capita Y 2,874 Dingxi income in 1970 (between Lithuania and Turkmenia); and a 2.5 to I gap in Indian state per capita domestic product in 1975-76 Qianzhou Township, Wuxi (between the Punjab and Bihar). 4. The gap in personal per capita income between Shanghai and Gansu (the richest and poorest provinces by this measure) is 2.4 to 1. This compares with 3.1 to I in the United States in 1950 * Share of agriculture in total output between Delaware and Mississippi; 1.9 to I in the U.S.S.R. in E Share of industry in total output 1965 between Estonia and Azerbaidjan; and 1.7 to I in India in 1967-71 between the Punjab and Orissa. 5. See Annex 2, Appendix B, for an interprovincial compari- Source. Data provided to the economic mission. son of cropping patterns in China and India. 76 Box 5.1 The Spatial Distribution of Dairy Production Akhough dairy production in China is still quite limited, optimal location for a dairy will thus depend crucially on the sector is characterized by a wide range of systems, whether it is more efficient to transport feed a long dis- from individual farmers grazing one or two dairy cows on tance and dairy products a short distance, or vice versa. low-quality natural pastures to large suburban dairies that International experience suggests that it may be more eco- rely on feed concentrates and are often closely integrated nomical to locate the dairy where the roughage is pro- with other agricultural activities. During the next two de- duced, bcause the quantity of feed is much greater than cades, demand for dairy products in China is expected to the quantity of dairy products. In fact, three or four tons increase rapidly. Some of this demand can and undoubt- of feed (roughage and concentrate) are generally required edly should be met through increased imports (in view of for every ton of milk (or about 100 kilograms of cheese, the very low prices of powdered milk products on the 131 kiloErams of whole milk powder, and 435 kilograms international market), but the Government also wants to of condensed milk). Unit transport and handling costs develop a more efficient domestic dairy industry. may, of course, be higher for milk and dairy products than International experience suggests that there are three for feed, but the difference is unlikely to affect the choice basic systems for dairy production: (a) the system common of location. in most developing countries, where dairy cattle obtain A more major issue may be the transport and marketing most of their feed nutrients from low-quality roughages, of dairy by-products, including manure, which has some- such as crop residues or natural pastures; (b) the system times been used as a justification for locating dairies in used in Australia and New Zealand, based mainly on im- suburban areas and integrating them with fish farming or proved high-quality pastures; and (c) the system used in the fruit growving. However, there are other sources of ma- United States, Western Europe, and Japan, which is based nure, and the costs and benefits of such integrated opera- on relatively high levels of concentrate feeding. The much tions need to be closely compared with alternatives such as higher unit production costs for system (c) over system (b) integrated pig or poultry raising and fish farming. have resulted in farmgate prices for milk ranging from Clearly there is no simple answer to the issue of the USO.12 per liter in Australia and New Zealand to optimal ;patial pattern of dairy production, and careful US$0.23 per liter in the United States and Western Eu- sector ani project analysis, to determine where alternative rope, and to as much as USS0.35 per liter in Japan. systems can be most efficiently developed, should precede In China, efficient expansion of dairy production is any major investment decisions. In addition, potential con- likely to involve increasing the quantity and quality of straints to the efficient spatial distribution of dairy produc- roughages, with roughage feed supplemented by concen- tion include the transport, storage, and marketing systems, trates. The location of dairies based on such a feeding which may make it feasible to develop only a suburban- system needs to be considered in relation to the sources of based dairy industry dependent on feed concentrates, and feed supplies and the locations of markets for milk and planning and management systems, which need to ensure other dairy products. Major markets are likely to be in that muricipalities do not develop integrated dairy opera- urban areas. Nevertheless, roughage, which has a very low tions when production in outlying areas for sale in the return per hectare, is likely to be most efficiently grown in municipa lity would be more economical. remote and hilly areas with few alternative land uses. The bilities for nonagricultural development. However, the towns, and toward more outlying plain areas. Finally, in economic cost of transport in areas close to cities and hilly or mountainous areas, pasture and livestock devel- towns will usually be lower than in areas farther away opment, coupled with tree crops on gentler slopes and because chemical fertilizers and other agricultural inputs forestry on steeper lands, will frequently be the most are frequently transported through cities, and the main economical pattern of agricultural development. markets for agricultural products are usually in urban It is not easy, however, to predict the precise directions areas. Accordingly, areas close to cities and towns will of efficient agricultural specialization, as Box 5.1 (on the need to specialize in activities (such as cultivation of vege- location of dairy production) indicates. The development tables and some perishable fruits) that have relatively high of a flexible and economically rational system of agricul- returns per hectare and high transport costs. Foodgrain tural management is therefore very important (see Chap- production, which is characterized by relatively low unit ter 3). The 1979-80 rural reforms, which gave farm transport costs and returns per hectare compared with households more autonomy in deciding what crops to activities such as vegetable growing, should probably be grow, have already done much to facilitate agricultural moved increasingly away from areas close to cities and specialization and should continue to do so. To ensure 77 that the resulting specialization is economically efficient, for example, has a sunny climate and fertile soil, which however, the incentive system must provide appropriate could provide the basis for high-value horticultural crop signals to farmers on what crops to grow and what other production: indeed, melons and deciduous fruit from agricultural activities to develop. Gansu sell at high prices in Hong Kong, while Chinese The most important incentives are output and input cabbage procured in Gansu for Y 0.02-0.04 per kilogram prices. On average across all regions, these should be such sells in northeastern cities for more than Y 0.20 per as to balance the demand and supply of particular com- kilogram. These price differentials suggest that better modities and to encourage optimal input use (see Chapter transport might well enable some farmers in Gansu to 3). In addition, variations in agricultural prices from one specialize in horticulture rather than grain production, region to another should reflect variations in the eco- purchasing their foodgrain from elsewhere. More gener- nomic costs of transport. The prices of products mar- ally, improved transportation would permit more rational keted through the collective and individual sectors al- use of suburban land, reduce input supply problems, dis- ready reflect prevailing differences in the actual costs of tribute perishable products more widely, and allow a transport, but there are differences between actual costs broader spectrum of China's farmers to reap the income and economic costs that need to be reduced by adjusting gains of high-value agricultural production. the prices of items such as fuel and trucks. In the state sector, in contrast, input and output prices may not even Industrial Specialization and Location reflect the actual costs of transport, because, for example, In contrast to agriculture, and despite recent policy procurement prices for agricultural products may be set changes and reforms, China's rather low degree of local at higher levels in more remote or mountainous areas industrial specialization does not seem to be increasing. than in high-productivity crop production bases. Such Specifically, although some localities already specialize to practices help to raise incomes in poor areas, but can also some extent in particular industrial products (Wuxi in hamper agricultural specialization. It may be more effi- textiles and clothing, for example), transport problems, cient to replace these practices with other policies and shortcomings in commerce and material supply, price dis- measures that benefit poor areas, including improved tortions, barriers to competition, and the fiscal system transport. combine to provide strong incentives for local self-suffi- Recentlv announced agricultural price reforms should ciency in a variety of other industrial products. Each make it possible to do without (from 1985 onward) most provincial and municipal planning bureau accordingly mandatory production planning-including quotas for continues to emphasize the development of a wide range sown area and procurement of grain, which have been of industrial activities on a scale tailored to local needs, major obstacles to agricultural specialization. In Jiangsu, rather than concentrating its resources on large-scale low- for example, farmers living close to urban areas want to cost production of a narrow range of commodities for specialize in activities that have a high value per unit of export to other localities. Because this could have a seri- land, and township authorities have had to pay them fifty ous adverse impact on China's overall investment effi- fen per jin of grain (three times the state price) to ensure ciency, there is an urgent need for measures to reduce the that they grew enough grain to meet the state procure- present conflict between what seems rational for a paric- ment target. But, while many such forms of directive ular locality and what is rational for the whole economy. planning can and should be abolished, economic levers Specialization of medium-size and smaller cities in par- will need to be supplemented by administrative measures ticular industries is usual in other countries (large cities in some situations. In hilly areas, for example, some types tend to be more diversified).6 But because in other coun- of annual cropping may produce large short-term benefits tries much of a city's labor force is engaged in commerce but have major long-term costs for the natural environ- and other services, the proportion of employment in the ment. The development and enforcement of regulations industry of specialization rarely exceeds 40 percent and is on land use should therefore remain an important respon- typically only 10-20 percent, though sometimes with a sibility of both national and local authorities. The Gov- significant percentage of other workers engaged in sup- ernment will also need to continue providing special assis- port activities for this industry. Even so, difficulties in the tance, including relief grain, to poor areas and to develop industry concerned could depress the whole local econ- special programs to stimulate their development (dis- omy, which means that specialization has risks as well as cussed later). benefits. In China, moreover, the past pattern of indus- Agricultural specialization in China will also be greatly trial investment will limit the speed of movement toward affected by the adequacy of transport and trade facilities greater specialization of localities, since it will often be (also discussed later). The development of individual and collective enterprises, as well as of state enterprises, in transport and commerce is already assisting specialization. 6. The experience of other countries is discussed in greater But transport infrastructure remains a bottleneck. Gansu, detail in Background Paper 2. 78 economically rational to keep a diverse range of existing Table 5.4 Extent and Composition of Rural enterprises in operation for some time, rather than to Nonagricultural Activities in Selected Countries close them all down straightaway (see Chapter 7). What (percent) is crucial, though, and what does not yet appear to be Other Asian happening, is that new industrial investment should be- Cbina, Thailand, Indonesia, developing I I' I Coanponent ~~~~~~~~1982 1975-76 1980 countries come much more specialized. Component Location decisions for heavy industry have in many Share ol'rural labor force countries been based as much on historical and political involved in nonagricul- considerations as on current economic conditions. But turala.ctivities 13.8 16.6 34.7 15-35 China's spatial pattern of heavy industrial development Composition of rural non- seems to have unusually little economic rationale. In re- agricultural labor force cent years, many of the small and inefficient heavy indus- Manufacturing 45.2 . 23.1 15-39 Construction 8.9 . . 7.5 4-14 trial plants have been closed. Yet industries such as iron Coiiimcrce 12.2 . . 29.7 12-29 and steel-where economies of scale are critical-still Tra sport 4.6 .. 5.5 5-10 have many small, locally operated plants producing at Oth,er services 29.1 .. 34.2 25-45 high cost (see Box 5.2). Moreover, medium-size and large ANote Da a for China have been adjusted to conforin with the adjustments heavy industry plants are still concentrated in very large to total employment made in other sections of this report. cities (see Table 5.3), despite the fact that these industries Source. Lackground Paper I and World Bank Reports 3906-TH on lhai- land and 3 586-IND on Indonesia have very large spatial needs, create substantial environ- mental problems, and do not always require the many services and other benefits that come with a large-city location. Specific decisions on relocation of heavy indus- approFriate for different types of light industry, because tries can be made only after a detailed industry-by-indus- size aff'ects economic costs and benefits. The economic try assessment. But measures to concentrate production costs of land, for example, are much higher in larger cities in fewer enterprises-located away from the largest cities (and in the center of the city than in the suburbs) because and often closer to resource deposits-could not only the land has a greater number of alternative, high-value, improve national efficiency, but also benefit poorer inte- uses. The economic costs of materials and of housing, rior provinces, which currently produce raw materials for commuting, and food for workers also tend to be higher processing elsewhere (see Chapter 5). in larger cities because of higher transport and land costs. Concentrating some types of heavy-industry produc- For some activities, however, there may be additional tion in fewer enterprises, as a means of realizing econo- benefits as well as costs to locating in larger rather than mies of scale, implies that many small rural enterprises smaller cities. These benefits (often referred to as urban- might eventually need to be closed-and some already ization economies) arise from having immediate access to have been. However, only some 6 percent of China's other ndustries, diverse labor markets, large local con- rural labor force is now employed in manufacturing, sumer markets for product testing, and the diverse envi- compared with 3 to 15 percent in other Asian countries ronment that accompanies increasing urban size. (see Table 5.4). And only 24 percent of rural manufactur- Light industrial enterprises producing standardized ing workers in China are employed in the food process- products such as bicycles (see Box 5.3) tend to benefit ing, textile, and clothing industries, compared with 40 to much ,nore from economies of scale arising from the size 60 percent in other Asian countries.7 Accordingly, while of their own industry than from urbanization economies; there may be a decline in output from some types of rural it is therefore likely to be more efficient to locate them in industry, both national efficiency and equity could be smaller cities. But high-technology industries and ad- enhanced by measures to promote rapid growth in other vancecl service industries benefit much more from urban- types of rural industry, including coal mining (Chapter 4), ization economies, which are frequently sufficient to off- and cement and bricks (which enjoy a high degree of set the additional economic and social costs of location in natural protection), as well as agroprocessing and some a largc city. Some of the activities now located in large other light industries. cities like Shanghai and Tianjin should thus be progres- The efficient location of many light industries depends sively transferred to smaller cities and towns. But these less on location of raw materials than on the cost of other activities should be replaced by new activities such as inputs and the location of markets. Those industries serv- electrcnics (and services), which are only now assuming ing a broad market and not subject to scale economies impornance in the national economy. As a result there may be widely dispersed, with rural as well as urban plants, while those that can benefit from scale economies will probably be most efficiently located in urban areas. A 7. See Background Paper I for further discussion of these critical issue, however, is the size of city or town most issues. 79 Box 5.2 Lctionalssues in the Irona tel Industry Chinras iron and steel indust consists o approxiately in Sh ai, which currently uses pig iron transported 2,000 state and collective enterprises, vaing from small from other parts of the country and which in the future enterprises that produce less;thaVn5,000 tons of steel per should probably switch to imported materials. There are, year to; th:e4 large :integratedS Anshan ;Steel; Company in however, many plants in Vother partsof the country which iIaaoning, which in 1982 produced 6.1 million tons of pig shuld ultimately be closed down unless they are produc- iron, 06.8 fmillion4tons of steel, and 4.4million tons ofsteel ing light, nonflat products and using iron and steel scrap. products. ThAe industr variesiwidely by technology used 0000 ;$if ;00: Decisionszon the 0structure and location of iron and steel (31 percent of steel prouction sttll comes from open plants ougt to he based on sound sector planning, which hearth furnaces) and costs of production (pig iron produc- takes account of the interdependence of decisions on the tlon costs, for example, vary frm Y I 38 per ton for large mix offinal products, the size of productive units, and the enterprises to Y 192 per ton forimedium-size enterprises coe ehnology, as well as the location of productive and Y 304 per ton for small erprises). Teindustryis units Efforts to strengthen planning in the industry are also very dispersed, with all provinces exceptXizang and therefore:critical. Several other constraints to efficient spa- Qinghai producing pig iron and all except Xizang produc- rial development need to be addressed. One major issue ing crude steel and steel producs. concerns the:relationship between different administrative Efficiency improvements in theIiron and steel industry levels of government in planning and managing the indus- depend in part on changes in theIstructure and location of try. Gansu, for exampl, has one national iron and steel enterprises, including increased concentration of produc- enterprise located close to major iron ore and coal de- tion in a few lar integrated enterprises. China already has psits, and one provincial enterprise inthe provincial capi- ten integrated enterprises with a crude steel-making capac- ta, which depends on supplies of pig iron transported ity of 1 .0 million tons per year or more, hut these account from the national enterprise over 1,000 kilometers away. for only 60 percent f total steel production. Moreover, From a national economic point of view, it would almost the minimum efficient scale for an integrated enterprise certainly be more efficient to close the provincial enter- producing steel by the basic oygn systemand rolling a prise, but because of uncertainties concerning expansion fairl comprehensive range of products is likely to be plansfor the national enterprise and the disposal of its nearer to8mllion tonsperyear.Internatiol experence ial government decided to continue suggests that tfere will also be economic justification for operating and even to expand the provincial enterprise. In some small nonintegrated steel mills producing liht, non- Hubei, because of problems of coordination between ad- flat steel products, and rig e nd steel ministrative leels, four iron and steel enterprises were scrap or on directly reduced iron if cheap sources of energy built in one town, each with similar output but reporting are available locally, to different administrative levels (national, provincial, mu- Efficient location of iron and steel plants depends criti- nicipal, and county). All but the national enterprise were cally on transportation costs for a few inputs (iron ore, small scale, experienced problems in obtaining raw materi- scrap, coal, and electricity) and for finished products. Inte- als, and showed poor economic results. grated steel enterprises are involved in a weight-reducing The material supply system also appears to be a con- activity, and transport cost considerations are likely to jus- straint to optimal spatial development in the iron and steel tify locating such enterprises close to sourcesd of raw mate- industry. This is particularly apparent in Jiangsu, which has rials, even though the economic costs of transport are a diversified and rapidly expanding industrial sector that generally higher for a ton of finished products than for a includes a significant iron and steel industry, even though ton of material inputs. China's largest iron and seel enter- the province has only minimal iron ore and energy re- prises tend to be located in the North, Northeast, and sources. Local authorities view the iron and steel industry Central South regions, close to abundant supphles of iron as critical to overall industrial development in the province ore, coal, and limestone, and future development of the and are even willing to subsidize the industry because it industry should probabltinvolve expansiono f existing ensures regular supples of steel products to the many com- enterprises and construction Of n these loca- : mune, brigade, and other enterprises that are outside the tions. The only major xceptions are theBaoshan iron and plan and have,no guarantee of receiving steel products steel plant, which will depend on imported ore once port through the tnormal materials allocation process. facilities have been developed, and another large enterprise may be no overall decline in the relative importance of tion, will require action on several fronts. Of fundamen- larger cities. tal importance will be systemic reforms (see Chapter 10) To increase industrial specialization, and more generally that increase the cost-consciousness of enterprises, espe- to improve the economic rationality of industrial loca- cially state enterprises, and expose them to stiff competi- 80 Box 5.3 Locational Issues in the Bicycle Industry China's bicycle industry, the largest in the world, has been activities to ensure their full utilization. expanding especially rapidly in recent years. Between The locational implications of each alternative are some- 1978 and 1982, total bicycle production increased from what different. If bicycle production is centralized in a few 8.54 million to over 24 million, and the number of enter- very large integrated enterprises, most provinces will no prises engaged in bicycle produaion grew from 3 8 to 140. longer be involved in bicycle production, but if enterprises All but three of China's twenty-nine provinces are now specialize in production of bicycle components, more engaged in bicycle production. From a national economic provinces may remain involved. In either case, however, point of view, however, the bicycle industry appears to bicycle enterprises should probably be concentrated more have been expanding very inefficiently. Bicycle production in small rather than large cities, because the economic costs offers major economies of scale, with integrated enterprises of land, labor, materials, and transport are much higher in producing fewer than 300,000-500,000 bicycles per year large cities, and enterprises producing a standard product likely to be unprofitable. But in China, average output per such a, bicycles are unlikely to benefit sufficiently from enterprise declined from 225,000 in 1978 to 173,000 in urbanization economies to justify paying such additional 1982, and fourteen provinces produce fewer than 500,000 costs. For the Shanghai bicycle factory, for example, the bicycles per year. Unit production costs range from Y 70 economic costs of production in its present location are at in Shanghai to more than Y 200 in smaller enterprises in least Y 10-20 per bicycle greater than if it retained the Jilin, Heilongjiang, and Nei Monggol. same excellent management but were located in a nearby International experience suggests two alternative types mediuri-size city. of structural change to help realize economies of scale and One means of facilitating optimal decisions on bicycle improve overall efficiency in China's bicycle industry. Bi- plant klcation would be to make enterprises pay some of cycle production could be concentrated in a few integrated the additional costs of locating in a large city, ideally enterprises; with an optimal capacity (estimated by the through a pricing system in which enterprises pay for the Ministry of Light Industry) of 3 million bicycles per year varying economic costs-of such factors as land and trans- and demand unlikely to exceed 30 million bicycles per port-of each location. Alternatively, the tax system could year, this would imply reducing the number of bicycle be modified to include higher tax rates for locations with enterprises from 140 to 10. The alternative would be to greater costs, such as the center of large cities. At the very realize economies of scale from improved production orga- least, investment planning should be improved and differ- nization-for example, by some enterprises specializing in ences in the economic costs of alternative locations taken production of bicycle components or sharing facilities, into account. equipment, and labor with enterprises involved in other tion, especially from enterprises in other localities. Also both inputs and outputs. Although most enterprises al- important will be improvements in transport and com- ready pay for transport, there are apparently large differ- merce (discussed further below), without which competi- ences between actual and economic tariffs. Recent in- tion would be limited and localities unable to rely on creases, in rail tariffs (including the latest decision to external sources of supply for essential goods. General charge more for short hauls) are a step in the right direc- price reforms and changes in the system of public finance tion, as is the pricing flexibility allowed to individual and are also needed, to reduce the incentive for local authori- collective transport undertakings. But studies are needed ties to set up (and preserve) their own small enterprises in to establish economic user charges for state-operated rail, profitable or highly taxed industries. road, and other transport services, based on long-run Since the changes now under way in the system of marginal costs. nonagricultural economic management will increasingly Perl2aps even more vital will be a system for charging delegate decisions to relatively autonomous enterprises, enterprises for the use of land that takes account of vary- price reform will also be essential to give enterprises ap- ing lard values among regions and localities. Experience propriate guidance on what industries to develop and from other countries suggests, for example, that the eco- where to develop them. This is true both of prices in nomic cost of land can rise by as much as 25 percent per general and of some specific prices that crucially influence kilomcter as one moves from agricultural land on the location decisions-transport tariffs, land use charges, and edge of an urban area to core land in the inner city. If the wages. economic cost of agricultural land on the edge of a Chi- As in agriculture, industrial enterprises must be sensi- nese city is assumed to be Y 0.8 per square meter, the tized to variations in the economic cost of transport, for economic cost of land in the city center would be about 81 Y 10 per square meter for Wuxi, for example, which has desirable, however, it is important to base them on care- a radius of about ten kilometers, but would be as much as ful investment planning including, where necessary, use Y 120 for Shanghai, which has a radius of twenty kilo- of planning models that take account of the interdepen- meters. dent nature of decisions on location of productive units, Enterprises must be made to feel these wide variations size of productive units, choice of technology, time phas- in the opportunity cost of land when making spatial deci- ing of the stages of the project, and mix of final products. sions. This could be done through a system of land taxes, (Planners also need to bear in mind that profitability under which the assessed tax varies according to the fea- calculations based on current wage, capital, and zero land tures of each locality (rural versus urban, small versus costs can be very misleading as indicators of the relative large urban area, suburbs versus center, and so on). Other economic efficiency of alternative locations.) methods consistent with public ownership of land, such as a system of rents determined by competition among those who need the land,8 could also be considered. In Transport, Commerce, and Other Services introducing either a land tax or a rental market, however, During the past three decades, China has made major the whole system of local public finance would need to investments in transport, especially railways, and has de- be reconsidered, to ensure that both enterprises and local veloped a substantial system of commerce and material governments were subject to consistent and appropriate supply based on administrative allocation of key com- fiscal incentives. For example, given their present direct modities. In the decades ahead, rapid economic growth dependence on enterprise profits, local urban govern- will undoubtedly require further expansion of these basic ments might have mixed feelings about levying a land tax supporting sectors. But rationalization of the spatial pat- that would drive some enterprises away. tern of economic activity, together with other structural The cost of labor can also powerfully influence enter- changes and systemic reforms, could significantly alter the prise location decisions. Even if there were no restrictions pace and pattern of development in both transport and on labor mobility,9 interlocality differences in the eco- commerce. nomic cost of labor would remain, corresponding to dif- ferences in the economic cost of living. Experience from Transport Needs other countries shows that these differences can be quite Difficulties and delays in transportation are already an substantial: in the United States, for example, wages must important concern in China, as is the longer-term possi- rise by 0.3 3 percent for each I percent increase in the bility that transport shortages may hold down the overall population of an urban area to maintain the same con- speed of economic growth, especially given the length of sumer standard of living. In southern Brazil, the corre- time needed to construct new railways, roads and other sponding ratio has been estimated at 0.66 percent, imply- infrastructure.'° It is extremely difficult, however, to pre- ing that wages must be at least twice as high in cities of dict how rapidly the total volume of freight transport will more than 200,000 as in cities of under 50,000 to ensure in fact need to grow, given any particular assumption similar living standards. In China, however, enterprises about the future trend growth rate of national income. are not yet affected by such issues because rent and food This is largely because by international standards China prices and wages are all fixed at fairly uniform levels. To currently uses an exceptionally large amount of freight the extent that price reform involves more local variation transport-more than three ton-kilometers per dollar of in consumer prices, enterprises will be obliged to reflect GNP, as compared with well under two ton-kilometers in at least some of this variation in wages. But if price India, Brazil, and the United States. The Soviet Union, reform does not involve such changes, then other mea- though, is even more freight-intensive (over four ton- sures (including local variations in tax rates) could be used kilometers per dollar of GNP) than China. to influence enterprise decisionmaking. Though the freight-intensity of particular countries is Even with prices, taxes, and wages such as to provide affected also by their specific geography, these compari- appropriate signals to autonomous, cost-conscious enter- sons suggest that future growth of transport demand in prises, some decisions on spatial development would still China could vary widely, depending on the pattern of have to be made by industrial bureaus or ministries. For economic growth, both in terms of sectoral structure (for example, in subsectors such as cement and fertilizers, it may be economically most efficient to determiie the location and scale of several plants simultaneously, to miimz trnpr plu prdcto cots In oThe sub- 8. A system of rents determined by competitive bidding suchmize transpot alut s aroducond ecostroIni , ner among enterprises would, in principle, be more efficient than a sectors, such as automobiles and electromc cmps, one land tax system and would avoid the administrative problem of very large plant might be able to satisfy much of the deciding on appropriate tax levels. national market, but with no obvious choice as to its 9. These restrictions are discussed in detail later in this chapter. locationi. In cases wlhere administrative interventions are 10. Annex 6 contains a fuller discussion of transport issues. 82 example, industry is more transport-intensive than ser- major program for paving and upgrading intraprovincial vices) and in terms of spatial location (which influences highways will be necessary. average transport distance). It will also be affected by Greater industrial specialization will also involve a very system reform, which will, on the one hand, tend to large increase in interprovincial and interregional trans- increase transport requirements because of increased spe- port of light industrial goods. Currently, most roads in cialization and exchange, but, on the other hand, could China radiate from provincial capitals, forming separate greatly reduce wasteful use of transport facilities, for ex- networks with few interconnections; rail is therefore the ample through less cross-hauling, more preliminary proc- only effective link between provinces and regions. As a essing of materials, and more rational location decisions. result, light industrial goods are usually transported short The possible magnitude of the combined impact of distances by road to the rail network and then by rail changes in the growth pattern and system reform is illus- across provincial and regional boundaries. However, light trated not only by international comparisons but also by industrial products, which often have high value-weight China's own experience: in 1952-77, freight transport ratios or are quite fragile (and liable to be damaged by volume increased 55 percent faster than national income, transshipment), are more efficiently transported by road but in 1979-84 about 26 percent slower than national than by rail, even over long distances. Developing the income. intercity and interprovincial road network for the trans- Within the framework of the macroeconomic projec- port of light industrial goods should therefore be a high tions introduced in Chapter 2, alternative projections of priority in transport investment. (Improvements in the freight transport volume are presented in Annex 6. In volume and quality of air freight services could also make QUADRUPLE, with continuing strong emphasis on material a significant, though much smaller, economic contribu- production, and with industry and agriculture using tion.) transport only moderately more efficiently than in the It is widely recognized in China that road transport past, freight volume could reach 3,200 billion ton-kilo- should and will play a larger role in the future. Yet there meters by the year 2000. In BALANCE, by contrast, with seems to be no strategic plan to bring this about. Indeed, the same growth of national income achieved through during the past few years, there has been little road con- faster expansion of services and slower expansion of in- struction, and the situation is unlikely to improve sub- dustry, and with somewhat greater efficiency in transport stantiall y so long as roads remain almost entirely a local use, freight volume in 2000 would be 2,400 billion ton- government responsibility. This is partly because the kilometers. Even in the latter case, China's projected present price and fiscal systems make directly productive freight-intensity (2.4 ton-kilometers per dollar of GNP in investment (especially in industry) generally more attrac- 2000) remains high by comparison with other countries, tive to local governments than investment in roads and which suggests that the possibility of an even slower other infrastructure. But it is also because the interest of growth of freight volume cannot be ruled out. each locality in contributing to improvement of the road But although the overall volume of transport require- network linking it with other localities is reduced in the ments in the longer term is highly uncertain, changes in absence of higher-level coordination and cost-sharing the industrial and spatial pattern of development will among localities. For this reason, in many countries, the certainly require a new and different mixture of transport finance, planning, and coordination of road development investments, as well as changes in organization and man- has been made a responsibility of higher (national and agement. Intraprovincial specialization in agriculture and provincial) levels of government, with delegation of re- industry, for example, will involve the movement in fairly sponsibility only for actual construction and maintenance small lots of a great variety of goods, many of which are to local ,overnments. perishable, over short to medium distances (5 to 100 As regards sources of finance for road development in kilometers), to and from dispersed origins and dcstina- China, the issuance of bonds (as in some other countries) tions. To help meet these diverse and growing demands as has recently been encouraged, and rural labor contribu- efficiently as possible, a major expansion of the transport tions are increasingly being commuted into cash pay- network linking rural and urban areas will be essential. In ments. But much more use could be made of road-related parts of southern China, richly endowed with dense net- taxation. Since direct road user charges (tolls) often have works of rivers and canals, water transport can continue undesirable effects on traffic patterns, most countries to play an important role. But in most parts of the coun- have adopted a system of road user charges based on try, roads will be the critical constraint. One major task vehicle a.nd fuel taxation. The tax component of gasoline will be to expand the rural road network, which cur- and diesel motor fuels is often very substantial (up to 50 rently totals 300,000 to 500,000 kilometers, compared percent, and in some countries even more), and vehicle with 900,000 kilometers in India, 1.2 million kilometers taxes based on axle loads recoup the high share of road in Brazil, and more than 100,000 kilometers in Thailand maintenance costs caused by heavy trucks. Such a system (which is smaller in area than Sichuan). In addition, a of road user charges in China could generate sufficient 83 Box 5.4 Trade, Pricing, and Transport of Cement In most countries, the raw materials for cement produc- Box Table 5.4A Local and Centrally Controlled Cement tion are widely available, and the industry consists of Prices many plants located close to major markets in order to fyuan p- minimize delivery costs, which are very high in relation to Local cement Centrally controlled the value of the commodity. In the United States, for delivered by cement delivered by rail example, the ratio of freight cost to production cost is Ciponene road, 50km 2,500km 3,000km about 0.5 for cement, compared with 0.04 for steel. Chi- Cement price 80.0 50.0 50.0 na's cement industry is also characterized by many plants Transport tariff 10.0 35.2 42.3 Delivered price 90.0 8 5.2 92.3 and dispersed locations, but average transport distances are Delivered_prlce_90.0 _8___.2 _92._3 significantly higher than in the United States primarily Source Data provided to World Bank staff because the North and Northwest have little local cement production and must trade with provinces that have sur- Improving the efficiency of cement distribution will de- pluses, especially Liaoning in the Northeast, which ac- pend in part on further dispersing production from its counts for 9 percent of national output. More important, original concentration in the Northeast, and efforts to however, cement production in China is divided between achieve this are already under way. But dispersed produc- centrally and locally controlled plants, with output from tion is efficient only if local demands are actually met by centrally controlled plants priced much lower than local local production. Some provinces with centrally controlled output and frequently of higher quality. As a result, and plants are now obliged to send most of their cement out- even without accounting for quality differentials, centrally put to other provinces, but then have to buy cement from controlled cement delivered by rail can compete with local locally controlled enterprises in other provinces. This cement delivered by road well beyond the distances found wastes a lot of transport. Reforms in the trade and pricing economical in other countries (see Box Table 5.4A). system for cement are therefore critical. funds for an accelerated road construction program, as output and on the extent of industrial specialization but well as contributing to an economically appropriate divi- also on how quickly the materials supply system can be sion of traffic betwcen road transport and other modes of improved. Currently, for example, centrally allocated ma- transport. terials at state prices are more competitive than local To develop China's road transport system to a suffi- materials available at market prices, even when delivered cient extent, various other steps would also have to be over long distances. The result is excessive interprovincial taken, some of which would require greatly increased movement of materials, because some provinces are horizontal coordination among different government obliged to send out materials under centrally adminis- ministries and agencies, and possibly the establishment of tered allocation and then purchase similar materials from a superior agency specifically charged with road transport other provinces at higher prices (see Box 5.4). In these development. These steps include production of different circumstances, a change from administered to market dis- (smaller as well as larger) and more fuel-efficient trucks, tribution of key materials could significantly affect inter- better management and utilization of truck fleets, upgrad- provincial movement of raw materials and hence the re- ing of fuel quality, improvement of fuel distribution sys- quired railway investment program. tems, and rationalization of the motor fuel price struc- Since increased energy requirements in China will have ture. For example, eliminating the large subsidy on diesel to be met largely by coal (see Chapter 4), and since water fuel for agricultural use would help to clear rural roads of and road transport of coal can increase only to a certain walking tractors, which slow down all traffic as well as extent, the amount of coal transported by rail will con- being highly energy-inefficient by comparison with small tinue to grow. Rail transport distances for coal are also pickup trucks. But this would be difficult because plans to likely to continue to increase, as coal mining will be develop the production of small trucks (and fuel for increasingly concentrated in the North, and in particular them) have lagged far behind growth of demand. in Shanxi province. With coal production increasing to Railways will remain the most efficient mode for trans- 1,400 million tons by the year 2000, the share of coal porting raw materials and most heavy-industry products would be likely to increase to about 40-50 percent of over long distances. But the extent of these movements, total rail ton-kilometers (up from 32 percent in 1980), and hence the railway expansion program, will depend depending on the extent of transport rationalization for not only on the overall growth and pattern of industrial other commodities-for example, processing of wood 84 tkit and ores before shipment and greater use of coastal and made r-garding interfacing of different modes, including inland water transport because of more appropriate loca- technical solutions and standards to facilitate intermodal tion of new heavy industry plants. transfer or avoid unnecessary transshipment. Examples Total rail freight transport in the year 2000 might grow include increased containerization for high-value prod- to about 1,300 billion ton-kilometers under the BALANCE ucts, self-unloading ships to reduce the need for port scenario, and 1,700 billion ton-kilometers under the facilities, and ocean-going barges to avoid transshipment QUADRUPLE scenario (see Annex 6), increases 2.3 and 3.0 betwee:i coastal shipping and inland water transport. To times 1980 levels, respectively. These increases in rail realize 1:he potential efficiency gains from intermodal co- transport capacity seem quite feasible, especially given the ordination requires prices and tariffs that reflect the full demonstrated capacity of the Ministry of Railways to cost of each transport mode, but also more competition secure finance and implement investment projects. For and specialization of transport modes. In market econo- example, with a modest increase (15 percent) in the total mies, where intermodal coordination is achieved mainly railway network, to 60,000 route kilometers by 2000, through pricing and competition, many specialized entit- the required increase under the BALANCE scenario could ies pros ide transport, storage, and transfer services. There be achieved if average freight density were to double, are also many joint ventures among enterprises operating from 12 million ton-kilometers per kilometer of rail route in different transport modes to facilitate intermodal trans- in 1980 to 24 million ton-kilometers per kilometer of rail fers (for example, jobbers, storage companies, and com- route by the year 2000. In the U.S.S.R., a comparable panies leasing containers, railcars, trailers, and other increase in traffic density was achieved between 1960 and equipment). 19 7 5 by switching from steam to electric and diesel trac- Despite a rapid rise in recent years, personal mobility in tion, combined with double-tracking of about one-third China is still low by international standards, and passen- of the railway network. Steam traction declined from 85 ger transport can be expected to grow fast. Rural eco- percent in 1955 to 15 percent in 1975 and has since nomic (development has lately generated a lot of demand vanished completely. By contrast, 70-75 percent of total for short-distance travel. Increased internal economic spe- traction in China is still provided by inefficient steam cializatian and trade will also require far more personal locomotives, and only 18 percent of the rail routes are contacts and business trips, and experience in other coun- double tracked. The investment cost of network expan- tries shows that people like to spend a considerable frac- sion, double-tracking, and electrification to meet pro- tion of income increases on travel to visit relatives or for jected traffic demand to the year 2000 would be large, recreation. This demand for greater mobility has invari- but not prohibitive-perhaps Y 5 billion to 7 billion per ably been accompanied elsewhere by demand for private year in the BALANCE scenario, and Y 7 billion to 9 billion vehicles (motorcycles as well as cars). Public passenger in the QUADRUPLE scenario. transport will shift toward (lower-cost and more flexible) Despite expansion in recent years, domestic water buses, as well as aircraft, with a decline in the share of the transport is still not used to its best economic advantage. railway;. This strengthens the case for improvement of The network of navigable inland waterways has de- China's road network, as well as of rural and intercity bus creased by over one-third since 1960, partly because of (and air) services, although rail passenger transport capac- dams built to generate electricity and provide irrigation ity will also have to expand quite fast. Within towns and under the responsibility of a different ministry than that cities, too, public transport will need to be strengthened. controlling water transport. With a great increase planned in hydropower development to the end of the Developing Commerce century, the use of waterways for transport must be rec- Like transport, commerce has an essential role in facilitat- ognized and included at an early stage in the planning and ing specialization, both of localities and of enterprises, as evaluation of hydropower sites. Coastal shipping-which well as providing the infrastructure of market regulation offers an economic alternative to north-south railways- (Chapter 1). To accomplish this, however, China's com- has likewise grown more rapidly in recent years, but is mercial system would need to become much more flexi- still underdeveloped, most notably by comparison with ble, responsive, and diversified, which would entail basic Japan. Much greater use of coastal and inland water ship- changes in organization and management. It would also ping could be made if more heavy industries were in need to become much larger, especially in terms of em- future located so as to avoid the need for inland transport ployment, but also in terms of fixed assets (although it to and from ports. might reed less circulating capital than at present-see Considerable efficiency gains could be made in the en- Chaptei 2). For example, in the QUADRUPLE projection, tire transport system through better intermodal coordina- where the pattern of development resembles that of the tion, which has thus far been impeded by the vertical and Soviet UJnion and of China in 1950-80, employment in largely self-contained organization of China's transport commerce remains exceptionallv small by comparison and other sectoral agencies. Better choices need to be with other countries-under 5 percent of total employ- 85 ment in 2000. In the BALANCE projection, reflecting there is also need for some legal process or institution more fundamental changes in the system of economic with the power to impose large fines and compel com- management, it increases to 10 percent of total employ- pensation, to which any enterprise or individual sub- ment. jected to local government interference could appeal di- Development of collective and individual commerce in rectly.11 recent years has already contributed to expanding and However, removing commerce from the control of improving retail trade, as has greater freedom of choice of localities may not be sufficient so long as localities in suppliers in the (still overwhelmingly dominant) state- China own many of the enterprises and can influence the operated retailing system. The efficiency of wholesale buying and selling decisions of enterprises in such a way trade-especially in key materials-has also been im- as to protect other local enterprises. Of course, in some proved by reductions in the scope of centrally adminis- circumstances, there may be good reasons from a na- tered allocation. Many enterprises are now allowed to tional viewpoint for providing special assistance to local sell some of their output themselves, and unregulated industries (Chapter 5). But the power to grant such assis- wholesale markets for producer goods and industrial con- tance, and the conditions under which it is granted, must sumer goods have been established in many cities. But the be very clearly specified by the central government- proportion of goods involved is still small. Centralized otherwise interlocality specialization may be seriously allocation has frequently been replaced by provincial- or harmed. county-level allocation, and shortages and bottlenecks re- Development of services other than commerce (dis- main, especially in the availability of materials for small cussed in Chapter 1 and 2) will also shape, and be shaped collective and commune and brigade enterprises. As a by, changes in China's spatial pattern of economic activ- result, many localities still feel obliged to develop local ity. This is so for social services and public administration, materials production capacity. There are also undesirable for personal services such as restaurants and hairdressers, pressures to enlarge the scope of administrative planning and for financial and other enterprise support services, to incorporate small urban and rural collectives. including law, telecommunications, and technical consul- In accordance with the Government's intentions, there tancy. Some of these service activities contribute to spe- should continue to be a steady shift from administrative cialization and exchange in material production, espe- to market allocation of both essential and other commod- cially by increasing the flow of technical, and commercial ities, in parallel with a decline in the proportion of indus- information. Some-colleges and centers of local govern- trial production subject to obligatory quotas. There ment, for example-in other countries constitute the should also continue to be a steady increase in the relative main specialized activities of particular small cities. And importance of individual and collective commerce, whose some sophisticated financial, technical, and commercial motivation, initiative, flexibility, and responsiveness to services benefit from clustering together in large cities. customer needs is generally superior to that of state-run But most services need to be spatially dispersed, close to commerce (which could, however, improve significantly their customers. As a result, most localities in other coun- if stimulated by competition). Individual and collective tries do not specialize in particular services, as they do in enterprises, moreover, should not remain confined to re- particular industries, but provide a wide range of services tail trade in consumer goods, but should probably also to meet local needs. eventually play a major role in wholesaling and material supply. To do this, they would need to be given much better access than at present to premises, credit, skilled Urban and Rural Development labor, and transport facilities. Despite controls on rural-urban migration, the distribu- Even with a more market-regulated system of com- tion of China's population between urban and rural areas merce, the powers of different levels of government to appears to be quite similar to other low-income countries control interlocality trade would remain a major issue. (see Table 5.5). The urbanization rate may even be above Despite the growth of collective and individual commerce the low-income country average if account is taken of and instructions from the central government, local gov- ernments continue to protect local industries by restrict- ing purchases of certain types of goods from outside the 11. The importance of reducing obstacles to interlocality trade locality. This sort of problem is by no means special to in China is well demonstrated by the recent national report of a China: local governments in most countries and historical member of a self-employed household in Sichuan who was asked to buv Lumber for his township's mechanized brickyard. It took periods (for example, the city guilds of medieval Europe) haveralso sought to protect local producers from external him fifty-four days and Y 1,453 for seven permits to transport have also sought to protect local producers fromexternal thirteen cubic meters of used building materials, which he bought competition. But to overcome it requires strong action by for Y 3,300, a mere sixty kilometers. The unreasonable expendi- the central government. In addition to general prohibi- ture raised the cost of the lumber, which was consequently rejected tions on administrative barriers to interlocality trade, by the brickyard. 86 Table 5.5 Urbanization and Economic Development in Selected Countries Percentage of tbe Percentage of the urban population Per capita GNP population living living in cities of Per 2 capzta GNEin urban areas 500,000 or more Country (dollars) 1960 1982 1960 1980 China 310 18 21 42 45 India 260 18 24 26 39 Other low-income countries 250 12 20 19 40 Average for lower-middle-income countries 840 24 34 28 47 Average for upper-middle-income countries 2,490 45 63 38 51 Source, World Bank, World Development Report 1984 (New York: Oxford University Press, 19 84). Table 5.6 Per Capita Income in Urban and Rural Areas, 1982 Urban Rural Total Per capita Percentage of Per capita Percentage of Per capita Percentage of Area income (yuan) population income (yuan) population income (yuan) population National average 574 21 269 79 33 3 100 Shanghai municipal average 668 63 437 37 582 100 Jiangsu provincial average 593 16 309 84 354 100 WuxiCounty' 572 7 412 93 423 100 Qianzhou Township' 560 100 560 100 Hubei provincial average 550 18 286 82 334 100 Gansu provincial average 648 15 174 85 245 100 Dingxi Countya 562 9 108 91 149 100 Dongye Townshipa 55 100 55 100 a. Data are for 1983. Source: Annex 5, Appendix l; data provided to the economic mission. differences in definitions."2 However, the differentials in to limit urbanization, emphasizing rural industrialization output and investment per capita between urban and instead. The current policy is much less restrictive, since rural areas appear to be larger than in most other low- it is now generally accepted that increased urbanization income countries and as a result urban areas account for will be an essential element of structural change and spe- an unusually large share of output and investment. cializa:ion. But some elements of the former policy per- The distribution of personal incomes between urban sist, especially regarding the growth of large cities and and rural areas of China is also somewhat different from movernent of population into formally designated urban that in other developing countries. The policy of national areas. The rationale for these continuing restrictions, uniform wages has resulted in extremely small income which could have significant economic and social costs, differences among (as well as within) urban areas. But merits careful consideration. So does the question of large differences in average incomes remain between ur- whether the restrictions should apply to people, or to ban and rural areas, especially in poorer regions (see Table directly productive investment, or to investment in infra- 5.6). In Jiangsu, for example, the ratio of urban to rural structure. More generally, it will be important to devise a per capita income is 1.9; in Gansu it is 3.7; and in Dingxi county, over 5. Moreover, these urban-rural income gaps reflect differences in real living standards and not just 12. China now classifies as urban both the agricultural and the differences in costs of living. The average urban house- nonagr cultural population of all towns with 3,000 or more perma- hold for example owns 1.3 bicycles, 2.1 watches, 1.2 nent residents, at least 70 percent of whom are engaged in nonagri- radios, and 0.4 televisions, whereas the average rural cultural activities. Previously, the agricultural population was ex- household owns 0.5 bicycles, 0.6 watches, 0.8 radios, cluded--hence the lower urbanization rate quoted in the World and 0.06 televisions. Bank's first economic report on China. However, the Chinese definition of "urban" may still he more restrictive than the defini- In most countries, the low- to middle-income transition tions used in many other countries. For a detailed discussion of has involved a significant increase in urbanization. In China'" urban definitions and how they compare with definitions China, though, the Government for many years sought used in other countries, see Background Paper 2. 87 Figure 5.2 Distribution of Employment, Capital, and Gross Output, Alternative Projections, 1981-2000 Employment Capital stock Gross output 5 18 3 1981 2 QUADRUPLE projection, 2000 17 2 MODERATE projection, 2000 23~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4 3 BALANCE projection, 2000 Rural agriculture * Rural nonagriculture El Urban nonagriculture * Urban agriculture Note For more detail, see Annex 4, Chapter 6. Source World Bank projections. 88 Table 5.7 Distribution of Employment and Population, Alternative Projections, 1981-2000 (percent) Measure and Employment Population projectin Rural Urban Totrl Rural Urban Total Shares of total 1981 75.8 24.2 100.0 79.8 20.2 100.0 2000 QUADRUPLE 73.2 26.8 100.0 75.0 25.0 100.0 MODERATE 79.7 20.3 100.0 81.1 18.9 100.0 BALANCE 67.8 32.2 100.0 70.0 30.0 100.0 Average annual growth QUADRUPLE 1.6 2.3 1.8 0.7 2.2 1.0 MODERATE 2.0 0.8 1.8 1la 0. 7 1.0 BALANCE 1.2 3.3 1.8 0.3 3.1 1.0 a. These growth rates are the same as the projected natural growth rates of the rural and urban population and therefore imply no rural-urban migration. Source. World Bank projections. management system and set of policies that generate an RUPLt. and MODERATE, would involve substantial rural- economically and socially efficient pace and pattern of urban migration. This is because large fertility urbanization. differentials between urban and rural areas have devel- oped in recent years, which will result in a natural growth Urban-Rural Balance rate of both the total and the working-age populations Although the macroeconomic projections discussed in that iS much higher in rural than in urban areas over the Chapter 2 are not uniquely associated with particular next two decades. Indeed, while the natural growth rate patterns of urban and rural development, Figure 5.2 illus- of the working-age population in rural areas will decline tratively connects each of these projections with a specific from about 3.3 percent at present to 1. I percent per year urban-rural distribution of employment, capital, and out- by the late 1 990s, the corresponding decline in urban put. 13 In QUADRUPLE it is assumed that the overall target areas will be from 2.1 percent to virtually zero. Accord- of quadrupling the gross value of industrial and agricul- ingly, only under the MODERATE scenario, with the urban tural output is accompanied by a strong emphasis on the share of total employment falling, would no migration be development of nonagricultural activities in rural areas. required (see Table 5.7, which assumes a rate of migra- This emphasis is reflected both in an increased rural share tion just sufficient to meet employment requirements). of the nonagricultural capital stock and in an increase in Even the modest growth in urban employment assumed the share of nonagricultural employment in total rural in QUADRUPLE would require substantial rural-urban mi- employment from 14 percent in 1981 to 25 percent in gration-and the pace of migration would have to be 2000, which is in the high range for lower-middle-income even gyreater if China were to achieve the urban employ- countries. However, urban employment is still assumed ment growth rate assumed in BALANCE. Even then, ur- to grow more rapidly than rural employment. In MODER- banization in China in the year 2000 would still be well ATE it is assumed that lower macroeconomic efficiency below the 34 percent average for lower-middle-income relative to QUADRUPLE is accompanied by greater restric- countries. tions on urban employment growth than in QUADRUPLE In order to efficiently manage the overall process of and no shift of capital away from urban areas. As a result rural- irban migration, the Government will almost cer- rural nonagricultural output growth is seriously con- tainly have to rely much more on pricing and other strained, while urban areas respond to labor shortages by economic measures, and much less on administrative adopting more capital-intensive techniques. In BALANCE, measures than in the past. At present the Government in contrast, it is assumed that the rapid growth in services provides only a relatively small number of people with and greater economic efficiency relative to QUADRUPLE official permission to move to cities of more than are accompanied by faster growth of urban employment, 200,000 inhabitants. It is also trying to limit the number which further reduces the need for accelerated (and from of people moving to small or medium-size urban areas by a national economic viewpoint inefficient) automation of stipulating that workers who move usually may not bring urban production. Growth of urban employment at the rate assumed in BALANCE, though more conducive to raising national eco- 13. The assumptions behind each of the urban-rural balance nomic efficiency than the lower rates assumed in QUAD- scenar os are discussed in more detail in Annex 4 (Chapter 6). 89 Table 5.8 Average Annual Growth of Urban and implications of such variations for urban-rural income Rural Income, Alternative Projections, 1981-2000 disparities, it is necessary to take account both of the ~,,,.oe. relationship between growth in output per worker and Measure and growth in earnings per worker and of the relationship projection Urban Rural Total between labor force and population growth. For example Earnings per worker if, as is assumed in the macroeconomic projections, inter- QUADRUPLE 4.5 4.5 4.6a sectoral variations in the growth of gross output per MODERATE 3.7 3 .7 3.5b worker are offset by relative price and tax changes, and BALANCE 4.5 4.6 5.0' earmings per worker increase at the same rate in all sectors Per capita income (see Chapters 2 and 3), urban-rural differentials in earn- QUADRUPLE 4.6 5.4 5.4 ings per worker would probably change very little (Table MODERATE 3.8 4.6 4.3 5.8). BALAN8'CE 4.6 s.s s.8 However, the labor force in relation to population will Ratio of labor force be increasing much more rapidly in rural than in urban to population 0.1 0.9 0.8 tenx w ail to_____population ______________________0___9 _________ areas over the next two decades, primarily because the a. The average earnings growth rate is above the urban and rural rate effects of family planning policies and slower population because the share of urban employment (which is higher paid) increases growth will be felt much later in rural than in urban areas overall. b. The average earnings growth rate is below the urban and rural rate (see Table 5.8). As a result, significant reductions in ur- because the share of rural employment (which is lower paid) increases ban-rural income disparities could occur even if urban overall. Source. World Bank projections. and rural earnings per worker increased at the same rate (see Table 5.9). Nonetheless, the projected income dispar- ities in 2000 are still large; additional measures to reduce them should be considered (see Chapters 3, 8, 9, and 10). their families with them or be granted urban registration (which deprives them of the subsidies normally available Development of Towns and Cities to urban residents). However, growth in employment Over the next two decades major changes will take place opportunities in collective and individual activities and not only in urban-rural balance but also in the number, other changes (such as the availability of foodgrains in size, and structure of rural towns and of small, medium, urban free markets) have made it much easier for people and large urban areas. The precise changes will depend in to ignore such administrative regulations; already an esti- part on the type of urban-rural balance strategy China mated 5 to 10 percent of the labor force in some large follows, but some major features are already apparent. cities are there unofficially. In these circumstances, eco- Within rural areas, small towns were neglected for many nomic efficiency as well as intraurban equity could proba- years and need to develop rapidly. In some cases this is bly be enhanced by replacing such administrative regula- already happening. In Jiangsu, for example, many small tions with measures to reduce the existing huge incentives towns actually experienced a fall in population during the for rural-urban migration to more economically rational 1950s and 1960s, as service and other activities were proportions. One way of doing this would be to increase increasingly taken over by the state. In recent years, how- the costs of locating activities and employing workers in ever, movement of people into these towns has been urban areas (especially large cities) through the land use actively encouraged, and many of the towns now qualify charges and other pricing measures discussed earlier. as urban areas. The alternative urban-rural balance scenarios also have major implications for investment flows and how these flows should be managed (see Chapters 9 and 10). Both QUADRUPLE and BALANCE, for exml,assume that there Table 5.9 Urban-Rural Income Disparities, , example, Alternative Projections, 1981-2000 would be an increase in the share of rural areas in the nonagricultural capital stock. This would probably re- Average Urban Rural Ratio of quire changes in the composition of accumulation, espe- perfcapta percaplta perscapta urbanrto cially as between government, enterprises (retained Year and income income income per capita profits, including those of collective and individual enter- projection (yuan) (yuan) (yuan) income prises), and households and in the institutions and policies 1981 290 554 223 2.5 that channel the savings of particular sectors into invest- 2000 ment in other sectors. QUADRUPLE 788 1,302 606 2.1 The alternative urban-rural balance scenarios result in MODERATE 645 1,125 524 2.1 quite large variations in the growth of output per worker BALANCE 846 1,326 617 2.1 between urban and rural areas. But, in order to assess the Sources World Bank projections. 90 Compared with other large countries, China also ap- cities ave the potential to grow in line with population pears to have a lower proportion of its urban population and incomes, as well as with service demand. in cities with fewer than 100,000 inhabitants (32 percent, In China, special attention needs to be given to the compared with 47 percent in India, 42 percent in Brazil, ways in which broader reforms (see Chapter 10) may and 41 percent in the Soviet Union), and the Govern- require or facilitate changes in the system of local public ment's emphasis on the rapid development of small and finance. At present, directly productive investment-es- medium-size urban areas seems appropriate. In addition, peciatLy in industry-absorbs an unusually large propor- though, an increase in the number (and major changes in tion of local expenditure, while economic and social in- the structure) of large cities will be required as part of the frastricture and related services are comparatively overall process of structural change and development of neglected. In the future, however, responsibility for fi- high technology and service activities. nancing directly productive investment may shift increas- The expansion and efficient development of towns and ingly to enterprises and to banks. At the same time, the cities will depend, among other things, on the quality and present social responsibilities of enterprises (for housing, quantity of public infrastructure and support services. For education and health) may be shifted increasingly to gov- example, if small towns are to become effective commer- ernment at various levels; increases in the complexity and cial centers for surrounding rural areas, transport links specialization of the economy may require more sophisti- must be greatly improved, more marketing facilities es- cated and expensive urban infrastructure and services. In tablished, and public utilities (including water supplies, these circumstances, to minimize duplication and gaps, it electricity, and telephones) greatly expanded. There will will be important to assign particular expenditure respon- also be a need to improve rapidly the quality of educa- sibilities to particular levels of government. tion, health, and other social services and to provide On the revenue side, greater independence of enter- more housing in these small towns if they are to consti- prises from administrative control may reduce the share tute a socially viable alternative to larger towns and cities. of receipts from direct enterprise ownership in local gov- In urban areas in general, the quality of education and emient budgets. To offset this, there could be greater health services tends to be much better than in surround- reliance on land taxes (or rents, as mentioned in Chapter ing rural areas, but the stock of housing is small'4 and 5), which are the main source of revenue for local govern- poorly maintained (in part because of low rents), the ments in other countries. Local income taxes, which are transportation system is very congested, and there are used e xtensively in the United States, might also be con- serious environmental problems. sidered. In addition, with the important exception of Developing appropriate financing mechanisms (includ- basic ,ocial services such as education and health (Chapter ing user fees) for public infrastructure and support ser- 10), the cost of providing urban services should as far as vices will be a major issue, particularly in rural towns. possible be recovered by charging fees to users and bene- Social labor-which rural households are still obliged to ficiaries. And to the extent that revenues from these vari- contribute-and the retained profits of local enterprises ous sources do not match assigned local expenditure re- are at present the main resources available. With the sponsibilities, arrangements for revenue sharing with development of the rural production responsibility sys- higher levels of government will need to be reconsidered, tem, the obligation to supply social labor is becoming as wiLl the possible powers of local governments to bor- increasingly inconvenient, and in some of the richer and row for capital expenditure. more rapidly growing rural areas can be avoided by pay- ing the village or town the daily wage equivalent. Im- proved systems of local taxation and pricing of services are needed, however, to strengthen the financial base of The -ural reforms and policy changes introduced in re- rural towns. cent years have already reduced rural poverty. Many of In county towns and larger urban areas, more formal the pDorest rural areas were among the first to introduce systems of resource mobilization already exist, including the production responsibility system and have benefited grants from higher administrative levels, surcharges on from rapid growth in crop production. Poor areas of industrial and commercial taxes, shares of industrial and eastern and southern China that were previously obliged commercial enterprise profits, as well as utility fees. But to grow foodgrains under unsuitable agroclimatic condi- the adequacy and appropriateness of these resources must tions have been able, in recent years, to switch to other be viewed in relation to likely future expenditure respon- crops (such as cotton), which produce much better re- sibilities. In many countries, efficient and equitable devel- opment of urban services has been seriously constrained either by fragmented and unclear allocation of responsi- 14. The average amount of urban housing per person is less bility among different administrative levels or by failure than ive square meters compared with twelve square meters in to ensure that revenue sources available to towns and rural areas. 91 Table S.10 Rural and Provincial Poverty, 1982 the Loess Plateau, where basic agricultural conditions are 99ercent) extremely difficult, and much further research and exten- National sion work may be necessary before suitable technical Type of area Jiangsu Hubei Gansu average packages can be developed and agricultural development Urban 0 2 1 stimulated (see Box 5.5 and Table 5.11). In the mean- Rural 3 3 41 13 time, cash incomes from agriculture could even fall, if Total 3 .. 35 11 there were decreases in output prices and increases in Note. Data are percentages of the population with per capita incomes below input prices to better reflect the economic costs of trans- Y 167 in urban areas and Y 140 in rural areas. port. The limited prospects for agricultural development Source, Annex 5, Appendix L will also constrain nonagriculural development, and im- provements in trade and transport will bring competition from outside that will threaten existing industrial activi- turns. Some poor rural areas, however, appear to have ties. There is also a danger that levels of human resource benefited much less, if at all; interprovincial variations in development in poor areas could fall: the adoption of fee- the incidence of poverty remain very large (see Table for-service medicine (following the introduction of the 5. 10); and there is increasing concern that the per capita production responsibility system) may lead to the relative output and income gap between the coast and the inte- neglect of preventive measures; and school enrollment rior could widen significantly. The Government would ratios and attendance may fall if parents have to pay more like to ensure that increased specialization and trade, and for, or see little advantage to, their children's education. changes in urban-rural balance, help to further reduce the Finally, a less direct role for the government in rural incidence of poverty and stimulate development in the commerce may make it more difficult to provide relief interior. But this will require both a detailed assessment grain to poor and disaster stricken areas. of prevailing conditions in poor areas and the develop- To ensure that poverty is reduced while national effi- ment of an overall strategy of systemic changes, invest- ciency is increased, all levels of government will need to ments, and other interventions. review very carefully both the prospects for poor locali- There are, of course, many circumstances under which ties and the measures that would most efficiently increase reforms and investments that facilitate specialization and the incomes of their populations. Of fundamental impor- trade will also contribute to poverty reduction. For exam- tance will be measures to improve the human resource ple, in some poor rural areas, investments in supporting base, including both health and education levels. China infrastructure for agriculture (such as irrigation and trans- now needs to extend to poor areas and regions the im- port) may show better returns than additional invest- provements in health that have occurred in most parts of ments in richer areas that already have relatively high yields. In many poor areas and regions, industrial devel- opment could be stimulated by the pricing and commer- cial reforms needed to enhance specialization and trade. Table 5.11 Rural Development in Dingxi County Changing the current practice whereby the state procures (Gansu) and Wuxi County (iangsu), 198 3 most agricultural products for processing in large state Measure Dingxi Wuxi enterprises would enable many poor areas to establish local agroprocessing industries. Further expansion and di- Rural per capita income (yuan) 108 412 versification of commerce, together with improvements in transport, would help some poor rural areas to com- Rural labor force pete better with state and collective enterprises in other Percentage in farming, forestry, or livestock 9 1 5 1 parts of the country. Reform of the material allocation Percentage in other activities 9 49 system should also help to ensure that mineral resources Arable land per agricultural worker (mu) 14 3 are no longer mined in the interior and allocated to Percentage irrigated 6 96 coastal provinces for processing when it would be more Percentage grain crops 79 87 efficient to process them locally. Finally, a system of eco- Grain yield (kilograms per mu) 85 400 nomic charges for land should help stimulate develop- ment in poor rural areas and in the interior, where land PermcapitanGviAo (yuan) 169 2,005 values will generally be much below national averages. Brigade and team industry 5 5 35 Other aspects of system reform and specialization, County and commune industry 46 1,113 however, may have a negative effect on poor areas and provinces, particularly those where the physical resource Primary school enrollment ratio (percent) 80 base is poor and the level of human resource develop- Hospital beds per thousand people 0.6 2.3 ment low. Dingxi County in Gansu, for example, is on Source Data provided to the economic mission. 92 Box 5.5 Agriculture on the Loess Plateau Chinese agriculture and civilization first developed on the cult topographical and agroclimatic conditions, but com- Loess Plateau,' but continuous farming over many millen- prehensive and reliable data necessary for such work are nia has totally changed the landscape and caused major often unavailable. More woody and forage species that are environmental problems. The original vegetation has been suited .o the varying conditions of rainfed areas need to be completely removed, and the loess (a dust of minute yel- found and developed, but worldwide searches for suitable lowish-grey grains that covers the entire plateau from materials and local programs of genetic selection and plant depths of a few meters to more than a hundred meters) breeding appear to have been very limited. Yields in both breaks down with alarming ease. As a result, hillsides, irrigated and dry land areas of the Loess Plateau could though extensively terraced, are affected by serious gully probably be increased by more appropriate fertilizer use, and sheet erosion problems. The region is subject to bit- but programs of field experimentation, soil testing, and terly cold winters, hot summers, and low rainfall (generally extension are still very weak. Perhaps the major constraint, below 400 millimeters per year). Cotton, wheat, and however, is a very serious shortage of the qualified techni- (more recently) corn are grown in irrigated areas, but these cal and professional personnel needed to develop research, account for only a small proportion of the total cultivated and extension programs that can form the basis of future area. In nonirrigated areas, millet and gaoliang are the agricultural development on the plateau. major crops, but yields even in a good year are very low. Programs of agricultural education, research, and exten- The incidence of rural poverty in these areas is as high as sion v ill take time to develop, so that improvements in anywhere in China, and most households survive only by agricultural and environmental conditions on the Loess having access to relief grain from the state. Plateau may only be realized in twenty years or more. An Efficient development of the plateau is likely to require overall strategy for raising incomes on the plateau should some investments in irrigation, if these can be shown to be take this into account. To achieve significant income in- economical, and, more important, reforestation and devel- creases in the meantime might require programs for outmi- opment of improved pastures in rainfed areas. But there gration. are many constraints to such developments. For example, irrigation projects in the Loess Plateau require especially 1. The plateau covers Shaanxi, parts of Gansu and Ningxia, west- careful design and appraisal because of the extremely diffi- ern Shaaxi, and westem Henan. the country. Mortality and morbidity due to infectious paren[s must be convinced by specific results that educa- diseases are frequently high in poor areas, substantial un- tion will lead to a socially and economically better life for dernutrition remains, and the level of health services is their children. In this regard, the contrast between Wuxi much below the national average. Overcoming these (Uiang3u) and Dingxi (Gansu) is striking. In Wuxi, parents problems will require careful analysis of prevailing health are determined to keep their children in school because problems in poor areas, development of plans to address the children need a basic education to get jobs in the those problems, and provision of the necessary man- rapidly growing rural industrial sector. In Dingxi, by con- power and financial resources to carry out such plans. trast, the rural industrial sector is not expanding, and Financing issues will be of particular importance, because parents see much less value in their children's education. of the collapse of the rural cooperative health insurance In many of the poorest rural areas, population levels systems in many poor areas and the need to ensure that are already high in relation to the quality, as well as the everyone can receive a minimum basic level of health care quantity, of natural resources and are increasing quite and that preventive activities (including health campaigns) rapidly (high birth rates more than offset relatively high are not neglected. death rates). For these areas, the Government needs to Primary and secondary school enrollment ratios are consider the potential role of outmigration in increasing unsatisfactorily low in many poor areas. This problem is income levels, both for those who move and for those of long standing, but has apparently been aggravated by who remain. International experience suggests that mi- the production responsibility system (Chapter 1): in a gration to urban areas and to other rural areas can play an village an hour's drive from the provincial capital of extremely important role in reducing poverty, providing Gansu, for example, the primary school enrollment ratio it is appropriately managed and does not involve move- in early 1985 was only 50 percent. There is thus an ment of too many of those with the best education and urgent need to improve both the quality of education and entrepreneurial talent (see Box 5.6). Although migration access to schools in poor areas (Chapter 8). In addition, has been more limited in China than in other countries, 93 Box S. 6 The Benefits andCotsofMigrontoCoammunnities of Origin The effects of migration on comniorigin app ear Experence from some developing countries, however, is to have varied greatly not lyamong ountries but also not as encouraging about the potential benefits of outmi- lwithin one country at different pe s, and even within gration for cromunities of origin. In particular, experience one country at the same time. Muchd s on the from northeast Brazil and certain regions of Mexico sug- characteristics of the comm a in sgests that the younger, more skiled, and better educated as on the scope, costs, and lts ofmirtion. W ei population groups tend to migrate, leaving behind an older very difficult to- generalize, so tantes emerge and les well-trained gro As a result, some communities from a comparison ofexperience in:different countries, of origin have lost the very resources that could stimulate From 1950 to 1 9h70,thestateoWsirin ithe their vitality anddelpme, and total (and even per United States experienced both a p percent (as outmigrants moreith a th e Clearly, the tent positive effects of outmigration of births over deaths) and a rate of eonomic and social oncommunities of origin are not realized automatically- development much in e s t ia r Pert adieed the ultimate impact of outmigration is likely to capita income went up 14 pert faste tanin the de very m on overal development policy. If out- United States as a whole school enrollments for youths migraton is one element in an overall strategy to raise the age sixteen to sevente ro 5 rt eathe l of human capial and labor productivity in poor ar- percentage of people em l irltrd ed by e p g ay gain greatly. But if appropriate 79 percent-9 percent ser t tenational aeg. It inve ntare not made in human and other capital in seems most unlikely t thepae fd t would prat may even suffer thrh outmigration. have been asgratit irants had not left. people were moved in the 1950s from densely populated higher in the interior than in coastal provinces as a result rural areas, such as Shandong, to the much less densely of efforts to establish uniform nationwide real wages in populated Northeast. Within Gansu, the possibility of the face of regional cost of living differences. Conse- moving people from Dingxi to Hexi, which has better quently, enterprises in the interior have little incentive to agricultural prospects, is being considered. Poor rural ar- employ more labor and face more difficulties in compet- eas are also being encouraged to form construction teams ing with enterprises on the coast. In the past year there that can work on a temporary basis in other parts of the have been indications that wage levels at the commune country, and the recent lifting of restrictions on the use of and brigade enterprise level are increasing more rapidly in hired labor may provide further opportunities for tempo- richer than in poorer areas (bonuses are much higher in rary migration from poorer to richer agricultural areas. Jiangsu than in Gansu, for instance). But the competitive- The Government also needs to assess the costs and ness of urban state and large collective enterprises in poor benefits of larger programs of permanent migration from areas also needs to be improved. One possibility would poor rural areas to urban areas or to richer rural areas, be to make changes in the current policy of national and to study how such programs might be implemented. For example, as labor market conditions tighten in more Table 5.12 Wage Levels, by Province and Type rapidly growing rural areas, and more and more people Unit, 12 move out of agriculture and into nonagricultural activi- (yuan per workerperyear) ties, national efficiency and equity could both be en- National hanced by some permanent rural-rural migration, from Type of unit Jiangsu Hubei Gansu average slowly growing to more rapidly growing areas. Even with some temporary or permanent migration of State-owned units' 748 760 936 836 unskilled labor out of poor rural areas and perhaps out of Collectively owned units unskilled labor out of poor rural areasmand perhaps out of in cities and towns' 626 657 649 671 the interior, surplus labor is stlfl likely to be greater, and Commune and brigade the economic cost of labor lower, in these areas for many enterprises 488 443 604 493 years ahead. Currently, however, large interregional dif- ferences in the economic cost of labor are not reflected in posation of the labor flrce and from interprovincial differences in the com- state and large collective enterprise wage levels, nor even of living. in commune and brigade enterprise wage levels (see Table Sourye State Statilstical BuHraud Sy and Fssheries Agicultural Yearbook 5.12). Indeed, in many cases wage levels may even be of China, 1983 (Beijing, 1983). 94 Map 5.2 Provincial Budget Expenditures and Revenues, 1 9 82 U . S. S. R . fi~~~~~~~~~~~~~~~~~~~ ~~~~~JAPAN A14GHANIST O OE --REPUBLIC9 KILOMETERS o 5~~~00 A NM 0 MILES 30 ~ AHI ~ AG#(J LIIEXPENDITURES EXCEED REVENUES 11ZREVENUES EXCEED EXPENDITURES iNKO,W( Province boundaries POL International boundaries ' /' DE &60(h chisa sea, PHILIPPNES THAILAND )RE SEPTEMBER 1985 uniform real wages for employees of these enterprises. Table 5.13 State Budget Revenues and Expenditures Alternatively, the policy could be maintained, but enter- in Wuxi and Dingxi Counties, 1983 prises in poorer areas could be provided with wage subsi- Measure Wuxi Dingxi dies. Efficient development of poorer areas and provinces Rural per capita income (yuan) 412 108 over the long term wil also require some major improve- Revenues per capita (yuan) 157.7 13.9 Expenditures per capita (yuan) 39.4 28.8 ments in the skilled manpower and technical resource Ratio of expenditures to base of such areas. Currently, for example, enterprises on revenues 0.25 2.07 the coast appear to be better managed and better able to Source. Data provided to the economic mission. take advantage of new technologies than enterprises in the interior. But this is more a consequence of manpower and technology policies than of location. There is a clear and even longer to have an effect. The effects of increased need to increase the availability of skilled labor in the internal trade and competition, however, are likely to be interior-through increases in the number of higher edu- felt much more quickly, with some activities being stimu- cation institutions located there and through increased lated while others are destroyed. Many of the existing incentives for skilled labor to work there. It may thus be activities in poorer areas could never be made economi- necessary to modify the wage system so that scarce skilled cally efficient and ought to be phased out, but others labor is paid more and surplus unskilled labor is paid less could in the long run be competitive, though requiring in the interior than on the coast. In addition, poorer areas some protection in the short and medium term. need to be given better access to technical developments, To protect local economic activities, as mentioned ear- both domestic and international. This could be accom- lier, poorer (as well as richer) provinces frequently pro- plished by providing more incentives for enterprises to hibit commercial organizations from purchasing goods cooperate with and diffuse technology to enterprises in produced elsewhere or force them to buy local products other provinces and by developing consulting and other that are inferior in quality and price. However, there may agencies that specialize in diffusion of technology (see be other ways of providing some temporary protection Chapter 7). More mobility of skilled labor across provin- for local activities at less cost in terms of national eco- cial boundaries would also help spread new ideas and nomic efficiency. Probably the most efficient measure technology. would be a straight subsidy for specific activities that The development of agriculture and industry and the could be phased out according to a prearranged schedule. improvements in transport, education, and other services Another possibility would be to impose taxes on goods that are essential to overall economic development and brought into a locality and to provide subsidies for goods income growth in poor areas and provinces will all re- leaving the locality, both on a temporary basis. If the tax quire substantial financial resources. Although some re- and subsidy rate were set at 10 percent of final product sources can and should be raised locally, large contribu- value, for example, all industrial activities in the interior tions from higher levels of government will also be with costs no more than 10 percent higher than else- essential. In recent years the revenue-sharing system ap- where would still find it financially profitable to expand pears to have been quite effective in transferring fiscal and even to export to other regions. Implementation of resources to poorer areas and evening out state budget such a system would, however, be administratively quite expenditures (see Map 5.2 and Table 5.13). But the pric- difficult, requiring all goods moving into and out of a ing system probably continues to transfer resources away locality to be checked and then taxed or subsidized. from poor areas (for example, because interior provinces The issue of the appropriate level or levels of govern- such as Gansu get a low price for their minerals). More- ment to be responsible for approving and implementing over, fiscal resources are only part of total financial re- such measures of local protection is also important. If sources: special funds outside the budget and funds chan- specialization and national efficiency are not to be seri- neled through banks and other financial intermediaries ously compromised, only the central government should are increasing in importance; the Government will need be vested with the power to approve subsidy or other to take steps to ensure an adequate net flow of nonbudge- protectionist measures, because only that level of govern- tary finance to poorer localities. ment can adequately assess potential tradeoffs between There are thus many measures that the Government national efficiency and local development. Moreover, could and should introduce that would help poorer locali- central government approval should be required for ties while at the same time facilitating specialization and measures to protect industries in specific counties as well national efficiency improvements. But some of them- as in specific provinces, since both affect national effi- such as improvements in education, expansion of the ciency. Appropriate lower levels of government could transport system, and development of agricultural re- still, however, be responsible for implementing protec- search and extension-will take many years to implement tionist measures. 96 International Economic Strategy In choosing among alternative international economic For the future, there are two large and related issues. strategies, China cannot avoid tension between the wish The first is how much, and in what directions, to further to guide and protect the character of internal develop- increase these contacts. The second is how best to man- ment and the wish to take advantage of the gains possible age such contacts, and in particular whether to establish a through contact with the external world. Greater in- truly 'open door" between China's economy and the volvement in the international economy could undoubt- rest o' the world, or whether to maintain the present edly raise efficiency and income in China. But it could "airlock" of administrative intermediation and separate also increase economic instability and regional disparities. price systems, which insulates the domestic economy A central theme of this chapter is that an economic sys- from the world outside. In both respects, the choice will tem that responds quickly and flexibly to changes in both depend partly on expected developments in the structure domestic and international conditions will be critical in and rnanagement of the world economy, and partly on realizing the benefits, as well as in minimizing the costs, internml economic objectives and reforms. of more trade and contacts with the outside world. The chapter begins by discussing the general issue of External Environment external economic contacts and management, in light of The general consensus at present is that over the next two expected international and domestic economic trends. It decades the industrial market economies, which still ac- then looks more specifically at China's foreign trade pros- count for about 60 percent of total world income and pects and policies, and at options for use of foreign capi- trade, will grow (in terms of real GDP) at a trend rate tal. It concludes with a discussion of the possible regional around 3.5 percent per year-slower than in the 1950s implications of a more open external orientation.' and 1960s, but somewhat faster than in the 1970s and early 1 980s. Inflation in these countries is also expected China in the World Economy to persist, but at lower rates (perhaps 5-6 percent per year) .han in the 1970s. Difficulties in adjusting the eco- After a long period of inward-directed development that nomin structure of those countries in the face of changing accentuated the normal tendency of all very large coun- economic and technological circumstances-particularly tries to export only a small fraction of national product the need to shift many industries to developing countries and import a correspondingly small fraction of national and to replace them with newer industries and services- supplies, China has begun in recent years to adopt a will also persist, and hence so will the tendencies toward much more open orientation toward the world economy. protectionism in these countries. This more open orientation is reflected in part in the Nonetheless, world trade is expected to grow faster rapid growth in foreign trade. Exports of goods and non- than production, implying further increases in economic factor services now account for 9-10 percent of GDP, a interdependence. Trade among industrial market coun- ratio above that of India and similar to large middle- and tries vwill still account for a large share (currently over 40 upper-income countries such as Brazil and the United percent) of total world trade, but developing countries States (see Figure 6.1). Inflows of direct foreign invest- will continue to increase their share, with particularly ment and foreign borrowing are also being actively en- rapid growth of manufactured exports (8-10 percent per couraged, and tourism, purchases of technology, overseas training of Chinese nationals, and other sorts of external contacts are increasing rapidly. 1. This chapter draws on Annex 4 and Background Paper 3. 97 Figure 6.1 Exports of Goods and Nonfactor Services uncertain.2 Moreover, around the projected trends there in Selected Countries, 1978-82 will undoubtedly continue to be cyclical fluctuations whose timing and amplitude are unpredictable. Percentage of GDP Such uncertainties and instabilities in the world econ- 40- South Korea omy should not discourage China from expanding its Hungary external economic contacts. China will always be a large and diverse economy, with a relatively small foreign trade 35 - /sector, which greatly reduces the possible damage that unexpectedly unfavorable external developments could inflict, relative to the more predictable costs of isolation. 30- However, international uncertainties should cause China to favor a flexible and responsive system for managing its . Romania , A.s external economic contacts. 25 -, Thailand Benefits and Costs of External Economic Contacts Indonesia To achieve rapid, sustained growth, China needs to intro- 20 ,, if ''''''' Industrial duce more modern technology (Chapter 7), but also- market economies more importantly-to increase the efficiency with which YugoslavX / all resources are used. External economic contacts are u s- lv Japan desired in China primarily as a source of new technology. But they can also contribute to greater efficiency. Indeed, --------Turkey if they do not, the transfer of technology in the broadest 10 United sense will itself be impaired. States Brazil Foreign trade can increase efficiency through specializa- ......... ____1 India tion and realization of economies of scale-concentrating sChina *^--' resources in low-cost production for export, and import- ing goods that can be purchased internationally more cheaply than they could be produced domestically. This 0 19 197 198 191 18 source of efficiency gains is less important for China than for smaller countries, because the domestic market is in many sectors large enough for several competing plants of Source rorid Bank, Wrld Development Report 1984 (Ne4 York Oxford optimal scale (although, as discussed in Chapter 5, small, high-cost, low-quality, protected local enterprises con- tinue to proliferate). China's potential gains from interna- tional specialization should not be underestimated, how- year). Partly as a result, production (real GDP) in develop- ever. Within agriculture, other countries with limited ing countries is expected to grow at an average trend rate cultivable land (the Netherlands being a prominent, if around 5 percent per year-faster than in the industrial small, example) have made themselves much better off by market economies, although this will be canceled out by exporting high-value, land-intensive products and import- faster population growth, with little narrowing of the ing in exchange lower-value products such as grain. large gap in per capita income. Some developing coun- Within industry, countries much larger than China (in tries, moreover, will grow much more slowly, because of economic terms) find it advantageous to engage in spe- persisting external debt difficulties or fundamental inter- cialization and exchange on a huge scale. nal problems, while developing countries in East Asia are There are many other ways in which increased external expected to continue growing more rapidly than the aver- economic contacts could be of benefit to China. Of par- age. ticular importance are the new products, new technol- These global projections should be treated with cau- ogy, new standards and ideas, as well as assistance in tion. Previous long-term projections have often been mastering them, that can come with a more open exter- wrong: in 1950, most economists envisaged recession in the following two decades, and there was a boom; in 1970 most economists projected the boom to continue 2. At present a multilateral system provides the institutional and iotd ecnotm ill oepend on rorm an contim framework (ruLes and regulations) for most transactions of indus- trial market economies and some transactions of developing coun- provement of the systems for managing the world econ- tries. A separate system (the Council for Mutual Economic Assis- omy, but the extent to which countries will be guided by tance) governs economic relations among East European multilateral or regional rules and procedures remains very nonmarket economies. 98 nal orientation. External trade can also provide some soften this tradeoff-minimizing the costs while securing much needed competition for domestic producers and many of the benefits. Different countries (Box 6.1) have thereby further stimulate improvements in domestic eco- opted for different management systems and for different nomic efficiency. sets of instruments, however, and China has to make its Increased external economic contacts could, however, own choice. The following paragraphs will focus on ways have some serious disadvantages in terms of China's in- of inclirectly regulating a decentralized external trade and ternal economic objectives. One potential disadvantage is finance system as a possible alternative to the present that immediate foreign competition could stifle infant administratively managed and relatively centralized sys- industries-activities in which China could produce eco- tem. This is not of course China's only option, but it is nomically, but only after a period of learning-or indus- the option that (in various forms) most other countries tries of strategic importance. Others include the possible have eventually chosen, and the one that seems most transmission into the domestic economy of primary consistent with the direction of reform of the system of product price volatility, general inflation, and fluctuations economic management within China. in world demand for industrial products; increased re- International experience strongly suggests that decen- gional inequalities; and the possible emergence of undesir- tralization to the enterprise level can be very important in ably large trade deficits or surpluses and unwanted in- helping realize the potential benefits of increased external flows and outflows of capital. economic contacts. Direct exposure of exporting firms to foreign buyers and competitors has proved elsewhere to A/lanaging an Open Economy be an extremely effective way not just of learning in the To achieve the benefits of increased external economic abstrt.ct about new and better products and processes, contacts it may be necessary to endure some of the costs. but also of learning how (and being put under pressure) But the establishment of a sound management system to introduce them in practice (see Chapter 7). Similarly, and skillful use of appropriate policy instruments can greater freedom for enterprises-including commercial Box 6.1 Contrasting Types of International Economic Strategy The table below (drawn from Background Paper 3) sum- detailed differences yielding widely varying results for marizes five orientations with country examples. The cate- countnies with the same general orientation. gories are not narrowly defined; each spans a multitude of Box Table 6.1 A International Economic Strategies Strategy Examples 1. Highly open economies with little protection against imports and few restraints on foreign investment. a. with little government intervention to limit poverty or guide investment Chile, 1973-82 b. with considerable domestic intervention for social objectives Singapore from 1967 2. Relatively open economies, emphasizing export promotion, particularly for industrial products, but Japan with considerable use of protection and government control of the economy Republic of Korea from 1962 Brazil from 1964 Yugoslavia from 1964 Hungary from 1968 3. Regional groupings with relatively free trade for industrial products inside tc e region, but with pro- European Economic tection of agriculture in each country and moderate protection from outside manufactures Communities 4. Highly protected economies without strong orientation coward exports, but with attempts to pro- Brazil to 1964 mote industrialization and some foreign investment Columbia to 1967 India 5. Relatively closed economies with tight restrictions on trade and investment Burma from 1962 Sri Lanka to 1976 China to 1977 U.S.S.R. Saurce. Background Paper 3. 99 enterprises-and consumers to choose directly between tent where the world market for a product is limited- imported and domestically produced goods could greatly rice and many of China's other traditional agricultural increase competitive pressure on Chinese producers to exports being examples (see Chapter 3). It may also be introduce new, better, and cheaper products. Direct con- appropriate to prevent short-term fluctuations in the tacts between domestic and foreign enterprises would world prices of primary commodities from influencing also facilitate the process of specialization, much of which the domestic economy, by interposing a system of stabi- takes place within finely classified product categories. lizing taxes (when prices are low) and subsidies (when Decentralization of decisionmaking would make it im- they are high) on exports and imports-although it is possible for the government to predict or control the often difficult to distinguish fluctuations from trends. actions of individual enterprises or the details of external Many countries, including such economic successes as transactions. To ensure that these actions were in accor- Japan and South Korea, as well as some conspicuously dance with China's overall economic interests, and unsuccessful ones, have supplemented indirect regulation broadly consonant with government strategy, internal re- through taxes and subsidies with direct quantitative con- forms to make enterprises more sensitive to costs, to trols on imports and exports. Explicit upper limits (in- customer requirements, and to profits and losses would cluding outright prohibitions) have frequently been im- therefore be essential, as would an appropriate system of posed on imports of particular products, while the import prices and other indirect levers. Of particular importance of others has been deliberately obstructed by government is the relationship between domestic and world prices, and state industry procurement practices, product safety which greatly affects the economic rationality of decen- laws, complex administrative procedures, and so on. Up- tralized export and import decisions. In many respects, per limits have sometimes also been imposed on exports, the more direct the linkage the better, since external and in some cases enterprises have been given formal or prices, mediated through an appropriate exchange rate informal export quotas (a crude but effective method of (discussed later), can provide appropriate signals to Chi- offsetting the bias noted above). nese producers and consumers about the value (for ex- China should thus not necessarily eschew direct quanti- ports) or cost (for imports) of particular goods to China in tative controls on foreign trade, but should use them world markets. They provide incentives to produce more sparingly. They are almost always a less economically (and use less) of items that can be exported at a profit to desirable means of achieving a given result than equiva- the economy or that it would be unnecessarily costly to lent taxes and subsidies (for example, because some ex- import; and to produce less (and use more) of items that ports under quotas are needlessly costly-either to pro- could be exported only at a loss or could be more cheaply duce or to withdraw from domestic consumption-and imported. Such direct linkages could also contribute in because some imports under quotas are either ineffici- China to a much-needed widening of price differentials ently allocated or resold at a high profit). Yet because they between low- and high-quality products, without which can be more directly and precisely applied, they can enterprise incentives to innovate and improve quality will sometimes be a useful means of implementing a strat- remain small. egy-such as building up a particular industry or obtain- Certain modifications to direct price linkages may also ing a foothold in a particular external market-that is be valuable, however (even when domestic prices and itself economically sound, in the sense that it would ulti- costs are economically rational, as is not at present the mately be viable without these controls. In most cases case in China-see below). Import tariffs are a widely elsewhere, however, quantitative controls on trade (as used and quite efficient means of providing infant or well as taxes and subsidies) have been applied without strategic industries with a certain degree of protection any such strategy. from foreign competition-though it is easy to miscalcu- A common reason for indiscriminate use of quantita- late the amount of protection provided.3 Their main dis- tive controls, as of tariffs, has been inappropriate ex- advantage is that they discourage infant industries from change rate policy, and in particular maintaining too high exporting, thus slowing their technological advance, un- an exchange rate (in terms of dollars per unit of domestic less complemented with export subsidies or replaced with production subsidies (both of which, however, are liable to provoke retaliation by foreign countries, unless consis- tent with international rules-which permit indirect tax 3. Even an apparently small nominal tariff can often provide a rebates). This bias against exports has in many countries high degree of effective protection to domestic producers, mainly been compounded, and other serious economic distor- because value added and profits are only a fraction of gross output. tions introduced, by indiscriminate use of high tFor example, if the ratio of material inputs to output (at world tlons mrde,ymtuefh faa prices) were 70 percent, a 10 percent tariff on output. with no substitute for exchange rate adjustments. Taxes on some tariffs on inputs, would provide an effective rate of protection of exports could also be used to discourage Chinese ex- 33 percent (by increasing value added per unit of output at domes- porters from driving down prices to an unprofitable ex- tic prices by this proportion). 100 currency), which encourages excessive imports, by mak- sudde.1 but enduring change in the price of a major com- ing them cheap, and discourages exports, by making modity (such as oil) or in the openness of a major market them unprofitable. By contrast, countries that have main- (such as might be caused by war or political realignment). tamned a competitive or even undervalued exchange rate Changes in the desired long-run level of net capital inflow have been able to intervene much more selectively and could also require alteration of the exchange rate. purposively and have not suffered from the chronic short- Temporary fluctuations in global economic activity are ages of foreign exchange that have plagued most develop- an additional source of problems, though one that is ing countries. However, too low an exchange rate can almost as serious with an administrative airlock as with create its own problems, including unwanted trade sur- an open door. They cause export revenues to fluctuate pluses, inflationary pressures, and discouraging the pro- (often requiring offsetting oscillations in reserves or exter- duction of useful nontraded goods and services. nal borrowing), which in turn can cause internal fluctua- Especially following the recent unification of foreign tions in economic activity, especially by affecting produc- exchange arrangements (the internal settlement rate was tion in export-oriented manufacturing. In principle, these abolished at the beginning of 1985), and given the cur- fluctuations can be damped by countercyclical changes in rent account surpluses of recent years, large reserves, and domemtic fiscal and credit policies (which, however, may the ready availability of external capital, China appears aggravate the fluctuations in the foreign trade balance) or extraordinarily well placed to move toward a selectively in the exchange rate and trade incentive system. In prac- and mainly indirectly regulated open door system, with- tice in other countries, this has proved hard to do on out the problems (of external insolvency or internal infla- more than a limited scale, at any rate without sacrificing tion or deflation) that many other countries attempting a longer-term objectives. Less stability in the domestic similar move havsimimultaneously had to contend with. econcmy may therefore be a concomitant of greater con- To do this immediately, at one stroke, would probably be tact through trade with the world economy. This is a unwise: most Chinese enterprises are not yet appropri- much less significant problem for China than for smaller ately motivated and price sensitive; even if they were, countries, but particular industries and localities could there are still distortions in domestic prices, taxes, wages, experience quite sharp swings in prosperity. and interest rates that could lead them in economically undesirable directions. But devolution of export and im- port decisions to enterprises could proceed in parallel Foreign Trade Prospects and Policies with internal economic reforms. Although the size of the foreign sector in relation to During this transitional period, and subsequently, the national income has increased considerably in most coun- suitability of the exchange rate should be carefully moni- tries over the past two or three decades, it tends to be no tored and necessary adjustments made. This is partly be- greater in large high-income countries than in large low- cause there is no way of telling in advance whether the inconie countries (in small countries, it is significantly present exchange rate would, following internal reforms greater at higher income levels). China's trade ratios have and dismantling of the external airlock, generate approxi- recen.ly risen to levels well within the 6-15 percent mate balance between imports, on the one hand, and range of other very large countries. In the macroeco- exports plus desired capital inflow, on the other. It might nomic- projections introduced in Chapter 2, these ratios turn out to be too high-with an excessive trade deficit- increase only modestly in the next two decades (see Table or too low-with too little net capital inflow or a trade 6.1), but the model used to make the projections is not surplus. It is also because China's optimal exchange rate well suited to addressing foreign trade issues,4 and would not usually remain constant over time, for two larger-or smaller-increases might turn out to be possi- reasons. One is differences between internal and external ble or desirable. inflation rates: to the extent that there are differences in China's present composition of foreign trade (see Fig- these inflation rates, the (nominal) exchange rate would ure 6 2) is in many respects quite similar to those of other need-other things being equal-to be changed regularly low-income economies, specializing in primary product by the difference between the average world inflation and textile exports to industrial countries and in such rate and the internal inflation rate. The other reason is fundamental changes in economic circumstances, which may gradually or occasionally alter the real exchange rate needed to attain any desired long- 4. In primary sectors, as explained in Chapter 3 (agriculture) run level of capital inflow (with reasonably full utilization and Annex 3 (energy), as well as in Annex 4, the trade balance is of domestic productive capacity). Possible examples of just a esidual-indicative of tendencies and problems to be solved, rather tha-n of likely actual outcomes. In other sectors, there are no such changes would be a steady improvement in China's explicit efficiency-seeking or learning mechanisms, and the results technological level relative to other countries, including are governed largely by assumptions about sectoral self-sufficiency, developing country competitors in export markets; or a openness and export composition. 101 Table 6.1 Total Foreign Trade Ratios, Alternative search on new forms of energy has been initiated. (This Projections, 1981-2000 does not mean, however, that energy prices will necessar- (8ercentage of GDP) ily remain stable or fall over the next two decades). In 2000 view of the relatively low cost of exploiting its huge coal Measure of trade 1981 Quadriple Moderate Balance reserves, China is likely to find it economically advanta- geous to remain basically self-sufficient in coal and petro- Exports 8.4 11.0 10.6 10.3 leum, though some two-way trade in crude oil and petro- Foreign trade deficit -0.3 0.4 0.3 014 leum products may be attractive (see Chapter 4). In the case of agricultural raw materials, better and Source. World Bank projections: Annex 4. 'wder application of existing technology and scientific discoveries in the area of biotechnology are expected to result both in greatly increased world production capacity services as tourism, which are exchanged for machinery and in continued falls in the real prices of most agricul- and transport equipment and other manufactured goods tural products. In the medium term, China's comparative and services. However, China, like India, has a high per- centage of exports of manufactured goods compared with other low-income economies and is highly self-sufficient in machinery. China is also not an oil-importing country. 1978-8 3 In the following paragraphs, some future possibilities and options for the growth and composition of China's foreign trade are reviewed, together with related policy PercentageofGOP Exports measures and reforms. The quantitative projections for specific products and groups of products that are men- tioned should be regarded as subject to a wide margin of 8 - error. Of much greater importance will be the capacity of China's system of economic management to generate an 6l efficient overall level and pattern of trade in the face of changing internal and external circumstances. Primary Products 4 China is currently a net exporter of primary products (a trade surplus on energy more than offsets a trade deficit 2 - on other primary products). Whether China should re- main a net primary products exporter will depend to a large extent on domestic cost and international price de- 0 velopments. Primary product prices on average, and in real terms, have been falling for at least the past fifty 2 - years, since developments in production technology have more than kept pace with the growth in demand; most recent studies of the world economy expect this trend to 4 _ continue. The markets for primary products will, how- ever, continue to be characterized by great price variabil- ity, instability, and uncertainty. 6 - In the case of energy, for example, most studies of the world economy now predict that the resource and tech- 8 nological capacity eithcr cxists, or is likely to be dcvel- Impotts oped, to meet expected growth in world demand. Incre- mental advances in technology could soon make wc l I exploitation of nontraditional oil and gas resources (such 1978 1979 1980 1981 1982 1983 as tar sands, bituminous shales, and frozen natural gas) * Nonfactor services economical, thereby extending use of oil and gas as major E] Manufactures energy resources well into the next century. World coal Primary products reserves are much greater than oil and gas reserves, and coal usage is likely to increase as oil and gas usage de- clines. Use of nuclear energy will also increase, and re- Source Bank of China, International Monetary Fund, and World Bank esti- mates. 102 advantage may still lie in expanding exports of some agri- Table 6.2 Share of Manufactures in Total cultural products such as rice (even if this contributes to a Merchandise Exports, 1960 and 1981 fall in world prices), in importing other agricultural prod- (percent) ucts including coarse grains, and in achieving an overall Macbinery and surplus on agricultural trade (see Chapter 3). But China is transpont not well endowed with agricultural resources. Thus in Country if Allmanufactures equipment the longer term, as domestic costs increase and the real origin 1960 1981 1960 1981 prices of agricultural products on world markets continue United States 63 70 35 44 to fall, China's comparative advantage may lie in becom- Germany, Fed. Rep. 87 86 44 45 ing a net importer of agricultural products. France 73 73 25 34 In an effort to make primary products trade more effi- Japan 79 97 23 57 cient and responsive to international as well as domestic Yugoslavia 37 79 15 29 economic developments, the Government has recently Uruguay 29 30 (.) 2 Hungary ~6 6 6 5 3 8 3 1 initiated a further stage of management decentralization. Hungary 4 6 3 1 Initially, management of primary products trade was de- Chile 4 10 0 2 centralized from the Ministry of Foreign Economic Rela- South Korea 14 90 (.) 22 tions and Trade (MOFERT) to national foreign trade corpo- Thailand] 2 27 0 S rations. Some of these corporations have proved to be Sri LanLa () 21 0 () very efficient. TIhe Chinese national food trading agency, China . 5 3 5 CEROIL, for example, is among the most astute grain buy- India 45 59 1 8 ing and selling operators in the world and has saved Burma I .. 0 China substantial amounts of foreign exchange. How- Source: Vorld Bank. World Development Report 1984 (New York: Oxford ever, other national trading corporations in China are Universiiy Press, 1984), pp. 236-37. much less efficient. The Government has therefore de- cided to decentralize trade in many primary products to local and often more specialized corporations. already beginning to be affected by market constraints for This will make it all the more necessary to develop an certain types of manufactured exports, and the rapid pace incentives system that provides appropriate signals to pro- of change in both domestic and international industry ducers and consumers, as well as importers and exporters, makes it difficult to forecast future trading patterns. as to the level and pattern of production and consump- One major issue for China is how open the manufac- tion and hence the level and pattern of primary products turing sector should become in terms of the ratios of trade. Internal pricing policies (see the discussion of agri- exports and imports to output and demand. The higher culture and energy in Chapters 3 and 4) will be especially these .atios, the more opportunities there will be for important. The Government will also need to closely domestic enterprises to participate in and benefit from monitor current and prospective world prices for key involvement in foreign trade. But higher ratios may also primary products and provide market information to increase the potential for world market uncertainties to prospective buyers and sellers. In addition, it will need to affect the domestic economy. In the macroeconomic pro- consider carefully the implications of alternative levels jections-whose limitations in respect of foreign trade and patterns of primary products trade for investments in were rioted earlier -it is assumed that the Government infrastructure and other support services, many of which aims for higher sectoral export and import shares in BAL- may be both large and product-specific. ANCE than in QUADRUPLE and MODERATE (see Table 6.3).i In QU NDRUPLE, however, even maintaining a ratio of Manufactures manufactured imports to domestic manufacturing de- In both large and small countries, rising per capita in- mand of 6 percent requires a significant increase in the comes are usually associated with a rising share of manu- ratio cf manufactured exports to domestic production in factures in total exports (see Table 6.2). Moreover, trade order co offset the projected net deficit on primary prod- in manufactures has usually been more effective than ucts trade (see Tables 6.4 and 6.5). In MODERAFE, in primary products trade in fostering diversification of pro- contrast, continued surpluses on primary products trade duction, in promoting new technology and ideas, and in would reduce the required increase in the ratio of manu- stimulating improved efficiency in the domestic econ- factured exports to output. And in BALANCE, also with a omy. In recent years, China has been able to rapidly expand exports of textiles and other manufactures despite the slow growth of the world economy, and trade sur- 5. The higher target shares are in metallurgy, chemicals, ma- pluses on these types of manufactures have partially offset chinery and other manufacturing; target export and import shares trade deficits on machinery and equipment. But China is in other sectors remain unchanged. 103 continuing primary products surplus, the objective of a sition of its manufactured exports very significantly over higher ratio of manufactured imports to domestic de- the next two decades. The textiles and clothing sector, for mand could be satisfied with the same ratio of manufac- example, has provided many countries with the first step tured exports to output as in QUADRUPLE. The growth on the ladder to export-oriented industrialization and has rates of manufactured exports in all three projections are been a key element in China's recent rapid growth of quite substantial, ranging from 7.3 percent per year in manufactured exports.7 As a result, China now accounts MODERATE to 10.1 percent in QUADRUPLE, but compara- for 5 percent of world trade in textiles and clothing and ble with the World Bank's projections of worldwide de- for a much larger share of world trade in certain (gener- veloping country growth in manufactured exports of ally lower-quality) textile products. Although it is difficult 7.5-9.7 percent per year for 1985-95.6 to predict future technological developments, China's Whatever the overall growth rate of manufactured ex- cotton production potential (see Annex 2) and its rela- ports, it seems that China will have to change the compo- tively low economic cost of labor are likely to give it a clear comparative advantage in textile and clothing ex- ports for many years to come. China could therefore take Table 6.3 Sectoral Foreign Trade Ratios, Alternative over a much larger share of the world market, if allowed Projections, 1981-2000 to do so. However, expansion of Chinese textile exports is already being curtailed by voluntary export restraint Measure agreements with the United States and member countries of trade 1981 Quadruple Moderate Balance of the European Economic Community and by the pro- Exports (percentage visions of the Multifibre Arrangement. There is still con- of gross output) siderable scope for boosting textile and clothing export Primary products 2 6 2.3 3.4 3.3 earnings by upgrading product quality, and China slhould Manufactures 6.0 8.3 6.5 8.3 also continue to press in international forums for relaxa- Machinerv 3.8 10.7 8.2 8.9 Textiles' 10.1 8.3 6.5 8.2 tion of the restraints on trade. But China's textile and Other 4.7 6.4 5.2 7.8 clothing exports will probably grow by no more than 6 percent per year (in real terms), and their share in total Imports (percentage manufactured exports will decline. demand) To offset the likely slow growth in textile exports, Primary products 2.2 3.9 3.5 3.0 China should more rapidly expand exports of other man- Manufactures 6.7 6.0 6.1 8.0 ufactures, including machinery and metal products, Machinery 8.3 8.0 8.0 12.0 'rextiles' 6.2 2.0 2.0 2.0 Other 5.9 6.4 6.8 7.9 6. See World Bank, World Development Report 1984 (New Note. Data are calculated at constant (i1981) domestic prices. York: Oxford University Press, 1984), p 35. a. In this and subsequent tables, textiles includes clothing. 7. Includes Standard International Trade Classification catego- Source: World Bank projections; Annex 4. ries 65 and 84. Table 6.4 Average Annual Growth of Foreign Trade, Alternative Projections, 1981-2000 (percent) Quadruple Moderate Balance Constant Current Constant Current Constant Current Measure of trade yuan dollars yuan dollars yuan dollars Exports Primary products 5.3 9.1 6.4 11.3 7.1 11.4 Manufactures 10.1 16.4 7.3 13.6 8.8 15.2 Machinery 15.2 21.6 12.1 18.4 12.8 19.1 Textiles 6.1 12.0 3.4 9.1 4.9 10.8 Other 9.8 16.3 7.4 13.8 9.6 16.3 Imports Primary products 9.3 t 5.2 7.4 14.4 7.4 14.2 Manufactures 7.3 14.4 6.2 12.6 7.9 14.2 Machinery 8.4 14.5 7.1 13.1 9.9 16.0 Textiles 0.8 6.5 -0.3 5.2 -0.2 5.4 Other 8.5 16.9 7.6 14.6 8.3 15.2 Source. World Bank projections; Annex 4. 104 Table 6.5 Structure of Foreign Trade, Alternative Figure 6.3 Direction of External Trade, 198 3 Projections, 1981-2000 Measure 2000 Billions ol U.S. dollars oftrade 1981 Quadruple Moderate Balance 8 _ er Agriculture Textiles Machinery Other Exports and other and manufactures Primary products 44.6 19.1 35.4 29.8 primary equipment Manufactures 55.4 80.9 64.6 70.2 6 products Machinery 10.5 35.0 26,9 25.0 Textiles 25.0 17.5 13.7 15.1 Other 19.9 28.4 24.0 30.1 4 - Imports Exports Primary products 24.4 38.3 32.6 26.3 Manufactures 75.6 70.3 74.6 81.7 2 - Machinery 29.4 31.4 32.0 42.7 Textiles 19.2 5.2 5.3 4.5 Other 27.0 33.7 37.3 34.5 _ Note. Data are percentages of total merchandise exports in current dollars. Source. World Bank projections; Annex 4. 2 which currently account for less than 20 percent of man- ufactured exports and which in many countries have 4 - | Imports played a key role in overall export growth. Developing countries currently absorb 85 percent of China's machin- ery exports (see Figure 6.3) and will probably remain an 6 - important market. Substantial increases in machinery ex- ports to industrial countries should also be possible. This, however, would involve manufacture of parts and com- 8 - ponents, and in some cases assembly of imported compo- nents, as much as sales of complete equipment wholly manufactured in China. An extensive and flexible net- It l Developingeconomies work of subcontracting and other cooperative agree- ments between Chinese and foreign enterprises would I Industrial market economies therefore be vital, but would be hard to establish through 12 * Eastern Europe China's present airlock of administrative intermediation. (East European exporters of machinery and equipment, Source \4inistry of Foreign Economic Relations and Trade; General subject to a similar airlock, have done much less well than Administation of Customs. the newly industrializing countries of Asia and Latin America-Background Paper 7.) Both within the machinery category, and in the wide tailed composition should change continuously as China and promising range of other nontextile manufactured learns to produce particular goods domestically, while exports, sales of high-value consumer goods to industrial constaltly adding new imports to provide the basis for and upper-middle-income countries constitute an espe- future learning. In recent years, China has placed most cially attractive market. But competition from other de- emphasis on intermediate and capital goods imports to veloping countries will be intense, and improvements in ease eaergy and other domestic supply constraints, to the quality of Chinese products essential. Experience else- stimulate technological development and improve effi- where suggests that quality improvements could be as- ciency in domestic industries. In metallurgy, for example, sisted by allowing Chinese exporters to import freely if domestic steel supplies have been increasingly supple- materials, components, and equipment available from do- mented by imports, which in 1983 accounted for 24 mestic suppliers are not up to international standards. percent of steel product demand. Much the same has Chinese enterprises will also need to develop the capacity happened in chemicals, and machinery imports have re- to meet individual buyers' requirements and to ship prod- sumed their upward trend. Rapid expansion of such in- ucts that are packaged and labeled, ready for sale. termediate and capital goods imports is likely to con- Rapid growth in manufactured exports will make possi- tinue, and quite large deficits in energy-intensive ble rapid growth of manufactured imports, whose de- metallurgy and chemicals subsectors may persist. 105 China's strategy with regard to consumer goods im- present practice of setting prices for manufactured goods ports still appears to be influenced more by consider- relatively high, as a means of generating large enterprise ations of self-sufficiency than of economic efficiency and profits and hence large profit tax and remittance revenues technological progress. Despite recent relaxation in some for the budget. These high domestic prices reduce the sectors where demand has far outstripped production ca- incentive for Chinese enterprises to export. They could pacity, imports of most consumer goods remain strictly be offset by subsidies for manufactured exports, but such controlled. This helps protect domestic consumer goods subsidies may result in charges of "dumping" by other industries. It also has the possible social objective of limit- countries and may make it more difficult for China to ing the availability of expensive goods that only a few participate fully in such multilateral trading arrangements consumers can afford. However, international experience as the General Agreement on Tariffs and Trade. This strongly suggests that a relatively closed orientation to- provides an additional reason for changes in China's price ward consumer goods imports is likely to result in the and fiscal systems. Among other things, movement away development of a highly inefficient domestic consumer from profit taxes and remittances and toward indirect goods industry, and that the competition and new tech- taxes (including a value added tax-see Chapter 9) would nology that come with a more open import orientation make it easier to address dumping charges, because re- may be just as important in consumer goods as in other bates of indirect taxes are generally more acceptable un- industrial sectors. der international rules than export subsidies. If China is to expand its trade in manufactures rapidly and efficiently and reap the full benefits from such trade, Services domestic firms will need to be allowed more freedom to In recent years, world trade in services has been increas- choose between domestic and foreign suppliers and more ing rapidly and in some-mostly industrial-countries, direct contacts with foreign markets and firms. A much service exports are now substantial relative to merchan- wider and more competitive range of foreign trade enter- dise exports and GDP They include activities ranging from prises-probably including collective and individual en- transport and insurance services, which are often closely terprises-will also be needed to provide specialized com- related to merchandise trade, to banking and financial mercial and other services. These changes will be services, which are closely related to investment and bor- important not only in enabling Chinese enterprises to rowing policies.8 But changes in the structure of the compete effectively in external markets but also in ensur- world economy (including the increased importance of ing that there is a real transfer of technology and other the information and service sectors) are expected to result benefits to China. In recent years the Government has in continued rapid growth of trade in services. tried to grant enterprises greater autonomy and has en- At present China's main service receipts are transport couraged foreign trade corporations to act more as agents and tourism. Chinese enterprises also have overseas con- than as principals in external transactions. But the pattern struction contracts, especially for projects in the Middle of exports and imports still appears to be determined East, and payments for labor services have increased more by administrative decisions at the industrial bureau sharply. In South Korea this type of service export has or foreign trade corporation level than by incentives at been very important for the overall growth of exports, the level of the individual enterprise, and direct contacts leading to other types of service export, such as licensing between Chinese and foreign enterprises remain quite and technical agreements and consulting services, and also limited. helping stimulate exports of machinery and equipment, It is important to reiterate that effective decentraliza- many of which are directly related to construction and tion of manufactured goods trade to the enterprise level other contracts. must be accompanied by and not precede the overall China could benefit from substantial further increases process of enterprise reform. At present many enterprises in imports of technical information and other services in China still appear to care relatively little about the from industrial countries. In industries in which China quality and cost of inputs and about whether or not they already has substantial technological capability, licensing are able to sell their output. In such circumstances the or purchase of technological assistance could help fill the development of direct contacts between local and foreign gaps in local knowledge and capabilities and could up- firms could easily result in inappropriate trade patterns and even in serious trade and balance of payments diffi- culties for China. 8. The principal services involved in international trade are: As the management of manufactured goods trade be- banking and financing services; insurance; freight transport and port handling; passenger transportation and tourism; architecture, gins to be decentralized to autonomous and price sensi- construction and engineering, repairs and maintenance; films, com- tive enterprises, the Government will need to consider munications, data processing, printing, professional services (tech- further the use of price, tax, exchange rate, and other nical, health, education, legal, accounting); and other services (fran- regulatory levers. One major issue concerns China's chising, leasing). 106 grade technology quickly. China has increased its use of Figure 6.4 External Balance, 1978-84 licenses and other disembodied technology sharply in the last few years, but still spends relatively little on these Billions of U.S. dollars items. Licensing often covers only older technology and 6 may carry restrictions on use, but it could help overcome critical shortcomings in domestic capability (for example, components for machine tools). Advanced countries also _ find licensing an effective means of upgrading technology: the United States and Japan, for example, conduct a large 4 - iupu mutual trade in licenses. Surplus 3 - External Balance and Finance Many developing countries, as well as many industrial 2 - countries at earlier stages of development, have used for- eign trade deficits and net capital inflows as a means of maintaining a higher long-run rate of investment than could be sustained using domestic savings alone.9 But such a strategy is economically beneficial only if the pro- o ductivity of the additional investment, at the macroeco- nomic as well as the project level, exceeds the cost of foreign capital. China's current account surpluses of the early 1980s (see Figure 6.4) may therefore have been an appropriate economic response to a situation in which 2 international interest rates were very high in real terms Deficit and the capacity of the domestic economy to absorb 3 - imports efficiently was limited by the lack of sound feasi- bility studies for major investment projects. The Government's longer-term policy, however, is to 4 7 1 8 1 i5 78 1979 1980 1981 1982 198 3 1984' incur moderate foreign trade deficits, with increased but cautious reliance on external finance. Accordingly, in the -Currnnt account balance macroeconomic projections it has been assumed that for- El Invisibles balance eign borrowing adjusts gradually upward toward a ceiling * Trade balance imposed by a 75 percent ratio of debt to exports.'0 How- ever, the ceiling is not reached within the projected pe- a. Preliminiary estimate. iod (see Table 6.6). In all three projections, trade and Source Baank of China, International Monetary Fund, and World Bank esti- riod (se Tabl 6.6).In allthree pojectins, trde and mates. current account surpluses change to trade and current account deficits, but the deficits in relation to GDP remain at or below I percent and fall to 0.7-0.8 percent by Table 6.6 Foreign Trade Balance and Debt, 2000. The projected levels of debt in 2000 appear large Alternative Projections, 1981-2000 at current prices, but at 1981 prices they would be only (billions of dollars) one-third as large (that is, $56 billion in QUADRUPLE, $46 2000 billion in MODERATE, and $ 54 billion in BALANCE). Mo Measure 1981 Quladruple A4oderate Balance Exportsf.o.b. 22.5 279.0 216.6 261.8 Importsc.i.f. 21.7 289.6 223.0 271.5 Trade balance +0.8 -10.6 -6.4 -9.7 9. In the nineteenth century net capital inflows to the United Net interest -0.1 -11.1 -9.2 -10.7 States hovered around I percent of GNP and reached 7-8 percent Net transfers and of GNP in Canada, Australia, and some other now developed other services 0.6 1.6 1.6 1.6 countries. World Development Report 1985 surveys the recent as Current account well as historical experience of many countries in using external balance +1.2 -20.1 -14.0 -18.9 finance. It also analyzes current problems and future prospects and Net debt a 0.9 158.3 128.4 151.9 options in this area. 10. If an average interest rate of 10 percent per year and an Note: Data are at current prices, assuming international inflation at an average of 5.6 percent per year. average repayment period of ten years is assumed, this corresponds a. Debt outstanding and disbursed, less foreign exchange reserves. approximately to a 15 percent debt service ratio. Source World Bank projections; Annex 4. 107 The appropriate future level of China's foreign trade appropriate for China, and foreign investors may operate deficit and net capital inflow will depend in part on the as an enclave within the domestic economy rather than as availability and cost of foreign capital. In the 1 960s and an integral part of it. Judicious policy intervention is thus 1970s, capital flows to developing countries at moderate needed to induce foreign investors to link their activities real interest rates increased significantly and in many to those of local enterprises. For example, many coun- countries played an important role in facilitating struc- tries have used domestic content legislation to force auto- tural adjustment and in accelerating economic growth. mobile companies to make greater use of domestic suppli- Since 1981, however, the supply of foreign finance has ers, and component manufacturers have thereby been declined, real interest rates have increased greatly, ' and assisted in upgrading product quality and increasing effi- the dollar (in which most developing country debt is ciency. Direct foreign investment can also aggravate some denominated) has appreciated significantly against other of the problems caused by domestic price distortions, currencies, causing serious debt servicing problems for including those associated with inappropriately high pro- many countries. These circumstances make it difficult to tective tariffs or quotas: these can allow foreign as well as predict the future supply and cost of foreign capital. But a domestic enterprises to earn high profits from activities significant reduction in current high real interest rates is that are economically inefficient (sometimes even with likely only if industrial countries can reduce their pres- negative value added at world prices, or a net loss of ently large budget deficits. foreign exchange). The appropriate size of China's foreign trade deficit Although the appropriate solution clearly lies in inter- and net capital inflows will also be affected by the ongo- nal price reform rather than restriction of foreign invest- ing process of economic reform aimed at improving the ment, official fear of problems of this kind seems to have efficiency of investment. It would not be in China's eco- impeded direct foreign investment in China. In addition, nomic interests to borrow at real interest rates of 5 per- and despite recent improvements, the number of admin- cent to undertake a project whose economic rate of re- istrative hurdles that prospective foreign investors must rurn was only 2 percent-indeed this would actually surmount is unnecessarily large. In this context, the re- make China worse off. And there are many examples of cent widening of the officially permitted scope for foreign countries (in Eastern Europe, for example) that were investment-now covering far more geographical areas, once, like China today, very cautious in their foreign as well as a broader range of consumer goods and ser- borrowing programs, but then moved too quickly to vices-seems wise. This is because, in China's circum- more active use of foreign capital and now face balance of stances, foreign investment is desirable not only for the payments and debt service problems. Careful control and foreign exchange and advanced technology that it may monitoring of China's increased use of external finance bring, but also for the demonstration effect of modern will thus be necessary. Of particular importance are the management techniques. The example of, and competi- regulations governing which entities may engage in for- tion from, well-run foreign companies can help domestic eign borrowing. In the long run, increasing the access of firms to identify weak links in management, product de- enterprises to foreign capital will probably be in China's sign, material supply, and quality of service and can spur national economic interests, but it is essential that such them to make changes they might otherwise never con- decentralization of foreign borrowing not precede effec- sider. The greater the dispersion of foreign firms among tive reform of enterprise management. Whatever the sectors and localities, and the more they are integrated rules of access, it is also essential that the central govern- into the domestic economy, the larger these demonstra- ment develop a debt information and monitoring system tion effects are likely to be. for provinces and enterprises. In recent years, China has increased its use of conces- There are, of course, many different types of capital sional and nonconcessional lending from official sources flow-including direct foreign investment, official lend- (mainly Japan and the World Bank Group), but the costs ing, and commercial bank lending-each with particular and benefits of such lending also need careful evaluation. advantages and disadvantages. As a result, some situations On the positive side, official flows are usually accompa- may justify net inflows of certain types of capital and net nied by technical and institutional assistance and some- outflows of others. Indeed, China has been in just such a times by policy assistance. In addition, official flows are situation in recent years, when current account surpluses usually of longer maturity than commercial borrowing allowed early repayment of commercial bank debt but and are often at lower interest rates. However, official the Government still encouraged net inflows of foreign lending can also have both direct and indirect costs; in- investment and official loans. Among the potential benefits of direct foreign invest- ment to China are increased access to export markets and 11. The average real London interbank offered rate, or LIBOR to the very latest technology in some fields. However, (deflated by the U.S. GNP deflator) was 5.6 percent during the latest product or process design may not be the most 1979-83, compared with 0.34 percent during 1974-78. 108 deed, even concessionary aid could prove costly to China in poorer provinces and regions. In addition, improve- if tied to a particular technology or project that was not ments in the transport system, particularly roads, will economically efficient. allow interior provinces to participate more in external In addition to direct foreign investment and official trade. lending, China would probably have to use commercial Chan;yes in external sector management, and particu- bank lending and other types of capital inflow to finance larly in the extent of regional autonomy, could also facili- part of the current account deficits projected in Table 6.6. tate greater involvement by interior provinces in external During the 1 970s, commercial banks played a critical role economic activities. At present the extent to which exter- in linking developing countries (particularly middle-in- nal sector management has been decentralized varies come developing countries) with international capital widely among regions. It is greatest in the four special markets, by providing syndicated loans and short-term economic zones established in 1980 (Shenzen, Shantou, trade finance, as well as medium-term export credits guar- Zhuhai, and Xiamen) and in the fourteen coastal cities anteed by governments. They also assisted many devel- and surrounding regions (including the four provinces oping countries to manage reserves and other commercial adjaceni to Shanghai) that were recently given many of activities. But since 1981, voluntary lending by commer- the same privileges. But some coastal provinces, such as cial banks has declined sharply, and the future availability Guangdong and Fujian, have more autonomy than oth- of commercial bank finance to developing countries re- ers, and the more remote interior provinces, such as mains very uncertain. Other sources of external finance Gansu, have the least autonomy. The degree of auton- could be considered, including the international bond omy varies not only in regard to trade in particular prod- markets, which traditionally offer long-term money at uct lines, but also to acceptance of foreign investment and fixed rates of interest, but which are very averse to risk signing of cooperative arrangements and to control of and require a consistently high level of creditworthiness. foreign exchange. International (as well as recent Chinese) experience Regional Issues confirms that the establishment of special zones that are less subJect to tariffs and other import restrictions, that China's more open orientation of recent years has bene- have gcod infrastructure and service facilities, and that fited coastal areas more than the interior. Procurement of speciali2e in production for export can be a useful initial goods for export has been increasing much more rapidly step in developing a more open external orientation for in coastal provinces likeJiangsu than in interior provinces the whole economv. However, because these zones are like Gansu, and in 1982 the level of export procurement enclaves within the domestic economy, the spread of new in relation to net material product reached 12 percent on technology and ideas tends to be limited. Policies and the coast (and as much as 31 percent in Shanghai) com- practices initially applicable only to special zones and to pared with only 4 percent in the interior. Provincial dif- export activities, where these prove to be beneficial to ferences in the degree of openness are partly a result of national interests, need subsequently to spread to other the transport advantages enjoyed by the coast, and so parts of the country and to other sorts of economic activi- long as most external trade continues to be waterborne ties, with a decline in the significance of special zones. the coast will retain these advantages. Nonetheless, vari- Long isolation and distorted domestic prices have prob- ous domestic and external policy measures could help ably increased the potential costs (as well as benefits) to spread the potential benefits of an open external orienta- China of a more open orientation. Initiating a more open tion to more regions and provinces. orientation on an experimental basis and in limited areas Many of the measures needed to stimulate develop- has therefore been appropriate. Equally, however, the ment in poorer regions (Chapter 5) would also enable recent major expansion of the geographical scope of these those regions to gain more from international trade. For open areas seems an appropriate response to their initial example, the introduction of a system of economic success (despite some unresolved problems), and further charges for land use could affect the choice between locat- expansion to include the interior provinces should be ing a land-intensive export activity (or foreign invest- considered. But to make this decentralization of external ment) in a large coastal city and locating it in a less economic management a success, in coastal and interior densely populated or interior region. Increases in un- provinocs alike, it must be closely matched by internal skilled wage differentials to better reflect interregional reforms to motivate and guide enterprises to respond variations in the economic cost of labor could also facili- rational.y to the challenges of a more open economy. tate the development of labor-intensive export activities 109 Managing Industrial Technology China's long-term economic objectives can be attained study estimates that onlv 20 percent of China's present only by combining extensive growth, which means fol- industrial technology is of 1960s and 1970s vintage, an- lowing the past pattern of expansion through duplication other 20-25 percent is backward but can still serve of existing production, with intensive growth, which in- present needs, and the remaining 55-60 percent should volves reducing costs, increasing productivity, and intro- be replaced. ducing new and better products. The need to use inputs Technological self-reliance has been sought not just at more efficiently is obvious. Improved products are the national level, but also by individual ministries, prov- equally important. Obsolete or low-quality producer inces, localities, and even enterprises (many of which, for goods raise costs and reduce quality in the industries that example, manufacture their own machinery and design use them. Inferior products also limit expansion of ex- their own products). This has been extremely wasteful of ports and lower living standards. both physical capital and human resources. Enterprise This chapter, which focuses on industry (though much machine shops are usually equipped with inefficient gen- of it applies equally to other nonagricultural sectors), first eral-purpose machine tools and are chronically underuti- discusses the role of technology in intensive growth. It lized. Equally serious, the talent of China's scarce techni- then deals with strategic planning for technological ad- cal personnel is wasted through duplication of effort. vance, means of acquiring technology from abroad, and Productivity performance has been disappointing. In domestic research and development (R&D). The discus- most rapidly developing countries, a considerable propor- sion turns next to ways of stimulating the application and tion of output growth is due to growth of total factor diffusion of innovations, and finally to policy on obsolete productivity (TFP), that is, to output growing faster than products and processes. A recurring theme is that the key the total of all inputs.2 But in China's state-owned indus- to intensive growth in China lies not only in better access try, TFP apparently increased in 1952-57, then stagnated to modern technology, but also in systemic reforms that or declined in 1957-82 (see Table 7.1). Increases in capi- would generate constant and spontaneous pressure on tal per worker raised labor productivity (output per enterprises to make sound investment decisions, to cut worker), but this was more than offset by a decline in costs, and to introduce new and improved products.1 capital productivity. Product innovation and quality improvement have also Technology in Perspective been slow in China. Many product designs have re- mained unchanged since their introduction decades ago. China has made tremendous strides in industry since For example, the ubiquitous Liberation truck is based on 1949. With one of the highest sustained industrial a Soviet model, which in turn was based on a U.S. model growth rates in the world (8.5 percent per year in 1957- from the 1930s; Chinese ball bearings last only one-fifth 82), a full range of production has been built up in nearly to one-half as long as the best of foreign ones; and 40 every industrial sector. This has been accomplished de- spite prolonged international isolation, by mastering 'lable~~~~~~~~~~ tehooy .mrvn it.n oiyn tt 1. Background Papers s-8 contain a fuller treatment, vwith available technology, Improvig It, and modifyig It to more references to other work, of the issues discussed in this suit local conditions. As a result, however, and despite chapter, including the experiences of other developing countries, some notable technological successes, industrial technol- the Soviet Union, and Eastern Europe, and technological develop- ogy in China has generally lagged behind that of industri- ment in electronics. alized and newly industrializing countries. One Chinese 2. See Chapter 1 and Annex 5. 110 Table 7.1 Index of Total Factor Productivity Figure 7.1 Output and Cost of Production of the in State-Owned Industry, 1952-82 JA 1-1 Sewing Machine, by Enterprise, 1980 Measure 1952 1957 1978 1982 I. Net output 37.6 100 673.3 798.4 Cost per nachine (yuan) 2. Laborinput 68.2 100 406.6 468.3 150-_- 3. Capital input 44.3 100 948.7 1,299.8 - 4. Total factor inputs (40% labor, 0 60% capital) 53.9 100 751.8 967.2 5. Total factor inputs (60% labor, 120 40% capital) 58.6 100 623.5 800.9 6. Labor productivity (I 2) 55.1 100 165.6 170.5 7. Capital productivity (I 3) 84.9 100 71.0 61.4 8. Total factor productivity 90 - * (I 4) 69.8 100 89.6 82.5 0 9. Total factor productivity (1 * 5) 64.2 100 108.0 99.7 0 Note Net output in constant prices was estimated by multiplying the ratio 60 _ of net to gross output in current prices of all industry by the index of gross output at comparable prices of state-owned industry. Capital is the original 0 value of fixed assets. Lines 4 and 5 present alternative estimates of total inputs, assuming different weights, and lines 8 and 9 present correspond- ingly different estimates of total factor productivity 30 Source. State Statistical Bureau, Statistical Yearbook of China, 1983 (Hong Kong: Economic Information and Agency), pp.7,13, 22, 126, 216. 0- I I 200 400 600 800 1,000 1,200 percent of Chinese nitrogen fertilizer is still low-quality Annual output (thousands of units) and unstable ammonium bicarbonate. It is widely believed in China that the way to solve a Includts production of both JA.1 andJA 2-1 models. these problems of stagnant productivity, high costs, and Source: Siate Economic Commission, Investigation and Research Office, low product quality is to introduce modern technology. "An Anaysis of the Economic Efficiency of the Sewing Machine Industry in China ''Jingji Diaocha (Economic investigations), no I (October 198 3), But in formulating policy, it is essential (a) to recognize p. 25 that in many cases new technology is neither necessary nor sufficient to improve productivity and product qual- ity, and (b) to have a clear understanding of how techno- logical improvement takes place and how it contributes causing waste and excessive wear on tools. Solving these to economic progress. basic problems would probably contribute more than the propos,-d introduction of sophisticated controls to achiev- Other Deterninants of Efficiency and Quality ing the joint venture's export targets. Productivity and product quality in Chinese industry Similarly, the large differences in performance among could often be dramatically improved without new tech- firms in a given industry often depend more on manage- nology, whose introduction may indeed divert attention ment and organization than superior equipment. Two of from more elementary problems. A case in point is the the largest, oldest, and best-managed watch factories in Tianjin Elevator Company's joint venture with the Otis China may be used as an example: the Tianjin Watch Elevator Company of the United States,3 where quality Factor), has more and better equipment and nearly as problems have prevented elevators from being exported. many 'orkers as the Shanghai Watch Factory, yet it Some officials of the Tianjin company believe that these produces only 35 percent as many watches at a cost 27 problems would be solved by replacing electromechanical percent higher. And the gap between the Shanghai fac- controls with sophisticated microprocessors. The imme- tory and many new factories is much greater. In the diate problem, however, lies more in work practices and sewing machinery industry, costs of production for one the quality of material supplies. Workers now hold metal model are nearly five times as great in the highest cost pieces for drilling rather than using clamps, measure enterprise as in the lowest (Figure 7.1). Even among the pieces manually rather than mechanically, and move par- five enterprises producing 124,000-135,000 units, tially assembled pieces by dragging them rather than us- ing a conveyor. The result is that holes do not line up and parts are damaged or do not fit together. Moreover, cast- ings supplied by other factories are of low quality, thus 3. SeeAsian Wall Street journal, August 10-11, 1984. 111 which probably have similar techniques of production, taken in the 1970s concluded that only one-third met the the cost differential is more than two to one. three criteria of short construction periods, high utiliza- Failure to exploit economies of scale (see Figure 7.1) tion rates after the start of operations, and good opera- has also denied-and continues to deny (Chapter 5)- tional results.5 (For some of the problems encountered in China an important source of productivity growth and steel, see Box 7. 1.) product improvement. For example, in 1978, 130 enter- Similar problems in absorbing foreign technology have prises, in twenty-six provinces and under the jurisdiction been encountered and studied in other countries. Tech- of several ministries, produced 150,000 motor vehicles. nology, it appears, cannot simply be transferred, since it Consolidation of production could have a major impact cannot be thoroughly codified in blueprints or embodied on product cost and quality even without sophisticated in capital equipment. Considerable effort and investment machinery and automated production, by standardizing are needed to master any technology and to make adapta- parts, allowing longer production runs for individual tions in complementary production, in management, and parts, and introducing more specialized machine tools. in the technique itself. Studies of "learning by doing" Production costs of bicycles, refrigerators, and watches have shown that improvements in productivity often are also unnecessarily high because of "small and compre- come through experience in production. There is a learn- hensive" production units in these industries. The solu- ing curve following the introduction of a new product or tion is not necessarily large and comprehensive units. In process, in which unit costs first fall steeply as total cumu- some industries, economies of scale in production may be lative production increases and then level off. Moving limited (for example, in the clothing industry) or may be down the learning curve does not happen automatically, offset by high transport costs (for example, in the cement however, but requires a conscious effort and investment. industry). Moreover, small specialized firms can be an Production experience needs to be complemented by sys- important source of growth (as will be discussed later). In tematic improvements in the basic technological knowl- the auto industry of industrial countries, a few very large- edge and skills of individual enterprises and specialized scale units produce key parts and undertake assembly, agencies. while numerous (often small-scale) units produce compo- Two other, related, lessons of technological advance in nents and provide specialized services. other countries also seem of great relevance to China. First, the cumulative effect of successive small innova- Economically Efficient Technologv tions (including modifications, adaptations, and changes Experience in China and other countries alike confirms of materials and work practices) on both production cost that there can be no quick technological fix to problems and product quality is larger than that of completely new of high cost and low product quality. New technology products or processes. It is therefore more important that improves economic results only when carefully selected enterprises should hunt every single day for small ways of and appropriately applied over a long period. improving performance than that they should make occa- Some innovations may actually increase China's pro- sional radical innovations. Second, a large gap exists in duction costs, especially if they substitute expensive and every industry and country between the newest (or electricity-intensive imported machinery, such as fully au- "best-practice") technique and the average technique in tomatic looms, for inexpensive and abundant labor use. Since it is the average that determines the economy's power. Associated improvements in product quality may overall performance, narrowing the technology gap be- sometimes justify the additional cost, but because of dis- tween the leaders and the laggards in an industry is as torted prices (including the cost of capital and product vital to cost reduction and product improvement as the quality differentials), enterprises now often make choices pace at which the leaders acquire new technology. that are wrong for the whole economy. Such investment in automation could also be economically rational in the Planning for Technological Development longer term if the machinery could be satisfactorily-and more cheaply-manufactured in China, and if the initial Good investment decisions-for both radical and small imports provided models for domestic manufacturers. innovations-are crucial for upgrading technology. An But this should not be assumed to be so without careful important element is sound investment project analysis, investigation. Deficiencies in China's existing technological capability may also prevent some new technology from being effi- 4. Ma Hong, "Do a Good Job in the National Economic c.ently absorbed. One Chinese study of large and me- Appraisal of Construction and Renovation Projects:" China Report: clently absorbed. One Chnmese study of large and me- diDm-sie poetcmltddiEconomicAffairs, no. 366 (July 20, 1983), pp. 1-7. translated from "among the nine projects imported from foreign Jisbu Jingjiyu Guanli Yanjiu, no. I (March 198 3), pp. 9-12. that 5among the nine proj s. Chen Huiqin, "The Orientation of Technologv Import countries, six have poor economic results."4 Another Must Be Changed," Economic Management (April 1981), pp. study of more than thirty major turnkey projects under- 22-2 5. 112 Box 7.1 Continuous Casting of Steel Development of continuous casting in steel production better-Iciown technology of steelmaking proved to be as offers an instructive example of how to open up a new important for raising productivity as attaining production technological path through selection of domestic and for- capabiliiy in the more radical innovation of continuous eign technological elements. The Wuhan Iron and Steel casting. Through an intelligent blending of foreign and Company (wisc) prepared design criteria and requirements domestic technological elements, wisc was able to master for a continuous casting plant in 1970. The Government the continuous casting process, but its success was marred approached a West German firm to help build the plant, by slow mobilization of domestic expertise to solve pro- and an agreement was signed in 1974. The agreement duction problems in the supposedly well-known steelmak- called for the German firm to provide all of the equipment ing process. and for the Chinese to construct the building, with collab- The initroduction of continuous casting was also success- oration in designing the building and erecting the plant. ful in ar,other way. China has used the experience of wisc More than 200 people were sent to Germany for training. to develop investment as well as production capability in Construction began in August 1975, the plant was com- continuous casting. The Echeng Steel Mill, a provincially pleted in October 1978, and trial operations were con- run enterprise near Wuhan, is planning an investment in ducted for four months, followed by full load acceptance continuous casting, for which the design work is being tests. The plant was approved for operation by the state in done by the Ministry of Metallurgy and the Wuhan Iron 1980. Initial problems with the control system were solved and Ste2l Design Institute on the basis of what has been by German technicians still in China. learned in wisc. The equipment will be jointly manufac- The continuous casting unit was thus operating tured by German and Chinese companies in China and smoothly from early 1980. However, the plant reached will be entirely installed by a Chinese construction com- only 50 percent of capacity in 1980, 60 percent in 1981, pany that participated in the installation of two other bil- 66 percent in 1982, and 79 percent in 1983. The bottle- let-casting units elsewhere in China. Training for the neck was in production of liquid steel because of a short- Echeng plant will be done in Chinese factories. Thus, age of oxygen supply for the converters and poor-quality through forward planning, China has been able to substi- refractory material, which resulted in excessive down time tute increasingly more domestic for foreign technological of the converters. These problems finally appear to have elements in continuous casting. There may have been been solved by bringing in a manager from the Capital some tradeoff between the development of investment ca- Iron and Steel Company in mid-1983. This illustrates the pability and diffusion of continuous casting production, limitations of using radical new technology to solve pro- however. duction problems. Upgrading production capability in the undertaken by efficiency-conscious enterprises and plan- mechaiiical watch production or only produce electronic ners, using methods that are well established in other watches for incremental demand? Should electronic countries.6 But major investment decisions should also be watches use domestic or imported components? These taken within a sectorwide strategic framework. And con- questions require careful analysis of existing cost differ- ventional cost-benefit analysis should be supplemented by ences and judgments about the future direction of tech- evaluation of alternative technological paths. This can be nological change. Because electronic watch production illustrated by an example from the watch industry. China might open up new technological paths for the economy, mostly produces mechanical watches, and its best factor- dual technological development (continued production of ies produce these efficiently by world standards. A large mechanical watches and experimental production of elec- domestic market and a modest export market for me- tronic watches) undoubtedly makes sense for a time. chanical watches will probably continue to exist for many China faces some unusual strategic choices. The nor- years. However, other countries have largely shifted to mal sequence of developing technological capability is production of electronic watches and production costs first to develop production capability, then investment are falling. At present, China could only produce elec- tronic watches cheaply by using imported electronic com- ponents, because of the high price and low quality of 6. The China Investment Bank has produced an 'Appraisal domestic components. Should China plan to expand me- Manual for Industrial Credit Projects," which adapts these meth- chanical watch production or shift to production of elec- ods to Chinese conditions. A similar manual has also bcen pre- tronic watches? If the latter, should China replace existing pared b) the Agricultural Bank of China. 113 tThe need toevluat hall options is0000 ilutatdb tea- f0orBFtiol:9 maufaturje alle. thR e ompnent Xof:n $the 300- t _ 2 i X W i g 00 thus SBFX had: a~~~~~~~~~300 proach tta kenir togup grx)sading ;itehnoog in bolilersfo powtaer00 S0Wr m egwtt U00boiler, ucninuetoimpr the drumur. PA= sCltnations. Th)e Shangh iBoilerFacetory (sBF) poduces sev- scl ond alt eirnativewouldbes ito manufacture ijfathe drtum in -. 00, \= to:f age, a new f~~~~~~~~~~~rogh 00 through-tmanyp0ietnboiler ftoroe :h staqtions. lStFfi was doissclartisfied40 new fatory inH ari o hspups.AtidalternativesfcgSFhokberaei waaiith, tne peformncel of innvtso 30-eawattboliler7 and Chias woul bagesae tohneypodcr upgradeihnuliyo hoc-hogthe 300- enteredy into anfl rangreemen winuthis afrSompmachney to pr- uegwtto boilgeera(wthutsa drm)awhinr(sch S as prodchiedtol fia rducs andrum-type o 30-mgawadett cnlg nnal n boiler.neh ge- frnyas. The Bonc7.2-throughdboiler hish protenialy.a everont, therUeS comprany will povid thes dirume oflorwhe energy -teffcintlogand usomewha schae tan teletois drum er brodcilers? whuhieSmnuacue impost comori ntentaord-y boier e th mtate in probiemrwthousin tly distigshat g beterato-e ingr tod a jroitd estign. haspab lsoy sent fifty peopluer trod the maticvcntrols m fust e intrdcd tof b pdabino letorlnkiup u nts ed Stattesnrone yar'st toupraining. Aftertheolthir set, dwith Iti improv ed ge rse ad ecturbincs. T ducs SBF had a Buildig (e bwhich d idyothe drum a itro i l pro duer oth once-through and factuig poetwaflwdintaitfietoeaut drmtp30-gwttboilers in the future. Parallel alterative proerly.One istane of his cncers the R&D o cometingtech olges is often justifiable, but full decisiono to uneimprtked cdompre whesiv dom ldesticsmanu productiondevelopmint oe adboth technologies may be un- turingl ofythe drumsraded foroug 3 megawattboilers; theren are seconomic . efrE vceeing to this final stage, a new manyh atrnauctlives thou thisopona, but ith isimot clearnthat ealuion odthies hal aves faciunge Sfshould be made in thsey htavegbeenifull evaeated Ones alternativedwould be light of experience gained through development thus fAr capability, and finally innovation capability.' But China is a large-scale machinery producer, upgrading the pro- already has a full range of industries, from machinery to duction of general-use machinery (such as machine tools final products, and needs to upgrade technology in nearly and boilers, see Box 7.2) should have high priority. every one. Where to break out of this circle of low In high-technology industries such as electronics, differ- productivitys Should China import foreign technology to ent strategic considerations apply, distinguishing between upgrade production capability for final consumer prod- the advantages of use and of production of electronics ucts, or give attention first to upgrading technology in products. It is the use of electronics products in many producer goods industries For which industries should industrial processes that offers the greatest potential for technology be importcd and for which could existing productivity gains. The advanitages of electronics produc- technology be upgraded through domestic R&D? And in tion are secondary. Even in the United States, high-tech- which product lines should China, for the time being, nology industries have accounted for only 3 percent of rely entirely on imports? Issues of strategic choice are best approached industry by industry, looking at all the stages of production from 7 aaiiisnee oaqie siiae s,aat producer goods to inemdaecmoet ofn change, or create technology can be placed in three categories: goods. An objective comparison should then be made of production, investment, and innovation. Production capability is the effects on production costs and product quality o needed to operate productive facilities. It is reflected in technical using existing domestic equipment, importing equipment efficiency and in the ability to adapt operations to changing market embodying a newer technology, and attempting to manu- circumstances. investment capability is needed to establish new facture more up-to-date equipment in China. For exam- productive facilities and expand existing facilities. It is reflected in ple, it makes sense in a relatively mature industry such as project costs and in the ability to tailor project designs to suit the 'les o cocentrte frst n upgadin prodctio ca- circumstances of the investment. innovation capability is needed to textile tocnetaefrto prdn rdcinc- create new technology or introduce existing technologies in a new pability in textile machinery. Improved domestically pro- environment. It is reflected in the ability to introduce processes or duced textile machinery could then be used to upgrade services that are less costly or to develop new products or services tetle production capability. More generally, since China that better meet specific needs, 114 the new jobs created during the past fifteen years. Many production base is broadened (see Box 7.3). For example, smaller advanced countries have almost no electronics use of computers in the Chinese economy could be in- production, but use imported components to reduce creased most quickly and effectively by importing fin- costs and improve products in other industries. ished systems. Domestic manufacture of a complete com- China may eventually be able to produce electronics puter system is unattractive, especially since it would not products competitively because of its large internal mar- function well unless all parts of the equipment develop- ket and low labor costs, and the low material intensity of ment program were successful. Moreover, a domestically production. Moreover, starting production now may produced system would be greatly out of date by the speed the process of learning and make it easier to catch time it had been developed. But for various reasons, in- up later on.' But China's technological capability is far cluding actual or potential restrictions on the supply of below that of world leaders, who in turn are still moving foreign -made computers, China may feel obliged to move down their learning curve, and Chinese electronics prod- ucts are now extremely high in price and often of poor quality.9 Obliging Chinese industry to use domestically 8. Flowever, it may also be possible-and cheaper-to wait manufactured electronics products would thus limit their until thc world industry learning curve has begun to level off and use and sacrifice many potential cost reductions and qual- then try to catch up, skipping some of the stages gone through ity improvements, in other countries. This would have to be evaluated industry by One way to help resolve the conflict between fostering industry 9. FDr example, the price of an advanced integrated circuit in the use and fostering the production of electronics prod- China is eighty times greater than the world price for the same ucts is to concentrate on mastering production of selected circuit. Other Chinese electronics products are typically five to ten products and solving key production problems before the times th, world price and of lower quality. Box 7.3 A Successful Strategy in Electronics Many developing countries have promoted their electron- ports is often held responsible for inadequate effort by the ics sectors over the past two decades; South Korea has industry to grow out of its "infant" status. been one of the most successful. By 1981, South Korea's The electronics sector in South Korea has been able to production was, for instance, estimated at around four reap economies of scale and break out of the constraints times that of India, a country with comparable levels of imposed by a small domestic market. The production of production and technology in the early 1 960s, and South consumer electronics in some other developing countries, Korea's technological expertise has moved decisively such as India, has been concentrated in small-scale enter- ahead. prises; limiting the scale of operation has been uneconomic In South Korea, the thrust was on rapidly acquiring for sorr,e activities and for many dynamic firms. More- production know-how, initially for assembly of final prod- over, small-scale enterprises are often ill equipped to under- ucts and later for the production of parts and subassem- take significant technological development. blies. The South Korean electronics industry has been free Choices about modes of technology acquisition and the to make product and technology choices and to respond to nature of capabilities acquired by the South Korean elec- market and technological trends. It concentrated initially tronics sector have apparently been economically appro- on consumer items and components, but has recently priate, given the needs of the sector as it evolved and the moved into industrial electronics, as well as some high- availabiLity of know-how within South Korea and abroad. technology semiconductor processing. In some other de- It has inported both "disembodied know-how" (for ex- veloping countries, such as India, the government re- ample, product designs and licenses) and technology em- stricted the growth of consumer electronics and promoted bodied in capital equipment, as well as parts and compo- the development of producer electronics. nents tl-at could not be manufactured domestically to The South Korean electronics industry has been heavily international standards. The government has encouraged export oriented. Import protection was provided for sev- diffusion, especially through the development of local sup- eral consumer electronic products, but competition in pliers. Attention was also paid to the development of ap- world markets gave enterprises a great incentive to reduce plied engineering capabilities, which is essential for the costs and improve efficiency and an opportunity to learn assimilation of more advanced foreign technologies as well from demanding users. In contrast, the Indian electronics as the rapid translation of know-how into better products industry, for example, has essentially been oriented toward and processes. In this regard, the South Korean Institute of the domestic market, and the heavy protection from im- Electronics Technology played an important role. 115 toward domestic manufacture. In these circumstances, it The Make or Buy Decision might be best to concentrate initially on developing pro- duction of certain key components (such as the central China has a wide network of research institutes with processing unit and monitor) and import those compo- proven ability to develop new technology. These have nents (such as the disk drive and printer) for which good- often matched or improved upon advances in other coun- quality domestic production would be difficult. Domestic tries, and they have also shown considerable skill in production could later gradually diversify to replace im- adapting techniques to China's special conditions. But ported components. The objective, as in other areas of excessive self-reliance has undoubtedly impeded China's electronics production, should be to match international technological progress in the past. China has preferred to price and quality standards in whatever is produced and develop known technologies by copying imported equip- to avoid producing items in which China cannot become ment or through completely independent research and competitive. development, rather than by importing technological Sectoral investment strategies have figured prominently know-how. The strategy of creating, and often recreat- in the development process in Japan, the only backward ing, technology contrasts sharply with the absorptive country to catch up with (and in some cases overtake) the strategy of Japan, which aimed to assimilate and adapt technological leaders. What lessons are there for China? foreign technology as fast as possible. The absorptive One lesson-abundantly confirmed by experience else- strategy has many advantages for a technological late- where-is that neither planners nor enterprises can pick comer, because selectively importing or borrowing tech- technological winners consistently. Japan's Ministry of nology generally involves lower risks, a shorter time lag, International Trade and Industry (MITI) is sometimes and a lower cost of acquisition. credited with extraordinary foresight and power; in fact The issue is not whether China should upgrade tech- MITT has just helped create a favorable environment for nology through its own research or through imports. technological advance, by facilitating information flows Studies in other countries have shown that these two among producers and reducing the risk of investments routes are complementary. Countries (and enterprises) that followed its consensual "vision" of future develop- with a strong R&D capability can make more productive ment. But it has not forced producers to conform to its use of imported technology because they can unpackage plans. Some large enterprises (such as Sony and Honda) the technology and be selective in importing. They can often pursued strategies at odds with official objectives: use the information developed from their own research private automobile producers, for example, made auto- to strengthen their bargaining position, use imports to mobiles one of Japan's leading exports despite MITT's ini- strengthen their own research, and through their own tial decision not to support development of autos as an R&D adapt foreign technology to local conditions. The export industry. There have also been failures, as when issue, then, is the right balance between domestic R&D both MITT and private producers wrongly expected petro- effort and importing technology. chemicals to become a competitive industry; and some- times private producers alone tried but failed to develop Importing Technology industries that MITI did not want to back (for example, Selectivity is the key to using imports in upgrading tech- commercial aircraft). A second lesson of Japanese experi- nology. In the past, China has tended to alternate be- ence is thus that planning for long-term technological tween the extremes of importing turnkey plants or rely- transformation need not, and indeed generally should ing wholly on domestic R&D. But there are many not, be mandatory. intermediate options. Elements of unpackaged technol- Two other features of Japanese industrial strategy have ogy that can be imported include licenses, designs, key been critical. First, the criterion of success has been ability equipment, and consultants to help solve management, to compete in the international market after a period of marketing, or engineering problems. Recent Chinese pol- infant industry protection, and second, enterprises them- icy has stressed the need to import "software" (know- selves have borne the risks of failure and reaped the re- how) in addition to, or instead of, "hardware" (equip- wards of success. The careful attention MITI has given to ment). This is an important advance in technology policy, developing internationally competitive industries is often because it recognizes that individual elements of technol- compared with the attention a mother gives a student ogy can be traded and that equipment is not necessarily preparing for an entrance examination. Japan has thus the most important element. had a flexible system of strategic planning that allows for Import policy should remain flexible, however. It is less mistakes and for experiments with alternative technologi- important bow foreign technology is obtained than what cal paths, and a system in which competition-domestic is being obtained and why. The appropriate mode of as well as international-provides a strong incentive to acquisition will depend on the objectives being pursued: seek out promising technological paths and retreat to gain production capability in a radically new technol- quickly from paths that lead nowhere. ogy, import of a turnkey plant may still be the cheapest 116 and most effective option; to develop investment capabil- turing equipment, by copying foreign equipment or using ity in producing better chemical equipment, it may be its own designs. The company imported a West German enough to purchase an overall design and certain special- electric resistance welding machine in 1962 and through ized parts. The objectives will determine the specific ele- its own R&D efforts has since tripled the machine's speed, ments to be acquired and the mode of acquisition; within essentially matching improvements made in Europe. The each mode of transfer, considerable scope exists for nego- company also developed a rim-making machine based on tiating conditions tailored to promote the objective. a study of the technical literature starting in 1976. Ac- Two important modes of acquiring technology, direct cording to company sources, this machine is superior in foreign investment and licensing or purchase of technical most respects to comparable equipment produced by a assistance, were discussed earlier (Chapter 6). Other leading French bicycle machinery manufacturer. China modes can also be effective. One is the "apprentice" has also begun to participate in international subcontract- pattern of development, in which the the first plant in an ing and is sending study tours and students abroad, as industry is built on a turnkey basis with substantial train- well a, attracting back overseas Chinese with technical or ing of, and observation by, local staff. In subsequent business experience. All of these international contacts projects, domestic technological elements can be substi- should increase the flow of information into China. tuted for foreign elements and assimilation of investment capability becomes more important than mastery of pro- Researnb and Development duction capability. Many developing countries have effec- In China, central government agencies and ministries, tively used this approach to upgrade local technological with their research institutes and superior resources, have capability in several industries, as has China in large-scale traditionally been considered the best agents for deter- nitrogen fertilizer production and in continuous casting mining what innovations are needed (though R&D pro- of steel (see Box 7.1). posals from below are in practice usually approved). For "Informal" means of acquiring foreign technology can example, the State Science and Technology Commission also be important. A survey of 1 12 South Korean export- selected thirty-eight key research projects and identified ing firms in 1976 showed that the most important several innovations to popularize during the Sixth Five- sources of process technology, both domestic and foreign, Year Plan. Central agencies can certainly play an impor- were overwhelmingly informal. Formal means (licensing, tant role in informing bureaus, corporations, and enter- technical assistance, and government-supported institutes) prises about technological developments abroad and in were important only 28 percent of the time. Informal progressive Chinese enterprises. They could also usefully means, particularly assistance from suppliers or buyers, monitor costs and the physical productivity of key proc- and hiring personnel with experience abroad (including esses and provide enterprises with feedback on their indi- foreign managers-for example, in shipbuilding) or in do- vidual performance in comparison to best (domestic and mestic firms, were much more important. The role of international) practice. overseas buyers in providing information on innovations There are severe limitations, however, to the top- is an important extra benefit of exporting and partly down, "technology-push" approach to innovation. Inter- explains why the fastest-growing economies in East Asia, national experience indicates that bottom-up, "need- with their strong export orientation, have been able to pull" innovation is generally much more successful. A upgrade technology quickly. study of major U.S. innovations by source during 195 3- Copying imported equipment is another informal 73 showed that 80 percent of new ideas came from means of acquiring technology, but one that must be used within the enterprises that produced the innovation, carefully. Even when it is technically possible to copy three-quarters of them from the producing division or foreign machinery, it may be uneconomic to do so. For department itself. Less than 5 percent came from univer- example, the Shanghai Boiler Factory and Shanghai sities and government laboratories. A European study Number 3 Machine Tool Plant jointly designed a special also showed that over 70 percent of the main ideas for deep-bore drill. The drill matches the technical quality of innovation came from within the innovating company an imported product and may even have cost somewhat and that by far the highest proportion of successful ideas less, but it took several years to produce and will not be came f1rom commercial rather than technical staff (Table produced again. This illustrates the tendency to concen- 7.2). Market feedback is thus the most important source trate only on the technical rather than the economic of ideas for innovation. aspects of innovation. But while copying is sometimes The traditional organization of R&D in China puts too uneconomic and can never be the only means of acquir- much emphasis on technology-push rather than respond- ing technology, it can be effective in some industries (par- ing to the demands of users for innovation. Only 10-30 ticularly simple mechanical industries). The Shanghai Bi- percent of research results are utilized in production. cycle Company has matched and even surpassed Moreover, research results passed to production units are advanced European standards for some bicycle manufac- often incomplete and prototypes require additional devel- 117 Table 7.2 Sources of Successful Innovations users. A danger, however, is that linkages between enter- in European Firms prises and research institutes will be emphasized at the Number Failure expense of linkages among enterprises. Moreover, re- of Commercial Commercial Technical rate search institutes often do not have sufficient production Source cases success failure failure (%) experience to make an effective contribution. Technical Pilot plants are being established in some industries to staff 31 8 19 4 74 link research and production. A prominent feature of Commercial Japan's early industrialization strategy was government- staff 60 33 21 6 45 sponsored pilot plants, although these were turned over management 6 1 4 83 to private industry once they reached a certain level of efficiency. China, too, has in the past established small- Source. The Economist (une 26, 1982), p. 96. based on data from the scale pilot plants for many products, some of which have European Industrial Management Association. been successfully scaled up for commercial production (for example, the production of benzine for use in polyes- opment work that the production unit cannot handle. In ter fiber). More recently, research institutes have em- addition, because of administrative compartmentalization barked on full-scale production of some products; for in China, potentially beneficial innovations are often not example, the Beijing Machine Tool Research Institute easily integrated into production if they require the coop- now assemblesJapanese numerical control devices for sale eration of other ministries to develop new materials, to machine tool enterprises. components, or ancillary machinery. The move toward commercially applicable R&D is com- User feedback is very important in the machinery sec- mendable, but should not be applied to basic research, or tor, but China's system does not usually allow sufficient even to all applied research. Most basic and some applied interaction between machinery users and producers. If research produces results, which-because they are either an enterprise's request for equipment is approved, a ma- universally or only indirectly applicable to production- chinery manufacturer is assigned to produce the equip- would not be ordered or purchased by fee-paying users, ment from a design of its own or of a design institute. even if they would benefit the whole economy. Stressing Users often cannot convey their special requirements to financial self-sufficiency might result in too few resources machinery producers, and many resort to designing or being devoted to research, or in research institutes be- manufacturing their own machinery, which is a far from coming mere production units, to the neglect of research satisfactory solution. Innovations by machinery manu- that has no immediate commercial payoff. facturers would be much more frequent if users had not Nonetheless, consideration should be given to transfer- only direct contact with suppliers, but also, and more ring some applied R&D from research institutes to enter- importantly, the freedom to choose their suppliers-in- prises. Most R&D in industrial capitalist countries takes cluding foreign suppliers if they can show that this would place within productive enterprises. Governments fund a be economically rational. This freedom should include large proportion of this research, partly to further defense the right to manufacture their own equipment, though objectives, but mainly because the risks involved, and the this option would undoubtedly be chosen less often than difficulty of patenting or commercially exploiting certain at present. In fact, during the readjustment period in research results, would otherwise lead to inadequate R&D 1981, equipment users did have much more freedom of effort. But even publicly funded research is usually lo- choice in suppliers, with clear beneficial effects (discussed cated in enterprises in order to respond to user needs and later); subsequently, as markets have become tighter, re- take advantage of the enterprises' production experience. strictions on choice have been reimposed. Locating applied research in enterprises is also consistent Several changes have recently been made in R&D policy with the greater importance, mentioned earlier, of tech- and organization. Most are potentially beneficial, but nological advance through cumulative small changes than their impact will depend largely on how they are imple- through radical innovations. To do this successfully in mented. Research institutes are now being encouraged to China, however, would require greater applied engineer- form closer links with production, through contract re- ing capability within enterprises, so that they could adapt search, technical assistance on a fee-paying basis, and and applv innovations introduced from research insti- joint research-production teams. Research-production as- tutes, foreign sources, or other enterprises. Enterprises sociations in the Soviet Union, however, have had a would also have to be given appropriate incentives. minimal impact on innovation; integration has been nominal only, and the activities of research institutes and production units have continued to be separate. Contract research and technical assistance offer considerable poten- Despite the extensive talent and tremendous drive to in- tial for making research more responsive to the needs of novate of its technical personnel, China's planning and 118 incentive system has traditionally been strongly biased ment, arid how they might be implemented, are discussed against innovation. An enterprise's most important pro- at length in Chapter 10. The same chapter discusses ways duction targets have been for physical output or gross of dealing with some of the social problems that might value of production. Few enterprises therefore have been arise w .th stronger positive and negative incentives for willing or permitted to interrupt (or risk interrupting) innovation-including closure of enterprises and high current production by introducing new products or proc- managerial and entrepreneurial earnings. esses, even when the longer-term benefits would be large. Secondary plan targets for cost reduction, material use, Difusion of Technology and product quality have done little to offset the bias Diffusion is one of the weakest aspects of China's tech- against innovation. Market pressures have been equally nology system. Although China is justifiably admired for ineffectual. Price differentials between high- and low- making economic use of older technologies, especially in quality products have been small or nonexistent. Enter- rural eniterprises, it also seems that innovations intro- prises have usually had no responsibility for sales. And duced in particular enterprises usually spread uneconomi- with goods generally in short supply-a seller's market- cally slowly to the rest of industry. Even in the steel almost any product has found a buyer. industrv, one of the best organized for spread of informa- The experimental reforms of industrial enterprise man- tion, Clhina has been rather slow to adopt energy-saving agement in the past few years have not basically changed innovations such as continuous casting and conversion the situation. Experience in the Soviet Union and Eastern from open-hearth to oxygen steel production. China has Europe confirms that minor modifications to the plan- also lagged in converting cement production to the dry ning and incentives system have little impact on innova- process (60 percent is still wet production), introducing tion and productivity. Special targets or bonuses for cost high-efficiency boilers, and upgrading quality or reducing reduction or new products do little to offset the over- costs in backward enterprises in consumer durable indus- whelming importance attached to quantitative produc- tries, such as watches, bicycles, and sewing machines. tion targets. Profit targets and profit retention schemes Diffusing best-practice techniques, including management offer only a weak incentive to innovate when enterprises techniques, already in use in China may, as discussed can sell everything they produce without innovating. earlier, be more critical to productivity growth than in- Moreover, distorted prices encourage enterprises to pro- troducing more advanced technology into a few leading duce the wrong products or to use scarce inputs waste- enterpr[ses. fully. Increased price flexibility for new products tends to China's institutions and policies for diffusion of tech- stimulate cosmetic product changes to justify price in- nology have followed the Soviet model, which stresses creases, generating inflation with little meaningful inno- the benefits of a free flow of knowledge. Specialized re- vation. The most significant change that has occurred in search institutes have usually developed innovations, China has been in marketing. Since 1980, not all output which zan then be passed to production units free of has been guaranteed an outlet through planned purchases charge. Similarly, approved new products and techniques by commercial departments or the material supply net- develoFed by enterprises can be diffused to other pro- work; part has had to be sold directly to users. This ducers at no charge. The Soviet Union and China have change, at least for a time and in some industries such as both recorded impressive technological achievements in machinery, has created a buyer's rather than a seller's areas ir which they have concentrated the considerable market. Enterprises therefore have had to pay attention resources of their R&D systems, but they have been much to product quality and customer preferences to maintain less successful in diffusing a wide variety of improvements sales and profits, which has led to some changes in prod- throughout industry. uct mix and an increase in product innovation in several One important reason for failure to diffuse technology industries. is the general lack of incentives for innovation discussed For China to catch up with the world's technological earlier. Other reasons are also systemic, but relate to the leaders, far more radical changes in the system of incen- nature of technology. Technology is not machinery or tives for innovation would be required-with far-reach- blueprints that can be simply passed from one user to ing economic and social implications. The changes re- another, but is rather a whole method for doing some- quired include: (a) greater enterprise and managerial thing, which requires understanding of associated work accountability for financial performance, with heavy and procedures and organization and which can be assimi- unrelenting penalties for poor performance as well as lated only with considerable effort, cost, and sometimes large rewards for success; (b) rationalization of prices, to substantial adaptation. Three examples of failure to dif- ensure that profit and loss reflect cost efficiency and the fuse innovations in China illustrate shortcomings of the economic benefits of what is produced, and (c) greatly present system. increased competition among enterprises. These essential and basic reforms of the system of economic manage- EXANIPLE 1. A Shanghai foundry producing small cast- 119 ings for machine tools successfully copied (achieved two- joint venture with the Suzhou Bicycle Factory for associ- thirds of the speed of) a Danish automatic molding ate production of the Forever brand bicycle. SBC will process. This innovation was not diffused. Some years provide technical assistance to improve the quality of the later, a research institute developed a similar machine, main parts produced by the Suzhou plant, as well as of which was then made available to other foundries. In this several small components made by other manufacturers. case, diffusion occurred from the research institute, but This should expand the market share of the superior not from the enterprise. In general, innovations have Forever brand at the expense of the lower-quality Suzhou sometimes been successfully diffused to subordinate bicycle, but without loss to the workers in the Suzhou units, but they are rarely diffused horizontally (there is factory. little communication with other ministries or regions pro- Examples I and II show that enterprises that innovate ducing similar products). successfully need incentives to diffuse new technology to other producers. A change in the right direction now EXAMPLE II. The Shanghai Bicycle Company (SBC) is being made is the introduction of patents, internal licens- not only one of the leading bicycle producers in China,"' ing of technology, and fees for technical assistance. but also a producer of bicycle machinery that is advanced Examples II and III suggest that greater labor mobility by world standards. However, SBC is not interested in would enhance diffusion. Hiring of experienced techni- producing bicycle machinery either for other plants in cians and managers is perhaps the most important source China or for export. The Shanghai Light Industry Bureau of new knowledge for enterprises in most countries. (Ex- has set up a specialized factory to produce bicycle ma- cessive labor mobility, however, can create a disincentive chinery, which may use SBC designs but will have only for enterprises to invest in training.) Japan is a notable minimal access to SBC specialists. Since much technologi- exception to this general rule, at least in the large-scale cal knowledge cannot be explained in manuals, but rather manufacturing sector. Japan apparently substitutes for is embodied in people, the bureau's factory is unlikely to this means of diffusion partly through MITI's efforts to be as proficient as sBc and is certainly less likely to have bring industry experts together to share experience and to the export potential of SBC. In other countries, SBC develop a consensus on future development and partly would spin off a specialized company to produce machin- through informal contacts among people who went to ery, or its engineers would set up their own company or the same school or university but now work in different be hired by the new producer, thus diffusing the technol- companies. Even so, Japanese enterprises have far more ogy of machinery production. flexibility in hiring than Chinese enterprises, so that a textile enterprise, for example, can hire electronics engi- EXAMPLE III. The Shanghai Number 6 Textile Mill neering graduates. China is now introducing some flexi- achieved national recognition for its success in developing bility in its hiring procedures, though far more is needed computer monitoring of looms. About fifty other textile (Chapter 8). Professional associations based on individual producers in China have built or will build a similar membership have also been revived, thus encouraging system, many with the assistance of the Number 6 mill. engineers and scientists in different industries and fields to Thus diffusion is occurring because of the organization make contact. and communications within the Ministry of Textiles. The These examples also show that China needs to encour- Number 6 mill now proposes to develop, in cooperation age specialization. Special agents can mediate between with a research institute, an online microprocessor con- producers and users of technology or provide missing trol system, but neither unit has any staff specially trained technological elements. In Example III above, consultants in microprocessors. Since it is difficult for the textile in- with expertise in electronics and some familiarity with dustry to hire an electronics engineer or to get support textiles could probably have provided a fast and cost- from the Ministry of Electronics, the two units will effective alternative to staff training or interministerial mostly rely on training their own staff through seminars coordination. Consultants and other specialized agents, given by the city. moreover, have an incentive to speed diffusion whereas competing enterprises may wish to limit it (for instance, CONCLUSIONS. Several improvements could be made in Example II, SBC might not wish to assist other bicycle to remove the barriers to technology diffusion in these manufacturers by selling them improved machinery, but a examples. Examples I and III illustrate the problem of specialized enterprise spun off from SBC would have a poor communications between ministries, regions, and strong interest in selling better equipment). enterprises. Though this problem can never be wholly overcome in an administratively regulated economy, joint ventures across administrative boundaries could diminish it. For example, as part of the plans for cooperation 10. Its Forever brand bicycles are in very high demand and must within the Yangtze Delta zone (established to break down be rationed in the domestic market. SBC also exports bicycles to provincial barriers to development), SBC has formed a Sears, Roebuck & Co. in the United States, 120 China has recently begun to encourage the formation of consulting companies, using experts from research and Obsolescence and Scrappig design institutes, universities, and ministries. But much Old and new processes and products coexist in all coun- more needs to be done to facilitate the creation of new tries, since it is rarely economic to scrap and replace enterprises and organizations that specialize in production recently acquired equipment or recently introduced prod- of technology-related goods and services, and indeed also ucts the moment new processes and products become of other things. In other countries, units providing ser- available. Thus, although China has deliberately chosen vices such as repair and maintenance often spin off from to use old and new technology simultaneously as one their parent firm to provide services on a contract basis to aspect of the strategy of "walking on two legs," it could both the parent and other firms, which helps make firms also be said that the typical economy walks on far more in most industrial capitalist countries less comprehensive than two legs: it is like a centipede, with the newest than Chinese firms."1 Large Japanese firms likewise often products and techniques at the front, the oldest at the help set up trusted employees in business as subcontrac- back. The creature moves forward steadily, innovating at tors. The specialization that results from spinoffs is one of the head, abandoning old products and processes at the the main vehicles of productivity growth. tail. But it always remains quite long: thc vintages of equipment in use may span two decades or so, with The Role of Small Enterprises corresponding variations in enterprise profitability, while Greater reliance on small enterpnrses for innovation could outputs of widely varying quality are sold at different also contribute more generally to China's technical prices to different segments of the market. progress. Studies in other countries have shown that Whai: is unusual about China is the length of the tech- small industrial enterprises are just as innovative as large nological centipede, and indeed its tendency to get longer ones, especially in sectors such as specialized machinery, and longer. As new processes and products are intro- electrical and other consumer goods, and in small com- duced, the head moves forward, but the tail does not puters and related software. Even in sectors where large- move. There is very little scrapping of equipment and scale production is optimal, such as basic chemicals and abandoning of old products, even when it makes good metallurgy (and, in electronics, semiconductors), small economic sense to do so. Antique vintages of equipment specialized firms have contributed many useful innova- are not merely used, but continue to be reproduced and tions, sometimes in the role of subcontractors to the embodied in new investment on an enormous scale by larger firms. (Chinese visitors to foreign countries are China's machine-making industry.'2 Equally remarkable often given a misleading impression in this regard, be- is the coexistence of high- and low-quality products sell- cause they are usually taken only to well-known, large ing at the same price, with the result that, for example, enterprises.) inferior bicycles accumulate in inventory while "famous In China, by contrast, it is generally assumed that small name" brands continue to be rationed. firms are necessarily technologically backward, while Therl are various reasons for this unusual pattern. large firms must be the technological leaders. As a result, Some of them are connected with aspects of China's collective and individual enterprises, which already have system of economic management mentioned in earlier strong incentives to innovate, are denied access to the sections of this chapter-few incentives for innovation, skilled manpower, modern equipment, and foreign ex- little competition, chronic shortages, irrational prices. change they need to innovate. State enterprises, which There is also a widespread, deep-rooted aversion to the have far less incentive to innovate, have exclusive access "waste' involved in scrapping equipment that still func- to the necessary resources. tions, or any product that might still serve some purpose. Diverting a substantial fraction of China's skilled man- This is coupled with intense reluctance to displace work- power, technology imports, and research support to small ers, largely because of fear that they would become un- enterprises on a competitive basis would unquestionably employed. These apprehensions have some foundation, speed China's technical advance. Moreover, the social but are in important respects misconceived and exagger- problems posed by strong incentives for innovation could be less serious in the context of small than of large enter- prises. High individual or collective earnings from suc- cessful innovation may sometimes be more acceptable than large payments to managers and researchers insu- 11. In Switzerland, watch companies do not have their own lated from risk in big organizations. Similarly, the failure machine shops or provide all of their components as in China. In of a small enterprise (in other countries, most small enter- Japan, steel companies contract out many services, such as chang- ofriasmfail within a year or so of starting up although ing refractory linings in furnaces, that are provided within Chinese prises fall within a year or so of starting up, although'enterprines. Some steel companies in less developed countries even those involved often try again) affects only a few workers contract for this difficult operation internationally. and a small amount of capital and has little effect on the 12. An eminent Chinese economist, Sun Yefang, was the first to prosperity of its locality. describe China's machinery industries as reproducers of antiques. 121 ated; they will need to be dispelled if China is to make problems with consumer goods, where the higher cost of rapid technological progress. the new item has to bc weighcd against its benefit to The Government, recognizing the outmoded state of consumers. Willingness to pay, reflected in the price dif- much of China's industry, is at present stressing technical ferential between new and old products, is the measure of transformation of existing enterprises rather than wholly consumer benefit in a market economy. But at present in new construction. In line with this approach, it needs to China, consumer prices are not allowed to move freely be more widely understood in China that the economic enough to serve this purpose. Even if they were, the life of equipment is usually shorter than its physical life.1 Government, like others in developing countries worried It would also seem worth disseminating certain well-es- about the spread of costlv Western consumerism, might tablished rules for deciding whether an old process is still not see them as an appropriate guide to the allocation of economically viable, or whether it should be upgraded or technological resources. scrapped, with or without replacement. Specifically, an The possible adverse effects of scrapping on employ- old process should in principle be scrapped and replaced if ment should also be put in perspective. Much of the the total cost of production using new equipment (includ- replacement process can occur within existing enterprises: ing interest and depreciation on the new capital em- inJapan, for example, lifetime employment in large firms ployed) would be less than present operating costs (mate- has proved compatible with high rates of product and rials, energy, and wages). It should be upgraded if the process innovation, and much the same has been true in saving in operating costs would be greater than the capital other countries. At the same time, the experience of cost of the upgrading investment. And it should be smaller firms in Japan and elsewhere makes clear that scrapped without replacement if both the operating cost evolutionary competition between old and new products with existing equipment and the prospective total cost and processes would inevitably lead to the closure of with new or upgraded equipment exceed the product enterprises, with consequent loss of jobs. In general, how- price. ever, the closure of old enterprises actively stimulates the These principles are logically powerful and could assist expansion of technologically more advanced enterprises, in choosing technical transformation projects, many of creating new jobs to absorb the displaced workers. This is which are now selected more on the basis of engineering because the continued existence of backward enter- enthusiasm than economic calculation. But they are not prises-as in China-acts as a drag on progressive enter- easy to apply in present Chinese circumstances, mainly prises, by depriving them of materials, skilled labor, capi- because prices, wages, and charges for the use of capital tal, and markets that they could put to more productive do not reflect the true value to the economy of the items use. concerned. Cheap fuel, for example, would on the princi- Over time, and looking at the economy in aggregate, ples outlined above encourage uneconomically slow re- both economic logic and the experience of other coun- placement of energy-intensive equipment, while capital tries suggest that high rates of scrapping and replacement grants, cheap loans, and unduly high wages for unskilled cause faster, rather than slower, growth of modern indus- labor would all lead enterprises to scrap and replace trial employment. This may require many workers to equipment sooner than economically desirable. Nor is change jobs more frequently, sometimes with transitory there an adequate market in used equipment, which periods of unemployment; it may also, more seriously, could allocate items abandoned by one enterprise either cause certain workers and localities to experience pro- to other enterprises that could use them economically or tracted unemployment. There are policies and institu- to the scrapheap. In this situation, in the near term, tions that could alleviate the economic hardship of dis- enterprises and their supervisory bureaus might be placed workers and assist in their retraining and obliged to use shadow prices set by some central agency reemployment. But, as in any society undergoing rapid to make economic audits of their equipment and deci- change, the losses of particular individuals and groups sions about replacement and other technical transforma- from technical advance would still have to be weighed tion. In the longer term, price reform and related mea- against the gains of the majority. sures would help enterprises spontaneously to make appropriate decisions on old equipment. 1 3. It is sometimes argued in China that raising the depreciation Similar principles exist for scrapping and replacement rate, which is at present only 3-4 percent per year. would bring of products, but are also not easy to apply in present about a more appropriate pace of scrapping. Such a step might Chinese circumstances. In the case of producer goods, the indeed help to overcome some irrational resistance to abandoning t oi things before the end of their physical life. But, as explained in the test is whether the higher cost (because of the additionai text, the magnitude of accumulated accounting depreciation does Investment required) of supplying a new and better Item not enter into an economically sound calculation of whether or would be more than offset by the cost reductions it not a particular item should be scrapped at a particular time. Only would permit users to realize. Again, though, the test is the operating costs of installed equipment are relevant to decisions valid only if prices are rational. And there are additional regarding its future use. 122 Human Development People are both the means and the ends of economic position (Table 8.1). The proportion of the population development. Their numbers, health and nutrition, edu- with primary education is exceptionally high. But, largely cation and training, allocation among economic activities, because of the Cultural Revolution, which stopped or effort, and initiative will largely determine the pace and hampered higher, upper secondary, vocational, and tech- pattern of China's development. People's well-being, nical education for ten years, the number of people with however, is the whole object of development-to enable advanced educational qualifications is small-smaller, in- them to enjoy greater consumption of material goods and deed, in the twenty-five to thirty-four age group than services, improved health, increased opportunities to among those over thirty-five. China's many years of iso- shape their own lives, and a greater sense of security and lation have also contributed to the smallness of the stock social worth. of people equipped to apply or teach the up-to-date skills China's future human development policies-especially essential for modernization. in education, employment, and wages-must take ac- count of the need for structural change. At the macroeco- Basic Educational Priorities nomic level, rapid growth will require and bring about These deficiencies in advanced, high-quality education substantial changes in the sectoral pattern of employ- have-rightly-been the focus of China's educational ment. At the microeconomic level, as discussed in Chap- strategy in the past few years, when great efforts have ters I and 7, structural change should be even more been made to restore academic standards, to expand both rapid, as newer and better processes and products displace formal and informal higher education, and to plan for its older ones, causing constant shifts in the relative effi- furthei substantial enlargement over the next two de- ciency of different enterprises and in the pattern of de- cades. These efforts should not, however, divert attention mand for labor. A central theme of this chapter is that from basic education, where problems have emerged in China's presently rigid, compartmentalized approach to recent years. the development and use of human skills could prove a major obstacle to rapid and equitable growth. INTERNATIONAL EXPERIENCE. China's long-standing The first part of the chapter addresses some key issues comm.tment to widespread basic education has a sound in education and training, including the role of man- economic, as well as social, foundation. Some other power planning.' The discussion then turns to employ- countries, such as Mexico or Brazil, have industrialized ment, productivity, and wages, with special reference to quite rapidly with relatively low levels of basic education. labor surpluses and shortages, allocation of workers, and But their development has been characterized by much payment systems. The final section deals with some po- inequality of incomes, partly due to the high earnings of tential problems of the twenty-first century arising from the educated minority, and has been unduly oriented changes in the size and composition of China's popula- toward capital-intensive, not very efficient, industry. By tion, whose solution would require action within the contra,t, widespread basic education in some East Asian next decade or so. Education and Training Issues 1. Annex I contains a fuller discussion of recent educational developments and of issues in primary and secondary education, By comparison with other low-income countries, China's technical and vocational education, teacher training, and education present stock of educated manpower has an unusual com- costs and financing. 123 Table 8.1 Educational Attainment of the Population, by Age and Sex, in Selected Countries Percentage ofpersons who bave completed at least Country Prinmary school Lower secondary school Upper secondary scbool Postsecondary school and age group Male Female Male Female Male Female Male Female China, 1982 Totai,l 5yearsormore 79.1 51.1 42.9 26.0 13.3 8.3 1.0 0.3 15-24 years 95.1 82.2 71.0 53.6 23.5 17.8 0.1 0.1 25-34years 88.8 61.9 48.0 26.4 13.4 7.6 0.8 0.4 35yearsormore 63.2 24.6 21.5 7.5 6.4 2.4 1.6 0.5 Other low-income countries India, 1971 Total,15yearsormore 37.2 14.7 21.3 7.1 10.3 3.0 1.6 0.4 15-24 years 53.6 27.4 35.5 1 5.1 15.6 6.4 1.2 0.7 25-34 years 39.4 14.4 23.4 6.5 13.1 3.1 2.7 0.7 35 years or more 26.4 7.1 12.0 2.4 5.9 0.9 1.3 0.2 Pakistan, 1973 Total,lS years or more 30.8 95 . 12.5' 3.9" 1.9 0.5 15-24years 44.8 18.0 . 17.0' 7.3" 1.5 0.9 25-34 years 34.8 8.8 . 15.2' 3.71" 3.4 0.8 35yearsormore 20.8 4.8 . 8.7' 2. 11 1.5 0.2 Sri Lanka. 1971 Total, 15 years or more 59.6 44.8 . 7.7T 6.6") 0.7 0.3 15-24 years 69.8 65.0 7.5' 7.71' 0.1 0.1 25-34 years 66.5 53.6 . 12.2' 10.8", 1.6 0.9 35 years or more 49.1 30.8 . 5.7' 3.6" 0.8 0.2 Middle-income countries Brazil, 1970 Total, 15 years or more 33.6 31.7 . 9.9' 9. 1 1.5 0.5 15-24 vears 41.5 41.6 . 10.8' 11.6" 0.1 0.1 25-34 years 35.0 31.8 . 11.8' 10.6' 2.1 1.0 35 years or more 26.7 23.5 . 8.2' 6.31" 2.3 0.6 Colombia, 1973 Total, 15 vears or more 80.7 79.0 . 26.0' 23.7" 4.1 1.6 15-19 years 87.5 89.8 . 36.0' 36.2" 0.7 0.7 20-29 vears 86.4 86.5 . 34.4' 30.7" 7.4 3.6 30-39 years 81.4 78.2 . 22.4' 18.2" 5.2 I1. 40 vears ormore 71.1 65.6 . . 14.6' 12.3" 2.8 0.4 Mexico, 1970 Total, l5vears or more 31.3 27.7 10.8 7.0 6.4 4.1 3.8 1.5 15-24years 44.4 39.7 1 5.1 10.4 6.8 5.1 3.2 1.7 25-34 years 31.5 27.5 11.6 7.1 8.5 5.0 5.6 2.0 35 years or more 20.5 17.8 6.9 4.2 5.0 2.8 3.3 1.2 South Korea Total, 15 vearsor more 82.3 66.3 44.5 22.4 23.7 8.9 6.2 1.6 *15-24 years 97.5 95.9 55.9 39.3 23.2 13.1 1.6 1.5 25-34 years 94.9 8 5.7 57.9 29.6 37.2 14.2 11.7 3.4 35 years or more 63.7 37.7 28.2 8.2 16.3 3.6 6.6 0.7 a. All secondary school, male. b. All secondary school, female. Source Ten Percent Sample of Cbina Population Census (Beijing State Statistical Bureau, 198 3); Country Demograpbic Profiles (Washington, D.C.: U.S. Bureau of the Census, Center for International Research Division), selected issues. economies has helped achieve both unusually high output and efficiency.2 Failure to achieve basic literacy is a con- per unit of physical capital and an unusually equal sharing demnation to utter poverty: strong evidence links pri- of the benefits of rapid developmenit. Western specialists now believe that developing coun- tries may have overemphasized higher education at the 2. See World Bank, World Development Report 1980 (Aashing- expense of primary education and in some cases also ton, D.C., 1980; reprinted by the World Bank in 1982 as Poverty secondary education, from the viewpoint of both equity and Human Development) for references to detailed studies. 124 mary education and the ability to earn an income, in unless Farents actually enroll their children in school and both rural and urban settings. Economic rates of return encourage their continued attendance. on educational expenditure, viewed as an investment, The campaign to publicize the benefits of basic educa- appear to compare favorably with rates of return on tion should clearly be continued. But the Government physical capital; they are generally higher for primary should also direct more financial support to schools in than for secondary education, and for secondary than for poorer areas. This money could be used to reduce some tertiary education. Furthermore, educating women en- of the obstacles to school attendance, including fees, dis- courages and assists them to have fewer children and tance (by building more schools or providing free trans- contributes to their children's health and educational at- portation), and the quality of staff and buildings. Finally, tainments. and especially insofar as there is a conflict between the interests of parents and children, serious consideration PROBLEMS IN CHINA. International experience should should be given to making attendance compulsory, as is thus reinforce concern about the drop (discusscd in Chap- normal in industrial countries and already practiced in a ter I) in primary and secondary school enrollment rates few places in China. This is not a simple solution, since it in China's rural areas in the early 1980s, mainly as a creates administratively (and often morally) difficult, and consequence of the production responsibility system costly, enforcement problems. But China's relatively high (PRS). The declines have been most marked in secondary degree of social organization would make it less difficult education and for girls in poor areas; this is apparently than for most other developing countries. related to employment opportunities. Thus school atten- dance in the richer parts of Jiangsu has not been much KEY CHOOLS. Educationdefinessubsequentopportu- affected, although the productivity of additional child nities to an unusual degree in China. Examinations deter- labor on a family holding might be quite high, because a mine whether a child will attend senior high school, and junior secondary education is necessary for a job in a local thus whether he may take the unified national university commune or brigade enterprise. Furthermore, education examination. This examination determines the institution can contribute more to raising agricultural output in areas to be attended, if any, and the course of study to be such asJiangsu where technological advance is rapid. followed, which in turn has normally determined the In the poorer parts of China, however, where there is graduat:e's life-long job assignment. Examination results little technological advance in agriculture or opportunity are thus of momentous individual significance; and they for nonagricultural employment, the economic return to may be greatly affected by the quality of the school a education must seem very low. Peasant households in child attends. It may therefore be questioned whether these areas tend to have large families, so a twelve-year- China's key-school policy, which deliberately increases old daughter can help around the house or look after quality differences among schools bv concentrating re- younger siblings without seeming to jeopardize her pros- sources in the better schools, is desirable, or whether it pects. But taking a longer-term view, the same daughter unnecessarily increases inequity and widens the gap in may be in the labor force forty years from now, when educational access between rural and urban areas. economic opportunities will be very different. Moreover, Research has shown that the educational system has an if there is to be significant migration out of the poorest inherent tendency to perpetuate the economic advan- parts of China (Chapter 5), a good basic education will be tages a-d disadvantages of particular families from gener- essential to enable migrants to adapt to a new economic ation to generation, which is unlikely to be absent in and social environment. China, even though differences in family background The Government has taken steps to reverse the enroll- may be smaller. This tendency is aggravated by the une- ment decline, apparently with more success in primary ven quality of secondary, as well as higher, education than in secondary education. But increased local financial even in the richest countries. Schools with the highest self-reliance in basic education seems to have aggravated achievement rates are located in cities and hence are more the already substantial problems of poorer areas (see accessible to children from urban families. All countries Chapters 5 and 10). Moreover, the factors that caused also eventually abandon the principle of equality in the enrollment decline could be a serious obstacle to higher education in order to concentrate students, staff, attaining the Government's target of making nine years and facilities in centers of excellence, because of the costs of basic education universal by the end of the century. involved in scattering them among all instittitions of This is an important target-essential if China's full po- higher learning. tential for growth is to be realized, and its benefits equita- Given the low enrollment ratios, poorly qualified or bly distributed between town and country and between over-age staff, shortage of library facilities, and obsolete backward and more advanced localities. It is also an af- equipment in China's higher education, some scarce re- fordable target, because of the prospective decline in the sources should probably be concentrated in a few "key" primary school age group. But it will not be achieved universities. But the gains in educational quality or cost- 125 effectiveness from such a policy at the secondary and attempt to provide all their own manpower; for example, especially at the primary level are questionable. Concen- the Ministry of Railways has a medical school to train tration of resources in certain privileged schools encour- physicians for employment in its hospitals and clinics. ages the educational system to classify children too hastily This "small but comprehensive" approach, even where it as high or low achievers, which leaves much potential does not result in overspecialized courses, is wasteful be- ability untapped. Many countries, capitalist as well as cause it passes up potential economies of scale. socialist, have rejected such policies in basic education as Given the apparent shortage of skills, the recent fall in contrary to both equity and economic efficiency. enrollments at-the mainly enterprise-run-schools train- ing skilled workers is surprising, especially because the Specialized venus General Education disincentive to enterprise training that worker mobility The Government's present policy is to reduce opportuni- creates in other countries is much less in China (an enter- ties to go to general senior secondary schools and to prise can select the best trainees from its own schools). promote vocational education, by adding vocational The causes of this decline, which could be related to the courses at some general secondary schools and wholly new emphasis on profitability (although some enterprises converting others into vocational secondary schools. This are using part of their retained profits to supplement the policy is a reaction to the heavy pressures on university 1.5 percent of the wage bill they may use for training), entrance and the general difficulties of absorbing new should be investigated, as should its consequences. This is labor force entrants into productive jobs. However, en- partly because upgrading and relearning of skills will con- rollment in general upper secondary education is not high tinue to be badly needed (the state of China's impres- in China compared with other rapidly growing develop- sively large and capable, but outmoded, stock of human ing countries, and its contribution to raising the produc- resources strongly resembles that of its physical capital tivity even of pupils who do not go on to university stock and requires just as much "technical transforma- should not be underestimated. tion"). It is also because enterprise-sponsored vocational Vocational secondary education, moreover, has been training, despite the danger of excessive narrowness, can controversial in many countries since the beginning of ensure close contact between the trainers and users of this century. In the 1 920s, for instance, Strumilin argued trained manpower and provide opportunities for training that sound general education supplemented by brief voca- on the latest equipment, thus overcoming problems of tional training was the most efficient form of training for irrelevance and obsolescence that have plagued vocational skilled workers.3 For China, this argument is reinforced training elsewhere. by the need for rapid structural change over the next few This advantage of enterprise-sponsored training can decades, which will require constant alterations in the also be captured by schools in close touch with enter- skill composition of the labor force, and by the slow prises, but funded by local governments, several of which prospective growth of the labor force in the twenty-first already exist in China. Representatives of local enter- century (discussed later), which will reduce the scope for prises can be on the boards of directors for the schools changing skill composition simply by changing the pat- and can arrange for on-the-job training using relevant tern of training of new labor force entrants. Providing equipment.4 These schools can provide broader training workers with good general education, which is the best .S. G. Strumilin, The Economic Significance of National foundation for retraining as circumstances change, is thus Education:, in Readings in the Economics of Education, edited by likely to yield large economic benefits. Mary Jean Bowman and others (Paris: Unesco-IIEP, 1968), pp. Any vocational training should be in skills adaptable to 413-50. Despite Strumilin's opposition, the Soviet Union did a wide range of occupations, but vocational training in eventually develop formal vocational education. Various attempts China suffers from one of the acknowledged weaknesses were also made to integrate vocational and general education, but of the Soviet system on which it is modeled: overspecial- none was successful. See Irene Blumenthal and Charles Benson, ization. China now offers 700-8 00 training programs Educational Reform in the Soviet Union: Implicationsfor Developing (ztheSoviet Union ino thewearly 9 offered70080g p rabo Countries, Staff Working Paper 288, (Washington, D.C.: World (the Soviet Union in the early 1970s offered about Bank, 1978), pp. 32-50. I 100). Most vocational training, and even higher educa- 4. In Romania, after several reforms of education and training, tion, is carried out by enterprises or under the auspices of a reportedly cost-effective, successful system has been developed. sectoral ministries and tends to be excessively narrow and All vocational schools are jointly administered and operated by a job specific. nearby enterprise and a government institution. The enterprise The administrative compartmentalization of training subcontracts the manufacture of product components to school also reduces quality and raises costs. Enterprises or sec- workshops. However, these must fit into the training programs toral agencies now seek self-sufficiency in key skills be- and modules prepared by the Ministry of Education, which en- sures that training is nor narrowly specialized. See Andreas C. cause trained manpower has been scarce, the assignment Tsantis and Roy Pepper, Romania: The Industrialization of an process does not meet their needs, and they cannot hire Agrarian Economy under Socialist Planning (Washington, D.C.: employees away from other enterprises. They may even World Bank, 1979), pp. 151-64. 126 than enterprises normally do and are less likely to lose doctors would find medical employment, while the oth- economies of scale. They could also fill the void created ers sought retraining or jobs requiring more general skills. by the proposed phasing out of apprenticeship programs By contrast, with administrative allocation, the forecast- (now involving more than 2 million young workers). ing error would be less conspicuous, but enterprises that Other countries, however, in particular both East and really needed a doctor and an engineer would find them- West Germany, have found apprenticeship schemes, with selves with two doctors and no engineer. a minority of time spent in the classroom, to be an effec- RecDgnizing the difficulties of manpower forecasting, it tive way of training skilled workers. would be desirable for postsecondary schooling to offer several routes to any given educational qualification. For Tbe Role of Manpower Planning example, rather than having an indivisible four-year uni- Planning of both the level and the composition of ex- versity course for engineers, an individual might become penditures on (especially advanced) education and train- a "technician" in, say, two years and an engineer in ing in China is linked to manpower forecasting. How- another two. Under this system, the requirements for ever, manpower forecasting elsewhere has proved very both technicians and engineers need be forecast only two unreliable (though less so in projecting needs for teachers years ahead. Revising forecasts annually or biannually and doctors than in other areas). Manpower forecasts in could steer the educational system in line with labor re- countries particularly active in the field some time ago- quirernents. Canada, France, India, Italy, Nigeria, Sweden, the Thc considerable flexibility of higher education in United Kingdom, the United States, and Zambia-were China, achieved for example through correspondence especially inaccurate more than two or three years ahead. colleges and television universities, is offset to some ex- Forecasts were usually based on estimates provided by tent by excessively specialized fields of study. It is less enterprises of the numbers of workers with particular difficu k, for example, to forecast the demand for chemists skills (especially in scientific and technical areas) they than for organic chemists, inorganic chemists, and indus- would require. But only very large firms regularly fore- trial chemists. The difficulties of manpower forecasting cast personnel requirements; moreover, they can only thus strengthen the case for broad curricula, with highly guess at their individual relative shares of the total mar- specialized skills acquired after, not during, formal study. ket, so that the forecasts of individual firms may be quite inconsistent and cannot be aggregated into a forecast for Employment and Productivity the whole economy. More complex manpower planning methods-which The preceding discussion revolved mainly around short- attempt to calculate labor requirements (by occupation, ages. Just as much thinking in China, however, has re- sector, and industry) based on output and productivity volved around surpluses, especially of less-skilled labor. forecasts and to convert them into an educational struc- China's labor force will continue to grow, though less ture through estimates of the level of formal education rapidly in the 1990s than in the 1980s. In absolute terms, required for each occupation-have been no more suc- the numbers are staggering: an increase between 1981 cessful. Converting information about occupations into and 2000 of about 250 million in the population of educational needs is difficult partly because it involves working age, and perhaps 180 million in the labor force, aggregating varying amounts and kinds of formal and requiring on average nearly 10 million additional jobs nonformal education. But the main problem of this ap- each year. proach is its neglect of the possibilities for substitution: Faced with these statistics, and with evidence of exist- depending on prices, wages, and availability, enterprises ing labor surpluses, many people in China are worried produce given amounts of output with widely varying that there could be-also partly because of the PRS- mixes of labor and machinery, of different occupational widespread open unemployment in the countryside, with skills, and (within occupations) of different educational seriouis social and political consequences. There is like- qualifications. As a consequence, extensive research has wise 'ear of urban unemployment, which, for example, been unable to relate output forecasts (which are them- has led the Government in the past five years to cram selves problematical) to educational needs. many young people (especially those who returned from Manpower planning in China is likely to suffer from all the countryside) into enterprises that were already grossly these problems, but the consequences of error could be overmanned-and whose labor productivity has deterio- much more serious, at least if skilled labor continues to be rated as a result. More generally, a strong desire to pre- administratively allocated (discussed later). In other coun- serve and multiply jobs is reflected in reluctance to take tries, inaccurate manpower forecasts are less worrisome, steps that would raise productivity. These legitimate and because both individuals and enterprises can respond important concerns raise many policy issues requiring more flexibly to errors: for example, if 1,000 doctors had careful analysis of both the facts and the principles in- been trained while only 500 were needed, the 500 best volved. 127 Productivizy and Growtb manager may complain that he is obliged to employ far Economic growth is, overwhelmingly, productivity too many veterans, females (especially in heavy industry), growth. Since the ratio of the labor force to the total and children of current or former employees, but also population will change little, growth of average per capita that he is short of competent, well-trained, able-bodied income will be almost entirely determined by growth of workers even at relatively low skill levels. Shortages and average output per worker. And, as in other countries surpluses of specific types of labor coexist in other coun- (chapter 1), although part of the increase in average labor tries, too, but, as with goods, the degree to which this is productivity will come from movement of labor out of so in China is unusual. agriculture into other sectors, most of it must come from The acuteness of the mismatch between labor demand higher productivity within each sector. For these basic and supply in China is partly due to rigid labor allocation reasons, measures that restrict labor productivity growth and wage policies (possible implications for these policies will slow China's overall economic growth. Slower are discussed later in this chapter). But it, and the general growth, moreover, means slower accumulation of capital predominance of labor surpluses in urban areas, also owes and expansion of demand in the modern sectors, and much to the unbalanced sectoral and institutional struc- hence over the longer term less of precisely the sorts of ture of production. employment that the Government is seeking to provide The smallness of employment in the service sectors in by holding down labor productivity. China, especially commerce and miscellaneous business This is not to deny that increases in productivity may and personal services, and the possible economic advan- lead to the loss of particular jobs. Indeed, as discussed in tages of its much faster expansion in the future, have been Chapter 7, rapid and intensive growth depends on this discussed in earlier chapters. The range of future possibili- happening on a large scale, with obsolete or uneconomic ties may be illustrated by comparing Chapter 2's QUA- activities and products being replaced by new and better DRUPLE projection, in which service sector employment ones. Displacement of old jobs by new jobs can take place increases by the year 2000 to 14 percent of the labor largely within existing enterprises-in other countries, force (90 million workers), with the BALANCE projection, employment tends to grow faster rather than slower in in which-with the same growth rate of GDP-it increases enterprises with above-average rates of innovation and to 25 percent of the labor force (155 million workers), productivity growth. But part of it must involve move- most of the increment being in commerce and miscellane- ment of workers from one enterprise to another (raising ous services. Service sector employment in QUADRUPLE possible issues of unemployment, discussed below), in 2000 is only 70 percent of industrial (manufacturing Nor is productivity growth incompatible with the "la- and mining) employment, which is similar to the Soviet bor-intensive" development path that some successful de- Union in 1959 (80 percent). In BALANCE, service sector veloping countries have followed. High labor intensity in employment is 150 percent of industrial employment, the relevant sense is not low output per worker. It is which is the same as Japan in the early 1 950s, though less making the best use of scarce capital, land, materials, and than in the average lower-middle-income country today skills by using them in conjunction with large amounts of (200 percent, which is the product of a similar-size service unskilled labor, thereby achieving the highest level of sector but less industrial employment). output (and hence average labor productivity) that avail- Service sector activities, being more labor intensive able resources permit. In countries that have, like China, (combining more workers with a given amount of capital concentrated scarce resources in capital-intensive enter- and materials), could create more jobs and hence reduce prises, labor productivity in those enterprises has some- the overall urban labor surplus. These jobs would, more- times been high, but average labor productivity in the over, disproportionately employ the categories of labor economy has been reduced. that are now in surplus in urban China. In the United Kingdom in 1970, for example, women made up 52 Urban Employment percent of the labor force in commerce, finance, business, The situation regarding current and prospective urban and miscellaneous services (as compared with 30 percent labor surpluses and shortages in China seems to be re- in manufacturing), partly because of the greater scope for markably varied. Despite general concern about surpluses part-time employment.5 The service sectors in other there is rapid absorption of labor into small rural towns. countries also absorb many males who lack the skills, Some larger towns and cities-for example, Changzhou physical strength, or tolerance of monotony needed for in Jiangsu and Shashi in Hubei-are also experiencing manual work in modern industry. Expansion of the ser- general labor shortages. And the projections discussed in Chapter 5 imply that this will happen more frequently unless greater migration is permitted. 5. In the U.K. data, 36 percent of the women in the service Within enterprises, in localities of general surplus and sectors worked part time, as compared with 22 percent in manu- general shortage alike, the picture is even more mixed. A facturing. 128 vice sectors could thus permit many people who now nonagricultural enterprise activities is of course not con- contribute little or nothing to output in China's factories fined :o urban areas. The degree to which such activities to move to more productive work. So could expansion of expand will also have a powerful effect-probably as transport (see Chapter 5), especially road transport, great an effect as the overall pace of growth-on the which is much more labor intensive than rail transport. number of workers in agriculture, which is likely to re- Closely associated with China's presently unbalanced main the residual employer of labor over the next few production structure is the smallness of the individual decades. sector-including self-employment, partnerships, and sole In the QUADRUPLE projection, with continuing rather proprietorships with a few workers-which in other slow growth of the service sectors, the proportion of the countries plays a dominant role in commerce and miscel- labor force in agriculture declines from 70 percent in laneous services, although it is important also in industry, 1981 to 59 percent in 2000; the number of agricultural construction, and transport. In China in 1984, this sector workers increases by 56 million. In the BALANCE projec- employed about 2.5 percent of the urban labor force. By tion, with faster growth of the service sectors, employ- comparison, to take some representative examples of cit- ment in agriculture drops to 52 percent of the total, but ies in other developing countries, the equivalent "infor- the number of agricultural workers still rises by 13 mil- mal sector" is estimated in the early 1970s to have em- lion. This is a normal pattern by international standards. ployed 31 percent of the urban labor force in Abidjan, 40 A marked decline in China's agricultural labor force over percent in Jakarta, and 3 0-40 percent in Lima.6 Even in the next two decades could be achieved only if nonagri- other socialist countries, the proportion is much higher cultural labor productivity grew unsatisfactorily slowly than in China: for example, the proportion of the indus- (Chapter 2). trial and industrial-handicraft labor force employed in Thus although the agricultural labor force could shrink private, semiprivate, and cooperative enterprises in the rapidly in the twenty-first century (Chapter 3), surplus early 1970s was 8 percent in Romania, 11 percent in East rural Labor in the next twenty years is probably unavoid- Germany, 17 percent in Hungary, and (in 1980, with able. The ratio of farmworkers to cropland in China is services included) 29 percent in Yugoslavia.7 very ligh indeed by international standards, with Chi- It is increasingly recognized in China that many ser- nese 5ources generally estimating that about one-third of vices and specialized products are more efficiently pro- the agricultural labor force is superfluous.9 In considering vided by small, flexible enterprises and that such eco- nomic activity complements, rather than threatens, the state sector. Since 1980, young job-seekers have been 6. Heather Joshi, Harold Lubell, and Jean Mouly, Abidjan: officially encouraged to organize small collective enter- Urban Development and Employment in the Ivory Coast (Geneva: prises. Self-employment and the development of family International Labour Office, 1976): S. V Sethuraman, "The Ur- enterprises have been promoted, and the taking of ap- ban Informal Sector: Concept, Measurement and Policy," Interna- prentices and hiring of labor (up to eight employees) have tional labour Review, vol. 114, no. 1 (July-August 1976). pp. 69- been permitted. Although still very small, the individual 81; C. Wendorff, "El Sector Informal Urbano en el Peru: sector has grown rapidly. In 1984, privately owned in- Interp^etacion y Perspectivas" (paper presented at a seminar on dustrial andcomeriaunerakigsememployment problems in Peru, Catholic University, Lima, Peru). dustria adcmecaunetkn7 The figure for Romania excludes small-scale private firms than 13 million people, of whom about 3 million were in and is therefore an understatement. The figure for Yugoslavia urban areas, a twentyfold increase over 1 9 7 8. The major- includes transport, communications, catering, tourism, construc- ity are in commerce, where the Y 28.8 billion turnover tion, and artisan work. David Granick, Enterprise Guidance in East- was 8.6 percent of total retail sales.8 ern Eirope (Princeton, NJ.: Princeton University Press, 1975), p. Further expansion seems desirable. Indeed, to take 486; lHarold Lydall, Yugoslav Socialism (Oxford: Clarendon Press, fuller advantage of the employment-creating and produc- 1984). p. 268. tive potential of individual and small collective activity in 8. China Daily. March 16, 1985. China would require few positiv . actions, since the sector 9. Labor use in China is very high, even compared with other China would require few positive actions, since the sector parts of East Asia, which in turn is usually greater than elsewhere tends to grow spontaneously to meet demand once re- in the world. For example. Japanese agriculture used nearlv 525 strictions are removed. It would be necessary, however, man-days per hectare in 1956, and well over 600 man-days in to further improve its access to premises, materials, some areas. See K. N. Raj, "Preface," in Labor Absorption in Asian skilled labor, and credit and to make sure that the work- Agriculture, by Shigeru Ishikawa (Bangkok: International Labour ers involved have access to public services such as hous- Office-Asian Regional Team for Employment Promotion, Asian ing, health care, and pensions. (Policy toward the individ- Emplcyment Programme, 1978), p.v. A recent study of thirty ual sector is discussed further in Chapter 10.) production teams in Nantong County, Jiangsu, suggested that a wheat-cotton rotation required 92.3 labor-days per mu (1.384 per Underutilized Rural Labor hectaro) and that a wheat-rice rotation needed 77 labor-days per mu (1,1 55 per hectare). See Song Linfei, "Village Labor Surplus The scope for expansion of service sector and individual (continuea) 129 what to do about this, it may be unhelpful to think of organizational problems of public works. The difficulties some specified proportion of agricultural workers as sur- encountered with extensive rural mobilization during the plus, the rest as necessary. Instead, the problem could be Great Leap Forward demonstrated the limits to this ap- viewed as most agricultural workers making a small incre- proach, mainly because workers were diverted from agri- mental contribution to production-or, as Western econ- cultural production. During the Cultural Revolution, ex- omists would say, generally low marginal productivity. cessive emphasis on large projects that did not justify the Because farmers are poor and ingenious, this contribution effort involved also led to a reaction against labor-inten- is rarely zero-if it were, parents would not (as they are sive construction methods. But aside from these two pe- now doing in China, and have always done in other riods, China has effectively used rural labor in the slack developing countries) keep their children out of school to season for such things as the maintenance and upgrading work on the land. Marginal productivity may even have of irrigation facilities and terracing. The obligation to been increased in recent years by the PRS-for example, perform "social labor" is a tradition that preceded the cotton used to be picked only once, but is now picked People's Republic and has not been ended by the PRS, repeatedly. And the continuing reorientation of the sec- although it can now often be commuted into a tax pay- tor away from grain toward other crops and animal hus- ment. bandry may raise the marginal productivity of agricul- Neither seasonal public works nor the general utiliza- tural labor over the decades ahead. But it will in most tion of very labor-intensive methods in rural construction places and activities remain exceedingly small. is likely to be a panacea for underutilized rural labor. But This does not necessarily mean low or stagnant agricul- China could, over the next two decades, use labor-inten- tural incomes. For although in other countries low mar- sive public works to improve rural infrastructure, espe- ginal productivity would mean low wages for landless cially roads. Even the roads linking substantial towns and laborers, peasant incomes in China depend instead cities are often poor, and the even lower quality of the mainly on average productivity-the amount produced rural road network poses a serious obstacle to economic on the household plot divided by the number of house- development (Chapter 5). The abundance of underuti- hold members. In some localities, and for some house- lized agricultural labor could, with proper management holds, average productivity also will be very low. But for and planning, provide an opportunity to tackle this prob- the agricultural sector as a whole, average productivity is lem, while contributing directly and indirectly to increas- likely to increase quite substantially over the next two ing rural incomes. decades (Chapter 3). From the viewpoint of efficiency, though, agricultural Allocating and Motivating Workers labor with very low marginal productivity is an underuti- lized resource. It has often been argued in other develop- To realize the full potential of China's human resources, ing countries that utilizing such labor for rural public it will be essential not only to provide appropriate educa- works could increase output and investment, as well as tion and training and to create the right jobs, but also to raising the incomes of some of the poorest members of society. There has also been much skepticism about this possi- bility, Although a few rural construction projects could blt.lhuh ferrlcn i pand Its Outlet," Chinese Social Science, no. 5 (1982), pp. 121-33. use unskilled labor and need very simple tools, most rural Even with this intensity, however, and with a significant amount of investment needs other equipment. Using labor-intensive sideline and collective enterprise activity, this particular county had methods can also be very slow, especially if the work is a great deal of surplus labor. There were only 1.6 mu per head of only done in the off season. Partly finished projects are the agricultural labor force; the study argued that about 4 mu per like unutilized capital assets-it may be worth using worker would be needed to avoid surplus labor. This is actually a higher-cost methods to complete them quickly. Some substantially higher estimate of labor requirements than many oth- ers used in China. For example, forecasts of national agricultural projects might be completed by using very large amounts manpower requirements by the Ministry of Agriculture, Animal of labor at once, but this requires considerable organiza- Husbandry and Fisheries use an average cropping intensity of 9 mu tional skills, which are not costless. Furthermore, labor (0.6 hectare) per worker in crop production as an estimate of full participation in public works projects requires the pay- employment. On official estimates of the cultivable area, this ment of reasonable wages, at least to justify the travel and would suggest that crop production itself could fully employ only work effort involved. If this is not to be inflationary, about 160 million workers. additional taxes need to be collected. In short, these pro- 10. See Shahid J. Burki and others, Public Works Programs in posals for using surplus labor require investmentdecisions Developing Countries: A Comparative Analysis, Staff Working Paper 224 (Washington, D.C.: World Bank, 1976); and Basil Coukis and based on careful evaluation of the benefits and costs in- others, Labor-based Construction Programs: A Practical Guide for volved.'° Planning and Management (New York: Oxford University Press, China has made considerable progress in solving the 1982). 130 be successful in matching workers to jobs and getting able; they are still obliged by the labor bureaus to accept them to perform well in those jobs. China's traditional "packages" containing both wanted and unwanted re- system of allocating and motivating workers has serious cruits; they cannot discharge redundant workers; and shortcomings, but changes to achieve greater economic they have been allowed to dismiss unsatisfactory workers efficiency will raise difficult issues, especially of equity. In only in a few instances, involving extreme absenteeism or agriculture, fundamental and highly beneficial changes in malfeasance. It appears imperative to consider further, the system of motivation have already tied household more radical, changes. earnings to household effort and initiative. This section will thus focus on labor allocation and motivation in INDI'VIDUAL CHOICE OF JOBS. A fundamental weak- nonagricultural and urban activities, which will predomi- ness of the present assignment system is its failure to take nate over the longer term. account of the the fact that individual job performance is as much a matter of individual interest and motivation as Employment Cboices ofEnterprises and Workers of quLlifications. Nor is monetary reward the only China's system of labor allocation allows individual em- motivator of effort and initiative. Freedom of choice ployers and employees uniquely little freedom of about where to work not only increases individual satis- choice-far less even than in the Soviet Union and East- faction, but also productivity, since people are likely to be ern Europe. Until recently, all young people were admin- more enthusiastic in a job they have chosen to do. There istratively assigned to particular jobs-college graduates are therefore many advantages to allowing all workers, by the central government, secondary school graduates skilled and unskilled, new graduates and experienced peo- by local labor bureaus-with little attention to their pref- ple, to choose and change jobs freely. erences or the preferences of employers. The assignment Most governments assign graduates only for compel- was typically for life: with few exceptions (generally dic- ling reasons. For example, a few with training in much- tated from above rather than a result of individual or needec skills might be asked to work in remote areas for employer preferences), workers could not move from a fixed period. Centrally planned countries assign some one enterprise to another. Enterprises, moreover, were graduates to their first job: for example, Hungary assigns not permitted to discharge workers, even if they had doctors and teachers; its chief manpower planner found more employees than they needed, and even if particular the idea of extending this procedure to all graduates "in- employees were habitually absent, lazy, or negligent. conceivable."'' Hungary, Romania, and the Soviet In the last few years, there have been some changes. Union all allow some choice of assignment, at least to the The system of unified assignment has been dropped for best stuidents, which provides an incentive for good per- youths with limited training, and enterprises may now formarice at school. Furthermore, these assignments usu- examine prospective employees assigned to them by labor ally last only three years. Many workers then choose bureaus. Some jobs are now advertised. It was recently another job (perhaps because they were poorly assigned, announced that the best graduates of two universities or perhaps because the assignment process is used to fill (Qinghua and Jiaotong) may choose their own employ- inherently unattractive positions). In the Soviet Union, ers. From mid-1984, an experimental "job invitation" the first year of assignment became a probationary pe- scheme has allowed a few organizations to offer jobs riod. The enforcement of assignments, which was once which they may later terminate to particular individuals, very strict, is now rather lax. the individual also having the right to refuse the offer or Allowing experienced workers to change employers subsequently to withdraw. Another experiment, covering also has advantages. It diffuses technological knowledge 300,000 workers in early 1984, involves "job contracts" (Chapter 7) and reduces overspecialization. Workers can between enterprises and new employees, often for three move :o jobs where their particular skills will contribute to five years, specifying mutual obligations in regard to more io production than in their present jobs. Further- wages, performance standards, contract renewal, and sev- more, people like variety and are often challenged and erance. Enterprises are now in principle allowed to dis- stimulated by new situations. (At present in China, the miss unsatisfactory employees. Finally, there has been employing unit has to give permission for an employee to rapid growth of the commune and brigade and individual move: it will obviously consider its own needs more than enterprise sectors, where employment decisions are sub- those of other enterprises or the whole economy and is ject to little administrative regulation. likely :o deny permission to those with the greatest en- These and other reforms, though beneficial, have not ergy and aptitude for tackling a new job.) Besides, work- basically changed the system. The unified assignment ers forced to remain in jobs where they are bored or scheme is still used for virtually all those with any postse- condary training and for some skilled manual workers. Enterprises can seldom hire badly needed skilled workers 11. Jinos Timar, Planning the Labor Force in Hungary (White from other enterprises where their skills may be less valu- Plains, N.Y.: International Arts and Sciences Press, 1966), p. 137. 131 unhappy are generally unproductive. Even large Japanese of workers threatened with dismissal. But the virtual im- enterprises, which typically provide lifetime employ- munity of employees from dismissal in China is unique, ment, though only to about a quarter of Japan's labor and the comparative experience of Hong Kong and for- force, encourage periodic reassignments among different eign businessmen who have established enterprises in departments. They also oblige their workers to "re- China abundantly confirms that it substantially reduces tire"-usually to other jobs-at about age fifty-five. worker effort and initiative. Positive incentives that kin- A possible disadvantage of individual freedom in job dle enthusiasm for work and reward diligence are essen- choice could be that the resulting pattern of employment tial. But it is vital to complement them with the threat of was not in accordance with national economic priorities, dismissal for poor (and not merely atrocious) perfor- particularly because many skills are now extremely mance. scarce. This is indeed a real danger in China's present Inevitably, some dismissals will be unfair. Because of circumstances, but mainly because enterprises are not yet mistaken perceptions or personal dislike, employers subject to appropriate motivation and economic signals. sometimes dismiss satisfactory workers unless they have Without enterprise responsibility for profit and loss, to go through a prohibitively elaborate administrative or product market competition, or rational prices, employer judicial review process. Moderate safeguards, such as the willingness to pay to secure or retain particular employ- right to appeal to a tribunal containing some worker ees would be a poor guide to their potential economic representatives, would be essential, but could not avoid contribution (the changes in the wage system that might all injustice. be required are discussed below). But these problems Dismissal will also often cause economic hardship- should be corrected through wider system reform, which, unemployment and loss of income-which, though fair except for a few projects of great scientific or strategic for the idle or negligent worker, is not so for his children. significance, could make it unnecessary and inefficient for For if dismissal involves no hardship, it loses most of its the Government to intervene directly in the allocation of force, as, for example, in the Soviet Union, where labor. chronic labor shortages mean that dismissed workers The other possible disadvantage of giving individuals have little difficulty in getting other jobs. Even in indus- more freedom to choose and change jobs is excessive trial capitalist countries, where a dismissed worker risks labor turnover, which could disrupt production and pre- protracted unemployment and may be ineligible for un- vent employees from acquiring enough experience in par- employment insurance benefits (which are commonly ticular jobs, as well as deterring employers from offering confined to those who lose their jobs through no fault of training (unless subsidized or obliged to do so). Though their own), he would usually be eligible for some form of there is controversy about its extent and causes, labor public assistance, especially if he had dependent children. turnover in the Soviet Union is generally believed to be Some such compromise between making dismissal a high by international standards and to impose substantial toothless sanction, and making it distressingly onerous, economic costs. By contrast, the Japanese model of mu- would have to be struck also in China. tual lifetime commitment between worker and employer Policy toward workers an enterprise would like to dis- in large enterprises is widely believed to have contributed charge (or reject), not because they are idle or negligent, to the country's economic success. Nevertheless, substan- but because they are simply not needed for production, tial labor turnover in most countries does not seem to involves yet harder choices. Indeed, even the principle have been particularly disruptive. Even in Japan, 13 per- that enterprises should be allowed to do this (or, more cent of all workers leave their enterprises each year, and drastically, to close down altogether) is not generally ac- South Korea has been economically successful despite cepted in China, mainly because it would inevitably lead very high rates of turnover, especially in the labor-inten- to open unemployment. This is seen as wasteful of hu- sive export sectors. man resources, and also inequitable because those who lose their jobs may have been conscientious workers ENTFERPRISE CHOICE OF WORKERS. The possible ad- (though clearly employers will let the least diligent go vantages of allowing enterprises to compete for workers first). It would also inflict unusual hardship in China, (new graduates or workers in other enterprises) have al- where there is at present not only no unemployment ready been mentioned. There would also seem to be insurance, but also workers depend on their enterprises advantages in allowing enterprises freely to reject poten- for housing, pensions, much education and medical care, tial recruits they do not want, to release redundant work- and so on. These objections deserve careful consider- ers, and to dismiss unsatisfactory workers. ation. Dismissal of unsatisfactory workers is already an ac- Long-term, large-scale unemployment is unquestiona- cepted principle in China. The issue is, rather, what con- bly wasteful, especially where, as at present in the indus- stitutes unsatisfactory performance. This varies from trial capitalist countries, large numbers of young people country to country, as do the legal and procedural rights are affected. In China, moreover, where the actual wages 132 of many urban workers may be above their potential The arguments for and against allowing enterprises to contribution to production in alternative employment, release or reject redundant labor hinge ultimately on the employers might on the basis of profitability calculations probability of released workers finding new jobs, paying want to release more workers than would be rational acceptable wages, within a short time. This probability- from the viewpoint of economic efficiency. For example, which, if high, would make unemployment much more if the wage were Y 100 per month, profit-oriented enter- acceptable-should not be underestimated in China. prises would release all workers contributing less than Y Many workers released by one enterprise would be able 100 to monthly production, even though the alternative to find work in other enterprises simply because em- jobs for these workers might raise production by only Y ployer needs for particular types of labor are now poorly 50-thus clearly lowering total national production. (The matched with labor allocations. Other released workers implications of this for China's wage policy are discussed would find work after retraining in accordance with the below.) revealed pattern of shortages and surpluses of particular Nonetheless, the waste associated with unemployment skills (which is at present hard to ascertain). The elimina- has to be weighed against the waste associated with the tion of old jobs and inefficient enterprises actively con- present employment system. To some extent, this system tributes to the creation of new jobs and the expansion of simply transfers unemployment from the streets into the employment in efficient enterprises (Chapter 7). There is factories, which is no less wasteful. The resulting sense of great potential for creating productive jobs in the service labor surplus in many enterprises, moreover, discourages sectors and the individual economy. managers from improving work habits and increasing the But even under the most optimistic assumptions, free pace of work. The rigidity of the system is also an obsta- release of labor would cause an appreciable number of cle to long-term technological advance and productivity unlucky individuals-and localities-to suffer long-term growth, both by impeding the replacement of old prod- unemployment, or (as is more usual in developing coun- ucts and processes with new ones (Chapter 7) and by tries without unemployment insurance schemes, and con- obstructing the movement of particular workers to jobs sequently with little open unemployment) to experience in which they could be more productive. For these rea- sharp wage reductions. The increases in unemployment sons, allowing enterprises to release or reject redundant in industrial countries since about 1970 are not encourag- workers could on balance lead to fuller use of China's ing: although unemployment rates are still only 2-4 per- human potential, despite transitory unemployment. This cent in Austria, Japan, Norway, and Sweden, they are would be particularly likely if the state provided displaced over 1 3 percent in Belgium, the Netherlands, Spain, and workers with retraining and assistance in finding suitable the United Kingdom (where 5 percent of the labor force new jobs. has now been without jobs for more than one year). There are also ways of reducing the hardship that un- There is also appalling urban poverty in some developing employment would otherwise impose on workers re- countries. Thus, even though most unemployment in leased or rejected by Chinese enterprises. These, which other countries is in fact temporary, freedom for Chinese are discussed in Chapter 10, include establishing a system enterprises to release or reject redundant-as distinct of unemployment insurance and transferring much re- from idle or negligent-labor would have both advan- sponsibility for housing, pensions, and social services tages and risks and should probably be approached cau- from enterprises to the Government. However, setting tiously and gradually. the level of unemployment benefits (including housing and social services for unemployed workers) involves an Efficiency, Fairness, and Stability in Wage Determination insoluble dilemma. If these benefits are set low, relative For most of the past three decades, China's wage system to wages, then people who become unemployed through has been just as rigid as its labor allocation system- no fault of their own, and their families, suffer unaccepta- though more similar to the Soviet and East European ble hardship. But if they are set high, people have little systems. There are centrally determined basic wage incentive to work and may even prefer to be unemployed scales, varying among occupations, industries, and locali- rather than take, to gain a little extra money, a monoto- ties. Enterprise managers have had virtually no discretion nous or arduous job. The strength of this disincentive to over where to place individual workers on, or how fast to work remains controversial. But most economists agree move them up, these scales-indeed, individual wages that generous unemployment benefits, both by providing might be frozen for a decade or so, causing, for instance, a disincentive to work and by raising wages in low-skill a rapicly promoted minister to be paid less than a longer- jobs, have contributed significantly to the high levels of serving junior official. Bonuses, piecework, and other long-term unemployment that now afflict, in particular, forms Df payment by results were for many years prohib- some West European countries. These benefits have thus ited. Moreover, a large fraction of worker remuneration also changed unemployment from a problem of poverty has be,en provided in kind (housing, social services, subsi- into one mainly of wasted human resources, dies), rnore or less equally to all employees. 133 In recent years, there have been some changes, most ther steps toward a more flexible, less egalitarian wage notably the reintroduction of piecework and bonuses, system, in which demand and supply, as well as individual which now have no upper limit for individual workers, contributions to production, would play a greater role in though enterprises making payments above specified lev- determining the earnings of particular employees and oc- els are in principle subject to stiff taxes. Under various cupations. It must be stressed that in few other countries "floating wage" schemes, most or all of the basic wage are wages freely determined by market forces: even in follows the standard eight-grade scale, but bonuses and a capitalist nations, there is usually some mixture of market part (usually 20-30 percent) of the basic wage are linked forces, collective bargaining by trade unions, minimum to the profit or output of the firm or to performance wage legislation, and-within large organizations-ad- targets for the individual job.'2 A related scheme is the ministered wage scales influenced by considerations of "floating wages and grade system," under which a hierarchy and fairness. (The rigidity that these nonmarket worker is promoted for good performance but can only elements introduce into the wage structure is widely retain the higher grade after three years of sustained good thought to have contributed to both unemployment and performance. Some state enterprises set aside up to 1.5 inflation in the industrial countries.) Moreover, for percent of the wage bill for "labor emulation" or "labor China, the economic benefits of a more flexible wage competition" awards, and honorific titles and privileges system would be relatively limited without more general are offered for good work performance or innovation. In reforms of enterprise motivation and prices, which would addition, there has been rapid expansion of commune align the financial interests of enterprises more closely and brigade and individual enterprises, which have con- with the economic interests of the nation. siderable freedom to set their own wages. Objections to more wage flexibility-which could be As with labor allocation, however, these reforms have accomplished partly by reform of the state enterprise not basically changed the system. Bonuses have become wage system, partly by relative expansion of the less regu- general wage supplements, distributed to workers and lated nonstate enterprise sector-are mainly connected staff in almost all organizations, often with little regard to with equity. Although the Communist Party's Central the performance of either the organization or the individ- Committee has recently stressed that China needs more ual. Managers still have little discretion regarding promo- than the present degree of wage inequality, " both very tion (or demotion). Floating wages apply to a tiny minor- high and very low earnings may be considered contrary ity-inJiangsu in early 1984, only 1.4 percent of workers to socialist principles. Another concern is that more flexi- were covered by such schemes. Moreover, even with bility might lead to wage inflation, with adverse conse- floating wages, the main determinant of earnings remains quences for the budget or the overall price level. These the centrally prescribed wage scales, which have changed possible disadvantages of further substantial reform of very little and are most egalitarian by the standards of China's wage system merit careful consideration. other countries, especially as regards the smallness of the differentials between manual workers and technical, man- HIGH WAGES. It is essential to distinguish between agerial, and professional staff. two quite different forces that would tend, with a greater These continuing characteristics of China's wage sys- role for market regulation, to increase the relative pay of tem cause serious economic problems. It is hard for enter- certain individuals and occupations. The first is the na- prises to reward good work-including, very importantly, ture and magnitude of their work: jobs that are particu- that of their managers-with higher pay or promotion, or larly arduous or unpleasant, or involve disagreeable risks to penalize poor work with lower pay. The low, fixed or responsibilities or long training, must be better paid pay of skilled or other scarce labor contributes to misallo- than other jobs requiring comparable skills to persuade cation and waste, because employers who need such la- enough people to work in them (this is true even now in bor cannot attract it, and because wages that do not China, where mining and remote or hazardous jobs com- reflect its economic value offer no disincentive against mand a wage premium); and people who work longer employing such labor unnecessarily. Conversely, the rela- hours, or faster, or more conscientiously will tend to be tively high wages that must be paid to unskilled workers paid more. These causes of high earnings seem fully com- are a deterrent to their employment and would become even more so if employers were given more freedom of 12. See, for example, Qiu Yang, "On the Basic Direction for the choice in hiring and firing. More generally, the rigidity of Reform of the Wage System of Enterprises,"' Jingi Guanli, no. 9 the wage structure is an important cause of the striking (September 5, 1983), pp. 20-23 (Joint Publications Research Ser- coexistence, mentioned earlier, of labor shortages and vice 84603, October 24, 1983). Also Zhuang Qidong, "Com- coexlstence,~~~~~~~~~~ ~ entsone onlr Flatn Wages Beingge Tried Ou in Chn, Rem surpluses in urban China. (The problems caused by lim- Rmebnts on Floatmg Wages Be9g Tried Out n Chia Renmin ited interregional differences in wage levels and structure 13. Communique of the Third Plenary Session of the Twvelfth were discussed in Chapter 5.) Central Committee of the Chinese Communist Party, October 20, The Government is therefore actively considering fur- 1984, Decision on Reform of the Economic Structure, Article VII. 134 patible with the socialist principle of payment according ary burden of subsidies to overmanned loss-making enter- to work. prises is also a disadvantage. The second source of high wages in a market setting is If un;skilled wages were allowed to float downward, the scarcity. This may arise from a limited supply of innate problerns of employment associated with a more flexible talent-outstanding artists, sportsmen, and businessmen system of allocating labor, discussed earlier, would be being examples. Or it may arise from bottlenecks in the reduced: profit-oriented employers would want to release education and training system, which cause shortages of fewer unskilled workers than at the present, higher, certain sorts of labor-at least for a while-even though wages; and it would be easier for those released to find more than enough people would be willing to undergo employment elsewhere. But this gain, and other associ- the necessary training. Unexpected changes in demand ated increases in economic efficiency, could involve a and supply-including those associated with other aspects socially and politically unacceptable drop in unskilled of system reform-can also cause temporary shortages of wage levels. There are various ways of easing this prob- particular types of labor in particular places. These sorts lem; ncne offers a perfect solution. of scarcity tend to generate high earnings, which, while One is to move gradually, so that there is little or no they may be economically efficient, are much harder to absolute reduction of unskilled wages even though they defend as fair. However, progressive personal income tax- decline in relative terms. This could be done on an indi- ation can help to reconcile efficiency with equity by driv- vidual basis, with the guarantee that workers' wages will ing an increasingly large wedge between the wage paid by not be reduced, even though new recruits will be paid the employer (whose high level encourages efficient use less. This would, however, create tension between "old" of scarce skills) and the wage received by the employee. and "new" workers. Another approach would be to es- China already has a progressive personal income tax, tablish a legal minimum wage at the present unskilled aimed primarily at resident foreigners; this could be more level, anid then (this would be the vital but difficult part) widely applied as greater wage flexibility is introduced. (A not to r aise it in real terms for perhaps twenty years, even progressive payroll tax paid by the employer on wages though average earnings-especially those of skilled above a certain level could achieve similar results.) But workers-were increasing. For example, if the unskilled the progressive tax would have to be levied on high earn- wage were initially three-quarters of the average wage, ings due to effort as well as on those due to scarcity, since and the average wage increased at 4 percent per year (as there is no administratively feasible way of discriminating in the QUADRUPLE projection in Chapter 2), a constant between them, and could thus deter hard work and initia- absolute unskilled wage would, after twenty years, have tive. For this and other reasons (discussed in Chapter 9), declined to one-third of the average wage. the personal income tax rate structure in China, as in all A re.ated strategy would be to increase the cost to countries, would have to be a compromise between con- employers of other inputs to production, thus making use flicting objectives. of unskilled labor relatively more attractive. Obvious ex- amples are increases in the cost of capital (higher profits LOW WAGES. The conflict between principles of effi- or taxes on machinery, higher interest rates, fewer ciency and of equity is generally sharper, and harder to grants), energy, raw materials, and of course skilled labor. resolve, for low relative wages than for high relative But beyond a certain point, such increases might be eco- wages. In China, as in other developing countries, un- nomically irrational, although they could perhaps be skilled labor is in such abundant supply that profit-ori- complemented with targeted payroll subsidies to employ- ented employers may only be willing to pay a wage ers of low-wage labor. (Payroll taxes or employer social below the minimum necessary for a decent existence. insuranse contributions on unskilled wages would of China's approach to this problem thus far has consisted course have the opposite effect.) mainly of keeping unskilled wages above their economic The association between low-wages and household level while obliging employers to take on more unskilled poverty is weakened by varying labor force participation. (and other categories of surplus) labor than they want. Some low-wage earners are members of households with This approach is not altogether inefficient, since many of reasonable average incomes, either because there are few the "unwanted" workers in state enterprises are engaged dependents or because other family members earn higher in low productivity activities, much as if they had been wages. Conversely, some households are poor even voluntarily hired at a lower wage. But it has some serious though their working members earn reasonable wages, disadvantages in the context of wider economic reform: it because they have many dependents. Rather than at- runs counter to the spirit of making enterprises manage tempting to boost or supplement low wages as such, it their own affairs and cut costs; and it makes it hard for may thus be more equitable-and more efficient-to pro- state enterprises to compete effectively in product mar- vide direct state income supplements to poor households. kets with nonstate enterprises, which in turn encourages This possibility-and some of its other advantages and administrative restrictions on competition. The budget- disadvantages-are discussed in subsequent chapters. 135 WAGE STABILITY. Recent experience in China, as well reduced by poor accounting and managerial indifference as in other socialist and capitalist countries, substantiates to financial results. the concern that a more flexible wage system might cause In addition, formal linkages between worker remunera- excessive increases in average and total wages-and hence tion and enterprise (as distinct from individual or group) budgetary or macroeconomic management problems. In performance should be approached with caution. This is capitalist countries, the main sources of wage-push infla- partly because irrational prices and limited competition tion have apparently not been market pressures, but insti- make profitability a poor measure of economic perfor- tutions such as collective wage bargaining and legal mini- mance at present in China, partly because alternatives mum wages. These can not only push up wage rates measures such as labor productivity are fraught with am- despite balance or surplus in labor markets, they also biguities of definition and calculation. But it is mainly make it difficult for any wage rate to fall in money terms, because there is no way to determine how much of the so that relative wage adjustments can be achieved only variation among enterprises in the level or growth of through wage increases. In socialist countries, workers profits or productivity is due to variation in worker effort also exert pressure on managers to raise wages, and man- and initiative, as opposed to variation in technology or agers are often disinclined to resist, partly for ideological managerial skill, which are generally more important. reasons, partly because they are not concerned to hold Some improvements in enterprise performance, more- costs down. As a result, there has always been some form over, are simply a reflection of unacceptably poor prior of wage control from above in the Soviet Union and performance. Larger wage increases for workers in enter- Eastern Europe, although its form has varied. prises whose profits or productivity increase more rapidly If China were to move toward a more flexible, market- are thus often (and justifiably) perceived as unfair by regulated, wage system, control of wages in the individ- workers in other enterprises, who then exert pressure on ual and collective sectors would probably not be needed their employers to secure similar increases by manipulat- for overall wage stability (although wage bargaining by ing the statistics and bending the rules. (For example, trade unions is a potential problem). But in the state rules relating wage increases to enterprise productivity sector, unless or until managers are made highly profit- increases seriously undermined an overall policy of in- conscious (see Chapter 10), wage control from above will come restraint in the United Kingdom in the late 1960s.) still be necessary. This is partly because workers in large Current government proposals to establish formal link- organizations can exert collective pressure for wage in- ages between average remuneration and financial perfor- creases, partly because, as the Hungarians have discov- mance (tax or profit remittances) in individual state enter- ered, it may be politically inexpedient to encourage con- prises could therefore cause problems. An alternative flict over wages between the managers and workers of approach might be to have rather uniform guidelines for state enterprises. In China, the case for continuing state average wage increases in state enterprises, related to an sector wage control is further strengthened by the large overall strategy regarding wage differentials between the size of the present gaps between wages (and benefits) in state sector and other sectors (including agriculture-see state and nonstate enterprises and between urban and Chapter 3), as well as to economywide labor productivity rural incomes, which should probably be gradually nar- increases. These overall limits should as far as possible be rowed. combined with stronger linkages within enterprises be- The precise form of wage control is itself a difficult tween the remuneration and performance of particular issue. The Soviet Union and Eastern Europe have moved individuals and groups, and with greater flexibility in increasingly toward limits on the enterprise's total wage setting the pay of scarce, skilled categories of staff and bill, leaving managers considerable discretion over how workers. But there will inevitably have to be compro- many workers of particular types to employ and how mises between the need for continuing administrative much to pay them, and toward linking the overall wage restraint of total wage bills and the need for more micro- bill to measures of the enterprise's economic perfor- economic flexibility and incentives. Moreover, the mana- mance, such as growth of net output. Such an approach gerial indifference to profitability that makes administra- could have advantages for China, too. But it would be tive restraint of total wages necessary is obviously also an desirable to retain limits on growth of average (per impediment to establishing an economically efficient and worker) earnings in each enterprise, as well as on the total nonegalitarian distribution of the wage bill within partic- wage bill. Workers and managers may, otherwise, find it ular enterprises. in their interests to restrict or reduce employment to an economically undesirable level, which would have more serious consequences in China than in the labor-short Soviet Union. These limits on earnings increases should, China's low rate of population growth, by comparison moreover, be enforced administratively, and not merely with most other developing countries, is one important through tax penalties, whose efficacy in China is much reason for China's unusually favorable per capita income 136 growth prospects (Chapter 2). It has been achieved Figure 8.1 Population, Alternative Projections, through widespread female literacy and primary health 1980-2;100 care (see footnote 2 in this chapter), as well as strong birth planning policies. The Government's target for the Millions of people year 2000 is a population of 1,200 million. This will 1i800 - require a substantial drop in rural fertility from its present _ A level, which will not be easy to achieve. But the target 1,600- seems broadly feasible, as well as desirable, and its attain- B ment has been assumed in the various economic projec- 1,400- tions presented in earlier chapters. A The focus of the present section, however, is on demo- 1,200- /__ graphic and related issues in the twenty-first century. Its objective is not to offer predictions or prescriptions, nor 1,000- even a comprehensive account of possible outcomes and "-. C their determinants and implications, but rather to draw 800- attention to certain issues that may need to be considered i98( 2000 2020 2040 2060 2080 2100 and acted on within the next few years. Figure 8.1 and Table 8.2 present and extend to the year Source. World Bank projections. 2100 a demographic projection (B) which almost exactly attains the Government's population target for the year 2000, as well as two alternative projections (A and C).`4 Projection C reaches the same low fertility rate as Projec- In Projection B, fertility falls below the replacement level tion B in the early 1990s and maintains it until the year during 1985-90, reaches a low point in the first half of 2040, when it returns to replacement level. Its popula- the 1990s, and rising again, remains at replacement after tion is 1,180 million in the year 2000, falling to about the year 2000. Its population of 1,196 million in the year 800 million by the time it stabilizes. (There are no good 2000 stabilizes by the year 2 100 at an unchanging level historical precedents for such a decline, except in condi- of about 1,500 million. Projection A assumes a rather tions oi acute economic distress, as in nineteenth century slower decline of fertility, never falling below replacement level; but it, too, assumes that replacement fertility will be achieved in the year 2000 and then maintained indefi- 14. These projections follow the World Bank's standard meth- nitely. Its population of 1,273 million in the year 2000 odology. See My T Vu, World Population Projections I984 (Wash- eventually stabilizes at about 1,700 million. In contrast, ington, D).C. World Bank, 1984). Table 8.2 Population Growth, Alternative Projections, 1980-2100 Projection A Projectt n B Proiection C Average Average Aferage annual annual annual Population growth Population growth Populationi growth Year (millions) TFR' (percent,)[ (millions) TFR (percent' (millions) TFR' (percent)" 1980 980.3 - - 980.3 -- - 980.3 - - 1990 1,112.4 2.3 1.3 1,093.8 2.1 1.1 1,093.8 2.0 1.1 2000 1,273.0 2.2 1 4 1,196.3 1.7 0.9 1,179.7 1.5 0.8 2010 1,392.1 2.2 0.9 1,3 12.2 2.2 0.9 1,232.7 1.5 0.4 2020 1,495.0 2.2 0.7 1,379.4 2 2 0.5 1,245.9 1.5 0.1 2030 1,583.4 2.1 0.6 1,436.8 2.1 0.4 1,224.8 1.5 -0.2 2040 1,626.3 2.1 0.3 1,461.1 2.1 0.2 1,158.1 1.5 -0.6 2050 1,646.8 2.1 0.1 1,449.7 2 1 -0.1 1,085.6 2.1 -0.6 2060 1,663.8 2.1 0.1 1,446.7 2.1 -0.0 1,002.6 2.1 -0.8 2070 1,679.3 2.1 0.1 1,454.8 2 1 0.1 938.4 2 1 -0.7 2080 1,682.6 2.1 0.0 1,461.0 2.1 0.0 893.9 2.1 -0 5 2090 1,685.7 2.1 0.0 1,459.9 2.1 -0.0 859.3 2.1 -0.4 2100 1,687.8 2.1 0.0 1,462.1 2.1 0.0 843.0 2.1 -0.2 a. Total fertility rate; i.e.. the average number of children born to women who experience the prevailing pattern of fertility during their child-bearing years. h. Growth rates are averages over the preceding ten years Source, World Bank projections. 137 Ireland, but some Chinese demographers have suggested smaller age cohort would offer an opportunity for Im- an ultimate population target as low as 700 million, to be proving enrollment ratios, but recent falls in enrollment reached as rapidly as possible).'5 suggest that capacity is not currently the main constraint In the next two decades, the economic implications of on enrollment. these alternative projections are quite limited. The differ- ence in population of about 93 million between Projec- POPULATION OF WORKING AGE. All three projections tions A and C in the year 2000 is of course striking, show growth in the population of working age (here especially since neither projection is based on extreme taken as age fifteen to sixty-four) until after the year 2010 assumptions. But, though very large in absolute terms (Table 8.3). The anticipated growth of the labor force is and by the standards of other countries, it is a difference slower under Projections B and C, but even under Pro- of less than 10 percent, and the difference in the average jection A, the rate is not high by past Chinese or develop- annual growth rate of the population during 1985-2000 ing country standards. is only 0.5 percent. The higher figure would cause per Eventually fewer people enter the labor force than re- capita income in the year 2000 to be somewhat lower, tire from it. In Projection C, the population of working and probably also the pattern of food consumption and age (and hence presumably the labor force) begins to agricultural trade to differ (Chapter 3), but not to a large shrink after the year 2010. Under the other projections, extent. this drop comes somewhat later and is less marked. The In contrasting Projections A and C over the longer problem is especially acute in 2030-40 when the large term, however, the differences in population size are as- cohorts of the mid to late 1960s and early 1 970s retire, tounding. A nation of 1.7 billion people would differ in and all three projections show declines in the working age very many ways-physical and social, as well as eco- population. Negative labor force growth might have nomic-from one less than half that size. To assess the some economic advantages. It could cause rapid increases long-run optimal size of China's population would be in productivity through faster abandoning of old plants, difficult and perhaps impossible, since there is no basis for and stronger incentives for employers to economize on predicting the state of technology more than a century the use of labor. In agriculture, the opportunities for ahead. But it is possible to examine the more predictable consolidating landholdings and mechanization, and more consequences of alternative population growth paths for generally the elimination of the rural labor surplus, would the age structure of the population and some of their be welcome. economic and social implications. These will probably be However, a shrinking labor force also inevitably causes quite small in the next fifteen years, but could be very problems of adjustment. To introduce new products and large in the twenty-first century. processes would require more than equal reductions of employment in older activities, which would not be eco- Cbanges in the Age Structure nomically or socially costless: the closure of some plants to preserve economies of scale in other plants might, for POPULAFION OF SCHOOL AGE. The difference be- example, require substantial geographical movement of tween Projections A and C in the number of children age labor. There would be a rise in the average age of the five to fourteen is 29 million at the end of this century; labor force, with shortages especially of younger, more thereafter, the gap widens until by the year 2040 there recently trained, more adaptable workers, and possibly are 106 million five- to fourteen-year-olds in Projection A surpluses of older workers, who find it harder to adjust to and only 50 million in Projection C. new production possibilities requiring substantially differ- A smaller number of children would offer the potential ent skills. These problems-whose solution would be im- for significant savings or quality improvements in educa- possible without a flexible and responsive system of labor tion. But the cost reductions would not be proportionate training and allocation, far removed from China's present to the difference in population size, since there could not system-would occur under all three projections, but be a proportionate elimination of classrooms and schools would be most serious and prolonged under Projection C. or dismissal of teachers. Even in industrial countries, with much more urbanization, better public transport, and a THE ELDERLY. Because all three population projections stronger tradition of dismissing redundant employees, it assume the same ultimate fertility and mortality rates, has proved difficult to consolidate educational facilities in the face of declining educational needs. China already has fairly modest primary student-teacher ratios (averaging 1.SnJa~"ouainDvlpetGasadPas'i 25:1ain 1983)t primand av uderag telasszes (athirfouerapil China's Population: Problems and Prospects, by Liu Zheng and others 25:1 in 1983) and average class sizes (thirty-four pUpllS (Beijing: New World Press, 1981), pp. 25-31; Hu Baosheng and per class in 1983), with an average of about five classes others, "Setting a Target for Our Country's Total Population:' per school, which, with a five-year curriculum, gives little Renkou Yu Jingii (Population and economics), no. 5 (1981), pp. opportunity for consolidation. At the secondary level, a 15-18, 64. 138 Table 8.3 Population of Working Age, Alternative Projections, 1980-2100 Projection A Projection B Projection C Percentage Average annual Percentage Average annual Percentage Average annual Population oftotal growtb Population oftotal growth Population oftotal growtb Year (millions) population (percent)` (millions) population (percent) (millions) population (percent)- 1980 584.8 59.7 - 584.8 59.7 - 584.8 59 7 - 1990 761.4 68.4 2.7 761.4 69.6 2.7 761.4 69.6 2.7 2000 850.3 66.8 1.1 850.3 71.1 1 1 850.3 72.1 1.1 2010 956.8 68.7 1 2 901.0 68.7 0.6 901.0 73.1 0.6 2020 1,020.1 68.2 0.6 944.0 68.4 0.5 894.2 71.8 -0.1 2030 1,027.6 64.9 0.1 935.2 65.1 -0.1 829.7 67.7 -0.7 2040 1,000.6 61.5 -0.3 866.2 59.3 -0.8 698.7 60.3 - 1.7 2050 1,029.0 62.5 0.3 877.6 60.5 0.1 620.8 57 2 -1.2 2060 1,025.4 61.6 -0.0 899.6 62.2 0 2 570.7 56.9 -0.8 2070 1,021.4 60.8 -0.0 883.8 60.8 -0.2 532.3 56.7 -0.7 2080 1,030.0 61.4 0.1 888.8 60.8 0.1 508.6 56.9 -0.5 2090 1,024.8 60.9 0.0 894.6 61.3 0.1 505.7 58.8 -0.1 2100 1,029.4 60.9 0.0 887.2 61.7 -0.1 510.5 60.6 0 1 Note Data are for the population fifteen to sixty-four years old. a Growvth rates are averages over the preceding ten y ears Source WVorld Bank projections Table 8.4 Population Sixty-Five Years Old or More, Alternative Projections, 1980-2100 Projection A Projection B Projection C Percentage Working- Percentage Working- Percentage .Working- Population oftotal age pop. Population oftotal age pop. Population oftotal age pop. Year (millions) population 65 + pop. (millions) population -65 +pop. (millions) population - 65+pop 1980 45.1 4.6 13.0 45.1 4.6 13.0 45 1 4.6 13.0 1990 67.2 6.0 11.3 67.2 6.1 11.3 67.2 6.1 11.3 2000 94.5 7.4 9.0 94.5 7.9 9.0 94 6 8.0 9.0 2010 118.5 8.5 8.1 118.5 9.0 7.6 118.5 9.6 7.6 2020 172.9 11.6 5.9 172.9 12.5 5.5 172.9 13.9 5.2 2030 234.8 14.8 4.4 234.8 16.3 4.0 234.8 19.2 3.5 2040 316.7 19.5 3.2 316.7 21.7 2.7 316 7 27.4 2.2 2050 307.7 18.7 3.3 307.7 21.2 2 9 307.7 28.3 2.0 2060 323.6 19.4 3.2 275.8 19.1 3.3 275.8 27.5 2.1 2070 348.8 20.8 2.9 299.7 20.6 2.9 256.4 27.3 2.1 2080 338.3 20.1 3.0 304.6 20.8 2.9 227.4 25.4 2.2 2090 347.7 20.6 3.0 294.0 20.1 3.0 200.1 23.3 2.5 2100 350.2 20.8 2.9 305.3 20.9 2.9 179.1 21 2 2.9 Source: World Bank projections. they all eventually converge-in the twenty-second cen- is likely to find aging issues more serious, because the rise tury-to the same, unchanging age structure (Table 8.4). in the Droportion of elderly people will begin at a com- People age sixty-five and over constitute about 2 1 per- parativ,2ly low income level. In addition, with an irregular cent of the "stationary" population in each case. The decline in fertility, the proportion of the elderly may even population of working age is about 61 percent of the tempotarily rise above its long-term stationary level. In total, only slightly greater than the 60 percent of 1980. Projections A and B, the proportion of the elderly in- The ultimate ratio of dependents to the labor force is thus creases slowly to its ultimate size with little fluctuation. not very different from today. But the demographic struc- By contrast, in Projcction C, the elderly becomiie an ex- ture of dependency is very different: in 1980 only about tremely large proportion (2 7-28 percent) of the popula- one out of ten dependents (excluding those age fifteen to tion for several decades after the year 2040. sixty-four) was sixty-five or more; ultimately, it will be How financial support of China's prospectively large about one out of two. elderly population might be organized-an issue that may Though all countries will probably eventually have a require important steps to be taken soon-is considered similarly substantial proportion of elderly people, China below. The other major problem posed by the aging of 139 Box 8. 1I The Second Health Care Rev tion The pattern of Chinese health problems I has already iment holds promise of results and can be afforded; they changed from that of a low-income developing country,; must pay careful attention to affordable plans for rehabili- ...where typically over 40 percent of deaths result from in- tation of individuals partially or wholly incapacitated by fectious, parasitic, and respiratory causes4 and mortalitys is noncommunicable disease; and they must be concerned particularly high among infants and young children, to one wvith humane care for the terminally ill (an area in which where the leading causes of deathf are heart Xdisease, there have been major and quite affordable advances in strokes, and cancer, as they are in high-incomne countrtes. Western medical practice). China now faces the difficult challenge of preventing and "A third ingredient of the second health care revolution treating these chronic diseases. The conclusions of a recent will be that of designing an insurance and financing struc- World Bank study of China's health sector are summarized ture that encourages prevention and discourages current in the following paragraphs: tendencies toward overuse of facilities. Such an insurance "While there are important steps [to prevention of structure should both provide strong disincentives for in- chronic diseases] that can be taken with little further analy- troduction of the high-ost procedures that would neces- sis-particularly control of salt intake and tobacco con- sarily be available to only a fraction of the population; and sumption-. . . much work needs to be done to identify it should (through inclusion of substantial deductibles and and field test appropriate preventive strategies. The task is co-insurance rates) create incentives for both patient and inherently far more difficult than is that of prevention of doctor to utilize health resources prudently. communicable disease. An essential ingredient of a second "Assembling the above three ingredients will, inevitably, health care revolution [to tackle emerging problems of prove to be a major challenge; but, to the extent that chronic disease] in China will be to recognize the difficulty success is achieved, China will have become a world leader and importance of that task and to commit substantial in the effective and humane handling of the burden of resources in a sustained way to its solution. chronic disease without succumbing, as other countries "Another important ingredient of the second health care have, to endlessly costly investments in medical technolo- revolution will be development and widespread implemen- gies of limited efficacy."' tation of strategies for dealing with the large number of _________:i _ cases of chronic disease that will, inevitably, occur. These I iDean T Jamison and others, Chbna: The Healtb Sector (Washing- strategies must include capacity for treatment where treat- ion, D.C.: World Bank, 1984), paras. 5A13-5.16. China's population will be health care: in the United trol the level of fertility precisely. Though couples are States, medical expenditure per elderly person is roughly urged to have one child, some will have more. Despite three times that of working-age people. Changing age substantial economic incentives, the one-child policy is structure is in fact only one of the components of China's not popular in a family-oriented culture, which has impending second health care revolution (see Box 8.1), greatly prized sons rather than daughters. but it provides the clearest illustrations of the terrible There are no official population targets beyond the choices that will have to be made about the-usually year 2000. But the preceding analysis suggests that while extremely expensive-treatment of chronic and terminal the achievement of replacement level fertility (2.1 chil- diseases. Britain, for example, has held down costs by dren per women) is clearly desirable, there could be dis- denying those above a certain age access to treatments advantages in trying to hold fertility well below this level, such as renal dialysis. Other countries, such as the United especially after the year 2000. Beyond a certain point, as States, have not done this, with the result that their illustrated by Projection C, the gains from achieving a health care costs have risen dramatically as a share of smaller population in the twenty-first century could be GNP-and the proportion of the elderly in China in the more than cancelled out by the adjustment and transfer next century will be far higher than the present propor- problems posed by the population's changing age tion in the United States. structure. In any event, an important issue is how eventually to Birtb-Planning Policy achieve an orderly transition out of the one-child policy. In the past decade, the Government has shown a remark- Abandoning the policy suddenly would probably lead to able ability to control individual fertility decisions. China an undesirable new baby boom. One measure already is in a good position to determine its demographic des- adopted, which is designed to emphasize that the one- tiny. Nevertheless, even the Chinese system cannot con- child policy is only a temporary measure, is to allow 140 children from one-child families who marry each other to reduces mobility of workers among enterprises, partly have two children. Another policy modification now ap- because it increases the difficulty and social cost of closing plied in rural areas is to permit parents whose first child is down or drastically reorganizing inefficient enterprises. a girl to have a second child. A possible extension of this Other approaches to supporting the elderly include would be to permit every family to have one son, but no having the elderly support themselves, through work and further children: average family size would then be about througii personal saving; the establishment by the state or two children. Whatever route the transition takes, the employers of funded pension schemes; and state noncon- present generous level of economic incentives to parents tributory (or nonfunded) pension schemes. Other coun- who have only one child should be maintained, at least tries ha.ve adopted varying mixtures of these approaches, until their removal would not excessively stimulate fertil- each of which has some advantages and disadvantages. ity. Indeed, the core positive incentive-a monthly pay- Except for household agriculture, the degree to which ment to one-child families-might well remain, even if the eld.rly support themselves through work depends on some of the negative incentives, which discriminate the age of retirement. In China, this is now normally against children from families with more than one child, sixty for men, fifty-five for women cadres, and fifty for were removed. women workers-a pattern that is not unusual by inter- national standards, and indeed has been shaped by similar Supporting the Elderly pressures, including a desire to move workers (especially Providing financial support to a large elderly population women, who actually tend to live longer than men) out poses two sorts of problems. One is its economic cost. of the labor market at times of general unemployment. While the ultimate overall dependency ratio may be Official retirement ages are not closely related to ability much the same as at present, elderly people have higher to work; despite increasing longevity and improving material consumption needs than children, especialiy for health, the retirement age was raised in only two out of food and housing space. The economic cost of supporting fourteen industrial countries between 1949 and 1977 the elderly could in principle be reduced by cutting their (nine countries lowered it). Moreover, forcing people living standards, but China, like other societies, will want with the capacity and desire to make a contribution to its older people to share in the increasing prosperity that stop wcrk at a certain age can be bad for their health and they have helped to bring about. The costs involved are self-resoect. In the longer term, and especially when Chi- already causing difficulties in rich countries, even though na's labor force is stagnating or shrinking in the twenty- they have a smaller proportion of elderly people than first century, providing more-especially part-time-em- China will have in a few decades. ployment for people above the retirement age, or raising The other problem is how to organize and provide this the retirement age, would be an option worth consider- support. Traditionally, in China as in other developing ing. countries, elderly parents have lived with, and been sup- Gre,,ter personal saving for retirement also has certain ported by, their children; China's 1980 marriage law advantages (governments elsewhere sometimes provide reiterates this filial duty. But in all countries, economic tax concessions to long-term saving schemes), including modernization-with more urbanization and mobility, smaller administrative costs and a greater sense of inde- rising incomes, and changing social attitudes-has re- penderice and responsibility for the beneficiary. But be- duced the number of multigeneration households. In cause many people are either too poor, or insufficiently China, the difficulty children have in supporting parents farsigh:ed, or without access to appropriate financial in- will be further increased by the one-child family policy. vestments, to save enough to support themselves decently China already provides some support for the elderly, in old age, personal saving will problably in China (as in but existing institutions will not be able to cope with the most other countries) be confined to a supplementary problems of the twenty-first century. In the countryside, role. the Five Guarantees provide food and clothing, shelter, Full)' funded (FF) pension schemes, sponsored by the medical care, and burial expenses. But the level is mini- Government or by employers, are similar to personal mal, and people are sometimes reluctant to accept these savings in that they receive and invest contributions dur- benefits, because of the shame of indigency or because ing the employee's working life, with the resulting pen- they want to retain the right to bequeath family property. sion strictly dependent on the amount contributed and In the cities, and in a few rural areas, enterprises and the returns earned from its investment. They may differ other employers now provide pensions for retired work- from personal savings, however, in receiving contribu- ers; the coverage of such schemes will increase as industri- tions from employers as well as from employees, or in alization and urbanization proceed. However, continued making employee contributions obligatory. heavy reliance on pensions paid directly by employers Employer-sponsored FF pension schemes would be su- would be an obstacle to intensive growth, partly be- perior to the present Chinese system of pensions paid cause-as experience in other countries has shown-this directly by the employer (sometimes called pay-as-you- 141 go), because the fund could and should be legally sepa- der government schemes), or by raising the contributions rated from the enterprise's general finances and could of current employees or their employers. The principle thus survive and meet its pension responsibilities even if that beneficiaries should have earned their pensions the enterprise closed down. Similarly, if employees were through contributions remains important, if only because allowed to retain or transfer their share of the pension the elderly thus seem (and feel) less of a burden on cur- fund to a new employer (which employers elsewhere rent workers. have been reluctant to permit, partly to encourage long There is increasing concern, however, that collecting service, partly because of the administrative costs in- these contributions through-in effect-a substantial tax volved), the employees could move more easily among levied on wages may have an adverse effect on employ- enterprises. But, clearly, government-sponsored pension ment. (This could also be a problem in China, especially schemes (whether funded or not) would also offer inde- as regards unskilled workers.) For this reason, but also pendence from enterprise finances and greater worker because the contributory principle is now often a matter mobility. of appearance more than substance, state pension FF pension schemes have the additional advantage of schemes financed straightforwardly through general taxa- providing pensions as a right, with less sense of burden tion are sometimes advocated, and have been imple- for both the recipient and the working-age generation. In mented in the Soviet Union and, on a supplementary other countries, moreover, pension funds play an impor- basis, in Japan and the United Kingdom. China's Five tant role in financial markets and could perhaps also do so Guarantees system is similar in principle to these pension in China, as well as providing a means for indirect worker schemes. ownership of enterprises (these possibilities are discussed Noncontributory pension schemes are sometimes criti- further in Chapter 10). cized as being excessively costly and as imposing an unjust However, FF pension schemes also have certain short- burden on current workers. This is fundamentally incor- comings. They provide little or nothing for those who rect, since, whatever the specific institutional arrange- retire within a decade or two after their introduction and ments, the elderly are ultimately always supported by the indeed become fully operational only after forty years or productive activities of current workers-the real issue so. Moreover, the pensions they provide are uncertain being at what relative level. Raising the substantial reve- and variable in relation to the wages of current workers. nues required through general taxation may, however, as This is partly because wages go on rising, while the pen- with payroll taxation, have adverse economic effects or sion depends largely on contributions from earlier-and be politically difficult. much lower-wages, partly because the return on finan- For China, a central issue would be whether, when, cial and other investments is somewhat unpredictable, and how to include peasant households in any pension especially when corrected for price inflation (which some- scheme. Reform of the urban pension system would be times makes real returns negative). comparatively straightforward. In rural areas, however, For these reasons, which would be relevant also in there would be not only more administrative difficulties, China, governments and employers have often supple- but also, and more fundamentally, the problems of an mented or modified FF pension schemes. To meet the initially large gap between average urban and average needs of those retiring before the schemes are fully opera- rural incomes and of widely varying rural incomes. In tional, some governments have-as in Japan-established these circumstances, should the contributory principle be parallel noncontributory pension schemes, while others- applied, should contributions be compulsory and related as in the United States shortly after the introduction of to income, and should pensions be related to contribu- the social security system-have waived or reduced con- tions-all of which would make pensions as unequally tribution requirements for these age groups. To ensure distributed as incomes and, in some cases, very low? Or that pensions maintain a stable and equitable relationship should a large and fiscally expensive noncontributory ele- with current wages, many government and enterprise ment, or cross subsidies, be used to make pensions less pension schemes have made pensions dependent on the unequally distributed than-or even unrelated to- number of years during which contributions were made, contributions? the individual's terminal salary, subsequent price infla- In planning for the support of the elderly in the tion, current wage levels, and so on-rather than on the twenty-first century, China has a wide range of options. individual's accumulated contributions. The choice among them is not simple, and none (and no Although benefits have increasingly been divorced mixture of them) is ideal. But it will be important to from individual contributions, the principle of an inde- consider them, to choose a system, and to put it into pendent, financially viable fund has generally been pre- operation, within the next few years. Other countries served, either through supplementary contributions from have found that the problems of an aging population can employers (under enterprise schemes) or the budget (un- be compounded by failure to look sufficiently far ahead. 142 Mobilizing Financial Resources In China, as in other countries, rapid economic and social ing. Conversely, total domestic saving comes partly from development will require (though not automatically fol- the government, partly from enterprises (state, collective, low from) high rates of both saving and government and individual), and partly from households. Of this sav- revenue raising. To mobilize the huge amounts of money ing, pan is used for self-financed investment-enterprises involved, and to do so in ways that promote efficiency, ploughirig back retained profits into expansion and peo- improve the distribution of income, and maintain macro- ple building houses. The rest is channeled into investment economic stability, will require much political will and indirectly, through government grants or loans to enter- administrative skill. The economic and institutional prises, through bank deposits being relent to enterprises, changes associated with system reform, moreover, will through bond sales, and so on. require constant adaptation of old financial instruments and policies, as well as experimentation with new ones. This chapter discusses some interrelated long-term is- A sues and options in government finance and national sav- The macroeconomic projections discussed in Chapter 2 ing-financial intermediation and investment decision- suggest :hat to attain its long-run growth targets China making are discussed in Chapter 10. After reviewing will need to save something like 30 percent of national alternative sources of saving, it looks at subsidies and income (the numbers mentioned in that chapter range transfer payments, then at overall government revenue from 26 percent to 3 6 percent, depending on the assump- needs and sources, and finally at tax reform.' (Though tions made about investment efficiency, foreign borrow- local public finance was addressed in Chapter 5, this re- ing, and the growth rate). This is a high saving rate by port does not deal fully with the complex and difficult most iniernational standards, though similar to those of financial relationships among different levels of govern- Japan ard the East European socialist countries (Annex 5, ment in China. And it touches only very briefly-in Table 3.2), and comparable to the rates achieved by Chapter lO-on short-term fiscal stabilization policies.) China over the past three decades (Chapter 2). It is thus The various topics and projections discussed in this important to consider, especially in relation to proposals chapter share a common macroeconomic accounting for further fundamental reform of China's system of eco- framework, summarized in Figure 9.1 (which includes nomic management, how this high saving rate might be estimates of the relevant magnitudes for China in 1981). achievecl, and in particular what might be the contribu- The economy is divided into three institutional sectors tions of different institutions and sectors. (the government, enterprises, and households), each of Up tc now in China, the Government has been the which has a distinct source of income (taxes and remitted main saver, which has meant a large overlap between profits, retained profits, and wages and other earnings, saving and public finance. This pattern of saving is proba- respectively). The diagram divides the uses of national income into three main catcgorics: household consump- tion, public consumption, and saving. It then shows how 1. Inernational experience in public finance and its implica- each of the three institutional sectors contributes to each tions for China are discussed in Background Paper 4. of the three uses of national income. Government reve- 2. Government finance of household consumption through nues, for example, finance some household consumption subsidies is not explicitly shown in the figure, since uses of national (through transfers and subsidies2), as well as public con- income are measured at purchaser prices (which already include sumption (to which enterprises also contribute), and sav- the price-reducing effect of subsidies). 143 Figure 9.1 Sources and Uses of Funds, 1981 (billions of yuan) National income at producer prices (442) Profits, (160) | Wages and earnings (2 82) Indirect Direct taxes Net profits after taxesb and levies' taxes and levies (60) (118) (46) Wages and earnings Income generation and distribution Government revenue (178) Retained Divi- after taxes Other uses of government profits dendsd (278) Government revenue ( 34) (45) (1) subsidies (44) National income at purchaser prices (458) Institutions Governmentf Enterprises' Households Institutions | (178) (45) (279) l Government Enterprise Household transfers to transfers and self-financed Household consumption (2 5 3) households subsidies to consumption households (4) (8) (241) Government- Enterprise- Household- financed financed financed r=. public public public Uses of income Public consumption (7 3) Si consumption consumption consumptiong E (65) (8) (0) 0 Saving (132) (65) Net financeI of investment through grants, loans, banks, bonds, and so forth l I Note This accounting framework can be applied to any country, but the numbers in each box are tentative estimates for China, 1981. The diagrammatic integration of the national accounts with government, enterprise, and household accounts is based on the United Nations System of National Accounts. 'I'hus, "national income" includes not only net material product, but also nonmaterial services (which compose a large fraction of "public consumption') and depreciation (which composes a large fraction of "saving"). National income (or gross domestic product, (.DP) can be measured at producer prices, reflecting the incomes actually received by the producers in the economy, or at purchaser prices, reflecting the amounts paid bv households, enterprises, and the Government for the goods and services they buy. The difference between the two is indirect taxes, net of government subsidies. The sum of incomes received by thc Government, enterprises, and households is equal to r.;oP at producer prices plus total indirect taxes (which form part of government revenue). a. Gross of depreciation and government subsidies, but net of indirect taxes. b. All taxes and levies on goods and services. c. Includes income and profit taxes, asset taxes, agricultural tax, profit remittances to the budget and to government organizations, remittances of dcprciation funds, and social insurance contributions. d. Workers' bonuses paid out of enterprise profits are considered part of wages and earnings. e. Includes extrabudgetary income and expenditures of government organizations, as well as the gov'crnment budget. f. State, collective, and individual enterprises are treated as a single sector. g. In the conventional national accounting framework, household expenditures on education, health, and other forms of public consumption arc defined as private consumption and therefore are included in the box above. Source: Annex 5, Appendix J. 144 Table 9.1 Gross Domestic Savings in Selected Countries, 1976-80 (percent) Cbina, Cbina, United United Soutb Measure 1978' 1981' States Xingdom Japan Korea India Share of total saving Government 73 49 7 0 9 26 13 Enterprisesb 12 22 58 56 37 35 22 Householdsb 15 29' 35 43 54 38 65 Total 100 100 100 100 100 100 100 Aggregate saving rated 37 29 19 19 32 25 23 Note: Data include depreciation funds. a, For China, government saving includes both the consolidated budget surplus of revenues over current expenditures and the nonbudgetary saving of government organizations. Enterprise saving includes state and urban and rural collective ente-prises. b. Rough adjustments have been made to the national accounts statistics in order to incluide the saving of unincorporated enterprises with that of other enterprises (rather than with that of households) c. In 1981 personal saving appears to have been unusually' high in China, because of large sto:k increases in rural areas. d. Total saving as share of GDP at purchaser prices. Source: Estimates based on State Statistical Bureau, Statistical Yearbook of Cbina. 1981 (Hong Kong: Economic Information and Agency, 1981), see Annex 5, AppendixJ; United Nations, Yearbook of National Accounts (New York, 1981). bly similar to that of the Soviet Union and Eastern Eu- heavy rcliance on government saving has actually been rope (though detailed data are not available), but is quite associated with underinvestment in economic and social different from those of most other countries. In Table 9.1 infrastrt.cture; in other countries a significant fraction of for example, despite a significant change between 1978 such investment is financed by the savings of enterprises and 1981, government saving in China is far more impor- and hoiseholds through the issuance of government tant than in any of the other countries. Correspondingly, bonds. .MAoreover, government finance of other invest- the shares of both enterprise and household saving in ment does not always imply close control: recent experi- China are the lowest in the table-although the division ence in China suggests that it may in fact be difficult for of saving between enterprises and households varies the central authorities to regulate the level and composi- widely among the other countries (enterprise saving being tion of local government investment, especially when it is substantially larger than household saving in the United financec from extrabudgetary funds. States, for example, but substantially smaller in Japan). One clear disadvantage of heavy reliance on govern- For the future, one possibility for China would be ment saving is the strain it imposes on the public finances. continuance of the past (or East European) pattern, with Tax rates must be higher than they would otherwise have government revenues as the predominant source of sav- to be, other government expenditures (on public con- ings. It is also possible, however, that increased reliance sumption or support of household consumption) lower, on market regulation and other systemic reforms might or some combination of the two. High tax rates in turn make it preferable to further increase the shares of enter- are disadvantageous because they weaken incentives to prises and households in total saving. How much to re- cut costs, innovate, and work harder and because they duce the share of government saving, though, and how to increase tax evasion and the administrative costs of tax divide the remainder between enterprises and house- collection. Reductions in other public expenditure are holds, depends on the advantages and disadvantages of disadvantageous because-despite some potential for each of these sources of saving. eliminating waste-there are pressing needs (mentioned frequently in this report) for greater government expendi- Government Saving ture in mnany areas of social and economic importance, Two clear advantages of government saving in the Chi- for which greater enterprise and household expenditure nese context are that it is a proven and reliable means of cannot or should not substitute. Another possible disad- achieving a high and reasonably steady aggregate saving vantage of keeping government saving as the main source rate and that it is unquestionably consistent with social of finance for investment is that this might conflict with ownership of the means of production created through expansion of enterprise autonomy in decisionmaking. saving. It could also be argued that government saving is But it might be possible to channel government saving to essential for financing infrastructural investment and that enterprises through an efficient network of financial in- it permits close control of other investment. These latter termediaries (discussed in Chapter 10). arguments are not quite so compelling. China's past Perhaps the best approach to determining the appropri- 145 ate level of government saving in China might be to think The potential future contribution of state enterprise of it as the sum of two components-first, as the main saving in China naturally depends heavily on the likely source of finance for vital basic investments, and second, future level of state enterprise profits-which will also as bridging the gap between the total amount of saving affect the level and pattern of taxation needed to achieve required for rapid growth and the actual saving of enter- government revenue targets (discussed later). In China, prises and households. But even this admits a wide range unlike some other countries (where poorly run state en- of possibilities. As regards the first component, for exam- terprises are a financial burden on the rest of the econ- ple, if government saving were to finance only 60 percent omy),' and despite the losses of many specific enterprises, of investment in energy and transport, 75 percent in the state enterprise sector has always generated substan- education and heakh, and 90 percent in public adminis- tial profits. A crucial question, however, is whether pric- tration and defense, its share in total saving in the year ing and other reforms will reduce the average profitability 2000 would be 20 percent (on the basis of the QUADRU- of China's state enterprises. This is hard to predict: it will PLE investment projections), but if vital basic investment be the net result of several conflicting influences. were construed to mean 100 percent in these three cate- Price reform (discussed in Chapter 10) is likely to entail gories plus 75 percent in industry and 50 percent in redistribution of profits among sectors-for example, commerce and housing, the government would account from manufacturing to mining and commerce (there has for 70 percent of total saving. Similarly, the amount of already been substantial redistribution toward agricul- gap-bridging government saving that might be needed ture). This would reduce overall state enterprise profits could-on the basis of experience in other countries- if-as seems possible, given the recent expansion of col- vary widely, depending on the amounts that enterprises lective coal mining and commerce-nonstate enterprises and households were able (or permitted) to save. were more common in the sectors whose profitability was increased than in the sectors whose profitability was Enterprise Saving diminished. Equally crucial, in all sectors, will be the In the four more developed countries in Table 9.1 enter- relative efficiency of state enterprise management and the prise saving accounts quite consistently for 10-12 percent relative wage level in state enterprises. If state enterprises of national income (roughly double the percentage in were less cost- or quality-conscious than nonstate enter- China). It is important, however, to distinguish among prises, or obliged to pay higher wages or employ surplus enterprises according to their ownership-and especially labor, increased competition among enterprises-some- whether they ultimately belong to the Government or to thing which would otherwise be highly desirable-could (individual or groups of) households. reduce the profits of state enterprises, and hence their capacity to generate savings. STATE ENTERPRISES. Government saving and state en- terprise saving are obviously closely related: both result in NONSTATE ENTERPRISES. Experience in other coun- state ownership of the means of production created tries suggests that collective and individual, as well as through saving, and one can be substituted for the private and mixed-ownership enterprises could poten- other-within limits-by varying the profit tax rate and tially contribute a large fraction of the savings required the rules governing remittance of after-tax profits (and for rapid growth in China, and at the same time could depreciation funds) to the budget. But state enterprise contribute to increasing efficiency through competition saving has certain advantages over government saving, and microeconomic structural change (see Chapter 1). To especially in the context of a reformed economic system: do so-as China's experience especially with commune it gives enterprise managers a stronger incentive to cut and brigade enterprises confirms-nonstate enterprises costs and increase profits, and it enables enterprises to must be allowed to make substantial profits (in aggregate, function as autonomous economic units-since such units though some will lose money) and to invest and expand must control at least some investment, and at least some freely. They will then have a strong incentive to increase of their investment must be financed internally, rather their profits through cost reduction and greater attention than by borrowing. State enterprise saving may also have to customer needs and, just as important, to save and disadvantages, especially if other aspects of system reform reinvest a high proportion of their profits. It is essential, lag behind: inappropriate motivation and distorted prices however, that the owners of these enterprises should be may lead enterprise managers to make bad investment confident that the Government will allow them to go on decisions, and profits intended for enterprise saving may operating indefinitely, without expropriation or punitive be diverted into bonuses and benefits for workers. (The taxation. In Hungary, for example, uncertainty about the advantages and disadvantages of giving more control of future course of government policy, which now encour- investment to state enterprises are amplified in Chapter 10; control of worker remuneration in state enterprises is 3. World Bank, World Development Report 1983 (New York: discussed in Chapter 8.) Oxford University Press, 1983), Chapter 8. 146 ages individual and small collective enterprises, has caused or with provision for retirement. Saving for housing is these enterprises to reinvest only a small (by international already important in China. Peasants have apparently standards) proportion of their profits in expansion and used a considerable part of their increased saving in the market development. past few years to build new houses and improve existing The main disadvantage of relying heavily on nonstate houses. In urban areas, there is also a substantial private enterprises as a source of saving in China would be its housing stock (mainly rather old), and some state-built potential consequences for income distribution. A consi- apartments have recently been sold to individual families. derable fraction of their after-tax profits (perhaps 50 per- In the future, the greater part of China's investment in cent, with the other 50 percent saved) must be distrib- urban housing could be financed by personal saving uted for consumption, or else the incentives for (partly through housing cooperatives), with far less provi- establishing and expanding these enterprises would be sion of workers' housing by enterprises. The Govern- reduced or even eliminated. Such distribution may be ment mlght provide housing directly only to the minority acceptable for collective enterprises, where consumption of people unable to pay, but could assist others by selling out of profits is used to benefit the local community, or off more existing urban housing, as well as by "sites and given to workers as bonuses, or as higher wages than they services" projects-planning and providing basic utilities could earn elsewhere. It poses greater problems in the for new individually constructed housing-coupled with case of individual enterprises, where much of the profit technical assistance and limited subsidies or tax conces- accrues to one person or family. Other countries have sions to housing cooperatives. (These could partially re- used taxation to influence the division of private enter- place the large existing housing subsidies, which would prise profits between consumption and saving (by higher need to be reduced or eliminated to provide an incentive rates of tax on profits distributed to owners than on for tenants to become owners.) Experience elsewhere reinvested profits). However, any resulting short-term re- suggests that such a system would be a powerful stimulus duction in consumption out of profits eventually tends to to hous2hold saving, as well as to better construction be offset by the higher rate of enterprise growth caused standards and maintenance. by more reinvestment of profits. A better approach to Saving for retirement (including contributions to minimizing adverse distributional effects is thus probably funded pension schemes) was discussed at the end of progressive taxation of personal incomes and wealth (dis- Chapter 8. Household saving for this purpose could cussed later and in Chapter 10). probably be substantially increased, though not as the sole or recessarily even the primary means of supporting Household Saving the elderly. It could be encouraged by organizing funded In low-income countries a large part of household saving pension schemes and other long-term saving instruments is for reinvestment in peasant agriculture and thus shares (commonly provided in other countries by life insurance most of the potential advantages and disadvantages of companies), and possibly also by payment of bonuses to nonstate enterprise saving discussed above. (It is normally workers once or twice a year-as in Japan-rather than treated as household saving simply because of the practi- monthly (this might also facilitate a closer linkage be- cal difficulty of disentangling the various sources and uses tween bonuses and profits). of income within farm households.) In China, too, the Saving for retirement by today's workers will of course introduction of the rural production responsibility sys- always be to some extent offset by the dissaving of retired tem, and repeated official confirmation of its indefinite people, although many bequeath something (often a continuation, have given farm households strong incen- house) to their children. The household sector's net con- tives to make their incomes grow faster through greater tribution to national saving is thus influenced by the saving and reinvestment. But the statistics currently avail- growth rate and age structure of the population, with a able, though suggesting a high rate of rural saving (proba- slower-growing, aging population (as China will have for bly increased somewhat by shortages of consumer the forsecable future) tending on balance to save less. Fast goods), do not permit an accurate assessment of the rate growth of per capita income, by contrast, tends to raise of reinvestment in agriculture, as opposed to nonagricul- the hous,ehold saving rate, by making the incomes of tural activities and housing (Annex 2, paras. 1. 13-1.18). today's savers higher than those from which today's dis- In particular, it is unclear whether increased household savers built up their savings. In this regard there is good investment has been sufficient to offset reduced state and potential for household saving in China-as in Japan, collective investment in agriculture. where the effect of rapid income growth appears to have Apart from agriculture, household saving in other more thin outweighed that of an aging population. countries is partly for purchase of consumer durables and The data in Table 9.1 suggest that household saving other occasional and unanticipated expenditures (and is may be an especially important potential substitute for largely cancelled out by dissaving or borrowing for such government saving. In particular, Japan, the only large expenditures). Most, however, is connected with housing nonsocialist country to have persistently saved about 30 147 Table 9.2 Subsidies, 1981 ond, for support of household consumption through sub- Type of subwsidy Billions of yuant sidies and transfers (income supplements and social relief). Typ- of subsidy Billions ohuan The Chinese Government will have to make important Price subsidies 37.8 decisions on individual items of public consumption in Industrial inputs to agriculture 2.2 the future. But for reasons discussed earlier (Chapter 1), Imported agricultural products 8.8 . . . a Grain 5.9 publc consumption i aggregate may not change greatly Other 2.9 as a share of national income over the next two decades. Domestically produced staple food 15.0 At least the possible range of options (in terms of de- Grain 12.2 mands on government revenue) seems much narrower Edible oil 2.8 than for either government saving or subsidies and trans- Other commodities 6.8 r Nonstaple foodstuffs 2.8 fers. Cotton 1.0 Cbina's Subsidies in Perspective Other 1.7 Large subsidies are at present the subject of widespread Housings 5.0 concern in China. But assessment of the problem is com- Subsidies to money-losing enterprises 10.2 plicated by confusion over definitions, magnitudes, and Industrial enterprises 4.2 economic impact. The term "subsidy" is used in China Commercial enterprises 6.0 to cover various kinds of wage supplements (for every- Total 48.0 thing from nonstaple foods to baths and haircuts), as well a. Estimated depreciation, maintenance, repairs, management, and interest as in its two more normal senses, namely (a) government (based on the cost of construction), less rent paid. About Y 1.3 billion is direct government expenditure; the rest is paid for by state enterprises. payments that reduce the prices of certain goods and Source Annex 5, AppendixJ, Worksheet D. services, and (b) government payments to keep inefficient enterprises in operation. Moreover, in China (as in some other countries) most subsidies are not explicit items of government expenditure, but instead reduce government percent of national income, has done so not mainly revenue (by lowering state enterprise profits), which through high enterprise saving, but through an extraordi- makes them hard to measure, particularly since there is narily high personal saving ratio (about 20 percent of uncertainty about what enterprise profits would or personal disposable income). In most other nonsocialist should otherwise have been. countries, where personal saving ratios are around 10 Table 9.2 shows Chinese estimates of the magnitude of percent, aggregate saving is usually only 20-2 5 percent of price subsidies and subsidies to money-losing enterprises national ihcome. In South Korea, for example, the aggre- in 1981. They totaled Y 48 billion, equivalent to about gate saving rate is as high as 25 percent only because the 3 3 percent of budget revenues and 11 percent of GDP.4 government contributes a quarter of the total. Over half the total is food subsidies, which have become Because Japan's personal saving ratio is so unusual, it much larger in the past few years as a result of substantial seems likely in China that government saving will have to increases in agricultural procurement prices with little or continue to be substantial if a high aggregate saving rate is no increase in the retail prices of staple foods. The next to be achieved. However, the Government's share of largest category (21 percent of the total) is subsidies to saving could probably be much lower than in the past, cover the operating losses of inefficient industrial and with greater reliance on both enterprise and household commercial enterprises. Another substantial item (over saving instead. What the precise balance among different 10 percent of the total) is the subsidy involved in keeping sources of saving in China could or should be in the the rent on urban enterprise and municipal housing at longer term cannot be assessed at present, since much will only a fraction of its capital and operating costs (this depend on the future course of system reform and experi- apparently is paid mostly by state enterprises rather than ence in mobilizing and using enterprise and household directly by the Government). saving (Chapter 10). But some of the possibilities are explored numerically later in this chapter, in connection with alternative projections of government expenditures 4. Budget revenue is adjusted as explained in Annex 5 (Appen- and revenues. dixj, Supplementary Worksheet 1). If government extrabudgetary funds were included, the share of subsidies in total government revenue would be 27 percent. GDP is measured at "producer Subsidies and Transfers prices" (that is, less indirect taxes plus subsidies-see Figure 9.1). This specialized national accounting usage of the term "producer In addition to mobilizing saving, government revenues prices" is confined to figures and tables in this chapter. Elsewhere are needed, first, for public consumption (mainly educa- in the report, the term is used to mean ex factory or farmgate tion, health, defense, and public administration) and, sec- prices. 148 With transfer payments (at present very small) added Table 9.3 Subsidies and Transfers as Percentages of to subsidies, the total as a share of budget revenue in Government Revenue in Selected Countries China is not much higher than the average for developing Country group 1975 1979 countries, and well below the average for industrial mar- ket economies, where transfer payments are very large Induistrial market economies' 54 56 (see Table 9.3). The difference between China and other Oil- mporting developing countriesa 30 31 developing countries would appear larger, however, if India 29 36 subsidies and transfers were expressed as a share of GDP, Pakistan 1e9 20 since budget revenues in China are over 30 percent of South Korea 29 34 GDP, as compared with about 20 percent in the average Si Lanka 32 26 developing country. Food subsidies, moreover, are a Tlhailand 17 13 higher proportion of government budget expenditure in Chiiab 33 China than in all but a few developing countries (includ- ing, in the mid-1970s, Egypt, South Korea, and Sri Note Data are for central government only. Capital transfers are excluded. a. Averages weighted by 1975 GDP. Developing country average excludes Lanka); as a proportion of GDP they are probably higher China. than in any other country except Egypt.' b. Based c,n a figure of Y 44.3 billion derived from Table 9.2 (the total minus Y 3.7 billion in housing subsidies paid for by enterprises), plus an The incidence (or economic impact) of China's subsi- estimated Y 2.0 billion of pensions and Y 1.0 million of social relief paid by dies can be established only on specific assumptions about the budget Data are for 1981. the situation without subsidies and may vary dramatically Source (ntsrnational Monetary F18 nd, Govenment Finance S4a6stcs Year- depending on the assumptions made. For example, Chi- na's food subsidies are often said to raise the real incomes of urban residents, but this is so only if the assumed alternative is higher retail food prices. If, by contrast, the subsidized coal discourages the purchase of fuel-efficient alternative were assumed to be lower procurement prices, stoves o0 when low rents cause inadequate housing main- then the beneficiaries would be not the urban population tenance. Subsidies may also be costly to administer, espe- but the rural population. An assumed alternative combi- cially if the item in question has to be rationed. nation of higher retail and lower procurement prices As instruments of income redistribution, subsidies are would imply that the benefits go partly to urban, partly at best crude, since they provide low prices to rich and to rural residents. Similarly, assessment of the incidence poor alike (as contrasted with income supplements tar- of China's urban housing subsidy is dependent on as- geted at the poor, which in principle can achieve the same sumptions about the level of wages without the subsidy. effect at lower cost). They may even have perverse ef- fects: in urban China, for example, lower-paid temporary Future Optionsfor Price Subsidies workers from the countryside and the lower-paid work- Price subsidies are not in principle a bad thing.6 If used as ers of small-town collective enterprises have less access to negative indirect taxes in the context of market regula- subsidised food and housing than higher-paid permanent tion, they can be useful economic levers for stimulating state ent-rprise workers, and in rural China, higher grain the production and consumption of goods and services of procurerment prices may have disproportionately bene- which there would otherwise be too little from an eco- fited richer households, which have a greater marketable nomic or social viewpoint. Research, training, books, and surplus, and perhaps even have harmed some poorer the arts are common examples in other countries. Price househo ds, which are net purchasers of grain. As instru- subsidies can also. contribute to improving the distribu- ments for price stabilization, too, subsidies are of dubious tion of living standards: food subsidies tend to accrue benefit, since they encourage governments repeatedly to disproportionately to lower income groups, who spend a postpone comparatively minor price increases until the higher proportion of their incomes on food; children's required price increases become disruptively large. clothing and school lunches are sometimes also subsidized Thus Ilthough China's subsidies perhaps are not such a for similar reasons. Price subsidies can be used to a lim- serious problem as is sometimes suggested, they might ited degree to stabilize living standards in the face of advantageously be reduced, modified, or replaced with temporary fluctuations in producer prices. Especially in the way China now uses them, price sub- sidies also have some disadvantages. They can complicate fiscal planning by fluctuating unexpectedly, particularly if 5. Jeffrey M. Davis, The Fiscal Role of Food Subsidy Pro- grams:' IM4F Staff Pap ers, vol. 24, no.1 (March 1977); Sadiq they arise from a commitment to keep certain prices Ahmed, lublic Finance in Egypt: Its Structure and Trends, Staff constant (as contrasted with, say, a fixed percentage sub- Working Eaper 639 (Washington, D.C.: World Bank, 1984). sidy). They can also have adverse effects on economic 6. The advantages and disadvantages of subsidies to inefficient efficiency by distorting price signals-for instance, when enterprises are discussed further in Chapters 1, 5, 7, and 8. 149 alternative instruments over the next few years. Food not accurately compensate for increased retail prices and subsidies could in principle be reduced and regulated by rents, but would need to be combined with supplemen- adjusting agricultural procurement prices. This is not an tary income transfers to high-dependency households. attractive option, partly because it would widen the agri- There should probably also be special supplementary in- cultural-nonagricultural income gap, but mainly because terest payments on saving deposits, which would other- it would in principle conflict with the use of agricultural wise lose part of their real value, financed perhaps by a producer prices to balance supply and demand for partic- corresponding special levy on borrowers, whose loan re- ular agricultural products (Chapter 3). In order to enable payment burden would otherwise decrease in real terms. agricultural producer prices to provide appropriate alloca- (See also Chapter 10) tive signals to farmers in changing circumstances without The specific mixture of general wage increases and sup- disrupting the budget, the Government might reduce its plementary income transfers chosen would depend on role as a buyer and seller of agricultural products, leaving the relative importance attached by the Government to the bulk of the market to collective and individual com- minimizing its outlays or revenue losses, preventing any- merce (Chapter 3). But it would also be essential to estab- one's living standard from falling, and widening the wage lish a rational relationship between the retail and pro- differences between skilled and unskilled workers (for ducer prices of food and other agricultural products. This employment and incentive reasons discussed in Chapter would involve restoration of normal distribution and 8). For example, equal proportionate wage increases processing margins, mainly through increases in retail would widen the real income differences between skilled prices. and unskilled workers (since on average the latter tend to At present producer prices, eliminating food subsidies belong to lower-income households, which spend propor- would apparently require a 50-60 percent increase in the tionately more on food and rent), but they would reduce urban retail price of grain and an 80 percent increase in living standards in low-income households unless there that of edible oil. To maintain living standards, these were wage overcompensation (wage increases that on increases, plus those necessary to eliminate other com- average more than offset price increases) or larger income modity subsidies, would require urban incomes to be transfers to poor households, both of which would in- raised by about 25 percent on average, but perhaps by as volve greater costs to the Government. Alternatively, at much as 45 percent for the poorest urban households, lower budgetary cost, the general wage increase could be which spend a higher proportion of their incomes on proportionately larger for low wages than for high wages, food.7 If housing subsidies were simultaneously elimi- which would maintain the purchasing power of low nated through rent increases, the required urban income wages but would leave the wage differential between increases would be roughly 30 percent on average and 50 skilled and unskilled workers unchanged in real terms, percent for the poorest households. although it could subsequently be widened gradually by The social problems that have sometimes followed faster wage increases for skilled workers (see Chapter 8). large retail price increases in other countries seem to have There is room for disagreement as to whether such a occurred mainly because the price increases were used to major realignment of retail prices and incomes should be cut real household consumption in the face of macroeco- done in one step or more gradually. If full compensation nomic difficulties. In China, however, there is no need were provided, a carefully prepared and well-explained for a cut in urban household consumption (although it one-step adjustment could be quite acceptable and would might be desirable for urban consumption to grow more avoid the protracted uncertainty and delays to other nec- slowly than rural consumption in order gradually to nar- essary reforms associated with gradual adjustment. But it row the urban-rural gap). Urban households could and would also obviously increase the cost of errors in prepa- should thus be compensated with income increases to ration (especially calculating the required increases in offset the required retail price increases-although this prices, wages, and other forms of compensation) or im- would mean no net improvement in the budget balance. plementation. In any event, the Government has thus far A major instrument of compensation would be general chosen to take limited steps, such as the May 1985 in- wage (and pension) increases for state and large urban crease in nonstaple food prices, which was offset by a flat- collective workers. The earnings of individual and small rate per capita increase in urban incomes. urban collective workers, which are not subject to admin- istrative regulation, could not be raised so easily, but they would tend to be pushed and pulled up by the higher wages of other workers, partly because of the need for 7. These figures are rough estimates based on the urban house- wages of other orkers, prtlybecaseoftheneedfor hold survey data in State Statistical Bureau, Statistical Yearbook of small enterprises to attract labor, partly because of the China, 1983 (Hong Kong: Economic Information and Agency, associated increase in money spending power. Because 1983), and on the data in Table 9.2, assuming 75 percent of the the ratio of dependents to workers (and pensioners) varies price subsidies in categories 1-4 are for goods sold to urban resi- among households, a general wage increase alone could dents. Urban incomes in 1981 totaled about Y 1 OObillion. 150 In the long term, once normal commercial margins mentioned above. Another might be extension of the have been restored, rises and falls in the producer prices social security benefits (pensions, sickness and disability of particular goods should be reflected in corresponding payments, maternity leave) currently enjoyed by govern- changes in retail prices. This would contribute to national ment and state enterprise employees to the rest of the economic efficiency by encouraging consumers to buy nonagricultural labor force and even to part of the agri- less of things in short supply or whose production costs cultural population. Even without this wider coverage, are increasing and more of things in abundant supply or the Government might assume direct responsibility for whose production costs are falling. Experience in other providing most or all of the benefits now paid by state countries suggests, however, that it would be inadvisable enterprises. Reform of the labor allocation system to in- to provide continuing automatic compensation for retail crease mobility would necessitate some form of govern- Table 9.4 Subsidies and Transfers, Alternative Projections, 1981-2000 (percent) 2000 Subsidies Subsidies Intermnediate Higb eliminated, eliminated, subsidies, subsidies, licw bigb law intermrediate transfers tranfers transfers transfers Type 1981 (1) (2) (3) (4) Subsidies 10.8 0.9 0.9 4.7 8.6 Government subsidies 10.0 0.9 0.9 3.4 7.5 Food 6.0 0.0 0.0 1.8 4.5 Other consumer goodsa 1.4 0.0 0.0 0.7 1.4 Housing 0.3 0.5 0.5 0.2 0.4 Money-losing enterprises 2.3 0.4 0.4 0.7 1.2 Enterprise housing 0.8 0.0 0.0 1.3 1.1 Transfers 1.8 3.1 12.6 3.1 7.3 Social relief 0.2 0.2 0.2 0.2 0.2 Employee benefitsb 1.6 2.9 9.3 2.9 4.3 Paidbygovernment 0.6 0.7 9.3 0.7 4.3 Paid by enterprises 1.0 2.2 0.0 2.2 0.0 Compensation for reduction in subsidies - - 2.5 - 2.2 Unemployment compensation - - 0.6 - 0.6 Total 12.6 4.0 13.5 7.8 15.9 Paid by government' 10.8 1.8 13.5 4.3 14.8 Paid by enterprises 1.8 2.2 0.0 3 .5 I. Note Data are percentages of GDP, at producer prices (see Figure 9. 1). a. Includes subsidies for agricultural inputs. b. Pensions, sickness and disability payments, maternity benefits, funeral expenses, and sirrmilar benefits. c. Possibly financed in part from mandatory contributions by employers or employees to government social insurance funds. Source World Bank projections. price increases in the form of general wage or income ment-sponsored unemployment compensation (Chapter indexation, since this tends to aggravate inflation. Even 8). Finally, the increase in the share of the elderly in selective indexation (of low incomes, for example, or for China's total population (see Table 8.4) will mean in- pension funds) should be approached cautiously. creased pension costs, paid for at least in part by the Government. Future Optionsfor Transfers Table 9.4 presents some illustrative projections of the Over time, the share of subsidies in China's national cost of subsidies and transfers in the year 2000. In 1981, income and government expenditure is likely to decline, transfer payments were very low, consisting of govern- But transfer payments could greatly increase over the ment and urban (primarily state) enterprise social security next two decades, for a number of reasons. One is possi- expenditures for employees and government social relief ble supplementary income payments to poor households, payments. They amounted to less than 2 percent of GDP, 151 with the Government and enterprises each paying about cult for the Government to capture all of these profits I percent.8 Subsidies, in comparison, were relatively high (unless perhaps state enterprises were obliged to contrib- (see Table 9.2). Overall, subsidies and transfers totaled ute to a Government social security fund in proportion nearly 13 percent of GDP, I1 percent covered by the to their wage bill or number of employees). In any event, Government and 2 percent paid by enterprises. much of the cost of extending social security protection In the year 2000, at one extreme (scenario 1), subsidies to people outside the state sector would probably have to might be virtually eliminated (their share in GDP cut from be borne by the Government. 11 percent to I percent), but social security coverage would not be expanded and no major new transfer Revenue Needs and Sources schemes would be introduced. The share of transfers in GDP would nonetheless rise to over 3 percent, because of Long-term government revenue needs depend on the de- the increasing proportion of retirees in the government sired levels of government-financed public consumption, and state enterprise sectors. Subsidies and transfers to- subsidies and transfers, and government saving. Some gether would account for only 4 percent of GDP, with less illustrative projections for China in the year 2000 are than half of this amount paid by the Government. If shown in Table 9.5, which focuses mainly on alternative subsidies are eliminated, however, it is more likely that possibilities for government saving and for subsidies and transfer payments will substantially increase their share in transfers. At one extreme (the High scenario), the Gov- GDP. This is shown in scenario 2, where government- ernment might account for as much as 70 percent of financed social security coverage is extended to the entire aggregate saving (mentioned earlier), while government nonagricultural population, as well as to at least part of subsidies and transfers might be as much as 15 percent of the agricultural population; low-income urban house- GDP (scenario 4 in Table 9.4). Government revenue holds are compensated for the elimination of subsidies in would then have to exceed 50 percent of GDP, compared part through supplementary income transfers (rather than with about 40 percent at present. At the other extreme higher wages); and increased labor mobility generates a (the Low scenario), subsidies might be virtually elimi- modest amount of temporary unemployment (people nated while government transfer payments remain lim- changing jobs and employees of failing firms laid off), ited primarily to civil servants' pensions and disaster relief necessitating a government-financed unemployment (scenario 1 in Table 9.4). At the same time, the Govern- compensation system (possibly funded in part by em- ment might mobilize only 20 percent of total saving-less ployer contributions). Overall, government spending on than half the present level, but still somewhat more than subsidies and transfers would be over 13 percent of GDP, in most nonsocialist countries. Total government revenue nearly all of it transfers. could then be less than a quarter of GDP, as compared In scenario 3, the Government is only moderately suc- with 20 percent in the average developing country in cessful in containing subsidies, but transfer payments are 1977. kept low. Total spending on subsidies and transfers in this Two intermediate scenarios are used in subsequent case would be around 8 percent of GDP, with enterprises analysis in this chapter. In the High Saving scenario, the covering nearly half of the costs. In scenario 4, at the Government accounts for 70 percent of aggregate saving opposite extreme from scenario 1, subsidies fall only (as in the High scenario), and the share of government slightly as a share of GDP while at the same time large new saving in GDP rises from 15 percent in 1981 to 21 percent transfer schemes (covering the entire nonagricultural pop- in 2000. Subsidies are to a considerable extent brought ulation) are introduced: subsidies and transfers total 16 under control and are not replaced by new transfer pro- percent of GDP, nearly all paid by the Government. grams (much as in scenario 3 in Table 9.4). As a result, Much additional spending on transfers would be offset government subsidies and transfers decline from 11 per- by reductions in other expenditure categories. This ap- cent of GDP in 1981 to 5 percent in 2000. These changes plies to income supplements to compensate for reduc- roughly offset each other, so the Government occupies tions in subsidies as well as to unemployment compensa- about the same share of the economy in 2000 as it did in tion (which at present is in effect paid by enterprises as 1981, but it saves more and spends less on subsidies and the wages of redundant employees). Nonetheless, the larger elderly population will clearly absorb additional resources. The degree to which higher transfer payments become an additional burden on government finances 8. Enterprise "transfers" exclude the many wage-like income also depends on financing mechanisms and the tax sys- supplements, bonuses, and "subsidies" presently paid by Chinese tem. Shifting responsibility for transfer payments from state enterprises to their employees. They refer only to payments Shifting ~~~~~~~~~~~~of a social insurance nature, such as pensions, sick leave, and state enterprises to the Government, for example, would maternity benefits. (Enterprise expenditures on education and cause increased government spending to be offset exactly health are considered public consumption rather than transfers, by increased state enterprise profits, but it might be diffi- though the dividing line is somewhat arbitrary-see Figure 9.1). 152 Table 9.5 Government Expenditures, Alternative Projections, 1981-2000 @ercent) 2000 Higb Higb Measure 1981& Higb LOw saving transfers Sbares of GDPb Total uses of government revenue' 41 52 24 42 35 Public consumptiond 15 16 16 16 16 Subsidies and transfersd I1 15 2 5 1 3 Saving 1 5 21 6 21 6 Ratio of government saving to total domestic saving 49 7) 20 70 20 Sbares oftotal government expenditures Publicconsumption 37 31 67 39 45 Subsidies and transfers 27 23 7 10 38 Saving 36 41 25 5 1 17 a. Data are estimates explained in Annex 5, Appendix J. GDP and saving for 2000 are from the QUADRUPLE projection discussed in Chapter 2. b. GDP is measured at producer prices (see Figure 9. 1). c. The size of the govemment deficit (equal to the amount by which government-finance(d investment exceeds government saving) is not projected. Therefore the figures are for total uses of revenue rather than total expenditure. d. Includes only the portion paid for by the Government; some public consumption, subsidies, and transfers are financed by enterprises (see Figure 9.1). Hence, for 1981 the underlyingpublic consumption figure is less than that in Annex 5 (16 percent of GDP before price adjustment), and the underlying subsidy and transfer figure isY 3.7 bilion less than in Table 9.2, this difference being the estimated portion of the housing subsidy financed by enterprises. Source World Bank projections. transfers. In the High Transfers scenario, government sav- Revenue Sources ing would be as low as in the Low case. Subsidies would There is considerable room for choice concerning the be drastically reduced, but this would be more than offset composition of government revenue, even if total reve- by a large increase in transfer payments, reflecting the nue requirements, the structure of the economy, and introduction of government-financed social insurance relative prices are given. Some of the issues and tradeoffs schemes (scenario 2 in Table 9.4). As a result, the share of in this area are presented in Figure 9.2, which is based on subsidies and transfers in GDP would rise somewhat, and the illustrative projections in Table 9.6. overall the share of the Government in GDP would fall Pie A in Figure 9.2 presents the situation in 1981. The from 41 percent to 35 percent. Government relies overwhelmingly on levies on profits China's experience during the past few years suggests and on indirect taxes to meet revenue needs. Agricultural that it will not necessarily be easy to achieve the revenue incomes form a large chunk of the economy but are only growth required even for the two intermediate expendi- lightly taxed (through the stagnant agricultural tax), while ture scenarios. Budget revenue grew only about 40 per- wage incomes are not taxed at all. The share of indirect cent as fast as national income in 1978-82 and grew taxes n national income is somewhat higher than in most faster in 1983 only because of a new levy on enterprise- other developing countries.'0 The "profits" piece of the retained profits and other extrabudgetary funds. None- pie includes two sectors that are taxed rather differently: theless, international experience does not rule out the state and large urban collective enterprises are taxed at a possibility of revenue growth fast enough to achieve the rather high rate (nearly 70 percent on average), while High expenditure scenario (tax revenues in a sample of rural collectives are lightly taxed (8 percent). Pies developing countries grew 40 percent faster than national B-E look at different possibilities for the year 2000, income from 1953-55 to 1966-68),9 although govern- based on the High Saving and High Transfers expendi- ment revenues in China are already high by comparison ture scenarios in Table 9.5, and on the macroeconomic with other developing countries. On balance, given Chi- projections introduced in Chapter 2 (QUADRUPLE for Pies na's political and administrativc capabilities, the level of B-D, BALANCE for Pie E). In all cases, structural change government expenditure seems unlikely to be limited causes the share of wages in national income to be higher simply by capacity to raise revenue. Instead, the central issue will be to weigh the economic and social benefits of higher expenditure against the various distortions and disincentives associated with higher taxation, which may 9. For some countries, revenue growth was even faster. See be more significant in a reformed economic system than Background Paper 4, Table 12. they have been in the past. Io. Background Paper 4, Table II. 153 Figure 9.2 Income Flows and Taxation, Alternative in 2000 than in 1981 and the share of agricultural in- Projections, 1981-2000 comes to be lower. Pie B assumes that current expenditure trends and poli- Wages cies will broadly continue (the High Saving scenario in Agricultural Table 9.5) and that the Government will continue to rely incomes on the same revenue sources (indirect commodity taxes A. 1981 and enterprise profit tax). Indirect taxes remain constant as a share of GDP The tax rate on enterprise profits Profits ) g _ remains relatively high, which means that if nonstate Indirect taxes enterprises continue to grow more rapidly than state en- terprises, profit tax rates for the former would have to be sharply increased (by 1983 the average profit tax rate for rural collective enterprises had risen to 14 percent). B . Traditional revenue, Pie C is based on the High Transfers expenditure sce- B. Traditional revenue, nario (low government saving,_virtual eliminationvoflsub- high saving sidies, and high government transfer payments), but on the revenue side there is continued reliance on traditional sources. Specifically, there are no new taxes on personal incomes, and indirect tax rates are increased to permit lower profit tax rates (needed because most saving must come from enterprises' after-tax profits-see below). As a C. Traditional revenue, result, the share of indirect taxes in GDP rises from 14 high transfers percent in 1981 to 19 percent in 2000, and the average profit tax rate falls from 68 percent to 39 percent. This would permit lower tax rates on the nonstate nonagricul- ture sector and more rapid growth of the sector (with the same profit tax rates as in Pie B). Given the large share of indirect taxes in national income (more than double the D. Personal income taxes, average in other developing countries) it would be crucial high transfers to levy such taxes in a way that minimized distortions (discussed later). Pie D is based on the same expenditure scenario as Pie C, but on the revenue side substantial taxes on wage incomes are introduced (part of which could be social security contributions by employers or employees), and E. BALANCE taxation of agriculture is increased (with the burden as- sumed to fall mainly on farmers with high incomes). This results in a tax structure in which all institutional sectors provide significant amounts of government revenue, and there is no increase in the share of indirect taxes. As in Government revenue Pie C, there could be rapid growth and relatively light After-tax ncome taxation of the nonstate nonagriculture sector. What would happen if the structure of the Chinese Note Each circle represents the economy, with the main income flows economy in the year 2000 were similar to that in the shown as pieces of a fixed "pie. "The colored area toward the center of each piece represents the portion of that type of income that becomes part of BALANCE projection rather than the QUADRUPLE projec- government revenue (through taxes and levies of various kinds.) Thus the tion? As can be seen from Pie E (which is otherwise based sum of all the colored portions represents total government revenue, and on the assumptions of Pie D) the share of wages in the the ratio of the colored area to the total area of the circle is equal to the t share of government revenue in the economy The sizes of the different economy would be considerably larger and those of pieces of the pie are determined by price and wage policies and by indirect profits and agricultural incomes somewhat smaller. tax rates, as well as by the structure of the economy. The share of govern- Lower saving requirements (resuling from greater effi- ment revenue in each piece is determined by the average tax rate on that L type of income (including income, profit, and asset taxes, as well as profit ciency) would mean a lower share of government reve- and depreciation fund remittances and social insurance contributions). The nue in the economy and therefore that tax rates could be piece representing indirect taxes is somewhat different from the others: inecnm its entirety it forms part of government revenue, and the size of the piece reduced somewhat. In Pie E, tax rates on wages and (rather than the share of the piece taxed) is determined by the average rate agricultural incomes are lower, permitting a higher stan- of indirect taxation in the economy. Source Table. 9.6. dard of lving. Alternatively, profit tax rates or indirect 154 Table 9.6 Illustrative Projections for Figure 9.2 (Percent) Agricultural Indirect Government Projection incomes Wages Profitsa taxes Total' subsidies' Pie A (1981) RatiotoGDP 37 25 38 14 114 10 Share taxed 3 0 68 100 - - Revenue as share of GDP I 0 26 14 41 - Pie B (traditional revenue, higb saving) Ratio to GDP 28 32 40 14 114 3 Share taxed 3 0 68 100 - - Revenue as share of GDP I 0 27 14 42 - Pie C (traditional revenue, bigb transfers) Ratio to GDP 28 32 40 19 119 1 Share taxed 3 0 3 9 100 - - Revenue as share of GDP 1 0 15 19 3 5 - Pie D (personal income taxes, bigb transfers) Ratio to GDP 28 32 40 14 114 1 Share taxed 6 1 5 3 7 100 - - Revenue as share of GDP 2 4 1 5 1 4 35 - Pie E (Pie D with BALANCE) RatiotoGDP 26 37 37 14 114 1 Share taxed 4 13 3 7 100 - - Revenue as share of GDP I 5 14 14 34 Note GDP is measured at producer prices (sum of wages, profits, and agricultural incomes) a. Gross of depreciation and subsidies, but net of indirect taxes. b. Equal to GDP measured at producer prices plus indirect taxes. Subtracting government s.sbsidies from this total yields GDP at purchaser prices (see Figure 9.1). c. Excludes enterprise housing subsidy (see Table 9.4). Source: For 1981 (Pie A), estimates based on Chinese financial statistics. For 2000, based en the projections introduced in Chapter 2 (QUADRUPLE for Pies B-D, BALANCE for Pie E), and on the government expenditure scenarios in Table 9.5 For further details, see Annex 5, Appendix J. tax rates could be reduced, or government spending on tax profits compared with historical experience in China public or private consumption could be increased. Thus and in other countries, where the reinvestment rate from improvements in efficiency can help ease budgetary con- after-tax profits tends to be 50-60 percent. Enterprise straints. saving rates of 50-60 percent in this pie would have perm' tted personal saving rates lower than 10 percent (or Implicationsfor Saving higher aggregate saving and growth rates). Implicit in the revenue scenarios discussed above are dif- Pie Dl is based on Pie D in Figure 9.2; it brings out ferent patterns of saving. Figure 9.3 (based on the num- one possible problem of less reliance on government sav- bers in Table 9.7) illustrates some of the points on alterna- ing to achieve a high aggregate saving rate. With the tive sources of saving raised earlier in this chapter. Government accounting for only 20 percent of total sav- Specifically, it shows the implications of some of the ing, and if households save only 10 percent of their dis- revenue scenarios in Figure 9.2 for the composition of posable earnings, the burden on enterprises becomes very aggregate saving and sectoral saving rates. great. With income shares as in the QUADRUPLE projec- Pie Al corresponds to Pie A in Figure 9.2 showing tion, enterprises would have to save 75 percent of their what the situation was in 1981. Profits are highly taxed, after-i1ax profits in order for the economy as a whole to so their share (after tax) in the economy is rather small. save amd invest enough to meet growth targets. State The Government accounts for a large proportion of total enterprises in China have typically not saved as much as saving, while enterprises save a little over half their re- 75 percent of their after-tax profits in the past (the figure tained profits. Pie B I is based on Pie B in Figure 9.2. The has been closer to S0 percent). Moreover, the nonstate share of the Government in total saving rises to 70 per- sector (where a substantial share of after-tax profits needs cent (almost as high as in 1978-Table 9.1); as a result, to be distributed as dividends or bonuses) will presumably household and enterprise saving can be relatively low. In account for a considerable proportion of total profits. particular, enterprises save a low proportion of their after- One way around this difficulty is to increase enterprise 155 Figure 9.3 Sources and Composition of Saving, profits to the point where the required saving rate out of Alternative Projections, 1981-2000 after-tax profits is reduced to a realistic level (which could be accomplished through price or tax adjustments). Pie Sources of saving Composition of saving D2 illustrates this solution. The share of after-tax profits in national income is raised from 25 percent to 37 per- Households cent, while that of personal earnings is correspondingly (3) EnteTrises reduced from 54 percent to 42 percent. This permits (1) i_achievement of a 30 percent aggregate saving rate with a Hous eh_ l Evernment saving rate from enterprise profits of only a little over 50 earnin _ Feue _ percent. But the higher share of profits and the relatively Government low saving rate out of after-tax profits mean that a large Al. 1981 (2) proportion of total consumption would be financed out of profits. Some of this might be in the form of dividends or wage supplements distributed to worker-owners in ] ! !(33_ collective enterprises, but large amounts might also go to high-paid workers in state enterprises, or to owners of individual enterprises (see above). This could lead to so- @ ) _cial problems. An alternative, in many ways more attractive, solution B I. Traditional revenue would be higher saving rates out of personal earnings. Pie D3 shows what happens if after-tax income flows are as in Pie Dl but households save 20 percent of their dispos- able earnings instead of 10 percent. The required saving rate from enterprise profits is reduced to about the same (2) (1) level as in Pie D2, but without the undesirable conse- quences of an increased share of profits. Pie D3 thus somewhat resembles the situation in Japan (where a high Dl. Low household saving aggregate saving rate is achieved through high personal savings). _ ~~~~~~~~~~(3) Tax Reform (2) The revenue scenarios in the previous section illustrate W2 . High profit~ some important choices among broad categories of taxa- D2. High profits tion: iniettxs ae netrrs rofits, and taxes on personal earnings. Within each of these broad catego- ries, there are equally important choices to be made about specific tax instruments, which will play an increas- _ \ (3) _ ing role in China's reformed economic system, both as indirect levers to influence the decisions of enterprises and consumers (see Chapter 10), and as redistributors of income. (2) Indirect Taxes * Saving D3. High household g China now relies heavily on indirect taxes, with a highly U l Consumption differentiated rate structure (3-66 percent). These taxes have several advantages. Revenues generally tend to in- Note. Each large pie chart represents the economy as a whole, but the pieces crease automatically with production, and taxes are rela- are after-tax incomes of households and enterprises and total government tively easy to collect because the number of taxpayers revenue (not just indirect taxes). Wages and agricultural incomes arc lumped (primarily industrial producers and wholesale commercial together, since it is assumed that the saving behavior of these two groups is the same. The colored area towvard the center of each piece of the large pies units) is relatively small. Moreover, rate differentiation represents the part of that type of income that is saved; the sum of all those can be used to achieve income distribution objectives colored areas is total saving, which is the same share of national incomc in each of the pies. Beside each large pie is a smaller one showing the composi- (with some reservations mentioned later), as well as to tion of saving in the economy, which can be compared directly with 't)le discourage the consumption of items such as cigarettes. So9rce Table9 7 China's indirect tax system also has its disadvantages, 156 however. Taxation of intermediate products encourages moderate rate for most consumer goods; and a high rate uneconomic vertical integration and can make effective for luKury goods or goods whose consumption should be tax rates (including the taxes paid at earlier stages of discouraged. If alternative means of reducing inequality production) on particular items diverge significantly from in living standards could be adopted, there might even be nominal tax rates. Moreover, taxation of intermediate a single rate for all but the last category of consumer transactions between state enterprises may not increase goods. In either case, the retail prices of specific goods budget revenue. (including the tax) would be quite closely related to pro- Following reform of producer prices and restoration of ducer prices, thus giving appropriate signals to consumers normal retail commercial margins (discussed earlier), about relative production costs and scarcities. Intermedi- some relatively simple changes could greatly improve the ate goods should not be taxed; nor, perhaps, should in- industrial-commercial tax. One would be a drastic de- vestment goods, except to restrain investment demand or crease in the differentiation of the rate structure. For discourage uneconomic automation of production. This example, there might be only three different rates for approach would allow continued use of tax instruments consumer goods: a low rate (possibly zero) for food; a that China is familiar with, that create few economic distortions, and that generate rapid revenue growth. One problem involves taxation of commodities used both as internmediate goods and for consumption or investment, Table 9.7 Illustrative Projectons for Figure 9 which requires either arbitrary classification or an admin- Housebold Enterprise Government istratively awkward system of tax exemptions for certain Projection earnings, profits revenues Total users. China is already experimenting with value added taxa- PieAl (1981) tion (VAT) for some commodities, but mainly to discour- Share of GDP 61 12 41 114 Saving rate 14b 54 36 - age vertical integration rather than to raise revenue. Share of saving in GDP 9 6 15 30c Wider use of VAT would have several advantages. Reve- Share of total saving 29 22 49 100 nue fiom VAT tends to increase at least as fast as national income. It avoids the distortions caused by taxation of Pie B) (traditional) interrmediate goods. VAT also can be rebated to exporters Share of GDP 60 12 42 114 and charged on imports, which provides incentives to Savingrate 10 24 51 - Share of saving in GDP 6 3 21 30' increase exports and economize on imports in an interna- Share of total saving 20 10 70 100 tionally accepted way (as with other indirect taxes-see Chap:er 6), with a minimum of domestic distortions. Pie DI (low household saving) These advantages may make VAT the most attractive indi- Share of GDP 54 25 35 114 Saving rate 10 75 17 - rect tax, perhaps supplemented by some commodity-spe- Share of saving in GDP 5 19 6 30' cific excise or sales taxes. VAT does, however, require that Share of total saving 18 62 20 100 the taxpayers (enterprises) keep fairly good accounts on input purchases as well as sales, perhaps making adminis- Pie D2 (high profits) tration and compliance more costly than for the indus- Share of GDP 42 37 35 114 trial-commercial tax.' Saving rate I 0 54 17 - Share of saving in GDP 4 20 6 301 An advantage of all indirect taxes over direct taxes Share of total saving 14 66 20 100 (particularly personal income taxes) is that they are more impersonal and less visible, which may make it easier to Pie D3 (high household saving) mobilize large amounts of government revenue. There is Share of GDP 54 25 35 114 also less room for tax evasion, exemptions, and corrup- Savingrate 20 53 17 - Share of saving in GDP 11 1 3 6 30c tion. The key weakness of indirect taxes is that they are Share of total saving 36 44 20 100 not wvell suited to reducing inequality in living standards Note Income flows for nongovernment sectors are gross of depreciation (since they tax all purchasers of a given commodity, not but net of all taxes and levies by the government. GDP is measured at just those who are relatively well-off). This is true even if producer prices. tax rates are differentiated, but it is much more so if there a. Wages and agricultural incomes, after taxes. b. In 1981 personal saving appears to have been unusually high, because of is littIe differentiation in the rate structure (which is desir- large stock increases in rural areas. able for efficient resource allocation). c. The aggregate saving rate in this table is 30 percent because GDP is measured at producer prices. With GDP measured at purchaser prices (as in the 1981 input-output table in Annex 5, and in the multisectoral model), the saving rate is 29 percent (see, for example, the discussion of QUADRUPLE in Chapter 2). The relationship between GDP at producer prices and at 11. However, when VAT is calculated by the crediting method, purchaser prices is explained in Figure 9.1. firms purchasing goods have an incentive to make sure that suppli- Source See Table 9.6. ers pay VAT so that they can receive credit. 157 Profit and Asset Taxes ate share of the surplus this sector generates is channeled Taxes and levies of various kinds on enterprise profits to society in general. Exemptions and differential profit form the other pillar of China's present tax system. Re- tax rates can also be used as economic levers to steer both cently, a proportional profit tax was introduced to replace nonstate and independent state enterprises in appropriate direct remittance of state enterprise profits to the budget. sectoral and geographical directions. In the long run, rates But there is a different effective tax rate for each enter- of taxation for state and nonstate enterprises should prob- prise, to offset the impact of price distortions and other ably be equalized, for administrative convenience and to factors. China has also experimented with a charge on prevent distortions or problems in taxing mixed-owner- state enterprises' fixed assets, but this was apparently ship enterprises. But the great difference in effective abandoned after running into obstacles also associated profit tax rates between urban enterprises and rural col- with widely varying enterprise profitability. In other lective enterprises at present (mentioned earlier) means countries, taxes on enterprise profits have increased in that such equalization should proceed gradually in order importance during economic development and are gener- to avoid large revenue losses on the one hand or excessive ally regarded as an administratively convenient and eco- tax increases for nonstate enterprises on the other. nomically efficient source of revenue (although it is some- Careful attention must also be paid to the relationship times argued that they are in fact passed on to consumers between taxation of nonstate enterprises and taxation of through higher prices). In China, the great variation in their workers and owners. In the absence of personal effective profit tax rates among state enterprises, and the income taxation (discussed below), taxation of profits en- fact that profit tax payments are determined largely by courages tax avoidance through payment of higher wages bargaining, make the present system highly unsatisfactory to workers, managers, or owners. However, to tax both from the viewpoint of generating appropriate incentives enterprise profits and personal incomes derived from af- for enterprises and ensuring steady growth of govern- ter-tax profits is a form of double taxation which can ment revenue. Reform of the price system would make it have undesirable disincentive effects if the rates are too possible to solve the first of these problems, but the sec- high. This, for example, led the United Kingdom to mod- ond also requires improvements in enterprise financial ify its enterprise profits tax system so that distributed discipline and a change in the relationship between state profits (dividends) were subject only to personal income enterprises and the government organizations that now tax, rather than, as in the United States, to both enter- supervise them (see Chapter 10). prise and personal income taxes. Experience in other Once these problems are dealt with, the main issues countries also suggests the desirability of simplified taxes would become the appropriate rate of profit tax and on small enterprises, in order to reduce collection costs whether part of the aftertax profits of state enterprises and to avoid complex bookkeeping. For very small enter- should still be remitted to the Government (perhaps in prises, a lump sum tax or license fee, in lieu of both the form of a "dividend" on the capital provided by the profits and personal income tax, might even be sufficient. Government). There is an obvious tradeoff between In both state and nonstate enterprises, an excess profits greater incentives for enterprises from low tax rates and tax can be used to absorb differential rents earned in the need to mobilize government revenue. In this con- activities like mining. China has already introduced a text, a charge on government-financed investment in "resource" tax, which is triggered once a mining or pe- state enterprises could both raise revenue and strengthen troleum extraction enterprise's profit rate on sales ex- incentives to use capital economically. The charge could ceeds a certain threshold. These taxes can be a useful be a tax on capital assets provided by the Government source of revenue, in principle without distorting incen- through grants, or interest on (and repayment of) gov- tives (indeed they may usefully discourage overly rapid ernment loans to enterprises. A loan or asset tax system depletion of resources). But caution is needed in setting could ensure a minimum, stable return to the Govern- the tax rates because of the difficulty of distinguishing ment on the funds it provides for investment in profit- accurately between differential rent and high profits due earning sectors of the economy. But the stability of such a to greater efficiency or the need to compensate for high charge in the face of varying enterprise revenues and costs risk (most mineral exploration is unsuccessful). would tend to make after-tax profits fluctuate more, com- plicating the financial planning of relatively independent Personal Income and Wage Taxation enterprises, especially if the tax rate were high or the loan Personal income taxes are immensely important in indus- repayment period short. trial market economies, both as a source of revenue and Profit taxes on nonstate enterprises (rural collectives, as a means of reducing income inequality, but they have small urban collectives, and individual businesses) are been largely neglected by socialist countries. China has likely to be an increasingly important source of revenue recently introduced a progressive personal income tax, in China, with the relatively rapid growth of the nonstate aimed mainly at foreign residents. If this tax were ex- nonagriculture sector. They can ensure that an appropri- tended to cover, say, the highest quarter of Chinese wages 158 and salaries (see Chapter 8), as well as high earnings from ing countries. Neither income taxes nor commodity taxes individual enterprises and agriculture, it could play a use- are very effective when the tax base is scattered and ful role in regulating the increases in income inequality largely self-employed, and a large proportion of produc- that are likely to result from reform of the economic tion is consumed in kind. Indirect taxes on the procure- system. In setting tax rates, there is a difficult tradeoff ment of agricultural products by the commercial system, between what are perceived as inequitably low rates and or artificially low procurement prices, coupled with high high rates that discourage effort and stimulate tax eva- taxes (or prices) for industrial inputs into agriculture, are sion. In other countries, the trend in recent years has familiat means of extracting revenue from agriculture. been to reduce rates (which peaked at over 95 percent in But they can severely distort incentives, which has led some cases), while at the same time narrowing loopholes China to reduce reliance on such instruments in recent for tax avoidance and extending taxation of benefits in years. iFurther steps in the same direction would leave kind, capital gains, and personal wealth (especially when Chinese agriculture subject only to the very low (on aver- transferred between generations). Renewed attention has age about 2 percent) direct agricultural tax, though of also been given to personal expenditure taxes (which tax course also to general indirect taxation of manufactured consumption but not saving) as an alternative to personal consumer goods and services. income taxes. It is important to distinguish between the desirable Extending the personal income tax in China to cover level of agricultural taxation and the choice of instru- the great majority of wage earners would make it a major ments for extracting it. So long as agricultural incomes source of revenue. The extent to which this should be are on average much lower than nonagricultural incomes done depends largely on the attractiveness (including the (as seenis likely for the next few decades-see Chapter 3), comparative administrative costs) and availability of alter- low direct taxation has obvious merits. But shortfalls of native revenue sources, especially indirect taxes. Experi- revenue from other sources might make heavier direct ence elsewhere suggests that income taxes on wages and taxation of agriculture unavoidable. Moreover, progres- salaries are cheap to collect, because employers can be sive taxation of unusually high agricultural incomes may obliged to deduct them from wages. Personal income be desired for reasons of equity. The best instrument for taxes, however, probably constitute a greater disincentive this last purpose would in principle be the personal in- to individual effort than equivalent amounts of indirect come tax, especially because it demonstrably applies the taxation. Moreover, they can be difficult to collect from same standard of equity to agricultural and nonagricul- individual enterprise proprietors and other self-employed tural iiicomes alike. In practice, however, it could be people, who in some countries evade most tax. difficul: to collect such a tax on agricultural incomes- Taxes on wages, including employers' and employees' even if it was limited to a relatively small number of high- social insurance contributions, are also an important incomc farmers. source of government revenue in industrial market econ- As a means of raising large amounts of revenue from omies: in 1981, for example, social security contributions agriculture, personal income taxation (because of its high were 25 percent of all tax revenues in the United States, admini,trative costs) is probably inferior to the present 39 percent in West Germany, and 44 percent in France. agricultural tax. The latter is in essence a land tax based In China, wage taxes (discussed in Chapter 8) could be on estimated income-earning potential, levied in absolute used to finance social insurance and pension schemes. But amounts that remain fixed for several years at a time. for employment and income redistribution reasons, it This tax, whose fixed nature provides stronger incentives might be desirable to exempt low-paid workers and to to increase production in efficient ways than an income levy higher rates on higher wages. In most other coun- tax, coild be levied at higher rates. It could also be made tries, by contrast, such taxes not only apply to all work- more flexible by expressing the tax as the product of a ers, but also have an upper limit such that no one pays standard percentage rate, common to all land, and a "tax- more than a certain absolute amount, in order to reduce able value" peculiar to each plot of land (this is the stan- inequality in contributory social insurance benefits. dard form of urban property taxes in other countries). The overall tax yield could then be varied by altering the Taxation of Agculture standard percentage rate, with the taxable value of partic- Agricultural taxation is a difficult matter in most develop- ular plots subject to periodic reassessment. 159 Development Management A recurring theme of all the preceding chapters is the usually also seeks to limit inequality in income distribu- critical importance of system reform to China's future. tion, and in particular to assist the poorest and least This is well recognized in China today. Over the past five privileged. China's achievements in these areas have been years, far-reaching reforms introduced in rural areas have most impressive and compare very favorably with those had a remarkable impact on production and income. of other countries at similar levels of development. Many While reforms in urban areas have been much less defini- of these social objectives, however, are still being met tive and comprehensive, the Central Committee Decision through policy instruments designed when the country of October 1984 indicates a firm determination to accel- was much poorer and the economy and society were less erate overall reform.' Together, the rural reforms and the stable. October Decision constitute one of the most far-reaching Perhaps more than in any other country, social welfare attempts to improve the functioning of a socialist eco- and services for the urban population in China are pro- nomic system. Moreover, this attempt is only the begin- vided by their enterprises or work units, with the state ning of China's overall strategy of building "socialism playing a limited role. Health and disability benefits are with Chinese characteristics," a system "integrating the provided through a labor insurance scheme, but pensions basic tenets of Marxism with actual conditions in China." are paid as part of current production cost and housing is System reform in China over the next two decades will usually provided by employers. Many social and distribu- require consistent and comparable progress in three areas: tional objectives are met by state manipulation of eco- (a) increasing economic dynamism and efficiency through nomic instruments. For example, wages are set relatively market regulation; (b) strengthening and reforming social equal and generally at such a low level that they have to institutions and instruments to safeguard and further im- be supplemented by price subsidies on essential goods and prove social welfare and equity; and (c) strengthening services. Economic security is provided to urban workers economic management through the coordinated use of through "lifetime" job guarantees. Full employment is indirect and direct controls. This final chapter, drawing achieved by forcing enterprises to hire more workers partly on earlier chapters, discusses general issues in these than they need, by restricting migration from rural areas, three areas, as well as some specific issues not covered in and sometimes by sending urban youths to rural areas. earlier chapters. The first section discusses the role and For the rural population, a subsistence level of grain is responsibility of the state toward the individual, focusing provided to the poorest households. on issues of income inequality, economic security, and Although egalitarianism was a guiding principle in the basic social services. The second section reviews the rela- past, the level of welfare or security provided has been tionship between the state and state-owned enterprises, as uneven. The quality of social services and housing pro- well as collective and individual enterprises, and proceeds vided to workers in different enterprises varies substan- to a discussion of investment and prices. The following tially and in ways unrelated to the performance of the section discusses the role of the state in planning and workers or the enterprises. State-owned enterprises gen- managing the economy. erally provide better services and greater benefits than The State and the Individual 1. Communique of the Third Plenary Session of the Twelfth In a modern society, the state generally provides a degree Central Committee of the Chinese Communist Party, October 20, of economic and social security for all of its citizens. It 1984. All quotations in this chapter are from this document. 160 collectives, while the self-employed receive few benefits. pable managers, hard-working farm households, and tal- In rural areas, welfare and social services are provided ented innovators in all fields. The resulting degree of mainly through the collective efforts of the local popula- income disparity might well be larger than socially desir- tion, with limited assistance from the state, so that the able. Eut rather than distorting or suppressing economic quality of service varies greatly among localities and is forces, it would be preferable, to the extent possible, to generally much below that in urban areas. deal with inequality through social policy instruments Two important questions must be addressed: Do Chi- designed for this purpose. Specifically, high incomes gen- na's social policy instruments still meet the social and erated by a more market-oriented economic system could equity objectives of a modernizing society rapidly ap- be reduced to socially acceptable levels by directly taxing proaching middle-income levels? And are they compatible the incomes of the rich more than the poor (progressive with the other changes that are taking place in the eco- income taxation), by making high-income earners con- nomic and social system? Unwillingness to confront the tribute more to social security, or by imposing high taxes possible social consequences of economic reform has been on goods and services consumed by the rich. Some op- a critical flaw in many East European attempts. The tions in these areas were discussed in Chapters 8 and 9. result is often an incorrect impression that reform must It is essential, however, that income taxation-or other entail the abandonment of social concerns. This risk ex- measures to reduce income disparity-should not be seen ists in China also, perhaps to an even greater degree than as a punitive measure against high-income earners. In- in other socialist countries, because income distribution deed, continuous education of the population about the and economic security objectives are being met in un- policy of the Government-and the Party-on the legiti- usual ways, and particularly because social instruments macy of high incomes will be necessary. It would be are so entangled with economic instruments. better for the central government to set a clear policy on this issue, and to take the lead in introducing the policy Issues of Inconze Dispanty instrurients necessary to reduce income disparity, than to It is recognized in China today that "prosperity comes allow local communities to set their own standards-as unevenly in the course of economic development:' and curren-ly seems to be happening in some places. that reform will require a new way of thinking about International experience suggests that it is better to different types of income and income inequality. An espe- confront the problem of high incomes directly. Faced cially difficult issue in this regard may be the legitimacy with the dilemma of having to reward very productive or and possible size of incomes from entrepreneurship. important members of their societies in the face of politi- Some people in China, unfortunately, still associate entre- cal pressure for equality, some countries have tried to preneurship with all the worst aspects of pre-Liberation obscure high incomes and consumption. In some coun- capitalism. Entrepreneurship, however, will be critical in tries, for instance, the wages of public officials have been achieving China's economic objectives, partly by causing kept low, but extensively supplemented through bribery markets to function more smoothly and efficiently. Entre- and ccrruption. In other countries, formal incomes are preneurs, for instance, buy goods at low prices in places low arid fairly equal, but senior officials and important where they are in surplus and sell them at higher prices in members of society receive special privileges such as bet- places where they are often urgently needed; or they ter housing, second homes, access to special stores provid- produce and sell necessary goods and services initially at a ing goods not available elsewhere, and sometimes even high profit, thus luring others into the same activities, unrecorded but regular supplements to their official sala- which increases supply and drives profits down.2 Al- ries. The benefits of such obfuscation are strictly tempo- though their motivation is obviously to make money for rary, s.nce these practices quickly become well known, themselves, they are nonetheless useful. Entrepreneurship and th, costs are large and long term, as they may set in is also essential for technical or economic innovation- motion an irreversible evolution toward a corrupt and the search for ways to produce something at lower cost hierarchical society. or of higher quality and to respond to changing market conditions when there are no assurances that the search will be successful (see Chapter 7). Because entrepreneur- ship involves taking risks and is successful only occasion- 2. Especially in the early stages of economic reform, reflecting ally, it will floutish only when successes are well re- the huge imbalances created by the previous rigidly administered warded. system, entrepreneurial incomes tend to be large even in some In a socialist economy, income disparity resulting from activities that would not normally result in such high incomes. ofpropersty wilabe reltiv limited. Some entrepreneurs in China are already earning annual incomes private ownership of property will be relatively limited. of Y 10,000, even Y 100,000, by raising chickens and selling eggs But considerations of economic efficiency and dynamism to the urban population, by opening small restaurants, or by per- will require that hgh incomes be permitted-and indeed forming badly needed household services, all of which were previ- encouraged-for successful entrepreneurs, unusually ca- ously prohibited. 161 Social Security and Welfare to build on. In urban areas, a household registration sys- At the other end of the income spectrum, a social secu- tem, which includes detailed information about each rity program consistent with the needs of a reformed household's situation, forms the basis for grain and other economy is urgently needed. Enterprises must function product rationing. This system could be used initially in increasingly as economic entities, which will inevitably administering social security programs. Even in remote conflict with their current role as providers of many so- rural areas, the former system of communes, brigades, cial services. Prices and wages need increasingly to reflect and teams proved highly effective in reaching every economic considerations and cannot be manipulated to household and in pursuing economic and social objectives the same extent as in the past to meet social objectives. with limited resources. Although the economic functions Rapid economic change will mean accelerated growth of of these institutions were eroded by agricultural reform, some enterprises but contraction and closure of others. they or their successors could continue to administer col- Some skills will become obsolete, and some workers- lective social services as well as state-run social security sometimes through no fault of their own-may become programs. at least temporarily unemployed. Thus, the absence of The major function of most social security programs is any organized means of supporting the unemployed and to supplement the income of households whose incomes others suffering from economic change will become in- are low either temporarily (for example, unemployed or creasingly unacceptable. Rising expectations among the ill workers) or permanently (for example, retired workers people as a consequence of economic growth will also without other pensions). In this connection, the challenge increase pressure for the state to assume a growing and is to provide for those in need while minimizing disincen- increasingly direct responsibility for social security and tives to work. Providing all households with a fixed in- welfare. come supplement (or "social dividend"), regardless of Especially in the last few decades, social security pro- their work and other sources of income, would minimize grams have been introduced in all industrial countries, any disincentive to work (since they would keep all other with the objectives of ensuring an adequate level of in- income), but would be far too costly an option for China come for all citizens and minimizing the risk of poverty to consider. A guarantee by the state to bring the incomes due to old age, illness, disability, unemployment, and of all households up to a given level would be less costly, other personal misfortune. These programs have grown but would remove incentives to earn income below the rapidly, and in most Western industrial countries they guaranteed level. Concentrating income supplements on now account for 15-20 percent of national income. families with special characteristics (such as retired people Many, however, have evolved from a series of smaller, with no other income, single-parent households, and fam- fragmented programs, each responding to particular so- ilies with many children) reduces budgetary costs and cial, economic, and political exigencies. Although individ- disincentives, but can increase administrative costs and is ual programs may be quite sensible, the social security inequitable for other impoverished households. Thus all programs as a whole are not necessarily so. In some programs involve some tradeoffs among budgetary and countries, for instance, complex income tax and welfare administrative costs, disincentives, and inequities. Careful regulations have combined to produce unintended ab- design is needed to suit a country's particular social and surdities. In some West European countries, generous economic conditions and priorities.3 unemployment benefits for young people have contrib- In most Western industrial countries, social security is uted to increased unemployment by deterring them from based largely on the insurance principle-citizens contrib- taking arduous, dull, or low-paid jobs. Despite these and ute to a fund that then compensates them for loss of other problems, however, there is little doubt that large income due to old age, unemployment, illness, and so and steadily increasing government spending on social forth. The level of individual benefits sometimes depends security, medical care, education, and other social services on the size of individual contributions. The main role of has contributed greatly to the reduction of inequality in industrial market economies. A major potential advantage of social security pro- 3. All options involve some compromise. Many economists grams is that benefits go only to those in need. By con- propose that the state should establish a basic (maximum) supple- trast, price subsidies-the most important way in which ment, whose amount would vary according to the size and other the state attempts to assist the poor in China today- consumption needs of the household. The amount actually paid to benefit the rich, as well as the poor, and are therefore an individual household would depend in some specified way on much more costly. Compared with price subsidies, com- the income earned by the household, so that a household with no prehensive social security programs are of course adminis- other source of income would receive the full basic supplement, while a household with income over a cutoff level (which would tratively much more complex. Yet in this regard, China have to be higher than the amount of the basic supplement) would has considerable advantages over most other countries, receive no supplement at all. This sort of scheme is sometimes particularly since it already has an institutional structure referred to as a negative income tax. 162 the state is to administer the program, sometimes aug- may be desirable to give the local administrators in rural menting its resources with other government revenues areas substantial freedom in organizing the entire range of when necessary, and to provide a "safety net" of noncon- social services-including social security, pensions, health, tributory income supplements for those who are not cov- education, and family planning-subject to some national ered by insurance or whose income would otherwise fall policy guidelines. There is also a need for what may be below an established minimum level. An alternative to termed "preventive welfare" (similar to preventive medi- this insurance approach, however, would be for the state cine), which involves social assistance of various kinds to to assume direct responsibility-as China has in the improve the earning capacity of individual families. For past-for ensuring a minimum level of economic security instance, the unemployed-in addition to income sup- and to finance this from general budgetary revenues. The port--may require retraining, help in obtaining new em- state-run and budget-financed scheme could then under- ployrnent, and possibly aid in organizing for community pin additional schemes, which would use the contribu- work programs. tory, insurance principle to provide extra benefits. Thus there will be a growing need for state agencies to A possible scheme following this second approach, and assist individuals in coping with the changing economic which would also address the problem of high incomes, environment.4 Institutions for this purpose will need to would be to establish a nationwide social security fund. be d2veloped and strengthened. One important step High-income earners would be required to contribute to migh: be to consolidate and extend the present enter- the fund a portion of their wages or incomes above an prise-based labor insurance and welfare programs into a established level, while households with incomes below a modern state-run social security program. Because of the specified minimum level would receive an income supple- man} complex issues involved-including administrative ment. Eligibility for the supplement would extend to arrangements, financing and use of funds, determination households afflicted by old age, unemployment, disabil- of beneficiaries and benefit levels, and relations with the ity, and other adverse natural or economic circumstances, state budget and other insurance schemes-it might be but not to those who simply refused to work. The costs desira.ble to create a high-level social security commission of these income supplements-which are likely at least to formulate a long-term plan to be implemented in initially to be considerably greater than the contributions stages. of high-income earners-would be financed largely from budget revenues. Such a scheme would minimize admin- Social Security and Services in Rural Areas istrative costs, because the majority of households, which A dif.icult issue in planning for a social security program, have middling incomes, would not be involved. Since which affects both equity and cost, is the proportion of high-income earners would be making a direct contribu- the population to be covered.' With regard to the urban tion to the relief of poverty, it might also alleviate the popuation, any state-run social security program should apparently growing resentment against them in some surely cover everyone, whether employed in the state, communities. Moreover, this kind of scheme would help collective, or individual sectors. To do otherwise would to ensure that reform and policy measures introduced be inconsistent with the current objective of promoting a later (for example, allowing enterprises to dismiss work- multiple-ownership system in the economy. Workers in ers, widening wage scales, and eliminating subsidies for town,hip and village enterprises, as well as in collectives food, rent, and other goods) would cause no families to in small towns, should probablv also be included. Over fall below a socially acceptable income level. the past few years, workers in collective enterprises in Such a national scheme for providing a low but uni- some rural townships have begun to organize their own form level of income support throughout the country pension and social security programs. Although these ini- could be augmented by provincial and local schemes fi- tiatives should be encouraged, local governments might nanced by local funds or contributions from participants. attempt to consolidate these fragmented efforts into Groups of workers (in particular occupations or localities) could also be encouraged to save for retirement and share risks in areas such as health and disability. Especially for 4. IRecently, for instance, after introducing an economic respon- retirement, a minimal social security program should be sibility system, the Capital Iron and Steel Corporation found that seen only as an underpinning for self-financing pension severa thousand workers were redundant. The corporation orga- schemes sponsored by local governments, employers, or nized these workers to form a service company and to work on other groups. (Options for organizing pension schemes environmental improvement projects, while some were retrained. have been discussed in Chapter 8). Such efforts are highly commendable, and more of them will be needed. Ultimately, however, responsibility for retraining redun- Other countries' experience indicates that management dant workers and helping them find new jobs will have to rest of social security programs can often be more effective if mainly with the Government. closely linked to the provision of other social services, 5. The possible future fiscal implications of social security such as health and education. In China, for instance, it schem s of varying comprehensiveness were outlined in Chapter 9. 163 larger programs, which would be more effective because money and effort should likewise be put into boosting of their larger numbers of participants. school attendance (Chapter 8). China has long experience Extension of a social security program to agricultural and proven capability in such matters, but action is ur- workers and their families would be a more difficult and gently needed. complex task. In all low-income countries, as in China, The problems of poor areas can be exacerbated by the clash between the desire for equality of social welfare misguided concern about equality-for example, paying for all people and the limitations of administrative and workers the same as in more developed areas, thereby financial resources has led to different treatment of the impeding the development of poor areas (Chapter 5). But rural and urban populations. The question is whether this it is also necessary to guard against mistakes of the oppo- compromise remains necessary or desirable in China- site kind. For instance, some localities in China are appar- especially because economic trends by themselves may ently making even the basic education and health systems not sufficiently narrow rural-urban income disparities financially independent, which is causing their scope and (Chapters 3 and 5) and because poorer areas are likely to quality in poor areas to deteriorate further. In some of fall further behind other parts of the country (Chapter 5). these localities, moreover, the generally sound concept of The state may therefore need to intervene more force- "cost recovery" -making beneficiaries pay the full cost of fully in the social sphere to alleviate emerging inequality services-is being incorrectly extended to basic social ser- and potential poverty. vices. These developments seem contrary to the objec- A rural social security program subsidized by budget- tives of system reform in China and need to be corrected ary revenues, for instance, would be an efficient means of through policy directives from the central government. transferring nonagricultural productivity gains to the agri- cultural population (Chapter 3). While the long-term tar- The State and Enterprises get should be to have a uniform basic social security program for the whole population, an interim measure The recent Central Committee decision has identified the might be to combine and augment the existing subsis- task of "invigorating enterprises as the key to reforming tence grain support and Five Guarantees into a rural the national economy," and acknowledges that "socialism social security program providing needy households and with Chinese characteristics should, first and foremost, be communities not only with basic food, but also with a able to instill vitality into the enterprise." Attention has small cash income. The amount of cash income and other been focused on state-owned enterprises (or more accu- benefits could then be increased over time as incomes rately, enterprises owned by the whole people); however, rise, with the objective of eventually merging the rural policy toward nonstate enterprises-collective and indi- social security program with the urban program. vidual-will also be critical in China's reform of the ur- A disturbing aspect of recent rural reform has been the ban economy and could indirectly contribute to reform weakening of the cooperative health system in poor ar- of state enterprises. eas; school attendance has also been adversely affected, and in some poorer localities is now rather low (Chapters Control and Management of State-Owned Enterptises 1 and 5). These trends should be of major concern to Over the past five years, various reforms in the manage- China's planners. There is overwhelming international ment of state-owned enterprises have been introduced, evidence that investment in human resources-in basic mostly on an experimental and piecemeal basis. These health and education-has not only been a most effective have included: enterprises retaining a proportion of way of helping the poor, but has also contributed directly profits for workers' bonuses, collective welfare expendi- to economic growth.6 Thus, neglect of basic education tures and small-scale investment; some freedom for enter- and health runs counter to the pursuit of both social prises in selling part of their output and in securing in- equity and economic efficiency. puts; some flexibility in setting prices; and increases in The state should probably finance an increased propor- managerial authority over production decisions and per- tion of the costs of social services in poor areas, although sonnel matters. These changes, similar to reforms tried in self-reliance may remain satisfactory for well-off rural ar- nearly all socialist countries, have had some good results. eas. Because cooperative schemes have proved effective in They represent only marginal changes, however, and the the past, assistance could take the form of matching fundamental problem remains of the proper relationship grants to townships or villages below a certain average between the state and the enterprise. income level (that is, the state would provide an amount In the past, "ownership by the whole people was taken equal to, or some multiple of, the amount of funds col- to mean direct operation by state organs." As a result, lected through the cooperative system). Assistance by the state could also include assigned government health workers, subsidized training of local health workers, and 6. See, for instance, World Bank, World Development Report priority to poor areas in state health investment. More 1980 (New York: Oxford University Press, 1980), Chapters 4 and 5 164 "the state exercised excessive and rigid control over enter- Giving direct control of state-owned enterprises to prises" and "enterprises became subordinate bodies of their workers would promote a strong sense of economic administrative organs." The relationship between the democracy and worker participation. Japanese experience state organs and the enterprises was strictly hierarchical, also shows the economic advantages of strong worker and the former frequently interfered in enterprise opera- commitment to the well-being of the enterprise, based tions. An important step has been taken, therefore, with not just on profit-related bonuses but also on constant the recognition in China that "ownership right can be education and frequent consultation. But the concept of duly separated from operating right." The direction of worker management suffers from the fundamental prob- reform will be toward a more complex system of manage- lem that state enterprises should operate for the benefit of ment, in which a multitude of state agencies, as well as the whole society and not only of those who work in the enterprise itself (both workers and managers), assume them--a particular problem in China where state enter- various responsibilities. Since the state retains the author- prise workers are already a relatively well-off group. Ex- ity to determine the division of rcsponsibilities among the perierLce in Yugoslavia and elsewhere, including that of various state organs, including enterprises, it will, of nationalized industries with powerful labor unions in cap- course, retain ultimate control. italist countries, suggests that worker control could result Though the principle that state enterprises should have in excessively high wages and worker benefits, inadequate much greater autonomy than in the past is accepted, labor discipline and effort, restrictions on employment important and difficult questions remain. One is the (partly though excessively capital-intensive investment de- propcr degree of autonomy: should it be confined to day- cisions), and indiffcrence to profitability-net of pay- to-day operating and marketing decisions, or should it ments to workers-beyond the minimum necessary for extend to appointment of managers, major investment enterprise survival. and diversification decisions, and the right to close down Experience in industrial market economies suggests part or all of the enterprise? Some possible answers to that giving control of state enterprises instead simply to this question are considered in the next section of this their managers would have important advantages, includ- chapter. ing a strong urge to expand, improve, and innovate. It Another question is how the state could or should also suggests, however, that completely independent en- regulate the activities of autonomous enterprises. With terpris;e managers sometimes choose a quiet life, or, more the exception of strategic and key enterprises (discussed commrronly, seek personal power through expansion, with below), it is impossible in a complex and rapidly changing insufficient attention to profitabilitv. In socialist coun- market-regulated economy for the state to know or, spec- tries, moreover, managerial control tends to differ only ify what individual enterprises should do, which may slightly from worker control, since managers often find it make it pointless or counterproductive for the state to hard to resist worker demands for greater benefits or to attempt to direct the activities of enterprises from within. insist on the often unwelcome changes in work practices Instead, the state should primarily seek to create an exter- needed for innovation and increased efficiency. This has nal environment such that the self-interest of enterprises been the Hungarian experience. It is also already a prob- guides them in directions consistent with the national lem in China: studies of experimentally reformed enter- interest, thus making it unnecessary for managers or oth- prises controlled mainly by their managers reveal large ers within enterprises to be charged specifically with rep- increases in worker benefits, but small increases in eco- resenting the interests of the state. Such an environ- nomic efficiency. ment-whose creation is the subject of much of the rest An alternative approach, common in nonsocialist of this chapter-involves appropriate prices, wages, and countries (and now being tried in Hungary), is to give interest rates, a high degree of competition, and a well- strategic decisionmaking authority in each enterprise to a specified legal framework, as well as the use of regulatory board of directors. The board could contain some repre- levers such as taxes, subsidies, and credits. sentatives of society at large, as well as of the workers. How enterprises are controlled and managed internally But to provide the necessary motivation, the board will largely determine what constitutes their self-interest. woulc mainly have to consist of (or represent) institutions Even with an ideal external environment, inappropriately with a strong interest in the enterprise's profits. This motivated enterprises-especially those that do not ac- woulc. be the case, for example, if the institutions were tively and continuously seek increased profits (discussed owners entitled to dispose of the enterprise's after-tax later)-may act contrary to national interests. Moreover, profits, either by reinvesting them in the enterprise itself experience in China and other countries confirms the or by withdrawing them for consumption or investment difficulty of devising appropriate internal management elsewhere. The board could then insist that the enter- arrangements for state enterprises. The various alterna- prise's managers behaved appropriately, partly by its au- tives to the past system of direct administrative control all thorit:v to appoint and dismiss them, partly by linking have weaknesses as well as strengths. their r emuneration to profitability. The managers might 165 in practice make most decisions, even major ones, but activities that are no longer wanted by the market or would be greatly influenced by their ultimate account- where they cannot compete with more efficient enter- ability to the directors. prises. Specifically, efficient enterprises should be allowed Experience elsewhere makes it clear that merely estab- to grow rapidly, through the use of their retained profits lishing boards of directors for state enterprises is not and other sources of finance (financial flows are discussed enough. What is necessary, in addition, is that these in a later section), in new activities as well as existing boards should be not just profit-oriented, but also free activities. They should be permitted to compete for mar- from direct intervention by state administrative organs. kets, for investment funds, for land, and for foreign ex- However, precisely because these are state enterprises, change. The objective should be to make the widest use this may be difficult to achieve. Even if an enterprise's of the superior management and entrepreneurial skills of board of directors were to consist of representatives of these enterprises. the Ministry of Finance or a new Ministry of State Prop- Poor performers, however, should bear the full force of erty, rather than representatives of the relevant sectoral market competition. They should be forced to adjust and ministry or the government of the locality where the improve, and if this is not possible, to reorganize, seek enterprise is situated, informal connections and pressures new kinds of activities, or close down if necessary. In this could effectively perpetuate direct government control of connection, China needs to begin formulation of laws the enterprise. and regulations on the treatment of enterprises, both A possible solution might be to spread the ownership state-owned and others, in financial difficulties. When of each state enterprise among several different institu- assistance through state intervention is clearly desirable tions, each in some way representing the whole people, for social reasons, this is best provided through financial but with an interest mainly in the enterprise's profits subsidies (including, for example, wage subsidies), so that rather than directly in its output, purchases, or employ- the costs of such assistance are explicit. Assistance ment. Examples of such institutions, in addition to cen- through restriction of entry for other enterprises' prod- tral and local governments, are banks, pension funds, ucts or other means of hampering competition is much insurance companies, and other enterprises. This possible less desirable. Financial subsidies should be given for a system of socialist joint stock ownership has no parallel fixed time only, so that frequent reviews can determine elsewhere, but contains elements found in other coun- whether the subsidies continue to be justified. tries, including extensive enterprise ownership by pension For enterprises in financial difficulties, but for which funds in most industrial market economies and by vari- state subsidies are not justified, explicit procedures for ous levels of government and socially owned banks in reorganization or closure are necessary. Most countries West Germany (see Box 10.1). In China, such a system have bankruptcy laws, for instance, stipulating a time of socialist joint stock ownership could perhaps be cre- period during which an enterprise would be allowed to ated initially by suitable dispersion of the ownership capi- stop paying its debts while it reorganizes its finances. tal of existing state enterprises. Over time, it could be re- These laws usually include the possibility of having a inforced by a more diversified pattern of investment court- or government-appointed administrator take over finance (discussed later), with a varietv of state institu- the management of the enterprise and define the rights tions acquiring financial interests in existing and new and responsibilities of the existing board of directors, enterprises. managers, and workers. They also usually specify legal procedures for closure of plants, disposition of assets, and Con2petition treatment of workers and managers, which would some- The critical importance of competition in promoting effi- times be handled by specialized bankruptcy courts. ciency and innovation is increasingly recognized in Any reorganization of enterprises that impedes compe- China. Competition means that "enterprises will be di- tition should be avoided. In China recently, and fre- rectly subjected in the markets to the judgment and eval- quently in the Soviet Union and Eastern Europe, enter- uation of the masses of consumers, so that the superior prises have been reorganized-partly in an attempt to will survive and the inferior be eliminated." The Party's separate enterprise management from state administra- unambiguous position on this issue is crucial, because tion-into large holding companies, trusts, or centrale. promoting competition among state enterprises will re- Often the holding companies are economic entities in quire not only changes in policy, but also fundamental name only, and the result is to substitute one form of changes in the attitudes of many people. bureaucratic intervention for another. Moreover, even if An essential policy to promote competition is allowing these companies truly function as enterprises, there is a free entry into all kinds of economic activities. Enter- high risk that they will quickly become monopolistic, prises (both state and nonstate) should be given the right which may be even less desirable than having enterprises and, indeed, encouraged to enter new kinds of activities if controlled by administrative organs. China is fortunate they have the capability to do so, and to withdraw from that its vast size offers the possibility of domestic compe- 166 tition in most sectors, but steps will have to be taken to Effective competition, and indeed most other aspects of discourage local governments from imposing restrictions reform, will also depend on expansion and improvement on internal trade and competition (Chapter 5). Im- of China's legal system. A decentralized economy, based provements in transport, commerce, and communica- primarily on horizontal linkages among enterprises, with tions, especially the transmission of market information, a lot of internal trade and complicated financial arrange- will also be needed to create a unified national market ments, cannot function properly without a comprehen- (Chapter 5). In many sectors, much greater competition sive system of commercial and contract law, and the from imports will be desirable, as will competition of institutions and personnel (accountants and auditors, as more Chinese enterprises in world markets (Chapter 6). well as lawyers) needed to implement it. Laws are also Box 10.1 Dispersed Social Ownership in the Federal Republi, of Germany West Germany has a market economy dominated by pri- have apex institutions (Landesbanken), which are some- vate ownership of the means of production. Public enter- times owned by groups of local savings banks, sometimes prises (those with at least 50 percent public ownership of partially or wholly by state governments. A long-term equity or voting rights) account for only about 10 percent development bank (the Kreditanstalt fur Wiederaufbau), of GDP. But the true extent of social ownership is much owned 110 percent by the federal government and 20 per- greater. Minority government ownership in enterprises is cent by states, plays a major role in financing government- widespread; for example, the nominally private energy and approved projects in developing countries, in export fi- industry giant VEBA is 44 percent owned by the federal nance, and in promoting regional development and aiding government. VEBA in turn has holdings in many other small businesses. firms, some of which own still other businesses. In some An example of the ownership ties that link West Germa- corporations like the Rhine-Westphalia Electrical Works ny's soc ally owned banking system with key sectors of the (RwE), local governments exercise voting rights that are a economy is the Westdeutsche Landesbank (WestLB), the large multiple of their nominal stock ownership. largest cf the regional apex institutions for savings banks In addition to being widespread, social ownership in and the third largest bank in the country overall. WestLB's West Gerrnany is highly dispersed, among federal, state largest shlareholder is the State of North-Rhine/Westphalia (provincial), municipal, county, and community govern- (43.2 percent); the other owners are associations of savings ments and various kinds of cooperative or publicly owned banks in Rhineland and Westphaiia-Lippe (16.7 percent banks. Multiple ownership is common, and in general each) arid associations of local governments in the same public ownership does not mean close operational ties two regions (11.7 percent each). WestLB is a significant between enterprises and individual government agencies. shareholder in thirty-three major corporations and scores Social ownership is pervasive in West German banking. of othei enterprises in such diverse fields as housing and In 1980 five of the ten largest banks were socially owned, real esta:e, finance, leasing, energy, engineering, mechanics as were all of the next tier (eleventh to twentieth in size). and optics, construction, machinery, textiles, and com- Socially owned banks account for around 60 percent of puters. Some of these companies in turn own other busi- the total volume of bank business. Banks have a considera- nesses. For instance, WestLB is a 32 percent owner of ble ownership stake in many firms. They also tend to have Kommunale Energie Beteiligungsgesellschaft, a large en- close credit ties with particular enterprises and are heavily ergy cor,glomerate whose other shareholders are a number represented on boards of directors. Most banks engage in a of local governments at various levels. This company in variety of activities; even local savings banks may lend to turn is a 32.6 percent owner of Vereinigte Elektrizi- or own industrial enterprises, directly or through their tatswerAe Westfalen, which itself has holdings in eleven apex institutions. other erterprises, including a mining company which is a The socially owned banking sector is characterized by a small shareholder in Ruhrkohle, West Germany's biggest diversity of forms. Urban and rural cooperative banks are coal-mir ing concern (whose largest shareholder is VEBA). permitted by law to lend 50 percent more than other Widespread social ownership in West Germany appar- banks with the same equity capital, which has enabled ently does not lead to inefficiencies. This is probably due them to grow rapidly in recent years. Trade union-owned in part to the fact that government and other social entities banks, originally established to manage the funds of trade owning enterprises generally act like ordinary stockhold- unions, have been amalgamated into the Bank fur Ge- ers, particularly in their detachment from the daily opera- meinwirtschaft, West Germany's ninth largest bank in tions of firms. Moreover, the dispersed and often indirect 1980. The local savings banks (Sparkassen) are not character of social ownership and the coexistence of a large 'owned" by anyone in the traditional sense, but are private sector mean that competitive pressures to maintain "guaranteed" by the lowest level of government. They efficiency are preserved. 167 needed to provide for the founding and closing of enter- profit is a more difficult issue. Some large enterprises in prises as well as to regulate the economic activities of capitalist countries have profit-sharing schemes for work- independent enterprises and individuals. In addition to ers, because they believe this to be good for incentives providing for product and worker safety, environmental and morale. But most do not, mainly on the grounds that protection, and so on, legislation should prohibit-as in variations in profitability over time are not much influ- other countries-specified types of monopolistic, anti- enced by worker behavior, as opposed to management competitive, or exploitative behavior. At present in decisions and external conditions. This would be true also China, the situation is unsatisfactorily vague, with enter- of most state enterprises in China, with the added com- prises and local officials free to place their own interpreta- plication that profit-sharing schemes, whether introduced tions on general guidelines from the center, which in by individual enterprises or uniformly prescribed by gov- some cases causes unacceptable abuse, in others (probably ernment regulation, could undermine macroeconomic more numerous) economically valuable activities to be wage control (Chapter 8). In the longer term, if state regarded as illegitimate. enterprises were controlled by boards of directors inter- ested in increasing profitability, administrative control of Motivation of Workers and Managers wage bills might cease to be necessary, and enterprises Appropriate enterprise motivation should include a could be allowed to introduce profit-sharing schemes strong desire to increase profits. Avoidance of losses is an where they felt this to be conducive to efficiency. But for important aspect of this, but by itself is insufficient, since the time being, the disadvantages of linking the remuner- China needs enterprises that not merely pursue a passive ation of workers directly to profits may well outweigh strategy of staying out of trouble, but rather actively seek the advantages. to increase production and sales of existing, improved, Managers-the staff responsible for major decisions on and new products and to cut costs of all kinds. With production, engineering, sales, and other activities- rational prices and competition, the best single measure should be held accountable, much more than the work- of enterprise performance in this regard is usually me- ers, for the enterprise's performance, both in the long and dium-term profits (especially after deduction of the cost- short run. Entrepreneurship is also part of the enterprise depreciation and interest-of the capital employed). The management function, and the criteria against which desire of peasant households to-in effect-increase managers should be judged and rewarded should include profits because their standard of living depends on it lies the enterprise's ability to respond to market demands, to at the heart of China's recent agricultural successes. Out- seek ways of producing goods at lower cost and of higher side agriculture, the same motivation is a natural feature quality, and more generally to innovate and introduce of individual and family enterprises and of small enter- technological changes. A large proportion of managers' prises owned and operated collectively, or as partnerships, incomes, and decisions concerning their promotion or by their workers. In medium-size and large enterprises, dismissal, should thus depend on enterprise profitability. however, establishing the necessary link between profits In China, this will require a broader view of the responsi- and individual rewards is more complicated. bilities of enterprise management and a willingness to A distinction needs to be made between managers and reward superior managers appropriately. workers, especially in large enterprises. Incentives for Obviously price reforms, including appropriate charges workers are mainly a question of wage policy (discussed for use of valuable assets (such as capital, skilled labor, in Chapter 8). There have been many recent attempts to natural resources, and urban land), are needed in order to introduce a production responsibility system, similar to make profitability an economically rational or socially that in agriculture, for individuals and units within enter- defensible basis for managerial remuneration. Moreover, prises. Progress has apparently been limited, partly be- no schemes for linking individual or group pay to enter- cause managers are not yet sufficiently motivated to pay prise performance will work properly without substantial workers by results and partly because many workers-as improvements in financial accounting and auditing. At in other countries-resist this, but also because it takes present in China, there is apparently little discipline in time and experimentation to devise incentive payment accounting, and enterprises are able to ignore state regula- schemes that fit the varying circumstances of individual tions by, for instance, recording bonuses as collective wel- enterprises. It will probably be necessary to allow each fare expenditures. Accurate accounts, subject to thor- enterprise to establish its own payment-by-results system ough, compulsory, and independent audits, with severe for workers, based on qualitative and quantitative evalua- penalties for noncompliance, will be essential if China is tion of performance, in accordance with its particular to establish a workable, efficient, and honest system of conditions but in line with general principles established independent enterprises. by the Government. The extent to which an enterprise's total wage bill Management of Strategic and Key Enterprises (including all bonuses and benefits) should be linked to State enterprises in China cover a vast range of activities, 168 from power stations and railways to neighborhood res- The xroportion of state enterprises designated as strate- taurants, and come in many sizes, from huge industrial gic or hey, and hence subject to direct government regula- complexes such as Anshan Steel employing several hun- tion, slhould perhaps be larger in the early stages of re- dred thousand workers, to small retail stores employing a form, when distorted incentives, bottlenecks, and few persons. It would be impractical to try to manage all uncerta.inties are more serious, and when mandatory pro- these enterprises in the same way. A more realistic ap- ductior planning and materials allocation still play an proach might be to identify various categories of enter- important role. But as other aspects of urban economic prises for which particular principles of state control and reform proceed, such directly regulated enterprises could management would apply. become a small (albeit very significant) minority, with Some enterprises will surely remain under the direct independent enterprises constituting the overwhelming supervision or control of state administrative organs, for bulk of the state sector. instance defense-related industries, as well as basic public services such as electric power and rail transport. Diversi cation of the Ownership System And in sectors where the economically optimal scale of Reform of state enterprise management is the central operation is so large that enterprises naturally became element of urban reform, but by itself is likely to be regional or national monopolies, direct state regulation insufficient. Many types of economic activities cannot be may also be necessary. But even in these strategic enter- efficiently undertaken by state enterprises, however man- prises, managers should be given increased authority, and aged. E'ven in socialist economies dominated by public incentive systems established to reward and penalize ac- ownership, collective and individual enterprises can play cording to agreed performance criteria and evaluation an important and irreplaceable role as discriminating pur- procedures. In this regard, the experience of some East chasers, as efficient suppliers of inputs, and as competitors European countries (most notably East Germany), as well to the state sector. Dynamic nonstate enterprises can thus as some nonsocialist countries, in evaluating managers of both assist and put pressure on state enterprises to im- noncompetitive enterprises may be useful. In France, for prove efficiency and upgrade technology. Perhaps as im- instance, many state-owned enterprises are managed on portant as the recognition of the need to reform and the basis of formal three- to five-year contracts between invigorate state enterprises, therefore, is the position the government and the enterprise, which set out a num- taken in the recent Party decision that "the initiative of ber of performance targets for the enterprise and the the stal:e, the collective and the individual should all be obligations of the government regarding financial support encouraged. We must work to develop diversified eco- and policies affecting enterprise performance. These con- nomic .'orms and various methods of management." tracts also distinguish between normal operational objec- Evidence from many countries, socialist as well as non- tives and any specific social objectives that the enterprise socialisr., illustrates the kinds of nonagricultural activities is expected to meet.7 that are difficult to organize in state enterprises (or even The state may also wish to exercise direct control over large corporations in capitalist economies). They include selected key enterprises, including those exploiting energy personal services (restaurants, repairs), much retailing and and other mineral resources on a large scale, very large wholesaling, small-scale construction and transport, high- enterprises in other sectors, and some enterprises in prior- quality consumer goods, specialized services to industry ity subsectors. Particular machine tool enterprises, for (technical consulting, research and development), and instance, might be selected for state-directed technical small-order parts and components not suitable for assem- transformation in order to compete in international mar- bly line production. In the course of economic develop- kets, or pilot plants in electronics could pioneer the use of ment, moreover, demand for activities provided by small- advanced technology. scale enterprises normally increases just as fast as demand Such selective intervention has been successfully pur- for the products of large-scale industry, partly because of sued even in countries where governments do not nor- increasing specialization in material production, but also mally play a direct role in industrial development. France because services (such as preparing meals, making clothes, and Japan, for instance, have been particularly successful repairs) hitherto organized within the household become in selective interventions in specific industries and enter- parts of the exchange economy, and because there is a prises, especially because they have insisted on increased growing need for business services to lubricate the in- efficiency and technological progress (by contrast, many creasingly complex economy (for example, communica- countries have nationalized declining or inefficient indus- tions, information services, consulting, financial services). tries, but have seldom succeeded in making them eco- nomically viable). Competition, from both domestic and foreign firms, is thus vital for key enterprises. Subsidies may be justified when these enterprises are still in an 7. For other examples, see World Bank, World Development Re- infant stage, but only for a fixed period. port 1983, Chapter 8. 169 Thus, in industrial countries such asJapan and the United from financial extortion by misguided or corrupt officials States, small-scale enterprises account for a large share of at lower levels-a problem common in other countries, national output, partly as subcontractors to large corpora- of which examples are now emerging in China. Regula- tions, and are also important sources of innovation tions that remain in force for many years are also essential (Chapter 7). to persuade nonagricultural entrepreneurs, like farmers, Even when collective and individual enterprises have to take a long-term view and invest in their businesses, been restricted, as in many socialist countries, such activi- rather than aiming at quick profits for immediate con- ties have nonetheless flourished, but have become part of sumption (Chapter 9). a "second economy." An estimate for Yugoslavia for Some state enterprises in small-scale industries and re- 1973-77 was that about 30 percent of services and re- pair and service activities might be contracted to groups pairs were performed illegally. In the Soviet Union during of workers who would then manage them as collectives the 1970s, in addition to large earnings by private hous- (as is currently being considered in China) or simply sold ing repairmen, using supplies and materials of question- to collectives or individuals. The latter option may be able legality, new urban housing construction earned especially relevant for state enterprises suffering persistent black market builders an estimated 1 billion rubles per losses (which would otherwise be closed down), but also year. Because the output and services of the second econ- for highly successful parts of some enterprises, which omy are in great demand, official restrictions have not might be spun off as specialized small-scale enterprises. only been rather ineffectual, but have also had unfortu- Chapter 7 gives the example of a machinery-producing nate side effects. The Soviet Union, Yugoslavia, and workshop within a bicycle enterprise, which in most Hungary have all found that making individual activity other countries would be spun off as a specialized com- illegal results in a tremendous amount of illegality, brib- pany to further develop its capability. Similarly, the re- ery, and corruption, waste of working hours, and theft of search or engineering departments of some enterprises state property. Potential tax revenue is lost, control over might be spun off as consulting companies, which could relative earnings is weakened, and respect for the system then serve more enterprises. of economic management and regulation is undermined. Small collectives share with individual enterprises the Promotion of collective and individual enterprises in major advantage of an extremely close link between the socialist countries requires above all the removal of re- enterprise's financial performance, both short and long strictions on their activities and of discrimination against term, and the personal rewards of its workers-who in their owners and workers. There is usually a vast store of most cases will also be the owners. This advantage is lost initiative that can be tapped once restrictions are re- in large collectives (such as those in urban China, which moved. Despite the recent Party decision, however, it are managed much as state enterprises) and more gener- will take time for the population generally to view em- ally in all collectives whose workers do not share heavily ployment in nonstate enterprises as comparable to state in their ownership. An interesting exception to this gen- employment, because of the long history of discrimina- eralization has been China's commune and brigade enter- tion against small collectives and individual activities, and prises. Although these enterprises are owned by the en- people will need to be convinced that the new policy will tire township or village and not by their managers or remain in force. workers, the latter all come from the township, whose An important step would be to provide equal access to population is highly immobile. There is therefore a much social services and welfare benefits for workers and entre- greater feeling of solidarity and a stronger incentive to preneurs in the nonstate sector. They should also be develop and expand these enterprises and, most impor- given access to credit from banks and other financial tant, to create employment opportunities for relatives still institutions. Large collectives may continue to be taxed working in agriculture. Experience in other countries, on the same basis as state enterprises, but taxation of including France, Spain, and the United States, confirms smaller collectives and individual enterprises should be that enterprises owned by communities work particularly simple to minimize bookkeeping requirements (Chapter well when the community is stable and well integrated, 9). Access to materials and premises is also crucial: collec- especially in remote areas. tive and individual enterprises are seriously handicapped But there are few examples in the world of success with by the present allocation system and would benefit from genuine large and medium-scale collectives (where the its relaxation or replacement by market regulation, espe- workers are the owners), especially in manufacturing. cially for key materials. Most successful collectives are small or in nonindustrial Most important will be a set of simple, unambiguous, lines of activity, especially commerce and distribution and stable regulations on the establishment and operation (where they are often consumer or producer, rather than of these types of enterprises. Such regulations, backed up worker, cooperatives). The problem seems to be in re- by legal arrangements for enforcement and appeal, taining a strong sense of ownership beyond a certain size; would, among other things, protect small enterprises the few successful examples of larger-scale industrial col- 170 lectives (for example, Mondragon in Spain) have devel- Investment Decisions and Financing oped elaborate procedures regarding contributions and In infrastructure, education, health, and defense, the withdrawals of capital. The advantages and disadvantages Government will want to retain direct control over the of the different forms of collective enterprises will thus bulk of investment decisions. Similarly, in agriculture, it require continuous review in the course of reform. will want to control investment in infrastructure such as Incentives to increase efficiency and innovate are obvi- irrigation and land development, though most other in- ously strong in enterprises owned and managed by indi- vestment will probably be undertaken by households and viduals or families. The issues in China are the maximum collectives. In industry, however, an issue that needs to be acceptable scale of these enterprises (and related questions considered is the extent to which investment decisions such as the number of workers they can employ) and the should se made at the enterprise level. acceptable share of these enterprises in overall economic An argument for investment decisions by enterprises is activity. To prohibit the growth of individual enterprises that a degree of responsibility over the future of the enter- beyond a certain size, or to restrict their numbers, would prise, in addition to responsibility for current production stifle a potentially dynamic force within the economy. An decisions, would give enterprise managers more meaning- alternative might be to require individually owned enter- ful responsibility for the enterprise. Enterprise perfor- prises to sell majority ownership to the state, or other mance depends, after all, not only on current manage- socialist institutions (such as state enterprises or banks) ment decisions, but also on investment decisions made in once they exceed a certain size. The minority owners the past. Enterprises, and their managers, cannot be held could continue managing the enterprises, perhaps super- accountable for current performance if investment deci- vised by a board of directors representing all owners. For sions were made by state administrative organs. this approach to work well, without discouraging promis- Enterprise responsibility for investment decisions is also ing small enterprises from expanding, the sale of owner- often necessary to improve the efficiency of investment. ship rights to the state would of course have to be at a fair All invcstment decisions involve risk, and a certain pro- price-diminishing the control of the individual, but not portion of failures-and unexpected successes-is inevita- confiscating his wealth. In other countries, many success- ble with any decisionmaking system. But, as with produc- ful small businesses are voluntarily sold after a few years tion, managers of enterprises should have a much better to larger firms. understanding than state administrative organs of the Socialism with Chinese characteristics is envisaged by benefits and costs of available investment opportunities. the Party to include "a cooperative relationship between This advantage of decentralized investment decisionmak- state, collective and individual enterprises." The next two ing is erhanced during intensive growth (Chapters I and decades-and beyond-will be a period of continuous 7), which requires continuous reassessment of the existing and experimental search for the mix of ownership and structur, of production, as well as a continuous search for management systems that would best serve China's eco- ways of lowering costs and improving quality. nomic, social, and political objectives. What seems cer- Moreover, higher-level administrative officials, how- tain is that there will be a variety of types of enterprises, ever intelligent and conscientious, have no financial stake each suited to certain sorts of activities. Diversified own- in the success or failure of the projects they are deciding ership arrangements are already emerging even within the on. By contrast, decisionmakers at the enterprise level- state sector, with enterprises being established by other the owrers in small businesses, managers appointed and enterprises, with provincial and local governments coop- supervise d by boards representing the owners in larger erating to found new enterprises, and with joint ventures concerns-have or can be given a substantial direct per- between state and collective enterprises and between Chi- sonal financial interest in the enterprise. Experience- nese state enterprises and foreign private enterprises. positive and negative-both in China and elsewhere sug- gests that personal financial involvement in the outcome Investment and Price Reform of an investment decision can be a very effective way of encouraging bold and innovative thinking combined with A central element of system reform is to expand the role thoroug7l analysis, hard-headed calculation, and the of market regulation. An important component of this avoidance of waste. This is true not only of the expansion effort is, of course, the reform of enterprise management or renovation of existing enterprises, but also of the estab- discussed in the preceding section. Other components, lishment of new enterprises. including changes in the labor allocation, wage, and tax Finally, delegating more responsibility for investment systems, stronger incentives for technological change and decision to independent enterprises could help in striking dissemination, and reform of internal and external trade, a better overall balance between infrastructural and other have been discussed in earlier chapters. This section dis- investmcnt. Economic efficiency in China, as in other cusses two additional issues-investment decisions and socialist countries, has suffered from a "medium-sighted" financing, and price reform. bias in favor of directly productive industrial investment, 171 which in practice still remains strong, especially among resources flow to the uses offering the highest economic local governments. This bias might be reduced if govern- returns. In Yugoslavia and Hungary, not only is the flow ment at all levels had less of a role in organizing and of investment funds among enterprises limited, but enter- financing such investment and could thus concentrate prises are constrained in their ability to enter new lines of more on infrastructural investment. In particular, making activity. As a consequence, enterprises have often made enterprises independent and giving them more invest- investment decisions that might be sensible from their ment responsibility could oblige local governments-as in own perspective, but inefficient from the national per- other countries-to improve infrastructure in order to spective. Since each enterprise's own resources are lim- attract industrial investment. ited, there has also been a bias toward projects below the Many socialist countries, including China in recent economically optimal scale. Delegation of investment de- years, have permitted state enterprises to retain a part of cisions to enterprises thus has to proceed in parallel with their profits for specified categories of investment. This reforms of the price system and other aspects of enter- has had some good results, but has also encountered prise management, and with increased mobility of invest- serious problems. In China, irrationality of prices, short- ment funds. comings of the material supply system, and protection of local industries have led enterprises and local govern- ALFERNAFIVE CHANNELS OF INVESTMEN-T FLOWS. In ments to make some investment decisions that are highly the traditional socialist system, mobility of savings is undesirable from the national perspective. There has largely vertical. Savings are mobilized by the state budget been overinvestment in sectors with high administered through profit remittance, taxation, and other fiscal in- prices, underinvestment in vital but unprofitable sectors struments and then allocated according to the plan either such as energy and transport, and continued proliferation through the budget as grants or through state-run banks of uneconomically small and low-quality projects aimed as credits. In addition, there is some compartmentalized at increasing local self-sufficiency in materials or equip- reinvestment of savings generated within particular sec- ment (Chapter 5). tors and localities. In a reformed socialist economic sys- Another problem has been inappropriate motivation of tem, the relative importance of different institutional enterprises and their managers (discussed in the previous sources of saving could be quite different (Chapter 9), as section). State enterprises have much to gain from expan- could be the ways in which saving is channeled into sion of fixed assets and production-increased bonuses investment. Vertical flows and compartmentalized rein- for staff and workers, more resources for collective wel- vestment, although they would remain important, should fare, and the enhanced prestige and power that come be increasingly supplemented and replaced with horizon- with larger size. The negative consequences of misguided tal flows. investments are also much attenuated by subsidies and First, enterprises could be allowed and encouraged, as other forms of assistance from state organs. This leads to mentioned earlier, to enter new kinds of activities, financ- "investment hunger," as in other socialist countries, with ing such new activities with credits or other external funds enterprises competing for investment resources even as well as internal resources (more profitable enterprises when the potential economic returns are low or negative. tend to have better access to both internal and external The resultant tension between central planners seeking to finance). This would not only create flows of investment control the level of investment and the hungry enterprises funds (as well as other resources such as management skills, frequently results in sharp investment cycles. Even in labor, and capital equipment) to new activities or sectors China, where the authority of enterprises over invest- where they could be more usefully employed (even though ment has been very restricted, this tension nonetheless they remained within the enterprise), but would also exists between central and local authorities (including the greatly increase competition in the economy. rural collective sector), with similar investment cycles. Second, horizontal flows of investment resources could In Hungary and Yugoslavia, which have experimented be achieved through direct investment between economic more than any other socialist countries with decentral- units. Enterprises could be permitted to invest in other ized investment decisionmaking, this has not improved enterprises, establish new enterprises, or participate in investment efficiency also because of the limited mobility joint ventures with other enterprises, local governments, of investment funds. Highly profitable enterprises tend to collectives, or even foreign companies. Such horizontal be efficient and dynamic and should therefore be allowed flows and cooperation are already emerging in China, in more investment resources-but not necessarily for in- many different forms, mostly from lower-level initiatives, vestment in their existing activities. They should also but with the Government's encouragement and support. consider investment opportunities elsewhere in the econ- (Ensuring the supply of key materials has apparently been omy, evaluating these against opportunities for either a major motive-for instance, investment in Shanxi coal new activities or expansion of existing activities within mines by industrial enterprises in energy-deficient their present fields. Only in this way could investment Jiangsu.) 172 These forms of direct investment between economic administrative. In addition to financial support, the finan- units should be encouraged and, with further reforms in cial int=rmediaries could also-as in other countries- other aspects of the economic system, will be increasingly provide technical assistance in financial management and guided by economic criteria. But these horizontal flows information in areas such as technology and marketing. can offer only a limited channel for investment resource Indeed, the activities of such institutions as highly moti- mobility. They are comparable to bilateral trade (barter- vated gatherers and disseminators of economically rele- ing) for commodities, since they involve the direct ex- vant information would be just as important as the finan- change of resources between suppliers and potential us- cial flowas themselves. ers. For more efficient and possibly larger flows of Diverse organizational forms and a multiplicity of over- investment resources, horizontal flows through financial lapping and competing institutions would be highly desir- institutions functioning as intermediaries between the able (see Box 10.1 on variegated social ownership in West suppliers and users of resources will also be necessary. German banking). If only a single financial institution One function of financial institutions would be to mo- existed, or several that had mutually exclusive areas of bilize investment funds from individuals, collectives, state responsibility, they would tend to become monopolistic, enterprises, and local governments, by accepting deposits bureaucratic instruments of particular levels (or agencies) and by issuing financial instruments such as bonds (either of government. In China, which has begun to diversify its by themselves or on behalf of the users of capital). They financia.. system since 1979, there could be many differ- would have to offer an interest rate (or other return) that ent formis of financial institutions-banks, credit coopera- reflected the scarcity of investment funds in the economy. tives, investment trusts, insurance and pension funds, and In this way, enterprises with internal funds available so on. In rural areas, for instance, some of the economic would be made aware of the potential returns to invest- institutions of the former communes and brigades, which ment opportunities elsewhere in the economy and would still exist following the reorganization of local govern- be able to choose rationally between direct investment ment administration, could be transformed into financial (internal or in another enterprise) and indirect investment institutions. Such township banks or investment compa- (by depositing the funds with financial institutions). nies could mobilize investment resources from collective The funds raised by these institutions would be made enterprises to establish new enterprises, help existing en- available to potential investors, particularly enterprises of terprises expand and modernize, and support agricultural all kinds. The funds could be provided as loans, with development. However, the operations of financial insti- fixed rates of interest and repayment periods. At least tutions should not be geographically restricted. The expe- some financial institutions, however, should be able to rience ol Yugoslavia confirms that regional banks tend to provide ownership capital, partly because many worth- impede rather than facilitate flows of investment re- while projects are too risky to be financed largely by sources between regions and thus contribute to regional loans, partly because directly sharing in the risks and autarky. benefits would motivate financial institutions to give State intervention in the management of financial insti- more assistance in project design and implementation. In tutions, and specifically in their allocation of funds, is a this case, the financial institution would be an investor more cornplex issue. There is obviously a need for state itself, much as local governments, bureaus, state enter- regulation. A major function of financial institutions is to prises, and collectives now participate in joint investment reduce risk for both savers and borrowers and thus in- projects. crease incentives to save and invest efficiently. Govern- These institutions would thus be similar to commercial ment regulations and intervention are required largely to banks, investment trusts, development finance compa- reduce the risk to savers, through deposit insurance nies, and other intermediaries in capitalist countries, but schemes, restrictions on the scope of various institutions' with the difference that they would be dealing with flows operations and the types of assets held, requirements on of funds that were to a large extent socially owned-by full disclosure of information, and monitoring the perfor- government organs, state enterprises and other state insti- mance arid viability of these institutions. Since the activi- tutions, or collectives. They would, however, also be ties of the financial sector may have a substantial impact fundamentally different from existing banks in socialist on the overall level of economic activity and prices, the countries, which primarily implement planned vertical Government clearly also has to intervene in financial flows. These institutions would in effect create a socialist markets in order to regulate the overall supply of money market for investment funds. and credit (see Box 10.2). All this would require broaden- Whether capital is provided as loans or equity, the ing and s:rengthening the functions of the People's Bank, relationship between the financial institution and the en- which is :o become a specialized central bank. terprise that receives the capital (or that was founded by The appropriate degree of government intervention in the financial institution, either alone or in cooperation credit allocation and interest rates is more controversial. with other institutions) should be economic, rather than Excessive intervention obviously risks turning financial 173 Box 10.2 Monetary Control Reform of China's system of economic management will ments of monetary control. greatly increase the importance of monetary control. In other countries, one pillar of monetary control is the When most goods are administratively allocated, the avail- imposition of mimimum reserve ratios on individual banks, ability of money and credit to pay for them is of secondary which require them to keep at least a certain proportion of importance, and monetary policy plays a passive, accom- their assets in the form of cash and deposits with the modating role. Moreover, when most enterprises are ad- central bank. This puts an upper limit on the "money ministrative organs, whose economic survival is guaran- multiplier"-the relationship between the stock of "high- teed, it is futile to attempt to restrain their spending powered money" (cash plus deposits at the central bank) through monetary anda credit restrictions. In Eastirn Eu- and the total stock of money, including all bank deposits, rope, for example,, state enterprises respond, to a tight which is several times larger. The second pillar of mone- credit policy by refusing to pay their bills on time, generat- tary control is central bank regulation of the stock of high- ing involuntary trade credit that cascades throughout the powered money, most cormmonly through purchases and economy as firms settle accounts outside the banking sys- sales of government bonds ("open-market operations") tem and formal or informal separate clearing mechanisms and varying the terms of central bank lending to commer- emerge. This creates great difficulties for the banking sys- cial banks (the "discount window"). tem, and sooner or later the authorities feel compelled to Monetary control of a market-regulated economy is by return to a relatively easy credit policy, no means easy. Opinions differ as to whether the main By contrast, in a market-regulated economy, with finan- object of control should be money (variously defined), cially independent enterprises, the supply of money and credit, or interest rates. The size of the money multiplier credit crucially affects the overall level of (monetary or may fluctuate unpredictably, for example, if banks keep nominal) demand for goods and services, and hence, espe- more than the required amount of reserves. Changes in the cially in the medium term, the general price level. For this money supply likewise do not cause precisely predictable reason, and particularly since the inflationary surge of the changes in the overall level of demand for goods and ser- 1970s, the governments of market-regulated economies vices, which in turn do not have precisely predictable ef- have attached increasing importance to monetary control. fects on the price level (or the real volume of economic In China, both the need to6control money and credit and activity). None of this vitiates either the need for, or the the existence of a connection between fiscal policy and feasibility of, strict medium-term monetary control. But monetary policy are well understood. But a separate cen- the short-term causes and consequences of changes in tral bank was established only in 1984, and recent experi- monetary magnitudes may be too complex and uncertain ence of unexpectedly rapid credit expansion suggests that to permit exclusive reliance on any simple or automatic there may be scope for improved techniques and instru- formula for monetary control. institutions into state organs, impeding improvements in trol was exerted indirectly, by offering low interest rates the efficiency of investment, without necessarily increas- and favorable repayment terms to particular activities of ing effective central control. In Yugoslavia, after the 1965 private enterprises. Unlike those in Yugoslavia, these en- reform, for example, the government kept the interest terprises were highly motivated to increase profits over rate (after allowing for inflation) at a level much below the medium term and were made conscious of the true that necessary to balance the supply of and demand for economic value of their subsidized credits by informal investment funds, and hence the investment hunger of parallel financial markets, which among other things "re- enterprises and local governments persisted, requiring fre- cycled" some of the subsidized funds at much higher quent administrative intervention by state organs in the interest rates. Moreover, as their economies have devel- financial system and causing continued cyclical invest- oped further, both Japan and South Korea have given ment fluctuations. greater independence to their financial systems. During periods of rapid growth in Japan and South Korea, in contrast, their governments played a major role EVOLUTIONARY CHOICE. For China, continued ex- in the allocation of credit to specific sectors and enter- perimentation and exploration in the area of investment prises. Since in both countries' credit accounts for an decisionmaking and financing will be necessary, especially unusually large proportion of enterprise investment because the experience of other countries provides no funds, this gave their governments a lot of control over precedent for a socialist financial market. The exact the allocation of investment, although much of this con- shares and relative importance of alternative forms of 174 investment allocation and financing should be allowed to selves constitute obstacles to price reform-including lack evolve with the lessons of experience and the develop- of competition, inadequate motivation to hold down ment of institutions. costs, and investment hunger, all of which contribute to Savings mobilized through the state budget would at a chronic shortages of many commodities. minimum continue to finance large amounts of invest- One ingredient of reform should thus be to minimize ment in infrastructure, energy, and the social sectors. In the nonallocative role of prices, through greater reliance addition, part of budgetary savings could be channeled on taxes, transfer payments, and other instruments dis- through financial institutions to enterprises to finance cussec[ earlier to achieve social and fiscal objectives. Prices other investment. The government would also indirectly could then be used mainly to signal relative costs, needs, regulate the level and content of the growing share of and scarcities to enterprises and households, guiding their total investment undertaken directly by enterprises and production, expenditure, and investment decisions in households. economically appropriate directions. To accomplish this, Financial intermediaries would mobilize an increasing as the Government recognizes, the price-setting system share of savings, tapping the potential for household sav- must itself be changed to give market supply and demand ing, as well as funds from enterprises with limited internal forces a greater role. Administered changes in prices set investment opportunities. These institutions would also by the state are also needed, especially in the near term. play an increasing role in investment decisionmaking, Several aspects of price reform were discussed in earlier through their evaluation of investment proposals by bor- chapters, including more flexible pricing of agricultural rowing enterprises and through their own direct invest- products (Chapter 3), altering the prices of various ment. sources of energy (Chapter 4), restoring the link between Enterprises would not only make an increasing propor- producer and retail prices of essential consumer goods tion of investment decisions, but would also generate an (Chap;-er 9), adjustment of prices that influence the spa- increasing share of savings from their retained profits, tial location of economic activities, especially transport supplemented (largely at their discretion) by loans or eq- tariffs and land use charges (Chapter 5), and stronger uity capital from other institutions. The speed with connections between domestic and world prices (Chapter which enterprise-level investment should increase in im- 6). This section concentrates on some general issues relat- portance depends on the speed with which it is possible ing to adjustment of energy and raw material prices and to establish, first, linkages between the personal interests to the reduction of administrative price control. of managers and the outcome of investment decisions, and, second, economically rational prices, interest rates, ADJUSTMENTS OF ENERGY AND RAW MAIERIALI and other signals that reflect relative costs in the econ- PRICES. The most obviously distorted of alL China's pro- omy. ducer prices are the state-set prices of energy and some While many of the vertical and horizontal flows of raw materials-far below opportunity costs' in domestic investment funds would be in the form of credits, some and world markets. But the sheer magnitude of these would be in the form of equity capital. The latter would distortions makes them hard to correct: to introduce the include flows between enterprises (both state and non- required price adjustments at one stroke would involve state), between financial intermediaries and enterprises, dramatic changes in the financial circumstances of many and from households to enterprises, either directly or enterprises and institutions; yet to make them in small through financial intermediaries such as pension funds. steps would undesirably prolong the period of adjust- Diversified flows of funds would therefore be both conse- ment, particularly because these key prices directly and quence and cause of a diversified pattern of enterprise indirectly influence almost all other prices, which are ownership. themselves in need of other sorts of adjustments. Whether in one administrative step, or in a number of Price Reform announced steps, or through rapid transformation of the It is recognized in China that "rational prices constitute present two-tier pricing system into a unified market pric- an important condition for ensuring a dynamic and stable ing system, major adjustments of energy and raw material economy, and reform of the price system is the critical prices seem urgently needed. Postponement of these ad- element of the entire program of system reform." This is justments over the past few years has led to the extension because the decentralized decisions of independent, of ad hoc state interventions in other areas, including profit-oriented enterprises would otherwise often be inef- ficient for the whole economy. Yet price reform is com- plex and difficult, partly because prices have up to now 8. Opportunity cost is the value of other economic possibilities been used largely for other purposes, including income created or forgone by producing or consuming a specific commod- redistribution and government revenue raising. More- ity. For goods that can be traded, consideration of opportunity over, deficiencies in other aspects of the system them- costs shculd include markets abroad as well as in China. 175 enterprise-specific subsidies and "adjustment" taxes and differing energy intensities and to eliminate other current retroactive and frequent changes in tax rates. Hungarian irrationalities in the relation of prices to production cost experience confirms that interventions of this kind tend and product quality). Enterprises operating in more com- to negate the benefits (and slow the futher progress) of petitive areas could be given greater flexibility in varying socialist system reform. their output prices, since competition and market pres- Possible adverse effects on state revenues are one of the sure would work against excessive price increases. concerns that have inhibited the government from in- creasing energy and raw material prices. In principle, of MARKET PRICING. Despite their potential usefulness course, since state enterprises constitute the great bulk of in correcting accumulated distortions, there is increasing both producers and users of these commodities, price recognition in China that administrative price adjust- adjustments should mainly simply redistribute an un- ments-which tend to lack the flexibility, complexity, and changed aggregate amount of profit and tax among state precision needed in a modern economy-are only part of enterprises. Even so, this could alter budget revenues price reform. More critical is to give market demand and unless there were simultaneous adjustment of the varying supply a greater role in price determination. There has rates of profit tax on individual enterprises. Moreover, been progress in this direction over the past few years: given China's weak accounting and auditing, enterprises the prices of many minor items, and of some transactions whose profits increased as a result of these price adjust- in more major items, are already determined by market ments might report less of a rise than they should, while forces. Major progress has, however, been impeded by others might exaggerate their additional costs. (Following the fear that, because there are chronic shortages of many recent increases in coal prices, some coal mines even goods, price decontrol might lead to a sudden accelera- claimed that their profits went down because of unau- tion of inflation. thorized increases in input prices!) Such leakages would In a properly functioning economy, excess demand in make enterprises better off at the expense of the state: if one sector would imply excess supply in others, and such there were tight control of worker remuneration and imbalances would be corrected through relative price benefits, this would mainly increase the proportion of changes induced by market forces. Generalized excess investment funds in the hands of enterprises-a conse- demand and price inflation would in principle arise only quence that would be less worrisome once prices more if there were macroeconomic mismanagement-for ex- accurately reflected relative costs. ample, budget deficits financed by money creation or The possible ripple effects of large energy and raw excessive expansion of bank credit. The solution to such material price increases on the prices of other goods are problems would thus lie not in price control, but in more another source of concern in China. But many of these restrictive macroeconomic policies. ripple effects should be welcome: higher energy prices, Experience in China and elsewhere strongly suggests, for example, cannot and should not be fully absorbed in however, that there are more deep-rooted causes of gen- energy-using enterprises, but should be partly passed on eralized excess demand in most socialist economies. in the form of higher prices for energy-intensive products, These include the inefficiency and rigidity of centralized whose use needs to be discouraged in order to improve material allocation and production planning, the lack of the overall energy efficiency of the economy. Moreover, concern among enterprise managers about cost and prof- increases in the prices of energy and energy-intensive itability, the absence of competition, and the "investment products 'would not necessarily raise the general price hunger" of enterprises and local governments (discussed level, since they would soak up purchasing power and later). The "seller's markets" that prevail for most com- hence tend to reduce the demand for, and prices of, other modities in such economies thus cannot be eliminated products. simply through tight fiscal and monetary policies. The state might nonetheless wish to manage the ripple The fact that chronic shortage appears to be a universal effects caused by major adjustments of energy and raw and permanent feature of the traditional system of social- material prices. For "strategic" and "key" enterprises ist economic management suggests, however, that reform (discussed earlier), a unified program for the adjustment itself would create the necessary preconditions for price of product prices, costs, and taxes could be worked out in decontrol. This view is supported by China's experience some detail by controlling administrative organs, includ- in the past few years, especially in agriculture, where ing the Price Bureau and the Ministry of Finance. In relaxation of direct controls has turned long-standing other areas, where competition is still minimal, enter- shortages into abundance. The same could happen in prises might be required to pass on only a specified pro- other sectors with the implementation of reforms in en- portion (larger in sectors with lower profitability, and vice terprise management, commerce, competition, and so versa) of their increased costs in higher average product on. Increases in specific prices could eliminate shortages prices. They could be allowed greater latitude in varying by stimulating supply and reducing demand. These price the relative prices of individual products (both to reflect increases, moreover, would tend to reduce purchasing 176 power over (and hence the prices) of other goods, pro- cause of the costs it might inflict on certain-especially vided that the Government did not give compensating poorer-groups. There is also fear that initial moderate tax reductions or subsidies and more generally kept strict price increases might be greatly magnified by panic buy- control of the budget balance and the availability of ing, especially because household bank deposits are now money and credit. very large (at the end of 1984 they equaled 36 percent of Implementation of the other reforms necessary to elim- the year's retail sales). inate chronic shortages will take time, however, as will The experience of other countries suggests that these some of the required demand and supply responses in apprehensions are not without foundation, but can easily nonagricultural sectors. Smooth adjustment could also be be exaggerated. Most increases in the prices of particular impeded by downward inflexibility of some prices and of consuner goods will be offset by decreases in the prices wages. In addition, China lacks experience of indirect of other consumer goods, provided that fiscal and mone- macroeconomic management through fiscal and mone- tary policies are prudent. Even if it is not possible to keep tary policy, which would become much more important the consumer price level absolutely stable, most people in a reformed economic system (see Box 10.2). Even are apparently adaptable enough to live easily with mod- countries with much greater experience in this regard still erate inflation, especially if their incomes are increasing suffer from rising prices (although rapid inflation has al- steadily in real terms. And although general indexation of most invariably been the result of political irresponsibil- incomrs is not to be recommended (Chapter 9), the living ity, especially unwillingness to restrain consumption or standards of some vulnerable groups can be protected by investment to the limits set by production and external indexation, as well as by allowing and encouraging greater borrowing capacity). For these reasons, price decontrol household ownership of real assets such as housing. probably has to be gradual though steady. Indeed, an early major increase in the extent of home owner,hip in urban areas could serve several related ob- PRICE DECONTFROL OF PRODUCER GOODS. The Gov- jectives. It could be accomplished by offering most state- ernment's strategy is to reduce administrative control of or enterprise-owned residences for sale to their occu- the prices of materials, machinery, and equipment in par- pants, while at the same time announcing a schedule for allel with the gradual dismantling of annual production substantial increases in their presently hugely subsidized planning and allocation. This has already happened to a rents. Though these rent increases should be compen- significant degree: the number of goods subject to pro- sated by wage and other income increases (Chapter 9), duction planning and allocation has declined. And for they would nonetheless give people an incentive to make many goods, only a part of total output remains allo- use of their savings to purchase their homes, probably at cated, at a fixed price, to meet requirements for high- a somewhat concessional price. This would lead to better priority production. Enterprises acquire their remaining maintenance of housing. It would provide many people requirements at the prevailing market price. This two-tier with an asset that in other countries has proved an excel- pricing arrangement has substantial advantages over the lent hedge against inflation. And it would enable the state former system, but also gives rise to certain problems, to absorb a large fraction of urban household bank de- including inequities among using enterprises and the com- posits. (Toward this last objective, especially in rural areas plexities involved in determining and regularly adjusting where nost housing is already privately owned, the Gov- the state allocations. It should thus probably be only a ernment might also encourage lump-sum contributions to transitional means of reforming the price and material pension funds-as mentioned earlier and in Chapter 8). allocation systems. One option would be to move the administered price of each commodity, say, halfway to- ward the market price in each time period (the length of E the period being determined by progress in other aspects An implication of the preceding discussion is that the of reform mentioned above), so that the two prices even- state should reduce its direct involvement in many eco- tually converge. Another option-which could be com- nomic areas, including production, commerce, prices, bined with the first-would be to gradually reduce the and employment. This would release resources, financial share of output subject to allocation and administered and human, for much needed increases in state activity in prices until all transacions were at market prices, other areas of economic and social importance. More- over, it will continue to be the primary responsibility of DECONTROL OF CONSUMER PRICES. Though the the state to guide the country's overall development. This prices of many minor items now float freely, the Govern- will, hcwever, require a broader and fundamentally dif- ment is especially hesitant about decontrolling the prices ferent cDncept of planning than in the past. of other-essential or major-consumer goods. The im- mense inflation before Liberation has not been forgotten, New ApDroach to Planning and even slow inflation is regarded as unacceptable be- Direct (ontrol over economic activities has been the sin- 177 gle most important feature of China's past management controls will lead to more uncertainty. The number of system. Three main instruments may be distinguished: economic decisionmakers is, after all, greatly increased, physical planning of production, centralized allocation of and their objectives are not mutually consistent, nor nec- materials, and budgetary grants for fixed investment. To- essarily in harmony with the Government's objectives. gether, these instruments constituted a system for manag- Much will thus depend on the clarity and stability of ing the economy mainly by controlling supply. What was medium-term policies, which constitute the core of the not controlled was not planned in any meaningful way. development plan; the better these are enunciated, the The central plan for coal production and allocation, for less the uncertainty surrounding decisionmakers or other instance, did not include output from coal mines con- economic agents, and the less the chances of their deci- trolled by provincial authorities; provincial plans for coal sions having a destabilizing effect. Even so, indirect con- did not include output from mines belonging to rural trol involves a relationship between instruments and ob- communes and brigades. Similarly, provincial investment jectives that can never be precise, since it depends on the plans did not take account of investment projects fi- responses of a multitude of individual decisionmakers, nanced either by the central government or by collective including farmers, enterprises, collectives, and consumers. enterprises. In fact, often the absence of control implied It is impossible to predict their behavior with complete the absence not only of planning, but also of knowledge: accuracy, and therefore the instruments used will never provincial authorities had little knowledge of where out- be exactly what is needed to achieve particular objectives; put from centrally controlled enterprises was allocated or the residual is a source of uncertainty that defies policy of the overall level and content of investment activities in intervention. This shortcoming of indirect control may the province. It is this notion that direct control equals be no greater than some of the shortcomings of the tradi- planning that causes some in China to fear that a reduc- tional system of direct control. But especially in the early tion of direct control must mean the erosion of planning. stages of reform, when the coordinated use of indirect Planning in a meaningful sense must extend beyond instruments of control is still a new concept, planning and activities under direct state control, and indeed in a re- policy will need a high degree of flexibility and respon- formed system a large part of the planning effort will siveness. involve measures of an indirect nature. Reforms that have Of major importance to the new concept of planning is already been introduced, as well as possible further the newly established view in China that planning and changes discussed in earlier sections, will mean that markets can coexist and develop harmoniously. Although farmers and enterprises have greater latitude to make the potential for clashes between direct and indirect con- their own supply decisions and to use various market trols should not be forgotten, and although the govern- channels for their sales. Central control over investment ments of other countries differ widely in their attitudes allocation could also be reduced. There will thus be a far toward markets and planning, international experience greater number of decisionmakers, each with some con- tends to support the present Chinese view. Planning has trol over a part of national resources. Planning will ac- been used to a differing extent between countries and cordingly and increasingly have to involve determination over time, and it is not easy to generalize about its effi- of the economic environment within which relatively cacy, especially since actual developments cannot be di- independent economic agents-farmers, enterprises, fi- rectly attributed to the extent or quality of planning. But nancial institutions-operate. where market mechanisms were neglected or not taken This raises the question of whether greater reliance on sufficiently into account, planning was usually less effec- indirect controls would reduce the Government's abilitv tive and national priorities became distorted, often with to manage the economy. In this regard, it seems impor- inefficient use of scarce resources. Conversely, there is tant to recognize that-as discussed above-direct con- considerable evidence that planning can be strengthened trols are inevitably limited in scope, while indirect con- if it is exercised in a way that allows market forces to trols generally influence the whole economy. An reinforce rather than oppose it. additional drawback of direct, as compared with indirect, Mutually consistent roles for plan and market imply control is that it must be shared among vertically distin- that planning must be based on managing demand as well guished sectors, or horizontally divided localities, or some as supply. This is essentially because it is the growing level compromise between the two, which leads-as is well and evolving structure of demand, reflecting the rising known in China-to autarkic tendencies and segmenta- prosperity of the society, that should mainly drive the tion of the economic management effort. On balance, changing composition of supply. In a large part of the then, indirect controls appear potentially at least as effec- economy, planning is thus mainly a matter of the state's tive as direct controls in economic management and al- managing and guiding changing demand, with supply re- most certainly more efficient. sponding through the decentralized decisions of peasant A second and closely related question concerns the households and enterprises. Mandatory planning of sup- extent to which a different mix of direct and indirect ply, involving directives from planners or other state ad- 178 ministrative organs, will still be essential in developing limited reliance on direct controls. It attempts to use infrastructure and in regulating part or all of the produc- available instruments in ways that influence the decisions tion and use of certain commodities. For a large and of market participants, bringing these into line with the growing proportion of economic activities, however, plan objectives of medium- and long-term development. Eco- objectives should be realized through guidance, including nomic policies need to form a coherent and internally manipulation of taxes and credits to influence the relative consistent package rather than to be formulated individu- profitability of different sorts of production and invest- ally. The translation of development objectives into a set ment. of polizies that directly and indirectly guide the develop- ment process thus requires considerable coordination ANNUAL PLANNING. The nature of annual planning within the government-more than now exists in China. in the future depends on what happens to the material Governments are not monolithic, and the interpretation allocation system. Without centralized material alloca- of general development objectives in terms of responsibil- tion, annual physical planning of production loses much ities and actions may differ among ministries and be- of its meaning. China has already reduced, and is contin- tween central and lower-level government organizations. uing to reduce, the scope of mandatory production tar- While it is essential for each agency to look at its own gets and material allocation. The question is now respon;ibilities in the wider context of national develop- whether they could be abandoned altogether in the near ment, it is equally essential for one government agency to future, weighing the possible benefits of retaining the have the analytical capability to develop comprehensive present system agaist its costs, including the incentives it policy scenarios and to present these as the basis for creates for local governments at all levels, as well as enter- policy decisions by all the agencies concerned. prises, to invest for self-sufficiency and assured supply. Policies cannot be expected to achieve the desired im- The experience of Hungary, which abandoned directive pact immediately, and policy planning must have a me- production planning in a single step in 1968, strongly dium-term (three- to five-year) horizon. Moreover, any suggests that such a change would not disrupt or destabi- set of policies agreed for the medium term will require lize the economy-though, of course, Hungary is a far phased implementation and may need to be adjusted in smaller country than China. But it would cause annual light of changing economic circumstances and assess- plans to become mainly the step-wise implementation of ments Df effectiveness. Examples abound in China and the medium-term plan. This should be closely linked to abroad of economic policies that were appropriate and the budget cycle. effective for several years, but lost their usefulness as the Annual planning and budgeting in a mainly indirectly economic structure gradually shifted. These same policies controlled economy also involves regulating the overall have frequently ended up as stumbling blocks for policy pace and stability of growth through fiscal, monetary, reform Particularly in countries expecting rapid struc- and exchange rate policies (rather than-as in the past- tural and technological change, as in China over the next through direct intervention in production, investment, few decades, an argument can be made for the promulga- prices, and foreign trade). Among other things, mainte- tion of policies for strictly limited periods, at the end of nance of short-term macroeconomic stability involves the which they expire unless explicitly renewed. manipulation of taxes, government expenditures, money In preceding chapters, the need for policy planning has and credit (see Box 10.2), and the exchange rate to offset been discussed in several contexts: guiding price trends temporary fluctuations in economic activity of domestic and ensuring price stability for agricultural products; lead- or external origin. ing the technological transformation of industry; correct- The experience of other countries shows that such indi- ing dis:orted energy prices; decreasing the rural-urban rect macroeconomic management is not easy, although income gap; changing and extending the social security China is fortunate in being a large and diverse economy, system; assisting poor and backward localities; and using with a relatively small foreign trade sector. But experi- foreign trade to promote economic efficiency. These are ence elsewhere has also provided much useful informa- issues chat cannot be dealt with through short-term tion concerning the coordinated use of fiscal, monetary, actions or with a single policy instrument. Also, the ef- and related policies, which have been perhaps the main fects oif thc necessary policics arc pcrvasivc and their subject of theoretical and empirical research among West- macroeconomic consequences difficult to assess. There em economists during the past half century. China could are moreover alternative ways of addressing these issues, benefit from this body of information and experience, each with its own costs and benefits. Comprehensive possibly in part through contact with organizations such analysis of alternative scenarios is needed, on which deci- as the International Monetary Fund. sions can be based, but the institutional capability for doing this does not yet seem to exist in China. POLICY PLANNING. The planning of economic policies The :nedium-term plan is thus the core of the entire is the central element in a planning system that places planning system, serving as the vehicle for policy adjust- 179 ment and change, and as the instrument to forge the planning efforts. Particularly in sectors where long lead consensus needed about those policies between all the times are required for new investments, or where a long- agencies concerned. And it is medium-term policy plan- term technological strategy is essential, there is a need for ning that should become the main responsibility of the a longer perspective, say, fifteen to twenty-five years. State Planning Commission, or of some other high-level Planning for these sectors will require the framework of a organ without specific responsibility for implementing longer-term view-not necessarily very precise or de- any particular set of policies. This in turn will require tailed-of the whole economy. The sector plans are un- development and strengthening of the statistical and ana- likely to be fully consistent with the overall long-term lytical capability to evaluate and study the current situa- framework; some may be subject to a significant margin tion, to offer comprehensive programs of policies to ad- of error or present alternative scenarios. The essence of dress specific issues, including time phasing for long-term planning is, rather, to anticipate possible basic implementing the programs, and to monitor program problems, so that timely policy adjustments can avoid implementation. otherwise insuperable bottlenecks or imbalances. France and Japan, for example, were selective in choos- PROVINCIAL PLANNING. The division of responsibility ing the sectors for planning; their objective was to make between central and local governments is a difficult issue these sectors more efficient and enable them to face inter- that cannot be resolved quickly. It appears consistent national competition. Both countries emphasized consul- with the direction of reform that the central government tation with the enterprises in the industries concerned would increasingly concentrate on policy planning while and the need for a consensus. France's planning agency planning of much infrastructure development and most was not sufficiently highly placed in the government to social services would take place at lower levels. have powers of direct enforcement, but merely tried per- Indeed, the consideration being given in China to dele- suasion, backed by investment subsidies. In Japan, the gating major responsibility for planning to provincial gov- leading role was assumed by the Ministry of International ernments, or perhaps even to municipal authorities, Trade and Industry, which could back up its strong views which would also have responsibility for surrounding ru- about the course and character of industrialization with ral areas, seems to fit quite well with the new approach to resources and policy instruments, but still relied mainly planning suggested above. Provincial governments should on consensus and allowed dissenting views. (Japanese be able to monitor and project the growth of production, experience in strategic planning of technological develop- the creation of new employment opportunities, and the ment was discussed in Chapter 7. Comparable South generation of incomes within provincial boundaries. Korean experience in electronics is described in Box 7.3.) Their proximity to lower-level authorities and markets Sector plans should not be limited to targets for output, should allow them to identify problem areas and reme- efficiency, and profitability, but should also address such dies quickly. Moreover, provinces are large enough to be issues as changing the organization of production; the able to remove bottlenecks by reallocating resources, to number, size, and location of enterprises in the sector; promote promising investments, and to ensure a fair and therefore the closure, merger, or expansion of exist- distribution of the benefits from growth within their ing production facilities and the establishment of new boundaries. ones. They should also address the creation or strength- Over the next two decades, interprovincial trade will ening of research capability and the training of skilled rapidly gain in importance, overcoming past autarkic ten- staff. Ministries responsible for formulating plans and dencies. Thus the national economy will be more inte- consulting with enterprises about their implementation grated, but the central government will correspondingly must be well informed on the state of technology in their need to promote more consistency among the various sector, both in China and abroad, and capable of dissemi- provincial production and investment plans. The Gov- nating this knowledge. Thus the role of sectoral minis- ernment will also need to monitor and regulate the levels tries in China could change greatly, yet remain of critical of saving and investment in different provinces and flows importance to the task of development management. of investment funds among provinces-especially be- tween those with widely differing income levels. These Toolsfor Planning flows will partly involve transfers through the central The quality of planning depends not only on the way it is budget, but the Government may also wish to participate organized, but also on the quality and relevance of infor- in investments across provincial boundaries and to influ- mation received. Adequate analytical capability is needed ence flows through the banking system. to handle inflows and outflows of information. LONG-TERM PLANNING. Although the medium-term INFORMATION SYSTEMS. The quality and timeliness plan can function as the core of all other planning activi- of decisionmaking depends on the information available ties, it may not provide a sufficient framework for all to decisionmakers. At present in China, both vertical and 180 horizontal information flows are inadequate for sound (consumption, investment, and external trade) on pro- policy planning. The process of aggregating data before duction, investment, and employment. It is not suited passing them to the next highest level of government either to analyzing most problems of specific sectors (it precludes analysis of more detailed data at higher levels, was supplemented with more detailed agriculture and even for the main economic indicators, as well as differ- energy projections), or to investigating some other mac- ent aggregations of data. Another problem is that data are roeconomic issues, such as urban-rural income distribu- passed up through vertical channels, but not adequately tion or government revenue-raising (which Chapters 5 distributed horizontally. Thus, a provincial planning com- and 9 addressed with simpler projections). mission may not be aware of the production, employ- Constructing models for these and other purposes, and ment, or investment data that state enterprises under maintaining approximate consistency among them, could central control are reporting to their ministry. be made easier if the various kinds of data collected in For proper planning and delegation to lower levels, China were based on an overall accounting framework further work is needed to manage existing and new data such as a social accounting matrix. Many statistics are flows in ways that provide meaningful information. collected for administrative purposes, and their coverage Comprehensive tabulations need to be designed to meet is limited by administrative needs and spheres of control. the particular requirements of each level of government Econcmic analysis, in contrast, requires data by sectors, for data on production and consumption, financial flows, commodities, or economic activities, irrespective of the interprovincial and external trade, and investment ex- agencies involved in their production, trade, or use. So- penditures. Most important, users-economic research- cial accounting could provide the necessary integrating ers, policy analysts, planners-should have a major voice framework and thus assist policy analysis and planning. in deciding what types of data are collected and should The potential contribution of economic modeling, or have access to the primary data base. other sophisticated analytical tools, should not be exag- In a decentralized economy, the behavior of the numer- gerated. The effectiveness of all tools depends not so ous economic agents controlling resources needs to be much on their technical complexity, but on how they are understood when designing economic policies. This re- used. The quality of economic planning in China will quires collection and analysis of data on the behavior of thus depend largely on the common sense and good judg- economic agents-a field of statistical activity barely ment jf planners, on improved economic training (not touched in China. The most obvious gap concerns con- only for planners, but also for decisionmakers throughout sumer behavior, in different parts of the country and at the economy), and on wider dissemination and discussion different income levels, in response to changes in incomes of economic and social information. and prices. Although household budget data are col- lected, they are not processed in ways that allow much economic analysis. Similarly, on the production side, little Overview is known about the supply response to prices and other Coordinated changes are required in many aspects of market forces, whether in the agricultural or the nonagri- China's economic and social system. One set of changes cultural sectors. involves increasing the dynamism and efficiency of the economy through greater autonomy for peasant house- ANALYTICAL TOOLS. The range of analytical devices holds and enterprises, more lively competition, and pro- for planning includes both the very simple and the very motion of collectives and individual enterprises to com- sophisticated, as well as both specific and comprehensive plement the state sector. International experience, as well tools. In many countries, planners use quantitativc as experience in China over the past few years, makes models for policy analysis. Only a few countries have, clear the need for coordination among the various ele- however, constructed models for medium-term planning; ments of this economic reform. To be efficient, for in- even fewer have succeeded in putting these models to stance, enterprises must be motivated to increase their timely and effective use. The task is far from easy, requir- profits and to respond to demand; they must be given ing the allocation of scarce talent, well in advance of the independence and freedom of maneuver; they must be planning process, to a demanding and even risky under- faced vith economically rational prices; and they must be taking. subjected to competition. None of these elements is indi- Models also have to be designed with specific objec- vidually easy to establish, and the absence of any one of tives in mind. A set of specifically focused, but not ex- them reduces or nullifies the benefits of the others: for actly mutually consistent, models is usually more helpful example, appropriate motivation produces bad invest- for policy purposes than a large multipurpose model. For ment cecisions if prices are irrational and exploitation of example, the multisectoral model described in Chapter 2 customers in the absence of competition. is useful mainly to analyze the long-term effects of An cnlarged role for markets and competition, though changes in the level and composition of final demand it will undoubtedly improve efficiency and accelerate 181 technological progress, could potentially also have unde- overcoming its weaknesses. This will surely not deter sirable economic and social consequences, including un- China-in the past, a successful pioneer in many areas- employment, unacceptably low (and high) wages, bank- from moving ahead. But it argues for a gradual advance, ruptcy of enterprises and dismissal of workers, and the with experimentation and evaluation at each step, even poor and the backward being left further behind in the though a one-stroke change would in principle involve development process. Increasingly, there will thus be ten- fewer internal inconsistencies. Experience in Eastern Eu- sions between concern for greater efficiency and dyna- rope also suggests the importance of moving steadily and mism, and concern about fairness and about the very of trying to avoid ill-judged steps in the direction of poor. These tensions could become especially great in market regulation that subsequently have to be reversed China, not only because its socialist ideology emphasizes or administratively tampered with, thus creating needless fairness and poverty alleviation, but also because many of uncertainty. Indeed, quite apart from the immense size of China's current social institutions and instruments are ill- China and the need to avoid major mistakes, the type of suited to deal with the emerging problems of a reformed reform envisaged by China's leaders requires a new way economy. Reform in China must therefore include of thinking, and a new pattern of behavior, from eco- strengthening policies and institutions to address social nomic decisionmakers at all levels that can come only and equity concerns, and especially to help the poor. gradually. Reform in China must also involve strengthening the Not all the steps need be small, though. In some cases, state's ability to direct the future course of economic and despite the greater risk of error, it may be best to intro- social development, combining more limited use of direct duce substantial packages of simultaneous reforms. Nor controls with much expanded use of indirect controls. need progress be slow. What has been accomplished in The range of instruments available to the Government China's rural areas in the past few years has provided not will need to be selectively used in a consistent, purpose- only an example, but also an excellent opportunity and ful, and effective manner. Many of these instruments will indeed a vital need for complementary and similarly rapid need to be designed to influence decisionmakers through- progress in the urban economy. Though in many ways out the economy, rather than to confine or prescribe the more complicated and troublesome than rural reform, behavior of specific economic units. urban economic reform probably has the advantage of On each of these fronts, there are promising ways for- not needing to be so uniform. Other countries-notably ward, but also problems and hard choices to be faced. In France and Japan in recent decades-have successfully addition, it is hard to overstate the importance and diffi- applied different management methods in different sec- culty of striking a correct balance among the three. Very tors and enterprises, and China should be able to do the few countries have combined state and market regulation same, while constantly seeking to refine and improve the in such a way as to produce rapid and efficient growth, mixture. and fewer still have also managed to avoid intolerable In system reform, and in the many other areas covered poverty among substantial segments of their populations. by this report, both the potential for progress and the On the contrary, there are far more countries in which problems involved are so large, and there is so much that unhappy combinations of plan, market, and social institu- is without historical precedent, that an even-handed and tions have produced neither rapid growth, nor efficiency, credible conclusion may be impossible. At a minimum, nor poverty reduction. though, China's long-term development objectives seem There is thus a vital need to guard against losing the attainable in principle, and if recent experience is any strengths of the existing system-its capacity to mobilize guide, there is a good chance that they will be attained in resources, as well as to help the poor-in the course of practice. 182 r- I f> I I I I 806 0 16o2- -1 4O u R~~~~~~~~~~~~~~~~~~~~~~~~~~~~. MONGOLIA Canhu ~/~MONGGOL LIAONI~ ~ EMOqF O AFGHANISTAN KO~~~~~~~~~~~~SH4N)KSANN KILOMETERS ~ ~ ~ ~ ~~~~~~~~~hi~zhu JIAN $ H AN 5 3 N~Wuanca RELIEF AND ADMINISTRATIVE DIVISIONS /Jf vh Elevations in meters: National capital Above 3000 a Province capitals 500-3000 __-Province boundaries LNni Below 500 ~~International boundari'es BU A Rivers flf PEOPLE SEPTEMBER 1985 The most recent World Bank publications are described in the annual spring and fall lists. The latest edition is available free of charge from Publications Sales Unit., Department B, The World Bank, Washington, D.C. 20433, U.S.A. I I I z fife sl08 o