A WORLD BANK COUNTRY STUDY - 76199 CHINA Agriculture to the Year 2 000 Annex 2 to CHINA 9H HD2097 .C462 1985 c.3 International Bank for Reconstruction and .thina, agriculture to the year 2000. JOINT BANK-FUND LIBRARY 9H H02097 .C462 1985 c.3 Ch;"liili~·~~~m~~~r· JLC067979 A WORLD BANK COUNTRY STUDY CHINA Agriculture to the Year 2000 Annex 2 to CHINA Long-Term Development Issues and Options The World Bank Washington, D.C., U.S.A Copyright © 1985 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 printing September 1985 World Bank Country Studies are reports originally prepared for internal use as part of the continuing analysis by the Bank of the economic and related conditions of its developing member countries and of its dialogues with the governments. Some of the reports are published informally with the least possible delay for the use of govern- ments and the academic, business and financial, and development communities. Thus, the typescript has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. The publication is supplied at a token charge to defray part of the cost of manufacture and distribution. 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 is 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. The most recent World Bank publications are described in the annual spring and fall lists; the continuing research program is described in the annual Abstracts of Cu"ent Studies. The latest edition of each is available free of charge from the Publications Sales Unit, Department T; The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from the European Office of the Bank, 66 avenue d'Iena, 75116 Paris, France. Library of Congress Cataloging-in-Publication Data Main entry under title: China, agriculture to the year 2000. (Annex 2 to China, long-term development issues and options) (A World Bank country study) 1. Agriculture--Economic aspects--China. 2. Agri- culture and state--China, I. International Bank for Reconstruction and Development. II. Series: Annex to China, long-term development issues and options ; 2. III. Series: World Bank country study, HD2097.C462 1985 338.1'0951 85-17914 ISBN 0-8213-0601-4 Preface This report ts one of six annexes to a main report entitled China: Long-Term Development Issues and Options. The complete list of annex volumes is: Volume 1 - China: Issues and Prospects 1n Education Volume 2 - China: Agriculture to the Year 2000 Volume 3 - China: The Energy Sector Volume 4 - China: Economic Model and Projections Volume 5 - China: Economic Structure 1n International Perspective Volume 6 - China: The Transport Sector (Note: Throughout this volume references to Annexes A, B, C and so on refer in that same order to Annex Volumes 1, 2, 3 etc.) The main report and annex volumes were prepared principally by members of an economic mission that visited China twice in 1984, for four weeks in February/March and for five weeks in April/May. In addition to Beijing, the mission went to three provinces: one coastal and relatively high-income (Jiangsu); one inland and average-income (Hubei); and one interior and low-income (Gansu). It received a lot of information, as well as numerous valuable comments and suggestions, from officials and others in these provinces, as well as from those in many central agencies and institutions, including: the State Planning and State Economic Commissions; the Ministries of Finance, Agriculture, Coal, Communications, Education, Foreign Economic Relations and Trade, Labor and Personnel, Petroleum, Railways, Urban and Rural Construction, and Water Resources and Electric Power; the State Statistical Bureau; and various universities and research institutes of the Chinese Academy of Social Sciences. A series of seminars was organized by the Technical-Economic Research Center under the State Council. The generous and thoughtful assistance of all these people in China contributed greatly to the preparation of the reports. The Bank mission was led by Edwin Lim (mission chief) and Adrian Wood (deputy mission chief), and also consisted of William Byrd (econo- mist), Mats Hultin (senior education adviser), Erh-Cheng Hwa (senior econo- mist), Timothy King (senior economist), Jacques Yenny (senior transport econo- mist}, Umnuay Sae-Hau (research assistant), Betty Ting (interpreter), Luc De Wulf (senior economist, International Monetary Fund), Benjamin King (consul- tant on statistics), Wouter Tims (consultant on planning and agriculture); and the following teams: Agriculture: J. Goering (team leader, April/May), Tom Wiens (team leader, February/March), Lang-Seng Tay (irrigation specialist), Lo-Chai Chen (fishery consultant) and Fred Bentley (consultant on arid agricul- ture); Energy: Roberto Bentjerodt (senior economist, coal projects), Weigong Cao (power engineer), Abdel El-Mekkawy (engineer, petroleum proj- ects), Robert Taylor (energy economist), and Darrel Fallen-Bailey (con- sultant); D.C. Rao (Assistant Director, Energy Department) led the team in the field; - 111 - Industrial Technology: Gene Tidrick (team leader), Anupam Khanna (indus- trial economist), Reza Amin (industrial specialist), and Josephine Woo (research assistant); Location and Trade: Ian Porter (team leader), Vernon Henderson (consul- tant on urbanization), John Sheahan (consultant on industrial location and trade) and Samuel Ho (consultant on rural nonfarm activities). The following also contributed to the preparation of the reports: Wlodzimierz Brus (consultant on socialist economies), Gerhard Pohl (energy and transport), Robert Drysdale (Annex Vol. 1); Helena Ribe, Nikhil Desai (Annex Vol. 3); Shujiro Urata (Annex Vols. 4 and 5); and Lily Uy (Annex Vol. 6). Larry Westphal, Carl Dahlman and Bruce Ross-Larson organized background work on technology. Behrouz Guerami-N, Tejaswi Raparla, and Kong-Yam Tan helped with the multisectoral model, the input-output table and data for international comparisons. Ann Orr, Kenneth Hill, Moshe Syrquin, J.V.S. Sarma, Kenneth Cochran, Chang Hsin, Liu Ying and Cai Jinyong undertook research. Linda Mitchell and Terrice Bassler edited the reports; and Helen Kung assisted in their processing. The reports also benefited from comments of a review panel consisting of Anne 0. Krueger, Luis de Azcarate, Kemal Dervis, Janos Kornai (consultant) and managers of the East Asia and Pacific Regional Office. In addition to the main report and annex volumes, the following background papers have been prepared and are being issued as World Bank Staff Working Papers: 1. "The Asian Experience in Rural Nonagricultural Development and its Relevance for China" 2. "International Experience in Urbanization and its Relevance for China" 3. ''Alternative International Economic Strategies and their Relevance for China" 4. "International Experience in Budgetary Trends during Economic Development and their Relevance for China" 5. "Productivity Growth and Technological Change in Chinese,Industry" 6. "Issues in the Technological Development of China's Electronics Sector" 7. "The Environment for Technological Change in Centrally Planned Economies" 8. "Managing Technology Development: Lessons from the Newly Industrializing Countries" 9. "Growth and Structural Change in Large Low-Income Countries" The main report, other annex volumes and background papers are available from World Bank Publications, P.O. Box 37525, Washington, D.C., 20013 or from World Bank distributors listed on the last page of this volume. Prices will be furnished upon request. - lV - Table of Contents Page No. EXECUTIVE SUMMARY..... • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • vi 1. PERFORMANCE SINCE 19 7 9 ........................................... . 1 A. Introduction................................................ 1 B. The Changing Policy Environment............................. 1 C. Growth and Productivity..................................... 9 D. Incomes and Consumption..................................... 13 2. FUTURE DEMANDS ON THE SECTOR ••••••••••••••••••••••••••••••••••••• 16 A. Present Consumption Levels ••••••••••••••••••• B. Projections of Consumption to 2000 ••••••••••• .............. . .............. . 17 18 3. DETERMINANTS OF FUTURE GROWTH •••••••••••••••••••••••••••••••••••• 28 A. Land and Water Development.................................. 28 B. Fertilizers and Seeds....................................... 34 C. The Policy Environment...................................... 40 D. Agricultural Support Services............................... SO 4. PROSPECTS FOR MAJOR AGRICULTURAL PRODUCTS •••••••••••••••••••••••• 57 A. Food Crops •••••••.• c • • • • • • • • • • 111 ••••••••••••••••••••••••••••• 57 B. Industrial and Specialty Crops •••••••••••••••••••••••••••••• 64 c. ForestrY•••••••••e•••••••••••••••••••••••••••••••••••••••••• 70 D. Livestock •.............. ~ .......•......................•.•.• 72 E. Aquatic Products ••••••• 0 0 0 II 0 0 0 0 0 0 0 0 0 0 0 0 0 e 0 0 0 0 0 e 0 0 G 0 0 0 0 0 0 0 0 0 0 87 s. OPTIONS AND ISSUES') •••••••••••••••••••••••••••••••••••••••••••••• 91 A. Projections of Agricultural Growth •••••••••••••••••••••••••• 93 B. International Trade••••••••••••••••••••••••••••••••••••••••• 101 c. Adjustments in Domestic Structure ••••••••••••••••••••••••••• 103 D. Geographic Differentials and Income Distribution •••••••••••• 109 E. Prices .................................................... . 113 List of Tables in the Text 1.1 Agriculture in the State Budget••••••••••••••••••••••••••••• 6 1.2 Estimates of Rural Gross Capital Formation, 1982 •••••••••••• 7 1.3 Effects of Price Reform on Returns to Crop Cultivation....................................... 10 v Page No. List of Tables in the Text (cont'd) 1.4 Production Changes, 1979-83................................. 12 1.5 Composition of Rural Income and Expenditure, 1979-83........ 14 2.1 Projected Demand and Consumption Targets for Selected Agricultural Products, 2000...................... 20 2.2 Estimated Requirements in the Year 2000 for Concentrates and Feed Grains.............................. 24 2.3 National Supply Requirements in 2000 and Comparisons with Historic Rates of Production Growth.................. 25 3.1 Cultivated and Irrigated Land, 1970, 1975, 1977-82.......... 30 3.2 Selected Data from Two Gullies on the Loess Plateau, Shaanxi Province, Undergoing Treatment for Erosion Control............................. 32 3.3 A Comparison of Application Rates for Chemical Fertilizer, 1980................................. 35 3.4 Comparison of 1982 Border Prices with Domestic Procurement Prices............................... 43 3.5 Ratios of Farm Level Prices for Nitrogen and Grain, Selected Countries............................. 45 3.6 Comparison of Price Ratios of Animal Products to Feed with Efficient Feed Conversion Rates.............. 46 3.7 Annual Investment in Capital Construction of Agricultural Development, 1984-2000....................... 48 4.1 Projections of Grain Production, Area and Yields by the Planning Bureau, MAAF.............................. 58 4.2 Gaps Between Provincial and Trial Yields of Wheat........... 62 4.3 International Comparison of Industrial Crop Yields.......... 65 4.4 Cotton Production, Consumption and Trade, 1980-83........... 66 4.5 Area and Volume of Commercial Forests....................... 71 4.6 A Comparison of Selected Aspects of Livestock Production in China and the United States, 1980........... 74 4.7 Food and Feed Value of Various Oilseeds, China, 1981, per Hectare.................................. 85 4.8 Approximate Protein Requirements per Animal or Bird for Efficient Production of Pigs, Poultry and Eggs•••••••••••••••••••••••••••••••••••••••••• 86 4.9 Indicative Costs and Returns of Fishpond Improvement........ 89 5.1 Preliminary Government Targets for Agricultural Production and Employment •••••••••••••••••••••••••••••••••• 94 5.2 Agricultural Growth Projections, 1981-82 to 2000............ 97 5.3 Comparison of Alternative Growth Patterns •••••••••••••••••• 100 vi Page No. Appendix A - Tables A.l Net Changes in Annual Lending to the Agricultural Sector, 1980-82.............................. 116 A.2 Indices of Weighted Average Prices Received by Producers, 1978-82..................................... 117 A.3 Composition of Gross Agricultural Output, 1978-82........... 118 A.4 Rural Income Distribution, 1979-82.......................... 119 A.S Selected Rural Indicators for Major Regions, 1982, and Changes, 1979-82................................ 120 A.6 Food Balance Sheet, 1982.................................... 121 A.7 Per Capita Food and Nutrient Availability in China and Other Countries.............................. 122 A.8 Characteristics of the National Grain Bases................. 123 A.9 Comparison of 1982 Border Prices with Domestic Procurement Prices (Y/ton)....................... 124 A.lO Timber Production, 1981..................................... 125 A.ll Investment Requirements for Hypothetical Agricultural Development Program.......................... 126 A.l2 Commodity Prices and Price Projections...................... 127 A.l3 Basic Characteristics of Comparator Countries............... 128 A.l4 Agriculture's Shares and Relative Economic Performance in Comparator Countries....................... 129 A.lS Per Worker and per Hectare Contribution to Agricultural GDP in Comparator Countries.................. 130 Appendix B - Specialization in Industrial Crops B.l Provincial Comparative Advantage and Cropping Patterns, India 1977-78 and China 1981.................... 133 8.2 Interprovincial Specialization in Industrial Crops in China and India.................................. 134 Appendix C - Potential for Ruminant Livestock C.1 Pastoral Herds and Grassland Area by Type, 1971............. 139 C.2 Metabolizable Energy by Source, 1980........................ 140 C.3 Potential Metabolizable Energy with Improved Management..... 141 C.4 Dry Matter Production Under Alternative Development Scenarios, 1980, 2000......................... 142 C.S Overall Ruminant Production and Dry Matter Requirements..... 143 vii MAPS 1. Growth of Gross Value of Agricultural Output, 1979-82 2. Rural Per Capita Income, 1982 3. Chemical Fertilizer Use per Sown Hectare, 1982 4. Major Commercial Crop Production Bases 5. Meat Production Per Capita 6. Pastoral Areas viii ACRONYMS AND ABBREVIATIONS ATEC Agro-technical extension center BGDH bao gan dao hu ("contracting all actions to the household") CAAS Chinese Academy of Agricultural Sciences CAS Chinese Academy of Sciences CBE commune and brigade enterprise cu m cubic meter cun village FCR feed conversion ratio GBSF General Bureau of State Farms gm gram GVAO gross value of agricultural output, including sidelines GVAOR a restricted definition of GVAO which excludes sidelines ha hectare IARC international agricultural research center ICOR incremental capital-output ratio K potassium kcal kilocalorie m million MAAF Ministry of Agriculture, Animal Husbandry and Fisheries MCI multiple cropping index MWREP Ministry of Water Resources and Electric Power N nitrogen NGB national grain base NSC National Seed Corporation NVAOR net value of agricultural output excluding sidelines NYJJLC Nong Ye Jing Ji Lun Ceng (Essays 1n Agricultural Economics) p phosphorous PRS production responsibility system SH specialized households SSB State Statistical Bureau xiang township ix CURRENCY EQUIVALENTS The Chinese currency is called Renminbi (RMB). It is denominated in Yuan (Y). Each Yuan is 1 Yuan = 10 jiao = 100 fen In early 1984 the official exchange rate of the Yuan to the US dollar was around Y 2 = US$1. The internal settlement rate (ISR) of Y 2.8 = $1, however, was used in most merchandise transactions. The official exchange rate is now about Y 2.8 = $1. On January 1, 1985, the Government abolished the ISR. WEIGHTS AND MEASURES Chinese statistics are usually in metric units; in addition, ~ and jin are often used: 1 mu = 0.1647 acres 0.0667 hectares 1 jin o.s kg FISCAL YEAR January 1 - December 31 TRANSLITERATION The Pinyin system is used in this report. Note: In tables, individual items may not sum exactly to totals because of rounding errors. X CHINA AGRICULTURE TO THE YEAR 2000 Executive Summary Performance Since 1979 1. In an effort to improve incentives and management at all levels of the agricultural economy, the Chinese Government since 1979 has moved ahead rapidly to implement a comprehensive restructuring of rural institutions. The major feature has been introduction of the production responsibility system (PRS), the generic name for various contracting arrangements which define the rights and responsibilities of owners of assets (state, collective, or pri- vate), on the one hand, and managers of these assets on the other. The common PRS variant used in agriculture is the bao gan dao hu (BGDH) or "contracting all actions to the household" system. Under BGDH, the individual household has replaced the collective (production team) as the basic farm management and production unit. At higher levels, the commune is being renamed township (xiang), with its economic role confined to managing commune and brigade enterprise (CBE) development, while the brigade is reverting to the designa- tion village (cun) and losing much of its managerial and technical staff. Management of collectively-owned land is now contracted to households, usually in proportion to household size or labor force. Other collective assets are divided up, sold or contracted to individuals or groups willing to manage them. The household is obligated to pay taxes, make contributions to collec- tive welfare funds, provide its share of state procurement requirements, and contribute labor to maintain or construct public infrastructure. All remain- ing output may be retained by the household. BGDH has provided substantial rewards to those farmers who are able to increase production since obligations to the collective have been fixed in absolute terms, rather than on a percent- age basis, although amounts vary with the productivity of assigned land. 2. The shift in control over resources from collectives to individuals has raised the question of the adequacy of investment for rural development. Collectives appear to be handicapped by diminished income from agriculture and growing difficulties in drafting team labor for public works projects. The uncompensated "social labor obligation" to maintain infrastructure was estima- ted in 1982 to amount to Y 12 billion out of a total of about Y 65 billion 1n rural gross investment. Recent budget cuts designed to reduce the central Government deficit have meant declining state investments in agriculture. Of the 1982 investment figure of Y 65 billion, investment by the state appears to have accounted for only 5%, though it represents perhaps one fourth of directly productive investment. Collectives accounted for Y 13 billion of fixed investment in 1982, of which about 60% was derived from the net incomes of the CBEs. Estimates suggest that private rural savings are large enough to be a major source of finance for rural development, if these savings can be effectively mobilized without undermining the newly-established production incentives. xi 3. In order to improve living standards for farmers and provide incen- tives for increased production of certain essential commodities, the Government in 1979 made major adjustments in the level and structure of farm prices. Prices were increased by about 21% for grain, 15% for cotton, 25% for oilseeds, and 25% for pigs. The farm price structure was built around quota, above quota, negotiated and free market prices. After fulfilling obligations for quota and above quota deliveries, producers may sell remaining supplies at negotiated prices which fluctuate at approximate parity with free market prices. Negotiated prices apply to sales to the state by individual farmers for production which may be legally sold on the free mar~et. If no quota exists for a product, the above quota or negotiated price normally applies. As a result, increased procurements generally take place at higher than average prices and impart an uptrend to the weighted average price received by producers. With this more flexible, if complex, price structure in force, the number of agricultural commodities subject to state-established prices is declining. Measures are underway to further simplify the structure of agricultural prices. 4. The agricultural sector has performed exceptionally well since 1979. The gross value of agricultural output (GVAO) in 1983 grew by 9.5%, following growth of 11% in 1982, 5.7% in 1981, and 2.7% in 1980. Grain pro- duction in 1983 reached a record 387 million tons, some 17% higher than in 1979, due largely to a remarkable 22% increase in average yields. Though Chinese agriculture continues to be heavily crop oriented, the period has seen a steady decline in the share of crop production, a marked increase in side- line production (gross value of hunting and gathering and industries run by brigades and teams) and more modest relative gains in livestock. 5. Both incomes and consumption have risen markedly since 1979. In nominal terms, per capita incomes of farmers in 1983 were 2.3 times those of 1979. Incomes of employees in state enterprises (mostly urban dwellers) rose by 66% during this period and were nominally 70% higher than rural incomes in 1983. In real terms, rural incomes may have increased by 70% while urban incomes were up by about 40%. Consumption gains have been substantial, espe- cially in red meat (up 53%) and vegetable oils (up 48%). By 1982 the daily average Chinese diet provided over 2,700 kCal of food energy. Urban residents generally have access to more animal protein, edible oils, fruits and vegetables, whereas rural residents consume considerably more grain per capita. While urban/rural and interregional disparities remain in per capita nutrient availability, China's food procurement and distribution system has provided basic food nutrients to most of the population most of the time. Future Demands on the Sector 6. Average per capita consumption levels in China markedly exceed those of other low income countries and an unusually high proportion of calories comes from grain. At 2,580 kCal (1980-82 figures), the daily energy avail- ability exceeds the 1975-77 average for middle income countries (2,560 kCal) and the world (2,570 kCal). While daily protein availability of 66 g exceeds the average for all developing countries (57 g) and approaches the world aver- age (69 g), only about 10% of protein intake is from animal sources, as compared to a world average of 35% and 21% for developing countries. xii Similarly, the availability of animal and vegetable fat in 1980-82 was 5% below the average in developing countries and more than 40% under the world average. At present, per capita availabilities in China exceed estimated requirements of food energy by more than 20% and of safe minimum levels of protein intake-by more than 85%. 7. If consumption patterns in China follow those observed in other countries, higher incomes will bring about a substantial restructuring of con- sumption over the coming decades. This would involve a marked reduction in direct per capita consumption of grain and a corresponding increase in con- sumption of animal products, along with other forms of dietary diversifica- tion. Analysis of food consumption demands in 2000, using alternative assump- tions of 3.5% and 5.5% per capita GNP growth, shows in both cases a radical adjustment from the present dietary pattern to a more varied diet, in which meat, eggs, fish, milk, pulses, fruits, sugar and vegetable oil are larger components. Direct consumption of cereals and tubers would decline by 15-20% overall, largely in the coarse grains and tubers. Within the fine grains category, it is likely that per capita consumption of wheat, in China's cities in particular, will rise steadily in response to higher incomes and preferences for more conven1ence foods. 8. The amount of indirect grain consumption depends heavily on the efficiency of production in the animal husbandry and food industry sectors. While estimates based on high and low feeding efficiency factors vary considerably, they clearly indicate that high future demand for livestock products would require correspondingly large increments of feed grains and protein meals. Assuming rapid economic growth, the difference between added feed-grain requirements under efficient versus inefficient livestock produc- tion systems would exceed 85 million tons of feed-grains annually by the year 2000. More difficult to project are the requirements of the industrial sector for non-food products and additional demands which might be placed on the sector to generate export earnings in support of overall economic growth. If the Government's income and population growth targets were met, total grain supply in 2000 would have to increase by about 50% from 1980-82 levels (2.2% p.a.) and more than a third would be committed to livestock feeding (versus 10% at present). Feed grain requirements would exceed Government projections of supply. On the demand side, price and foreign trade restrictions can be used to encourage consumption behavior to conform to available production. Other countries, both developed and developing, have used these measures to hold food consumption to levels below those expected or "normal" for a given level of GNP per capita. Present Chinese price policy has probably had the reverse effect, because food prices have been kept low relative to prices of industrial consumer goods and the supply of urban services (e.g., housing) has been subsidized and restricted. Because basic nutritional requirements have already been met on the average in China, some restraint of consumption may be acceptable from a nutritional point of view. On the supply side a key consideration may be the nutritional and resource tradeoff between direct consumption of grain and its use in livestock feeding. xiii Determinants of Future Growth 9. Land and Water Development. Unique features of China's agriculture include the relatively small portion of the total land area which is culti- vated (about 10% as compared with 75% in India) and the large portion of the total cultivated area which is irrigated (45%, as compared with 23% in India). Uncertainties exist regarding the actual cultivated area in China. Although official figures of just under 100 million ha continue to be used, the actual figure may be a quarter to a third higher (and actual yields correspondingly lower). Official figures indicate that the cultivated area has declined marginally in recent years. The reported area of cultivated land in 1983 (98.4 m ha) was only about half a million ha larger than it was in 1949. Future availability of agricultural land largely depends on (a) the extent to which reclamation will compensate for land lost to non-agricultural uses and (b) the extent and effectiveness of future irrigation development. 10. The Government assumes that China's cultivated area will remain essentially unchanged over the next 15-20 years, as land lost to roads, urban- ization and industrial development is approximately offset by newly-reclaimed land. Estimates show perhaps 3-5 m ha as suitable for development in the medium-term for sustained production of annual crops. These figures may be compared with the reported loss of about one m ha annually to non-agricultural uses in 1959-78. Annual losses at present are not known but, outside major municipalities, there are few legal restrictions on conversion of farm land to non-agricultural uses. The current surge of rural housing construction suggests that losses of agricultural land will continue to be substantial, despite government efforts at prevention. 11. Given China's land scarcity, irrigation and drainage take on par- ticular significance as means to increase the sown area. MWREP projects additions to the irrigated area by 2000 of 8-10 m ha (from 45 m ha to 53-55 m ha). The amount of land actually irrigated has not increased significantly in 1978-83. About half (4.3 m ha) of MWREP's planned expansion of irrigated area would fall in the rice region south of the Huai River, largely through upgrading of existing facilities. Another large component (1.6 m ha) would occur in the North China Plain, primarily through the project to divert Yangtze River water northward. The remainder would be split between the Northeast and the arid areas of the Northwest loess region as well as the two provinces of Xinjiang and Gansu. In addition to expansion of irrigated area, an estimated 23 m ha of cultivated land require drainage facilities to prevent waterlogging and to reduce salinity. 12. Fertilizers. Increased use of chemical fertilizers has been an important factor in China's impressive agricultural performance in recent years. Application rates for chemical fertilizers doubled between 1977 and 1981 and by 1982 averaged 169 kg of nutrients per cultivated ha. Since 1979 chemical fertilizer use has been stimulated by greater availability and improved fertilizer:product price ratios resulting from recent price reforms. China also has a long history of using organic fertilizers and organics will continue to be important sources of nutrients, particularly for phosphorus, potassium and some trace elements. MAAF estimates that half of nutrient offtake will come from organic sources by 2000. This is likely to be xiv optimistic if utilization patterns ~n China follow those in other countries as agriculture modernizes. 13. About 85%, or 13.8 million tons, of China's total consumption of chemical fertilizer is produced domestically in some 2,200 plants. Production is characterized by a small number of product types and low nutrient content. Recently China has developed the capacity to design and construct efficient, large-scale plants for the production of nitrogenous fertilizers, but the four plants expected to be operational by 1990 will increase total production by less than a million tons. For both potassium and phosphate production, achieving rapid growth will require an early commitment of funds and manpower to master new technologies in the design and construction of large plants. By the year 2000 MAAF expects consumption of chemical fertilizers to reach 30 million tons of nutrients or half of total fertilizer requirements. It assumes domestic production of 24 million tons, elimination of imports of nitrogenous fertilizers, but continued imports of phosphorus and potassium to meet large domestic shortfalls. In recent years, fertilizer imports have averaged ten million product tons (2-3 million nutrient tons) at a foreign exchange cost of over US$ 1 billion p.a. 14. About 20% of domestic fertilizer production is distributed through the state system at centrally-estaolished prices. This system of allocating fertilizer has been used as a policy instrument to encourage production and marketing of key commodities. In the past it has tended to favor food crops relative to feed grains and pulses; the high, stable yield areas relative to lands of lower quality; irrigated areas relative to non-irrigated areas; and the state farms relative to the collective sector. This pattern of allocation appears much less efficient today. At the new higher levels of application, marginal yield responses in the high yield localities are reportedly below marginal yield responses in the low yield areas. There is therefore reason to reallocate available supplies, perhaps by allowing the market to play a large role. Important tasks for the Government are to improve the efficiency of distribution and strengthen fertilizer testing programs to guide production and utilization. 15. Seeds. Recognizing that high quality seeds in adequate quantity are critical inputs to increased crop yields, the Government since 1979 has given special support to strengthening China's seed industry. Now in place is a National Seed Corporation (NSC) under MAAF which functions as China's chief regulatory and policy/planning agency for seeds. There are also about 2,300 NSC-affiliated but locally-financed and administered seed companies which handle multiplication, processing, storage, certification and distribution of stock and certified seed. Breeder seed is supplied primarily by provincial agricultural research institutes. Multiplication of stock and certified seed is carried out by seed farms either managed by or under special contract to the local seed companies. The Central General Bureau of State Farmer (GBSF), also under MAAF, operates a parallel system of seed companies and farms. Though the basic mechanisms are in place, both processing facilities and qual- ity assurance procedures are inadequate. NSC and GBSF-affiliated companies supply about 40% of total national seed requirements (17 million tons), but less than 1% of this seed has been properly processed and certified using modern methods. As a result, farmers for the most part are forced to rely on XV poor quality seed which translates into high seeding rates, high production costs, and less than optimum yields. The Government's current program of seed industry development is aimed at modernizing over 300 seed production, processing and distribution centers by the year 2000. 16. The Policy Environment. Introduction of the PRS has significantly improved agricultural production incentives and can share credit with price reform for the outstanding growth record of Chinese agriculture since 1979. Recent movement towards an enlarged role for prices and market forces can be expected to enhance the climate for agricultural growth. There is a concern within the Government, however, that for a farm sector comprising a vast number of independent decision makers, such moves would lead to imbalances in the quantities and mix of crops grown relative to national requirements. Quotas for sown area of grain thus reportedly have been retained in order to prevent farmers from shifting a large portion of land under their control from grain to more profitable industrial crops. Analysis of domestic procurement prices relative to border prices for major commodities suggests the need for continued adjustments as the basis for promoting growth and efficiency in the sector. A number of important commodities, including timber, fish, and pork, appear to be underpriced relative to international prices and in fact may be too low to stimulate expanded production. On the other hand, some products such as sugar crops, natural rubber and edible oils may be overpriced. With regard to production inputs, prices of chemical fertilizers appear to be near economic levels. Prices of electric power in rural areas are relatively cheap while some farm machinery remains costly by international standards. 17. A key question is whether over the next two decades productive investment in the sector will be adequate to sustain agricultural growth at rates needed to achieve national growth and income targets. State investment in agriculture, according to current Chinese estimates, is not expected to grow by more than Y 1-2 billion annually. Collective investment (excluding that in CBEs) may continue to decline. Therefore, the ability to effectively tap private savings for productive investment is likely to be a critical determinant of the adequacy of agricultural investment. Measures to attract private savings might include granting private use rights over collective or state property, making time deposits attractive through higher interest rates, instituting a tax on idle assets, and encouraging rural credit institutions to devote a larger proportion of credit to longer-term lending. 18. The Government's investment priorities in agriculture appear to differ substantially from present subsectoral shares. The share of proposed investment which would benefit crop cultivation (about 30%) is less than half of the crop share in GVAOR (about 75%), whereas the other subsectors' shares are at least twice their present product shares. While these investment "weights" are generally consistent with the structural change that the Govern- ment supports in agriculture, they may not be an effective means of achieving production and income objectives. More than half of total investment would be absorbed by dairy and forestry development. Given the current high costs of dairying operations, massive investment in dairy development appears difficult to justify on efficiency grounds. And, while forestry development in general warrants priority on need and environmental grounds, expenditure on afforesta- tion will add relatively little to annual product in the next few years because of the slow maturing nature of this type of investment. xvi 19. Agriculture Support Services. Because yield increases will be the source of most of China's agricultural growth, a strong research program aimed at improving production technology is of particular importance. Although it is difficult to prescribe an optimum level of investment in research, esti- mates elsewhere suggest that building an effective national research capa- bility requires an investment of up to 2% of the annual value of agricultural output. For China, this would imply an annual expenditure of up to Y 4 bil- lion, considerably more than current levels. The bulk of agricultural research in China today is carried out by 390 research institutes at the pro- vincial level and above which operate under the general professional guidance of MAAF's Chinese Academy of Agricultural Sciences (CAAS). Research personnel in these institutes number about 18,800. Another 15,300 researchers work in 49 crop and livestock research institutes run by the provinces. Some research is also carried out in China's 85 agricultural colleges and universities. Despite some very notable research achievements, work at many research insti- tutes is hampered by inadequately qualified staff and poor research facili- ties. A large part of the intellectual leadership for agricultural research in China is provided by staff who are more than 60 years of age. 20. China has some 85 institutions of higher education in agriculture and related fields, with a student enrollment (1981/82) of about 100,000. This represents only 8% of those enrolled in institutions of higher learn- ing. Agriculture is generally not the first choice of candidates for college and those admitted into agricultural college and universities frequently have lower scores on admission tests. There is also a problem of retaining agri- cultural technicians in this sector after training is complete. Special quotas for admission of rural students to agricultural training institutions and additional salary incentives for those working in rural areas have been implemented to ease these problems. 21. The restoration of a farming system organized around individual households has placed new demands on agricultural extension services. Whereas previously the various levels of collective organization could serve as major foci in the delivery of production inputs and technical assistance, now the extension organization must establish direct links to nearly 200 million pro- duction units which operate much as do small family managed farms elsewhere. In response to this requirement, the Government is in the process of consoli- dating previously separate local agricultural services, including seed production, fertilizer distribution, and soil analysis units, into a single, county-level agrotechnical extension center (ATEC). Efforts also are underway to link extension more closely with agricultural research and education. ATECs already in existence - about 300 out of a projected 2,300 - are working more closely with local research institutes and colleges in such areas as varietal testing, extension methodology, and training and upgrading of local farmers and technicians. xvii 22. Provision of agricultural machine services has also been affected by development of the household farming system. Reductions in field size, as well as the farmer's desire to obtain maximum control over timing of farm operations and to minimize cash expenditures, have caused a decline in demand for large and medium-scale machinery, a corresponding rise in demand for draft animals, and continued strong demand for small-scale pedestrian tractors. About one third of all tractors in the collective sector now are privately owned. The emerging organizational form for provision of machine services involves private or collective custom land preparation by specialized operators or teams whose earnings are supplemented by short-haul transport work. Successful development of custom hire machine services would help reduce grain and forage consumption by draft animals. 23. Efforts to restructure agriculture and increase specialization are constrained by the limited capacity of the rural transportation, processing, storage and distribution systems. Inland transport - especially the rural road system - and port facilities must be improved in order to handle the projected expansion in traded commodities. Relative neglect of transportation networks under past policies of local self-sufficiency has made rural trans- port a particularly acute problem. Solutions have been slow in coming in large part because of inadequate coordination between agricultural planning agencies and agencies responsible for transportation. Grain storage is becoming an increasingly serious problem largely because of inadequate investment in the past and the surge in production in recent years. Cold storage capacity for perishable fruits, vegetables and animal products will have to grow very rapidly if the changing pattern of consumer preferences is to be accommodated. Improved inland transportation and storage facilities 1n turn would provide a basis for agro-industrial development which requires assured deliveries of high-quality, uniform raw materials. Prospects for Major Agricultural Products 24. Food Crops. The Planning Bureau within MAAF has targeted grain production to grow by about 2% annually from the average 1980-82 base while the area sown to grains would continue to decline marginally by about 0.1% p.a. The roughly 40% projected increase in rice production is expected to come entirely from higher yields. A major factor expected to contribute to increased yields is the projected expansion in area planted to hybrids from the 1983 level of nearly 6.8 million ha to 13.3 million ha by 2000. With a net yield advantage (taking into account the larger areas required for seed production) of about 20% over conventional varieties, an increase from current levels of the hybrid area by the implied 7.7 million ha, would generate nearly a third of the projected increment in rice production by 2000. Area expansion for hybrids would depend on successful introduction over the next few years of new early maturing varieties in double cropping systems. This seems feasible from technical and organizational viewpoints. With a yield advantage for hybrids of 0.75-1.0 tons per ha, the incremental benefit-cost ratio of about 5:1 suggests that hybrid rice is economically attractive to producers. 25. Wheat is the second most important grain in China after rice and accounts for about 20% of grain production and 25% of the area sown to grain. MAAF's Planning Bureau projects both wheat and corn production to grow xviii more rapidly (2.6% p.a.) than other major grains. With an assumed small decline in the sown area, this implies annual growth in wheat yields of 2.6% p.a. to 3.3 tons per ha. This would require continued attention to wheat disease problems, good water control in irrigated areas, and a proper nutrient balance in fertilizer applications. 26. Corn production accounts for about three quarters of coarse grain output and 18% of total grain output. Production since 1979 expanded at 3.2% p.a., or about one-half as fast as wheat. The Government projects virtually no change in the cropped area to 2000, but yields are to grow by about 70%, or 2.8% p.a., 1979-81 to 2000. China's corn yields (adjusted for likely under- reporting of the planted area) are slightly higher than average yields in developing countries, but less than half those in developed countries. Much the same pattern holds for other coarse grains. Factors restraining yields of coarse grains include poor seed quality, inadequate irrigation during times of moisture stress, low application of chemical fertilizer and poorer cultural practices compared with e.g., wheat and rice. Growth of total coarse grain production at a substantially higher rate than the Government target would be required to provide the feed grain necessary to meet output targets for livestock and poultry. 27. Industrial and Specialty Crops. With the significant exception of cotton, national average yields and per capita production of industrial crops (oilseeds, fibers, sugar crops, tobacco, tea) on the whole appear to be lower in China than in a number of comparator countries. The demand for these crops as well as for specialty crops such as fruits and vegetables is expected to grow rapidly as incomes rise. In the case of cotton, special encouragement given to production by price and procurement reforms and the planting of improved varieties stimulated a 22% expansion of cropped area and a 72% increase in yields during 1978-83. As a result of these area and yield gains, China has moved from being a large importer of cotton to basic self-suffi- Clency. Prospects seem good for further yield increases. 28. Production of oilseed crops (including rape seed, peanuts, sesame and sunflower seeds, ~ut excluding soybeans and cottonseed) more than doubled during 1978-82 due to both increased area and yield. Soybean production has grown by about 5.2% p.a. since 1978, largely due to increased yields, bu~ still is the only major crop grown in China in which the reported national average yield is less than that for developing countries as a whole. A steady increase in soybean yields will be required if China is to meet future needs for high quality protein meal for livestock feeding. Prospects for tea are clouded because of possible marketing constraints at home and abroad. Produc- tion of tobacco is likely to be increasingly influenced by the public percep- tion in China of the health dangers inherent in its use. Caution may be required also with other crops in this category- e.g., sugar cane and sugar beets - for which China may have little overall comparative advantage and domestic procurement prices exceed world prices by substantial margins. For these crops the primary emphasis probably should be to improve yields through better cultural practices, improved planting materials and careful choice of planting location. xix 29. Forestry. Development plans call for the area under forest to increase from the present 13% of China's total land area to about 20% by the year 2000 to meet expected rapid growth in demand for forest products. Annual per capita wood consumption is about 0.05 cum (excluding fuelwood), compared to about 0.02 cu m in India and 1.5 cu m in the US. Reflecting local short- ages and increased demand from population and income growth, imports of forestry products rose from US$242 million in 1976 to peak figures of US$642 million in 1980 and about US$540 million in 1981. Major thrusts of future development include afforestation, increased exploitation of mature and over-mature forests, expansion of forest industries and strengthening of forest research, education and extension. Investment programs by the state forest farms, of which there are some 3,900, have been hampered by the relatively low price they receive for industrial round wood sold to the Government. Despite recent increases, state procurement prices for timber are only about one-half negotiated prices. 30. Livestock. By the year 2000, the share of the livestock sector in GVAO ts projected by the Government to rise from the present 15% to 25%, with meat consumption more than doubling. Achievement of targets will depend heavily on use of animals with higher genetic potential, better production and management systems, improved veterinary services and the availability of much larger supplies of grass, grain and protein supplements. Low grain feeding systems and seriously inadequate levels of protein supplementation are major factors contributing to present technical inefficiencies in China's livestock sector. Solution of the toxicity problems in China's rape seed and cottonseed meals would greatly enlarge potential supplies of protein meal, although soybean meal is still the most appropriate meal for poultry and swine rations because of its amino acid balance. In the absence of adequate local supplies, the development of a capacity to produce synthetic amino acids warrants consideration. 31. China's extensive grazing areas of the North are viewed as a poten- tial source of much of the large incremental supplies of beef, mutton and wool in future years. However, overgrazing problems in areas such as Nei Monggol and northern Hebei have been exacerbated by policies since 1979 which have stressed expansion in herd numbers and the rapid introduction of the PRS. While the PRS has encouraged individual care of animals, with the resuLt that mortality rates in some areas are down, it also appears to have destroyed the mechanism by which collectives formerly were able to control animal numbers and to implement pasture management systems. The combination of privately managed herds on a common grazing resource has added pressure to an already overburdened range land. The most important immediate requirement is to match livestock numbers with carrying capacity through herd reduction and improved range management. Although the focus of ruminant livestock development in China has been on the pastoral areas of the North and Northwest, more than 40% of the large animal population, including about half of China's cattle, is located in 12 southern provinces. While these areas may be technically suitable for pasture development, the financial and economic viability of ruminant development in South China also needs careful assessment. 32. Several different production systems are being tried in the live- stock subsector. Official policy increasingly favors animal production by XX specialized households (SH), although most of the dairy development heretofore has been on state-owned enterprises. The term SH refers to small-scale commercial animal farms, which differ from state or collective enterprises in their household management and/or ownership of animals and use of private funds and capital saving production techniques. A major problem faced by the SH system is ensuring adequate supplies of balanced feeds. In some areas, authorities have encouraged SH animal producers to subcontract their land to other farmers who serve as SH feed producers. This form of specialization avoids some of the costs of transporting feed grains, though it is not a fully reliable mechanism for providing adequate feed supplies. China also has a growing number of large-scale, commercialized pig and poultry operations, mostly in the suburbs of major cities, to serve urban or export markets. These operations employ confined feeding systems and modern production technology with varying degrees of success. Local management in some cases ts inexperienced and a few joint ventures with overseas interests are being organized. 33. China's dairy development policy is to promote production initially near large cities, stressing production by collectives and individual house- holds and using state farms as sources of improved animals and modern tech- nology. With high investment and operating costs, many of the larger dairy enterprises which are already operational do not appear to be financially viable, despite local prices for dairy products which exceed world prices by a substantial margin. Viability of some operations has been improved by integrating dairying with more profitable activities such as fish farming or fruit growing which use the manure as feed or fertilizer. But pig or poultry raising is likely to be a less costly source of nutrients for these opera- tions. The economics of large-scale dairy operations in suburban areas warrants careful study. Smaller production units owned by SH and located closer to cheap forage supplies may be preferable. For the inland areas, higher productivity from improved breeds and better pastures may establish the basis for economic expansion of dairy production. 34. Aquatic Products. China's aquatic products sector (marine and freshwater) accounts for only about 1.5% of GVAO. China's fisheries rank third internationally in total production, but one hundredth in per capita production. Production in 1983 totalled about 5.3 million tons, of which more than two-thirds was accounted for by marine capture fisheries. The Government plans to expand aquatic products output to 11 million tons by the year 2000 by emphasizing intensive fish cultivation, particularly in freshwater ponds. There is good potential to increase fish yields in many of the fish farming areas. Factors which have contributed to low yields include poorly designed pond layout, inappropriate breeds of fish or stocking rates and poor feeding systems. With regard to the marine resource, additional consideration should be given to the production of fish meal for the livestock and poultry industries by expanding the collection and processing of fish species now considered to have little economic value. Options and Issues 35. Between the present and the turn of the century, China's agriculture will be required to feed at better nutritional levels a population growing by xxi at least 10 million per year and to supply the raw materials to meet rapid growth in industrial demand. This will be a challenging task. At best, the land area under cultivation will not increase and the sown area is likely to remain at present levels as land reclamation and irrigation are approximately offset by lands converted to nonfarm uses. Competition for available water in some parts of China between agriculture, industry and human consumption will increase and make agricultural expansion more difficult. Most of the produc- tion gains will have to come from increased yields on presently cultivated land. The general direction of development is clear, viz., the shifting of more resources into activities (1) which are less constrained by fixed resources (land and water); (2) for which improvement of management and labor skills has a high payoff (specialty crops, modernization of livestock and poultry production); and (3) for which the gap between existing and advanced technology is greatest (agro-industrial production). 36. It may be suggested that two decades (1980-2000) is too short a time to bring about major structural change in the sector, but examples exist and some lessons may be learned from them. Structural change in Korean and Japanese agriculture has been striking over a roughly similar period of time (1960-80) and began from levels of development in some respects comparable to China today. These structural shifts were not ''planned,'' but were largely the response of agricultural sectors in those countries to high rates of indus- trial growth and the attraction of agricultural labor to nonfarm employment. The comparison clearly suggests that rapid growth of the nonfarm economy is a requisite to rapid structural change in agriculture. 37. Projections of Agricultural Growth. Preliminary planning by the Government calls for rapid growth in livestock, poultry and sideline produc- tion and a relative decline in the crop share of GVAO by 2000. The farm labor force would decline sharply, while the rural nonfarm work force would rise extremely rapidly. The area sown to grain would remain unchanged, although wheat and rice production would grow slightly less rapidly than coarse grains. No significant increases are planned for areas planted to nongrain crop, although yields are expected to increase. 38. If the present generally favorable policies in agriculture continue, agricultural growth to the year 2000 is likely to be more rapid than long-run historical growth rates and structural change would be more rapid. However, this change may not be as rapid as indicated by Government supply targets or consumption demands under targeted income growth. Growth of crop production value is unlikely to greatly exceed the 3.7% p.a. achieved in 1965-83. Potential annual growth rates of grain production would be about 2.4% and nongrain crops 6.8%. Among grains, the MAAF target growth rates for fine grains seem feasible, but corn and soybean production would have to grow more rapidly to provide the feedstuff required to meet targets for the production of livestock and poultry. Nongrain crops would be required to grow at annual rates exceeding historical rates by 1-2 percentage points. Other components of GVAOR (excluding sidelines) are less subject to diminishing returns because of the land constraint and their growth rates will be determined more by the extent of Government and private investment, including that for agricultural support services. xxii 39. Though a wide range of alternatives is possible, under one set of reasonable assumptions regarding technical progress, agricultural investment, use of industrial inputs and prices, GVAOR would expand by about 4.4% per annum while NVAOR would increase at a rate of about 3.8% yearly. If the farm labor force expands by about 0.6% p.a., these increases in output would translate into an approximate doubling of gross labor productivity in agriculture by the year 2000. The share of crops in GVAOR would decline by about eight percentage points and the livestock share would increase from 18% in 1980-82 to 24% by 2000. While this projection of potential growth is slightly higher than the growth rate of total supply needed to satisfy consumption requirements in 2000, the structure of potential production is quite different from that projected for consumption. The value of potential crop production would significantly exceed consumption requirements (particularly of rice), while that of livestock and fisheries would fall short of requirements. Faced with such potential imbalances, the options could include measures to (1) adjust price and area for the crops in surplus, (2) encourage the production of feedgrains in the rice areas; (3) attempt to develop export markets for crops in surplus; {4) trade them for products in short supply (livestock products, forest products, etc.) or (5) import sizeable quantities of feedgrain and high protein meals to support livestock production. A challenge to Government policy would be to convert China's potential for rice and nongrain crop production into increased feed supply either through external trade or a major restructuring of domestic cropping patterns. 40. International Trade. Planning in China continues to be premissed on basic self-sufficiency in agriculture, although it is acknowledged that trade will continue to play a role in balancing supply and demand for farm products and in generating foreign exchange. However, a self-sufficiency orientation may well conflict with satisfaction of future consumption patterns and raises the question of the extent to which international trade could balance supply and demand at lower cost to the economy. Global supply and demand considera- tions suggest that in the next decade overall world trade in grains is not likely to grow as rapidly as in the 1970s. With generally abundant supplies, world prices for the major grains in real terms are expected to decline about 10% from 1976-82 levels. Because of the large volume of world trade in feed grains, it seems unlikely that a gradual increase in imports of these grains by China to support rapid growth of the livestock and poultry industries would have a significant effect on world prices. China's potential impact on the world rice market could be much larger because of the "thinness" of that market relative to the quantities which China could supply. The best export prospects in agriculture are likely to be for a wide variety of labor- intensive, high-value products for which international income elasticities of demand are high. But the successful production, processing and marketing of them in world markets would require strengthened support services in research, extension, market infrastructure and market development. xxiii 41. Adjustments in Production. Reflecting expected changes in domestic consumption patterns and world markets, the needed adjustment in cropping patterns might involve some reduction in the area planted to rice and expan- sion of the area under crops such as corn and soybeans required by a much larger livestock program. This might well include conversion of some of the uplands of the south from the production of low-yielding rainfed rice to feedgrains and oilseeds. Such a shift would allow peri-urban areas to plant higher value crops on lands now under grain and forage crops and provide a basis for more diversified production systems in the low income upland areas where rainfed rice is a major source of income. For the nongrain crops, bottlenecks in transport, storage and processing are major constraints to expanded production. For the oilseeds, the most urgent tasks, in addition to expanding production of high quality protein meals, are to improve varieties and processing technology so as to increase the suitability of these meals (particularly from rapeseed and cottonseed) as livestock and poultry feeds. China seems likely to maintain its position of basic self-sufficiency in cotton and may have supplies for export. Study is needed to compare the relative economic merits of further expansion of cotton for export or use of the land to, e.g., increase production of animal feedstuffs. For high value fruits and vegetables, the major challenges lie in expanding production to meet the expected sharp increase in domestic demand and exploiting possible export markets. It is in the processing, packaging and marketing of semi- perishables such as these that China's existing systems are most deficient. This appears to be an area where joint ventures with established foreign firms would be useful as a source of both technology and market access. 42. Central planners have posed ambitious targets for meat, dairy and egg production. Beef, mutton and wool production from the northern grasslands are likely to grow no faster than in the recent past, and possibly slower, because of serious overgrazing and deterioration of range lands. Shortage of feed and forage are also likely to limit livestock production in the southern grasslands and the agricultural areas. A solid technological base exists to expand pork production, but the rate of growth will depend heavily on the adequacy of high quality energy feeds and protein. China's poultry industry should be accorded high priority because (1) poultry meat and eggs are pre- ferred products among Chinese consumers, (2) modern production systems are efficient converters of plant materials to animal protein, and (3) technology to substantially improve production efficiency is available internationally and readily transferable. In light of the high investment costs, the objec- tives and means in China's proposed dairy development program should be recon- sidered. Production by state enterprises in suburban areas is increasingly handicapped by high economic costs of land and scarcity of abundant forage. In the medium term, increased milk supplies for urban consumers might well be provided at lower economic costs through the importation and reconstitution of relatively low cost milk powder. Greater emphasis on specialized households and smaller-scale production units would be expected to improve efficiency in the dairy industry. This would require improved access by these groups to better milking stock, good veterinary services and high protein feed sources to supplement local feed supplies. The technology to double the output of aquatic products is available, particularly for freshwater fish cultivation, but realization of that target will depend heavily on whether adequate supplies of feeds (manure, grass, some grain) can be made available at suitable prices. xxiv 43. Income Distribution. Recent rural reforms have significantly narrowed urban-rural income differentials, although in 1982 average per capita income in urban areas continued to be about twice that in rural areas. In the 1978-82 period, interpersonal income differentials in the rural areas almost certainly increased as a consequence of policies to improve production incen- tives and foster entrepreneurship. Interregional income differences originate in differences in land and water availability, access to markets for inputs and outputs, and tax, procurement and price policies of the Government. In 1979-82, some of the poorest areas were among the first to introduce the BGDH system and consequently realized the earliest benefits from the reforms in crop production (e.g., Anhui, Sichuan, Guangxi and Guizhou). Some of the better endowed areas (e.g., parts of Jiangsu and Guangdong) benefitted from expanded production from sideline industries and access to free market and export prices. Relative poverty remains characteristic of much of the northwest and parts of the north and southwest - areas with poorer soils, little irrigation and difficult access to modern inputs such as fertilizers or to urban or export markets. Although one element of Government policy is to provide special assistance to the poorer areas, significant results may be difficult to achieve. The Central Government controls directly only about 20% of chemical fertilizer production and current plans for water resource devel- opment continue to emphasize the better endowed provinces. On the other hand, pricing policy for farm products and production inputs has tended to favor the poorer, more isolated areas and criteria for assessing returns to irrigation in such areas are frequently less rigorous than elsewhere. The emphasis on dairy and poultry production around urban centers probably tends to widen interregional income differences, as does rural industrial development which is faster near major urban areas. 44. Policy options which would substantially change these general trends are difficult to identify. The areas of extreme poverty are typically resource poor; investments to substantially alleviate this situation may not be justified on economic grounds alone. In some of the poorest areas, outmigration is certain to be one of the required policy measures. Elsewhere, improved transport and access to markets for agricultural products and inputs are likely to be among the most important measures needed to address the relative poverty issue. But to be effective, these must be buttressed by stronger agricultural research and extension oriented to difficult agro- climatic circumstances. Issues related to inter-regional income distribution are likely to remain important in China. 45. Prices. The budgetary cost of the continuing upward drift of average procurement prices has encouraged changes in the multi-tier pricing system. An approach, currently applied to oilseeds and cotton and soon to be introduced for grain, is a system whereby fixed proportions of farm sales receive quota and above quota prices. Marginal prices received by farmers for additional sales are reduced to a weighted average of quota and above quota prices, and state procurement agencies cease to serve as buyers of last resort. With this approach, incentives to sell to the state and to use pur- chased inputs might decline and budgetary costs would be little affected unless the system were applied to reduce average procurement prices as well. A preferable alternative might be to replace this system with a single market price structure, with government intervention as necessary to maintain price XXV stability. The indirect tax resulting from low procurement prices could be replaced with direct taxes based on the productivity of land and other resour- ces. Procurement agencies would purchase only the quantities needed to meet urban and rural resale requirements and to maintain emergency reserves and price buffer stocks. 46. Some changes in relative prices for specific farm products and inputs seem to be required in order to enhance incentives and encourage effi- ciency. Some of the required changes emerge from the fact that much larger output of, e.g., livestock products, requires new production technologies (e.g., more intensive grain feeding) which are not profitable if grain is priced at current market (or marginal) prices. In pork production, for example, the policy has been to subsidize grain and other costs for producers who use grain intensive systems. A more efficient alternative would be to allow both product and input prices to rise and employ direct taxes as necessary to meet efficiency and equity objectives. 47. To stimulate necessary investment, comprehensive price reforms in agriculture are likely to require price increases for livestock, forestry and fisheries relative to most crops. This selective increase in farm prices might permit an increase in interest rates on agricultural loans and in the rates paid on deposits, thereby helping to mobilize more savings for produc- tive investment. While non-Government funds are likely to be the major source of investment in the sector, an increase in Government resources also will be required. xxvi 1. Performance Since 1979 Introduction 1.01 China's agricultural sector has undergone substantial change since the visit of the World Bank's first Economic Mission in November 1980. The introduction of the "production responsibility system" (PRS), price reforms and other changes in the organization and management of production units and support services have had very positive effects on the performance of the sector in recent years. This chapter of the annex describes these changes ~n some detail and examines agriculture's performance since 1979. The second chapter explores future demands for food and other agricultural products, while the third and fourth chapters look at determinants of future growth and production prospects for major products, respectively. A concluding chapter considers some of the issues and options which face the sector over the next 15-20 years. 1.02 The basic objective of this report is to bring together the material available to the Bank which bears most directly on future development of China's agriculture and, from that base, to discuss development prospects and options. The size and diversity of the agricultural sector makes this a formidable task. The work has been made somewhat easier by extensive discuss- ions with authorities in Beijing and through visits by the Economic Mission to Hubei, Jiangsu and Gansu Provinces during the first four months of 1984. The Bank's ongoing agricultural project work also has deepened our knowledge of the sector. The material which follows should be viewed as part of the World Bank's continuing effort to understand and assist an agricultural economy which provides basic sustenance to more than one billion people and liveli- hoods for some 800 million individuals. The Changing Policy Environment 1.03 Institutional Reforms. Beginning with the experimental abolition of collective farming in the more impoverished areas, the Chinese Government has moved swiftly since 1979 to implement a comprehensive restructuring of rural institutions. The reforms, a major feature of which is implementation of the PRS, were designed to improve incentives and management at all levels of the agricultural economy. By 1983 the farm household had become the fundamental unit of management and production in the agricultural sector, within a frame- work of collective or state ownership of land and major fixed assets. The continuing process of reform is directed at shoring up the new structure, dis- mantling redundant collective institutions, and facilitating a higher degree of commercialization and specialization. - 2 - 1.04 The PRS is based on a system of contractual relationships which define the rights and responsibilities of owners (state, collective, or pri- vate) and managers of assets. The contract system is also applied to define vendor-consumer or supplier-procurer relationships. It is intended to provide scope for increased, albeit constrained, independence of decision making and encourage relationships based on negotiation rather than administrative fiat. A wide variety of contractual forms have been tried, and in agriculture one system has become nearly universal: bao gan dao hu (BGDH). This, loosely translated, means "contracting all actions to the household". 1.05 In BGDH, management of collectively-owned land is contracted to households, usually in proportion to household size or labor force. Other collective assets are divided up, sold or contracted to individuals or groups willing to manage them. The household is obligated to pay taxes, make contri- butions to collective welfare funds, provide its share of state procurement requirements, and contribute labor to maintain or construct public infrastruc- ture. All remaining output may be retained by the household. Food rations and the work point system are abolished. Under this system, teams of 20- 40 households function largely as administrative bodies, make periodic adjust- ments in the distribution of land and other assets, collect taxes, procure production quotas on behalf of the Government, and implement family planning act1v1t1es. Teams may continue to promote the development of service or subsidiary production activities, but this is done usually through subcontrac- ting rather than direct management. 1.06 Some teams have retained less radical forms of the PRS and continue to play a major role in management or distribution. Usually these are teams in areas where extensive indivisible collective assets exist. The state farm system initially tied bonuses to contractual performance of individual workers on fixed basic wages, but has now begun to implement a scheme similar to BGDH. 1.07 As part of the trend toward a reduced role for administrative bodies in economic management, higher level institutions are also undergoing restruc- turing. The commune is being transformed into the township (xiang) with its economic role confined largely to fostering commune and brigade enterprise (CBE) development. The brigade is being renamed the village (cun) and is los- ing much of its managerial and technical staff. The Supply and Marketing Cooperatives are being restored to their initial role as dividend-paying, shareholder-managed institutions owned largely by farm households. At higher levels, Jiangsu Province is experimenting with the integration of cities, towns and surrounding countryside into single administrative units in order to reduce administrative obstacles to commercial development. State farm reforms include the establishment at provincial level of semi-autonomous Agriculture, Industry and Trade Corporations which will be responsible for, inter alia, interprovincial and international trade in products from state farms. These corporations also will be permitted to engage in construction work outside of the state farm system. 1.08 BGDH has provided substantial rewards to those farmers who are able to increase production. Tax, welfare fund contributions, and quota procure- ment obligations have been fixed in absolute terms, rather than on a - 3 - percentage basis, although amounts vary with the productivity of assigned land. The total obligation to the collective and the Government, counting the underpricing of ~locurements, appears to be in the range of 20-30% of gross value of output.- 1.09 A State Council regulation of early 1984 permits the transportation and marketing of most agricultural products by collectives and individual producers after quota and above quota deliveries have been made.~/ After licensing by local commercial bureaus, producers may transport agricultural products, including livestock, aquatic products and processed foods, for sale across county and provincial lines. Similarly, private merchants in urban areas may cross jurisdictional boundaries to purchase eligible products directly from producers. Prices are to be established through negotiations and may vary within the bounds, if any, specified by local or national policy. In the case of inter-county or inter-provincial movements of goods, itineraries of transport vehicles must be approved by authorities at county level or above. 1.10 The shift tn control over resources from collectives to individuals has created new policy issues, among them how to sustain a high rate of investment in the farm sector. Perhaps reflecting uncertainty over the dura- tion of the PRS policies, households were initially inclined to devote their new surpluses to housing construction and the accumulation of consumer durables or commodity stocks. The Government has responded with policy changes designed to encourage private investment, notably by lengthening land contracts to up to 15 years in the case of annual crops and 50 years for tree crops, machinery and equipment. Since the inception of the PRS some 1.2 mil- lion households have purchased pedestrian tractors. 1.11 Part of the new policy environment is to encourage specialization and commercialization, thereby reversing the self-sufficiency orientation of Chinese collective agriculture. This requires better functioning markets in labor, purchased inputs, services and output. These have begun to develop with official sanction. A clear example relates to labor and land. It is widely held that about one-third of the rural labor force is not required in agricultural production with existing technology. To encourage more efficient use of rural labor, recent Government decisions have allowed farmers to hire labor at negotiated wage rates and to surrender or subcontract land use rights in order to engage in off-farm activities. 1/ Assuming that 50-60% of total sales are at quota prices and the "market equilibrium price" is perhaps 50% higher than the quota price. 2/ State Council, "Regulations Issued by the State Council Concerning the Transportation and Sale of Farm and Sideline Produce by Collective Com- mercial Units and Individuals," February 25, 1984 (World Bank translation of Xinhua release of March 8, 1984). - 4 - 1.12 The growth of BGDH and the focus on farm households as production units has required new structures for agricultural extension and research. Here the Government is using new concepts such as the model farmer and is applying the contractual features of the PRS to extension work by experiment- ing with a fee-for-service system to provide technical assistance to producers. The farm household is the new target for support services, and the "specialized household" (SH) is considered to be the major vehicle to demon- strate improved technology. Some 14% of farm households now possess suffi- cient degrees of production specialization and/or commercialization to be classified as SH. The SH at present is something of a hothouse product, nurtured with subsidies and often headed by individuals with some technical tra1n1ng. Government policy reflects an increased awareness of the need to establish an institutional framework for agriculture which provides a "complete service structure supporting commercial production, including such requirements as technical assistance, investment, marketing, storage, I processing, transport, commercial credit, and management assistance."~ Achievements are encouraging, but much remains to be done in each of these areas. 1.13 Rural Reforms and Investment. Spread of the PRS appears to have been accompanied by a contraction in rural investment in production infra- structure. In the past, accumulation by collectives, together with investment in kind by collective labor, represented a major part of agricultural invest- ment. In recent years, in-kind investment by collective labor has become much less important, although this may have been offset to some extent by increased labor investment by households on lands contracted to them. The transfer of control over resources from collectives to farm families has resulted in an increase in private rural savings which now are equal to about 20% of annual income. Until recently, most of these savings had been funneled into housing construction or stock accumulation, possibly reflecting a reluctance to make long-term on-farm investments due to uncertainty about tenure arrangements or the permanence of reforms. In early 1984 the Government reacted to this pro- blem by issuing decrees to guarantee long tenure and to encourage private investment in productive activities. The rapid increase in crop marketing in 1983 suggests that on-farm stock accumulation may have reached its limits. 1.14 The collectives now appear to be handicapped by diminished income from agriculture and growing difficulties in drafting team labor for such works as irrigation, land improvement and local road projects. Compensation for labor on large projects must be paid in cash or kind, at rates which com- pete with the earnings possible from private farming and sideline activi- ties. However, an uncompensated "social labor obligation," perhaps averaging 20 work days per farmer per year, may be tapped to maintain infrastructure, construct small projects with local benefits, and fight natural disasters. Casted at about Y 2 per day, this may amount to as much as Y 12 billion in potential rural gross investment per year. In addition, rural collectives 3/ China Farmer's News 16 (February 1984), p. 1, describing the contents of the 1984 No. 1 Party Directive. - 5 - accounted for Y 13.1 billion of fixed investment in 1982, of which about 60% was derived from the net incomes of CBEs. 1.15 State investment has been adversely affected in recent years by increased budgetary stringency, designed to reduce the central Government deficit. Investment by the state in agriculture appears to have suffered more than some other sectors. The Y 3.0 billion budget in 1982 for fixed capital investment in agriculture (including forestry, water conservancy and meteoro- logy) represented a decline in this component to 9.7% of total Government investment in basic construction, compared to the 11-12% reached at the end of the 1970s (Table 1.1). Actual total investment in agriculture, including extra-budgetary funding by state-owned units at all levels, fell from 10.5% of the total during the Fifth Five-Year Plan to 5.8% in 1983. Much of this was absorbed by major projects receiving state support, e.g., Heilongjiang drain- age and land reclamation, Tianjin municipal water supply, and various projects in Guangdong, some of which appeared to be only indirectly related to agricul- ture. The source of financing of water conservancy projects has shifted away from state budgets to provincial and local funds. In 1983 only about a quar- ter of expenditure for this purpose originated in the state budget. 1.16 The agricultural credit system also may have been used to restrain investment. Available data suggest that increased rural deposits exceeded the net increase in rural lending in recent years (although more rapid loan repay- ment due to good harvests may account for this phenomenon). The portfolio of rural lending institutions also was undergoing restructuring, as individual farmers received a growing proportion of new loans and the financing of state farms shifted increasingly from central government grants to loans by the Agricultural Bank (Appendix Table A.l). 1.17 A rough estimate of the magnitude of rural gross capital formation in 1982 is about Y 65 billion (Table 1.2). Investment by the state appears to have accounted for only 5% of this total, but represents perhaps one-fourth of directly productive investment in agriculture (Y 3.4 billion of Y 14.5 bil- lion, the latter being the difference between the Y 26.5 billion in "other use" and the Y 12 billion for uncompensated labor). About one-half of collec- tive investment was directed to CBE expansion from CBE profits, which also provided more than one-quarter of the financing for other collective invest- ments (excluding "social labor"). The value of net increases in pedestrian tractors, draft animals, and pigs alone, presumably components of private investment, totalled nearly as much as state investment in agriculture. 1.18 The estimates in Table 1.2, notwithstanding their imprecision, suggest that by 1982 only a limited proportion of rural savings was being directed into investments which would contribute directly to agricultural productivity. A sizeable proportion of this investment undoubtedly was used to replace aging infrastructure or was offset by abandonment of collective assets such as large-scale farm machinery. On the other hand, the estimates suggest that private rural savings are large enough to be a major source of finance for rural development, if the savings can be mobilized without under- mining the newly-established production incentives. - 6 - i Table 1.1: AGRICULTURE IN THE STATE BUDGET (Y billion) ~ 1957 1965 1975 1978 1979 1980 1981 1982 Total budgetary allocations 1l 30.4 46.6 82.1 111.1 127.4 121.3 111.5 115.3 Of which: Basic construction investment - ' /c 12.4 15.9 32.7 45.2 51.5 41.9 33.1 30.9 Budgetary allocation for agri- culture, forestry, water conser- vancy and meteorology ~ Basic construction investment 1.1 2.4 3.6 5.1 6.2 4.9 2.5 3.0..6!_ Administrative expenditure 0.9 1.7 3.3 6.1 7.9 7.0 6.5 6.2 Working capital expenditure 0.1 0.3 0.3 0.8 0.5 0.4 0.2 0.2 Subsidies to communes 0.1 0.9 1.6 1.1 1.2 0.9 1.0 Rural relief assistance 0.3 0.7 0.7 0.7 1.0 0.7 0.7 0.9 Total 2.4 5.1 8.9 14.3 16.8 14.2 10.8 11.3 Allocation for basic construc- tion in agriculture as % of: Total budget 3.6 5.0 4.3 4.6 4.9 4.0 2.2 2.6 Total basic construction 8.8 14.8 10.9 11.3 12.1 11.6 7.4 9.7 Budget allocation for agriculture as% of total budget allocation 7.9 10.9 10.8 12.9 13.2 11.7 9.7 9.8 ~ Relates to magnitudes in the unified national development plan and therefore includes those allocations by the central, provincial and local governments which are normally included in the national plan. 1l From 1982 Ch:J.na Statistical Yearbook, p. 449. Refers to budgeted expenditure. ~ From 1982 Yearbook, pp. 452-53 • ..6!_ Actual expenditure, which was probably higher than budgeted expenditure. - 7 - Table 1.2: ESTIMATES OF RURAL GROSS CAPITAL FORMATION, 1982 Amount Percentage (Y billion) (%) By origin: State /a 3.4 5 Collectives .tE... 13.1 20 Private /c 48.5 75 (of which uncompensated labor) (12 .0) (19) Total 65.0 100 By use: Commune & brigade enterprise /d 6.8 10 Private housing /c -- 15.7 24 Stock accumulatiOn /e 16.0 25 Other uses - 26.5 41 Total 65.0 100 ~ From Table 1.1. Includes investment by state-owned enterprises and farms outside the state budgeto .tE... Report in State Statistical Bureau, Outline of Chinese Statistics, 1983 (Beijing, 1983), p.57. ~ The sum of estimated components, including investments in private housing and other buildings (Y 16.9 billion, from source in note b), stock accumu- lation (Y 16 billion, based on difference between income and consumption in kind as reported in rural income surveys), net investments in produc- tive assets except buildings (Y 3.5 billion, net of transfers from cQllec- tive to private ownership). The increased numbers of pedestrian tractors, draft animals, and pigs alone would account for Y3.1 billion, and uncompensated "social labor" (assumed to average 20 workdays per year at· Y 2/day for each of some 300 million farm workers) would have a value of about Y 12 billion. ~ The portion of retained earnings of commune and brigade enterprises which was not distributed or used to support agriculture. Includes Y 4.7 billion reinvested in CBE expansion andY 2.1 billion for "other uses," primarily infrastructure such as roads, bridges, and housing related to CBE production as well as some "administrative expenses". As reported by the CBE Bureau, MAAF. ~ Accumulation of grain stocks and other agricultural commodities, including private working capital. - 8 - 1.19 Price Reforms. The Chinese government in 1979 made major adjust- ments in the level and structure of farm prices because of concern that the existing structure did not provide sufficient income and incentives for the crop mix which the Government wished to encourage. Because these reforms forced sharply higher budgetary costs for consumer subsidies, the Government has indicated its unwillingness to consider further significant increases in the farm price level over the medium term. However, as seen below, this pricing system guarantees an upward drift in the weighted average price received by farmers as long as production and procurement continue to grow. 1.20 Prior to further reforms announced in late 1984, the farm price structure encompassed four distinct prices: (a) quota; (b) above quota; (c) negotiated prices; and (d) free market prices. Quota prices applied to crops sold in fulfillment of procurement quotas; above quota prices were paid for that part of crop sales in excess of quotas. Quantities to be delivered to the state at quota and above quota prices were stipulated for each producer. After fulfilling these two obligations, producers could sell remaining supplies at negotiated or free market prices. Ratios of above quota prices to quota procurement prices, which were 30% for grain prior to 1979, were increased to 50% for grains and oilseeds, 20% for bast fibers, and 30% for cotton (which previously had no above quota price premium). Procurement prices for soybeans were increased 15% in 1979 and 50% in 1981. Quantity norms or obligations for sales at above quota prices have been typically determined for a three-year period, but no penalty is attached to underful- fillment, as currently there is little differential between above-quota and free market prices. The number of agricultural commodities with state- established prices is declining and the pricing structure is being simplified. In Gansu Province, for example, the number of agricultural and sideline prices controlled by the price bureau declined from more than 100 to about 40 by early 1984 (of which 30 were for medicinal herbs). Beginning with the 1984 crop, all cotton in China was to be procured without regard to quota fulfillment under a one-price system in which a weighted average of quota and above-quota prices will apply, with weights varying by region but not over time. It is expected that a similar system would be applied to grains beginning in 1985~ 1.21 Negotiated prices fluctuate at approximate parity with free ~arket prices. They apply to sales to the state by individual farmers for production which may be legally sold on the free market. In 1980, negotiated prices were 10-20% higher than above quota prices, reflecting procurement shortages and high free market prices. Recently, negotiated prices for some grains have fallen in some areas to less than above-quota prices, reflecting bumper har- vests. Overall, in 1979 the average procurement prices (each type of price weighted by the share of sales transacted at that price) rose about 25% for grains, 40% for oilseeds, 17% for cotton and 37% for pigs. 1.22 Procurement quotas for the two major commodity groups, grain and cotton, were last fixed in the early 1970s and have remained substantially unchanged. Quotas for other crops such as oilseeds varied from year to year. If no quota existed for a product, the above quota or>negotiated price normally applied. As a result, increased procurements have generally taken place at higher than average prices and imparted an uptrend to the weighted average price, even with an unchanged price structure. The effect can be - 9 - sizeable. Of the total value of net procurement of agriculture and sideline output in 1983, 48% was at quota prices, 28% at above quota prices, and 13% at negotiated or market prices. Just under 11% represented free market sales to the nonfarm population. The increased volume of sales in 1983 over 1981 at quota prices was only 9%, but was 49-78% for each of the other three categories. As a result, average weighted prices in agriculture appear to have continued to grow in recent years at about 5% p.a (Appendix Table A.2). Reforms announced in late 1984 are expected to effectively abolish procurement quotas for most agricultural commodities. Under the new system Government procurement would be limited largely to contracted amounts needed for urban consumers. Excess production is to be retained by the farmer or marketed by the farmer on the free market. 1.23 Price reforms of recent years were designed to improve levels of living for farmers and to correct incentive imbalances among crops. However, there has been little public information on how these objectives were translated into specific prices. Some inferences can be drawn from post- reform surveys of farm production costs (Table 1.3). It appears that quota prices were set at levels which roughly equalized the net return (wages, rent, profit) per labor day for different crops (with the exception of sugar, where the intent apparently was to provide an incentive for expanded production), with the value of rural consumption per labor day which has been reported at about Y 1.40. This leads to significant differences among crops per sown ha, although these differences are greatly narrowed if cropping systems are con- sidered which would typically occupy land for an entire year (e.g., rice, corn or soybeans combined with wheat or rape, and compared with crops such as cotton, jute, tobacco or sugar cane where only a single crop per year is possible). 1.24 Average quota procurement prices appear to bear a reasonable rela- tionship to historical production costs for the "average" region or situa- tion. However, the new price structure is based on past levels of input use which, in turn, were influenced by the previous price structure. Thus, price adjustments which restored profitability relative to past production costs tend to perpetuate any inefficiencies existing in the previous price struc- ture. Further, average cost-price relationships mask regional variations in subsistence costs, land quality, the extent of multiple cropping, etc. Some local observers regard the new level of quota prices as too low to promote increased production and biased against grain production. However, agricul- ture's good performance since the 1979 price adjustments does not provide support for these assertions. Some anomalies, such as the low return per ha for soybeans (which are single cropped in the main producing region), have been corrected through subsequent price or quota adjustments. Growth and Productivity 1.25 The period since the 1979 reforms has been characterized by excep- tionally good performance of the agricultural sector. The gross value of agricultural output (GVAO) in 1983 grew by 9.5% following growth of 11.1% in 1982, 6.6% in 1981, and 3.9% in 1980. Since 1978, GVAO has grown at an annual average exceeding 7%, or more than twice the rate achieved over the previous two decades. Weather during much of this period was normal, but unusually good growing conditions in 1982 and 1983 made important contributions to high - 10 - Table 1.3: EFFECTS OF PRICE REFORM ON RETURNS TO CROP CULTIVATION ~ Net return per Net return Rate of return Crop labor day .fl!... per ha .LE_ on cost J.!:... Rice 1.56 782* 47 Wheat 1.13 411* 19 Corn 1.31 500* 33 Soybean 1.37 352 39 Cotton 1.34 1,087 40 Rapeseed 1.17 491* 25 Peanuts 1.27 570* 45 Jute 1.39 1,289 42 Tobacco 1.36 1,540 36 Sugarcane 1.98 1,769 66 Sugar beet 3.37 810 106 ~ Computed from survey data covering over 800 production teams and reported in the Handbook of Agrotechnical Economics (Beijing, 1983), pp. 642-57. Data refer to prices and costs which prevailed after the 1979 reforms • .fl!... Net return defined as gross value of main product at quota prices ~ess material costs (net of value of byproduct); that is, rent+ wages +pro- fit. Refers to crop area, not cultivated area. An asterisk designates crops which frequently occupy land in one of two crop seasons per year • .is:... Net returns exclusive of "wages," valued at Y 0.80/labor day (an account- ing wage used in cost surveys since the 1950s); as a percentage of mater- ial costs +"wages." - 11 - growth in those years. During the 1979-83 period, China's cropped area declined by 3.0%. The area sown to grains fell by more than 4%, while the area under non-grains rose by 20%. 1.26 Grain production in 1983 reached a record 387.3 million tons, some 17% higher than in 1979 (Table 1.4). Grain production in 1979-83 increased at an annual average rate of 3.9%, as the steady decline in sown area was more than offset by a remarkable 22% increase in average yields. Factors contri- buting to yield gains included more rational selection of areas planted to grains, improved production incentives and greater use of purchased inputs. The largest production gains have been centered on the fine grains. Compared with 1979, wheat production in 1983 was up by 30%, rice output was some 17% greater, but output of coarse grains increased by only 9%. An important con- tributor to grain production was hybrid rice, which now accounts for about 26% of China's rice production and has a yield advantage over conventional varie- ties of 15-20%. 1.27 Oilseed production in 1983 (excluding soybeans and cottonseed) totalled 10.6 million tons, down 10.7% from 1982 but 66% more than production in 1979. In contrast to grains, the area planted to oilseeds increased stea- dily until recently, with most of the increase coming in rapeseed. The area planted to rapeseed was cut in 1983 in response to market surpluses and state prices were reduced slightly. Good harvests permitted steady gains in per capita availabilities of edible oils to about 4 kg per annum, thereby helping to alleviate scarcities of this important food item. Soybean production in 1979-83 varied markedly from year to year in response to weather conditions, but trended upward. Imports of soybeans for crushing declined while soybean exports for food use have moved up slightly. Production in 1983 was 9.8 mil- lion tons, up 7.7% from the 1982 crop. 1.28 China's cotton crop more than doubled in 1979-83, reflecting both higher yields (up 56%) and larger planted area (up 36%). The 1983 crop was up more than 28% from the 1982 level. The rapid growth in production is due in large part to specialization incentives and the introduction of new varieties. These incentives are encouraging expansion in the traditional areas well- suited to production in North China, while the planted area declined in less well-adapted areas where production had been encouraged by previous policies. The planted area in 1983 in Shandong Province, the largest producer, was 40% greater than it was in 1981, while the cotton area in Sichuan Province was down about 50% from 1981 levels. Cotton imports continued to decline and in 1983 were only about 10% of levels in the peak years of 1979 and 1980. Stocks of cotton in early 1984 were about 1.6 million tons, equivalent to nearly half a year's consumption requirements, and export markets were being sought. Rationing of cotton textiles ended in 1983 in conjunction with a price adjust- ment (cotton up 20%; synthetic textile prices down 30%) designed to balance supply and demand and pass on increases in raw material prices. 1.29 Sugar production increased by 52% in 1978-82, in response to attrac- tive price incentives. Despite these gains, sugar imports continue to increase rapidly as part of the Government's efforts to improve per capita consumption levels of key commodities. Annual per capita consumption of sugar of 4.5 kg remains low by international standards (e.g. 20 kg in Thailand). In the 1979-83 period, tobacco production rose by 43%, tea by 45%, and silk cocoons by 26%. - 1Z- Table 1.4: PRODUCTION CHANGES, 1979-83 % annual Percent growth change rates Production Unit 1979 1983 1983/1979 1983/1979 Rice (Padd~) Sown area mln ha 33.9 33.1 -2 Yield tons/ha 4.3 5.1 +20 Output mln tons 143.8 168.9 +17 4.1 Wheat Sown area mln ha 29.4 29.1 -1 Yield tons/ha 2.1 2.8 +31 Output mln tons 62.7 81.4 +30 6.7 Corn --sown area mln ha 20.1 18.8 -6 Yield tons/ha 3.0 3.6 +20 Output mln tons 60.0 68.2 +14 3.3 Others~ Sown area mln ha 35.9 33.0 -8 Yield tons/ha 1.8 2.1 +15 Output mln tons 65.6 68.8 +5 1.2 Total Grain Sown area mln ha 119.3 114.0 -4 Yield ton/ha 2.8 3.4 +22 Output mln tons 332.1 387.3 +17 3.9 Cotton Sown area mln ha 4.5 6.1 +36 Yield kg/ha 88 763 +56 Output mln tons 2.2 4.6 +109 20.2 (ginned weight) Oilseeds /b Sown area mln ha 7.1 8.4 +18 Yield kg/ha 913 1,260 +38 Output mln tons 6.4 10.6 +66 13.4 Sugar (refined) mln tons 2.5 3.8 +52 11.0 Tea '000 tons 277 401 +45 9.7 Meat /c mln tons 10.6 14.0 +32 7.2 Tobac~ (cured) '000 tons 806 1,151 +43 9.3 1.!. Soybeans, pulses, sorghum, millet, barley, oats, tubers, miscellaneous grains • .!l!.. Excludes soybeans and cottonseed. 1.£ Pork, beef, and mutton only. Source: State Statistical Bureau. - 13 - 1.30 Meat production (pork, beef, mutton) in 1983 was 32% greater than in 1979, with most of the gains coming in pork production. Per capita availabi- lity of meat in 1983 reached 13.7 kg. Cattle numbers have increased sharply in recent years, partly in response to the PRS which now permits farm house- holds to own draft animals. The PRS in the grazing areas also is providing strong incentives to increase animal numbers, but thereby exacerbating the problem of overgrazing. Milk production is in the range of 1.8 million tons annually, but consumption remains low by international standards (about 1.7 kg per capita in China; 10 kg in Malaysia; 103 kg in the United States). The Government has just begun to collect statistics on poultry production. House- hold surveys suggest that per capita egg consumption has grown by nearly 50% since 1978. Modern enterprises for the production of eggs and poultry meat are being established near some of the larger coastal cities and are providing larger shares of total broiler and egg production. 1.31 Output of aquatic products in 1983 totalled nearly 5.5 million tons, of which the marine catch was 3.6 million tons and fresh water production 1.8 million tons. Output of fisheries products, after a decline in the late 1970s, has begun to expand as rapidly as most other subsectors of agricul- ture. Total fisheries production in 1983 was about 27% greater than the 1979 level. Sluggish growth of salt water fisheries is attributed to over-fishing of the marine resource. In contrast, fresh water fish culture has grown by 64% since 1979 and now accounts for about a quarter of total production. 1.32 Output from sideline activities (gross value of hunting and gathering and industries run by brigades and teams) increased by 110% since 1979 and 20% in 1983 alone. Sample surveys of households indicated that in 1983, 36% of farmers' net incomes came from private household activities. The comparable figure for 1978 was 27%. Output of commune-run industries (which are not included in GVAO) in 1983 was 87% higher than in 1978. Appendix A, Table A.3 notes the continued decline in the relative importance of crop pro- duction, the marked increase in sideline production and more modest relative gains in livestock. Chinese agriculture continues to be heavily crop orien- ted, with a crop:livestock ratio of about 92:8 (net value added) as compared with 52:48 in the United States and 83:17 in India. Incomes and Consumption 1.33 Recently published household survey data provide some measure of the extent of income and consumption gains in rural China (Table 1.5). In nominal terms, per capita incomes of farmers in 1983 were 2.3 times those of 1979. Incomes of employees in state enterprises (mostly urban dwellers) rose by 66% during this period and were nominally 70% higher than rural incomes in 1983. In real terms, rural incomes may have increased by 70% while urban incomes were up by about 40%; however, the urban/rural income ratio in real terms is still well over 2:1. Consumption gains have been substantial. Between 1979 and 1983 annual per capita consumption of grain in rural China increased by only 1% while per capita consumption of vegetable oils rose by more than 48%. Consumption of red meat (excluding poultry) on a per head basis was up by more than 53%. Consumption patterns in China's rural areas continue to differ markedly from those in urban areas with rural residents consuming - 14 - Table 1.5: COMPOSITION OF RURAL INCOME AND EXPENDITTURE, 1979-83 Percent increase, 1979 1983 1979-83 Rural net income (Y/capita) 160.2 309.8 93 Composition (%): Collective income 64 55 Private income 27 36 of which from: Crops 10 12/a Livestock 12 13Ta Other private 6 13Ta Remittances & other 9 9- Rural consumption expenditure (Y/capita) 134.5 248.3 85 Composition (%): Food 64 59 Clothing 13 11 Fuel 6 5 Housing 6 11 Other commodities 8 11 Services 3 2 Consumption of (kg/capita): Grain (unprocessed) 256.5 260.0 1 Edible vegetable oils 2.4 3.5 46 Red meat 6.5 10.0 54 Poultry 0.3 0.8 159 Fish and shrimp 0.7 1.6 128 ~ 1982 figures; 1983 not yet available. Source: State Statistical Bureau. Commodity income in 1979 valued at 1978 quota-procurement prices; in 1983 at 1983 procurement prices. - 15 - substantially more grain and less meat on a per capita basis, but the patterns have been converging. 1.34 It would be natural to expect that the PRS and associated reforms provided the incentives to accelerate agricultural growth but, in doing ~o, may have increased the inequality of income distribution in the sector.~/ The rationale behind this expectation is that ability and ambition are unequally distributed in the population and the earlier rural distribution system rewarded neither. Indeed, it put a "cap" on distributed income whenever abil- ity, effort and access to unusually productive agricultural resources would have led naturally to increased income differentials. Initial analysis of data through 1982 on the distribution of rural income suggests that the PRS in fact may not have increased the concentration of rural incomes (Appendix Table A.4). Rural income surveys suggest that between 1979 and 1982 poorer house- holds shared more than equally in the benefits of the PRs.21 A possible explanation for these results is that the PRS, and in particular the BGDH ver- sion, was extended first to the poorest teams and most backward areas which therefore realized the earliest income benefits from the reforms. Only in 1982 and 1983 did the wealthiest teams and areas adopt BGDH. Improved incen- tives for growing cotton in the relatively poor North China Plain and sugar- cane in the poorer areas of South China may also have narrowed regional dif- ferences. Moreover, BGDH was initially applied to crop production, and was of greatest benefit to families or areas with few non-crop activities. Recent reforms (1983-84) promote development of commercial sideline and CBEs, and will be of most benefit in wealthy areas. In short, the geographical reduc- tion in inequality through 1982 may have "swamped" the effect of reforms on inter-family income differentials. On this rg,soning, the full effects of BGDH on the richer areas have yet to be felt.- 4/ Egalitarianism has been officially discredited. A recent article in the Beijing Review stated: "For more than two decades after the cooperative movement, China has tried to bridge the income gaps between the peasants with large-scale egalitarian collective management and egalitarian methods of distribution. By arbitrarily equalizing peasants' income, these methods only crippled the peasants' enthusiasm for work and held back the growth of agricultural production which even came to a standstill in some places". (September 19, 1983). 5/ The Gini coefficient estimated from these data declined from 0.257 in 1979 to 0.225 in 1982. 6/ A factor complicating the interpretation of these results is that size of the sample surveys more than doubled between 1978 and 1982 and composi- tion also may have changed. Other statistical phenomena which would influence the computed coefficients include: (a) family divisions which would increase the weight of particular income groups seem more likely to occur among middle income families, rather than the poorest (who lack savings) or the richest groups (comprised mostly of young, nuclear families); and (b) migration out of the rural areas - and the rural statistics - may be highest among the poorest groups. - 16 - 1.35 While the full impact of rural reforms on income growth and distri- bution remains to be seen, the rapid income gains of all groups clearly reduced absolute poverty. Various estimates document the decline in absolute poverty.l/ In the 1977-79 period, about 200 counties were designated as poor. By 1982 only 42 counties were classified as in poverty. Measured somewhat differently, Appendix Table A.4 shows that families with yearly per capita incomes of less th1n Y 100 dropped from about 20% of the total in 1979 to less than 3% in 1982.~ 2. FUTURE DEMANDS ON THE SECTOR 2.01 Consumption of agricultural products in China will be strongly influenced by future trends in growth of per capita income and population. At the same time, consumption will be affected by policy measures intended to 'balance supply and demand. At any level of per capita income and population size, different price structures, rationing, or foreign trade patterns can be used to achieve balance at quite different levels and structures of food con- sumption. The projections of consumption made below are of potential demand in the particular situation where relative commodity prices are similar to those 1n other Asian countries at comparable income levels. 2.02 Peculiarities of the present Chinese consumption pattern, involving relatively high per capita levels of direct grain consumption, suggest that higher incomes would bring about a substantial re~7ructuring of the Chinese food consumption pattern over the coming decades.- The directions and magni- tude of this restructuring are accepted in China as long-term economic goals. Broadly speaking, it involves a substantial reduction in direct per capita consumption of grain and corresponding increase in consumption of 7/ A commune (township) is officially designated as poor if its per capita distributed collective income for three successive years is no higher than Y 50. Since the standard is not adjusted for price changes, use in intertemporal comparisons is misleading. 8/ Part of this change may reflect changes in accounting practices. In poor teams or areas, most of the distributed income is in kind. Official sta- tistics on distributed income are very sensitive to changes in accounting prices. In 1981-82 distributed commodities were revalued upward in income surveys by about 30%. Presumably, officially-reported distributed income was similarly revalued. This may suggest that a substantial num- ber of previously poor teams were pushed above the poverty line artifi- cially by this change in accounting prices. 9/ Direct consumption of grain refers to consumption by humans of processed rice, wheat, corn, other minor grains and tubers (at one-fifth fresh -·weight). Consumption of soybeans and other pulses is excluded from this definition. Indirect consumption of grains would be grain consumed by livestock and poultry, the products of which are then eaten by man. - 17 - animal products, along with other forms of dietary diversification. This restructuring of consumption would require changes in the composition of domestic production, including its regional dimensions, and perhaps increased foreign trade in agricultural products. Present Consumption Levels 2.03 After severe reversals and famine in 1960-61, China has made substantial progress in improving average per capita levels of available energy and protein. By 1982 the average Chinese consumed (directly) about 205 kg of cereals, 12.8 kg of red meat (pork, mutton, beef) and 4.2 kg of vegetable oil. This average diet on a daily basis provided over 2,700 kCal of food energy, 68 g of protein, and 38 g of fat (Appendix Table A.6: Food Balance Sheet, 1982). These availabilities exceed estimated requirements of food energy by 24% and of safe minimum levels of protein intake by 87%. Declines in per capita availabilities were registered in the late 1950s, the 1966-69 period, and again in the early 1970s. Generally rapid increases have occurred since the 1978 reforms. 2.04 These national average figures do not mean that nutritional levels throughout China are fully satisfactory. Interregional differences in per capita nutrient availability exist, although they are not well documented, and malnutrition is undoubtedly present in some localities from time to time. But China's food procurement and distribution system has been successful in provi- ding basic food nutrients to most of the population most of the time. Because interprovincial shipments of food are relatively small, data on provincial meat and grain production provide crude measures of differences in per capita nutrient availabilities. These data suggest that in 1981 per capita produc- tion of grain varied from 202 kg (Guizhou) to 420 kg (Jiangsu), while in 1980 average per capita availability of protein ranged from about 38 g (Guizhou) to more than 100 g (Heilongjiang). Average per capita meat production in 1981-82 ranged from 19.3 kg in Sichuan to 5.4 kg in Ningxia. 2.05 Survey data for 1979 suggested that in most relatively high-income locations such as urban Beijing, there was virtually no malnutrition among children. Rural-urban differences were substantial, although somewhat less so in weight-for-age than in height-for-age. The data indicated that more than a third of 7-year old boys in Sichuan were stunted, as compared with less than 4% in Tianjin. The rural-urban gap in per capita energy and protein levels may have narrowed by 1982. Urban residents generally have access to more animal protein, edible oils, fruits and vegetables, whereas rural residents consume considerably more grain per capita. 2.06 At the national level, average per capita consumption levels markedly exceed those of other low income countries, but in China an unusually high proportion of calories comes from grain. Average availability of energy in China on a daily per capita basis in 1980-82 (2,580 kCal) exceeds the 1975- 17 average for middle income countries (2,560 kCal) and the world (2,570 kCal). Protein availability of 66 g exceeds the average for all developing countries (57 g) and approaches the world average (69 g). However, in China only about 10% of protein intake is from animal sources, as compared to a world average of 35% and 21% for developing countries. Similarly, the - 18 - availability of animal and vegetable fat in 1980-82 was 5% below the average in developing countries and more than 40% under the world average. Direct per capita consumption of grain (about 209 kg p.a. in 1980-82) is among the world's highest and exceeds that in India by 60% and in Indonesia by more than 30%. In summary, by consuming relatively more grain and less animal products per capita, China achieves a diet substantially higher in energy and total protein content than that of countries with comparable income levels (Appendix Table A.7: Per Capita Food and Nutrient Availability in China and Other Countries). Projections of Consumption to 2000 2.07 Assumptions. The projections of consumption discussed below assume that relative commodity prices would be similar to levels elsewhere in Asia with sizeable Chinese populations, and examine the implications of changes in real per capita incomes for future consumption patterns. Three scenarios are examined: (a) a "target growth" case, wherein present Government targets for increased per capita GNP and population growth are achieved (requiring an average annual growth of 5.5% in per capita GNP and 1% in population); (b) a "low growth" case, where annual per capita GNP growth is only 3.5% and the Government's population growth target of 1% p.a. is met; and (c) a "high growth" case, in which the Government's targeted per capita GNP growth is com- bined with population reaching 1.3 billion in the year 2000 (1.4% p.a. growth in 1980-2000). 2.08 Ideally, future changes in consumption resulting from these proj- ected growth rates of per capita income could be estimated from income elasti- cities of demand derived from Chinese household consumption data. At present, the only data available permitting such estimates are for Beijing con- sumers.!Q/ Analysis of this data indicates consumer behavior similar to that in other countries, viz., as per capita incomes increase from relatively low levels, preferences move sharply toward animal products, fruits, vegetable oils and nonfood consumer goods, and away from direct consumption of grain. However, the narrowness of the statistical base, its urban orientation, and possible distortions arising from rationing make these data unsuitable for projecting demand at the national level. Moreover, these data are inappro- priate as the basis for projecting future demand because future incomes are expected to substantially exceed the range observed in Beijing today. 2.09 In view of the difficulties referred to above, the approach employed here assumes that 1consumption patterns of Chinese populations outside of the Chinese mainland at higher per capita income levels would approximate those which might be reached in China by the year 2000. For these populations the income elasticity of total grain consumption appears to have been about 0.25 over a considerable period of income growth. Applied to 1980-82 base year per capita total consumption of grain in China, this elasticity produces estimates 10/ Basic data are from the Beijing City Offices of Statistics, "Year-end Data from the Survey on the Livelihood of Working Families, 1982;" (Beijing, 1983). - 19 - of total per capita grain consumption in the year 2000 of/about 390 or 430 kg, if per capita GNP growth is 3.5% or 5.5%, respectively.!! The analogue Chinese population had roughly similar per capita consumption levels in 1970- 72 and 1976-78 respectively. The pattern of direct consumption of a selected number of commodities at these times is examined in Table 2.1. 2.10 Per Capita Consumption. In either the target growth or low growth scenario, little or no change from average 1980-82 levels would occur in per capita consumption of energy and protein.~/ However, compared with 1980-82, consumption of fat would roughly double, and the proportion of protein derived from animal sources would more than triple. The difference between the high and low income variants in per capita consumption in 2000 of various food items is less significant than the radical adjustment from the present dietary pattern to a more varied diet, in which meat, eggs, fish, milk, pulses, fruits, sugar and vegetable oil are larger components. Direct consumption of cereals and tubers would be expected to decline by 15-20% overall, reflecting a sharp decline in direct consumption of coarse grains and tubers and a slight increase in consumption of fine grains (rice and wheat) in the low growth case or during a transition period in the target growth case (Table 2.1). 2.11 Within the fine grains category, there is a question of how per capita consumption of wheat and rice in China will change in response to higher incomes. In developing countries, higher incomes typically lead to a significant shift toward wheat consumption relative to rice. In India, for 11/ Projections of total grain consumption per capita are: Assumed Growth in per Total grain annual growth capita grain consumption in GNP/capita requirements per capita Analogue 1980-2000 1980-2000 /a in 2000 /b years /c -------------- (%) ---------------- (kg)-- 3.5 18.7 394 1970-72 5.5 30.9 434 1976/78 /a Based on an income elasticity of demand for total grain of 0.25. 7b In terms of unprocessed grain; excludes tubers and pulses. The com- parable figure in 1980-82 was about 332 kg/capita. /c Years in which analogue Chinese populations had similar levels of total per capita grain consumption. 12/ Compared to SSB estimates of daily nutrient intake in 1982, the high income growth projection would involve equal per capita energy intake and a slight decrease in protein intake. SSB, Statistical Yearbook of Chinaz 1983, p. 509. - 20 - Table 2.1: PROJECTED DEMAND AND CONSUMPTION TARGETS FOR SELECTED AGRICULTURAL PRODUCTS, 2000 (kg per capita per year) Consumption in 2000 with per capita GNP growth of:/a CAAS /b 1980-82 3.5% 5.5% - 2000- Crops Cereals /c 193.0 164.8 153.1 132.0 Fine grain 146.0 160.1 144.8 Coarse grain 47.0 4.7 8.3 Tubers /d 16.0 4.1 2.1 7.2 Pulses /e 11.3 15.3 15.5 18.0 Sugar 4.1 12.4 19.4 6.0 Fruit 5.7 44.2 58.0 48.0 Vegetables 92.0 89.1 118.6 120.0 Vegetable oil 3.2 5.7 7.0 6.0 Cotton /f 3.4 4.1 4.5 3 .1.1...£ Synthetics 1!.. 0.8 1.3 1. 8 Animal Products Red meat 12.7 19.8 24.1 24.0 Pork 12.0 19.2 22.8 Beef 0.3 0.5 1.1 Mutton 0.5 0.1 0.1 Poultry 0.9 6.5 10.3 Eggs 2.1 4.3 6.6 12.0 Fish 4.5 34.6 35.6 6.0 Milk 1.5 14.4 27.6 24.0 Grain Requirements iJl 340 404 445 440 ~ Consumption patterns of analogue Chinese population in 1970-72 and 1976-78 respectively. Ll Figures represent the most optimistic of targets con~dered by MAAF. From Wu Daxin, "Briefing on Agriculture of the PRC", in E. M. Reisch ed., Agri- cultura Sinica (Berlin: Duncker and Humboldt, 1982). ~ In terms of milled rice, whole grain wheat, maize and other miscellaneous coarse grains. /d Tubers at one fifth wet weight. ~ Pulses include soybeans, soybean products, peanuts and other pulses. /f Estimates from the income elasticities of demand projections, based on the Beijing consumption survey. 1Ji Derived from the CAAS production target of 3.75 million tons of cotton, assuming a population of 1.196 billion. Cotton imports of 0.74 million tons (i.e., the average annual imports for 1979-81) have been included to facili- tate comparison with the figures for cotton consumption 1979-81. /h On Chinese definitions, viz. unprocessed weight, including tubers at one fifth wet weight, soybeans and pulses other than peanuts. - 21 - example, wheat's share in total grain consumption rose from about 20% in 1960 to 30% in 1980. In Indonesia, which grows no wheat, the share of consumption accounted for by wheat rose from about 1% to 5% during that period. In the Chinese analogue population, per capita consumption of wheat increased very rapidly until recently, when consumption of both rice and wheat declined. In Japan per capita consumption of milled rice fell from 118 kg to 76 kg p.a. in the 1962-82 period. It is likely that per capita consumption of wheat, in China's cities in particular, will rise steadily in response to higher incomes and preferences for more convenience foods. 2.12 These projections of per capita consumption are compared in Table 2.1 with a projection by CAAS (the most optimistic of three sets of pro- jections for the year 2000). The CAAS projection is based on total grain supply (unprocessed weight) of 440 kg/capita (including tubers, soybeans and pulses other than peanuts). Even though the CAAS number for total grain supply available for direct and indirect consumption is only slightly below our high estimate of consumption, the CAAS estimates assume smaller direct consumption of grain as well as less fat and animal protein. One result is 14% less energy in the average diet. In part this is explained by CAAS' assumptions of less efficient conversion of feedgrain into animal products and rapid growth of nonfood requirements. Equally important, the CAAS estimates derive from international standards of qutritional "requirements", rather than the preferences of Chinese consumers.ll 1 Judging by the consumption patterns of the analogue Chinese population, it is unlikely, ceteris paribus, that Chinese consumers at GNP levels of US$800 per capita would be satisfied with direct consumption of grain 30% below present levels if this would result in the significant decline in total food energy intake implied by the CAAS figures. 2.13 Feed Requirements. The estimated per capita consumption demands, in combination with the alternative population growth estimates, are used to estimate total supply requirements in the year 2000. In addition to direct consumption requirements, it is necessary to add indirect consumption via conversion of grain to animal products, as well as industrial and other demands. The amount of indirect grain consumption depends heavily on the 13/ The CAAS targets are reportedly based on the concept of "objective" nutritional needs, taking into account resource availability and produc- tion efficiency. Underlying the CAAS targets is a daily nutritional standard for an average adult of 2,400 kCal and 50 g of protein, below nutrient availabilities already attained in China. - 22 - efficiency of production in the animal husbandry and food industry sectors. 14 / At one extreme, one could suppose that production efficiency would be as high as that of the analogue Chinese economy used here to estimate direct consump- tion requirements. In that case, the per capita direct and indirect require- ments shown in Footnote 11 could be used to estimate total needs. 2.14 Alternatively, we could examine the effects on grain requirements of different efficiencies in livestock feeding as measured by FCRs. Little detailed information is available on FCRs in China and available data must be interpreted cautiously. In these circumstances, a reasonable approach to estimating additional concentrate required for additional meat production is to employ alternative FCRs, representing modern, efficient production systems on the one hand (low FCRs) and less efficient production practices (high FCRs) 14/ The largest additions to the per capita requirements for direct grain consumption are the agricultural products used in feeding of animals. The major components of animal feeds, together defined as concentrates, include feedgrains, by-products of grain milling and high protein meal or cake from oilseeds. The efficiency with which concentrates are converted into meat, milk and eggs can be measured crudely by feed conversion ratios, defined in this analysis as the kilograms of concentrate per kg of animal product (dressed carcass weight, eggs, etc.), with dressed weight assumed to be 72% of liveweight in the case of pigs. FCRs depend on: (a) the amount and quality of concentrate and, in particular, the amount and quality of protein meal; (b) the genetic potential of the animals being fed; and (c) general standards of animal husbandry in these production systems and, in particular, the quality of disease control measures. Little detailed information is available on FCRs in China and available data must be interpreted cautiously. Information does not always distinguish live weight from dressed weight, and marginal from average quantities, nor clearly specify whether, and how much, noncon- centrates are being fed. - 23 - on the other. 15 / Neither necessarily reflects China's existing household production systems, which typically minimize grain and protein meal use, but may apply to specialized households or state enterprises which together are expected to supply a large part of incremental production (Chapter 4). 2.15 The implications of alternative feeding systems for concentrate demand in the year 2000 are explored in Table 2.2. The contrasts in grain requirements between high and low efficiency systems of livestock production are striking. Under the high growth scenario the difference between added feedgrain requirements under efficient versus inefficient livestock production systems would be about 89 m tons of feedgrain. However, an efficient live- stock system would itself require more than 30 million tons of high protein meal, as contrasted with about four million tons used annually in livestock feeding in recent years. Despite the imprecision of these estimates, they clearly indicate that high future demand for livestock products would require correspondingly large increments of feed grains and protein meals for live- stock feeding. Furthermore, increased protein meal availability could lead to improved FCRs and large potential savings of feedgrains. 2.16 Total Reguirements. Grain requirements for direct and indirect consumption may be added to projected requirements of other products to pro- vide indicative estimates of total agricultural requirements in the future. These requirements for the year 2000 are shown (Table 2.3) under the most optimistic assumption of efficient animal production systems (FCRs as defined in footnote 15), and with the three income and population scenarios defined in para. 2.07. If the Government's income and population growth targets were met, total grain supply in 2000 would have to increase by about 50% from 1980-82 levels (2.2% p.a.) and more than a third of supply would be committed to livestock feeding (as contrasted with perhaps 10% at present). Supply of red meat would have to grow at 4.6% p.a., poultry at 14% and eggs at 7%. Very high growth rates would also be registered for sugar, fruit, fish and milk. A 15/ This analysis assumes an FCR (grain to meat) in pork production of 8:1 as the lower efficiency bound characterizing traditional systems and 5.1:1 as the higher efficiency FCR indicative of modern production systems. The 8:1 FCR is considered typical of traditional pork production systems in developing countries; the 5.1:1 figures represents a high standard of feeding efficiency somewhere between average commercial results (5.3:1) and "best practice" in the United States (4.1:1). An FCR of about 6:1 is reported for Japan's pork industry. Alternative FCRs assumed for incre- mental production of poultry meat are 5:1 and 2.9:1. For incremental production of beef and eggs only modern system FCRs are assumed viz., 10:1 and 3.5:1, respectively. The FCR assumed for milk (converted to powder at one-tenth fresh weight) is 5.0:1 (modern) and 10:1 (tradi- tional). On these assumptions, a "composite modern system FCR" for the year 2000, estimated by weighting FCRs for individual livestock product shares by projected shares of livestock products consumed in 2000 under the low GNP/capita growth assumptions (3.5%), would be about 4.2:1. Similarly estimated, the "composite traditional system FCR" in 2000 would be about 6 • 5 : 1. - 24 - Table 2.2: ESTIMATED REQUIREMENTS IN THE YEAR 2000 FOR CONCENTRATES AND FEED GRAINS (Million metric tons except where noted) Requirements under assumed feed conversion rates in feedins szstems which are: Modern Less efficient Average Per capita GNP growth Per capita GNP growth 1980-82 3.5% 5.5% 3.5% 5.5% Output Requirements of Livestock Products Pork 12.0 23.7 28.1 23.7 28.1 Mutton 0.5 0.1 0.1 0.1 0.1 Beef 0.3 0.6 1.4 0.6 1.4 Subtotal (red meats) 12.7 24.5 29.7 24.5 29.7 Poultry meat 1. 0 8.2 12.9 8.2 12.9 Subtotal (all meats) 13.7 32.6 42.6 32.6 42.6 Eggs 2.3 5.4 8.3 5.4 8.3 Milk 1. 6 18.2 34.7 18.2 34.7 Concentrate requirement 65 169 231 226 320 Of which: high protein meal 4 22 33 Availability of grain milling byproducts and oilseed cake 32 45 49 45 49 Feed grain requirement 33 124 183 182 271 Total cereal requirement 288 431 473 494 571 Feed grain as a % of cereal production 12 29 39 37 47 Notes: An average population growth of 1% p.a. is assumed for 1980-2000. Figures include a waste coefficient of 3% for pork, beef and mutton, and 5% for poultry, eggs and milk. Concentrate requirements for 1980-82 calculated as 10% of total grain pro- duction plus grain milling byproducts and oilseed cake. (Only 40% of oilseed assumed available as animal feed.) Concentrate requirements for the year 2000 are estimated as the sum of 1980-82 concentrate requirements plus incremental require- ments calculated as the product of incremental meat production and meat conversion rates. Feed conversion rates are as defined in Footnote 15. Seventy percent of oilseed cake production is assumed available as animal feed in the year 2000. Total cereal requirements are the sum of rice, wheat and coarse grain for direct human consumption and indirect consumption as animal feed. Figures are expressed in unprocessed form and include seed and manufacturing requirements plus an allow- ance for waste. Table 2.3: NATIONAL SUPPLY REQUIREMENTS IN 2000 AND COMPARISONS WITH HISTORIC RATES OF PRODUCTION GROWTH Actual average 1980/82 Target growth scenario Low growth scenario High growth scenario CAAS target annual growth, (actual) (Inc. 5.5%i PoE• 1.0%) (Inc. 3.5%i PoE· 1.0%) (Inc. '>.5%i PoE• 1.4%) for 2000 1955-57 (mil t) mil t % growth mil t % growth mil t % growth mil t % growth to 1980-82 .a. • a. .a. .a • CroE Product Requirements Grain /a Cereal /b 291.0 473.0 2.6 431.0 2.1 509.0 3.0 2.7 Fine grain 208.0 259.0 1.? 287.0 1.7 271l.O 1.5 1.8 Coarse grain i.E_ 83.0 215.0 5 .l 144.0 2.9 232.0 5.6 ( Feedgrain) (33 .0) (183.0) (9.4) (124.0) (7 .2) (198.0) (9.9) (293.0) (12.2) Tubers /e 26.0 3.0 -10.7 6.0 -7.4 3.0 -10.7 0.9 Pulses /f 15.0 22.0 2.0 22.0 2.0 23.0 2.3 1.7 Subtotal, Grain 332.0 498.0 2.2 459.0 1.7 536.0 2.6 530.0 2.5 1..:1. Sugar i..f... 2.8 23.7 11.9 15.1 9.3 25.4 12.3 7.2 5.1 6.4 Fruit 6.8 77 .o 13.6 58.7 12.0 82.6 14.0 57.6 11.9 3.4 Vegetables 101.0 157.6 2.4 118.4 0.8 168.9 2.7 144.0 1.9 Vegetable oil .!1!_ 3.6 9.1 5.0 7.5 3.9 9.8 5.4 8.0 4.3 3.0 Cotton 3.5 5.9 2.8 5.3 2.2 6.3 3.1 3.8 0.4 3.4 Synthetic fibers 0.8f.l.._ 2.3 5.7 1.7 4.0 2.5 6.2 N VI Animal Product Reguirements Pork 12.0 28.1 4.6 23.7 3.6 30.1 5.0 Beef 0.3 1.4 8.4 0.6 3.7 1.5 8.8 Mutton 0.5 0.1 -8.1 0.1 -8.1 0.2 -4.7 Subtotal, Red Meat 12.7 29.7 4.6 24.5 3.5 31.8 4.9 31.4 4.9 4.7 Poultry 1.0 12.7 14.3 8.0 11.6 13.6 14.7 Subtotal,· All Meat 13.7 42.4 6.1 32.5 4.7 45.4 6.5 Eggs 2.3 8.3 7.0 5.4 4.6 8.9 7.4 14.4 10.1 Fish 4.8 44.8 12.5 43.6 12.3 48.0 12.9 8.5 3.1 3.8 Milk 1.6 34.7 17.6 18.2 13.7 37.1 18.0 28.8 16.4 /a Total grain requirements correspond to China's "total grain production" figures which include fine and coarse grains, tubers, soybeans and pulses. 71) Cereals include rice, wheat, corn and other coarse grains, but exclude tubers and pulse production. ~ Wheat and rice. The 1955-57 - 1980-82 growth rate is for rice. ~ Corn, sorghum, millet and other minor grains. ~ At one-fifth fresh wet. ~ Soybeans and other pulses. The historic growth rate is for soybeans. i..f... Historic growth rate is for sugarcane. /h Historic growth rate is for oilseeds. /i Figures are for average availability, 1977-81. - 26 - lower income growth rate would lead to more modest changes in grain require- ments (1.7% per year} and use of grain as feed (to 27% of total supply), but high growth rates would still be required for sugar, fruit, poultry, fish and milk to reach a pattern of consumption commensurate with consumer demands at higher per c~pita income levels. 2.17 The assumption of both high income (5.5% p.a.) and high population (1.4% p.a.} growth rates adds some 38 million tons to total grain requirements over and above the Government target scenario, and raises the necessary growth rate of supply to 2.6% p.a. Feed grain requirements would be about 198 mil- lion tons. This scenario approximates the CAAS "optimistic" estimates for total grain and red meat supply (although the latter presumes a lower popula- tion growth rate). However, the CAAS estimate assumes a doubling in the number of draft animals, a substantial increase in industrial food demand, and less efficient feed conversion rates, so that more grain is required for feed and nonfood uses and less remains for direct human consumption. 2.18 Each of the three projections based on consumption patterns of the analogue Chinese population reflects much higher consumption levels of sugar, fruit and fish than might reasonably be expected in China (as is apparent from the CAAS projections). These food items account for substantial fractions of total energy and protein supply. To maintain the same energy and protein levels in China with much lower consumption of these products, direct or indirect grain consumption would have to be increased relative to the projected levels. 2.19 Grain requirements above those projected in these scenarios also would result if China were not able to increase the efficiency of feed conver- sion as livestock production expands. As noted in Table 2.3, the difference in feed grain requirements between high and low efficiency feeding systems is in the range of 60-90 m tons annually by 2000. This is some 35-50% greater than the projections by MAAF's Planning Bureau for grain production excluding rice and wheat. Improvements in FCRs will depend largely on supplies of high- energy and high-protein feedstuffs, which in turn are likely to require changes in future cropping patterns, increased transport of feed and/or live- stock among regions, and greater dependence on international trade. 2.20 This analysis has not taken into account the requirements of the industrial sector for nonfood products or any additional demands which might be placed on the sector to generate export earnings in support of overall economic growth. There are many opportunities for substitution of industrial intermediate products for agricultural raw materials, notably fibers and oils, which makes it difficult to project future requirements of these materials. China is now an exporter of both animal products and a variety of agricultural specialty products, and the potential for expansion of these exports will be considered below. Both industrialization and foreign trade policy could either augment or diminish net requirements fo the agricultural sector. 2.21 The overall potential for increasing the production of various types of agricultural products will be examined more closely in Chapters 3 and 4. The analysis of this Chapter suggests that future domestic supply potential may be insufficient for the radical restructuring of consumption implied by - 27 - projections based on consumption patterns of analogue Chinese populations. The opinion of MAAF's Planning Bureau is that growth of grain production will be more modest than assumed by CAAS' optimistic projections and will reach a level of about 480 million tons by the year 2000. Within this total, the 300 million tons projected for fine grains (rice and wheat) would exceed estimated requirements for direct consumption in all scenarios above. On the other hand, domestic supplies of coarse grains would be deficit relative to projected requirements. At best, availability of feed grains would be suffi- cient for the low growth projection. Under the high growth scenario at least 30 m tons of high protein meal would be required to meet projected consumption of livestock products (Table 2.3). If 70% of oil meal production was fed to livestock in 2000, China's oilseed meal supplies for livestock might equal about 15 m tons. Another 15 m tons of imports at 1984 prices would cost some US$3.3 billion. 2.22 The question emerges of how to reconcile possible future imbalances between domestic supply of agricultural products and consumption patterns which reflect likely income and population growth. Policy measures which both constrain demand and encourage supply are likely to be required. On the demand side, the instruments of price and foreign trade restrictions can be used to shape consumption patterns to better conform with available produc- tion. Other countries, both developed and developing, have used these measures to hold food consumption to levels below those expected or "normal" for a given level of GNP per capita.~/ Present Chinese price policy has probably had the reverse effect, because food prices have been kept low relative to prices of industrial consumer goods and the supply of urban services (includ- ing housing) has been subsidized and restricted. Because basic nutritional requirements have already been met on the average in China, some restraint of consumption may be acceptable from a nutritional point of view. On the supply side a key consideration may involve the nutritional and resource tradeoffs between direct consumption of grain as opposed to its use in livestock feed- ing, the products of which are then consumed by the human population. If con- sumed directly, the cereals or pulses from one ha in China yield at least five times as many calories and two and one-half times as much protein (albeit of lower quality) as if those grains w~re first fed to animals or poultry and the products then consumed by humans.!21 The amount of China's agricultural resources committed to livestock and poultry production is likely to be a key consideration in balancing future supply and demand for farm products. 16/ Among developed countries, Japan is a good example of how price and trade policy influence consumption patterns. Despite income levels roughly similar to those in the United States, per capita consumption of red meat, poultry and fish in Japan is only about half as high. Reflecting high food prices, 32% of consumer expenditure goes for food. The comparable figure in the US is 20%. 17/ Assume the 1982 mix of grain crops with a weighted average yield (rice on a milled weight basis) of about 2,800 kg and each kg containing 3,550 kCal and 95 grams of protein. With an FCR of 4.2:1, this grain mix is adequate to produce some 670 kg of animal products (assume largely pork and poultry) with 2,700 kCal/kg and 97 kg of protein. - 28 - 3. DETERMINANTS OF FUTURE GROWTH 3.01 This section discusses those factors considered to be the most important determinants of growth of Chinese agriculture through 2000. They include the availability of key production inputs such as agricultural land, irrigation, fertilizers, and improved seeds, as well as less tangible institu- tional factors related to the sector's research and extension, rural credit and transportation, storage and marketing services. Perhaps the most impor- tant determinant of agriculture's future performance is the economic environ- ment in which China's farmers will operate. As agriculture's rapid growth since 1979 graphically illustrates, Chinese farmers respond positively to a favorable economic environment characterized by attractive input-output price ratios, assured access to land and markets and stable policies upon which to base production and investment decisions. Land and Water Development 3.02 The unique features of China's agriculture include the relatively small portion of the total land area which is cultivated (about 10% as com- pared with 57% in India) and the large portion of the total cultivated area which is irrigated (45%, as compared with 23% in India). At present more than a quarter of China's cultivated lands of just under 100m ha is classified as high and stable yield areas, with average grain yields in excess of 2.2 tons per ha per crop. Another 40 m ha are medium yield lands with grain yields in the range of 1.2 to 2.2 tons per ha. One third of the cultivated area is con- sidered low-yielding (less than 1.1 tons per ha) and subject to drought, waterlogging or flooding. Because of past efforts in irrigation, drainage and flood control, adverse weather now affects agricultural production to a smaller extent than it did in the years immediately after 1949. 3.03 There is considerable uncertainty regarding the actual cultivated area in China. Although official figures of just under 100 m ha continue to be used, considerable evidence, based on satellite imagery and a cadastral survey in progress, suggests that the true cultivated area is substantially higher. In 1981 the State Agricultural Commission provisionally revised the cultivated area to 133 m ha. A general view within the Government is that the actual cultivated area is a quarter to a third greater than the official figure. The largest discrepancies between official and actual cultivated areas appear to be in North and Northwestern China where crop yields are considerably lower than in the eastern and southern areas. Assuming that total output figures are accurate, a larger cultivated area would imply significantly lower yields than official figures and lower per ha usage levels of key inputs such as fertilizer. Notwithstanding, this analysis uses official production, yield and input data unless otherwise specified. 3.04 Official figures indicate that the cultivated area increased rapidly during the 1950s, but declined marginally in recent years. The reported area of cultivated land in 1983 (98.4 in ha) was only about half a million ha larger than it was in 1949. The irrigated area grew rapidly through about 1973, but has shown no significant increase since that time. The Government assumes that China's cultivated area will remain essentially unchanged over - 29 - the next 15-20 years, as land lost to roads, urbanization and industrial development is approximately offset by newly-reclaimed land. The national multiple cropping index (MCI) is projected by MAAF to increase from about 1.47 at present to 1.60 in 2000 in response to irrigation development, although the MWREP considers this to be optimistic. The major questions determining future availability of agricultural land include: (a) the extent to which reclama- tion will offset land lost to non-agricultural uses; and (b) the extent and effectiveness of future irrigation development. The MCI appears to have de- clined marginally in 1979-83, reflecting a reduction in triple cropping in the South and double cropping in the North. This seems likely to continue. 3.05 Land reclamation. Bringing into cultivation unused or underutilized land has been a major thrust of China's agricultural development effort over the past 25 years. In the 1957-77 period, the reported gross loss in cultiva- ted area of some 33 m ha of generally high quality land was partially offset by some 21 m ha reported of newly-reclaimed land, much of which was located in less favorable agro-climatic areas. Chinese sources suggest that about 33 m ha of "waste land" are "reclaimable" for some type of agricultural use, with perhaps 60% (20 m ha) of this suited primarily for pasture.~/ Of the remaining 12-13 m ha, about 6 m ha are in the Northeast, 5 m ha in Xinjiang and the balance is in Nei Monggol and elsewhere. Much of the area in Xinjiang may be of questionable quality and land use conflicts with herdsmen are important in other places. In short, perhaps 3-5 m ha are suitable for devel- opment in the medium-term for sustained production of annual crops. These figures may be compared with the reported loss of about one m ha annually to non-agricultural uses in 1959-78. Annual losses at present are not known, but outside major municipalities, there are few legal restrictions on conversion of farm land to nonagricultural uses. The current surge of rural housing construction suggests that losses of agricultural land will continue to be substantial, notwithstanding government efforts to restrict conversions of this type. In Jiangsu Province, where pressure to convert agricultural land to nonfarm use has been particularly intense, the cultivated area declined by about 1% p.a. in 1979-83. A national average figure half as high would imply losses of about half a million ha annually. Although some conversions may be economically justified, institutional reforms might be devised which would make disincentives for land conversion as costly as the economic loss to society. For example, if a land tax were to replace the agricultural tax and quota system, it should apply with at least equal incidence to land in agricultural and non-agricultural use. 18/ Kang Qingi, "Problems of Rationally Developing China's Wastelands Suitable for Cultivation," NYJJLC, 5 (1984), pp. 259-262. - 30 - Table 3.1: CULTIVATED AND IRRIGATED LAND, 1970, 1975, 1977-82 (1) (2) (3) (4)=(2 )1'(1) Cultivated Crop Irrigated Multiple land area land /a cropping index million ha 1970 101.1 143.5 36.0 1.42 1975 99.7 149.5 43.3 1.50 1977 99.3 149.5 45.0 1.50 1978 99.3 150.1 45.0 1.51 1979 99.6 148.5 45.0 1.49 1980 99.3 146.4 44.9 1.47 1981 99.0 145.2 44.6 1.47 1982 98.4 144.7 44.2 1.47 1983 98.4 144.0 44.6 1.46 /a Figures for irrigated areas refer to "effective irrigation," i.e., command areas with adequate water and engineering structures to permit irrigation appropriate to crop requirements. Source: Various official sources. 3.06 Grain bases. Much of the reclamation emphasis of recent years has been on the relatively productive areas designated as national or provincial commercial food grain bases (Appendix Table A.8 and Map 4). The 15 national grain bases (NGBs) together account. for more than a quarter of total grain production and about 40% of the grain procured by the state. The southern grain bases typically are densely-populated, have high and stable yields from good water control and benefit from high levels of input use and good management. Most of these lands are double cropped and potential for expansion of the sown area appears to be limited. The grain bases in the North have lower yields, but are sparsely-populated and thereby serve as important sources of commercial grain. The new NGBs, located mostly in the north, have benefitted from large-scale water conservancy works and land development efforts in recent years and are important producers of commercial grain. 3.07 Another 15-20 areas are designated as provincial grain bases because of their potential for commercial grain production to meet intra-provincial - 31 - needs. Further development of these national and provincial grain bases would allow other areas to increase their specialization in industrial crops. Based on incomplete data, it appears that another 3 m ha could be developed in the NGBs. All of this is located in the Northeast or Northwest where only single cropping is possible. 3.08 The loess plateau. The loess soils, covering some 50 m ha in North and Northwest China are uniform and generally fertile, with a pH level of 8.0- 8.5 which permits establishment of soil renewing crops such as legumes. Most of these areas are now used in cropping or livestock, but at low levels of productivity. Rainfall, if carefully husbanded, can sustain annual cropping at low yields over a large part of the plateau. Precipitation varies from about 500 mm in the southeast, through a 400 mm zone of former grassland to a northern desert zone of 200 mm or less. In the southern parts of the plateau rainfall and irrigation permit two crops per year, but in much of the region the short growing season precludes double-cropping and deep water tables make irrigation from groundwater uneconomic. The rationale to rehabilitate these lands includes the need to reduce erosion and the transport of sediments to reservoirs and canals below, and to increase the productive capacity of these lands which continue to support, albeit at low levels, a sizeable human population. 3.09 In some parts of the plateau where labor is abundant, conventional soil conservation measures have been used successfully. In the more rugged and sparsely-populated areas, new technology involving, e.g aerial seeding of grasses and legumes show considerable promise. With available technologies it seems possible that productivity of up to 20 m ha of loess soils might be substantially improved. Major elements in the improvement of such lands include the conversion of the most erosion-prone croplands to grass and trees, control of grazing particularly by sheep and goats, and the planting of better adapted crop varieties. Sizeable production gains within a relatively short period seem possible (Table 3.2). Whether such development would be economi- cally justified is a matter deserving further study. - 32 - Table 3.2: SELECTED DATA FROM TWO GULLIES ON THE LOESS PLATEAU, SHAANXI PROVINCE, UNDERGOING TREATMENT FOR EROSION CONTROL 1979 1983 a/ (before treatment) Chuen Jia Gull;t Cultivated area (ha) 234 163 Grass land (ha) 16 79 Forested area (ha) 60 100 Livestock numbers Large animals 51 89 Sheep 72 257 Swine 169 172 Goats 280 9 Income from livestock (yuan) 8,500 21,300 Crop production b/ (tons) 249.5 337.4 Crop yield (tons7ha) 1.1 2.1 Ratio of cropland:trees:grass 15:3.8:1 2:1.5:1 Gao Jia Gull;t Cultivated area (ha) 168 147 Grassland area (ha) 12 50 Forested area (ha) 31 43 Crop production b/ (tons) 169 241 Crop yield (tons7ha) 1.0 1.6 Agricultural income (yuan) 65,000 119,000 Ratio of cropland:trees:grass 14:2.6:1 2.9:1:1 a/ Some data for Chuen Jia gully refer to 1981 or 1982. b/ Assumed to be largely grain and oil seeds. Source: Reports from Comprehensive Control Experiment Station, Mizhi County, Shaanxi Province 3.10 The red soils. Another group of problem soils widely regarded to have some development potential are the red and yellow soils in the warm, humid regions of South and Southwest China. Most of these soils are acidic, of low fertility and, although originally forested, some of these areas now support little more than shrubbery. About 8 million ha are under cultivation at low levels of productivity and perhaps 2-3 million ha more can be reclaimed for agricultural purposes. The challenge of the red soils is to find a range of cultural practices which permit sustainable crop, tree crop or livestock production systems which are economically viable. The necessary technology is - 33 - available and has been practiced with good results in parts of the south- eastern United States, Brazil and Peru. Success in these instances is highly dependent on sophisticated soil testing, fertilization, soil management and plant protection measures, all guided by skilled scientists supported by adequate laboratory equipment. Terracing of sloped lands, crop rotations including legumes, general applications of lime and other chemical fertilizers and careful monitoring and treatment for deficiencies of soil micronutrients are components in rehabilitation. The possibilities seem sufficiently promising to warrant the beginning of development efforts. 3.11 Irrigation and Drainage. In China's environment of land scarcity, irrigation and drainage take on particular significance as means to increase the sown area. The MWREP assumes that future irrigation development will permit an increase in the total irrigated area from the present level of about 45 million ha to something over 53 million ha by 2000. Other estimates are somewhat higher. These differences reflect different assessments of trends in conversion of irrigated area to nonagricultural use, availability of budgetary funds for water conservancy, and investment requirements for rehabilitation of aging existing works. Expanded irrigated area is not necessarily more important than improved water management in existing installations, a high proportion of which lack adequate facilities for distribution to the farm level. Under present average standards, the area actually irrigable falls short of official total area in one year out of four, due to water shortage. 3.12 Over half (4.3 m ha) of MWREP's planned expansion of irrigated area 7 WOuld fall in the rice region south of the Huai River, largely through upgrad- ing of existing facilities. Another large component (1.6 m ha} would occur in the North China Plain, primarily through the project to divert Yangtze River water northward, which ultimately could add 2.5 m ha of newly iryigated area and supplementary water for 1.8 million ha now under irrigation.~] The remainder would be split between the Northeast (1.0 m ha, including Nei Monggol) and the arid areas of the Northwest Loess region (0.6 m ha) as well as the two provinces of Xinjiang and Gansu (0.7 m ha). Additions to irrigated area will generally be costly, even in the South where improvements would come in hilly- or mountainous areas. MWREP estimates total costs to average Y 4,500-6,000 per ha where new reservoirs or major drainage works are not required, and Y 6,000-15,000 along the upper reaches of the Yellow River, where a lift of 100-300 m is usually required. The economics of such development may be questioned. 19/ Data from MWREP and the Yangtze River Basin Commission indicate that some 4.3 m ha would benefit from the project. Of the 2.5 m ha to be newly irrigated, most is in Hebei, Henan and Shandong Provinces. Hebei and Jiangsu would benefit most through the supplementary irrigation of already-cropped areas. Beneficiaries of the first stage of the water transfer project would be northern Jiangsu and southern Shandong. The middle diversion, which is not included in the··present project but which might be developed at a later date, involves another 3.8 m ha of newly irrigable land. - 34 - 3.13 It is estimated that about 23 m ha of cultivated area require drainage facilities to prevent waterlogging and, often associated ~~ith these areas, there exist some 7 m ha of salinized land, including 3.3 m ha in the North China Plain. Over several decades, some drainage improvement has been made to 18 m ha (including 4 m ha subject to salinity), of which 10 m ha now meets a protection standard of 1:5 years. Government ojectives fo1r 2000 call for increasing the area thus protected to nearly 17 m ha, eliminating salini- zation problems in the North China Plain, and making on-farm drainage improvements to an additional 1-2 m ha of low-yield fields. Areas benefiting would include the North China Plain, the Jiang-Han Plain, Northeast and South China. However, failure to combine irrigation with on-farm and higher level drainage works is a persistent problem in North China, and frequently leads to further salinization. The shortage of investment funds has contributed to the problem of inadequate drainage. Fertilizers and Seeds 3.14 Present use. Improved irrigation, better crop varieties and higher levels of chemical fertilizer use have been the most important mate!rial inputs behind China's good agricultural performance in recent years. Application rates for chemical fertilizer doubled between 1977 and 1981 and by 1983 aver- aged 169 kg of nutrients per cultivated ha (equivalent to about 115 kg per sown ha). About 70% of available fertilizer is used on grains and 30% on other crops. Consumption growth rates of over 11% p.a. in the 1970-80 decade are among the fastest registered for any developing country over a similar period of time. Current levels are more than three times those in India, about equal to levels in the United States, but a third those in J~lpan or South Korea, where price subsidies encourage use in excess of econc1mic levels (Table 3.3). Fertilizer use has been stimulated by greater availability and improved fertilizer: product price ratios growing out of the 1979 price reforms. 20t The N:P:K ratio for China (1981) of 100:31:4 may be co1mpared with a world average of 100:52:10. This serious nutrient imbalance reduces syner- gism and the benefits from the relatively high levels of nitrogen use. 20/ In 1976 the ratio of the price of nitrogen from urea to the price of paddy was about 4.4: 1, the highest of any major rice producing country in Asia. The average price for rice in 1981 was about 40% higher, reflecting a higher quota price, an above quota premium, and larger average deliveries to the state at these higher prices. This increase in the price of paddy was a major factor behind the decline in the ·· nitrogen:rice price ratio to about 3.2:1. N. Lardy, Agricultural Prices in China, World Bank Staff Working Papers No. 606, 1983. - 35 - Table 3.3: A COMPARISON OF APPLICATION RATES FOR CHEMICAL FERTILIZER, 1980 K~/ha of arable land /a N:P:K Country N P205 K20 Total ratio China 87 27 2 116 100:31:2 India 21 6 4 31 100:29:19 United States 56 26 30 112 100:46:54 Japan 126 141 105 372 100:112:83 South Korea 204 89 83 378 100:44:41 /a Includes cultivated area under temporary crops, temporary meadows for hay or pasturing, land under market and kitchen gardens, land temporarily fallow and land under tree crops. Source: FAO for all countries except China. Estimates for China were calcu- lated from application and sown area data from MAAF and SSB. 3.15 China has a long history of using organic fertilizers and, until very recently, organics probably supplied a major share of total nutrients from fertilizers. Estimates of nutrients from organics must be interpreted with some caution, since quantities, particularly of nitrogen, available for plant use depend heavily on the type of raw material and methods of storage and handling, etc. Estimates in China of the amount of N supplied by organics in 1980 range from 5-10 m tons. Possibly 4-5 m tons of P2o5 and S-6 m tons of K2o came from this source. On this basis, total nutrients from organics (14-21 m tons) would have accounted for at least 40% of all nutrients applied in 1980. While these estimates may exaggerate the importance of organics, organic fertilizers continue to be important souy~es of nutrients in China, particularly for P, K, and some trace elements.~/ MAAF estimates that only half of nutrient offtake will come from chemical sources by 2000. 3.16 Production and imports. About 85% of China's total consumption of chemical fertilizer is produced domestically (89% for N, 72% for P and 6% 21/ Based on the results of Chinese experiments comparing plant offtake of nutrients from chemical and organic sources applied during the same crop season, the Mission estimates that chemical fertilizers accounted for 71%, 24% and 4% of 1981 total offtake of N, P and K respectively. The remainder came from either organic fertilizers or mining the soil, and soil mining of P and K may be significant at present offtake levels. - 36 - for K). In 1983 some 2,200 plants in China produced 13.8 m tons of fertilizer (nutrients), of which 11.1 m tons were N, 2.7 m tons were P2o5 and about 29,000 tons were K2o. Domestic production 1s characterized by a s1mall number of product types and low nutrient content. Nitrogenous fertilizer production is about 60% ammonium bicarbonate (16-0-0) and aqueous ammonia; 33:% is in the form of urea (46-0-0) and the balance is largely ammonium nitrate. About 70% of P production is single superphosphate (12-20% P2o5 ) and 30% is calcium magnesium phosphate (14-18% P2o5 ). Users of local products complain of low quality, lack of uniformity and, occasionally, poor bagging. Urea and diammo- nium phosphate (18-46-0) are commonly preferred products on quality and price grounds. Imports of fertilizer have averaged 10 million product tons (2-3 m .nutrient tons) in recent years at a foreign exchange cost of more than US$1 billion p.a. 3.17 Distribution and use. Farm sales of chemical fertilizer are made by ~he Agricultural Means of Production Corporation, although the local distribu- tion network is essentially that of the National Federation of Supply and Marketing Cooperatives which handled the job until 1982. Most imports are handled by SINOCHEM, but small amounts are brought in directly by provincial trading companies. About 20% of domestic production is from relatively large plants owned by the central and provincial Governments and is usually distri- buted through the state system at centrally-established prices. In the 1973-79 period, the portion of chemical fertilizer distributed through the state system declined from 48% to 23%. That portion has probably c:ontinued to decline. A substantial share of state supplies is allocated as re~1ards or exchange for items procured by the state and typically goes to intemsi vely cultivated areas where application rates and yields are high. Fertilizer produced by county owned or lower level units is allocated locally under a state-recommended pricing system which is not always honored if mat·ket imbal- ances exist. This distribution system, and the use of fertilizer BlS rewards in pursuit of production objectives, has spawned a complex bureaucracy. In 1982 some 16 government agencies at state level, and more than twic:e this number at lower levels, were involved in the production and distribution of - 37 - chemical fertilizer. The system is cumbersome and seem~ ]ikely to result in high marketing costs and, commonly, delayed deliveries.~ 3.18 The state system of allocating fertilizer has been used as a policy instrument to encourage production and marketing of key commodities. Fragmen- tary information suggests that the allocation process in the past has favored food crops relative to feed grains and pulses; the high, stable yield areas relative to lands of lower quality; irrigated areas relative to non-irrigated areas; and the state farms relative to the collective sector. Surveys of crop production expenses (1981) indicate that per ha costs of chemical fertilizer applications averaged about Y 150 for paddy rice, Y 120 for wheat and corn, Y 50-75 for other grains and tubers, Y30-85 for soybean and major oilseeds, Y 175 for cotton, Y 160 for other fiber crops, Y 250 for sugar cane, and Y 45 for sugar beet. Differences in application rates among crops have narrowed considerably since the mid-1970s. Regionally, application levels of nutrients per sown ha (Map 3) ranged from less than 50 kg (Xinjiang, Gansu, Nei Monggol, Heilongjiang) to over 150 kg per ha (Fujian, Jiangsu, Shandong and Liaoning). 3.19 Consumption and supply prospects. Fertilizer, both chemical and organic, will obviously continue to be a major factor contributing to future agricultural growth within an environment of unchanged cultivated area. The question is whether inability to increase supplies rapidly enough can be anticipated to be a major constraint on growth. Based on crop offtake of nutrient at production levels projected for 2000, MAAF expects consumption of chemical fertilizers to reach 30 m tons of nutrients, equivalent to about 240 kg per cultivated ha (adjusted for underreporting) - or more than twice current US levels, yet well below those of Japan and South Korea. This projection assumes, perhaps optimistically, that use of organic fertilizer would continue to grow, so that only half of total nutrient requirements need be supplied from chemical sources. If the targeted N:P:K ratio of 100:50:40 were achieved, chemical fertilizer applications would consist of about 16 m tons of N, 9 m tons of P2o5 and 5 m tons of K20 • 3.20 Unofficial production targets for 1990 by the Ministry of Chemical Industry place total nutrient production at 17.1 m tons, of which 13.0 m tons are N, 4 m tons are P and 0.1 m tons are K. Thus, growth rates of N produc- tion would slightly exceed 2% p.a., while P would grow by 6% p.a. and output 22/ These factors are costly in economic terms. Delayed delivery and application reduces potential benefits from fertilizers. Multiple handling increases physical loss and raises distribution costs. The International Fertilizer Development Center (IFDC) estimates average physical loss in developing countries at 10% each time fertilizer is warehoused. Marketing costs are often 50% or more of the retail price in these countries. China appears to be no exception. Price data for 1979 indicate factory to farmgate markups of about 30% for urea and 40% for ammonium nitrate. These logistical difficulties explain in part the continued popularity of fertilizers from smaller plants outside the· state system, despite their problems with product quality and process inefficiencies. - 38 - of K would quadruple by 1990. MAAF assumes domestic production of 24 m tons of nutrient by 2000, elimination of imports of nitrogenous fertilizers, but continued imports of P and K to meet large domestic shortfalls. If the chemical industry were to accept these goals, they might aim to produce 16 m tons N, 1 m tons P and 0.6 m tons K by 2000. This would leave a supply short- fall of 2 m tons P and 4.4 m tons K which could be imported at a projected cost of about $1.7 billion (1983 dollars). As this only exceeds the current bill for fertilizer imports by 50%, the projected burden on foreign trade is not large. 3.21 Adjustments in organic fertilizer use are also warranted. Due to the scarcity of rural fuel supplies, a high proportion of crop residues are burned for heat and cooking. In the south, straw from the early rice crop is commonly burned in the fields even in areas where the soil is short of organic material, while oilseed meal is often used as fertilizer where feedstuffs lack protein. While there are usually technical or economic reasons why such practices are difficult to change, the situation could be improved through, for example, development of alternative rural fuel sources, including coal, fuel wood, and biogas (in association with commercial livestock production); and chemical replacements for nutrients and microelements supplied by oilseed meal fertilizers. 3.22 China has developed a capacity to design and construct efficient, large-scale plants for the production of nitrogenous fertilizers. Two new nitrogen production facilities began trials late in 1983, each with an annual capacity of 220,000 tons of N. Two additional nitrogen plants are planned to come into production before 1990. Total increased capacity in all four plants would be less than a million tons by 1990. If raw material supplies are made available (coal, fuel oil or gas), the rather modest production targets for N would seem realizable. The higher growth rates for P and K raise more ques- tions, although they are from a low base production level. Existing phosphate deposits are located in the central south and the mountainous southwest, in areas where transportation facilities are poor. Plans are to est.ablish phos- phate fertilizer plants near mines and produce high analysis products (such as TSP or DAP) for shipment to distant markets. Because China has limited exper- tise in the design and construction of large plants for these mat1erials, it expects to rely on foreign technology and advice. Domestic deposits of potash are limited to the Qaidam Basin, where a 1 m ton potassium chloride plant -will eventually be constructed. For both P and K production, achieving year 2000 targets demands an early commitment of funds and manpower to a process requiring mastery of new technologies. 3.23 Questions remain regarding the availability of the 30 m tons of nutrients projected annually from organic sources by 2000. If th1e projected increase in organic fertilizer use does not occur, proportionately more foreign exchange will have to be committed to chemical fertilizer purchase, or else fertilizer availability will be inadequate to meet requirements of targeted production. Assuming that organic fertilizer on average contains about one percent nutrients by weight, these targets imply the ne~ed to collect, store, transport and apply some 3 billion tons of organic materials annually. Although these materials may be available if high growth rates of livestock are achieved, would the use of such large quantities of labor be - 39 - economically desirable? It would almost certainly not be consistent with declines in the farm labor force projected by MAAF. Willingness of sufficient labor to undertake these frequently onerous tasks is also a consideration, as indicated by e~perience in other countries where commercialization of agriculture and growing opportunities for labor have resulted in steady declines in the use of organic fertilizers. Much depends on the future oppor- tunity cost of labor. If projected increases in labor productivity in the rural economy materialize, through opportunities for slack season and non- agricultural employment, it should be asked whether farmers in the year 2000 would be willing to put in the work time required to apply organic fertilizers at a rate approaching 25 t per ha. If alternatively, this were forced by, for example, tying chemical fertilizer distribution to organic fertilizer use, it would have undesireable consequences to the efficiency of fertilizer distribution. 3.24 Adjustments tn policy and practice regarding fertilizer use and distribution seem warranted. To favor particular commodities (food grains) or production systems (state farms) or agro-ecological zones (high, stable yield areas) may have been more economically rational when fertilizer application rates were relatiy~ly low. This pattern of allocation appears much less efficient today.~/ At the new higher levels of application, marginal yield responses in the high yielding localities are frequently below marginal yield responses in the low yield areas. CAAS studies suggest that marginal yield responses in some high yield areas on average declined from 15-25 kg of grain per kg of nitrogen in the 1960s to 5-10 kg of grain today. These results would imply a decline in the incremental benefit-cost ~4tio from nitrogen application (urea derived) from about 3.4:1 to 2.3:1.~ 1 The decline in incremental return in the high yield areas is reported to have stimulated illegal resales of fertilizer to areas where marginal returns are higher. 3.25 The general objective for fertilizer distribution is to facilitate the use of economically-optimum quantities in a timely manner. In the past, with general scarcity of chemical fertilizers and only nitrogenous fertilizers to distribute, a system of rationed distribution was easier to administer and could serve efficiency objectives. Today, with more abundant nitrogen but continuing scarcities of P, K and microelements, there is strong reason to allow the market to play a large role in allocating fertilizers. The most critical task for Government is to develop fertilizer testing programs and related extension work, involving both economic and technical considerations, which could guide the processes of production, distribution and utilization. 23/ Guo Jinju and Lin Bao, "A Study of China's Chemical Fertilizer Problems," Turan Tong Bao (Soils Bulletin), No. 2, April 6, 1983, pp. 25-27. 24/ Assume an average farmgate price for paddy of about Y0.17 per kg in 1965. By 1981 the paddy price had increased to about Y0.31 per kg (Y0.20 as the 1973 average quota price for japonica and indica rice, adjusted for 20% price increase on quota price received by producers in 1981). The price of nitrogen from urea at farmgate is estimated at about Y0.98 (Y450 per ton of product @ 46% N.). - 40 - Fertilizer import priortttes should be closely aligned with the nutrient needs established by research. 3.26 Seed supplies. Good quality seeds in adequate quantity are critical inputs in high-yield agriculture. At the central level, MAAF's National Seed Corporation (NSC) is responsible for national seed policy and planning. The NSC regulates international trade in seeds, coordinates domestic SE~ed produc- tion and distribution, and provides technical guidance to provincial and county seed companies. The 2,300 local companies are responsible for the multiplication, pro~~?sing, storage, certification and distribution of stock and certified seed.-- The breeding and development of pure strains is carried out primarily by provincial agricultural research institutHs. Multi- plication of stock and certified seed is done by seed farms operatHd by local seed companies. Special contracting arrangements also exist where brigades, teams and households, under supervision of center technicians, produce certi- fied seed which is then returned to the company for processing and ultimate distribution to users. The Central General Bureau of State Farms also oper- ates a system of seed companies and farms which produce seeds for both state farms and collectives. The 2,600 seed farms under both NSC and GBSF serve about 90% of all counties. 3.27 Although basic mechanisms are in place, the national seed system has a number of shortcomings. Supervision and technical services are inadequate and storage and processing facilities are in short supply. The national supply of seed produced under NSC supervision totals about 2.5 m tons and meets less than 12.5% of total seed requirements. Less than one third of this supply has received proper treatment and certification. Most farme!rs there- fore continue to rely mainly on retained seed of uncertain quality and paren- tage. In recognition of these shortcomings, the NSC has establishe!d pilot projects in 120 counties, and the provincial governments in 200 counties, to improve seed processing, storage, and distribution. The national target is to produce up to 10 million tons of seed under NSC supervision by the year 2000. The government is also considering developing a system of SE!ed testing centers, independent of NSC control, which would certify seed quality. The Policy Environment 3.28 The policy environment refers to the economic circumstanc:es in which Chinese agriculture will operate and includes a wide range of polic:y measures which influence levels of work effort, production, consumption, savings and investment by farm families. The prospects for agricultural growth in any country are strongly influenced by both the structure of production incentives and the level and form of investment in the agricultural sector. The promulgation of the PRS has significantly improved agricultural production incentives, and can share credit with price adjustments for the outstanding growth record of Chinese agriculture since 1979. Its benefits have been 25/ Stock seeds are progeny of genetically pure strains. Certified seeds are improved seeds available for distribution to farmers and conform to specified purity and quality standards. - 41 - manifested primarily in greater labor effort and improved production manage- ment. Additional benefits might be gained by improved allocation of resources among productive activities, often associated with various forms of special- ization. The Government recognizes this, and is conscious of the need to develop markets in productive resources to facilitate specialization. As noted in Chapter 1, recent directives open the way for development of a labor market and allow subcontracting of land. 3.29 China still relies to a considerable degree on administrative measures to distribute production inputs and assign land to major crop catego- ries. Despite considerable loosening of the system through recent reforms, the Government continues to circumscribe the decision-making freedom of farmers beyond the requirements for efficient use of common infrastructure. Further measures to permit an enlarged role for prices and the market were announced in late 1984 and should provide an enhanced environment for agricul- tural development. However, there appears to be a continuing concern within the Government that the present price system may not provide appropriate sig- nals which would lead a farm sector, composed of a very large number of inde- pendent decision makers, to grow and market the quantities and mix of crops needed to meet national requirements. Thus, quotas for sown area of some crops are reportedly being retained in order to prevent farmers from switching a larger portion of land under their control from grain to more profitable industrial crops. 3.30 Production incentives. Despite substantial increases since 1979, average procurement prices appear to be lower than the level required to in- duce voluntary sales of an equivalent amount of produce and therefore involve a tax on the farm sector. However, since in theory the procurement quota is fixed in absolute amount but varies among teams in proportion to land produc- tivity and cultivated land per capita, it satisfies the criteria which econo- mists have set for the ideal land tax. Eliminating the quota procurement system without instituting some alternative tax scheme would reduce revenues available to the Government, either directly through increased procurement costs or indirectly, as an increased urban cost of living forced compensating wage increases, leading to decreased profits from state enterprises. 3.31 It may be asked whether the new farm price structure provides an appropriate set of incentives to encourage efficient production. In reviewfng the price structure, the marginal price received or paid deserves attention, because adjustments in cropping patterns and input use are influenced by relationships between marginal revenues and marginal costs. Because many of China's agricultural products and modern inputs are internationally traded goods, domestic prices for these items can usefully be compared to border prices (world prices adjusted for transport, trade and processing margins to the farmgate level). The comparisons may suggest needed price adjustments to encourage efficiency. If the ratios of marginal domestic prices of outputs to input prices differ from the corresponding relative border prices, the total supply of a product from domestic and international sources could be increased without increasing its procurement costs. For example, as long as the domes- tic pig:feedgrain price ratio is lower than the border price ratio, it should be possible to increase the domestic pig price, exchange part of the induced additional pig procurement for imported feedstuffs, and increase national - 42 - income in the process. Since quotas are based on levels of local production which have been attained in the past, crop producers can normally expect that any increments in production can be sold at above-quota prices or higher, or else (in part) retained, in which case the product may be sold at negotiated prices or on . the free market. Hence, marginal prices are not quota procure- ment prices, but are some average of the other three prices. Table 3.4 there- fore compares the full range of 1982 domestic price levels for major agricul- tural products and inputs with farmgate border prices.~/ 3.32 It was suggested in Chapter 1 that China's farm price structure is designed largely on the basis of internal considerations, viz., to give a reasonable return to labor or land above domestic material costs. Therefore, it is not surprising that the relationship of marginal domestic prices to border prices shows little uniformity, although domestic prices for a number of crops and inputs do not seem far out of line with border prices. Some important products appear to be underpriced relative to international prices. These include rice, timber, pork and fish which together undoubtedly account for a high proportion of agricultural procurement. Three of these (rice, timber, fish) are products for which one would expect production cost as defined in China to be a poor guide to pricing, because it does not reflect rent or returns to past investment in, e.g., paddy fields or fishpond con- struction and water control systems, or depletion allowances for forests. Underpricing of pork may reflect the grain-saving nature of traditional Chinese production methods compared to grain-intensive production systems abroad and the fact that pig manure in China is a valuable product, the benefits of which accrue to the pig producer. 3.33 On the other hand, certain products seem relatively overpriced, including edible oil, sugar crops, natural rubber and milk. For some of these, climate or natural resources may put China at a comparativE~ cost dis- advantage, yet the Government has sought to stimulate domestic production. The overpricing of edible oils does not extend to all oilseeds. The implicit subsidy may accrue to the processing enterprises rather than the farmers. Oilseed meals in China tend to be priced well below international prices for similar products. 3.34 Pricing of industrial inputs has been generally influenced by a Government policy of maintaining a modest profitability of manufacturing and passing along cost reductions to farmers through reduced input prices. Thus fertilizer prices, which benefit from underpricing of energy and raw materials used in the manufacturing process, have fallen to levels which approximate 26/ Exported or imported products may differ in quality from average domestic production. For this reason, as well as possible estimation errors in conversion to farmgate and unit weight equivalencies, conclusions drawn from comparison of border and domestic prices should be taken as indica- tive rather than definitive. Moreover, absolute values of indices, which would change nearly proportionately if other than the official yuan-to- US$ exchange rates were used, are less meaningful than price ratios between products. - 43 - Table 3.4: COMPARISON OF 1982 BORDER PRICES WITH DOMESTIC PROCUREMENT PRICES ~ Farmgate Percentage of farmgate border price border Quota Above-quota Free market Average Product price price price price price (Y/ton) Crops Paddy rice 435 53 80 Wheat 411 77 115 122-170 Corn 316 68 102 114-152 Soybean 590 117 117 119-136 Lint cotton 2,817 105 136 177-199 115 Oilseeds Rapeseed 927 77 116 Peanuts 738 131 196 Edible oil 1,046 374 265 Jute 660 56 67 73 Tobacco (smoked) 2,927 59 52 Vegetables 460 Sugar cane 26 162 191 Tree Crops Fresh fruit Apples 571 79 68 Mandarin orange 671 92 72 Tea (processed) 3,720 345 95 Natural rubber 1,637 372 Timber 211 41 Animal Husbandry Pork 3,562 50 47 56 Beef 2,435 67 120 Chicken 1,942 113 185 Eggs 1,755 173 106 Milk 213 282 Wool 6,378 56 Honey 1,417 133 Fisheries Fresh fish 3,151 22 67 29 Inputs Urea 428 117/b Ammonium sulphate 213 1197b Phosphate (P205) 898 104/b Potassium chloride 255 1037b Soybean meal 501 66/b /a From Annex, Table 3.3. /b State fixed price. - 44 - international prices for equivalent products, but could turn sharply upwards if price reform increases production costs. Farm machinery remains costly by international standards, but diesel fuel is relatively cheap. On the other hand, the price of electric power for irrigation pumping and fees for surface irrigation water both reflect substantial Government subsidies and would have to be tripled to cover average costs (including operations and maintenance costs but not depreciation for the irrigation system). 3.35 Additional evidence on the appropriateness of the present price structure can be derived from comparisons of ratios of prices of output to major inputs. The ratio of the farm-level prices of nitrogen to crop pric~s/ is a commonly used indicator of incentives for high input use (Table 3.5)._2 Assuming that above-quota prices approximate marginal prices in China, this ratio for Chinese wheat production is identical to the border pric,e ratio and more favorable than in several other major wheat producing countri,es. The ratio for rice is substantially less favorable, although not worse than those of some other Asian producers (with lower crop yields and fertiliz1er use rates). This is significant in view of the high crop yields and l1evels of fertilizer use, and diminished response to incremental nitrogen applications which appears to characterize much of Chinese rice production. 3.36 Not all producers in China face price signals which are as attractive as those considered above. Quota prices appear to vary consider- ably among geographical areas, as suggested by the considerable differences between reported national averages and local survey data. Quota lE~vels appear to discriminate against areas which had achieved high productivity before introduction of the PRS, and this may be inducing sizeable conversions of cropland to non-agricultural use as in, e.g., Southern Jiangsu. Farmers who normally cannot meet their procurement quotas face quota procurement prices at the margin, and, hence, rather unfavorable incentives to increased input use. Domestically-produced nitrogenous fertilizer varies widely in price per unit of nutrient, and farmers who depend heavily on high-priced fertilizers are penalized. Solutions to these problems might include adjustment of fertilizer prices to reflect quality, nutrient content and transport costs, and reduction of excessive quotas. 3.37 A similar comparison can be made of the ratios of animal product prices to prices of feed input, assuming use of a balanced feed concentrate (Table 3.6). The price ratios may be compared with FCRs which relate concen- trate input to resulting meat output: production systems using concentrate- intensive techniques are only profitable when product:feed price ratios sub- stantially exceed conversion rates. In terms of border prices, this favorable relationship exists for pork, chicken and eggs. Border price ratic1s for beef and milk appear to be below levels necessary for profitable production with concentrate-intensive technologies, possibly because international prices are influenced by exports from subsidized production. 27/ It should not be taken to indicate absolute profitability, as low procurement pricing may be a substitute for taxes, which are ignored in the comparison. - 45 - Table 3.5: RATIOS OF FARM LEVEL PRICES FOR NITROGEN AND GRAIN, SELECTED COUNTRIES Selected Price ratio Selected Price ratio wheat of nitrogen r1.ce of nitrogen producers to wheat/a producers to rice /b Mexico 1.3 Japan 0.5 Bangladesh 1.5 Korea, Rep 0.7 Egypt 1.6 (Border price) 1.5 China: above-quota price 2.3 Indonesia 1.6 (Border price) 2.3 Sri Lanka 1.7 Canada 2.4 Bangladesh 2.1 Pakistan 2.6 Nepal 2.2 USA 2.6 China: above-quota price 3.1 Turkey 2.7 Malaysia 3.1 France 3.0 Philippines 3.2 Brazil 3.1 Thailand 3.4 China: quota price 3.5 Pakistan 3.6 India 3.6 India (Orissa) 4.3 Australia 5.4 China: quota price 4.7 Ia International statistics from CIMMYT, 1983 World Wheat Facts and Trends, Report Two (Mexico, 1983), Table 6, p. 18. Chinese statistics are from Annex Table 3.3. /b International statistics for 1979 or 1980 are from IBRD-EAPD, Rice Hand- book, February 1981, Table VII-B1. Chinese statistics are from Annex Table 3.3. 3.38 Domestic product:feed pr1.ce ratios appear to exceed border price ratios for all products except pork. Aside from pork and beef, they are also more than adequate to make concentrate-intensive production financially attractive. Indeed, they may be too attractive to encourage efficiency. For poultry products, where most sales are at free market prices, this high ratio may be a combined reflection of strong consumer demand and low-productivity domestic production systems. The data suggest a need to increase pig procure- ment prices if increases in production are to come from a concentrate- intensive production system. Because of low domestic productivity, the profitability of beef cattle raising and dairy production seems questionable in China even at output price ratios more favorable than border prices. Further study of these price relationships is necessary: given China's agricultural diversity and size, it is important to examine price relation- ships location-by-location. - 46 - Table 3.6: COMPARISON OF PRICE RATIOS OF ANIMAL PRODUCTS TO FEED WITH EFFICIENT FEED CONVERSION RATES Ratios of eroduct erices to feed erices Ia Border Domestic Kg feed per prices pr1ces kg product /b Pork 10.1 6.2 4.7-5.8 Beef 6.9 9.0 10.8 Chicken 5.5 11.1 2.5-3.9 Eggs 5.0 5.8 3.5-3.7 Milk 0.6 1.9 0.4 Ia Product prices are border or average procurement prices per ton of dressed weight. Feed prices are a weighted average of maize (80%) and soymeal (20%). The above-quota price is taken to be the domestic price of maize. /b Conversion rates shown (kg feed:kg dressed weight) are average1i for countries with efficient production systems. 3.39 Though some agricultural products appear relatively overpriced, the marginal procurement prices of other products may be too low to elicit expanded production. As together these represent a sizeable proportion of total agricultural procurement, increases in average procurement prices would be costly to the government budget or to urban consumers. Higher p1~ices would create larger profits for existing efficient producers and would improve incomes most in areas which are already well off, as the presence of rice and pigs on the list would imply. These effects could be minimized if, e.g. (1) the entire price structure, including quota prices, were adjusted upward, but income effects were offset by quota adjustments and/or resource-based taxes; or (2) if the tiered pricing structure were retained, only the above-quota price was increased. 3.40 Agricultural Investment. In every country the linkage between agricultural investment and growth of the sector is obscure. And, as the Chinese experience since 1980 demonstrates, agricultural growth may accelerate while public sector investment declines as long as production incentives remain strong. But this phenomenon of rapid growth in the presence of declin- ing public investment poses the question of whether productive inves1tment in the sector is adequate to sustain agricultural growth at rates neede!d to achieve national growth and income targets. As noted in Chapter 1, available information on agricultural investment in recent years relates large!ly to budgeted funds at the state and provincial levels. Little is known of public sector investment outside the state budget or on-farm investment by individual households. Yet it is at these lower levels that investable savings: are growing most rapidly. At the national level, little analysis appears to have - 47 - been done which estimates the size and type of investment in agriculture required to achieve specific production and income targets. Ministries have reached a general consensus on the size of technically-feasible programs in agriculture to 2000 and have rough figures on unit development costs. But the major analytical effort relates to five-year plans which ultimately determine the allocation of central Government funds among ministries. 3.41 In these circumstances, the present analysis is limited largely to a consideration of investment in capital construction required for major development programs noted by Government agencies in discussions with World Bank staff. Because these programs have been developed by separate agencies and have not necessarily been approved at higher levels, they should not be construed as an integrated investment program for agriculture to 2000. Rather, they provide the basis for developing the annual estimates of public investment in agriculture in several development activities accorded priority by the Central Government. As provincial and local governments have diffe- rent, more localized sets of priorities, and control over investment resources is shared among governmental levels, an all-inclusive program listing might require investment funds which are substantially larger. Moreover, these estimates are partial in the sense that they relate largely to capital construction in the public sector and exclude most capital expenditure for critical support services such as agricultural education, research, extension, marketing infrastructure and transport. Further, they exclude increased working capital requirements for agriculture. The estimates in Table 3.7 should therefore be regarded as illustrative only and as reflecting assump- tions which are not necessarily shared by all ministries concerned. 3.42 The subsectoral structure of this "program" differs greatly from present sectoral shares of output. The share of investment primarily benefit- ting crop cultivation (about 30%) is less than half of the crop share in GVAOR (about 75%), whereas the qther subsectors' shares are at the least twice their present product shares.~/ While these investment "weights" are generally consistent with the structural change sought in agriculture by the Government, an investment pattern of this nature may not necessarily be desirable from an economic viewpoint, or be an effective means of achieving production and income objectives. More than half of total investment would be absorbed by dairy and forestry development. The advisability of massive investment in dairy development will be questioned below. And while forestry development tn general warrants priority on need and environmental grounds, expenditure on afforestation will add relatively little to annual product in the next several years. 28/ Chinese statistical convention defines the gross value of agricultural output (GVAO) to include the value of output from production of crops, livestock, fisheries, forestry and "sidelines." Sidelines include industries operated by brigades and teams, as well as hunting and gathering activities. In keeping with more conventional accounting definitions, this analysis defines GVAOR to include only the gross value of production from crops, livestock, fisheries and forestry. - 48 - Table 3. 7: ANNUAL INVESTMENT IN CAPITAL CONSTRUCTION FOR AGRICULTURAL DEVELOPMENT (1984-2000) ~ Total area Annual Purpose affected expend:lture (ha '000) (Y millions, 1982 prices) Crop cultivation Irrigation 6,900 2,264 Drainage, desalinization 11,000 688 Land reclamation 5,970 1,594 Red and yellow soils 3,000 1,042 Subtotal 5,588 Forestry Afforestation 60,000 1,408 Maintenance of immature plantings 60,000 4,388 Subtotal 5,796 Animal Husbandry Pasture improvement 33,330 816 Feed processing facilities 666 Dairy development 5,008 Poultry development 530 Subtotal 7,020 Fisheries Fishpond construction 585 562 Fishpond improvement 800 482 Subtotal 1,044 Grand total 19,448 ~ The projected expansion in irrigation is closer to the opinions of MWREP than to those of MAAF, but land reclamation on the scale envisioned by MAAF would often further enlarge irrigated area (investment included in reclamation costs). Pasture improvement investment excludes costs of wells or cultivated pasture. Development of feed processing is assumed to emphasize dispersed, small-scale facilities in order to minimize investment costs. For most other animal and fisheries investments, requirements derive from MAAF production targets for the year 2000. See Appendix Table A.11 for additional detail. Source: Mission estimates, based on information provided by various agencies. - 49 - 3.43 The rate of long term growth will depend heavily on the adequacy of agricultural support services, including the agricultural research and exten- sion systems. Because yield increases will be the source of most of agricul- tural growth, a strong research program aimed at improving production techno- logy is of particular importance in China. Although it is very difficult to estimate an optimum level of investment in research, a study of international experience ha~ ~ecommended an amount up to about 2% of the value of agricul- tural output.~/ For China, this would imply annual expenditure (capital and recurrent) of up to Y 4 billion. Comprehensive information is not available on research expenditure in China, but available ~vidence suggests current levels are considerably less than this amount.lQ/ 3.44 Suppose that the Y 20 billion average annual fixed investment level of Table 3.7, which could mean something like Y 11-12 billion at present, rising to Y 33 billion by 2000, may be taken as a minimum requirement: how easily could this investment level be financed? Although at first glance, estimated 1982 rural investment levels of Y 65 billion (Table 1.2) seem ample, a part of the former would be absorbed by replacement investment and working capital requirements. Moreover, only Y 16.5 billion of this total represents investment resources in public hands (excluding obligatory public service labor), of which Y 6.8 billion is now being committed to CBE development. The opinion in China is that state investment in agriculture will not be increased in the short run by more than Y 1-2 billion annually and collective investment (excluding that in CBEs) may continue to decline. Therefore, the ability to effectively tap private savings for productive investment is likely to be a critical determinant of the adequacy of agricultural investment. 3.45 Several approaches could be taken to capture a larger proportion of private savings for productive investment: (1) by bringing more types of developmental activity within the scope of the private sector through the granting of private use rights over collective or state property; (2) by inducing more private direct investment by ensuring stable tenure for families which invest in resource development, assuring access to production inputs, adjusting price levels where necessary, and by offering loans to supplement private investments; (3) by making time deposits attractive through higher interest rates; or (4) by imposing additional taxes on idle assets, offset with tax credits for the productive use of assets. As noted in Chapter 1, the Government has made a start on some of these measures in recent months. But even in the best of circumstances, it will take time for rural institutions to completely abandon preferential treatment for collective or state-sponsored activities and for farmers to become convinced that the present favorable climate for private investment will remain. 29/ World Bank, Agricultural Research: Sector Policy Paper, (June, 1981), PP• 102-106. 30/ As noted below 1n 1982 expenditure by agricultural research institutes at provincial levels and above was Y206 million, of which Y45 m was for constructing and equipping of research facilities. - 50 - 3.46 The banking system will need to expand its role in directing investment to more productive uses. In particular, rural credit institutions should be encouraged to devote a larger proportion of credit to long-term lending, which could bid resources away from consumption or nonproductive use and finance use of rural labor for productive investments. The present trend towards an increasing proportion of lending to individuals seems desirable as a means to support the development of specialized household production. Consideration also should be given to, e.g., a higher interest rate structure and the strict repayment policy necessary to ration credit, supported by measures to improve management of credit institutions. Agricultural Support Services 3.47 The restoration of a farming system organized around individual households has placed new demands on agricultural support services. Previ- ously, the various levels of collective organization could serve as major nodes in the delivery of production inputs and technical assistance, with direction below the production team level left to team cadres. At present the support system requires direct links to nearly 200 million production units which increasingly operate like small family-managed farms elsewher,e. In response to this requirement, the system has begun to undergo a restructuring based in part on an innovative extension of PRS concepts. The future development of Chinese agriculture will depend to a large extent on the success of this restructuring in providing small-scale farms with the full range of support services needed by production systems which are increasingly complex technologically and more closely linked to markets for production inputs and outputs. 3.48 Agricultural Research. Of critical importance will be th1e continued generation of yield-augmenting production technology. Much of the agricul- tural research in China today is carried out in the 390 agricultural research institutes at provincial level and above which operate under the general professional guidance of the Chinese Academy of Agricultural Scienc1es (CAAS) within MAAF. The MAAF also supports other smaller research agencie1s such as the China Aquatic Sciences Research Institute (under the general guidance of the Aquatic Products Bureau) and the South China Tropical Crops Research Institute (under the General Bureau of State Farms). The Ministry of Forestry is the parent agency for the Academy of Forestry Sciences which provides professional guidance for provincial forestry research institutes. The Ministry of Water Resources and Electric Power oversees 33 affiliat1ad research organizations which carry out some research related to water use in agriculture. 3.49 The various provincial research institutes under CAAS guidance are staffed by some 18,800 research personnel, along with support staff and workers. The CAAS itself has about 3,500 research scientists. Thi!; network has available about 13,000 ha of land for research and demonstration work. Another 49 crop and livestock research institutes operate at province level or below, are staffed by some 15,300 scientists, and are under provincial leadership. The Chinese Academy of Sciences (CAS) supervises the work of more than 100 research institutes, some of which carry out research in the agricultural sciences. About 11% of the research conducted by institutes - 51 - under the guidance of CAAS is classified as basic research, while the balance is designated as applied (71%) or development research (18%). The distinction among these categories is not always clear, although the latter two generally relate to specific problems encountered in agricultural production and processing. In 1982 expenditures by agriculture research institutes at provincial level and above were about Y206 mil}ion, of which Y45 million were for construction and equipping of facilitiesl! • 3.50 China's agricultural research system continues to suffer from the effects of the Cultural Revolution when many major research institutions were closed and staff dispersed to the countryside. Work at many research institutes is hampered by inadequately qualified staff and poor research facilities. Experimental fields and greenhouses need to be established or upgraded and there is a general need for better housing and educational, health and recreational facilities, particularly for research centers in isolated areas. Qualified staff, especially at the middle and senior levels, are in short supply. Of the total research staff noted above, less than 3% hold graduate degrees and more than 35% have not earned a college degree. A large part of the intellectual leadership for agricultural research in China is provided by research staff who are more than 60 years of age. 3.51 Linkages between research institutes and the agricultural colleges and universities tend to be weak. At the ministry level there is some inter- action between bureaus of science and technology responsible for research institutions and bureaus of education which oversee higher educational insti- tutions. At the provincial level, the ties between teaching and research institutions are commonly based on personal contacts and ad hoc arrangements for joint work on specific projects. Cooperative research arrangements are being increasingly encouraged by central funding agencies which may respond to bids on research projects submitted by colleges and research institutes by making joint awards. Linkages between research and extension are to be improved through the work of extension committees which are being established at the provincial level. 3.52 Agricultural Education. China has some 85 institutions of higher education in agriculture and related fields with a total student enrollment (1981/82) of about 100,000. This represents only 8% of the 1.2 million students enrolled in China's 715 institutions of higher education. At the secondary level, some 740 institutions provide training in agriculture in 2-4 31/ Among the agricultural research achievements of international acclaim in China were the early development of stiff-strawed, fertilizer-responsive wheat and rice varieties, the development of hybrid rice varieties, methods to control stripe rust in wheat and, more recently, the develop- ment of high-yielding or shorter season cotton varieties. In 1983 the State Scientific and Technological Commission granted a number of awards for outstanding research achievements by the CAAS system, including the development of a vaccine against equine infectious anemia, the develop- ment of wilt-resistant cotton varieties, and the development of an effective vaccine against swine fever. - 52 - year programs. About 185,000 students were enrolled in these programs in 1981/82. At the university level, agriculture is not generally th'e first choice of candidates and those admitted to the agricultural colleg,es and universities frequently have lower scores on admissions tests than candidates entering other fields. There is also a problem of retaining agricultural students after training is completed. Since the early 1950s, some 500,000 of the technicians trained in agriculture have left the countryside for jobs 1n urban areas. Through new policies - a special quota system for admissions of rural students to agricultural colleges and secondary schools and :salary incentives to encourage graduates to return to the countryside to 1NOrk - the Government hopes to develop a much larger group of committed agricultural professionals and technicians. 3.53 Although the precise dimensions of the skilled manpower needs in agriculture must await further detailed studies,~/ recent projections indicate shortages of appropriately-trained personnel at both the professional and technical levels. Training institutions are striving to overcome such shortages, but efforts are hampered by limited enrollment capacity and inadequate quality of instruction and range of curricula. Based on present enrollments of postgraduates in the agricultural sciences totaling about 1,200, the Government estimates that requirements between 1983-90 for per- sonnel with advanced degrees will exceed availability by more than 15,000. At the college graduate and technical levels, shortages are most acute in spe- cialized fields such as forestry and meteorology where the present flow of 5,000-6,000 graduates per year can satisfy only half of the requirt~ments over the decade. Overall, it is estimated that there are currently about four agricultural technicians per 10,000 agricultural workers. 3.54 There is a further need to upgrade the qualifications of already- employed personnel, both professionals and technicians. Extension personnel number about 170,000 of whom 35% are graduates of colleges or technical schools and 65% are experienced farmers with no formal technical training. According to MAAF/FAO data, 80,000 of the technicians in the exten!;ion network are in need of skill upgrading, as are most of those in the animal husbandry and veterinary service stations. The demand for in-service training in agri- culture for secondary school teachers is also of growing concern as the Government seeks to strengthen the agricultural curriculum in the secondary schools. 32/ Manpower requirements in agriculture and implications for educ:ational planning are being examined in a pilot study in Jiangsu Province launched in 1982 as part of the World Bank-assisted Agricultural Education and Research Project (CR 1297-CHA). One of the preliminary results is that the present deployment at the ratio of about 1:1 between colle!ge-trained personnel and technicians trained at the secondary level makes inappropriate use of higher level personnel. This investigation, some version of which may be undertaken in other provinces as part of a nationwide agricultural manpower survey, is to be completed in 1985. - 53 - 3.55 Agricultural Extension. China's past agricultural successes in such areas as the rapid expansion of hybrid rice production can be attributed in large part to the working of a well-structured extension network. Organiza- tional reforms in the rural areas, and the sharply diminished roles for brigades and teams in agricultural production decisions, have necessitated new approaches to extension. In 1982 a decision was taken to convert the existing county research institutes into county agro-technical extension centers (ATECs) which combine the formerly-separate extension, research, plant protec- tion, seed production, fertilizer distribution and soil analysis, farm machin- ery and animal husbandry stations. Each ATEC is expected to include 10-15 agricultural technicians and support staff. Physical facilities may include some 3-5 ha of land for experimentation and demonstration, 2,000-6,000 sq m of floor space for laboratories, class rooms, audiovisual materials and exhibi- tions. More than 300 such centers have been established, generally in the better developed agricultural areas. About a third of these have been created with MAAF funding, while costs of the other centers were met from provincial, prefectural or county sources. The Government intends to establish ATECs in each of the country's 2,300 counties over the next 10 years. 3.56 The ATECs' major functions include training and demonstration and the provision of several support services to farmers such as plant protection, machinery services, soil analysis, weather forecasting and supplies of seeds, fertilizers and irrigation water. Training is provided through a variety of courses offered to farmers and technicians at county and township levels and ranges from field days for farmers to short courses of two weeks' duration to two-year courses for farmer and township technicians. Demonstrations may relate to fertilizer and variety trials, pest and disease control measures and irrigation practices. Higher level technical support to the ATECs in the form of, e.g., more basic research, development of improved crop varieties, and production of teaching materials is provided by agricultural colleges or provincial academies of agricultural science. 3.57 Contracts between the center (or a subcenter thereof) and individual households are other means of providing agronomic advice, plant protection or machinery services. While farm machinery is typically owned by the machinery subcenter, it may be operated under contract by a farm machinery team which provides services to farmers on a fee basis. The system specifies bonuses to be received (or penalties paid) by technical assistance teams in the case of over (under) achievement of yield or production targets. The specialized households or 11 keypoint 11 households (similar to model farmers in other extension systems) also are expected to be important means of extending improved technologies to producers. 3.58 The new arrangements to provide agro-technical services vary widely among locations and continue to evolve. Although not yet adequately tested through operation, they are likely to foster greater coordination among units which provide agricultural support services. The close integration of training, demonstration and the provision of production inputs such as seeds and fertilizers should be particularly beneficial to the increasingly complex production systems which are emerging in Chinese agriculture. The consolida- tion of services also should improve the means of ensuring a two-way flow of information between the research institutes and agricultural colleges on the - 54 - one hand and the farmers on the other. MAAF's recently-created Leading Group on Extension has the responsibility to coordinate at the national level all agricultural education, research and extension activities. 3.59 Machine services. Since 1979, Government policy has changed from active promotion of agricultural mechanization to a market-oriented approach in which the future rate of mechanization will depend largely on farmer demand and farmer investments. Following the development of BGDH management systems, the reductions in field size, as well as farmer desire to obtain maximum control over timing of farm operations and to minimize cash expenditures, have caused a decline in demand for large and medium-scale machinery, and a corresponding rise in demand for draft animals. The demand for small-scale pedestrian tractors continues to be high. Excluding the 10% of tractors owned by state farms, about one-third of tractors are now privately owned, another one-third privately operated, and the rest collectively owned and operated under contract to mechanization teams or service companies. Mechanization continues to be important in irrigation, pest protection, processing and transport, as well as on state farms, where low labor/land ratios and short growing seasons typically militate against labor-intensive techniques. 3.60 The emerging organizational form for provision of machine services involves private or collective custom land preparation by speci,alized oper- ators or teams, whose earnings are supplemented by short-haul transport work. The provision of transport services which normally accounts for 40-60% of farm labor input, is far more profitable than custom land preparation for regulated fees. This may contribute to the decline in mechanized plowing and increased demand for draft animals. Tractor hauling is reportedly as effi- cient as truck transport on poorly maintained farm roads, but contributes to clogging of the rural road system. In the long run, the introduction of fuel- efficient, small scale trucks to replace off-farm tractor hauling, would produce important savings in fuel and transport time. In the medium-term draft animals, which now consume over 20 million tons of concentrate per year, are likely to continue to grow in importance. In view of the competition this poses for limited concentrate supplies, successful development of custom machine services would help free resources for production of livestock products for human consumption. 3.61 Marketing and transport. One of the most important constraints to the restructuring of Chinese agriculture and increased specialization is the limited capacity of the rural transportation, processing, storage and distri- bution system. Most options for accelerating agricultural growth and increasing productivity have as prerequisites investments and organizational improvements in this area. Efforts should be made to both expand the capacity of the rural marketing system and increase its operational effici,ency. The system could be improved by policy measures· which simplify pricing and pro- curement, reduce administrative barriers to the flow of farm output and production inputs, curtail the number of public intermediaries in the market- ing chain, and encourage larger roles for individuals or small groups in the management and ownership of purchasing and selling points, processing and storage plants and transportation equipment. The demand for mark,eting ser- vices (transportation, storage, processing) will grow rapidly as commer- cialization of agriculture expands. Experience in other countries I - 55 - demonstrates that individuals and small groups can provide these services efficiently and can usefully complement the Government's role in these activities. 3.62 There appears to be inadequate linkage between the work of the agricultural planners in China and agencies responsible for transportation, with a resulting tendency to consider such bottlenecks as largely insuperable constraints. With improved transportation, agricultural specialization could expand to more fully exploit comparative advantage and facilitate the transi- tion to a more commercialized agriculture. For an extreme example, Gansu's Hexi Corridor is less distant from China's Northeastern provinces than California is from the east coast of the United States and shares with some areas in California the dry, sunny climate and fertile soil which provide the basis for high value crop production. At present, melons and deciduous fruit from Gansu sell at high prices in Hong Kong and Chinese cabbage procured in Hexi for Y 0.02-0.04 per kg sells in Northeastern cities for more than Y 0.20 per kg. These price differentials suggest that a reliable system of refriger- ated rail transport might well generate enough revenue to more than cover the cost of supplying additional grain to horticultural producers in Gansu. More generally, there will be little increase in the average national per capita consumption of high value, perishable foods such as dairy products, fruits, vegetables and fish if their commercial production and distribution are confined largely to the vicinity of major municipalities as at present. Improved transportation would permit more rational use of suburban land, reduce input supply problems, distribute perishable products more widely and improve the income distribution effect of high value agricultural production. 3.63 Recent policy decisions to permit individual ownership of trucks and provide for the relatively free movement of privately-owned farm products across jurisdictional boundaries are major contributions toward solving the rural transport problem. More transport vehicles will put additional stress on China's rural road system which is already over-burdened. Because of past policies emphasizing local self-sufficiency and relative neglect of local transportation networks, the problems of rural transport in China may be more acute than in many other developing countries~'. Improvement of the rural road network is essential to a better marketing system. A substantial amount of internal trade in agricultural commodities could probably be moved advan- tageously by barge in terms of cost per ton km, particularly if the cost of new rail or road lines is considered. 3.64 At the production level, the stress on holding local grain reserves, the lack of adequate investment in the past for storage, and the surge in production in recent years is creating an increasingly serious grain storage 33/ The Sahelian countries of West Africa have more trucks per capita than do the rural areas of China. In Northwest India expansion of the rural road network was a major factor behind the farmers' adoption of new production technology and the five-fold increase in commercial grain supplies. These roads, which reduced marketing distances by half, were financed in part by levies on grain sales. - 56 - problem. Reliable measurements of post harvest grain losses are not readily available in China, but post harvest losses in grain in the range of 5-10% of production have been reported and storage problems are likely to be even larger for the more perishable fruits, vegetables and animal products. Cold storage capacity will have to grow very rapidly if the changing pattern of consumer preferences is to be accommodated. Chinese sources suggest that cold storage capacity in the whole of rural China is less than that available in Northwest India which has a total population of only about SO m. To ensure their efficient use, these cold stores would have to be integrated into retail arrangements for the commodities being handled and would require similar investments in urban areas. 3.65 At the national level, a major problem is an inadequate physical infrastructure for trade in agricultural products, particularly bulky products such as grain. Without major improvements, existing port faciliti,es appear to be capable of handling not more than 16-18 m tons per year. This assumes that improvements at Dalien, Zhanjiang and Qingdao are completed. Little of the unloading appears to be handled by continuous flow equipment and none of the ports can accommodate the largest and most efficient bulk grain carriers. Shipping charges for these carriers from North American ports are :$4-5 per ton below those of the smaller vessels which can be accommodated in Chinese ports. Some idea of the grain handling problem at the major ports may be indicated by demurrage charges which have been estimated at nearly $30 per ton in 1982 for those grain imports which were carried on non-Chinese bottoms (about 20% of total grain imports). 3.66 Agro-processing. Improved inland transportation and sto1rage facilities would provide a basis for agro-industrial development which requires assured deliveries of high-quality, uniform raw materials.. For cereals and oilseeds, the investment priorities include: (1) improved oilseed milling facilities to increase extraction rates and provide processes to reduce toxic substances in processed products; (2) feed milling and mixing facilities to provide balanced feeds for poultry and livestock; and (3) improved wheat milling and processing facilities, including equipmEmt for the manufacture of convenience foods. 3.67 Substantial investment will be needed to establish a feed milling industry which meets China's requirements. In the rural areas, where trans- port of large volumes of raw materials will continue to be difficult, live- stock producers are likely to grow substantial quantities of their feed. In this setting, large scale feed mills would be less appropriate than an indus- try organized around the manufacture of feed premixes (vitamins, minerals, some high protein supplements) which are relatively low volume matE!rials and can be purchased and transported by small- and medium-scale produce!rs to be mixed with their supplies of grains. For the large-scale animal oz~ poultry production systems, well served by roads and located near major urban markets, the modern, high capacity feed mill using large volumes of grains, protein supplements and feed additives, is likely to be the most efficient. 3.68 A viable feed milling industry can only be established if the product is demonstrably superior to current feeding materials and justifies the higher cost. In China measures which would contribute to this objective - 57 - include development (or expanded production) of synthetic amino acids, fish and bone meal, and vitamin, mineral and antibiotic supplements. The oilseed processing industry will have to increasingly regard the production of oilseed meal as of equal importance with edible oil. The quality and price of feed concentrates will have to be adjusted to the point where production is profitable and the use of protein meals as fertilizer becomes uneconomic. 3.69 To provide adequate services, China's agricultural processing and distribution system will have to be much more specialized, technically complex and more closely integrated than is the existing system. This modern system will require a high level of technical and managerial skills. Much of this expertise, or necessary training facilities, is not currently available in China. The experience of other countries in developing modern food distribu- tion and agro-industries would be useful and an active program of information exchange and study tours should be encouraged to permit China to benefit from this experience. Foreign technology, including managerial expertise, could make a large contribution to meeting these skill requirements. 3.70 Local management capacity can be augmented through joint-venture contracts with foreign agro-industrial firms. In addition to facilitating the transfer of improved technology, joint ventures often provide the means to break into a competitive export market. The advantages of such links in terms of market access, packaging technology, quality control and assured delivery, often are more important in the export market than being the low price supplier. 4. PROSPECTS FOR MAJOR AGRICULTURAL PRODUCTS Food Crops 4.01 Grain production in China increased by 3.9% annually in 1979-83 as the annual yield increases of 5.1% more than offset the decline in sown area of 1.1% per year. The most rapid growth rates were reported for wheat (6.7% p.a.) and rice (4.1%). Corn production rose by 3.2% per year while out- put of other grains (pulses, soybean, tubers, other coarse grains) rose by only 1.2% per year. Looking forward, the Planning Bureau within MAAF has targeted grain production to grow by about 2% annually from the average 1980-82 base while the area sown to grains would continue to decline marginally by about 0.1% p.a. This would require average yield gains of 2.1% yearly over the two decades, with individual rates ranging from 1.3% annually for grains other than rice, wheat and corn to 2.8% annually for corn (Table 4.1). Rice yields would increase by only 1.8%. - 58 - Table 4.L~ PROJECTIONS OF GRAIN PRODUCTION, AREA AND YIELDS BY THE PLANNING BUREAU, MAAF Ave:rage annual 1980-82 Pro- growth rate (%) average 1983 jections 1980-82 (actual) (actual) 2000 to 2000 1979-83 Sown Area (million ha) R1ce 33 33 33 0.0 -0.5 Wheat 28 29 28 -0.1 -0.3 Corn 19 19 19 -0.2 -1.7 Others 34 33 33 -0.1 -2.1 Total 115 114 113 -0.1 -1.1 Yields (tons/ha) R1ce (paddy) 4.4 5.1 6.2 1.8 4. 7 Wheat 2.1 2.8 3.5 2.6 7.0 Corn 3.1 3.6 5.3 2.8 s.o Others 1.8 2.1 2.3 1.3 3.3 Total 2.9 3.4 4.2 2.1 5.1 Production (million tons) R1ce (paddy) 148 169 208 1.8 4.1 Wheat 61 81 99 2.6 6.7 Corn 61 68 98 2.6 3.2 Others 59 69 75 1.2 1.2 Total 329 387 480 2.0 3.9 Source: Derived from projections of aggregate growth rates provided by the Planning Bureau, MAAF. Other grain includes soybean, lentils, and tubers. As noted in Chapter 3, these figures are considered to understate the actual sown area and overstate yields. 4.02 Rice. MAAF's Planning Bureau assumes that the area so~1 to rice would remain virtually unchanged from current levels, with the roughly 40% production increase (2000 over the 1980-82 average) coming entirely from higher yields. A major factor expected to contribute to increased yields is the projected expansion in the area planted to hybrids from the 1983 level of nearly 6.8 m ha to 13.3 m ha in 2000. With a net yield advantage (taking into account the larger areas required for seed production) of about 20% over con- ventional varieties, an increase from current levels of the hybrid area by the implied 7.7 m ha (wihout yield increases from 1982 levels for all rice), would generate nearly a third of the projected increment in rice production by 2000. The Government also plans to stress measures to improve rice yields in low- yield areas where considerable potential seems to exist. 4.03 The major constraint to area expansion for hybrids is the relatively long maturation period for current hybrid varieties in China (at least 135 - 59 - days), a fact which limits hybrid rtce largely to the second crop in rice-rice sequences or the single rice crop in wheat-rice sequences in the South. Although not yet in commercial production, promising materials are being tested in China which have maturities of 110 days or less. The successful introduction of these varieties in China over the next few years, which seems possible from technical and organizational perspectives, would permit exten- sion of the hybrid rice area and possibly a cropping system which includes two crops per year of hybrids. China's hybrid rice varieties were developed from a narrow genetic base, and remain somewhat vulnerable to pests and disease, but research efforts in recent years have successfully incorporated much greater resistance in the newer varieties. 4.04 Progress also has been made in reducing hybrid seed costs by improving production technology. Hybrid seed yields in China have been increased on average from about 500 kg per ha in the mid-1970s to 1.0 ton at present. Yields of up to 1.5 tons have been reported in the better areas and genetic materials are available with physical characteristics (long stigmas and anthers, shorter flag leaves) which should permit further increaes in seed set and yields. The cost of hybrid seed in China is up to ten times the cost of conventional seed, although seeding rates are less than 20% of the non- hybrid types. Thus, hybrid rice cultivators spend somewhat more on seed than do cultivators of other conventional varieties. Because conventional seeds are usually produced by the farmer, cash costs for users of hybrid seed are substantially increased. With a yield advantage for hybrids of 0.75-1.0 ton per ha (valued at $100-120), the incremental benefit-cost ratio of about 5:1 (abstracting from incremenal costs of labor and other purchased inputs, which also would be expected to rise), suggests that hybrid rice is economically attractive to producers. Consumers appear to respond less positively. In markets where both hybrid and non-hybrid varieties are available, hybrid ri4:e sells for about 10% less than traditional varieties. 4.05 China's rice yield in 1983 of 5.1 tons (paddy) per sown ha (about 4.1 t/ha if adjusted for acreage underreporting) lies about halfway between the average for all developing countries and yields in developed countries, but yields are constrained by the need to use short maturity varieties in double cropping systems. Water control in most of China's rice-growing areas is already quite good and fertilizer levels are at relatively high levels. Cultural practices, such as weed control, are outstanding. A major determi- nant of future yield gains is likely to be the speed at which hybrids with shorter maturation periods can be introduced. The steadily-improving syste1n for rice research in China augurs well in this regard. Targets for production and yield increases seem modest relative to recent achievements. 4.06 Wheat. China is among the world's largest producers of wheat. In China whea~the second most important grain after rice and accounts for about 20% of grain production and 25% of the area sown to grain. MAAF's Planning Bureau projects both wheat and corn production to grow more rapidly (2.6% p.a.) than other major grains. With an assumed small decline in the sown area, this implies annual growth in wheat yields of 2.6% p.a. to 3.5 tons per ha. China's 1980-82 wheat yield per ha of 2.1 tons (1.7 t/ha, adjusting for underreported acreage) is close to the average yield for all developing countries and well-below yields in developed countries. Among developing - 60 - countries, Mexico, South Korea and Egypt have yields higher than China, due 1n in part to the fact that virtually all of the wheat area in those countries 1s irrigated, as contrasted with a figure of perhaps SO% in China. 4.07 About 60% of China's wheat area is double-cropped, generally with corn (in the North) or rice (in the South).~/ Nearly all winter habit wheat on the North China Plain has been double cropped with corn. Most intermediate habit wheats in the Yangtze Valley and spring wheats south of the Yangtze are double-cropped with rice. The spring ~heats of the Northeast, Northwest and Southwest are mostly single-cropped.~/ Although the area sown to wheat is projected to decline marginally, some increase in the double-cropped area can be expected if the Government's efforts are successful to increase the intensity of land use through expansion of the irrigated area. 4.08 Development of early maturing varieties is changing these double- cropping patterns and has important implications for production. A case in point is the cropping system in parts of the Yellow River where, prior to the 1980s, available varieties of cotton were of a duration which required early spring planting, thereby precluding double-cropping with winter wheat. Cotton breeders recently introduced cotton varieties for that area with maturation periods of less than 120 days, or a duration sufficiently short to permit double-cropping with winter wheat. One planning estimate suggests: that short season cotton varieties can be combined with winter wheat on up to 1.7 m ha previously idle during the winter months. At yields of about 3.0 tons per ha, this development could add 5 million tons of wheat (6% of 1983 production) with no increase in cultivated area. 34/ Multiple cropping refers to three practices: (a) sequential cropping--the sequential growing of two or more crops on the same land in the same year; (b) relay cropping--a form of double-cropping in which a second crop is planted between rows of the preceeding crop. In China this is widely practiced in a wheat-maize cropping system in which maize is seeded between the wheat rows about 30 days before wheat harvest; and (c) intercropping--two or more crops planted in the same field at the same time, but in alternating strips. Intercropping is widely practiced in Northeast China for various crops, including wheat. 35/ Wheat breeders define winter wheats as varieties which require a resting period induced by cold weather to flower and set seed. Spring wheats require no resting period and are generally planted in late winter after the last killing frost. Intermediate wheats must demonstrate cold tolerance, but require no resting period. About 60% of China's wheat crop is of the winter growth habit; 20% is of the spring habit; and 20% is intermediate-type wheat. In contrast to these definitions, the SSB defines winter wheat as that which is in the ground during the winter months, thereby including wheats grown near Beijing as well as in Guangdong. Wheat breeders would define the varities grown in the South as intermediate types. - 61 - 4.09 Cultural practices in wheat are generally of high standard, although a number of improvements seem possible. Weeds are controlled by hand pulling and the measure of control is quite high. Little is known of the rate of fertilizer applied to wheat, but for this crop, like others in China, applica- tion levels for phosphorus and potassium are likely to be well below economi- cally-optimum levels, thereby losing some of the benefits from applied nitrogen. The degree of water control and the efficiency of water use could be improved in many of China's wheat growing areas. Seed quality also could be improved. Mechanization plays only a small part in soil preparation, seeding, fertilizing, threshing and transportation to storage or market. While this system employs abundant labor, it bears the disadvantage of slow turn around time between the June harvest (North China Plain) and the planting of corn, cotton or other warm weather crops. To speed these operations and avoid damage from summer rains, some mechanization of harvesting, threshing and hauling may be needed. 4.10 Work by Chinese wheat breeders has produced notable results in such areas as earliness (the best winter wheats in the Beijing area ripen by June 10, i.e. several days ahead of foreign winter wheats), successful spring X winter wheat crosses, and high yields (the world record wheat yield of 15 tons per ha was realized under optimum conditions at the Qinghai Plateau Institute of Botany in 1977). All varieties released for commercial produc- tion since 1965 have yield potential of five tons or more and two varieties released in Shandong Province have a yield potential of about 9 tons per ha. However, the gap between average provincial yields and yields realized in experimental and demonstration trials remains large. Provincial yields are generally less than half the yields achieved in trials, suggesting both th1e need and opportunity for improved practices relating to disease and pest control, irrigation and fertilization and timeliness of seeding, weeding, irrigation and harvest. 4.11 A particular concern for the future relates to wheat diseases, especially scab which may cause annual losses of 10-15% in the Yangtze Valley and losses of up to 25% in that area in two years out of ten. The most severe losses seem to be associated with heavy spring rainfall in localities with poor drainage. Through joint efforts between Chinese scientists and the international agricultural research centers (!ARCs), scab resistant varieties are being identified or developed, but the process may require five to ten years to produce results of economic importance. It is possible that wheat research in China has not paid sufficient attention to developing disease resistance. Wheat losses due to disease may be higher than is commonly recognized in China. Good progress has been reported in dealing with the problem of stripe rust. 4.12 The target wheat yield for 2000 of at least 3.5 tons per ha is likely to require progress on several fronts. Breeding programs might use- fully stress greater disease resistance, using local and exotic germ plasm available from !ARCs. Some wheat varieties in China lack plumpness of kernels, a characteristic which is manifested in lower yields, but which can be ameliorated through careful selection of parent materials. ··Lodging also remains a problem, especially in areas with high levels of fertilizer use. It - 62 - may be necessary for wheat breeders in China to make larger number of crosses each year to identify a few outstanding progeny with stiff-strawed attributes.36/ Table 4.2: GAPS BETWEEN PROVINCIAL AND TRIAL YIELDS OF WH:EAT Provincial Provincial average as Trials at provincial average Minimum % of trial research institutes Year Yield yield gap yield (t/ha) (t/ha) (t/ha) (t/ha) Shijiazhuang, Hebei 1977 5-7 2.2 2.8 44 Jinan, Shandong 1977 7-9 2.6 4.4 40 Nanjing, Jiangsu 1977 5-7 3.6 1.4 72 Gongzhuling, Jilin 1977 5 1.0 4.0 20 Guangzhou, Guangdong 1981 4 0.9 3.1 23 Taigu, Shanxi 1981 5 1.7 3.3 34 Wugong, Shaanxi 1983 4-5 2.3 1.7 42 Source: SSB and reports by visiting scientists. 4.13 Agronomic research, particularly that related to fertilizer and water use, has been undervalued and, hence underemphasized in China. It would seem useful to undertake more complete fertilizer trials to examine production and economic effects of using alternative types of nitrogenous, phosphatic and potassic fertilizers and of varying the timing and quantity of applications. Further applied research would be useful on effects of alternative timing, amounts of water applied and number of irrigations. As part of the Government's emphasis on crop development in rainfed areas, more research 36/ Visiting scientists have stated that the germ plasm base in Chinese wheat varieties is unduly narrow. This may be incorrect. Through the IARCs, China has access to the world's germ plasm collections, but brteeders may not be using effectively the full genetic range of this material. - 63 - seems needed in the semi-arid areas of the North~est, including moisture conserving techniques such as minimum tillage.lr 1 4.14 Corn and other coarse grains. Corn production accounts for about three-quarters of coarse grain output and 18% of total grain output. Produc- tion since 1979 expanded at 3.2% p.a., or about one-half as fast as wheat. The Government projects virtually no change in the cropped area to 2000, but yields are to grow by about 70%, or 2.8% p.a., 1979-81 to 2000. China's corn yields (adjusted for acreage underreporting) are slightly higher than average yields in developing countries, but less than half those in developed coun- tries. Much the same pattern holds for other coarse grains~ In comparing wheat, rice and corn yields in China with those achieved in developed coun- tries, the largest difference is in corn. One explanation is that corn in China is usually relay cropped following wheat and therefore may suffer from a shorter growing season and initial competition with wheat for nutrients and moisture. Other factors restraining yields of coarse grains include poor see~d quality, inadequate irrigation during times of moisture stress, low applica-· tion of chemical fertilizer and poorer cultural practices compared with e.g., wheat and rice which, as principal food crops, benefit from greater care by producers. Although hybrid corn accounts for about 70% of total corn area, quality of much of this seed is questionable. Deficiencies of nitrogen and phosphorous are commonly observed in growing corn. 4.15 Growth of total coarse grain production at a substantially higher rate than the Government target would be required to provide the feed grain necessary to meet output targets for livestock and poultry. Much of China's development effort in grain has focused on the high-yield irrigated areas. One of the major challenges of the future is to raise crop productivity and farmer incomes in upland areas. Much can be done with the traditional grain crops, but other crops such as cassava also have great potential in some of these less-favored areas. Cassava's potential lies in its multiple-use character and its efficiency as a converter of soil nutrients, sunlight and moisture to carbohydrate energy. With proper inputs, further expansion of this crop may be possible on lower-quality tropical soils not suitable for sustained cropping of annual food crops. Output could be used to meet energy requirements in the local market for mixed feeds. Exports also may be possible to net importers of feed grains, starches and sweeteners. 4.16 Regional Trends. The 6.4% overall increase in grain production in 1979-82 reflects rapid rates of increase in east and south China and near- stagnation in the north (Annex A, Table A.S). There was a modest production increase in the northwest, a significant crop loss in the northeast due to flooding in Heilongjiang, and a decline in acreage in low-yielding coarse grains in the north offset by increased yields for fine grains. Southern and eastern production increases extended to most types of cereals, tubers and 37/ Some countries such as Turkey have been very successful in expanding dryland production of wheat. Wheat production in Turkey expanded by 3.9% p.a. over the past 20 years. The experience of Turkey may be useful to China. - 64 - soybean. Reportedly, 80% of the 1983 grain production increase averaging 7.5% was attributable to provinces in the north, restoring regional balance to the growth pattern. 4.17 The Government does not project any major changes in the regional pattern of the area sown to grain. At present, only the northwest: has an overall grain deficit, which the Government hopes to reduce by developing commercial grain bases within the region (Map 4). Lower projected growth rates for rice relative to wheat and corn would imply higher growth rates for grain production in the north. Within the south, there may be a decline in tuber and rainfed rice area and corresponding increase in corn andl soybean area, mostly in the hilly or mountainous sections. Broadcast (rather than paddy) rice is expected to expand in the north and two-crop paddy rice area 1n the south. Industrial and Specialty Crops 4.18 Measured by percent of sown area, national average yields and per capita production, the industrial crops (oilseeds, fibers, sugar crops, tobacco, tea) on the whole appear to be less well developed in China than in a number of comparator countries (Appendix B, Tables 8.1 and 8.2). Cotton 1s a significant exception. The demand for most of these crops tends to grow rapidly as incomes rise. Some of the specialty crops, such as fruit and vegetables, bring high returns to producers and typically become more important as agriculture modernizes. Long-term Government plans stress improvement of yields of these crops, in part by careful selection of planting location to take advantage of suitable environmental conditions, but do not call for significant expansion of the planted area. 4.19 One of the issues in a longer term perspective 1s the extent to which expansion of these crops in China could be economically justified. An international comparison of average yields of these crops relative to grain yields may provide a rough measure of comparative advantage (see Appendix B for details). If average yields are indexed to cereal yields and compared, the most important producing areas in China typically appear to be at least as efficient as comparator countries in the production of oilseeds, fibers and tobacco. Chinese producers appear to be less efficient in, e.g., the sugar crops. These indices may also be taken as rough measures of production inefficiencies in these crops and, therefore, of potential for improvement. - 65 - Table 4.3: INTERNATIONAL COMPARISON OF INDUSTRIAL CROP YIELDS (1979-81' tlha) Cereals Peanut Rape Seed Jute Tea Tobacco Sugar Sugar seed cotton beet cane China Ia 3.0 1.5 0.9 1.3 3.5 0.3 1.9 12.8 47.8 Korea 4.8 1.5 2.1 0.9 9.0 0.3 2.1 Japan 5.3 1.9 1.7 1.5 1.5 2.4 52.2 64.6 India 1.3 0.8 0.5 0.5 1.2 1.5 1.1 26.1 51.8 Pakistan 1.6 1.2 0.6 1.0 0.8 1.5 37.9 Thailand 1.9 1.1 1.4 1.1 0.6 37.0 Indonesia 2.9 1.6 0.7 2.9 1.1 0.5 98.7 Egypt 4.0 1.6 2.7 2.3 83.9 Index of Yields Relative to Cereals: China 100 50 30 44 177 10 63 427 1,593 (Leading province) lb 100 73 43 68 160 NIA 104 836 1,896 Korea 100 31 44 19 0 6 44 0 0 Japan 100 36 32 0 28 28 45 985 1,219 India 100 62 38 38 92 115 85 2,008 3,985 (Leading state) lc 100 87 37 10 66 101 64 NIA 3,644 Pakistan 100 75 38 63 50 0 94 0 2,369 Thailand 100 58 0 74 58 0 32 0 1,947 Indonesia 100 55 0 24 100 38 17 0 3,403 Egypt 100 40 0 68 58 0 0 0 2,098 Ia Cereals yield according to FAO definition (excluding tubers and pulses). Chinese yields are likely to be overestimates due to underreported acreage. lb In 1981, province with largest acreage in each crop. Tc In 1977178, yields of rice adjusted to those of paddy. 4.20 Cotton and other fiber crops. Special encouragement given to cotton production by price and procurement reforms, the increased freedom to deter- mine cropping patterns under the PRS, and the planting of improved varieties stimulated a 22% expansion of cropped area and a 72% increase in yields during 1978-83. In Shandong Province, the leading producer, the area planted to cotton rose by 150% between 1979 and 1983 and stimulated some concern of increased grain deficits for the province. As a result of these area and yield gains, the cotton production and trade situation has undergone a remarkable transformation. China has moved from being a large importer to basic self-sufficiency (Table 4.4). With cotton consumption presently below - 66 - production, imports have declined sharply to less than 100,000 tons1 per year and stocks may be approaching two million tons (equal to about half a year's consumption). Table 4.4: COTTON PRODUCTION, CONSUMPTION AND TRADE, 1980·-83 Con- End of Year Production Imports sumption year stocks mil tons ------------------------ '1980 2.7 0.7 3.3 0.6 I1981 3.0 0.5 3.5 0.6 1982 3.6 0.2 3.6 0.7 1983 4.6 0.1 3.6 1.6 Source: 1980-82 data are from SSB. The 1983 figures are from the United States Department of Agriculture. 4.21 Prospects for continued yield increases for this crop seem good for several reasons. First, the yield gap between average and best practice production is large. In Shandong, where yields of seed cotton have reached an average 2.0 tons per ha, cotton specialists believe that a further 50% yield increase could be obtained with existing varieties through extension of improved cultivation practices and better pest control. Second, cotton, along with vegetables and peanuts, is benefiting from the use of plastic sheet (93,000 ha in 1982), a technology which permits earlier planting, conserves moisture and facilitates weed control. This technique, which reportedly leads to cotton yield gains of more than 50%, appears profitable in labor-surplus regions, although the required cash outlay is a limiting factor. Finally, the CAAS Cotton Research Institute recently released a new early maturing (112-114 days) variety of summer-sown cotton which can be sequentially-cropped with winter wheat. Although yields and quality of this variety are likely to be less than those of longer-maturity varieties, introduction of wheat-cotton double cropping in both the Huanghe and Yangtze River basins may help ease the fear of grain deficits associated with increased cotton acreage. This variety is targeted in the near term for nearly 30% of the total cotton area. 4.22 Future consumption of cotton will depend increasingly on the size of the local market for textiles. In 1983 sales of textile products d,eclined and cotton rationing was abolished. Exports of cotton textiles, which provide a market for less than 10% of production, are likely to depend on China's ability to improve quality and styles. Competition with synthetic textiles also may become increasingly keen as the large new capacity for synthetic tex- tile production comes on stream. Some exports of raw cotton may be possible and have been included in trade agreements with East European countries. - 67 - 4.23 In China the demand for bast fibers such as jute and hemp has declined rapidly due to the availability of cheaper and more serviceable synthetic substitutes. Following a 50% yield increase in 1978-82, domestic consumption requirements have been met and sown area has declined sharply. Prospects are that the area sown to these crops will continue to decline as China's synthetic fiber industries mature. 4.24 Oilseeds. During 1978-82 production of oilseed crops (including rapeseed, peanuts, sesame and sunflower seeds, but excluding soybeans and cottonseed) more than doubled, due about equally to increased area and yield. Area expansion was partly due to resumption of specialization patterns abandoned in the 1960's during the drive for local grain self-sufficiency. New patterns also emerged such as the rapid adoption of sunflower in the well suited areas of the Northeast. There is also evidence that oilseeds received larger supplies of chemical fertilizers. As a consequence of these develop- ments, the previous extreme shortage of edible oils was eliminated and rape- seed oil grew rapidly to nearly one-half of edible oil supplies. 4.25 Soybean production has grown by about 5.2% p.a. since 1978, largely due to increased yields. Acreage has shown a small uptrend. Variable weather in major production zones resulted in sharp inter-year fluctuations in output. Soybeans appear to have been benefited less than the major grains from research and relatively small amounts of chemical fertilizers are allocated to this crop. Soybeans are the only major crop grown in China in which the reported national average yield was less than that for developing countries as a whole. Chinese yields (adjusted for acreage underreporting) are at least 20% below average yields in developing countries and 40% below those in developed countries. A steady increase in soybean yields would be required if futujg;needs for high quality protein meal for livestock feeding are to be met.-- 4.26 Processing of oilseeds could be much improved. Most existing facilities are small-scale operations using inefficient machanical presses and are characterized by low extraction rates, high levels of impurity in the product, and inability to remove toxic substances in rapeseed and cottonseed oil and meal. Less than 20% of China's oil seed processing capacity uses the more efficient solvent extraction methods. 38/ Soybeans originated in the temperate areas of Northeast China and most of the world's production takes place at latitudes more than 30 degrees from the equator. Efforts to develop tropical soybeans have been underway for a number of years in various countries. Researchers in Brazil recently released for commercial production three varieties of tropical soybeans suitable for planting within 15 degrees of the equator. Yields may average about 1,850 kg per ha, compared to about 1,100 in China, or 2,000 kg in the United States. Adaptation and introduction of soybeans into China which perform well at lower latitudes could improve the outlook for high quality protein meal and, indirectly, the livestock industry. - 68 - 4.27 Other industrial and specialty crops. These include sugar cane and beets, vegetables, fruits, tobacco, tea, coffee and various tree crops. Per capita production of these crops as a group is low in China relative to mo~t comparator countries. Income elasticities of demand typically are high. 391 Although the projected growth rate of demand suggests that the area sown to several of these crops should expand over the next several years, caution may be required with crops such as sugar cane and sugar beets where the evidence noted earlier suggests that China has little overall comparative advantage and domestic procurement prices exceed world prices by substantial margins. For these crops the primary emphasis probably should be to improve yields through better cultural practices, improved planting materials and careful choice of .planting location. It may be that overall efficiency gains can be made by replacing these crops in some areas with crops more suited to the environment. 4.28 Prospects for tea are clouded because of possible marketing constraints at home and abroad. Tea production in 1983 declined by 3.8% from 1982 levels, the first decline in 20 years, reflecting smaller sales at home and abroad. In the major producing province of Zhejiang, 1,000 ha of tea reportedly were converted to other crops because of a difficulty in selling the output. Proposals in China to decentralize tea marketing and permit free market sales would probably encourage larger domestic consumption. Average tea consumption in China of about 150 gm per capita per year is the lowest among tea producing countries and far behind the world average of 430 gm. The World Bank projects the world price of tea to show no change in real terms between the present and 1995 (Appendix Table A.l2). 4.29 Production of tobacco is likely to be increasingly influenced by public perception in China of the health dangers inherent in its use. Chinese authorities in the past have been reluctant to acknowledge that increased cigarette smoking may be linked to lung cancer. More recently, smoking has been increasingly recognized as a health hazard, although trends continue to show very rapid growth in tobacco consumption. In 1981 the retail price of cigarettes was increased by about 30% and limits were placed on th'e land allocated to production in response to the emergence of surplus leaf supplies. Tobacco production rose rapidly in 1981 and 1982 (to about 2.08 m tons) but declined in 1983. 4.30 Increased production of vegetables and fruit present special problems because of their perishability. A primary constraint in the develop- ment of these products is the weakness of the processing, transport and 39/ Vegetables at first glance appear to be an exception. Beijing consump- tion surveys suggest an elasticity as low as 0.20. Meat and fish are substitutes for vegetables in Chinese consumption habits. Re]placement of lower quality vegetables such as cabbage or mustard greens with higher quality vegetables and greater consumption during the off season also is associated with increased incomes. Both trends are associated with decreased crop yields. Hence, if the income elasticity for vegetables were defined in terms of value of expenditure, rather than quantities consumed, it would be more meaningful and also probably quite high. - 69 - marketing system. Particularly critical are the shortages of modern process- ing facilities and the lack of refrigerated transport and cold storage. Development of commercial production of perishable products (livestock products, fruit.s and vegetables) has been largely confined to the peri-urban areas partly in response to the transport and storage problems. However, the high and rising economic value for land near urban areas makes this a less viable solution. This concentration of production also contributes to income inequality within the agricultural sector, inasmuch as suburban producers reap the major benefits from high value production. With improved processing and transport facilities, production of perishable commodities could expand outside municipal boundaries. 4.31 Government plans give considerable emphasis to development of fruit and other tree crop production on lands too steep for annual crop cultiva- tion. The lag between planting of tree crops and first harvest implies that much of the supply of fruit between the present and 1990 has already been determined. Tree crop investment programs in the next few years will largely determine supply after 1990. 4.32 Potential for further specialization. Recent agricultural planning in China has been influenced by the view that a higher degree of specializa- tion, based on economic incentives and careful land use planning, is feasible and promises continued economic gains. Provincial statistics on sown area indicate the expansion of industrial crops since 1979, with the largest increases in provinces with relative comparative advantage. However, these increases in the industrial crop areas have come primarily at the expense of the area sown to grain. Few examples have been found of declining plantings in areas lacking comparative advantage in industrial crops. This may be related to marketing and transport deficiencies, which encourage self-suffi- ciency, and to a unified price structure which shifts much of the burden of differential transport costs from the producer to the marketing organizations and, ultimately, to the Government. 4.33 Without special studies or access to highly disaggregated data, it is difficult to draw conclusions about whether significant efficiency improve- ments are to be gained from further specialization. This discussion is there- fore limited to suggestive evidence from provincial data in the context of international comparisons. Appendix B examines the extent to which provincial cropping patterns are related to comparative advantage in China and India. Because provinces are large geographic entities with diverse micro-environ- ments, and because transport costs inevitably encourage local production, con- siderable production diversity is to be expected, and indeed is found, at the provincial level in both countries. However, the statistics suggest that pro- vincial specialization in particular industrial crops according to comparative advantage is more characteristic of China (1981) than India (1977-78), even though overall China devotes a smaller proportion of the sown area to such crops. This finding may be explained in part by the observation that many of India's farmers are closer to subsistence and such producers tend to use diversification as a measure to reduce risk. Further, China's rural produc- tion and welfare systems continue to -·reduce risk to the individual producer, permitting a higher degree of specialization. Unfortunately, the analysis provides no firm guidelines as to how far specialization could go if the I - 70 - weaknesses of China's transport and marketing system were remedied and the residual concerns for regional self-sufficiency in basic foods were relaxed. 4.34 Regional Trends. Increased production of industrial crops in the last few years has come primarily from specialized, commercial production bases (Map 4). Thus Shandong and Hebei Provinces account for 76% of increased cotton production in 1979-82, with traditional producing areas in provinces such as Xinjiang, Shanxi and Jiangsu accounting for smaller increases and southern area declining. Growth of oilseed crops has been broadly based, including rapeseed in all the provinces crossed by the Yangtze River, peanuts in Shandong, Guangdong and other provinces with areas of sandy soil, and sunflower in the northeast. Expansion of sugar cane and beet have also centered on commercial production bases. 4.35 In the Government's view, any further adjustments of industrial crop area should be marginal. Because southern cotton is of higher quality, only slight decreases in acreage should occur, whereas acreage in Xinjiang would further expand if transportation conditions permit. Emphasis in the north will be on quality improvement. Future locational trends for oilseeds are not expected to differ from those of recent years. Additional expansion of sugar cane is to be limited to that permitted by grain yield increases; sugar beet area would be increased, primarily in Neimonggol, Xinjiang and Gansu, in areas in which grain yields are low. Sugar beet, like sunflower, is especially targeted for areas with saline soils. In short, past patterns of industrial crop specialization have been largely restored, and future stability is expected. Forestry 4.36 Output and employment. Forestry's share in GVAO remains relatively small, but has increased steadily from about 1% in the early 1950s to just over 4% at present. Employment in forestry and forest industries totals 2.2 m (less than 1% of the rural work force), with 60% in wood processing, 30% in land preparation and planting and 10% in nursery operations, logging, etc. Industrial roundwood harvested and marketed within the state plan in 1982 was 50 million cu m, an increase of 2% over the 1981 level, but below the peak figure of 54 million cu m in 1979. (Appendix Table A.10: Timber Production, 1981). A volume of about the same magnitude was produced outside of the pian. Production of timber products includes sawn timber (2 million cum in 1981), plywood (0.35 million cu m), fiberboard (0.57 million cu m) and particleboard (76,700 cum). Paper and paperboard output was about 5.7 m tons in 1982. 4.37 The resource base. About 120m ha, or 13%, of China's land area ts under forest cover, with about a third of this in virgin forest, nearly half in secondary (natural) forest and the balance under replanted forest. Since 1949 some 86 m ha have been afforested but plantings on only about one-third of that area are reported to have survived. Survival rates in recent years have been higher. Total growing stock is estimated to be 9.5 billion cu m of which 55% is in coniferous species and 45% is in broadleaved types. About 21% of the forested area is in Heilongjiang Province alone and five provinces (Heilongjiang, Yunnan, Jilin, Guangdong, Sichuan) account for 48% of the total. In contrast, the five provinces with the smallest areas under forest I - 71 - (Ningxia, Qinghai, Nei Monggol, Jiangsu, Shandong) make up less than 2% of the total cover. Heilongjiang accounts for 24% of the country's growing stock of timber, while three provinces (Heilongjiang, Sichuan, Xizang) together make up 54% of the total. As a percent of provincial land area, forest cover ranges from 35% (Heilongjiang) to 0.3% (Qinghai). A significant feature of China's forest resource is the high proportion of the forested area and the standing volume of timber accounted for by mature forests (more than 30% and 60% respectively). This is due in part to the inadequate network of forest roads which restricts exploitation. 4.38 Consumption and imports. Annual per capita wood consumption 1s about 0.05 cu m (excluding fuelwood), compared to about 0.02 cu m in India and 1.5 cu m in the US. Reflecting local shortages and increased demand from population and income growth, imports of forestry products (logs, sawn timber, veneer, plywood) rose from US$242 m (CIF basis) in 1976 to a peak of US$642 million in 1980, then declined to about US$540 m in 1981. Between 1978 and 1981, pulp and paper imports rose from less than 600,000 tons (US$150 million) to 1.4 m tons (US$630 m). Expected income gains will alf81t certainly result in continued rapid growth in demand for forest products.-- Table 4.5: AREA AND VOLUME OF COMMERCIAL FORESTS Area Standing volume % of billion % of Age mil. ha total cu m total Not classified 0.2 2 0.2 2 Immature 3.8 39 0.5 7 Middle age 2.6 27 1.8 23 Mature 3.1 32 5.2 68 Total /a 9.8 100 7.7 100 /a Excludes 0.8 m ha in protective forests, 0.9 m ha in economic forests, - 0.3 m ha of fuelwood forests and 0.4 m ha in bamboo and other special use forests. Data refer to 1976. Source: Ministry of Forestry. 40/ Rural housing has been relatively neglected for much of the past 30 years. This is being corrected. Surveys suggest that the share of expenditures on housing by rural families in 1982 had tripled from 1978 levels. The housing boom resulted in a 32% increase in the average floor space per rural resident in 1978-82. I - 72 - 4.39 Future development. Forestry development in recent years has been accorded high priority by the Government but has been handicapped by depletion due to over-harvesting. Major thrusts of future development include affores- tation, increased exploitation of mature and over-mature areas, expansion of forest industries and strengthening of forest research, education and exten- sion. Government plans call for the land area under forest to be increased to about 20% by 2000, an increment of some 60 m ha. Annual harvesting of timber included within state plans is to increase from about 55 million cu m in 1983 to 100 million cu m by 2000. However, much of this would be accomplished by enlarging the scope of the plan to include most of an additional SO million cu m harvested annually, mostly for self-use by collective units and private farmers (not counting about 70 million cu m of firewood). "Real" production increases are not expected to exceed a few million cum, including under 1.5 million cu m from exploitation of mature forests and the product of some 200,000 ha to be developed as intensive tree farms. While total timber production volume would thus remain static, processing industries are to be rapidly expanded. For example, production of wood-based panels (now largely fiber board and plywood) may expand from 1.0 million cu m in 1980 to 3.5-4.0 million cu m by 2000 if the industry receives the Y 4.5 billion investment required. 4.40 The achievement of these objectives will require a very large effort by the agencies concerned. Afforestation of another 65 m ha would require planting of S-6 m ha annually with survival rates exceeding 60%. Some 4-5 million ha have been afforested annually in recent years, although survival rates reportedly have been less than SO% on average. The application of PRS in forest management is expected to increase this figure substantially. Funding of forestry development continues to be a problem. Although stumpage fees are paid in most areas, the revenues fall well short of development costs of the timber. Efforts are being made to shift a part of afforestation costs from budgetary grants to loans, although financial institutions such as the ABC have been wary of this type of lending operation because of the long gestation period involved. Investment programs by the state forest farms, of which there are some 3,900, have been hampered by the relatively low price they receive for logs sold to the Government. Despite recent increases, state procurement prices for timber are only about one-half the negotiated price. New policies announced in late 1984 permit timber produced by collectives or on private plots to be sold at negotiated or free market prices. Livestock 4.41 Livestock's share in GVAO was virtually unchanged during the 1979-83 period, although since 1979 growth of this sector has been faster than that of crops. Exports of live animals and livestock products exceeded $1.1 billion in 1982 and accounted for about 5% of China's total export value that year. Exports of live hogs and chilled, frozen and canned pork are in the range of Y 1 billion annually, a value 50% greater than 1982 coal exports and larger than tea and raw silk exports combined. Discussion of the livestock sector is conveniently divided into ruminants (cattle, sheep, goats, etc.) and non- ruminants (pigs and poultry). China has the largest swine population in the world (some 300 million head, or more than 40% of the total), but only about 5% of the world's cattle, 12% of its sheep and 18% of the horses. The I - 73 - country's ruminant herds and flocks utilize some of the world's most extensive grasslands (Map 6), variously es4i7ated at 319 m ha to 340 m ha (of which 225 m ha are defined as usable)._! Less is known of poultry and dairy production, although modernization of those industries is beginning. Red meat production (pork, beef, mutton) increased by about 12% p.a. in 1978-82, per- mitting per capita meat availability to grow from 8.9 kg to 13.3 during the period. Poultry meat may add about another kg to annual per capita consump- tion. More than 90% of total meat production (including poultry) is accounted for by pork. 4.42 China's agricultural area is divided into grazing lands (55% of the total) and crop lands (45%). The grazing areas are lightly populated by humans (6% of the total), but provide about half of the mutton production and 40% of the beef output. The crop lands support more than 70% of the large animals and almost all of the poultry and pigs. Patterns of livestock ownership and management have changed sharply since 1981 as a result of the introduction of the PRS. More than 90% of all hogs and a larger share of poultry are privately owned. In the grazing areas livestock previously managed collectively have now been allocated to individuals under sometimes vague agreements which require repayment of live animals to the collective. This system continues to evolve and is resulting in a sharp increase in livestock numbers, with accelerated pressures on already over-grazed lands. In the cropped areas, most large animals also have been allocated to individual households or specialized production groups. 4.43 Management systems tend to minimize feeding of grain and emphasize use of milling by-products, other processing wastes, cut forage produced on waste or marginal lands or natural grasslands. Concentrate consumption in 1981 was probably equivalent to about 37 m t of processed grain and 58 m t of grain and oilseed milling by-products. Most of the poultry flock depends on scavenging, although confined feeding systems are beginning to appear near larger towns and cities. Low grain feeding systems and seriously inadequate levels of protein supplementation are major factors contributing to various technical inefficiencies in China's livestock sector. Dressing rates for hogs and cat417 are only about 70% and 40%, respectively, of levels in the United States.~ Slaughter rates in the United States are more than twice in China and the ~~riod from farrow to finish in China is two to three times that in the us.~/ Milk production per cow in the US exceeds the Chinese figure by a similar margin. But these production systems are not without merit. The ability of native breeds of swine to ingest and utilize large quantities of low quality roughage, aquatic plants and agricultural processing wastes should be considered when comparing production efficiencies. The swine population 41/ The SSB gives a grassland area of 319 m ha, of which 225 m ha are "usable." These figures compare with rangelands of about 455 m ha in Australia, 370 m ha in the United States and 140 m ha in Argentina. 42/ The ratio of dressed weight to live weight. 43/ Annual slaughter as a percent of inventory numbers. - 74 - acts as a means of converting low value organic materials into meat and valuable organic fertilizer which has played a key role in maintaining soil tilth and fertility in China. 4.44 Ambitious targets have been established for livestock production as part of the Government's efforts to improve consumption patterns. By the year 2000, the contribution of the livestock sector to GVAO is projected to rise from the current level of about 15% to 25%. Meat consumption is to more than double to about 27.5 kg per capita (33m tons) by 2000, while the share of Table 4.6: A COMPARISON OF SELECTED ASPECTS OF LIVESTOCK PRODUCTION IN CHINA AND THE UNITED STATES, 1980 Item China United States Ratio of value of livestock to crop production 18/82 48/52 Year-end inventories (million) Hogs 305.4 67.4 Cattle 71.4 111.2 Dairy cattle 0.6 10.9 Sheep and goats 187.3 14.2 Animal product output (million tons) Pork 11.3 7.5 Beef 0.3 10.0 Mutton 0.4 0.1 Poultry meat 1.0 6.6 Cow milk 1.1 58.3 Per capita consumption (kg/yr) Red meat 12.3 72.6 Poultry meat 1.0 27.7 Eggs 2.5-3.0 15.7 Milk 1.4 103.2 Efficiency indicators Farrow to finish for hogs (months) 18-24 6 Slaughter rate, hogs (%) 62 143 Slaughter weight (kg) 94 110 Dressed weight (kg) 54 79 Dressing rate (%) 57 72 Pork production/head in inventory (kg) 37 111 Milk production/cow (kg) 1,800 5,300 Source: Adapted from USDA, China: World Agriculture Regional Supplement, 1983, P• 15. - 75 - pork in the total (pork, beef, mutton) would decline to 70% (23m tons). This implies that production of beef and mutton would increase from 680,000 tons (1979-81) to 9.9 m tons (14.2% p.a.), while pork output would expand by 3.7% annually. Production targets for other livestock products in 2000 include: eggs 13.8 kg per capita, wool about 1 kg per capita and milk 42.5 kg per capita. These figures imply a 5.5-fold increase in per capita egg consump- tion, a 7-fold increase in wool consumption and a more than 15-fold increase in milk consumption. Assuming these products are locally-produced, these figures imply annual growth in output ranging from more than 11% in the case of eggs and wool to more than 20% for milk. 4.45 Major means proposed to achieve the livestock production targets include more efficient production and management systems, based on higher levels of grain feeding, pasture improvement and use of animals with higher genetic potential. With these improvements, the Government projects FCRs to fall, and slaughter rates from herds and flocks to increase. Over the longer- term, slaughter rates are to increase from about 5% at present to 25% for cattle, 68% to 100% for hogs and 15% to 60% in the case of sheep. 4.46 Technical feasibility of the livestock targets depends critically upon the availability of much larger supplies of grass, grain and protein supplements. Major efforts will be required to solve the institutional and logistical problems of developing and servicing a modern livestock industry, based on nutritionally-balanced rations, comprehensive veterinary services, and sustainable systems of forage and fodder production. Supplies of coarse grains would fall short of requirements if production expands no faster than it has in recent years. 4.47 Inadequate supplies of high quality protein meals are perhaps the major causes for the low productivity levels in much of China's livestock industry. Although China produces some 8-10 m tons of oilseed cake and meal annually, at least 70% of this amount has been used as fertilizer. Factors contributing to the heavy use of these materials as fertilizers include local shortages of chemical fertilizers in situations where oilseed cakes are readily available, relatively low prices for oil cakes, transportation bottle- necks, and the ~resence of toxic substances in cakes which limits their use in animal feeds.~ 7 Solution of the toxicity problems in China's oilseed meals would greatly enlarge potential supplies of protein meal. However, soybean 44/ Gossypol pigments in cottonseed meal have generally limited use of this meal to feed for ruminant livestock and for fertilizer. Because of its toxicity and high fiber content, only small quantities of China's rapeseed cake can be used in feeding of non-ruminant livestock and poultry. The oil from traditional rapeseed varieties typically has an erucic acid content of at least 25% while the meal contains 6-18% of glucosinolates. Improved varieties, developed in Canada and Europe, contain no more than about 2% erucic acid in the oil and less than 2% of glucosinolates in the meal, which is therefore usable in livestock feeding. Work is underway in China and elsewhere to develop cotton varieties with lower gossypol content. - 76 - meal is the most appropriate meal for poultry and swine rations b;ecause of its amino acid balance and its relatively high content of the essential amino acid, lysine. In the absence of adequate lysine supplies for pork and poultry feeding, it may be useful for China to consider developing a capacity to produce synthetic lysine, as other countries have done in similar circum- stances. Priority should also be given to strengthening research programs aimed at introducing oilseed varieties with low toxicity levels. 4.48 The grasslands of the North. China's extensive grazing areas (Map 6) are commonly viewed as underutilized resources which can be the source of much of the large incremental supplies of beef and mutton in future years. The potential value of this resource is indisputable, but the extE!nt to which it can support larger livestock populations under present policies is open to question. Available evidence suggests that many of the grazing areas in Northern China are already seriously over grazed and in danger of progressive and g5~haps irreversible degradation unless corrective measures are implemen- ted.--/ Pasture conditions in the southern and southwestern areas; of China are generally in better condition, but productivity there also tends be low because of poor soil fertility and low feed value of natural vegetation. Inadequate forage and fodder supplies, particularly during the stressful winter months, are major factors behin~ }ow slaughter rates and winter mortality rates in the range of 7-15%.~ Lambing and ~alving rates of no more than 70-80% are reported in some of these areas.~/ Livestoc:k productivity in the northern grasslands is low because most of the: available dry matter is required for animal maintenance and little remains for production. 4.49 Overgrazing problems in areas such as Nei Monggol and northern Hebei go back to the 1960s, but have been exacerbated by policies since 1979 which stress expansion in numbers and the rapid introduction of the PRS. Herds and flocks, formerly managed collectively, have been allocated to individual households or specialized production groups. While grasslands also are even- tually to be divided up for individual use, in many areas they remain in use as commons. The PRS also has significant positive effects. It has encouraged individual care of animals, with the result that mortality rates in some areas are down. But it appears to have destroyed the mechanism by which collectives formerly were able to control animal numbers, at least to some extent, and to implement pasture management systems. The combination of privately-managed 45/ For example, large animal numbers reportedly are growing by 4:t p.a. in Gansu's grasslands and the numbers of goats and sheep, while declining, are estimated to exceed carrying capacity by at least 25%. 46/ Winter losses in the grazing lands of the American West, a region with ecological similarities to parts of China's northern grasslands, are generally less than 2% p.a. 47/ Ratio of lambs/calves to number of breeding females. - 77 - herds on a common gi3'ing resource has added pressure to an already over- burdened rangeland.-- 4.50 The policy of encouraging livestock numbers appears to be based on inadequate appreciation of the existing condition and potential of much of China's grassland area. The most important immediate requirement is to match livestock numbers with carrying capacity through herd reduction. Controlled grazing and the introduction of improved grasses and legumes could permit higher dry matter yields and ultimately higher meat offtake. Failing this, other elements of an improvement program may not be worthwhile. Measures to improve genetic potential of the livestock population without first increasing feed supplies will not increase animal productivity and might in fact lead to higher mortality rates for the less hardy improved breeds. The economics of introducing irrigation to pasture or forage in the short season areas of north China, or of permanent fencing of rangelands, are open to question. 4.51 Policy objectives should move from those of expanding animal numbers to increasing the output of animal products per unit of land area. Destocking is likely to be required, but herdsmen are not likely to reduce animal numbers unless they can be convinced that it is in their economic interest to do so. Demonstration farms and ranches, backed by sound technical research and management, will be required in this regard, as will successful implementation of the grassland laws which are to shift the management of grassland to indi- viduals or groups who would benefit from better management. It also seems useful to offer price premia for younger animals to encourage destocking. Many of these required changes are socially complex and institutionally difficult. In view of these difficulties and the existing condition of much of China's grazing areas, it seems unlikely that the large targeted increments of meat from beef and mutton are likely to be achieved. 4.52 Ruminant livestock in South China. Although the focus of ruminant livestock development in China has been on the pastoral areas of the North and Northwest, more than 40% of the large animal population, including about half of China's cattle, is located in twelve southern provinces which are con- sidered as "agricultural," rather than "pastoral." The demand for draft animals has been increasing and will remain strong for several years. As the basis for ruminant livestock development, attention has been called to the hilly and mountainous areas as the only underutilized land remaining in South China. In the past, indiscriminant clearing of forest and expansion of cropping has contributed to a serious erosion problem. Much of this cut over area is under grass and is used, or usable, for pasture. 48/ The comment of a livestock technician from Nei Monggol, after v1s1t1ng the sand hills grazing areas of western Nebraska, is perhaps illustra- tive: "Your sand hills are covered with grass; ours are blowing sand dunes. These people are concerned first with their grass and hay for the winter; we look first to the numbers of our animals. I can remember when our plains were lush grass areas. Now we have desert. We must change our attitude and grow grass first, then get just enough animals to eat it. We are destroying ourselves." - 78 - 4.53 Chinese sources (1980) indicate about 27 m ha of usable hilly or mountain pasture in South China. However, this figure is generally acknow- ledged to be exaggerated, and fails to take into account the quality of grasses, soils and topography, and availability of infrastructure. Aside from the low feeding value of present ground cover, a major obstacle to development is that the areas with high potential are patchy and isolated, while the largest areas are located in sparsely-populated mountain areas with poor access. 4.54 Southern pasture areas may be divided into three subareas of uncertain size, each with distinct development potential and problE!ms. The most promising areas, the "southern mountain and hilly areas," are charac- terized by relatively high altitude and a mild climate, but with steep slopes. Soils are deficient in phosphorus, but are generally suitable for legumes and grasses. At higher elevations, temperate zone pastures such as clovers and ryegrass do well. At lower elevations and latitudes, the climate approaches subtropical and temperate zone legumes perform poorly. The "tropic:al hilly areas" have less potential because of poorer soils (also generally deficient in phosphorus) and higher temperatures. The area may be suitable for tropical legumes, although an appropriate mix of varieties is yet to be identified. The third subarea, the hill areas north and south of the Yangtze River appear to be least amenable to pasture development. High summer temperatures reduce yields of cool season grasses and legumes, whereas tropical legumes cannot survive winter temperatures. In each of these subareas, research is needed to identify suitable grasses and legumes. 4.55 Pasture development generally would be coupled with fruit and other tree crops on gentler slopes and afforestation on steeper lands. Because of the longer growing season and more adequate rainfall, dry matter yields tend to be substantially higher in these areas than in the North and Northwest. Improvement of these pastures could be carried out through hand or aerial seeding with legumes and grasses and applications of phosphate fertilizer. Costs of this technique appear to be under Y 150 per ha and, if done properly, the approach seems as effective as more costly alternatives involving soil preparation. In the southern latitudes, the long frost-free season and high dry matter yields minimize the amount of winter feeding required and make sub- stantial investment in cattle housing unnecessary. Overgrazing is not as pervasive as in the North, although it occurs in more densely-settl