At China's table : food security options

漱！결；，·―겸l。―떼ㅎ｀---&-｀업계―．계 .．떼， & 식，\;-．큰·「 李「――룐｀! /[ с н i и л г о г о АТ CH1NA'S TABLE China 2020: Development Challenges in the New Century Clear Water, Blue Skies: China's Environment in the New Century At China's Table: Food Security Options Financing Health Care: Issues and Options for China Sharing Rising Incomes: Disparities in China Old Age Security: Pension Reform in China China Engaged: Integration with the Global Economy T H E W 0 R L D B A N K 01 W A S H I N G T 0 N D . C . ÉmT CHINÆ'S TÉ BLE F O O D S E C U R I T Y 0 P T I O N S T H E W O R L D B A N K W A S H I N G T 0 N D C Copyright © 1997 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 1997 The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemina- tion of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, Massachusetts 01923, U.S.A. Cover photograph by Joe Carini/Pacific Stock. Cover insets (from left to right) by Vince Streano/Aristock, Inc.; Claus Meyer/Black Star; Serge Attal/Gamma Liaison; Dennis Cox/China Stock; Dennis Cox/China Stock; Erica Lansner/Black Star. ISBN: 0-8213-4046-8 44 Contents :.. Acknowledgments vii Map of AgricuLtural Regions ix Overview i Chapter 1 Introduction 7 Chapter 2 Consumption and Demand 9 Consumption projections 10 Clarifying assumptions 11 Chapter 3 Past Growth, Coming Constraints, and SoLutions 13 Accounting for past growth 14 Agricultural research 15 Fertilizer 16 Land 18 Water resources 19 Infrastructure 20 Policies and institutions 22 Chapter 4 Food SuppLies 29 Domestic grain projections 30 The international grain market 31 v Chapter 5 Food Balance Options 35 Options 36 Domestic grain supplies 36 The international market 38 References 41 This report uses Hong Kong when referring to the Hong Kong Special Administrative Region, People's Republic of China. vi At China's Table: Food Security Options Acknowledgments he scope of this report was determined in sions with the director and staff of the Department of Rural Development, Development Research Center of the State Council, during a prepara- tory visit to China in September 1996. The contents are based largely on the findings of a World Bank mission that visited China in November-December 1996. The mission members were Albert Nyberg (World Bank, mission leader), William Martin (World Bank, trade economist), Peter Glenshaw (World Bank, fertilizer specialist), Wen Poh Ting (consultant, agronomist), William Sheldrick (consultant, fertilizer specialist), Christian Bach (consul- tant, econometrician), Zhengxuan Zhu (World Bank, operations officer, agriculture), and Biliang Hu (World Bank, economist). Other World Bank staff and consul- tants-including Donald Mitchell (World Bank, agricul- tural economist), Daniel Gunaratnam (World Bank, irrigation engineer), Gary Kutcher (consultant, irrigation vii economist), Jikun Huang (consultant, agricultural Agricultural Sciences for providing several hours of use- economist), and Wen Chern (consultant, economist)- ful discussion. The mission also benefited from discus- assisted in preparing background papers. Bonita sions and exchanges of correspondence with Shenggen Brindley (World Bank) provided invaluable editorial Fan, research fellow, International Food Policy Research assistance. Institute, and Scott Rozelle, assistant professor, The mission worked closely with counterpart staff Department of Economics at Stanford University, par- from the Development Research Center, particularly ticularly on agricultural research issues. Both scholars with Chen Xiwen, director, and researchers Yu Bao Ping gave generously of their time and knowledge of China's and Cui Xiao Li, who participated in numerous discus- agricultural research structure and assisted the mission sions and provided logistical support. The mission ben- in better understanding the research issues. efited from fruitful discussions with officials from seven The book was edited by Bruce Ross-Larson, government ministries and several other government designed by Kim Bieler, and laid out by Glenn McGrath agencies. Particular recognition is due to Professor Mei and Laurel Morais of the American Writing Division of Fangquan and other staff of the Chinese Academy of Communications Development Incorporated. viii At China's Table: Food Security Options ．化： Overview hi . v54. hina can remain food secure over the next two to three decades and domestic food production will largely keep pace with population growth-if it can overcome the obstacles to growth in agriculture and infra- structure and implement several policy reforms. An objec tive of the government's food policy is to maintain 95 percent grain self-sufficiency. But this study shows that greater integration with international markets would per- mit food security at a lower cost-and that cereal self-suf- ficiency could decline to 90 percent. This conclusion is consistent with the expectation that China will join the World Trade Organization and will rely more on market forces to trigger investment, production, and consump- CIL~ tion decisions. As a result China and China's tradin part- ners will benefit, through lower production costs and higher farm incomes, as each country's comparative advantage is exploited. China is also likely to attract more foreign direct investment and import more technology. The best estimate of China's grain demand in 2020 resulting in record outputs. But marketing reform to is about 697 million tons of trade grain (608 million allow greater private participation was rescinded. tons of milled equivalent). Of this, China will likely "Average" food consumption is relatively high. Per produce 90 percent and import the balance-about 60 capita cereals consumption is among the world's high- million tons. The major grain-exporting countries can est, and consumption of vegetables and livestock prod- readily supply this amount without a large price ucts is similar to that of countries with much higher increase. But China will have to invest heavily in port incomes. Continuing income growth will lead to further facilities and bulk logistical systems-otherwise, dietary changes, but because consumption of "supe- imported grain, if handled by the current outmoded rior" foods is already high, these changes will be less systems, will be very costly. And meeting 90 percent of than current income elasticities imply. China's grain demand through domestic production will be achievable only if China increases its invest- Consumption and requirements ments in infrastructure, agricultural research, and land and water development. If not, imports will have to Cereal requirements for all uses will be about 697 mil- increase. And even with large investments, irrigation is lion tons of unmilled grain in 2020, just a 50 percent likely to become a serious constraint to continued increase over twenty-five years (table 1).1 And with ade- growth in agriculture. quate investments in agriculture and infrastructure, Among the required initiatives: domestic production should reach about 636 million * Improve agricultural research and extension-and tons by 2020, requiring about 60 million tons of wheat balance fertilizer application. and coarse grain imports. * Improve the efficiency of water distribution and increase aggregate irrigation water supplies, including Domestic production reuse and water transfers from the Yangtze to the Yellow River Basin. Projections of domestic grain production are based on * Reclaim and develop land to maintain the current investments in agricultural research to maintain total fac- stocks of arable and irrigated land and increase the effi- tor productivity growth, water resources to expand irri- ciency of multiple cropping. gation water supplies by 0.5 percent annually, land * Develop dedicated and integrated grain and oilseed reclamation to maintain the existing cultivated land base, transport facilities. and domestic fertilizer manufacturing, coupled with for- * Reduce government intervention in the cereal and eign exchange availability, to provide balanced fertilizer fertilizer sectors, encourage market-determined prices, supplies. It was also assumed that policies would change and open marketing and trade competition. to encourage competitive prices and efficiencies. Since the early 1960s China has made remarkable Irrigation water supplies would limit production in progress in national food security, though household the northern corn and wheat areas, but in the more food security remains a problem for low-income groups. Through the late 1960s and 1970s, large invest- ments in agricultural research led to the production and TABLE 1 . . .Estimated cereal consumption and total grain adoption of new technologies. Fertilizer production and requirements in 2020 (milled) application increased rapidly-at an average of 18 per- cent annually between 1962 and 1980. New crop vari- Rice Wheat Coarse etie boosd ggrain Total eties boosted grain yields by an average of 5.1 percent Per capita consumption (kilograms) 105.7 99.4 53.2 258.3 a year. Over the same period the irrigated area Direct cereal consumption 66.5 51.5 23.1 141.1 expanded by one-third (by 11 million hectares). Manufactured food consumption 39.2 47.9 30.1 117.1 Then in the late 1980s and early 1990s agricultural Requirements (million tons) 208.6 174.8 224.5 607.9 research investments declined, and growth in fertilizer Direct and processed consumption 150.9 142.5 75.0 368.3 Feed grains 46.3 21.6 138.2 206.1 use slowed. Pricing reform in the mid-1990s brought Seed and Losses 11.5 10.7 11.3 33.5 domestic grain prices close to international levels, Source: World Bank data and staff estimates. 2 At China's Table: Food Security Options water-abundant south, supplies would be sufficient to new technology, agricultural growth will lag behind the expand rice production. Consequently, rice supplies general economy, and food imports will be larger. What will be more than adequate to meet direct consumption investments in agricultural research would be optimum needs, and rice will be used increasingly as a feed is difficult to determine. But it is essential to reverse the grain-primarily the hybrid rice varieties that are con- declining trend. sidered inferior as a food grain. Because water scarcity limits agricultural growth Based on these assumptions, domestic grain produc- most, large investments are required to develop addi- tion in 2020 is estimated at 667 million tons under the tional water supplies-an increase of at least a 0.5 per- optimistic scenario and 636 million tons under the cent a year in supplies-and to transfer water from the "most likely" scenario. The supply-demand balance surplus south to the deficit north. Public investments are indicates wheat and coarse grain deficits requiring needed to improve irrigation water distribution sys- imports of 30 million tons of wheat for food and 30 mil- tems, and private investments are needed to improve lion tons of coarse grain for feed and industrial uses. on-farm use. Irrigation is a residual water user after other requirements are met, and in the future more International supplies water will be diverted from agricultural to urban and industrial uses. Wastewater treatment and water recy- A survey of unused production capacity in the tradi- cling are essential to expanding irrigation water sup- tional exporting countries indicates ample scope to plies and producing uncontaminated food supplies. expand production enough to meet China's incremental China must import more potash to balance fertilizer requirements and those of other importing countries application. The constraint to balanced fertilizer appli- without major price increases. For example, in 1996/97 cation-whether poor policies, monopolistic market world grain production expanded 7.5 percent in structures, or lack of farmer knowledge-must be response to higher prices, but for the five largest quickly identified and resolved. Balanced fertilizer exporters production increased 20 percent, and for application would immediately increase grain yields by Argentina, 40 percent. Long-term estimates of world 12 to 15 percent-vegetable yields even more. For cereal production range from 2.7 billion tons for 2020 example, if China had imported and applied another (Rosegrant 1995) to 3.1 billion tons for 2030 6.7 million tons of potash in 1995 at an approximate (Alexandratos 1997). The second estimate had a base cost of $1.5 billion, the value of incremental grain pro- case for world trade of 400 million tons in 2030 but duction would have been more than $9 billion, and the that could be higher if China imports more than 50 mil- value of all incremental crop production would have lion tons. The greater difficulty will be China's capacity been about $18 billion. to import this much grain. Without large investments in China imports 20-25 percent of internationally deep-water berths, port handling equipment, transit traded fertilizer to meet about 30 percent of its fertilizer storage, bulk grain transfers, and bulk rail equipment to nutrient needs. In 1995 about 36 million tons of fertil- improve grain handling efficiency, handling costs will izer (by nutrient content) valued at 125 billion yuan be very high. were applied to agricultural crops, representing the pri- mary agricultural cash input. But poor application Investments methods and low-quality fertilizers make fertilizer use inefficient and crop production costs unnecessarily high. To ensure that agricultural productivity continues to International joint ventures in raw material mining grow, China must invest more in research. Since 1978 and fertilizer manufacture could resolve some of the agricultural growth has averaged 6.1 percent a year, investment financing constraints, accelerate fertilizer and agricultural research has contributed more to plant construction, and ensure that the latest technolo- growth than any other factor. Although nominal and gies are employed. New factories using modern tech- real research budgets have increased over the past fif- nologies to produce high-nutrient nitrogen and teen years, de facto research budgets have declined by phosphate fertilizers are under construction, and others several measures. Unless there is a continuing flow of are planned-but there are delays. Large additional Overview 3 investments in new plants will be required to produce ports now exceed foreign imports and exports, while the fertilizer needed and to replace nitrogen and phos- transfers through inland ports have remained stable. All phate imports. these flows are transported in bags-requiring slow and Removing obstacles to the reclamation of nonstate costly intermodal transfers. Storage requirements must commercial land would allow large construction com- be carefully calculated to avoid dead storage. Inland panies to invest in reclaiming and rehabilitating land. waterways, used intensively in other parts of the world About 18 million hectares-barren lands, tidal lands, for low-cost transport of bulk grain, have been neglected and wastelands-have the potential to be reclaimed. in China because of low rail tariffs. Market-determined The cost of reclamation varies from about 15,000 yuan freight rates would encourage the development of more per hectare of barren land to 150,000 yuan per hectare efficient bulk transport modes such as barges. of tidal land. Much of this land is in remote or arid areas, but about 25 percent is in the water-abundant Policies southern provinces. Official records indicate that cultivated land per Today's grain policies may assist in short-term food capita is very low, but satellite imagery shows the area security, but they distort resource efficiency and impede under cultivation to be 132 million hectares, 40 percent long-term security. The major grain policy objectives larger than officially reported. About 190,000 hectares are to maintain 95 percent grain self-sufficiency; of agricultural land are lost annually to urbanization, increase control over production, pricing, and market- industrialization, and infrastructure. These losses are ing; and continue controlling stocks and international more than mitigated by land reclamation averaging trade. Reducing government control and permitting about 240,000 hectares annually, but reclaimed land is more reliance on competitive market forces for deter- less productive than the land lost. mining prices would allocate resources more efficiently Infrastructure investments, though substantial, are and improve long-term food security. insufficient for a dynamic economy. The grain industry Commercial activities must be separated from policy. suffers from the lack of bulk-handling logistical sys- Responsibility for the policy functions should go to an tems. The volume of domestic grain transported appropriately budgeted government entity, knowing between provinces will probably double over the next that it will incur losses. Government grain enterprises twenty-five years, putting substantial pressure on trans- and input suppliers must operate under the same con- port capacity. Investment in bulk rail wagons and bulk straints, efficiency incentives, and commercial stan- intermodal transfer systems would improve the capac- dards as the nonstate enterprises. Monopoly and ity to transport grain and increase rail wagon efficiency monopsony privileges and subsidies currently enjoyed by increasing use and turnover. The nongrain food by the state grain bureaus must be discontinued. industries have better processing and storage facilities Countryside procurement restrictions on nonstate but weak transport infrastructure. enterprises must be relaxed. And enterprises not com- With grain imports projected to triple over the next mercially viable must be allowed to go bankrupt. twenty-five years, large investments will be needed in If the government wishes to maintain grain reserves, it port facilities. Without new facilities, massive vessel should designate a government bureau to manage congestion and demurrage charges will be incurred. them-and accept that it will incur losses. This could be China has only one deep-water berth dedicated to bulk the bureau responsible for policy functions. Government grain handling, with one other under construction. costs could also be reduced by procuring and storing less Most grain is now imported on 45,000-50,000 dwt grain. The reduction in procurement should be phased, vessels, costing at least $10 more per ton in transocean allowing nonstate enterprises to handle a larger share of freight charges than 80,000 dwt vessels. Without more the marketed surplus. Indonesia and India stabilize grain deep-water facilities, trebling imports could add $400 prices by procuring only 20-25 percent of marketed million in transport costs. grain. High throughput storage is crucial in a bulk grain To improve the efficiency of water use, including irri- logistical system. Domestic grain flows between coastal gation, the government must raise prices. Low water 4 At China's Table: Food Security Options prices encourage overuse and contribute to low effi- Planned trade and single company monopolies ciency and to scarcity. Immediate price increases should should be discontinued. State monopolies plan grain at a minimum cover short-term supply costs, and gov- imports and exports far in advance, and although their ernment should schedule regular increases to eventually theoretical purpose is to increase market stability, their cover long-run marginal costs. transactions have amplified supply fluctuations. Government should provide incentives for land recla- Greater reliance on market forces would improve fertil- mation and productivity increases. Possible incentives izer import efficiency and enhance grain security. might include income compensation through cultivation leases or parallel development of nonagricultural land Note for leasing to commercial-industrial interests. Farmers could be offered longer cultivation rights to ensure that 1. Itis difficutto project consumption because of uncertain base- Line data, uncertain income elasticity coefficients and income growth they will benefit from terracing and soil amelioration. rates, and uncertain physical factors, such as feed-to-meat conversion And allowing farmers to lease land to and from others ratios. Within these constraints, future human consumption was esti- would permit part-time farmers to lease land to those mated using a general equilibrium model. Feed grain consumption estimates were derived separately for various livestock products interested in full-time farming-and increase agricul- assuming gradually improving feed conversion rates. Per capita cereal tural output. consumption and total requirements were estimated to 2020. Overview 5 Introduction n 1995 Lester Brown of the Worldwatch Institute forecast that China's grain import needs would likely be between 200 and 369 million tons by 2030.1 Brown's estimates shocked the research community and alarmed Chinese officials. His estimates also provided the impetus for this study of China's national long-term food security, the latest in a series of World Bank reports on China's grain sector.2 China has been famine prone, and as recently as the late 1950s and early 1960s famine claimed millions of lives. That is why government officials remain concerned about grain balances, questioning the reliability of the interna- tional market to meet the country's needs. And that is why they have pursued self-sufficiency policies and a program of large domestic grain storage. The strategy is costly and inef- ficient, but it ensures China's autonomy in domestic grain. Government's distrust of domestic (and international) grain markets has meant pervasive intervention in grain 7 production, marketing, trade, and consumption. in food. To be otherwise would be prohibitively expen- (Noncereal food marketing-fruits, vegetables, and sive in terms of transport costs. Estimates of annual livestock and aquatic products-are relatively free of import requirements fluctuate between 20 and 50 mil- government intervention, and thriving competitive lion tons for the early decades of the 2000s.3 markets have developed.) Grain policy reform and This report focuses on how China will avoid liberalization have been gradual and deliberate, national chronic food insecurity, but it also evaluates bringing domestic input and output prices close to the storage program and other alternatives to address international levels and consolidating multiple farm- the problems of transitory food insecurity from drought gate prices. But marketing reforms for greater private or other seasonal calamity. It discusses national food participation have been short-lived, with frequent security constraints and investments required to main- retrenchment. tain total factor productivity growth of 1.0 percent a China's economic growth over the past two decades year. The overall objective is to identify constraints to has been remarkable. But its agricultural growth has increased domestic food production and to suggest how lagged behind, and within agriculture, output of high- China could efficiently achieve food security in the early value commodities grew at 9 percent a year or more, decades of next century.4 The study models and projects while cereals grew at only 3.2 percent. food supply and demand to 2020. Secondary objectives Rapid income growth has altered food consumption of the study were to identify and evaluate: patterns. Per capita (direct) consumption of grains has * Factors that most influenced agricultural growth declined, and consumption of higher-quality fruits, veg- over the reform period. etables, and livestock products has increased, particu- * Physical and technical constraints and solutions to larly among urban consumers. In this process, grain use domestic production increases. has increased primarily through indirect consumption * Policy and institutional constraints to efficient (the conversion of grain to livestock products), magni- domestic marketing and international trading. fying the importance of feed grains. As incomes grow * Infrastructure constraints to domestic and interna- over the next two decades, consumption will continue tional marketing. to shift toward higher-value food items. But future con- * The potential of the international market to supply sumption growth will likely be less than current income incremental grain. elasticities suggest-because Chinese consumption of "superior" foods is already high. Notes China is the world's largest cereal consumer, ranking fifth in per capita cereal consumption, and cereals will 1. Brown (1995). 2. See, for example, World Bank (1991, 1993a). continue to supply the bulk of calories, as they do in 3. Fan and AgcaoiLi-Sombilla (1996). most countries. Expected changes in consumption and 4. Grain production and consumption were evaluated using a world- production will require China to rely more on the inter- wide general equilibrium model to capture all indirect and cross-effects of rapid income growth and international market impacts. The general national market to balance cereal supply and demand. equilibrium model was supplemented by a physical constraints model, Nevertheless, China will be substantially self-sufficient and investment requirements were identified to relieve these constraints. 8 At China's Table: Food Security Options Consumption and Demand "asarpwoood consumption levels in China are difficult to lql`establish because data differ from source to source. W Cross-section data from the China State Statistical Bureau's annual urban household consumption and expenditure surveys clearly show consumption increases for all food items. But the urban survey time series show just as clearly that per capita consumption of grain has been declining over the past decade, while income has been increasing rapidly-implying that grains may have become an "inferior" food (a good whose consumption falls as incomes rise). These data also indicate that per capita meat consumption has increased only marginally. Rural household consumption trends are similar, but at different levels. The Food and Agriculture Organization (FAO) uses a balance sheet approach to estimate consumption, deriving consumption as a residual from production and net trade, adjusting for nonfood uses and losses (table 2.1). TABLE 2.1 to-meat conversion rates are very high (low feed-meat China's per capita food consumption ratio), because most livestock subsist on grazing, green (kilograms per capita) fodders, and household kitchen wastes-except for com- mercial poultry raised in cages. This results in very efficient State Food and Statistical Agriculture apparent" conversion of grain to meat, but very ineffi- Bureau Organization cient production in the time required to produce meat.' household survey ba[ance sheet There is potential to expand the production of grass- and Food (1992) (1992-94 average) forage-fed livestock, but most future incremental meat Grain (cereals) 235.9a 224.2 Pork, beef, mutton 20.3 27.7 production will require more grain-intensive production. Poultry 2.4 5.0 Aquatic products 9.2 13.5 Fish 2.0 6.0 Consumption projections Vegetable oil 6.1 6.4 Vegetables 116.9 86.6 There have been numerous efforts to estimate rural and Fruit 38.8b 28.5 Fruit38.8128.5urban income, expenditure, and price elasticities of a. Deriving overall grain consumption by combining rural and urban consumption of individuaL grains yieLds a nationaL per capita estimate of 216 kilograms. demand for major food commodities. But the estimates b. Urban consumers only. Source: China Statistical Yearbook 1994; FA0 1996. for foodgrains are very unstable and cover the entire inelastic range from 0.0 to 1.0, and those for livestock The major differences in the two estimates are for products range from about 0.8 to over 2.0. Some insta- meat products and vegetables-other differences could bility comes from the many years of rationing, when con- be attributed to measurement error. The State Statistical sumption was barely influenced by economic factors. But Bureau data capture average household purchases, not recent estimates using sophisticated multiple equation food consumed in restaurants, at work sites, or other models and data for 1993-94, when rationing was sus- nonhousehold locations. And the bureau's data for meat pended, have been equally unstable. and poultry consumption are inconsistent with produc- A recent study determined that consumption patterns tion data, implying losses that are not credible. are influenced by structural as well as economic factors.4 Alternatively, the FAO estimates appear to overstate Residential location (rural, small town, large city) and meat production-so consumption is also overstated. age are important determinants, and they will become Scholars have argued that both estimates overstate cereal more important as the Chinese population urbanizes and consumption by as much as 15 percent.' ages. Other studies have determined that income elastic- The balance sheet approach shows an average ities for food staples (grain) decline at higher income lev- Chinese diet that is very high in cereal and meat con- els-and have projected Chinese income elasticities sumption, which is more consistent with diets in coun- turning negative for grain by 2010.s tries with much higher incomes. Daily caloric Long-term food consumption projections are diffi- consumption is about 2,730, or 95 percent that of cult because of unstable elasticity coefficients and Malaysians, whose per capita GNP is 550 percent because data on base consumption levels vary. As a greater.2 Meat consumption, at 32.7 kilograms per result, projections often imply consumption and caloric capita, is essentially the same as in the Republic of Korea, intake of unrealistic proportions. Projections for indi- where per capita GNP is fifteen times that in China. rectly consumed items, such as feed grain, are even Except for "high fat" pork, which is 70 percent of meat more complicated because feed conversion ratios have consumption, the Chinese diet is relatively healthy. Fat, not been empirically derived and are based on approx- oil, and sugar consumption is low-and vegetable con- imations and estimates. Alternative approaches-such sumption high-compared with other Asian diets. as making time-lag and income-lag comparisons with Rice and wheat are 85 percent of direct cereal con- other nearby Asian countries-also are inappropriate sumption, but coarse grains are important in some rural given China's extraordinarily high per capita consump- locations, particularly in areas unsuited for rice or wheat tion and relatively lower income. production. Otherwise, coarse grains are mainly con- Following a literature review and adjustment of sumed indirectly in meat and livestock products. Grain- income elasticities to reflect an average annual GDP 10 At China's Table: Food Security Options growth rate of 7.0 percent between 1995 and 2020, the assumption that technology would improve feed conver- Global Trade Analysis Project (GTAP) model was esti- sion by 2020. The total estimated cereal requirements mated (using the elasticities in table 2.2).6 Utilization were 608 million tons (table 2.3). coefficients, developed from China's input-output The requirement for food grain (direct and indirect) tables, allocated grain production to direct consump- will be about 368 million tons, for feed grains 206 mil- tion, manufactured and processed products, and alter- lion tons, and for other uses 34 million tons. Some native uses, such as livestock feed.7 A relatively large coarse grains will continue to be used as a food grain, proportion of cereal is consumed as manufactured and and some fine grain will be used as feed. Wheat bran, as processed products. a flour milling by-product, will continue to go into feed. With very large increases in meat consumption, About 12 percent of rice production is now used to feed China's feed grain requirements will expand rapidly. For livestock, because of the low price of hybrid rice and the pork, the major meat component, 70 percent is produced constraints on transporting coarse grains into rice under "backyard" conditions using very little grain or areas. The model suggests that rice used in livestock manufactured feeds. But unless much of the incremental feed will increase to 22 percent of production in 2020, production comes from more "commercial" operations, but better market integration would improve food and the animal numbers will be extraordinarily high. The feed grain allocations. level of aggregation in the model (as in all general equi- Two other studies also looked at China's foodgrain librium models) hides the details of meat consumption requirements. The International Food Policy Research (pork, poultry, other), but future consumption growth Institute (IFPRI) study explicitly addressed the cereal will likely be greater in more feed-efficient poultry. requirements of only food and feed.8 If "other" uses were The model estimated feed grain needs using a single added, the total requirements would be 587 million tons, conversion ratio. But to determine feed grain require- similar to the World Bank estimates. The FAO study esti- ments more accurately, feed needs were estimated sepa- mated requirements for 2030 and allocated "200 kilo- rately for poultry, pork, and "other red" meat under the grams per capita for all other uses," which, if intended to include nonfood and nonfeed uses, is about 6 percent TABLE 2.2 below the other two estimates.9 The different models use Estimated direct food consumption, 2020 quite different methodologies, yet all reach the conclusion that about 600 million tons of cereals (687 million tons of Income consumption consumption unmilled trade grain) will be required in 2020 (table 2.4). Food elasticity (kilograms) (million tons) Rice -0.05 66.5 94.6 CLarifying assumptions Wheat 0.07 51.5 73.9 Coarse grains -0.18 23.1 31.9 Meat 0.62 51.3 73.6 Estimates of total cereal requirements are sensitive to sev- Source: World Bank data and staff estimates. eral factors. Changing the annual average growth rate to 7.5 or to 6.5 percent would increase or decrease total TABLE 2.3 Estimated cereal requirements for 2020 TABLE 2.4 (million tons) Alternative estimates of cereal requirements (million tons) Coarse Grain use R icea Wheat grain Total In 2020 Cereals for food 94.6 73.9 31.9 200.3 This In 2030 Manufactured products 56.3 68.6 43.1 168.0 Grain use study IFPRI FA0 Feed grains 46.3 21.6 138.2 206.1 Food 368.0 298.4 263.3 Seed requirements 1.1 3.5 1.6 6.2 Feed 206.0 232.0 ne Postharvest Loss5 10.4 7.2 9.7 27.3 Other 34.0 ne 292.6 Total 208.6 174.8 224.5 607.9 Total 608.0 530.4 555.9 a. Total paddy equivalent is 297 million tons. ne Not estimated. b. Postharvest losses are assumed to be 5 percent of production. Source: WorLd Bank staff estimates; for IFPRI, Rosegrant, Agcaoili-Sombilla, Perez Source: World Bank data and staff estimates. (1995); for FAO, Alexandratos (1997). Consumption and Demand 11 cereal needs by 17.5 million tons, with most of the GNP would be 230 percent greater if based on purchasing power par- change occurring in coarse grain requirements. But ity estimates. 3. Pigs produced under "backyard" conditions use very small China's large population means that small changes in per amounts of grain but require 12-14 months to reach marketable capita consumption would result in large changes in total weight; "specialized households" typically market their animals in 6-7 estimates. Small changes in income elasticities or price months. 4. Huang and Bouis (1996). elasticities would also result in significant changes in total 5. Chern (1997); Huang, Rosegrant, and RozeLLe (1997). cereal requirements. Similarly, small changes in feed con- 6. The Global Trade Analysis Project (GTAP) model is a global gen- version ratios produce large changes in total estimates. eral equilibrium model maintained at Purdue University by a consor- tium of research and policy institutions including the World Bank, So any estimate should be carefully examined to ensure World Trade Organization, European Commission, OECD, and numerous that the assumptions are understood and acceptable. bilateral agencies. The model and its underlying database, documented in Hertel (1997), are widely used for analysis of agricultural and gen- eral trade policy worldwide. Notes 7. China State Statistical Bureau, "Input-Output Table of China" for 1987 and 1992. 1. Garnaut and Ma (1992). 8. Rosegrant, Agcauili-SombiLLa, and Perez (1995). 2. Based on the World Bank Atlas method. Malaysian per capita 9. Alexandratos (1997). 12 At China's Table: Food Security Options Past Growth, Coming Constraints, and Solutions gricultural growth since 1978 can be divided into the Household Responsibility System imple- mentation period (1978-85) and the post-implementation period (1985-95). In the first period agricultural growth averaged 7.4 percent, and total factor productivity' growth was 6.6 percent.2 In the second period growth declined to 5.8 percent, created primarily by additional inputs and investments, and total factor productivity growth declined to 1.1 percent (1985-89). These slow- downs in both total factor productivity and agricultural growth raise concern about the future growth of agricul- ture and food production. The various components of agriculture and food pro- duction grew at very dissimilar rates, particularly after 1985. Cereal production increased the least, reflecting its low profitability. Nonstaple foods, livestock, fisheries, and vegetables and melons grew at 9 percent or more after 1985 (figure 3.1). 13 FIGU)RE- 3.1 duction growth, a dynamic multisector output response Index of gross output of agricuLtural model was specified to evaluate growth during the components, 1978-95 reform period (table 3.1). 1984=100 Growth was analyzed separately for southern and 500 -Tota northern China using different crops to account for the - Livestock different cropping patterns and agroclimatic environ- Cereals ment. Variables in the analysis were the Household 400 -Fisheries -- Crops Responsibility System (institutional reform), invest- 300 - Vegetable and melon ments in research and water control infrastructure, input-output price ratios, land and labor prices, and 200 environmental factors. Land use and markets and agri- cultural extension were excluded because of the lack of 100 -change in the land factors and the inability to quantify changes in extension. Their contribution to growth is a1 undetermined and by default is in the residual. 1978 1981 1984 1987 1990 1993 1995 The analyses confirmed that the Household Source: China Statistical Yearbook various years. Responsibility System was important, contributing more than 2 percentage points to agricultural growth in southern China during 1978-84. But its impact in Accounting for past growth northern China was mixed, contributing 3.8 percent- age points annually to the growth of wheat production Various studies have identified investments, primarily but nothing to the growth of corn production. Other public investments, and institutional innovations as con- important contributors to growth were investments in tributing to rapid growth. One study concluded that the agricultural research, the engine of agricultural growth institutional reform reintroducing the Household during the reform period. Irrigation infrastructure Responsibility System was the major element influenc- investment, another important contributor to agricul- ing agricultural growth in the early 1980s.3 The fact that tural growth, may have added more than the coeffi- growth rates declined after 1984, when 99 percent of all cients indicate because many of the investments were villages had adopted the Household Responsibility for infrastructure rehabilitation and compensation (to System mode of production, supports this view. To ana- compensate for irrigated land lost to urban and indus- lyze the contribution and impact of factors on crop pro- trial growth) rather than for capacity expansion. TABLE 3.1 Contribution to crop production growth Southern China Northern China 1978-84 1984-95 1978-84 1984-95 Contributor Rice Other grain Cash crops Rice Wheat Corn Wheat Corn Research stock 1.61 5.03 8.51 1.65 3.30 5.84 3.43 6.07 Irrigation stock 0.13 0.22 0.00 0.39 0.43 0.46 0.47 0.50 Institutional reform (Household Responsibility System) 2.10 2.71 2.23 0.00 3.86 0.00 0.00 0.00 Input-output prices 0.44 1.18 0.99 -0.11 0.86 1.94 -0.75 -1.14 Land prices -0.03 0.00 -9.14 -0.01 -0.04 1.83 -0.01 0.40 Labor prices -0.96 -2.95 1.29 -0.21 -1.30 -5.90 -0.08 -0.36 Environmental factors Disaster 0.06 0.20 0.02 -0.07 0.30 0.36 -0.02 -0.03 Erosion-saLinization 0.00 -0.37 -0.86 0.00 0.01 0.13 -0.02 -0.35 Residual 1.20 1.35 11.55 -1.28 0.21 -1.17 -0.09 -0.19 Growth rate 4.54 7.37 14.59 0.36 7.63 3.48 2.12 4.89 Source: World Bank staff estimates. 14 At China's Table: Food Security Options Agricultural research Another budget issue is the inefficient use of funds. Research could be made more cost effective by defining By creating new technologies and crop varieties and research priorities better. Given the increasing demand improving agronomic practices, China's research sys- for feed grains, research to develop and adapt corn vari- tem has raised total factor productivity in agriculture eties for the traditional rice areas should have a high pri- and expanded crop production in the face of declining ority. In addition, livestock research receives a output-input price ratios. The growth accounting disproportionately small share of the budgetary analysis was consistent with the conclusions of other resources. Restructuring the research system, basing it scholars who have found that growth in Chinese agri- on agroecological zones rather than administrative cultural productivity was overwhelmingly attribut- zones, would also improve efficiency-because crop able to research investments.4 The high social returns research is often duplicated by research institutes at dif- to agricultural research have been well documented ferent levels in different regions. for many countries.5 Government investment in Equally important is the decline in agricultural research and development resulted in 20 percent agri- research intensity (table 3.2). Between 1975-79 and cultural productivity growth between 1965 and 1990-94 research spending slipped from 0.5 percent of 1994.6 Other work has estimated that the internal agricultural GDP to 0.4 percent (the coefficient excludes rate of return to agricultural research over the same research funded from provincial budgets). This ratio is period averaged 94.4 percent.7 This high return slightly below the average of other developing countries means that China is underinvesting in its agricultural but only one-fourth of the investment ratio of industrial research, a public good that should be fully funded economies. Without consensus on appropriate agricul- from the public treasury. tural research intensity, it is recommended that China Research in any given year builds on past research reverse the declining trend and raise the agricultural and adds to the foundation for future research. research intensity above the developing country average. Research investments accumulate and last for several The de facto decline in the research budget has been years before fully depreciating. Thus, research stock is exacerbated by the research commercialization of the more important than investments in any particular year past decade. Research institutes have been encouraged (except for continuity of research projects). to commercialize their activities and earn a portion of their income through technology sales. Ideally, the insti- Areas for concern tutes would license the technology to manufacturers and commercial entities, but weak protection for intel- Although China's agriculture has been well served by lectual property rights made this less attractive than the research, three issues create concern for the future. The commercialization option. Commercialization thus primary issue is the budgetary allocation for agricul- evolved into unrelated activities with higher capturable tural research. National agricultural research budgets returns, including manufacturing, restaurants and have quadrupled in real terms over the past three decades, and the number of research scientists TABLE 3.2 employed has increased sixfold.' But these numbers are Central government investments in agricultural employedresearch deceptive. Funding per scientist has declined, and funds Average annual per scientist influence output (new varieties) more than Number of Agricultural expenditures (1990 yuan) research research Total Per total budgets. Research costs have generally increased Period scientists intensitya (millions) scientist more rapidly than inflation-genetic research is turning - 1965-69 10,166 0.33 464 46,001 to more costly biotechnology. And research has 1970-74 10,618 0.41 720 68,518 expanded to incorporate pest resistance and drought 1975-79 19,319 0.49 1,022 52,729 1980-84 33,111 0.44 1,404 42,482 tolerance along with the former focus on higher yields 1985-89 50,330 0.40 1,763 35,336 and shorter seasons. One study concluded that funds 1990-94 61,835 0.39 2,063 33,276 actually spent on rice research per active scientist-year a. AgriculturaL research intensity is the ratio of agricultural research expenditures to agricultural gross domestic product. declined during the 1990s.9 Source: World Bank 1997b (annex table A.7.1). Past Growth, Coming Constraints, and Solutions 15 hotels, and commercial trade. Diverting researchers The overall conclusions, given the overwhelming from their primary task seriously detracts from the contribution of research to agricultural growth, are that institutes' research objectives, especially since only real investments in research must be better managed to about 15 percent of the commercial revenues are allo- ensure that efficiency and long-term food security cated to the institutes' research budgets.10 objectives are met. And research funding must increase The third issue relates to the competitive grants pro- enough to maintain annual total factor productivity gram, initiated in 1982 by the Chinese Academy of growth at 1.0 percent in grains-otherwise, declining Sciences, which allocates funds directly to research sci- production growth will necessitate more imports. entists. Conceptually, competitive grants are an efficient method of allocating research funds, but they may favor Fertilizer established scientists and institutes and make it difficult for younger, better trained scientists to participate. The The widespread use of chemical fertilizers has been a competitive funding has shifted resources to national major factor in the remarkable increase in grain and institutes at the expense of prefectural institutes. food production. Domestic production, using both Whether this is efficient remains to be determined-pre- indigenous local technologies and modern international fectural institutes introduced more than half of the new technologies, provides 70-75 percent of the fertilizer rice cultivars between 1990 and 1994. nutrient supply. Fertilizer use has quadrupled since 1978, spurred by fertilizer-responsive crop varieties. Productivity and output of research Fertilizer used in 1995 was valued at 125 billion yuan, making it the major cash input in crop production. Research has shown that the number of new rice culti- Worrisome, however, is the apparent declining effec- vars developed, released, and commercially used in the tiveness of incremental fertilizer application. This has first half of the 1990s was similar to that of the first half led some analysts to conclude that the potential for fur- of the 1980s but 25 percent below the output in the sec- ther yield increases, using existing cultivars, is limited.13 ond half of the 1980s.1' Note, however, that yields con- The Chinese Academy of Agricultural Sciences has tinued to increase despite the reduction in new cultivar identified four causes of fertilizer's declining effective- releases. Are the productivity and output of research ness: unbalanced and underapplication of nutrients, declining? The evidence is insufficient to conclude that especially underuse of potash; poor quality of fertiliz- they are, but more consideration must be accorded to ers; poor application methods; and poor distribution. the possible weakening of the agricultural research sys- tem. And it could signal the inadequacy of technology Unbalanced supply and use of nutrients transfer. Research staff state that only 30 percent of the technology developed is regularly applied in farmers' Even though the importance of proper nutrient balance fields and that the balance remains on the shelf, even is well known, and generalized ratios of 100:50:25 though new-variety adoption rates are about 80 per- (nitrogen, phosphate, potash) have been recommended cent. This might, however, simply reflect high-risk or for several years, China's application ratios in 1995 were financially unattractive technology. 100:47:16. (Korea applies a 100:120:80 ratio, and Japan Although Chinese counterpart agencies have 100:40:40.) This study analyzed macro-nutrient bal- access to technology generated by the institutions ances on the basis of a detailed investigation of the input affiliated with Consultative Group on International and uptake for the most important crop and livestock Agricultural Research, China benefits little from agri- products. cultural technology developed by private interna- The investigation found that in 1995 nearly 3 million tional corporations such as transnational seed tons of applied nitrogen and just over 2 million tons of companies.12 Greater internationalization of research applied phosphates were not used by crops and live- and technology through increased cooperation with stock, but potash removal was 6.7 million tons greater international public and private institutions would be than the amount applied (table 3.3). The excess nitro- mutually beneficial. gen and phosphate, valued at 18 billion yuan, was 16 At China's Table: Food Security Options TABLE 3.3 TABLE 3.4 NationaL nutrient balances, 1995 Average yieLd response to potash in baLanced (million tons) fertiLizer trials (kilograms per hectare) Nutrient Nutrient Assessed Fertilizer inputV (crop and animal) Losses Balance Yield Nitrogen 32.5 19.5 10.1 2.9 Potash with without Average yieldb Phosphate 12.5 6.8 3.5 2.1 Crop applieda potash potash Increment response Potash 12.6 17.5 1.9 -6.7 Rice 98 6,038 7,020 982 10 a. Includes chemical fertilizers, crop residues, animal feeds, organic wastes, atmos- Wheat 98 2,790 3,900 1,110 11 pheric deposition, and nitrogen fixation. Corn 113 5,048 6,570 1,522 13 Source: World Bank staff estimates. Tomato 165 23,318 30,773 7,455 45 Broad beans 120 2,145 3,233 1,088 9 essentially wasted-possibly contributing to environ- Watermelon 150 31,290 38,430 7,140 47 mental pollution. The underapplication of potash Citrus n.a. 40,148 53,783 13,635 n.a. diminished the efficiency of nitrogen and phosphate Pineapple 375 22,530 28,590 6,060 16 n.a. Not available. uptake, and crop production was lower than it would a. Nitrogen and phosphate fertiLizers were balanced. havebee wit baance fetilzer pplcatin. his b. Output per kilogram of potash. have been with balanced fertilizer application. This o Stauffer and Denton 1995. analysis is consistent with a recent study of nutrient input and output on farms by the China National Rice recent satellite imagery indicates that the cultivated Research Institute. land base is about 40 percent greater than land statistics In the short run nutrients are applied only to the indicate and there is double counting of about 10 per- extent that there is an economic yield response since cent of the single-nutrient fertilizers blended into com- some nutrient uptake is provided by soil nutrient pound fertilizers. Correcting for these statistical reserves. But for long-run sustainability, crop and ani- anomalies reduces the application rate to a very modest mal nutrient uptake must be replaced by natural 155 kilograms per hectare. China's average fertilizer processes (atmospheric deposition, biological nitrogen application rate is therefore below the East Asian devel- fixation) or by fertilizer. oping country average and far below that in Japan and Field trials undertaken by Chinese agricultural the Republic of Korea, where yields are higher. This low research entities indicate how yields would respond to application rate may result from the limited supply. balanced potash, nitrogen, and phosphate application Planning guidelines call for nitrogen self-sufficiency (table 3.4). The potential increases in crop productivity and 75 percent self-sufficiency in phosphates by 2000, from a balanced application of fertilizers range from 16 with the increased production to come from large to 50 percent, even more than the estimates of soil fer- plants that would cost more than $5 billion. This 2000 tility and fertilizer researchers, who state that properly objective will not be met. The manufacturing plants balanced fertilizer application would increase yields by now under construction that will be completed by 2000 12 to 15 percent. It is not clear why more potash is not will have a capacity of about 2.1 million tons of nitro- used in China despite solid evidence of the benefits. One gen equivalent. But during the late 1990s nitrogen con- reason could be that fertilizer use in China has evolved sumption will grow by 3 million tons, increasing the on the basis of local supplies. Nitrogen is the most deficit to about 5 million tons by 2000-if ammonium abundant locally sourced fertilizer and is used most. bicarbonate production remains constant. To replace Local phosphates have become available on a large imports with domestic production would require scale only in the last decade, and phosphate use has twelve or thirteen additional plants-with an invest- increased in line with domestic production. Very little ment of about $4.4 billion, based on cost estimates of potash is available locally, and it is costly to produce. $350 million for a 400,000-ton nitrogen plant. Whether such an investment would be efficient is diffi- Fertilizer supply and underapplication cult to judge because the international price of urea is volatile. But if domestic energy costs can be maintained The average application rate of plant nutrients is offi- below $2 per million British thermal units, domestic cially given at about 240 kilograms per hectare. But manufacture should be cheaper than importing. Past Growth, Coming Constraints, and Solutions 17 Fertilizer imports are centrally controlled and have replace the low-grade fertilizers. New incentive struc- averaged 8.6 million tons (nutrient content) annually tures and special efforts are needed to accelerate the con- over the past decade. Urea has been the major import, struction of new, modern factories, which require large followed by potash and compounds. SINOCHEM, a investments and gestation periods of a decade or more. Ministry of Foreign Trade and Economic Cooperation company, has the exclusive right to import fertilizer, Poor methods of applying fertilizer leveraging its position to obtain very favorable prices. It is inflexible, however, and has not responded to demand Because fertilizer is broadcast, it is estimated that half changes. China must be prepared to import large quan- the nitrogen applied to irrigated fields is lost to evapo- tities of potash because local supplies can meet only a ration. Applying the fertilizer directly to the root zone small fraction of requirements. by using fertilizer drills or slow-release fertilizer tablets would reduce this problem. The nitrogen in ammonium Poor quality of fertilizers bicarbonate is particularly volatile, even evaporating in storage. Because the fertilizer cannot be stored for long About half of China's supply of nitrogen and phosphate periods, ammonium bicarbonate plants often cannot be consists of low-grade ammonium bicarbonate, single operated in the fallow seasons when fertilizer is not superphosphate, and fused magnesium calcium phos- needed, and that increases costs. Evaporation is not a phate, produced in small factories. Using ammonium problem with the low-grade phosphate fertilizers, but bicarbonate as a nitrogen nutrient source is unique to not all of the phosphate is water soluble and thus is not China. Its manufacture involves a relatively simple tech- immediately available for plant uptake. nology using widely distributed anthracite coal deposits (or coke) as feedstock. This permitted China to build up Land production capacity and expand nitrogen fertilizer application. But its manufacture is relatively energy inef- China's cultivated land per capita is low, but land is not ficient (mitigated by the abundance of anthracite) and a limiting food security constraint given the potential highly pollution prone, and it decomposes in storage. for higher yields under balanced fertilizer application, Similarly, phosphate fertilizer production remains for expanding agricultural land through land reclama- primarily low-analysis single superphosphate, which tion, and for long-term investments in land if farmer can be manufactured from widely distributed low-grade incentives are improved. rock using simple, low-cost technology. Single super- Official sources list China's cultivated land at 95 mil- phosphate provides traces of the minor element sulfur, in lion hectares and sown areas at 145 million hectares. which Chinese soils are deficient, and would have to be However, satellite imagery indicates cultivated area of applied separately if triple superphosphate were widely some 132 million hectares. If crops are distributed pro- used. These low-grade materials have low nutrient con- portionally on the more than 37 million hectares of cul- tent (less than 20 percent), which makes them costly to tivated land not in the official estimates, individual crop transport and handle, and have poor agronomic efficacy. yields and fertilizer application rates are 40 percent The small-scale production of low-grade nitrogen lower than officially reported. and phosphates using local technologies is a remarkable International yield comparisons are easier to interpret feat, but it has left China with inefficient, polluting fac- for rice yields than for other crops because rice is almost tories and an antiquated industry structure. As applica- always irrigated, while other crops are grown under both tion volumes increase, it will become more important to rain-fed and irrigated conditions. Under the revised esti- improve production and transport efficiency by shifting mate of cultivated land, rice yields remain higher and to higher-analysis fertilizers. wheat and corn yields similar to the Asian average. Cereal New factories using modern technology and produc- yields are well below those of many major producing ing high-analysis, high-quality products are under con- countries and less than half those of the top-ranked coun- struction, and others are planned. But they will meet try. This does not imply that average Chinese yields can be only incremental needs, and they will not immediately increased to those of other countries (say, rice yields equal 18 At China's Table: Food Security Options to those in Australia) because climatic conditions are dif- amenable to irrigation, especially the "red soil" areas. ferent. But it does show the substantial potential for One-third of the potentially reclaimable tidal lands are improvement, particularly for corn (table 3.5). in the coastal provinces on the Bohai Gulf (Liaoning, The land base is steadily diminished by residential and Tianjin, Hebei, and Shandong). The cost of reclaiming industrial encroachment and infrastructure construction. these lands varies from a low of 15,000 yuan per Since 1988 construction has removed 190,000 hectares hectare for barren land to 150,000 yuan for tidal land. of agricultural land from cultivation annually, about a Land productivity is reduced by floods, erosion, water third of the annual decrease. (Some recent studies suggest logging, and salinization and alkalinization, but the areas the actual cultivated area may be substantially greater.) vulnerable to those conditions are increasingly being These losses have been concentrated in the Shandong and controlled and improved. The cropping index ranges southeastern coastal provinces, where high multiple from less than 100 in some northeastern and northwest- cropping compounds the land loss. Land reclamation, ern provinces to more than 250 in Hubei. Short frost-free averaging 245,000 hectares annually, has partially miti- periods limit the potential for increasing the index in the gated those losses (though newly reclaimed land is less northernmost and high-altitude provinces, except for productive than that lost to capital construction). Some greenhouse production of vegetables, and water short- cultivated land is lost by transferring it to alternative agri- ages limit multiple cropping in other northern locales. cultural uses such as pastures, forests, and fish ponds. But improved water control, short-season varieties, and Other fluctuations come from marginal land that is cul- appropriate production incentives will stimulate addi- tivated or lies fallow, depending on prices and profits. tional multiple cropping in the southern and middle An estimated 13.5 million hectares of barren land, provinces. 24.0 million hectares of tidal land, and 2.0 million hectares of wastelands are potentially reclaimable, Water resources although infrastructure and farm buildings would reduce the cultivable area by 25-30 percent.14 These Water scarcity is China's most limiting agricultural pro- areas are widely distributed around the country, but duction resource, particularly in northern corn- and with almost 50 percent in Xinjiang and Heilongjiang, wheat-growing regions. Investments in improved water where water shortages would constrain reclamation. control and delivery systems would conserve water and And some of these lands are in remote areas or are low modestly improve yields. in fertility-but could still be productive as pasture or forest. About 25 percent of the land lies within, or south Irrigation efficiency of, the Yangtze River Basin and would be more Only about 30 percent of the water diverted into irriga- TABLE 3.5 tion canals is actually delivered to crop root zones, due International yield comparisons, 1991-94 average to delivery inefficiencies and inefficient on-farm water (kilograms per hectare) use. Delivery improvement will require lining the canals Country or group Rice Wheat Corn and other investments (pipe and hose systems), and on- China (official estimates) 5,793 3,317 4,777 farm efficiency could be improved with appropriate China (revised estimates) 4,138 2,369 3,412 water pricing. Efficiency gains in delivery and applica- Asian countries 3,668 2,500 3,462 Developing countries 3,514 2,424 2,670 tion must be pursued, but achievable gains are limited. Industrial countries 5,671 2,556 6,249 Even with improved efficiency, productivity growth will Argentina 4,471 2,158 4,289 Australia 8,284 1,581 4,853 require incremental supplies. Egypt 7,566 5,093 6,116 France 5,035 6,559 7,775 Italy 6,558 3,497 8,236 Incremental supplies Korea, Rep. of 6,060 3,007 4,798 Thailand 2,185 - 2,871 Annual average water runoff in China is considerably United States 6,421 2,510 7,522 below the world average, and only about a third of the - Not avaiLable. Source: UN Food and AgricuLture Organization database. runoff is exploitable. Actual water exploitation was Past Growth, Coming Constraints, and Solutions 19 only 60 percent of the potentially exploitable supply in grounds, water deficits will instead be shared, with con- 1993. Capturing more will require large investments in sequent reductions in production. storage, diversion works, recycling, pumping, and con- veyance systems. Water conservation investments Major increases in the supply of exploitable water to northern China must await the completion of one or Government investments in water conservation have more of the three components of the south-north trans- fluctuated over the past four decades but have consis- fer scheme from the Yangtze River. Prefeasibility studies tently claimed about 70 percent of agricultural invest- on the eastern and middle of the planned routes con- ments. Real investments were relatively low during the cluded that 20 to 25 billion cubic meters a year could be 1980s, reflected in the small marginal increase in irri- transferred to the arid north with investments of $10 to gated area. Both investments and irrigated area increased $15 billion. sharply in the 1990s, with irrigation investments trebling Wastewater is discharged into waterways for down- (in 1990 yuan) and irrigated area expanding by 10 per- stream use, but much of it is untreated. The health and cent between 1989 and 1995. The marginal cost of this environmental consequences are unknown. Considering expansion was 10,000 yuan per hectare, representing the large anticipated water allocations to municipal and very efficient investments. Investment should emphasize industrial users over the coming decades, investments in water-saving technology. wastewater treatment and recapture for irrigation are The marginal cost of further irrigation expansion necessary to prevent outbreaks of waterborne diseases, will be considerable because the less costly projects have ensure uncontaminated food supplies, and support food already been completed and water will be increasingly production. Also, mixing irrigation drainage water with diverted from agricultural to nonagricultural uses, fresh water for reuse can expand the available water making it costly to maintain the irrigated area. Cost supply by as much as 25 percent. estimates for increasing water supplies through treating and recycling industrial water and for improving distri- Water consumption bution systems are unavailable. Industries, municipalities, rural residents, and irrigated Infrastructure agriculture are the major users of captured water. Water for irrigation is a residual, determined by alternative Most grain continues to be consumed on the farm or in uses that are expected to grow rapidly during the next the village where it is produced. Only a third of the two decades, especially municipal and industrial uses. cereal produced is marketed outside the village of pro- Increases in industrial water requirements are closely duction. But with increasing urbanization and use of linked to industrial growth, projected at a sustained 6 cereals for livestock feed, the marketed surplus will percent a year through 2020. Municipal water con- increase rapidly. However, lack of transportation and sumption is determined by the urban population size supporting infrastructure are serious constraints to (although it is rationed in some northern cities), which food marketing and distribution. Especially deficient is projected to grow 4 percent a year in the 1990s and 3 are facilities for bulk grain transportation, intermodal percent in the 2000s. Water demand by rural house- transfers, handling, and transit storage. Refrigerated holds is projected to stabilize, because small per capita transport and storage for perishable products are also increases are balanced by a shrinking population. lacking. The inability to rapidly move large quantities Requirements for irrigation water depend on the of grain creates fragmented markets and impedes the effectively irrigated area, efficiency of the delivery sys- development of rational grain markets. tems, and cropping practices. But as a residual user, irri- gation requirements may not be met. For food security Bulk grain logistical systems it is more efficient to fully irrigate as much as possible with the water available and to idle (or cultivate as dry- Bulk logistical systems create benefits that would be land) the balance. Since this is unacceptable on equity unavailable if the system components were introduced 20 At China's Table: Food Security Options individually. To market grain more efficiently and could amount to as much as $400 million extra per year. rapidly China needs more dedicated bulk-handling Looked at another way, potential savings of $400 mil- systems in high-density grain corridors. Because of lion could be achieved if adequate investments are made their high cost such systems are justifiable only on in deep-water berths and port-handling equipment. high-volume corridors. China's inadequate number of To efficiently accommodate future grain flows, deep-water ports, dedicated grain berths, and port- China must add a new 4-5 million ton capacity bulk handling facilities make external grain flows unneces- deep-water grain berth every two years over the next sarily costly. Lack of domestic bulk transport facilities two decades. Berth and bulk-handling investment costs makes transfers between the interior and the coast would be roughly $60 to $85 million per berth, depend- similarly difficult. Bulk intermodal transfers are lim- ing on existing ancillary facilities. ited to a bulk grain shuttle system that transfers Grain flows through inland ports have remained imported wheat from Tianjin to Beijing and low- remarkably constant during the 1990s, averaging 6.5 volume transfers from Dalian to inland northeastern million tons a year of bagged grain, a slow and costly cities.15 pace. Although neglected in China, inland waterways are intensively used in other parts of the world as a Ports. Fourteen major ports handle nearly all of low-cost transport mode for bulk grain. One reason for China's internationally traded grain. But only Dalian, the neglect is the administratively determined rail tar- Lianyungang, Qinhuangdao, Shanghai, Zhanjiang, and iffs which, unlike in other countries, are below barge Tianjin have specialized grain-handling berths. Others tariff rates. Market-determined tariff rates for all trans- are under construction at Dalian, Fengcheng, and some port modes would increase the demand for bulk barge Yangtze River ports. High-capacity bulk unloading transport along the major inland waterways. equipment is available at some major import terminals, but bulk loading facilities are generally lacking. For Rail transport. Rail movements of grain rose from 45 most ports bulk unloading consists of low-capacity million tons in 1985 to 67 million tons in 1995, thanks grab cranes, conveyor hoppers, and often improvised to improved efficiency. Further efficiency gains will be portable conveying equipment that cannot be operated difficult, however, as rail wagon utilization is now 98 in inclement weather. The lack of dedicated all-weather, percent and load factors for grain have reached 99 per- high-capacity, off-loading equipment increases port cent of wagon capacity. Some modest efficiency transfer costs. Discharge and demurrage costs range increases will be possible through increased train speeds from $4,000 to $12,000 a day per vessel (depending on and shipping distances. Ninety-nine percent of the grain the size and age of the vessel). Daily loading charges are is transported in bags; only 1 percent is transported on generally double the discharge cost. the Tianjin-Beijing bulk shuttle. There are 394 coastal berths classified as "deep- China has made very large investments to expand water berths" capable of handling vessels of 10,000 railway capacity, and more are planned, but limited deadweight tons (dwt), but no data are available on transit capacity seriously constrains feed grain ship- the number of berths capable of accommodating ments from northeast China to southern destinations. larger vessels. Most grain vessels calling at Chinese Grain-specific investments are difficult to justify ports are 45,000 dwt to 50,000 dwt. There are no because grain is only 4 percent of rail freight, although grain berths capable of handling 80,000 dwt grain requirements of bulk users (food millers and man- ("Panamex") vessels. Because transocean shipment is ufacturers, feed millers, and industrial manufacturers) more costly on smaller vessels, ocean freight costs on will increase. Special-purpose bulk wagons would China-destined grain are higher than for countries boost efficiency by increasing wagon utilization and receiving the larger vessels.16 turnover per wagon. Bulk wagons would reduce load- Port facilities have been adequate if not efficient. ing and unloading time to minutes from the one to two Importing grain on smaller vessels and slow off-loading days now allowed by the Ministry of Railways. at poorly equipped ports cost an estimated $10 extra per Bulk rail wagons should be used only on high-vol- ton. If grain imports treble in the next two decades, this ume rail corridors, preferably with backhaul opportu- Past Growth, Coming Constraints, and Solutions 21 nities, with bulk loading and unloading facilities at rail- a basically free market, but government intervention heads. Individual bulk hopper wagons with annual remains the norm for cereals. During the late 1980s and turnover capability of about 4 million tons per kilome- early 1990s the government gradually liberalized grain ter annually (based on a 1,400 kilometer average dis- policy and reformed grain institutions, primarily tance of grain shipments) cost about $50,000 each. A through liberalized retailing, but rapid price increases in $625 million investment in 12,500 bulk wagons would late 1993 led to retrenchment in 1994. Subsequent handle half of the grain traffic, but ancillary loading and reform has been modest, focused on decentralization. unloading facilities would require further investments The major objectives of grain policy are to: to reap the full benefits of bulk transport. * Maintain 95 percent grain self-sufficiency as stated in the government's White Paper by the Information Storage. Grain storage requirements increase as grain Office of the State Council.'7 production and domestic trade expand, but much of the * Increase control of production, pricing, and market- storage capacity is constructed on-farm. The state adds ing-government quota procurement of 50 million tons about 1.7 million tons of storage space annually, mainly plus 40 million tons of additional procurement to con- long-term storage for reserves. Virtually all of these addi- trol 70-80 percent of the marketed grain. tions are flat warehouse storage, which does not lend * Continue controlling stocks and international trade. itself to rapid movements in and out. China could reduce government storage by relying on nonstate enterprises to Governors Responsibility System and the market and distribute grain if more flexibility were per- pricing-marketing framework mitted for variations in price by season and region. High-throughput transit storage is crucial to efficient In 1995 the central government delegated responsibility intermodal interfacing in a bulk grain logistical system. for balancing local supply and demand and for some Transit storage requirements at intermodal transfer pricing and marketing of grain (including financial points are highly site-specific and must be carefully cal- responsibility) to the provinces. Market transactions culated to avoid unused storage. Transport arrivals and between government grain enterprises in surplus and departures must be well timed to ensure high through- deficit provinces replaced centrally planned inter- put. provincial grain transfers. Government grain enter- prises from deficit areas are forbidden to procure grain Bulk fertilizer logistical systems in the countryside in grain surplus areas, and they must buy from wholesale markets at the county level or With most food and agricultural production on small above. Surplus-producing provinces must maintain a farms, bulk handling of fertilizer between manufacturer three-month supply of grain stocks-and deficit and farmer would be impractical, though it could be provinces a six-month supply. Provinces are free to sub- efficient for some state farms. And as larger domestic sidize farmers' inputs and consumers' purchases as their fertilizer manufacturing plants are constructed, bulk revenues allow. Private traders cannot procure grain in delivery to bagging centers in agriculture-intensive the countryside until state grain quotas have been filled. areas should be considered for some of the output, to The state retained the grain quota (for below-mar- avoid repetitive and costly bag handling. ket-price grain, which must be delivered to grain bureaus), national stock responsibility, and the interna- PoLicies and institutions tional grain trade monopoly. The effect of the Governors Responsibility System policy was to decen- Most grain policies are control oriented, urban biased, tralize responsibilities, replacing the nationwide policy focused on the short term, and impede long-term food with a multitude of provincial policies. But the policy security. Managing and supporting the grain sector is content changed only marginally. Decentralization, by complex because it directly involves five ministries and itself, had little influence on the record grain production several agencies. The nongrain foods-fruit, vegetable, of 1995 and 1996, attributable instead to higher prices and livestock products-are produced and marketed in and good weather. 22 At China's Table: Food Security Options Quota procurement and farm income stocks, and financial accounts. Quota grain is sold as higher-priced negotiated grain. Reserve grain is used for In 1995, 142 million tons of grain, 30 percent of total commercial stocks. Consolidated accounts make it easy production, were marketed outside the village of pro- to reallocate commercial losses to policy losses and then duction. Government grain enterprises procured 65 to claim government reimbursement. percent of this quantity-purchasing 46.2 million tons Although not a pure monopsony the grain bureaus of quota grain and an additional 46.3 million at negoti- procure grain at below-market (and largely govern- ated prices (table 3.6). Quota grain prices averaged 60 ment-determined) prices before competitive procure- percent of the free market price; negotiated prices 90 ment is permitted. They have little incentive to be percent. These below-market prices imposed a total efficient because their losses are reimbursed, and they implicit tax of 40.7 billion yuan on grain producers. have minimal competition for raw materials and only Recent grain price and production increases show that modest competition in marketing (grain bureaus pro- farmers respond to price; conversely, artificially low cure 70-80 percent of marketed surplus). Their reform average prices depress production and marketed sur- has focused primarily on retail marketing. pluses and make food security more difficult. Because Most consumers now purchase the major share of incomes of grain farmers are typically below those of their grain from the private sector, but private retailers aquatic, livestock, fruit, and vegetable producers, the purchase most of their stocks from grain bureau enter- tax is discriminatory as well. prises. An element of competition was formerly gener- The government proposes to procure 1997 grain ated by grain bureaus from various locales competing quotas at market-determined prices and to discontinue to procure farmers' surpluses. But this competition has consumption subsidies through the grain enterprises. It been reduced by the Governor's Responsibility System will retain procurement quotas to control grain stocks. and accompanying regulations. Unifying prices at the market level will improve farmer Pricing policy, combined with the grain bureaus' incomes, but the purchase price will remain state-deter- inefficiency, is costly. IOU procurement is prohibited, mined, albeit with competition between various gov- but there are reports of delayed or closed procurement ernment grain enterprises. Nongovernment grain due to lack of funds or storage space." Grain bureau enterprises will participate in the market only after the net losses continue to accumulate, rising to 19.7 billion state satisfies its procurement needs. yuan in 1996, and government consumption subsidies for "grain and edible oil" have averaged 20 billion yuan State grain enterprises a year over the past five years.19 Grain policy and enterprise inefficiency cost 80 bil- The grain bureaus have two conflicting roles: to per- lion yuan a year (implicit farmer taxes, losses, and con- form regulatory and policy functions and to operate sumption subsidies), so reform is imperative. Full commercially profitable enterprises. This two-track sys- market competition must replace the near monopsony tem allows enterprises to mix staff, activities, grain that the grain bureaus now enjoy. To achieve and main- TABLE 3.6 Government grain procurement and value, 1995 State procurement (miLion tons) Production Price (yuan per ton) Implicit tax Quota Negotiated (miLLion tons) Quota Negotiated Market (billion yuan) Rices 18.4 10.3 185.2 1,107.5 1,729.0 1,897.5 16.3 Wheat 17.1 14.1 102.2 1,080.0 1,528.0 1,688.0 12.7 Corn 9.3 15.0 112.0 855.0 1,385.0 1,580.0 9.7 Soybeans 1.0 4.2 13.5 1,814.0 2,422.0 2,711.0 2.1 Other 0.4 2.7 56.7 na na na na Total 46.2 46.3 469.6 na na na 40.7 na Not applicabLe a. Average for indica and japonica (unmitted). Source: World Bank 1997b (annex table A5.1). Past Growth, Coming Constraints, and Solutions 23 tain competitive marketing efficiencies, government be effected with smaller reserves. Provincial authorities enterprises must be subject to the same constraints, and other jurisdictions maintain their own reserves, and incentives, and standards as private traders, collective farmers are known to retain large stocks. But informa- farms, and grain processing companies. Competition tion on stocks and releases is not available, so the costs could result in some government grain enterprises and inefficiencies cannot be evaluated. declaring bankruptcy and exiting the industry, leaving Large reserves protect against transitory insecurity, but some of the 4.1 million grain bureau workers unem- less costly alternatives should be sought. Futures hedging ployed. But many of the grain bureaus would remain in is an efficient management tool when annual imports are business-with fewer workers. The cost of providing required to meet consumption requirements but are inap- social assistance to all 4.1 million employees would be propriate as insurance for possible imports. Futures far less than the current inefficiency costs of 80 billion options are relatively inexpensive theoretical alternatives, yuan. but there are too few market transactions in these instru- ments to be a viable alternative for China. However, if Grain reserves port and handling capacities are increased, greater reliance on the international market would be more effi- Industrial country governments have been getting out cient than maintaining stock for the most severe contin- of grain storage because the cost of holding reserve gency. From 1990 to 1996 the average cost of stocks exceeds the perceived benefits of grain price sta- internationally procured grain would have been $35 a ton bility, leaving reserves to be maintained as commercial less than procuring (at an international price equivalent) stocks by more efficient and competitive private enter- and storing domestic grain for subsequent use (table 3.7). prises. Food grains in these countries comprise only a Policy proposals call for: small portion of agriculture incomes and consumer * Separate storage for incremental grains added to the spending, so price instability is relatively unimportant. reserve, and gradual stock removals from grain bureau But there are economic arguments for price stabiliza- storage to self-controlled storage. tion in developing countries where food grains are a * New grain warehouses, to be constructed in readily major cost element in the food basket.20 accessible grain-deficit locations. The China State Administration for Grain Reserves * First-in, first-out inventory management. manages reserve stocks on a noncommercial basis. It * Reserves to be maintained independently of grain directs and pays provincial grain bureaus to handle bureaus, using separate storage, staff, and accounts. grain on its behalf. Separating the management and * Future reserves to be added as grains are available, operational functions has hindered the government's not procured according to plan. ability to effectively stabilize prices. Price stabilization These changes would permit the State Administration releases can be triggered by a 20 percent price increase, for Grain Reserves to implement policy decisions and "within a short period of time" with State Council perform its stabilization role. approval. Releases have been difficult to activate because the reserve grain is physically mixed with grain Input pricing and marketing bureaus' commercial stocks, and it is sometimes sold and thus unavailable when needed. In addition, the Fertilizer is the major cash input for food production, grain bureaus are reluctant to release stocks because and its marketing remains a planned component of the storage fees provide good income, and a substantial input supply sector. Various government jurisdictions portion of the reserve stocks are unavailable. plan their fertilizer requirements and submit them to the The state grain reserve target of 40 million tons- next administrative level for consolidation, which must together with reserves maintained by provincial and be approved by the State Planning Commission. The local authorities plus 90 million tons of commercial and plan indicates the sources of domestic manufacturing semicommercial grain procured by state grain enter- and import requirements-including timing and port of prises-is an extraordinary and costly quantity of grain delivery. The fertilizer supply market is basically a single to maintain price stability.21 Price stability could surely channel monopoly operated by the China National 24 At China's Table: Food Security Options before the beginning of the year and implemented by TABLE 3.7 Cost of maintaining versus importing wheat reserves monopoly trading companies. Given China's impor- (U.S. doLLars per ton) tance in the international market, these companies Purchase and carryover to impose considerable leverage in pricing and delivery Year following year' Unit value of imports terms, typically importing fertilizer and grain at c.i.f. 1990 172 (cost, insurance, freight) prices very near to interna- 1991 220 118 tional price levels. 1992 157 142 The monopoly trading companies and "planned" 1993 185 129 1994 170 131 import-export systems are effective, but they operate 1995 172 174 with uncertain efficiency. Planned grain trading has 1996 223 229 exacerbated fluctuations in rice, wheat, and corn sup- a. Assumptions incLude: 10 percent interest, 5 percent storage Losses, and storage cost of $18 per ton a year. plies over the past decade. More open trade regimes Source: Wor[d Bank staff estimates. might result in higher farmgate prices for corn, lower- cost wheat flour to consumers, and more stabilizing Agricultural Means of Production Corporation, which trade transactions. takes domestically manufactured fertilizer from the manufacturers each week and receives imported fertil- Grain pricing izer at the ports for onward distribution. Only an esti- mated 10 percent of the domestically manufactured After a long period of very stable subsidized cereal prices, fertilizer is traded outside this channel. Prices and mar- the government decided in 1993 to reduce its fiscal costs keting margins received by the corporation are govern- by increasing administratively set prices and moving ment controlled. Imported fertilizer, which typically has toward market-determined prices. Domestic prices higher nutrient content, is priced higher. began to move upward in 1993, increasing sharply in The Agricultural Means of Production Corporation 1994 before leveling off in 1995 and declining in 1996. at each level receives and sends fertilizer onward to the By the end of 1996 wholesale cereal prices had returned next lower level, eventually arriving at the farmers' vil- to early 1994 levels. Throughout the period price move- lages. This system has been effective in supplying farm- ments were relatively smooth and without the monthly ers with some nitrogen and phosphate fertilizers because volatility experienced in the international market-par- the 1,500 small factories are widely distributed through- ticularly with rice. Nor were the international price out the country and marketing basically involves mov- spikes in late 1993 and early 1994 transmitted to domes- ing local supplies over short distances. Interprovincial tic markets. But during 1994 domestic (wholesale) prices trade in fertilizers is primarily from imports. of corn increased 65 percent, rice 75 percent, and wheat This existing marketing structure has failed to deliver 60 percent-while international prices of corn and rice the types and combinations of fertilizer nutrients rec- were declining and wheat prices were stable. Both ommended and needed. Nor is it flexible enough to domestic and international prices peaked in early 1996, respond to the large and rapid changes in fertilizer with domestic prices above international prices, and quantities and composition required in the next cen- average domestic wholesale prices remain above inter- tury. Independent dealers, whose incomes depend on national prices. But there is a very wide range of whole- helping farmers to optimize their fertilizer input, work sale prices between the surplus and deficit provinces well in many countries of the world-and they should because the transport infrastructure is inadequate to be an improvement in China. move the large volumes required. The combination of administratively determined International trading companies procurement prices and planned imports and exports has hindered internationalization of the domestic mar- China is the world's largest fertilizer importer-and in ket. With domestic prices now higher than international some years the world's largest grain importer. Imports prices, domestic market liberalization and international of both commodities are "planned" up to six months integration are difficult. Reducing quotas and negoti- Past Growth, Coming Constraints, and Solutions 25 ated prices would hurt rural incomes and maintaining lioration, degradation prevention). Such farmers are domestic prices above international levels would less interested in agricultural productivity than in require import protection with either quotas or tariffs. retaining their land-use right as a social safety net. Maintaining the current price differences implicitly sub- Other farmers have begun to produce more profitable sidizes producers and requires large fiscal subsidies. commodities, such as fruits, vegetables, livestock, and Moreover, higher domestic grain prices encourage gov- aquatic products. That leaves farmers who produce ernment grain enterprises to import rather than procure grain, oilseeds, and cotton at a serious income disadvan- domestically. Thus China faces the prospect of unsanc- tage-not because they lack scale economies but because tioned cross-border trade-particularly with adjacent of an inappropriate land and labor resource combina- rice-exporting countries, tion. Rural incomes will diverge even further unless grain producers can gain access to additional land. If they had Water pricing access to the land of part-time farmers, with appropriate compensation, national productivity would increase. Water is the most limiting resource in food production, The various land-use experiments under way, includ- and current low water prices encourage overuse and ing shareholding and leasing village land, need to be greater scarcity. The Water Financial Directive of 1988 expanded to facilitate the rapid development of a land- called for marginal-cost water pricing by 1997, but this use market. clearly has not occurred. Water prices charged to most farmers remain below supply costs, ranging from zero Notes to 0.40 yuan per cubic meter in the Yellow River Basin. The marginal value of water in agriculture ranges from 1. Total factor productivity is defined as the ratio of output to the 1 yuan per cubic meter for grains to 4 yuan per cubic weighted sum of inputs. It is typically expressed as a ratio of index numbers measuring the change in output relative to the change in meter for vegetables.22 In the few locations where total inputs. rational irrigation water pricing has been imple- 2. Wen (1993, pp 1-41). mented, water demand has declined without affecting 3. Lin (1992). 4. Huang, Rosegrant, and RozeLLe (1996). yields. 5. For example, Fan (all agricuLture-China), RozeLLe (rice-China), Water charges should thus be immediately increased GriLiches (hybrid corn-United States), Evenson and Jha (all agricuL- to cover supply costs-with a timetable for the move to ture-India), Pray and Ahmed (all agriculture-Bangladesh), and Thirtle and BottomLey (aLL agriculture-United Kingdom). marginal cost pricing. Increasing prices may be difficult 6. Fan and Pardey (1997). unless irrigation services and infrastructure are ade- 7. Fan (1996). quately maintained and system performance is 8. Fan (1995). 9. Rozele, Pray, and Huang (1996). improved. 10. RozeLLe, Pray, and Huang (1996). 11. Roselle, Pray, and Huang (1996). Land use 12. The Consultative Group on International Agricultural Research is a Loose-knit group of donors that support eighteen institutes engaged in international agricultural research. Cultivation contracts for farmland were recently 13. Brown (1994). increased from 15 to 30 years. That improved incen- 14. Barren land may be either virgin land or abandoned cultivated tives, but incentives remain inadequate to stimulate the land. Tidal Lands are inundated during high tide and reappear during Low tide, and wastelands are primarily areas where rocks and other investments needed to improve land productivity. And waste materials from mining operations have been discarded. because land might be redistributed before contract 15. Additional systems are under construction in the northeast, expiration as rural households grow, land tenure Yangtze River, and southwest areas under a World Bank Group-financed . . 23 project. remains insecure.23poet 16. Lloyd's Maritime Information Services, Inc. Also, during a Since many farmers, particularly in the southeast recent 90-day period the average cost per ton for various bulk-grain coastal provinces, now engage in agriculture part time shipments between the U.S. Gulf Coast and China were: 30,000 dwt, and earn only a minor portion of their income from $31.00; 55,000 dwt=$25.00; and 80,000 dwt, $16.00. Fertilizer is fre- quently shipped in even smaller ships (15,000-25,000 dwt) with per farming, they have little incentive for productivity- ton rates in excess of $40.00. enhancing investments or practices (terracing, soil ame- 17. Information Office of the State Council (1996). 26 At China's Table: Food Security Options 18. Xinhua News Agency reports in various newspapers. 21. Indonesia's BuLog stabilized domestic rice prices during the 19. An estimated annual subsidy of 3 billion yuan for cotton was past three decades by relying heavily on private sector traders while subtracted from the price subsidies for grain, edible oil, and cotton itself performing only a marginal market role. BuLog's maximum annual Listed in China Statistical Yearbook 1996. rice purchase was 25 percent of production and carryover stocks are 1 20. Rice price stabilization by Indonesia's Bulog (Badan Urusan million tons. Logistik) is estimated to have contributed almost 1 percentage point 22. World Bank (1996). to GNP growth in its early years, 1969-74 (Timmer 1996). 23. Prosterman (1995). Past Growth, Coming Constraints, and Solutions 27 Food Supplies onomic model and a constraints model were ed to estimate future food production. The Global Trade Analysis Project (GTAP) general equilib- rium model was modified to evaluate seven agricultural subsectors: rice, wheat, coarse grain, nongrain crops, livestock, meat and milk, and other food products. The extensive aggregation required to reduce general equilib- rium models to manageable proportions results in some loss of detail, but the process ensures overall consistency (for details of the model, see World Bank 1997b). In this model agricultural growth is driven by investments and total factor productivity (TFP) growth, with production estimated under two TFP growth rates-1.0 and 0.5 per- cent. Lower investment would lead to a lower TFP growth rate. 29 Domestic grain projections gain experience in corn cultivation and new varieties specific to southern climates are developed, corn pro- GTAP model results for 2020 indicate a grain produc- duction will expand in the south. But producing rice as tion range of 707-769 million tons, depending on TFP feed grain will likely continue--despite its lower feeding growth, and meat production of 74 million tons- value-as hybrid rice yields are twice the corn yields in implying that meat production would be more efficient this region. than meat imports. Coarse grain production would The constraints model was also used to evaluate the increase the most, reaching 258 million tons-with impact on agriculture of alternative water allocations paddy rice reaching 361 million tons, and wheat, 150 for nonagricultural (municipal and industrial) use, million tons (table 4.1). These production estimates are alternative yield growth rates (as a proxy for investment extraordinarily high. in research), and changes in effective irrigated area (as a Because water is so severely limited in northern proxy for water investments). The rate of industrial China, and the TFP growth rate may be reduced further growth is the most influential factor in determining even with increased investments in research, a physical future irrigation water availability-and food produc- constraints model was developed to evaluate the impact tion. In any given area the faster industry grows, the less of various water options on agricultural output in the water remains for agricultural use. Rapid industrializa- nine water regions. Containing no economic variables, tion in some water regions would exacerbate irrigation it was estimated using three yield growth rates under water shortages. Alternatively, industrialization in the conditions of favorable water availability. A continu- water surplus areas of southern and southeastern China ously increasing yield growth rate of 1.0 percent annu- would not deprive agriculture of water. But the impact ally (the high rate); a yield growth rate of 1.0 percent of transferring cultivated land to industrial use would until 2010, with a subsequent growth rate of 0.5 per- be greater because the multiple cropping index is higher cent (the base rate); and a yield growth rate of 1.0 per- in the south than in the drier and colder north. cent up to 2000, with a subsequent growth rate of 0.5 Results from the model clearly indicate the need for percent (the low rate). investments in water development, water saving, and Estimated total grain output for 2020 in the con- water recycling to meet municipal and industrial straints model is 636 million tons for the base case- requirements and to provide agriculture with incremen- considerably lower than the GTAP model estimates. tal water supplies. In 2020 the marginal agricultural Based on rainfall and water runoff statistics, the con- impact of the proposed south-north water transfer straints model concluded that irrigation water will be schemes (eastern and middle routes) would be about 15 limited in the northern corn-producing regions but million tons of grain, 10 million tons of vegetables, and remain sufficient in the southern rice-producing various tonnages of other crops with a 1995 value of 57 regions. That means that rice could become an impor- billion yuan ($6.7 billion). That yields a crude internal tant feed grain as well as a foodgrain. In many parts of rate of return of 20 percent for agriculture (no effort southern China, low-quality hybrid rice is cheaper than was made to determine incremental benefits to munici- coarse grains-and is already used for feed. As farmers pal and industrial users along the canal routes). The impact could be greater if scarce water were applied TABLE 4.1 optimally, but for equity reasons, it is assumed that Grain production estimates for 2020 water shortages would be shared. (millions of tons) Besides the need for greater investment in water, the GTAP model model also indicates the need to invest in land reclama- TFP growth Constraints model . . m Grain 1.0 0.5 High Base Low tion to mamtam the cultivated and irrigated land base, Ricea 361 342 313 298 283 agricultural research to ensure rising crop yields and Wheat 150 119 151 144 137 expanded multiple cropping, and agricultural extension Coarse grains 258 256 203 194 186 to transfer research results to farmers. The extension Total 769 707 667 636 606 issue is particularly important for improving the fertil- a. Measured as paddy equivalent. Source: World Bank staff estimates. izer nutrient balance, because this would immediately 30 At China's Table: Food Security Options increase productivity and provide a higher base for market for private traders and government agencies, it future productivity gains. If the fertilizer nutrient bal- is highly segmented by strong preferences for rice ance is not improved, grain supply deficits will likely classes and qualities based on incomes, prices, and cul- double. tures. The world wheat market is less segmented but In the base case scenario for investments and water, still has many classes of wheat-ranging from feed cereal supply is estimated to be 636 million tons of rough wheat for livestock to high-quality wheat for bread, grain in 2020, if the cultivated land base remains con- pasta, and pastries. Coarse grains, used primarily for stant; 486 million tons of fruits and vegetables and animal feed, are traded based on feed value. 50 million tons of oilseeds would also be produced. This falls short of the expected needs of 695 million tons of Potential grain export supplies grain, calling for imports of about 60 million tons of wheat and feed grains in about equal shares. About half The potential supply of grain from the world market is the imports will be feed grains. If investments are inade- substantially larger than today's trade of 200 million quate to maintain the resource base and increase pro- tons, but stagnant world import demand since the late ductivity, the lower production will demand even more 1970s has delayed development of this potential. The imports. traditional grain exporters accounted for 90 percent of world grain exports in the first half of the 1990s.' The international grain market Domestic per capita consumption is already high in these countries, and, except for Thailand, population In the early 1990s world grain production averaged 1.7 growth is low-so production growth is primarily a billion tons, half of it coarse grains, and wheat and function of their ability to export. With reduced export milled rice accounting for the rest. About 11 percent of demand, they reduced grain crop areas by 34.5 million that production enters international trade-with wheat hectares from the highs of the early 1980s. The produc- the most widely traded-followed by coarse grains and tion potential of this land was roughly 115 million tons, rice (table 4.2). China accounts for about 20 percent of and it could be returned to production almost immedi- world grain production, but its participation in world ately if there were incentives to do so, as shown by markets has been modest, at no more than 5 percent of increases in 1996 cereal plantings and harvest. international trade. However, the world grain market In addition to the cropland that could be returned to has the potential to supply China with substantially grain production in these exporting countries, there are larger quantities. other potential sources of export supplies: The world markets for wheat and coarse grain are * The former Soviet Union has since the early 1980s the largest and best integrated-with active cash and reduced the cropland used for grain production by 27 futures markets for wheat, maize, barley, oats, and rye. million hectares, and due to fuel, fertilizer, and machin- Smaller and less well integrated, the world market for ery shortages and to marketing and transportation con- rice is primarily centered in Asia. Operating as a cash straints, its total production declined by 86 million tons between 1990 and 1995. * Argentina has 142 million hectares of permanent TABLE 4.2 pasture, flat delta land used primarily for cattle grazing, Annual average world grain production and trade, adol2. milo hetrso'emnn rp. 1992/93-1996/97 and only 2.2 milhon hectares of permanent crops.2 (millions of tons) With only a small portion of the land receiving fertiliz- China's share of ers, pesticides, or herbicides, much could be used for World China's world production World crop production under improved production practices. Grain production production (percent) trade . Export grain stocks in the major exporting countries Coarse grain 839.7 100.6 12.2 90.3 are about 100 million tons, some 40 million tons above Wheat 551.1 115.0 20.8 98.7 t Rice (milled) 367.0 128.7 35.8 12.1 the previous low of 60 million tons in 1995-96.3 So Total 1,759.6 344.3 19.8 201.1 world trade could immediately expand by roughly 40 Source: USDA/FAS 1996. million tons, and over the longer term the potential for Food Supplies 31 increased grain supplies is large. For example, in is less. But the major grain-exporting countries have response to higher prices in 1996-97 production in the been highly responsive to international prices and the five largest exporters increased by 20 percent. level of world demand. World grain price volatility Large-country effect Historically, world grain prices have declined over long If China were to increase grain imports significantly and periods relative to overall consumer prices, but they rapidly, its actions would push up world grain market have also increased sharply over short periods of two to prices. But if imports were gradually increased over sev- three years. Real average annual prices for wheat, eral years, the price impact would be marginal. maize, and rice declined 50 to 60 percent between the The level of grain stocks is important because it 1970s and the 1990s.4 But between 1971 and 1974 real reflects the ease of expanding exports. During the mid- prices of all major grains increased by 50 to 100 per- 1980s, the major grain-exporting countries had carry- cent, due primarily to a rapid rise in energy and fertil- over grain stocks of 260 million tons, while world grain izer prices. Also, between 1993 and early 1996 real trade was less than 200 million tons. This large reserve prices increased 33 percent for wheat and 46 percent for could expand exports without significant price corn but then declined (table 4.3). increases. In contrast, current world grain stocks of Real grain price declines are forecast over the next about 100 million tons could not meet a significant ten to fifteen years, according to recent studies by increase in exports. The price elasticity of export supply researchers at the World Bank, the UN Food and is lower now than during the mid-1980s-roughly 1.1.' Agriculture Organization, and the International Food In the longer term (say, after five years), export supply Policy Research Institute.5 These forecasts assume that is more elastic because producers have more time to historical yield increases continue, that population expand production. growth slows, and that demand increases due to rising incomes. Market access Future grain price volatility is likely to increase because of policy changes in the major exporting coun- Concern about market access has often led to policies tries-those due to the Uruguay Round Agreement on aimed at a high degree of self-sufficiency in traditional Agriculture and those made unilaterally in the United food-importing countries, stimulated by the poor policy States and the European Union for budgetary reasons. record of the major grain exporters. In 1996 the Policy changes in the major grain exporting countries European Union imposed a wheat export tax in an were largely responsible for the sharp declines in world attempt to reduce exports and prevent domestic prices grain stocks between 1986 and 1995. These changes from rising. During the 1970s the United States embar- will keep world grain stocks low because the United goed either soybeans or grain in three successive years States government will no longer hold significant grain to prevent higher consumer prices. Targeted export stocks, and stock-holding is expected to be lower in the embargoes have also been used for political reasons, as European Union. Lower stocks could lead to greater in 1980, when the United States embargoed grains, price volatility because the buffer against a poor harvest other foods, and agrochemical exports to the former Soviet Union to protest the invasion of Afghanistan. But TABLE 4.3 this sixteen-month embargo was largely ineffective World grain prices, selected periods 1991-96 because alternative supplies were available. (nominal U.S. dollars) Exporting countries often limit food exports if prices Period Wheat Rice Maize rise sharply, so unless an importing country has a long- 1991 (annual average) 128.66 314.40 107.40 term agreement, supplies from individual countries are 1993 (annual average) 140.24 270.00 102.10 not assured. Further, those supplies are least likely to be 1996 (January-June average) 231.40 349.60 183.00 1996 (December average) 175.70 319.20 117.70 available when prices rise sharply-precisely when Source: World Bank staff estimates. market access is most important. The best defense for 32 At China's Table: Food Security Options importing countries is to diversify imports among com- meet this requirement, production in the exporting peting suppliers and to enter long-term agreements that countries will need to grow by 1.1 to 1.4 percent a specify conditions for market access. year-readily attainable if investments in research are maintained. Meeting future demand Notes With appropriate incentives, grain output in the export- ing countries (including the former Soviet Union) could 1. The traditional grain exporters and their share of total world a bexports are as follows: the United States (42 percent), the European increase by 200 million tons and bring them to their pre- Union (22 percent), Canada (11 percent), Australia (7 percent), vious maximum output levels. But most regions of the Argentina (6 percent), and Thailand (2 percent). world are net importers and will require increased 2. FAQ (1995). imports. So the issue is whether the traditional exporters 3. USDA/FAS (1996). 4. World Bank (1997b). (including countries of the former Soviet Union) can meet 5. Islam (1995). all the incremental import requirements without large 6. Carter and Gardiner (1988). Consider the export supply elastic- price increases. ity of the five major grain exporters. The 1990-95 average grain pro- duction of these countries was 583 million tons, domestic In the early twenty-first century most of the incre- consumption was 432 million tons and exports were 185 million tons. mental demand will be met from domestic production, If we assume a domestic demand and supply elasticity of 0.2 and that except for Japan, the Republic of Korea, and Taiwan stock holdings will remain constant, then the export supply elasticity is 1.10. If we assume a domestic demand and supply elasticity of 0.1 (China). But worldwide import requirements will triple then the export supply elasticity is 0.55, and if the domestic demand (increase 200 percent) between 1990-94 and 2030. To and supply elasticity is 0.3 then the export supply elasticity is 1.65. Food Supplies 33 Food Balance Options hina has the potential to remain food secure over the next two to three decades if it carries out reforms and makes investments in agriculture and infra- structure. China is expected to rely more on market forces to signal investment, production, and consumption deci- sions. And once it joins the World Trade Organization, it will integrate more fully with the international market- place. If all this is successful, domestic food production will keep pace with population growth, but household food security may remain problematic for the poor. With a more open trade regime and competitive market, China and its trading partners will benefit through exploiting their comparative advantages. Increased incomes and urbanization will change diets: direct grain consumption (per capita) will decline, and the consump- tion of livestock products (indirect grain consumption), vegetable oils, sugar, fruits, and vegetables will increase. 35 China can substantially increase domestic grain pro- Domestic grain supplies duction, but irrigation shortages will limit expansion, and the country will have to rely more on the interna- Government would like to maintain 95 percent grain tional market for wheat and feed grains. self-sufficiency, but to limit future imports to 5 percent of consumption would be inefficient and costly. To achieve Options 95 percent self-sufficiency, import tariff protection of about 25 percent would be required and domestic whole- The best estimate of China's 2020 cereal demand is sale prices would be about 50 percent higher than inter- about 608 million tons (697 million tons of unmilled national prices. grain), but actual requirements depend on several uncertain factors. Domestic production can meet a Pricing and marketing major part of the incremental needs if investments are made to: To improve food security, grain production, and market- * Improve agricultural research and extension to ing efficiency, China must reduce government interven- maintain total factor productivity growth and rapidly tion and increase its reliance on market forces to achieve balanced fertilizer application. determine prices of both inputs and outputs. * Develop water resources to enable aggregate sup- plies of irrigation water-including recycling and More reliance on market forces could improve effi- reuse-to increase by an average of 0.5 percent annu- ciency. To achieve and maintain marketing efficiency, ally between 1995 and 2020. Achieving this will require government grain enterprises and nongovernment mar- water transfers from the Yangtze to the Yellow River keting agencies must operate under the same constraints, Basin. the same efficiency incentives, and the same commercial * Reclaim and develop land to maintain the current standards. To ensure that markets are integrated and stocks of arable and irrigated land. farmers receive proper price signals, the government * Develop dedicated port and rail facilities for inte- must discontinue the spatial and temporal monopoly- grated bulk-handling and transport of larger quantities monopsony privileges of its grain enterprises and elimi- of domestic and imported grains. (Without such invest- nate below-market-price procurement. It must also relax ments even greater investments would be needed in cov- countryside procurement constraints on the grain enter- ered rail wagons, bag-handling equipment, and flat prises from deficit areas and on private traders. Instead, storage-as well as in suboptimal agricultural produc- the government should procure grain for the poor, mili- tion and water resource development throughout tary forces, civil servants, and strategic reserves at com- China.) petitive "market" prices and transfer its policy functions A series of policy reforms centered on less govern- to other noncommercial agencies. Similarly, monopoly ment intervention in the cereal sector, market-deter- fertilizer marketing must be discontinued. mined prices, and open competitive marketing and Water is the major constraint to expanding food pro- trading regimes for inputs and cereals must accompany duction in China. But inappropriately low irrigation these investments. With these investments and reforms, water prices have encouraged inefficient water use and China will likely produce 90 percent of its grain require- have led to water shortages. To improve water-use effi- ments and rely on international suppliers for the bal- ciency, including irrigation, the government must ance-about 60 million tons of grain by 2020. But increase water prices to cover supply costs and then without them, more imports will be required. The establish a timetable to increase prices to long-run mar- major grain-exporting countries could readily supply ginal cost levels. Also, market-determined freight tariff these added imports, but unless China invests heavily in rates would encourage the development of more efficient port facilities and bulk logistical systems, imported transport modes, such as bulk barge carriers. grain will be extremely costly because even greater investments would be needed to improve the present Price stabilization can be less costly. Price stabiliza- outmoded handling systems. tion is possible with considerably less market participa- 36 At China's Table: Food Security Options tion than the 70-80 percent of marketed surpluses that sity has declined over the past fifteen years, raising government grain enterprises currently procure. If the doubts about its ability to sustain long-term growth in state wishes to stabilize prices and maintain strategic total factor productivity. grain reserves, these activities should be managed and Research indicates that balanced fertilizer applica- operated by a government agency completely separate tion would increase land productivity by 12 to 15 per- from the commercial operations of government grain cent-increasing cereal production by 50 to 60 million enterprises. The stabilization programs should operate tons. Whether the failure to apply balanced fertilizers within price bands that generally follow long-term lies within the agricultural extension system or the fer- international price trends. Otherwise, costs will become tilizer market structure, the constraints must be identi- exorbitant and unsustainable. fied and resolved. Developing China's fertilizer industry will require Changes in land policies could boost food produc- large investments with long lead times. China has suffi- tion. Two changes in land policies could provide incen- cient feedstock material to develop large-scale ammo- tives to increase land productivity and food security and nia-urea manufacturing plants to meet new nitrogen to improve rural income distribution. One is allowing requirements and replace a portion of the low-analysis rural residents to lease their land cultivation rights to ammonium bicarbonate plants. Also, the richer phos- others (when off-farm activities provide most of their phate deposits are being exploited, but earlier planned income) or to lease cultivation rights from others when investments in accompanying triple superphosphate underemployed on their own farms (which also would manufacturing plants have not materialized. improve rural income distribution). The other is Water is the most limiting food production encouraging the private sector to invest in land recla- resource. Increasingly, water will be diverted from mation. In addition, the rural land experiments now agricultural to urban and industrial uses, and large under way should be evaluated, consolidated, legalized, investments will be required to maintain and increase and publicized-to enable farmers and land developers irrigation water supplies. A first step is to improve effi- to understand the alternatives and the opportunities. ciency by treating and recycling municipal wastewater, lining canals, and improving on-farm application. International trade policies should be relaxed. The Municipalities and industries must be required to monopoly trading corporations have procured and invest in wastewater treatment to permit environmen- imported fertilizer and grains at prices very close to tally safe recycling. international prices. But farmers have not been pro- Water resources development is costly and has high vided with the appropriate balance of fertilizers. And social returns but low financial returns. Except for small because supply-demand conditions change between pond- and reservoir-based irrigation systems, only the planning and execution, planned international transac- government has the capacity to undertake new invest- tions have amplified the volatility of domestic supply. ments like the water transfers from the Yangtze to the Planned trade and trade monopolies should be discon- Yellow River Basin, which require long-term planning tinued so that open trading can facilitate appropriate and huge investments. imports, permit timely exports, prevent unnecessarily China's per capita cultivated land base is quite small, large corn stocks from accumulating in the northeast but satellite imagery shows that it is 40 percent larger provinces, and ensure that imported commodities are than official statistics indicate. This means that fertilizer procured and handled efficiently. application rates and crop yields are 40 percent below reported levels-representing tremendous potential for Research and resource investments increasing productivity and production. Continued investments in land reclamation are required to main- Agricultural research is an excellent investment, but tain the land base. Though newly reclaimed land is less research must have priorities, and research budgets productive than the prime agricultural land it replaces, must be increased. China has had a dynamic agricul- it is needed to compensate for the land lost to urban tural research system, but agricultural research inten- encroachment and infrastructure construction. Food Balance Options 37 Rising urbanization and cereal-based livestock pro- rity. Incentives will be needed for nongovernment duction mean that more grain will be marketed, handled, investments: and transported. Investment in improved handling infra- * International seed companies might invest in structure is essential for moving these increased quantities research in China if permitted and if intellectual prop- of grain efficiently. This includes dedicated bulk grain- erty rights were strengthened, but most agricultural handling logistical systems involving inland and coastal research will need to be funded from domestic public waterways, rail, transit storage, and intermodal transfers. resources. Large investments will be required to efficiently pro- * International fertilizer companies might invest in duce 636 million tons of cereals and import another 60 China's domestic industry. This would partially resolve million tons and to undertake routine maintenance and the fertilizer investment constraint and ensure that the capital replacement. Water resource development involv- latest mining and manufacturing technology is ing south-north transfers, improved irrigation canals and employed. distribution systems, and municipal-industrial waste- * Land reclamation and long-term productivity- water treatment will require the largest investment. Other enhancing measures require incentives for nonstate sec- large investments will be needed in transportation and tors to invest. These might include extended cultivation handling infrastructure, including ports, bulk rail, and rights or the privilege of leasing out cultivation rights waterway transport; ancillary bulk-handling equipment; and compensation for land reclamation companies and fertilizer manufacturing. Also, investment and cur- through land sales or leases (with government retaining rent budgets will need to be increased for agricultural ownership) perhaps for nonagricultural use. Longer- research and extension. Rough estimates of incremental term contracts for farmers might induce them to terrace investment requirements are indicated in table 5.1. or undertake other land improvements. * Incentives for nonstate investors in domestic market- Incentives for nonstate investments ing and transportation infrastructure are difficult to design, but large domestic and international corpora- The government alone will not be able to finance the tions (such as Chai Tai) might invest in bulk transit stor- large investments required to provide future food secu- age and bulk barge or rail wagons, if these were dedicated to their operations. And, if the grain market TABLE 5.1 were more open, international grain corporations Estimated incremental investment requirements, might invest in ports and bulk-handling equipment. 1995-2020 Without major private participation in grain markets, (millions of 1995 U.S. dollars) transport and logistical investments in high-volume grain corridors would probably depend on the public Water resources 64,000 South-north water transfer 25,000 treasury. Irrigation distribution improvementa 39,000 Municipa-industrial wastewater treatment n.a. The international market Infrastructure 1,385 Ports 600 Bulk rail wagons 625 International grain stocks have been rebuilt to about nulakn trinals, with bulk-handling equipment 100 million tons from the lows of 1995-96, creating a Fertilizer manufacturing plants 5,000 small margin for increased trade. But the major grain- Agricultural researchb 25,660 exporting countries could rapidly produce and export Agricultural extension n.a. much larger quantities of grain if they expected steady Total 96,045 growth in export demand. n.a. Not available. a. To maintain the trend of the past decade, investments in water conservation International grain prices are expected to continue should incrase by about $120 million annually; expenditures in 2020 should be $3 their long-term decline. But short-term volatility is billion more (in 1995 terms) than in 1995.thilogtr dein.Btsr-emvlaltys b. To maintain the trend of the reform period, agriculturaL research investments likely to be greater than during previous decades shouLd increase by 5 percent annually; investment expenditures in 2020 shouLd be $1.75 billion more (in 1995 terms). This may or may not be enough to maintain because the exporting countries have discontinued or total factor productivity growth at 1.0 percent a year. Source: World Bank staff estimates. reduced government storage, leaving a smaller buffer to 38 At China's Table: Food Security Options mitigate shortages from drought or calamity. It would sels and to link bulk grain logistical systems with port be to China's advantage to enter long-term contracts for facilities. specified quantities of grain with grain traders from As China continues to open to the world, its options exporting countries. for achieving food security will increase. Both China and China's fourteen major grain ports have a theoreti- its trading partners will benefit from exercising their cal import capacity of 23 million tons-they were at 85 comparative advantages in an increasingly dynamic percent of capacity in 1995 when total imports were 20 world market. Investments in agriculture will enable million tons. Although these and other ports could China to increase its domestic food production, and handle more grain, it would be very costly to import investments in infrastructure mean that food surpluses beyond design capacity because marginal costs and food imports can be efficiently transferred through- increase rapidly. That makes it crucial to construct out the country. Expanded options for food security will deep-water ports capable of handling 80,000 dwt ves- mean greater variety and abundance at China's table. Food Balance Options 39 References Ahuja, Vinod, and D. Filmer. 1995. "EducationaLAttainmentin Developing Countries: New Estimates and Projections Disaggregated by Gender." Background Paper for the World Development Report 1995. World Bank, World Development Report Office, Washington, D.C. Alexandratos, N. 1997. "China's Consumption of Cereals and the Capacity of the Rest of the World to Increase Exports." Food Policy. 22(3): 253-69. Anderson, K., B. Dimaranan, T. Hertel, and W. Martin. 1996. "Asia-Pacific Food Markets and Trade in 2005: A Global, Economy-wide Perspective." CEPR Discussion Paper 1474. Center for Economic Policy Research, London. Aubert, C. 1996. "China's Grain Distribution and Price Policies." Paper presented at a workshop on Agricultural Policies in China, OECD headquarters, December 12-13, Paris. Barichello, Richard. 1996. "The Nature of State Trading in Indonesia: The Case of BULOG." Paper presented at the International Agricultural Trade Research Consortium annual meeting, December 15-17, Washington, D.C. Brown, Lester R. 1994. "Can China Starve the World? Its Booming Economy Is Consuming Global Food Supplies." Washington Post, August 28. -._ 1995. Who Will Feed China? Wake-up Call for a Small Planet. New York: W. W. Norton & Company. Cao, Mingkui, Shijin Ma, and Chunru Han. 1994. "Potential Productivity and Human Carrying Capacity of an Agro-Ecosystem: An Analysis of Food Production Potential of China." Agricultural Systems 47(March): 387-414. Carter, Colin A., and Walter H. Gardiner, eds. 1988. Elasticities in International Agricultural Trade. Boulder, Colo.: Westview Press. Chern, Wen S. 1997. "Survey of Food Demand Elasticities in China." Background paper for this report. World Bank, East Asia and the Pacific Region, Rural Development and Natural Resources Sector Unit, Washington, D.C. 41 China State Statistical Bureau. Various years. China Statistical Yearbook. GTAP (Global Trade Analysis Project). 1996. GTAP Database, version Beijing: China Statistical Press. 3.3. Purdue University, West Lafayette, Ind. . 1987. "Input-Output Table of China 1987." Beijing. Guanzhong, James Wen. 1993. "TotaL Factor Productivity Change in . 1992. "Input-Output Table of China 1992." Beijing. China's Farming Sector: 1952-1989." Economic Development and Chu, A.C.P. 1996. "Sustainable Agriculture: Integrating Pastoral and Cultural Change 42(1): 1-41. Arable Land in Southern China; A Model to Increase Arable Land- Hertel, T. W. 1997. Global Trade Analysis: Modeling and Applications. What Can the New Zealand Experience Contribute?" Paper presented Cambridge: Cambridge University Press. at a workshop on Agricultural Policies in China, OECD headquarters, Hong, Y. 1996. "Sources of Productivity Disparities in Regional Grain December 12-13, Paris. Production in China." Paper presented at the China Workshop- Crook, F. W. 1996. "How will China's 'Rice Bag Policy' Affect Grain Global Agricultural Science Policy for the 21st Century, August Production, Distribution, Prices and International Trade?" Paper 26-28, Melbourne, Australia. presented at a workshop on Agricultural Policies in China, OECD Huang, Jikun. 1996. "China's Integration into the Global Economy: headquarters, December 12-13, Paris. Longer Term Implications for Agriculture and the Agro-Food Sector." Du, Ying. 1996. "Perfection and Reforms of Chinese Food Production Center for Chinese Agricultural Policy, Chinese Academy of and Marketing Policies." Paper presented at the International Agricultural Sciences, Beijing. Symposium on Food and Agriculture in China-Perspectives and Huang, Jikun, and Howarth Bouis. 1996. "Structural Changes in the Policies, October 7-9, Beijing. Demand for Food in Asia." Food, Agriculture and the Environment Evenson, R. E., and D. Jha. 1973. "The Contribution of Agricultural Discussion Paper 11. International Food Policy Research Institute, Research Systems to Agricultural Production in India." Indian Washington, D.C. Journal of Agricultural Economics 28: 212-30. Huang, Jikun, and Scott Rozelle. 1995. "EnvironmentaL Stress and Grain FahLbeck, E., and Z. Huang. 1996. "The Property Right Structure of Yields in China." American Journal of Agricultural Economics 77 (4): Farm Land and Family Farming in China-An Analysis of Developing 853-64. Options." Paper presented at a workshop on Agricultural Policies in - 1996. "Technological Change: Rediscovering the Engine of China, OECD headquarters, December 12-13, Paris. Productivity Growth in China's Rural Economy." Journal of Fan, Shenggen. 1991. "Effects of Technological Change and Development Economics 49 (JuLy):337-69. Institutional Reform on Production Growth in Chinese Agriculture." Huang, Jikun, and L. Zhang. 1996. "The Macroeconomic and Public American Journal of Agricultural Economics 73(2): 266-75. Expenditure Implications of Grain Policies in China: Some Selected - 1995. "Data Survey and Preliminary Assessment in China." Issues for Discussion." Paper presented atthe workshop on Agricultural Report to the United Nations Food and Agriculture Organization, Policies in China, OECD headquarters, December 12-13, Paris. January 30, Rome. Huang, Jikun, M. Rosegrant, and Scott RozeLLe. 1996. "Public . 1996. "Research Investment, Input Quality, and the Economic Investment, Technological Change and Reform: A Comprehensive Returns to Chinese Agricultural Research." Paper presented at the Accounting of Chinese Agricultural Growth." Paper presented at the China Workshop-Global Agricultural Science Policy for the 21st China Workshop-Global Agricultural Science Policy for the 21st Century, August 26-28, Melbourne, Australia. Century, August 26-28, Melbourne, Australia. Fan, Shenggen, and Mercedita AgcaoiLi-SombiL[a. 1996. "Why - . 1997. "China's Food Economy to the Twenty-first Century: Projections on China's Future Food Supply and Demand Differ." Supply, Demand and Trade." Food, Agriculture and the Environment Paper presented at a conference on Food and Agriculture in China: Discussion Paper 19. International Food Policy Research Institute, Perspective and Policies, October 7-9, Beijing. Washington, D.C. Fan, Shenggen, and Philip G. Pardey. 1992. Agricultural Research in Huang, Yanxin. 1995. "Current Demand for Grain in China and Forecast China: Its Institutional Development and Impact. The Hague: Until 2000." Paper presented at a conference on Grain Market International Service for National Agricultural Research. Reform in China and Its Implications, East and West Centre, . 1997. "Research, Productivity and Output Growth in Chinese September 16-19, Honolulu. Agriculture." Journal of Development Economics (June). Huffman, Wallace E., and Robert E. Evenson. 1989. "Supply and FAD (Food and Agriculture Organization). 1995. "The State of Food and Demand Functions for Multiproduct U.S. Cash Grain Farms: Bias Agriculture." Rome. Caused by Research and Other Policies." American Journal of - 1996. Balance Sheets, 1992-94. FAD Statistical Series 131. Agricultural Economics 71: 761-73. Rome. IFPRI (International Food Policy Research Institute), Environment and Feng, Lu. 1996. "Feed Demand and its Impact on Grain Economy in Production Technology Division. 1995. "Data Survey and China." Paper presented at a workshop on Agricultural Policies in Preliminary Assessment of Agricultural Investment in China." A China, OECD headquarters, December 12-13, Paris. Report to the United Nations Food and Agriculture Organization. Food Studies Group. 1996. "Global Cereal Markets and Food Security." New York. Paper presented at the International Development Center, Queen Information Office of the State Council. 1996. The Grain Issue in China. Elizabeth House, April, Oxford. Beijing. Garnaut, Ross, and Guonan Ma. 1992. Grain in China. Canberra, Islam, Nurul, ed. 1995. Population and Food in the Early Twenty- Australia: Department of Foreign Affairs and Trade, East Asia first Century: Meeting Future Food Demand of an Increasing Analytical Unit. Population. Washington, D.C.: International Food Policy Glenshaw, Peter A. 1996. "The Role of Fertilizers in Food Security- Research Institute. TheCase of China." Paperpresented by Sulfur Institute atfifth bien- Islam, N., and S. Thomas. 1996. Food Grain Price Stabilization in nial symposium: Sulfur Markets-Today and Tomorrow, March 28, Developing Countries: Issues and Experiences in Asia. Food Policy Washington, D.C. Review 3. Washington, D.C.: International Food Policy Research Griliches, Z. 1958. "Research Costs and Social Returns: Hybrid Corn and Institute. Related Activities." Journal of Political Economics 66(October): Johnson, G. L., and L. Quance. 1972. The Overproduction Trap in U.S. 419-31. Agriculture. Baltimore, Md.: Johns Hopkins University Press. 42 At China's Table: Food Security Options Kairajan, K. P., and Yiping Huang. 1996. An Alternative Method of Rae, Allan N. 1996. "China's Development, Food Consumption and Measuring Economic Efficiency: The Case of Grain Production in China. Livestock-Feeds Policy Interactions." Paper presented at a work- Adelaide, Australia: University of Adelaide. shop on Agricultural Policies in China, OECD headquarters, Lardy, Nicholas. 1983. Agriculture in China's Modern Economic December 12-13, Paris. Development. Cambridge: Cambridge University Press. Rosegrant, M. W., M. AgcaoiLi-Sombilla, and N. D. Perez. 1995. "Global _. 1992. "Rural Reforms and Agricultural Growth in China." Food Projections to 2020: Implications for Investment." Food, American Economic Review 82(1): 34-51, Agriculture and the Environment Discussion Paper 5. International Lin Bao, Lin Jixiong, and Li Jiakang. 1996. The Significance of Balanced Food Policy Research Institute, Washington, D.C. Fertilization, Based on Results from Long-Term Fertilizer Experiments. - 1997. "Global Food Projections to 2020: Implications for Balanced Fertilizer Situation Report 2. Beijing: Chinese Academy of Investment." 2020 Vision Discussion Paper Series 5, International Agricultural Sciences, Soil and Fertilizer Institute. Food Policy Research Institute, Washington, D.C. Lin, Justin Yifu. 1992. "Rural Reforms and Agricultural Growth in Rozelle, Scott, Carl Pray, and Jikun Huang. 1996. "Agricultural China." American Economic Review 82: 34-51. Research Policy in China: Testing the Limits." Paper presented at Lin, Justin Yifu, and Minggao Shen. 1994. "Rice Production the China Workshop-Global Agricultural Science Policy for the Constraints in China: Implication for Biotechnology Initiative." 21st Century, August. 26-28. Melbourne, Australia. Paper presented at the International Workshop on Rice Research Schwartz, L.L., A. Sterns, J.F. Oehmke, and R.D. Freed. 1989. "Impact Prioritization in Asia, February 15-22, Los Bafios, Philippines. Study of the Bean/Cowpea Collaborative Research Support Program Lin, J., J. Huang, and S. Rozele. 1996. "OECD Forum for the Future." (CRSP) for Senegal." Michigan State University, Department of Paper presented at the Conference on China in the 21st Century: Agricultural Economics, East Lansing. Long-term Global Implications, China's Food Economy: Past Scobie, M., and I.R. Posada. 1978. "The Impact of Technical Change Performance and Future Trends, January 8-9, Paris. on Income Distribution: The Case of Rice in Colombia." American Lin, X., C. Yin, and S. D. Xu. 1996. "Maximizing Sustainable Rice Yields Journal of Agricultural Economics 60(1): 85-92. through Improved Soil and Environmental Management." Paper pre- Sicular, Terry. 1991. "China's Agricultural Policy During the Reform sented at the International Symposium on Food and Agriculture in Period." In Joint Economic Committee, Congress of the United China's Perspectives and Policy, October 7-9, Beijing. States, ed., China's Economic Dilemmas in the 1990s: The Problems Liu, Shaoqiu. 1992. "The Impact of Irrigation on Agricultural of Reforms, Modernization and Interdependence. Washington, D.C.: Productivity in China." Ph.D. Dissertation, Department of U.S. Government Printing Office. Agricultural Economics, University of the Philippines, Los Baios. Simpson, J. R. 1996. "Animal Feed Requirements and Availabilities in Lucas, Robert E., Jr. 1967. "Adjustment Costs and Theory of Supply." China-Projections and Policies." Paper presented at a workshop on Journal of Political Economics 75(August): 321-34. Agricultural Policies in China, OECD headquarters, December 12-13, Ma, X. 1996. "Grain Price Variance in China and Its International Paris. Comparison." Paper presented at the International Symposium on Smil, Vaclav, 1993. China's Environmental Crisis: An Inquiry into the Food and Agriculture in China: Perspectives and Policy, October 7, Limits of National Development. New York: ME Sharpe. Beijing. Song, H. 1996. "China's Agricultural Infrastructure Construction and McMillan, J., J. Whalley, and L. Zhu. 1989. "The Impact of China's Investment Policy." Paper presented at a workshop on Agricultural Economic Reforms on Agricultural Productivity Growth." Journal of Policies in China, OECD headquarters, December 12-13, Paris. Political Economy 97: 781-807. Stauffer, Mark D., and James D. Denton. 1995. "Importance of Plant Mei Fangquan. N.d. "Sustainable Food Production and Food Security Nutrients in Increasing Agricultural Productivity-Chinese in China." Chinese Academy of Agricultural Sciences, Beijing. Experience." Paper presented by the Potash and Phosphate Institute - . 1996. "Prospects and Strategic Options of Food and of Canada at the Fertilizer Association of India's annual general Agricultural Production in China." Chinese Academy of Agricultural meeting, December 7-9, New Delhi. Sciences, Beijing. Stone, B. 1993. "Basic Agricultural Technology under Reform." In Y.Y. Nehru, V., and A. Dhareshwar. 1993. "A New Database on Physical Kueh and Robert F. Ash, eds., Economic Trends in Chinese Agriculture: Capital Stock: Sources, Methodology, and Results." Revista de The Impact of Post-Mao Reforms. Oxford: Clarendon Press. Andlisis Econdmico. 8: 37-59. Tang Renjian. 1996. "The Reform on Grain Policy: Difficulties and Pardey, Phil, R. Lindner, E. Abdurachman, S. Wood, S. Fan, W. EveLeens, Objectives-Reflections on China's Grain Problem." Paper presented B. Zhang, and J. Alston. 1992. "The Economic Returns to Indonesian at the International Symposium on Food and Agriculture in China: Rice and Soybean Research." Report prepared for the Agency for Perspectives and Policy, October 7-9, Beijing. Agricultural Research and Development (AARD) and the - . 1996. "The Reform of the GCS in China: Present Situation, International Service for National Agricultural Research (ISNAR). Target and Ways." Paper presented at the International Symposium Potash and Phosphate Institute and Potash and Phosphate Institute on Food and Agriculture in China: Perspectives and Policy, October of Canada. 1996. "Interpretive Summaries: North American and 7-9, Beijing. International Research Projects." Saskatoon. The People's Republic of China. 1996. "Food Security in China." Pray, C.E., and Z. Ahmed. 1991. "Research and Agricultural Contributed to the World Food Summit, November, Rome. Productivity Growth in Bangladesh." In R.E. Evenson and C. E. Pray, ThirtLe, C., and P. BottomLey. 1988. "Is Pubcly Funded Agricultural eds., Research and Productivity in Asian Agriculture. Ithaca, New Research Excessive?" Journal of Agricultural Economics 31(1): York: Cornell University Press. 99-111. - 1991. "Research, Productivity and Incomes in Asian Timmer, C. Peter. 1990. "How Does Indonesia Set Its Rice Price? The Agriculture." In R.E. Evenson and C. E. Pray, eds., Research and Role of Markets and Government Policy." Paper presented at the Productivity in Asian Agriculture. Ithaca, New York: Cornell International Conference on the Economic Policy Making Process in University Press. Indonesia, September 6-9, Bali. Prosterman, Roy, and Tim Hanstad. 1995. "Land Reform: Neglected, - . 1991. "Food Price Stabilization: Rationale, Design and Yet Essential." Rural Development Institute Report 87. Seattle. Implementation." In D. H. Perkins and Michael Roemer, eds., References 43 Reforming Economic Systems in Developing Countries. Cambridge: . 1995b. "World Population Projection 1994-95." Washington, Harvard Institute for International Development. D.C. . 1996. "Does BULOG Stabilize Rice Prices in Indonesia? Should - . 1997a. "China-Forward with One Spirit: A Strategy for the It Try?" Bulletin of Indonesian Economic Studies 32(2): 45-74. Transport Sector." Report 15959-CHA. East Asia and Pacific Regional Tin, Nguyen, Cheng Enjiang, and Christopher Findlay. 1996. Land Office, China and Mongolia Department, Infrastructure Operations Fragmentation and Farm Productivity in China in the 1990s. Adelaide, Division, Washington, D.C. Australia: University of Adelaide. . 1997b. "China: Long-Term Food Security." Report 16419-CHA. USDA/ERS (U.S. Department of Agriculture, Economic Research East Asia and Pacific Regional Office, China and Mongolia Department, Service). 1996. "Long-Term Projections for International Rural and Social Development Operations Division, Washington, D.C. Agriculture to 2005." ERS Staff Paper 9612. Washington, D.C. -. Various issues. Commodity Markets and the Developing USDA/FAS (U.S. Department of Agriculture, Foreign Agriculture Countries. International Economics Department. Washington, D.C. Service). 1996. "Grain: World Markets and Trade." GF1O-96. Wu, Harry X., and Christopher Findlay. 1996. "Grain Supply and Washington, D.C. Demand in China and Its Implications." Paper presented at the Wang, Zhi, and J. Kinsey. 1994. "Consumption and Saving Behavior International Symposium on Food and Agriculture in China: under Strict and Partial Rationing." China Economic Review 5(1): Perspectives and Policies, October 7-9, Beijing. 83-100. - .1996. "China's Grain Demand and Supply: Trade Implications." Wen, G. J. 1993. "Total Factor Productivity in China's Farming Sector: Paper presented at a workshop on Agricultural Policies in China, 1952-1989." Economic Development and Cultural Change, 42:(1) 1-41. OECD headquarters, December 12-13, Paris. World Bank. 1991. "China-Managing an Agricultural Transformation, Wu, Y., and H. Yang. 1995. "Growth and Productivity in China's Grain Sector Review." Report 8652-CHA. East Asia and Pacific Agriculture: A Review." Working Paper Series 95/2. University of Regional Office, China and Mongolia Department, Agricultural Adelaide, Chinese Economy Research Unit, Adelaide, Australia. Operations Division, Washington, D.C. Zhang, L., J. Huang, and S. Rozette. 1996. "Land Policy and Land Use --. 1993a. "China-Grain Distribution and Marketing Project." in China." Paper presented at a workshop on Agricultural Policies in Staff Appraisal Report 11671-CHA. East Asia and Pacific Regional China, OECD headquarters, December 12-13, Paris. Office, China and Mongolia Department, Agriculture Operations _.1996. "China's Agro-Environmental Policies and Measures for Division, Washington, D.C. Sustainable Agriculture." Paper presented at a workshop on - 1993b. "China-YeLLow River Basin Investment Planning Agricultural Policies in China, OECD headquarters, December 12-13, Study." Report 11146-CHA. East Asia and Pacific Regional Office, Paris. China and Mongolia Department, Agriculture Operations Division, Zhu, D. Qu, and Xing Naiquan. 1996. "New Development in Agricultural Washington, D.C. R&D Policy in China." Paper presented at the China Workshop- - .1995a. Global Economic Prospects and the Developing Countries Global Agricultural Science Policy for the 21st Century, August 1995. Washington, D.C. 26-28, Melbourne, Australia. 44 At China's Table: Food Security Options 2 0 2 -0 THE W ORLD BANK 1818 H Street, N.W. Washington. D.C. 20133 USA Telephone: 202-477 1234 Facsimile: 202- 477 6391 Telex: MCI 64145 WORLDBANK MCI 248423 WORLDBANK Cable Address: INTRAFRAD WASHINGTONDC World Wide Web: http://www.worldbank.org/ E-mail: Books@world bank.org Cover design by Glenn Pierce/The Magazine Group