UNN-103 International Finance for Food Security Barbara Huddleston D. Gale Johnson Shlomo Reutlinger Alberto Valdes .wORLD BANK PUBLICATION FILE C0trY / ; 1, Z iL- / International Finance for Food Securit,y A WORLD BANK PUBLICATION International Finance for Food Security Barbara Huddleston D. Gale Johnson Shlomo Reutlinger Alberto V'ldds Publishedfor The lWorld Bank The Johns Hopkins University Press Baltimore and London Copyright © 1984 by the International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W., Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America The Johns Hopkins University Press Baltimore, Maryland 21218, U.S.A. The views and interpretations in this book are the authors' and should not be attributed to the World Bank, to its affiliated organizations, or to any individ- ual acting on their behalf. First printing April 1984 EDITOR Virginia deHaven Hitchcock FIGURES Pensri Kimpitak COVER DESIGN Joyce C. Eisen Library of Congress Cataloging in Publication Data Main entry under title: International finance for food security. Bibliography: p. Includes index. 1. Food Supply-Developing countries-Finance- International cooperation. I. Huddleston, Barbara, 1939- . II. World Bank. HD9018.D44I57 1984 338.1'81 83-48109 ISBN 0-8018-3070-2 Contents Preface ix 1. Introduction and Summa7y 3 The Financial Facility Concept 7 A Simulation Model 8 The IMF Facility 10 2. Sources of Food Insecurity 14 Increased Variability of International Grain Prices 14 Fluctuations in Domestic Food Production 19 Variability of Food Import Bills 23 Food Import Bills and Foreign Exchange 28 3. Alternative Financing Facilities 36 The Simulation Model 37 Alternative International Financing Facilities 41 Comparison of the Alternative Facilities 43 Interaction of Buffer Stocks and Fi:nance 52 4. An IMF Facility to Finance Excess Food Imrkport Bills 56 The Alternative Proposals 56 Estimated Aggregate Drawings 59 Limitations and Benefits of a Food Financing Facility 60 Justification for the Alternative Selected 64 Determination of Need 66 Effect on the Use of Forward Contracts 68 5. Effects of the IMF Facility on International Market Prices 70 Likelihood of Increased Demand 71 Determinants of the World Price of Wheat 75 Predicted Effect of the Facility 77 Appendix. Methodological Notes 81 The Simulation Model 84 Data and Parameters Used in Simulation Experiments 88 v vi CONTENTS Changes in the World Price of Wheat Caused by the Facility 89 References 93 Index 97 CONTENTS Vii FIGURES 1. Variability in Real Export Prices for WVheat and Rice, 1950 to 1979 15 2. Variability in the Production of Staple Foods, for Selected Countries, 1961 to 1976 21 3. Probability of Consumption Falling below 95 Percent of Trend without Any Financial Facility 40 4. Probability of Consumption Falling below 95 Percent of Trend with Financial Facilities, for Country Policy A 44 5. Probability of Consumption Falling below 95 Percent of Trend with Financial Facilities, for Country Policy B 45 6. Stabilizing Effect of Financial Facilities on Average Food Consumption 46 7. Expected Annual Drawings of Six Countries from International Financial Facilities 47 8. Total Expected Annual Drawings from International Financial Facilities 48 9. Expected Imports under Various International Financial Facilities, for Country Policy A 49 10. Expected Imports under Various International Financial Facilities, for Country Policy B 49 11. Total Expected Imports and Additional Imports under Various International Financial Facilities, for Country Policies A and B 50 12. Drawings from International Financial Facilities in Excess of Stated Amounts at the 10 Percent Level of Probability 51 13. Drawings from International Financial Facilities in Excess of Stated Amounts at the 5 Percent Level of Probability 52 14. Indian Food Consumption under Alternate International Financial Facilities 53 15. Brazilian Food Consumption under Alternate International Financial Facilities 54 16. Relation of Ratio of Stocks to Demand and World Price of Wheat, Net of Time Trend 78 17. Consumption with and without Government Intervention to Stabilize Price and Consumption 86 ViiZ CONTENTS TABLES 1. Variablity in the Consumption of Staple Foods, 1961 to 1976 4 2. Variability in the Production of Staple Foods, 1961 to 1976 20 3. Correlation between Cereal and Total Staple Food Production and between Staple Food Production and Consumption, 1961 to 1976 22 4. Role of Cereal Imports in Meeting Shortfalls in Domestic Cereal Production for the Greatest and Smallest Shortfalls, 1965 to 1977 24 5. Variability of the Cereal Import Bill, 1961 to 1978 26 6. Ratio of Food Imports to Total Export Revenues, 1965 to 1977 31 7. Foreign Exchange Cost of the Excess Food Import Bill, 1967 to 1977 32 8. Hypothetical Withdrawals from the Food Financing Facility under the Two Alternative Proposals 58 9. Composition of the Gross Food Import Billfor Selected Countries, 1961- 65, 1968-72, and 1976-78 62 10. Actual and Trend Volumes and Values of Cereal Imports in Developing Countries, 1965 to 1976 71 11. Actual Volumes of World and Developing-Country Cereal Imports and Unmet Developing-Country Requirements, 1961 to 1978 74 12. Indicators of Food Import Vulnerability for Six Countries, 1976 to 1978 82 13. Basic Country Data Used in Simulation 89 14. Estimated and Actual Wheat Prices Associated with Different Ratios of World Cereal Stocks to World Demand 91 Preface THIS STUDY WAS CONCEIVED as a follow-up to research on the costs and benefits of an insurance scheme to provide food se- curity for developing countries. A food financing facility would provide funds to a developing country to help it pay for cereal imports in years when costs are relatively high because of poor crops or high international market lprices. The study's objective was to consider the merits and probable effect of alternative designs for a food financing facility and to assess the one finally adopted by the International Monetary Fund (IMF) in May 1981. The work was carried out in three stages. First, each of the authors conducted preliminary data analyses and informal interviews with responsible officials in six sample countries expected to benefit from creation of a food financing facility. Second, a simulaiLion model was constructed to analyze the effect of different methods of providing financial insurance against high food import bills in the sample countries. The effects were analyzed under three sets of assumptions about the countries' domestic food policies. Third, a price-forming equation for wheat was derived and used to test the likely effect of the financial facility in world markets under different condi- tions of supply and stocks. The results support the decision of the IMF to integrate a food financing facility with its previously existing compensatory fi- nancing facility for export earnings, but they show that the de- gree to which individual countries will benefit depends greatly on their domestic food policies. The facility is expected to have negligible effect on world markets, except in years when the ix x PREFACE ratio of cereal stocks to anticipated demand is low and drawings from the facility are high. However, individual countries could significantly improve their food security position by using the facility to offset adverse local conditions, even when the world situation is normal. The study was carried out with funding from the Rockefeller Foundation, and the support is gratefully acknowledged. The study seeks to shed light on issues that are pertinent for policy discussion and to contribute to the debate on the merits of the facility. World Bank publishes works that seek to draw out the policy significance of a particular topic, as this study tries to do for the financial facility. With the Rockefeller Foundation, the Bank is a co-funder of the Consultative Group in International Agricultural Research, to which the International Food Policy Research Institute (IFPRI) belongs. The Bank, the Foundation, and the authors therefore agreed that the Bank could properly publish a study that represented a joint effort by one of its staff, two IFPRI staff, and a project leader selected by the Foundation. D. GaleJohnson, professor at the University of Chicago, served as project leader; Shlomo Reutlinger, research economist at the World Bank, developed the country impact simulation model; and Barbara Huddleston and Alberto Valdes, research fellows at IFPRI, prepared the price analysis. All shared equally in pre- paring the final report. The views expressed are those of the authors and not those of any of the institutions with which they are affiliated, nor of the Rockefeller Foundation. The authors gratefully acknowledge the assistance of S. Yalamanchili with the simulation model and of Stephen Haykin for the price-forming equation. International Finance for Food Security 1 Introduction and Summary NATIONAL FOOD POLICIES in developing countries are increas- ingly influenced by global events and internatioial policies be- yond the control of individual governments. During the 1960s and 1970s developing countries stepped up their imports of food substantially, and thus their food supply policies have become more dependent on imports. For some countries the variability in food supplies and consumption presents a serious problem. A country experiences food insecurity when no measures are taken to cushion the effects of production and price variations on consumption. Conversely, food security is the assurance that supplies and financing will be available to mrreet minimally ade- quate consumption requirements without domestic price in- creases, regardless of world market conditions. The main causes of food insecurity in developing countries are fluctuations in domestic production and in the price of im- ported cereals. Taken together, the level of domestic production and the level of imports determine the supply available to meet consumer demand. When supplies are short, domestic prices rise, and people respond by reducing the amount they consume. Although part of the variability in food production is counter- acted by changes in stocks held by farmers, consumers, and marketing agencies as well as in net. trade flows, some instability in food supply remains. This instability will be aggravated if increases in the world price of cereals prevent countries from importing additional quantities to offset domestic shortages. The 3 4 INTRODUCTION AND SUMMARY Table 1. Variability in the Consumption of Staple Foods, 1961 to 1976 Instability in consumption of staple foods Standard Probability of deviation' Coefficient actual consumption Region and (thousands of of variation falling below 95 country metric tons) (percent) percent of trend Asia Bangladesh 1,013 7.6 26 India 5,570 5.3 17 Indonesia 1,204 6.1 21 Korea, Republic of 531 6.5 22 Philippines 192 3.3 6 Sri Lanka 163 8.3 27 North Africa/Middle East Algeria 667 24.6 42 Egypt 1,164 12.6 34 Jordan 88 21.2 40 Libya 115 16.7 38 Morocco 933 19.3 40 Syria 360 18.7 39 Sub-Saharan Africa Ghana 134 6.1 21 Nigeria 965 5.6 19 Senegal 319 15.7 37 Tanzania 517 14.6 37 Upper Volta 126 9.5 30 Zaire 172 4.1 11 Latin America Brazil 1,955 5.8 20 Chile 386 14.4 36 Colombia 147 4.7 14 Guatemala 69 6.9 24 Mexico 757 5.3 17 Peru 110 3.9 10 a. Defined as the standard deviation of the consumption variable C,- C. Source: Alberto Vald6s and Panos Konandreas, "Assessing Food Insecurity Based on National Aggregates in Developing Countries," in Alberto Valdes, ed., Food Security for Developing Countries (Boulder, Col.: Westview Press, 1981), p. 30. INTRODUCTION AND SUMMARY 5 country's balance of payments position may also have an effect on food security if additional demand for foreign exchange for food imports coincides with reduced export earnings. The effect of fluctuations in domestic production on producer incomes is another cause of food insecurity. A poor harvest reduces the ability of rural people to purchase food and causes prices to rise when reductions in domestic output are not offset by other sources of supply. Relatively minor shortfalls in the incomes of small farmers resulting from production shortfalls often considerably reduce food consumptiorn. The bulk of the adjustment to smaller food supplies falls on the poor, whose consumption levels already are low. For them, a drop in con- sumption results in energy-deficient diets, with long-run con- sequences for health and productivity. Unstable levels of food consumption in developing countries have received much attention in recent years. Although there are no precise measures of annual changes in per capita food consumption, the available data, sunimarized in Table 1, indicate a large degree of variability in several developing countries. For example, the last column shows that in Algeria and Morocco per capita food consumption may be 5 percent or more below trend in two years out of five. In India and Nigeria staple food con- sumption will fall 5 percent or more below trend in one year out of five. Variable consumption is the direct consequence of variability in consumers' incomes and of food prices. Thus, the extent to which variable domestic production and world prices result in variable consumption depends very much on the do- mestic policies that a country pursues. In the past, the most frequently advocated remedy for in- creasing food security was to build up large buffer stocks, both on an international scale and in each of the countries with highly variable production. On closer examination, however, the buffer stock option has not proved cost effective or realistic for most countries. The developed countries did not wish to underwrite 6 INTRODUCTION AND SUMMARY the high costs of international stocks that would be large enough to stabilize global supplies and the international price of grains. In any case, they could not agree on a mutually satisfactory formula for sharing costs. National stocks large enough to sta- bilize domestic supplies in each developing country against fluc- tuations in domestic harvests and foreign exchange availability would have required huge investments, which these countries could ill afford to make. To effectively counter fluctuations in consumption among the rural population, countries need to do much more than stabilize the national supply and price of food. When incomes are de- pressed because of a poor harvest, special programs are also needed to maintain incomes, such as assured credit and public employment and procurement at prices that compensate for the reduced volume of sales. When it is not feasible to stabilize incomes, food should be distributed at subsidized prices. Another essential instrument for assuring food security is a well- functioning internal trade sector which minimizes price differ- ences among regions within countries. Thus, on-farm and re- gional stocks, low-cost transport, a smooth flow of information, and a trading sector that responds quickly to imbalances in sup- ply are also essential to food security. Freer trade could contribute significantly to stabilizing inter- national prices. If international trade in agricultural products had been free in the 1970s, price variability in international markets would have been sharply reduced. But neither the in- dustrial nor the developing countries are ready to significantly reduce barriers to trade in food products. Consequently, in- creased food security must be attained by other means. A developing country now can stabilize its food supply in several ways. One is to reduce fluctuations in domestic food production, primarily by increasing irrigation. A second is to build up stocks in years of high production for use during poor years. Both of these options involve a substantial commitment of resources. An alternative is to vary the volume of food imports INTRODUCTION AND SUMMARY 7 inversely with changes in domestic production.' Under existing institutional arrangements, however, developing countries may not have sufficient foreign exchange resources to use this option. A food financing facility would minimize this constraint. The Financial Facility Concept The financial remedy examined and advocated in this mono- graph grows out of a realization that the malady is not so much the international instability in food supplies, but the inability of the poorest and most financially strained countries to secure a minimally adequate food supply. A food financing facility pro- vides the means for such countries to maintain or increase im- ports when domestic supply is inaclequate. The concept of a foreign exchange constraint has meaning for food imports and food security because it is assumed that the exchange rate is not and should not be adjusted to offset a short-run unexpected increase in the food irnport bill. Further- more, since low-income countries often cannot borrow on short notice in international capital markets at average lending rates when food import bills are abnormally high, international fi- nancial arrangements are needed to avoid reductions in consumption. Many countries, particularly middle-income de- veloping countries, can draw on foreign currency reserves in years when their aggregate import bill increases sharply. For the poorest developing countries, with low reserve ratios, however, the flexibility to use reserve holdings is, naturally, limited. Of course, a scheme to cover the excess cost of cereal imports can stabilize consumption for low-income families only if their purchasing power is maintained. A financial facility to reduce 1. D. Bigman and Shlomo Reutlinger, "Food Price Stabilization: National Buffer Stocks and Trade Policies," American Journal cf Agricultural Economics, vol. 61, no. 4 (November 1979). 8 INTRODUCTION AND SUMMARY constraints on food imports is by no means the only thing nec- essary to achieve food security. Fluctuations in consumption will persist unless incomes and food prices are stabilized. Income stabilization requires such measures as public works, crop in- surance, an adequate supply of credit, and publicly sponsored food distribution schemes. Price stabilization is promoted by improved transport and communication, which facilitate trading within a country and the imposition of subsidies for imports when world prices rise or of taxes when they fall. Yet even where policies to stabilize consumption are still rudimentary, deterio- rating nutritional status in urban and rural areas can be alleviated if the country can import enough food to maintain stable do- mestic prices, regardless of fluctuations in domestic production or world markets. It is primarily in this context that international arrangements specifically designed to facilitate adequate food imports are being advocated. More cost effective than buffer stocks, a reliable financial facility enables countries to invest more of their scarce capital in agricultural and general devel- opment, and less in the construction of storage buildings and the maintenance of large grain inventories. A Simulation Model This study uses a simulation model to analyze the probable benefits and the pursuant demands on an internationally spon- sored scheme for insuring against high food import bills. In this model, the instability of food consumption has been assumed to result primarily from variations in the domestic price of food. These prices are assumed to fluctuate either because the world price fluctuates or because the domestic supply plus imports differs from the amount that people will consume at the pre- vailing world price. In both cases, government policies and bal- ance of payment constraints play an important role. To insulate the domestic price from world market prices, governments must INTRODUCTION AND SUMMARY 9 subsidize domestic food prices or tax imports. To permit enough imports to satisfy demand at the world market price, there must be unrestricted allocations of foreign exchange. The model allows for random fluctuations in the domestic harvest, the world price, and availability of foreign exchange. The incidence of food insecurity (that is, the probability of food consumption falling below a specified level) is estimated under different specifications of government policy affecting the in- sulation of domestic price from the world price and the allocation of foreign exchange to food and nonfood imports. It is also estimated for different possible schemes for reducing foreign exchange constraints. Three possible schemes to reduce foreign exchange con- straints have been specified: compensation for shortfalls in for- eign exchange earnings (similar to the original compensatory financing facility, CFF, of the International Monetary Fund, IMF); compensation for excessive food import bills; and compensation for any shortfalls in foreign exchange arising because the al- gebraic sum of export earnings and food import bills is less than normal (approximating the expanded CFF recently created in the IMF). For each financial scheme the model also estimates the amount of funds that would be withdrawn from a financial facility and the additional amount of imports that would be necessary. The simulation experiments for six sample countries illustrate the strong influence of a government's policies on its food se- curity and on its likely drawings from an international financing facility. Differences in food security among countries are caused more by different policies pursuecl by their governments than by differences in the stability of food production and the avail- ability of foreign exchange. The simulation experiments further illustrate that food se- curity will be increased if countries have access to an integrated CFF, whether or not countries pursue policies to increase their food security. If a country does not insulate its internal food price from fluctuations in the international price and gives low 10 INTRODUCTION AND SUMMARY priority to food imports, however, food insecurity will persist despite access to a financial facility. The gains in food security will be most pronounced if the existence of the financial facility encourages countries to adopt policies to increase their food security. For the six sample countries, the expansion of a facility from one that compensates only for shortfalls in foreign exchange earnings to one that compensates for the algebraic sum of short- falls arising from variations in foreign exchange earnings and food import bills implies an approximate doubling of expected annual drawings. Both the gains in food security and the ex- pected drawings were calculated on the assumption that draw- ings are not constrained by a quota. Without quota limits the expansion of the CFF is expected to increase annual food imports from 20 to 45 percent of current levels. The smaller gain applies if countries were already pursuing policies to increase their food security before a financial facility was introduced. In this case, drawings from the facility are to a larger extent used to reduce the disruption in nonfood imports caused by these policies. The simulation experiments yielded a set of predicted changes in food security that would result from new financial arrange- ments. Since the results are sensitive to the domestic food poli- cies of the individual governments, the predictions are only con- ditional. They must be supplemented with speculations about the prevailing policies with and without a financial facility in the respective countries. In any case, the simulations anticipate the possible consequences of the financial facility and suggest com- plementary steps to achieve food security. The IMF Facility In May 1981 the IMF created an integrated financial facility for food imports by amending its CFF for export earnings to cover excess cereal import bills. Under the old rules the facility pro- INTRODUCTION AND SUMMARY 11 vided foreign exchange credit to cover shortfalls from trend in export earnings. Under the new rules, the shortfall in net export earnings would be calculated as trend exports minus actual ex- ports plus actual cereal import costs minus trend cereal import costs. Credit would then be provided to cover the shortfall in net export earnings. The total amount of t:he drawing is con- strained to 100 percent of quota for each of the two components and 125 percent of quota for the combined total. Countries can choose whether to draw under the new or the old rules, de- pending on which seem more beneficial. Once having chosen, a country cannot select the other option for three years. Re- payment at below market rates is required within three to five years after the initial drawing. A second proposal for creating a separate food financing fa- cility within the IMF was rejected by the Fund's board. Under this proposal, the new facility would have provided credit equal to the excess of current cereal import bills over their trend value. This study indicates that the integrated facility as adopted is preferable to a separate facility with comparable quota limits. The integrated scheme would result in lower average drawings for food imports than a separate facility because high food import bills are sometimes offset by high export earnings. However, the integrated scheme allows for substantially larger drawings in years when cereal production is short, cereal import prices are high, and foreign exchange is not readily available. In such years, the integrated scheme would supply most of the required foreign exchange, providing that quota limits are noit unduly restricting. This arrangement could provide food secturity to developing countries when they most need it, but could still keep operating costs reasonable. The financial facility adopted by the IMF will provide assistance only when import bills are above normal for reasons beyond control of the governments requesting help. Benefits will not be available to maintain or increase per capita consumption when government policies and economic trends lead to declines in the 12 INTRODUCTION AND SUMMARY country's normal degree of self-sufficiency. A financial facility would help create conditions that favor food security. To realize these advantages, however, many countries would need to sta- bilize food consumption and prices. Because of the difficulties in achieving these objectives, use of the facility will probably grow slowly. The facility will complement existing food aid, which is likely to decrease in volume when cereal prices rise. Food financing is less subject to political criteria than food aid. An international agreement on holding grain reserves would not be required. Undoubtedly, larger grain reserves could reduce the variations in world price, but the tangible effects of the agreement envi- sioned under the Wheat Trade Convention are questionable.2 This analysis indicates that world price reacts to reductions in world stock levels only when the initial level is relatively low compared with estimated world demand. As the ratio of stocks to demand falls below about 17 to 20 percent, prices start to increase at a faster rate. During the 1970s world stocks declined relative to world consumption. Depending on the production policies of major exporters and the consumption policies of centrally planned importers, stocks may not increase relative to consumption for the foreseeable future. If this is so, greater variations in world prices can be expected. The IMF food facility will not prevent international price in- creases caused by emergencies. If such price increases occur, however, the facility should ease the strain for cereal-importing countries whose costs suddenly rise sharply. Even when cereals are in short supply worldwide, the increase in imports financed by a facility would be small relative to the volume of international trade. At such times, however, the ratio of world stocks to de- mand will probably be low, and the shock of an additional 10 2. Daniel T. Morrow, The Economics of the International Stockholding of Wheat, Research Report no. 18 (Washington, D.C.: International Food Policy Research Institute, 1980). INTRODUCTION AND SUMMARY 13 to 20 million tons of demand at such times could affect prices noticeably. In most years neither the size of stocks nor the shift in demand are likely to be extrenme, and t]he effect on world price would be small. For individual countries receiving assis- tance, however, the additional imports might be significant and could mean the difference between adequate consumption or deprivation for many poor people in the developing world. 2 Sources of Food Insecurity DURING THE TWO DECADES that ended in 1972 low-income countries whose food production varied significantly from year to year could be assured that grain would be available at prices that changed little from year to year. For example, between 1960 and 1971 the largest year-to-year percentage change in U.S. export prices of corn and wheat was 16 percent. The highest annual average export price for wheat was US$67 per metric ton in 1966, and the lowest was US$53 in 1969. For corn, the range was from US$61 in 1971, when the corn blight reduced IJ.S. production, to US$47 in 1960. World market prices were stable even during the early and mid-1960s when world grain production was significantly below normal. Except for 1951 and 1952, during the Korean War, world market prices for grains also were stable during the 1950s. Increased Variability of International Grain Prices The situation has changed significantly since 1971 as inter- national grain prices have become highly variable. The U.S. export price of wheat more than doubled in current dollars between 1972 and 1974 and then declined 71 percent by 1977 (more than 50 percent in constant dollars). Between August 1977 and August 1979 the export price of wheat again increased 80 percent. For corn, the price rose more than 130 percent between 1972 and 1974, and then declined. During the calendar year 14 SOURCES OF FOOD INSECURITY 15 1980 corn export prices increased by almcst 50 percent, pri- marily in response to a reduced U.S. corn crop and a second year of low yields in the Soviet Union. In addition, variability in ocean shipping rates also appears to have increased, although significantly less than that of grain prices. The coefficients of variation shown in Figure 1 drama- tize the increased variability of export prices for wheat and rice. During the 1970s wheat prices varied more than eight times as much as during the 1960s. For rice, the export price variability more than doubled. Figure 1. Variability in Real Export Prices for Wlheat and Rice, 1950 to 1979 -z 40- 30 > 20 to 0 1950-59 1960-69 1970-79 .Vote: O wheat; C3 rice. Wheat Rice Coefficient of Standarcd Coefficient of Standard variation deviation (U.S. variaiiorn deviation (U.S. Years (percent) dollars per ton) (percent) dollars per ton) 1950-59 11.2 26.0 11.4 59.0 1960-69 3.6 7.0 17.5 89.0 1970-79 30.0 56.0 39.0 187.6 Source: World Bank data. 16 SOURCES OF FOOD INSECURITY International Price Stability up to the Mid-1970s The major reasons grain prices were stable during 1952-72 were the support policies of the United States and Canada. Very large quantities of grain were either owned or controlled by the two governments. In addition, the major exporters could pro- duce more grain than was demanded at politically acceptable prices. The tendency for stocks to grow was somewhat abated by efforts to limit output, especially in the United States, and by extensive use of food aid. Price stability was not the objective of these farm price policies, but was instead a consequence of agricultural price and income policies designed to bring grain production into better balance with demand. This goal was largely achieved by the early 1970s, although it was perhaps not fully appreciated at the time. Prices were stable during the 1960s despite unusually unstable grain production throughout the world. Shortfalls in grain pro- duction below normal during 1961/62-1965/66 exceeded those for 1970/71-1974/75 in absolute and relative terms. In the earlier period the algebraic sum of the departures above and below trend was 72 million metric tons compared with 62 million metric tons in 1970/71-1974/75. The relative shortfall also was greater in the 1960s because grain production grew by one third in the early 1970s. Consequently, the year-to-year fluctuations have gradually become smaller. Several factors have been re- sponsible for this greater stability in production: mechanization in industrial countries, which has permitted more timely farming operations; summer fallow and other moisture-conserving meth- ods of cultivation; and improved herbicides and insecticides. By the beginning of the 1970s the major grain exporters were no longer willing to carry stocks as large as those of the early 1 960s. As grain stocks increased from 70 million metric tons in 1]967 to 105 million metric tons in 1969, Australia, Canada, and the United States made major efforts to reduce the production of wheat through programs to limit output or acreage. Before 1970 neither Canada nor Australia had made significant efforts SOURCES OF FOOD INSECURITY 17 to limit wheat production. Wheat and feed grain stocks held by the major exporters in 1961/62 were 14.4 percent of world production, compared with 9.6 percent in 1970/71 and 8.6 per- cent in 1972/73. International Price Instability since the Mid-1970s A change in U.S. grain storage policies in the 1977 farm leg- islation will contribute to increasingly variable prices in the years ahead. The change reflected the belief that setting price supports near market prices had forced the United States to carry a large share of the world's grain stocks and, at the same time, to become a residual supplier in export markets. This was believed to have resulted in larger-than-desired expenditures for storage and sup- ply management. In addition, farmers concluded that U.S. stor- age policies that led to stable prices also meant low prices. The 1977 U.S. farm legislation may have been the first in modern times that was deliberately designed to increase price instability. The farmer-held grain reserve created by the 1977 act sub- stantially increased the spread between the loan or support level and the prices at which grain either could be or would be re- moved from U.S. reserve stocks.' The Comrmodity Credit Cor- poration, which previously owned or corntrolled most grain stocks, was prohibited from selling any of its grain at less than 150 percent of the support price as long as there was any grain in the farmer-held reserve. And once grain was put into the farmer-held reserve, the government couldl not call the loans until the market prices were 40 percent above the support price for coarse grains and 75 percent above it For wheat. Thus the range of price variability provided for in the reserve program is 1. The term "farmer-held reserve" refers to stocks of grain owned by farmers but obligated to the government as collateral for price support loans. Previously, the government took possession of such grain when loans came due. Now, farmers retain the grain and receive a storage subsidy from the government. In exchange, they agree not to sell the grain until prices reach a prescribed level. 18 SOURCES OF FOOD INSECURITY substantially greater than that of the 1950s and 1960s. Some proposals call for widening the price ranges further. No other major grain exporter has been willing to increase its stocks to offset the lower U.S. level, and there is no particular reason to expect any to do so. National policies to stabilize domestic prices of grains and other agricultural products also have increased price instability in international grain markets. Except for Canada, Australia, the United States, and, in recent years, India, governments have not been willing to hold sufficient stocks to offset variations in do- mestic supply and demand. Instead, governments have sought to achieve price stabilization by varying net trade. Thus, the internal instability of these nations has been reflected in the international market and in the economies of countries that permit their domestic prices to vary with the international prices. During the 1960s and 1970s an increasing share of the world's grain was produced and consumed in nations that achieved in- ternal price stability through managed trade. Basic agricultural policies changed little, but the ability and the will to pursue more effective domestic price stabilization policies increased. For ex- ample, the basic features of the price policies for food and ag- riculture in the Soviet Union were essentially the same in 1972 as in 1963. Prices paid to producers, prices of farm inputs, and consumer prices were fixed. In 1972, however, there was a much greater effort to establish internal prices at levels that more nearly equated supply to demand. In the earlier period sub- stantial shortfalls of supply relative to demand were tolerated; but in the later period serious efforts were made to eliminate or minimize them. Thus both the mean level and the year-to- year variations in net grain trade became much greater during the 1970s than during the 1960s. Similar changes in price stabilization policies occurred in the European Economic Community (EEC). In addition, an increas- ing number of developing countries have either instituted price stabilization policies or made existing programs more effective. SOURCES OF FOOD INSECURITY 19 There is no reason to expect that. the international price in- stability caused by various national policies will diminish during the 1980s. If anything, it is likely to increase. Nor is there any reason to anticipate that the storage policies of the major grain exporters will change in such a way as to increase stock levels. There now appears to be a rough balance between supply and demand for grains produced by the four major grain exporters- the United States, Canada, Australiai, and France-at politically acceptable price levels. As long as this situation prevails, there is no compelling reason for the major grain exporters to bear the costs of larger grain stocks.2 The increased price instability in international grain markets will cause greater variability in the grain and food import bills of the developing countries. In contrast to the 1960s, when food import bills varied primarily because of quantity changes, vari- ation since then has been caused by both quantity and price. Fluctuations in Domestic Food Production Food production varies greatly foir some individual countries but relatively little at the global level. This is the case for certain developed countries, such as the Soviet Union, as well as for many developing countries. Table 2 indicates that in ten of the twenty-four countries analyzed, production fell below 95 percent of trend approximately once every three years. Production has been more stable in large countries, such as India, Indonesia, Brazil, Bangladesh, and in those where a significant portion of cropland is irrigated, such as Egypt. The coeflicient of variation for production in these countries is about 6 percent. In contrast, in several Arab countries the coefficient of variation is about 20 2. D. Gale Johnson, "The World Food Situation: Developments during the 1970s and Prospects for the 1980s," in AEl Studies on Contemporary Economic Problems (Washington, D.C.: American Enterprise Institute, 1980). 20 SOURCES OF FOOD INSECURITY Table 2. Variability in the Production of Staple Foods, 1961 to 1976 Instability in production of staple foods Standard Probability of deviationa Coefficient actual production Region and (thousands of of variation falling below 95 country metric tons) (percent)b percent of trend Asia Bangladesh 765 6.4 22 India 6,653 6.4 22 Indonesia 1,040 5.4 18 Korea, Republic of 445 7.1 24 Philippines 346 5.7 19 Sri Lanka 107 9.3 29 North Africa/Middle East Algeria 531 28.9 43 Egypt 282 4.5 13 Jordan 119 65.6 47 Libya 56 28.0 43 Morocco 1,156 27.2 43 Syria 702 38.8 45 Sub-Saharan Africa Ghana 121 5.8 20 Nigeria 958 5.7 19 Senegal 325 18.6 39 Tanzania 430 12.7 36 Upper Volta 128 9.8 30 Zaire 190 4.9 15 Latin America Brazil 1,631 5.2 17 Chile 215 11.1 33 Colombia 126 4.4 13 Guatemala 56 6.5 22 Mexico 1,060 7.7 26 Peru 197 9.8 30 a. Defined as the standard deviation of the production variable Q, -Q. b. Defined as the standard deviation of the variable (Q, - Q)/Q, *100. Source: Alberto Valdes and Panos Konandreas, "Assessing Food Insecurity Based on National Aggregates in Developing Countries," in Alberto Valdes, ed., Food Security for Developing Countries (Boulder, Col.: Westview Press, 1981). SOURCES OF FOOD INSECURITY 21 percent. Similarly, it is high in several sub-Saharan African coun- tries, such as Senegal and Tanzania. The variability in the pro- duction of staple foods in ten selected countries is shown in Figure 2. The historical record is unequivocal as to the effect of unstable cereal production on food consumption. As shown in Table 3, changes in the production of cereals are highly correlated with changes in total staple production in almost every country, although the proportion of cereals to other staples varies by country. Cereal imports generally have been too small to compensate for shortfalls in production. Consequently, variability in con- sumption of cereals is highly correlated with changes in food production (see Table 3). Table 4 provides data on the level of the shortfalls that are made up each year by commercial imports Figure 2. J"ariabilitv in the Production of Staple Foods, for Selected Countries, 1961 to 1976 80 60 ='40 20 I Bangladesh Algeria 1o.rdanI Sy ria Brazil India Egvpt Libya Senegal Peru 22 SOURCES OF FOOD INSECURITY Table 3. Correlation between Cereal and Total Staple Food Production and between Staple Food Production and Consumption, 1961 to 1976 Correlation Correlation coefficient between Correlation coefficient between total staple food coefficient between cereal production production and export earnings Region and and total staple staple food and the food country food production' consumption' import billb Asia Bangladesh 0.99 0.90 0.32 India 0.99 0.89 0.55 Indonesia 0.94 0.92 0.23 Korea, Republic of 0.96 0.20 0.21 Philippines 0.99 0.97 0.32 Sri Lanka 0.91 0.56 0.57 North Africa/Middle East Algeria 1.00 0.78 0.76 Egypt 0.96 0.29 0.49 Jordan 1.00 0.63 0.63 Libya 1.00 0.62 0.31 Morocco 0.96 0.98 0.32 Syria 1.00 0.92 0.13 Sub-Saharan Africa Ghana 0.93 0.98 0.38 Nigeria 0.92 0.99 -0.27 Senegal 0.81 0.99 -0.08 Tanzania 0.09 0.98 -0.65 Upper Volta 0.99 0.95 0.31 Zaire -0.21 0.96 -0.15 Latin America Brazil 0.60 0.92 0.59 Chile 0.99 0.54 0.60 Colombia 0.85 0.51 -0.11 Guatemala 0.99 0.51 0.40 Mexico 1.00 0.53 0.15 Peru 0.97 0.37 -0.23 a. Staple food is defined as cereals, pulses, roots and tubers, and groundnuts. b. Export earnings include goods and services. Source: Alberto Vald6s and Panos Konandreas, "Assessing Food Insecurity Based on National Aggregates in Developing Countries," in Alberto Valdes, ed., Food Secunty for Developing Countries (Boulder, Col.: Westview Press, 1981). SOURCES OF FOOD INSECURITY 23 and food aid. Because of variations in production and stocks, this amount of imports exceeds requirements for some years, whereas in other years it is relatively low. There are several possible reasons why consumption has been highly correlated with production in most of the countries studied: cereal production may be highly correlated with con- sumers' disposable income; cereal production rnay have a strong negative correlation with cereal import prices; market channels may not be flexible enough to handle large year-to-year changes in imported cereals; and, of course, foreign exchange allocations to food imports may have been overly restrictive. Because prices have been relatively stable for most of the period covered, nega- tive correlation between production and import prices would not appear to be a significant factor. Variability of Food Import Bills Although changes in production are the main factors that induce variations in a country's food imports, the import bill is determined by both prices and quantities imported. During the 1960s, when prices were stable, variable import volumes were responsible for almost all of the variability in the cereal import bill for a sample of the developing countries (Table 5). In five of the eighteen countries, changes in import volume were re- sponsible for at least 75 percent of the variability in the cereal import bill. In contrast, during the 1970s, price changes were responsible for more than 50 percent of the total varialbility of the food import bill in ten out of the eighteen countries. Variability in import volume was still substantial and clearly dominant for several countries, such as Tanzania, Upper Volta, and Mali. In both periods, however, food import bills would have been much higher if countries had sought to obtain food security during 24 SOURCES OF FOOD INSECURITY Table 4. Role of Cereal Imports in Meeting Shortfalls in Domestic Cereal Production for the Greatest and Smallest Shortfalls, 1965 to 1977 Greatest shortfall Level of Commercial shortfall imports Food aid (thousands (thousands (thousands Region and of metric of metric of metnic country Year tons) tons) tons) Asia Bangladesh' 1971 3,536 1,164 n.a. India 1965 17,820 1,063 6,814 Indonesia 1967 2,112 497 108 Philippines 1966 983 578 47 Sri Lanka 1975 1,121 977 188 North Africa/Middle East Egypt 1973 4,470 1,858 14 Jordan 1977 329 238 147 Somalia 1977 110 120 17 Syria 1971 1,009 794 8 Sub-Saharan Africa Mali 1972 136 30 40 Senegal 1977 338 405 15 Tanzania 1974 322 430 1 Upper Volta 1977 116 45 6 Latin America Bolivia 1977 284 230 9 Brazil 1966 3,490 2,088 365 Chile 1975 822 401 314 Haiti 1977 222 107 80 Peru 1975 811 1,130 13 .Vote: The shortfall is defined as the difference between constant consumption of cereal per capita (using a 1961-65 base) and the domestic supply of cereal. The supply includes available data on changes in stocks as well as in the levels of domestic cereal production. Data on stock changes may not always be reliable, as they are often imputed from other supply data rather than reported directly. Cereals include wheat, wheat flour (in wheat equivalents), rice, maize, rye, and cereals not elsewhere specified. SOURCES OF FOOD INSECURITY 25 Smallest shortfall Level of Commercial shortfall' imports Food aid Number of (thousands (thousands (thousands years with of metric of metric of metric uncompensated' Year tons) tons) tons) shortfall 1973 1,871 1,921 935 5 1977 -9,997 30 939 2 1977 - 1,605 1,874 924 2 1977 -475 838 30 1 1968 718 774 180 4 1968 1,551 2,416 1 4 1965 104 123 16 5 1972 17 53 15 4 1976 120 195 77 9 1965 5 22 0 7 1967 185 191 46 4 1966 -68 74 5 6 1968 -15 20 0 8 1965 148 68 99 7 1977 -4,414 2,644 5 1 1971 -73 475 13 4 1967 42 36 1 9 1967 430 547 7 4 n.a. Not available. a. A negative shortfall indicates a surplus supply of domestic cereal over the constant level of cereal required per capita. b. Includes 1971-77 data. Source: Data compiled by Grant Scobie and Alberto Valdes, IFPRI. 26 SOURCES OF FOOD INSECURITY Table 5. Variability of the Cereal Import Bill, 1961 to 1978 (percent) Cause of variability of import bill 1961-70 Region and country Volume Price Interaction Asia Bangladesh 97.9 2.2 -16.3 India 98.7 1.3 -43.3 Indonesia 90.6 9.4 5.4 Philippines 78.2 21.8 108.3 Sri Lanka 91.4 8.6 -20.2 North Africa/Middle East Egypt 91.7 8.3 183.1 Jordan 92.0 8.0 31.8 Somalia 90.1 9.9 56.7 Syria 82.0 18.0 -58.2 Sub-Saharan Africa Mali 77.4 22.6 103.8 Senegal 42.5 57.5 58.4 Tanzania 93.3 6.7 16.6 Upper Volta 44.4 55.6 182.8 Latin America Bolivia 73.0 27.0 99.2 Brazil 86.2 13.8 358.6 Chile 90.6 9.4 -20.0 Haiti 84.9 15.1 118.1 Peru 78.6 21.4 94.6 NVote: These figures are obtained by dividing the variance of the actual food import bill by the variance of the import volume (times the average price) and the variance of the import price (times the average volume of imports). The methodology is found in Valdes and Konandreas, "Assessing Food Insecurity." The quoted prices of imports used here represent their world prices and not SOURCES OF FOOD INSECURITY 27 Cause of variability of import bill 1970-78 19,61-78 Volume Price Interaction Volume Price Interaction 52.7 47.3 -12.5 57.3 42.7 20.1 83.3 16.7 21.5 65.0 35.0 -21.3 41.7 58.3 -61.3 40.5 59.5 -52.0 34.8 65.2 -76.6 31.7 68.3 -59.4 13.0 87.0 -22.7 11.2 88.9 -36.1 35.4 64.6 3.3 25.8 74.2 22.8 41.3 58.7 -45.1 34.7 6'5.3 -41.6 39.6 60.4 -62.9 39.7 60.3 -45.1 52.3 47.7 -83.9 49.0 51.1 -75.2 84.6 15.5 96.9 82.6 17.4 263.1 20.7 79.3 -55.1 13.7 86;.3 -57.3 76.3 23.7 117.7 74.9 25.1 229.6 59.1 40.9 -48.4 53.2 41.8 -16.6 14.1 85.9 -39.7 9.4 99.6 -50.7 51.4 48.6 -20.7 32.9 67.1 -16.2 57.2 42.8 60.9 47.3 52.7 118.2 36.8 63.2 -15.7 29.3 70.7 11.4 26.7 73.3 55.6 22.1 77.9 69.8 the true cost of the portion imported under iFood aid programs. It is assumed that import volumes including food aid can be treated as the minimum desired level that developing countries would have imported even without food aid. Source: Data compiled by Grant Scobie and Alberto Valdes, IFPRI. 28 SOURCES OF FOOD INSECURITY the period. In fact, preliminary calculations suggest average an- nual increases for 1965-76 of about US$0.2 to US$1.1 billion.3 An increase in world carryover stocks, presumably the objec- tive of a new Wheat Trade Convention, could conceivably benefit developing countries by decreasing the variability of world prices of wheat. Nevertheless, even if the variation in world prices is reduced, the variation in a country's own levels of production must still be dealt with. Food Import Bills and Foreign Exchange The availability of foreign exchange could be the most critical factor in determining whether a country can import enough food to stabilize consumption. The average ratios of the food import bill to total export revenues (goods and services) presented in Table 6 for twenty-four developing countries indicate the pres- sure of food imports on their balances of trade. The ratios underestimate the pressure exerted by the food imports that would have been needed to stabilize consumption in 1965 to 1977 because these imports already had been affected by short- ages of foreign exchange. 3. Alberto Valdes and Panos Konandreas, "Assessing Food Insecurity Based on National Aggregates in Developing Countries," in Alberto Valdes, ed., Food Security for Developing Countries (Boulder, Col.: Westview Press, 1981), table 2.8. Estimates of the actual increase in the net import bill depend on country coverage and assumptions about the responses of these countries to fluctuations in do- mestic production and world market price. The net increase in the import bill is defined as the increase in the import bill offset by any increase in export earnings during the period in question. Both estimates shown are calculated for sixty-seven developing countries (excluding eight major oil exporters). The lower estimate is calculated using historical import levels for these sixty-seven countries during this twelve-year period. The larger estimate is calculated assuming that the import levels of the countries would have stabilized consumption at levels consistent with the historical growth rate. This means that imports are adjusted upward for years when there is a shortfall in actual consumption, but are adjusted downward when actual consumption levels rise above trend levels. SOURCES OF FOOD INSECURITY 29 A large food import bill could be simall relative to a country's total export revenues, or vice versa. Similarly, the fluctuations of a country's food import bill may coincide with or diverge from fluctuations in its export earnings or with other important im- ports, such as petroleum products. The higher the food import bill relative to the supply of foreign exchange, or the lower the correlation between the food import bill and foreign exchange earnings, the more severe is the problem of food imports. In general, food import bills are positively correlated to export earnings in most countries. This implies that, for those countries, in years when the food import bill is higher, foreign exchange earnings are also higher. The mean ratios in Table 6 for the normal years between 1965 and 1977 do not suggest a severe constraint fo:r some countries, particularly in Latin America. But the high ratios for many low- income countries, notably Egypt, Sri Lanka, Bangladesh, and India, do. The fact that food aid is used by some countries to ease the constraint does not lessen the irnportance of the problem. In years of higher import prices, clomestic production short- falls, or lower export revenues, the pressure of the food import bill on a country's overall import capacity was considerable. For some countries with low average ratios, such as Syria, the ratio increased by a factor of three or four in exceptionally unfavorable years. For several low-income countries, particularly in Asia and Africa, this ratio was intolerably high in some years. For a few developing countries, such as Nigeria, Libya., and Colombia, however, the ratio was remarkably low even at its maximum value. The short-term implications of a large and unexpected in- crease in the food import bill can be portrayed in terms of the foreign exchange cost of the excess food import bill relative to the supply of foreign exchange.4 Table 7 shows this for eighteen 4. Defined as [ P,, (R,, - (QP,, - AS,) -M,,) 1 /F, where P,, is the unit import 30 SOURCES OF FOOD INSECURITY developing countries. For several of these countries the extra cost of cereal imports represented 10 percent and sometimes as much as 20 or 30 percent of the supply of foreign exchange. This was over and above the trend value of cereal imports in those years and represents unexpected short-run excess demand for foreign exchange. A similar calculation for foreign exchange reserves showed that in some years the extra cost of cereal imports represented from 50 to 80 percent of the year's foreign exchange reserves (in India in 1975, Sri Lanka in 1973, and Haiti in 1977). In 1975 it even reached more than 100 percent for Tanzania, Sri Lanka, and Haiti. It remained a small fraction of reserves throughout the period for Peru, Chile, and the Philippines. Two qualifica- tions are relevant here. First, as a proxy for import capacity more weight is given to the supply of foreign exchange than to re- serves. A constant level of reserves is compatible with wide fluc- tuations in export earnings. Second, foreign exchange earnings, ideally, should be adjusted for changes in the nonfood import bill. Thus, although all of the countries are food-deficit countries, the sample includes countries whose individual problems and prospects are far from identical. The flexibility of some of the countries in their use of reserve holdings has naturally been rather limited, particLlarly for the poorest developing countries whose reserve ratios have been low. Thus, the brunt of the adjustment is made in the allocation of foreign exchange among imports. In many developing countries the problems created by the increase in the world price of cereals were compounded by the shortfalls in the domestic production of food. For some coun- price in country i in year t; R,, represents requirements in metric tons of cereals in country i in year t reflecting annual population and income growth using FAO estimates of income elasticity of demand for cereals in 1975; QP,, and AS,, rep- resent domestic production and changes in stocks, respectively; Mj, represents a three-year moving average of cereal imports (both commercial imports and food aid); and F represents the supply of foreign exchange as defined in Table 7. All food imports are listed at commercial prices, including food aid. SOURCES OF FOOD INSECURITY 31 Table 6. Ratio of Food Imports to Total Export R,evenues, 1965 to 1977 (percent) 1965-77 Region and country 1965-67 1970-72 1975-77 Mean Maximum Asia Bangladesh n.a. n.a. 67 77a 123 (1975) India 40 10 24 23b 45 (1966) Indonesia 7 13 6 11 21 (1968) Korea, Republic of 13 16 7 13 22 (1969) Philippines 7 5 4 5 9 (1965) Sri Lanka 25 25 36 30 50 (1975) North Africa/Middle East Algeria 8 6 9 7' 10 (1973) Egypt 23 12 20 17 32 (1974) Jordan 13 17 7 12' 18 (1972) Libya 2 1 2 2c 3 (1975) Morocco 9 6 2 8 14 (1975) Syria 8 14 7 9 19 (1971) Sub-Saharan Africa Ghana 5 3 4 4c 6 (1977) Nigeria 3 2 2 2' 3 (1977) Senegal n.a. 10 10 lod 20 (1973) Tanzania 3 3 10 6' 22 (1974) Upper Volta n.a. 8 9 9' 19 (1974) Zaire 3 3 6 3f 6 (1975) Latin America Brazil 10 4 4 6 11 (1967) Chile 4 4 6 6 14 (1974) Colombia 3 3 4 3 5 (1974) Guatemala 3 2 3 3 4 (1975) Mexico 1 2 5 3' 10 (1975) Peru 6 5 11 7' 16 (1975) Note: Export revenues include goods and services (except Brazil for 1965-67). The food import bill includes cereal and olther staples. All food imports are listed at commercial prices, including food aid. n.a. Not available. a. 1973-77. c. 1967-77. e. 1968-75. b. 1965-76. d. 1968-76. f. 1965-75. Source: Ammar Siamwalla and Alberto Valdes, "Food Insecurity in Developing Countries," Food Policy, vol. 5, no. 4 (NovemLber 1980). 32 SOURCES OF FOOD INSECURITY Table 7. Foreign Exchange Cost of the Excess Food Import Bill, 1967 to 1977 (percentage of the supply of foreign exchange) Region and country 1967 1969 1971 Asia Bangladesh n.a. n.a. n.a. India 1.6 -7.5 2.4 Indonesia 28.7 8.7 11.3 Philippines 1.0 1.3 1.8 Sri Lanka 0.9 2.2 3.4 North Africa/Middle East Egypt -3.8 2.4 1.7 Jordan 6.7 14.5 10.7 Somalia 6.5 0.7 -5.1 Syria 0.7 2.4 7.7 Sub-Saharan Africa Mali 7.5 -407.2 5.7 Senegal n.a. -2.2 1.9 Tanzania 9.6 10.9 6.1 Upper Volta n.a. 7.5 7.4 Latin America Bolivia 1.7 3.9 2.8 Brazil -0.8 6.8 7.0 Chile -1.9 -0.2 -2.5 Haiti 1.3 4.3 3.6 Peru -0.1 0.5 2.0 Note: The supply of foreign exchange includes total export receipts plus private unrequited transfers plus net capital inflow. Food includes cereals, meats, dairy products, fruits and vegetables, sugar, pulses, root crops, oilseeds and vegetable tries, however, the increase in their food import bill coincided with an increase in their export earnings. In sixteen out of twenty-four countries during 1965-76, the absolute variability of the food import bill was smaller than the variability of export earnings. However, the opposite occurs in some large low-in- come countries (Bangladesh, India, and Sri Lanka). In general, food import bills were positively correlated with export reve- nues, reflecting positive correlations in world market price SOURCES OF FOOD INSECURITY 33 Maximum value 1973 1975 1977 (1967-7J7) 14.0 86.2 55.4 114.1 (1976) -5.9 17.0 -1.3 17.0 (1975) 8.0 13.7 11.8 28.7 (1967) 3.0 1.7 0.9 3.4 (1974) 11.2 23.7 6.1 23.7 (1975) 4.2 -12.0 -5.3 4.2 (1973) 15.2 10.5 5.7 17.0 (1974) -3.3 -1.5 -5.6 6.5 (1967) 10.7 5.3 5.1 12.7 (1968) -9.9 -19.4 13.6 24.9 (1968) 0.0 1.6 0.7 3.2 (1968) 10.2 20.8 22.8 22.8 (19'7) 5.1 7.0 13.5 14.5 (197'6) 4.5 4.2 3.7 4.9 (1974) 10.4 12.2 5.0 12.4 (1974) 0.0 -1.9 -1.4 0.7 (1972) 7.8 17.5 11.7 17.5 (1975) 2.3 2.3 0.6 2.3 (1975) oils, cocoa beans, and derivatives. n.a. Not available. Source: Data compiled by Grant Scobie and Alberto Valdes, IFPRI. fluctuations. There is not a one-to-one relation between the observed changes in the supply of foreign exchange and the actual imports of food. A reduction of X percent in export revenues will not necessarily lead to a reduction of X percenit in food imports. Similarly, an increase of X percent in the supply of foreign exchange from export revenues or foreign borrowing is unlikely to increase food imports by the same percentage. 34 SOURCES OF FOOD INSECURITY The observed instability in import volumes in developing countries is in itself some measure of success in domestic food security policies. Stabilizing trade is the right solution for most countries. But the indirect effects of import instability could be detrimental to productivity in the nonfood sector. The possibility of food imports crowding out the imports of essential raw ma- terials and capital goods has been noted by several researchers, including Lofchie for Tanzania, Behrman for Chile, and Pitt for Indonesia.5 Grant Scobie has recently made a rigorous quantitative analy- sis of import demand for Egypt during the past twenty-five years. He observed that the decline in the volume of imports accom- panying a fall in foreign exchange supplies has the greatest effect on industrial raw materials and capital goods, and he concludes that "the inelastic demand for food imports engendered by the food subsidy scheme is associated with fluctuations in output and investment in the industrial sector."6 Thus, in countries already committed to policies to stabilize food consumption, shocks in the current account could be ab- sorbed primarily by adjusting nonfood imports. In other coun- tries, however, the major adjustment could be in food imports. The adjustments will vary from one country to another, reflect- ing differences in food consumption policies, in the share of food imports in the current account, and in access to foreign borrowing. 5. M. F. Lofchie, "Agrarian Crisis and Economic Liberalization in Tanzania," Journal of Modern Afirican Studies, vol. 16, no. 3 (1978), pp. 451-75;J. R. Behrman, Foreign Trade Regimes and Economic Development: Chile, Foreign Trade Regimes and Economic Development series, vol. 8 (New York: National Bureau of Economic Research, 1976); and M. M. Pitt, "Alternative Trade Strategies and Employment: Indonesia," in A. 0. Krueger, H. B. Lary, T. Monson, and N. Akrasanee, eds., Trade and Employment in Developing Countnes, vol. 1, Individual Studies (Chicago: University of Chicago Press, 1981), pp. 181-237. 6. Grant M. Scobie, Food Subsidies: Their Impact on Foreign Exchange and Trade in Egypt, Research Report no. 40 (Washington, D.C.: International Food Policy Research Institute, August 1983). SOURCES OF FOOD INSECURITY 35 Adjustment to variability in food import bills also is affected by income. The greater ability of higher income countries to vary their level of imports of food lies to a large extent in their greater access to the international capital markets and in their foreign currency reserves. In contrast, low-income countries can- not borrow on short notice in international capital markets at average lending rates. Furthermore, the poorest developing countries, with low reserve ratios, have a limited ability to draw on reserve holdings. Thus, the brunt of the adjustment takes place in the allocation of foreign exchange among the various items, including food. In summary, for many developing countries, particularly the poorest, the extra cost of food impor-ts (net of changes in export revenues) has been significant in terms of losses in their real income and deficits in their trading accounts. Quantitative as- sessments of above-normal additions to the food import bill between 1965 and 1976, adjusting for changes in export reve- nues, indicate that this extra cost was particularly significant for low-income developing countries. If the countries had followed a policy to stabilize food consumption instead, the calculations indicate that the net increment in the import bill resulting from increased food imports (net of fluctuations in export revenues) would have been considerably higher. To restore the balance in their current accounts, many de- veloping countries have had to reduce the volume either of food imports or of other essential goods. Furthermore, the lack of responsiveness of food aid to widespread production shortfalls observed during the 1970s, which was aggravated in years of high world market prices such as in the early 1970s, clearly added to the burden of imports during years of sharp increases in the food import bill.7 7. Barbara Huddleston, "Responsiveness of Food Aid to Variable Import Requirements," in Alberto Valdes, ed., Food Securityfor Developing Countries (Boul- der, Col.: Westview Press, 1981), pp. 287-306. 3 Alternative Financing Facilities FLUCTUATIONS IN FOOD CONSUMPTION in developing countries are basically caused by variations in people's purchasing power and in the price of food. For the purpose of analyzing the effect of a food financing facility it is assumed that fluctuations in income are either of minor. significance or that countries inde- pendently pursue policies to stabilize income. In food importing countries food prices fluctuate either be- cause world prices fluctuate or because imports are more or less than what people are willing to consume at the prevailing price. In both cases, government policies and balance of payment con- straints play an important role. To insulate the domestic price from world market prices, governments must either subsidize or tax imports. Permitting all the imports that people are willing to consume at the world market price requires unrestricted al- locations of foreign exchange. At best, international financial assistance can alleviate the bal- ance of payment constraints a government faces if it wants to institute policies to insulate food prices and if it wants to allocate enough foreign exchange to food imports to stabilize consump- tion. But the extent to which food security is actually achieved also depends significantly on each government's propensity to pursue policies that insulate the domestic food price from fluc- tuations in the world market price, and to actually allocate scarce foreign exchange to food rather than other imports. 36 ALTERNATIVE FINANCING FACILITIES 37 The Simulation Model A simple normative model can be constructed, based on plau- sible determining factors, to estimrate a country's food con- sumption, including probability distributions of domestic food production, the price of imported food, the availability of foreign exchange, and the country's policies for insulating the domestic price from the world price and for allocating foreign exchange to food imports. A country's ability to maintain desired con- sumption is then simulated on the basis of 3,000 randomly drawn observations from these probability distributions, with and with- out a financial facility operating under various rules. The ex- periments were repeated with different sets of parameters reflecting different government policies, and the results were used to assess the effects of international assistance schemes. A detailed description of the model and of the data and parameters used in the analysis is provided in the Appendix. The model describes a set of outcomes in relation to circum- stances presumed to exist at present. The assessments of con- sumption stability can be extrapolated to future years to the extent that shifts in the demand and supply of food and the availability of foreign exchange occur at the same rate over time, to the extent that there is no perceptible upward or downward trend in the world price, and to the extent that the relative variance of the variables remains unchanged. Calls on a financial facility in this case would increase iin proportion to the growth of consumption. Alternative assumptions concerning trends in domestic or international prices would, of course, change the outcomes. The random variables affecting a country's aggregate food consumption in the model are domestic cereal production, world cereal availability, and the total doimestic supply of foreign ex- change. On the basis of a linear, kinked world demand function for cereals, a normal probability distribution of world cereal 38 ALTERNATIVE FINANCING FACILITIES production is transformed into a skewed probability distribution of the cereal import price. Similarly, on the assumption that the desired level of nonfood imports is constant, the probability distribution of the total supply of foreign exchange is trans- formed into a probability distribution of foreign exchange avail- able for food imports without having to reduce nonfood imports. The mean of this distribution is the food import bill, when actual domestic cereal production is equal to mean domestic cereal production and when the actual cereal import price is equal to the mean cereal import price. Consumption stabilization in the model is determined by the extent to which desired consumption and desired imports are responsive to price and by the extent to which actual consump- tion and actual imports are reduced from desired levels when foreign exchange is scarce. Desired imports are the positive difference between desired consumption and food production. Desired consumption, according to the formulation presented here, has no normative connotations. To the extent that desired consumption is highly price responsive, it could be far below a level consistent with food security. Food security is measured by the likelihood of consumption falling below some prespecified level. A scarcity of foreign exchange affecting food imports is as- sumed to prevail whenever the desired food import bill exceeds the amount of foreign exchange available for food imports with- out having to reduce nonfood imports. When this happens, for- eign exchange will be apportioned according to the propensity to restrict foreign exchange allocations for food imports (m), and the related propensity to restrict foreign exchange alloca- tions for nonfood imports (1 - m). The value of m can vary between zero and one. A country that accords high priority to food security will give first priority to food imports (m = 0), and actual food imports will always equal desired imports. At the other extreme (m = 1), the food import bill will not be allowed to exceed the supply of foreign exchange available for ALTERNATIVE FINANCING FACILITIES 39 food imports when nonfood imports are at the desired level, and food imports will be restricted to the maximum amount that can be purchased given the available supply of foreign exchange for food, the price of the food to be imported, and the relevant exchange rate. In the model, government policy can affect the stability of consumption either through internal price-stabilizing policies affecting the desired level of consumption, or through the pro- pensity to restrict the amount of foreign exchange made available to food imports (m) when not enough foreign exchange is avail- able to achieve the desired levels of both food and nonfood imports. The model indicates the extent of food security with and without price stabilization policies, and with high, medium, and low propensities to restrict food imports when foreign ex- change is scarce (m = 1.0, 0.5, and 0.2, respectively). These propensities, as well as the implementation of internal stabili- zation policies, reflect governments' priorities and commitments to food security, but they also reflect uncontrollable constraints. Low-income countries and countries that even in normal years must devote a large share of their export earnings to food im- ports may not have the required financial reserves or access to capital markets and, thus, the flexibility to reduce nonfood im- ports such as oil. The effect of different international financial assistance schemes operated in conjunction with a range of country policies is illustrated by reference to data reflecting instability and current levels of the world food price, production, imports, and foreign exchange availability in six countries: Bangladesh, India, Egypt, Senegal, Brazil, and Peru. The data and patrameters used are approximations and thus only illustrate the range of outcomes from instituting different schemes for international financial assistance. The selection of countries was to some extent arbitrary, but the countries included do represent a diversity of situations with respect to geographical region, population size, income level, 40 Al,TERNATIVE FINANCING FACILITIES degree of self-sufficiency, degree of financial constraint, and variation of domestic policy. India was included as an example of a more or less self-sufficient country, but is atypical in that it was previously a large importer and has created a distribution network through which food imports can be channeled if needed. Such networks do not exist in many African countries that are also relatively self-sufficient and need to import only when do- mestic production fails. See the Appendix for more detail on the methodology and data. Figure 3 presents the results of simulation experiments for selected countries. It illustrates the relative importance of pol- icies to stabilize prices and to allocate foreign exchange (FE) to food imports. "Restricted FE" and "unrestricted FE" refer to the propensity of the country to restrict the use of foreign exchange for food imports; "market price" means that the domestic price Figure 3. Probability of Consimption Falling belouw 95 Pefrcent of Trend without .4rn Finanicial FacilitY 6() 50- 40- 30- 20 Bangladesh India Egypt Senegal Brazil Peru .\ote. El restricted FE, market price: 0 restricted F', stabilized price: 2 unrestricted FE., miarket price: &5 unrestricted FE, stabilized price.- Restricted FF." and "unrestricted F.'' refer to the propensitv of' the country to restrict the use of' foreign exchange for food imports. "Market price" means that the do- tnestic price of food fluctuates withl the world market price, while 'stabilized price" refers to a policy whereby the domestic price of food is insulated from the international price. ALTERNATIVE FINANCING FACILITIES 41 of food fluctuates with the world market price, while "stabilized price" refers to a policy whereby the domestic price of food is insulated from the international price. If unstable import prices are transmitted to the domestic market, food insecurity remains high despite the priority given to food imports in allocating foreign exchange. If the propensity to restrict foreign exchange allocations for food imports (m) is high, consumption cannot be stabilized during a shortfall in production because the country will not allocate the additional foireign exchange required to import any food needed to cover the shortfall. A high level of food security is possible only through a combination of policies to stabilize domestic prices and a low propensity to restrict for- eign exchange allocations for food imports. Alternative International Financing Facilities The effect of an international financial arrangement to reduce or eliminate the foreign exchange constraint depends on the degree of instability in the domestic food supply, import prices, and the availability of foreign exchange; the rules governing drawings from the facility; the country's price stabilization policy; and its food import priorities when foreign exchange is scarce. Depending on the combination of circumstances, the effect on food security of a financial facility can range from almost nil to quite large. Scarcity of foreign exchange may be caused by a rise in the world price of food imports or a decline in domestic food pro- duction, so that the desired import bill exceeds the trend level of foreign exchange available for food imports without reducing nonfood imports. Or scarcity may be caused by a shortfall in export earnings or diversion of export earnings to nonfood im- ports, so that the actual amount of foreign exchange available for food imports is less than the trend amount. Three types of facilities are considered for dealing with food insecurity. The 42 ALTERNATIVE FINANCING FACILITIES first would allow drawings based solely on shortfalls in foreign exchange availability, the second would cover overruns in the food import bill, and the third would cover both. A Facility to Stabilize Available Foreign Exchange (SAFE) The SAFE facility would correspond closely to the CFF operated by the IMF before the 1981 decision to broaden coverage to include food imports. It would contribute to food security by raising the availability of foreign exchange when there is a short- fall in export earnings, so that nonfood and food imports could be maintained as long as the value of desired food imports does not exceed the trend value of imports. When the cost of desired imports exceeds the trend value, however, this facility would not assure financing of the full quantity of desired imports. The IMF facility compensated only for shortfalls in export earn- ings. Thus, because nonfood imports sometimes varied along with food imports and because drawings were constrained by an upper quota limit, the IMF facility provided less food security than would the simulated SAFE facility. A Facility to Stabilize the Food Import Bill (SFIB) The SFIB facility would permit countries to draw credits to cover the difference between the actual food import bill and the average food import bill when the former exceeds the latter. This corresponds closely with the proposal for a separate IMF facility to compensate countries for cost overruns in their cereal import bills. In the model the actual food import bill is estimated for given levels of desired food import bills and foreign exchange availability, including potential drawings from the facility. Coun- tries would in general draw credits when the desired food import bill exceeds the average food import bill, except when a reduced amount of available foreign exchange and a high propensity to restrict food imports leads them to reduce food imports below ALTERNATIVE FINANCING FACILITIES 43 the desired level in spite of the opportunity to draw credits from the facility. An Integrated Stabilization Facility (][NTEG) The INTEG facility would eliminate all foreign exchange con- straints on food imports, whether they are caused by a higher than normal food import bill or by a shortfall in available foreign exchange. Drawings would equal the excess of the desired food import bill over the available foreign exchange for food imports. Thus, this facility would cover boti the difference between the desired import bill and the available foreign exchange for food imports and the difference between mean and actual foreign exchange available for food imports. Like the new IMF facility that would integrate compensation for shortfalls in export earn- ings and for overruns in cereal import costs, INTEG would allow a surplus in export earnings to reduce the coverage for cereal import overruns, and a reduction in cereal import costs to reduce the coverage for shortfalls in export earnings. Thus, this facility would permit countries to fully cover desired food imports and desired consumption, provided that the bill for desired nonfood imports is constant or, if not, that the propensity to restrict foreign exchange expenditure for ifood imports is high. Unlike the new IMF facility, this scheme is not constrained by quota limits. Comparison of the Alternative Facilities Figures 4 and 5 illustate the effect on food security and the expected drawings from the three foreign exchange assistance schemes given two sets of country policies. Country policy A is characterized by low concern with food security: there is no insulation from world market or domestic production fluctua- tions, prices and consumption levels are unstable irrespective of 44 ALTERNATIVE FINANCING FACILITIES Figure 4. Probability of Consumption Falling below 95 Percent of Trend with Financial Facilities, for Country Policy, A 60 50 - 40- 30- 20 10 0 Bangladesh India Egypt Senegal Brazil Peru Aole a none; E SAFF:E SFIB;s INUF:(;. Countrv policv A is de- fined in Figure 6. foreign exchange constraints, and the propensity to restrict for- eign exchange allocations for food is high when the value of desired imports is greater than the amount of available foreign exchange. Under country policy B there is relatively high concern about food security: prices to consumers are stabilized, and about half of the shortfall is made up by reducing other imports or by drawing down foreign exchange reserves when the value of de- sired imports is greater than the amount of foreign exchange available for imports. The integrated scheme (INTEG), which by definition eliminates any shortfall of foreign exchange, has an impressive effect on food security in both cases. However, food insecurity remains high when the country does not also stabilize internal prices. There are, of course, many reasons why a government may not wish to stabilize aggregate consumption by stabilizing prices, or may find it difficult to do so. One reason may be that the gov- ALTERNATIVE FINANCING FACILITIIES 45 ernment would be unable to stand by this policy when foreign exchange constraints make it impossible to secure the needed volume of imports. In such cases a government may consider it politically unwise to build up expe,ctations it might not be able to fulfill. Thus a reliable arrangement for international foreign exchange assistance might encourage the adoption of a policy to stabilize prices. For instance, under an INTEG scheme, a coun- try like Peru might be encouraged to reduce the probability of consumption falling below 95 percent of ncrmal from 50 to 3 percent. Figures 4 and 5 also indicate that the SAFE scheme would have a large effect on food security for countries in which the foreign exchange available for food imports varies mnore than does the bill for desired food imports. In some cases the SAFE scheme would have a greater effect on security than the SFIB facility. Food security resulting from an internatiornal facility would be much higher if countries would aLlso try to stabilize internal Figure 5. Probability of Consumptiotn Falling belouw 95 Percent of Trend with Finiancial Facilities, for- (,ountry Policy B 40 30 20 10 Bangladesh India Egypt Senegal Brazil Peru Note: 2 none; O SAFE; SFIB; INIE(;. Country policv B is de- fined in Figure 6. 46 ALTERNATIVE FINANCING FACILITIES prices. This is illustrated in Figure 6, which is based on the average food insecurity calculated for the six countries com- bined. The graph shows that food security improves dramatically if countries receiving foreign exchange assistance also adopt policies to promote food security. Figure 6 also illustrates the impressive effectiveness of an export earnings facility such as SAFE. In fact, under an unstable price policy, the effect is larger than that of a separate food import bill facility such as SFIB. But, as might be expected, an integrated approach is the most effective way to reduce unstable food consumption. A change in domestic policy increases sta- bility most dramatically when combined with any of the three facilities, as shown by the last set of bars in the graph. The expected annual drawings by country from the alternative international schemes are shown in Figure 7, while total expected Figure 6. Stabilizing Effect of Financial Facilities on Average Food Consumption 100 80 60- 40- - btL 20 0 Policy A: Policy B: Country moves unstable prices stabilized prices from policv A and high propensity and medium propensity to policv B to restrict use of FE to restrict use of FE for food imports for food imports \ote: [ SAFE; U SFIB; 0 INIEG. Averages are calculated for six coun- tries: Bangladesh, India, Egypt, Senegal. Brazil, and Peru. ALTERNATIVE FINANCING FACILITIES 47 Figure 7. Expected Annual Drawings of Six Countnies from International Financial Facilities - 500 -e / 400 300 2 00 0 D a Bangladesh India Egypt Senegal Brazil Peru Aote: SAFE; policy A. SFIB: E policv A, IN'I'( policy B, SFIB; E policy B, INTIG;. Policies A and B are defined in Figure 6. annual drawings are shown in Figure 8. Drawings from SAFE are unrelated to the country's import policies to promote food se- curity, but expected drawings from INTEG also are not very sen- sitive to the country's policy. This is in contrast to the large differences in the relative reduction in food insecurity provided by SFIB under different country policies. The sum of the estimated drawings from the SAFE and SFIB schemes would underestimate the expected total drawings if the two schemes were operated separately. The reason for this is that expected drawings from the SFIB scheme were calculated on the assumption that there was no SAFE facility. If both schemes were in effect, however, the combined expected drawings needed to achieve a given level of food security would be much higher than those for an integrated scheme. For inslLance, assuming that the countries are pursuing policies to increase food security, the expected drawings from the two separate schemes for India and 48 ALTERNATIVE FINANCING FACILITIES Figure 8. Total Expected Annual Drawings from International Financial Facilities N 1,200 0 1,000 = 800 600 400 200 05 SAFE SFIB IN'I'EG SFIB IN 1F; Policv A Policy B Brazil would exceed US$551 million and US$405 million, re- spectively, as compared with US$402 million and US$300 mil- lion, respectively, from the integrated scheme. Figures 9, 10, and 1 1 illustrate why new or expanded foreign exchange assistance schemes might interest major food export- ing countries. Under an INTEG scheme, average annual imports by the six countries would increase by from 2.4 million to 5.1 million tons above expected levels without a facility, depending on the domestic price policies. Although these additional imports would not be large in relation to total world supplies or even to total exports, they would represent substantial increases over expected import demand without a facility. Barring a foreign exchange constraint, policy B implies mnore iinlports than policy A when the world price is high, and less ALTERNATIVE FINANCING FACILITIES 49 Figure 9. Expected Imports under Various International Financial Facilities, for Country Policy A 4,000 300 200 1 3000 0 A Bangladesh India Egypt Senegal Brazil Peru Note: F- none; El SAFE; 0 SFIB; P INIrE(,. Policv A is defined in Figure 6. Figure 10. Expected Imports under l'arious International Financi'al Faci'lities, for C"ountry Policy B 400 0~~~~~~~~~ M. 2.000 0~~~~~~~~~~~~~~ -e1 000 0 Bangladesh India Egypt Senegal Brazil Peru .vote:. El none; a SAFk E SFIB; O IB E. Policy B is defined in Figure 6. 50 ALTERNATIVE FINANCING FACILITIES Figure 11. Total Expected Imports and Additional Imports under IVanious International Financial Faczlities, for Country Policies A and B - 12,000 10,000 _ 8,000 _ -E 6,000 - o 4,000 - 0. _2,000_ /1m0 Expected Additional expected Expected Additional expected imports imports under facility imports imports under facilitv Policy A Policv B \ote: E policv A, SAF'E; * policv A, SFIB, ; policy A, IN' It.; O policv B, SAFE; 0 policy B, SFB; E policy B, INE>. Policies A and B are defined in Figure 6. imports when the world price is low. In the model, expected overall imports are higher under policy B because the probability of above average prices is higher than the probability of below average prices. In the model that specifies linear demand func- tions, foreign exchange constraints arising from low export earn- ings or poor harvests are more likely to keep imports below desired levels under policy A than under policy B. Hence, the SAFE and INTEG schemes, which exclusively or predominantly relax foreign exchange constraints arising from below normal export earnings or poor harvests, provide for more additional imports under policy A than under policy B. Contrarily, the SFIB scheme, which relaxes foreign exchange constraints partly caused by high food import prices, provides for more additional imports under policy B than under policy A. Clearly, the value of additional imports resulting from a fi- ALTERNATIVE FINANCING FACILITIES 51 nancial facility would be less than the drawings from the financial facility, as countries would have imaported part of the deficit anyway. This would be particularly true for countries with a low propensity to restrict the use of foreign exchange for food im- ports. For such countries, the facility could nevertheless provide important balance of payments relief and prevent economically harmful reductions in nonfood imports occasioned by increases in the food import bill. For a particular country, the model can predict the likelihood of various intensities of drawings from a financial facility and the resulting increases in cereal imports. Generally, there is a 50 percent or higher chance that the country will not use the facility at all. Figures 12 and 13 show the 10 and 5 percent probabilities, respectively, that drawings will exceed stated Figure 12. Drawings from International Financial Facilities in Excess of Stated Amounts at the 10 Percent Level of Probabili't 2,000 r 1,500 1,000 a E 500 0 _ Bangladesh India Egvpt Senegal Brazil Peru Aote: E SAFE 0 SFIB; 0 INIF FG. 52 ALTERNATIVE FINANCING FACILITIES Figure 13. Drawings from International Financial Facilities in Excess of Stated Amounts at the 5 Percent Level of Probability 2,500 0 Bangladesh India Egypt Senegal Brazil Peru .Vote 20 SAFE0 E SFIB0 INTEG, amounts. These probabilities are based on the assumption that countries pursue a price stabilization policy and give medium priority to allocating scarce foreign exchange for food imports. Tlhe probability that the combined drawings of the six countries wi'II exceed the sum of the drawings at the stated levels will be much lower. How much lower depends on how much production shortfalls in the participating countries are correlated and on whiether drawings are the consequence of production shortfalls or of a higher than normal world price. Interaction of Buffer Stocks and Finance Two other measures frequently discussed in connection with food security are stabilization of domestic and global food sup- ALTERNATIVE FINANCING FACILITIES 53 plies. For this discussion, a 50 percent reduction of the standard deviations of domestic and global cereal supplies is postulated. The particular buffer stock arrangements and the size of the stocks required to achieve these levels of stabilization were not considered. Reduction in domestic instability could be achieved through domestically held buffer stocks, while higher global sta- bility could be obtained through some combination of stocks and trade liberalization policies in the major exporting and im- porting countries. A reduction in the variability of global supplies of such size would imply a reduction in the mean import price Figure 14. Indian Food Consumption under Alternate International Financial Facilities 30 bc 25 15 0 More More More More stable stable stable stable Base domestic import Base domestic import case supply price case supply price _9 _- Policv A Policy B Note: E none; B policv A, SAFF; 0 policv A, SFIB; 0 policv A, IN I,E(;; 2 policy B, SAFE; E policy B, SFIB; O policy B. INTEG. 54 ALTERNATIVE FINANCING FACILITIES Figure 15. Brazilian Food Consumption under .41ternate Internatlonal Financial Facilities 50 -S --0 4 30 20 10 0 More More More More stable stable stable stable Base domestic import Base domestic import case supply price case suppiv price Policv A Policv B Notle: E none; E policv A, SAFE; El policv A, SFIB;: policy A, IN I F(G 0 policv B, SAF'F: policv B, SFIB; g polici B. IN I EG. from US$225 to US$215 per metric ton, and its standard de- viation from US$50 to US$25 per metric ton. The effect of such stabilization measures on food security in India and Brazil is illustrated in Figures 14 and 15, respectively. The more stable domestic supply and the more stable import price scenarios correspond with a reduction in the respective standard deviations of 50 percent. As noted earlier, under policy A instability in world prices is transmitted to the domestic mar- ket, and there is a high propensity to restrict foreign exchange for food imports. Under country policy B prices are stabilized, and there is a medium propensity to restrict foreign exchange. ALTERNATIVE FINANCING FACILITIES 55 In general, even a much more stable domestic food supply or import price would far from eliminate food insecurity if foreign exchange is constraining food imports. But financial schemes to relieve foreign exchange problems would greatly improve food security. More important, Figures 14 and 15 indicate that financial arrangements to reduce foreign exchange constraints are an effective substitute for the much more costly buffer stock ar- rangements. For example, for India under country policy A, the base level of consumption instability is reduced from 28 to 19 percent under an INTEG. Without a facility, it would take a large amount of buffer stocks to obtain the same improvement in food security. For Brazil, the base level of consumption instability is reduced from 41 to 18 percent under an INTEG. Without a facility such an improvement is not achievable with even a 50 percent reduction in the instability of supply. The same pattern appears under country policy B, although the stabilized price policy and lower propensity to restrict food imports generally result in less consumption instability. The differences between the two coun- tries may arise in part because foreign exchange earnings com- prise a much larger share of total GNP for Bra[zil, while India is more nearly self-sufficient in food production. 4 An IMF Facility to Finance Excess Food Import Bills IN RESPONSE TO AJOINT PROPOSAL of the World Food Council and the Food and Agriculture Organization of the United Na- tions, the IMF began to consider proposals for an international food financing facility in early 1980. After considerable discus- sion, and with the support of several national and international bodies, the Board of Directors decided in May 1981 to provide such assistance by broadening the terms of reference of the existing CFF for export earnings. In reaching its decision, the IMF board considered two alternatives: one that would fully in- tegrate the food financing facility into the existing CFF and an- other that would create a separate food financing facility and link it to the existing facility by a joint quota limit. The Alternative Proposals Under the first alternative, cereal imports would be treated as negative exports, and countries would be compensated only when there was a net export shortfall. The procedure would resemble the one already used to compensate shortfalls in export earnings, except that excess cereal imports would be added to shortfalls in export earnings to obtain the net amount of the compensable shortfall. Thus if cereal imports were excessively 56 AN IMF FACILITY TO FINANCE EXCESS FOOD IMPORT BILLS 57 large and merchandise exports fell below normal, the require- ment for extra cereal imports would be treatecl as a shortfall in net export earnings, and the country would qualify for additional compensation from the CFF. If, however, merchandise exports fell below normal but cereal import requireiments were also small, net export earnings might not be short, and the country would not qualify for compensation. This alternative would pro- vide additional balance of payments support linked specifically to cereal import costs, but only at the expense of some support now provided to offset shortfalls in export earnings. At present, an allotment of special drawing rights (SDRS) iS allocated to each member country unconditioinally and is used to purchase other members' currencies and in certain transac- tions with the IMF itself. In addition each country is assigned an SDR quota, which it purchases with its own currency and against which it can borrow according to the rules and regulations of each of the IMF's different accounts and facilities. Normally, bor- rowings against quota must be repaid within three to five years at moderate interest rates, although some funds offer easier terms to low-income countries with serious payment imbalances. Each fund or facility within the IMF places an upper limit on the amount that can be borrowed. This limit is expressed as a per- centage of the country's quota. For the export: earnings facility the limit has been 100 percent of quota. According to the proposal to integrate a cereal component into the existing CFF, the export earnings comnponent and the cereal component would each be subject to a separate quota limit, and the net amount would be subject to a limit lower than the sum of the two. Both the quota limits and the "netting out" procedure that determines the compensable amount would re- strict how much countries could draw. The second alternative would retain the CFF unchanged and would create a separate food financing facility, which would operate under a joint quota limit applicable to, combined draw- ings from both. Countries could draw from the CFF as they do 58 AN IMF FACILITY TO FINANCE EXCESS FOOD IMPORT BILLS now, subject to the existing quota limit of 100 percent. They could also draw from the new food financing facility, which would compensate them for the amount that actual cereal import costs exceed the trend for the year of the drawing. These drawings would also be subject to a limit of from 50 to 100 percent of quota. If countries were eligible for drawings from both facilities in the same year, the combined total of these drawings would be limited to from 125 to 150 percent of quota. This alternative would allow countries to treat separately their requirements for compensation arising from export earnings shortfalls and those Table 8. Hypothetical Withdrawals from the Food Financing Facility under the Two Alternative Proposals (millions of SDRS) Item Basic data Export shortfall 29 29 29 0 -11 -11 Cereal import excess 14 -16 0 14 14 -16 Calculation of export component: lesser of export shortfall or 100 percent of quota' 20 20 20 0 0 0 Calculation of cereal component: lesser of cereal import excess or 100 percent of quota' 14 0 0 14 14 0 Integrated approach Sum of export and cereal components (A) 34 20 20 14 14 0 or Net export shortfall (B) 43 13 29 14 3 0 Drawing equals the lesser of (A) or (B) with a quota limit of 125 percent' 25 13 20 14 3 0 Separate approach Sum of export and cereal component (A) with a quota limit of 125 percent' 25 20 20 14 14 0 a. Where 100 percent of quota equals 20 million SDRs. Source: Constructed by authors from hypothetical data. AN IMF FACILITY TO FINANCE EXCESS FOOD IMPORT BILLS 59 from excess cereal import costs. Thle joint quota limit would prevent a country from making extrernely high withdrawals from both facilities in the same year. The difference between the two approaches is shown in the hypothetical example in Table 8. Estimated Aggregate Drawings If there were no quota limits, aggregate drawings would differ considerably under the two alternatives considered by the IMF. Without constraints, the second alternative, which allows for full compensation for both export shortfalls and cereal import ex- cesses, results in a significantly higher aggregate level of draw- ings because the drawings from the food facility are in addition to drawings from the CFF. Under the first alternative, countries that have a high correlation between the value of merchandise exports and the cereal import bill would not receive the full additional benefit of the new facility because their cereal draw- ings would be constrained by their export excesses, and their export earnings drawings would be constrained by low cereal import bills. A separate food facility would allow these countries to receive the full amount for which they are eligible from both facilities. When quota limits are introduced, aggregate drawings under both alternatives are sharply reducecl, but the second still results in a somewhat higher total than the first. The quota limit strongly inhibits drawings from the existing CFF for export earnings but does not have much effect on drawings to fully compensate countries for excess cereal import costs. Thus, with quota limits, the main difference between the two alternatives is that there is less than full compensation for cereal import excesses because of the offsetting effect of export surpluses under the first alter- native, as opposed to full compensation under the second one. As a rough approximation, additional annual average drawings for cereal imports in a noncrisis year might amount to 100 million 60 AN IMF FACILITY TO FINANCE EXCESS FOOD IMPORT BILLS SDRS under a fully integrated scheme and 390 million under a separate scheme with joint quota limits. In a crisis year these amounts could rise to 590 million and 730 million, respectively, for the two alternatives. Limitations and Benefits of a Food Financing Facility Under either approach an IMF food financing facility would provide balance of payments support only when the costs of cereal imports rise sharply. Concern for the consumption short- falls of developing countries in periods when domestic produc- tion was short and world prices were high led to interest in creating such a facility. But maintaining minimum levels of per capita consumption for all segments of a country's population is beyond the purview of an international financial institution. Many countries cannot fully implement domestic policies to maintain minimum consumption for everyone until they achieve full employment and minimum income standards for all wage earners. These countries will require consumption subsidies or targeted food distribution programs to counter chronic mal- nutrition for vulnerable groups for the foreseeable future. Such food subsidy programs are better supported by grants of food aid and other highly concessional resource transfers than by a balance of payment facility. These transfers enable countries to finance immediate consumption needs of low-income groups without diverting resources from productive investment. Despite these limitations, balance of payments support for excess cereal import costs can serve a number of useful purposes. A financial facility can relieve many developing countries of foreign exchange constraints, such as those that occurred in the mid-1970s when price instability increased. The facility could enable them to bid competitively for the food imports needed to meet effective demand in years of scarcity. AN IMF FACILITY TO FINANCE EXCESS FOOD IMPORT BILLS 61 For countries that do not now attempt to stabilize domestic cereal prices, a food financing faciliity could provide a cushion that would enable them to introduce such policies. For countries that currently try to stabilize prices for urban consumers but allow prices to fluctuate sharply in rural areas, financial support for extra imports in years of domestic production shortfall could provide an incentive to develop rural procurement and market- ing programs that would channel additional imports to rural consumers. Finally, for countries that try to stabilize prices and consump- tion by allowing nonfood imports to drop when cereal import costs increase, a food financing faciility coulcl help maintain a smooth flow of investment into infrastructure development and productive enterprise. Although the effect of thte facility on levels of cereal imports may not be very great in such cases, the overall economic benefit may be considerable. It wou]ld allow the coun- try to achieve the longer run economic goals that make adequate consumption levels possible. It seems likely that low-income countries generally will gain relatively more from a food financing facility because fluctuations in the costs of cereal imports are a more important balance of payments problem for them than for middle-iincome countries. In addition, low-income countries cannot readily borrow on short notice in international capital markets at average lending rates, whereas most middle-income countries can and do borrow extensively in the private capital market. The benefits would be greater for low-income countries that have a policy to stabilize food consumption and a low marginal propensity to restrict foreign exchange expenditures for food imports. Even though some low-income countries could not make fuill use of the pro- posed facility in the immediate future, it would establish a cush- ion against future needs. One consideration that could be increasingly important during the next decade is the role of noncereal items in the total food 62 AN IMF FACILITY TO FINANCE EXCESS FOOD IMPORT BILLS Table 9. Composition of the Gross Food Import Bill for Selected Countries, 1961-65, 1968-72, and 1976-78 (percent) Region and Other Von- country Period Wheat Rice cereals cereals Asia Bangladesh 1961-65 33 30 0 37 1968-72 42 28 0 30 1976-78 41 22 0 36 India 1961-65 68 22 2 9 1968-72 53 24 3 19 1976-78 39 5 2 54 Indonesia 1961-65 7 88 1 5 1968-72 19 67 3 11 1976-78 8 66 1 25 Philippines 1961-65 28 32 0 40 1968-72 34 11 2 52 1976-78 47 2 8 43 Sri Lanka 1961-65 13 38 0 48 1968-72 29 29 0 42 1976-78 51 32 0 17 North Africa/Middle East Egypt 1961-65 64 0 8 28 1968-72 59 0 2 39 1976-78 50 0 7 43 Jordan 1961-65 29 11 2 57 1968-72 21 6 3 70 1976-78 20 5 6 69 Somalia 1961-65 11 37 6 46 1968-72 20 35 8 36 1976-78 13 21 14 53 Syria 1961-65 18 11 1 70 1968-72 37 8 0 55 1976-78 25 9 2 64 AN IMF FACILITY TO FINANCE EXCESS FOOD IMPOIRT BILLS 63 Region and Other Non- country Period Wheat Rice cereals cereals Sub-Saharan Africa Mali 1961-65 ID 0 2 87 1968-72 11 21 10 58 1976-78 16 14 9 60 Senegal 1961-65 12 33 5 50 1968-72 15 34 6 45 1976-78 17 38 8 37 Tanzania 1961-65 19 11 13 56 1968-72 11 7 14 68 1976-78 15 21 14 50 Upper Volta 1961-65 19 13 1 67 1968-72 36 3 0 61 1976-78 26 17 10 47 Latin America Bolivia 1961-65 5,7 2 0 41 1968-72 56 0 0 44 1976-78 63 0 0 37 Brazil 1961-65 83 0 1 16 1968-72 64 0 2 34 1976-78 56 2 10 33 Chile 1961-65 29 3 1 67 1968-72 2 4 8 66 1976-78 (iO 2 5 33 Haiti 1961-65 43 0 0 57 1968-72 3 1 0 0 68 1976-78 32 13 2 53 Peru 1961-65 49 7 1 43 1968-72 47 3 1 48 1976-78 47 5 13 35 Note: Food is broadly defined to include meats, dairy products, cereals, fruits and vegetables, sugar, pulses, root crops, oil seeds and vegetable oil, cocoa beans, and derivatives. Source: Data compiled by Grant Scobie and Alberto Vald6s, IFPRI. 64 AN IMF FACILITY TO FINANCE EXCESS FOOD IMPORT BILLS import bill. For many countries noncereal foods, particularly vegetable oils and sugar, comprise a significant proportion of the diet of poor people and of the total food import bill. When minimum requirements for cereal protein have been satisfied, these two items are among the first to be added to a diet to provide extra calories and nutritional balance. For many low- income countries, these items are important dietary necessities for the poor, not just luxury goods for the middle classes. On average, noncereal items accounted for more than 40 per- cent of total food imports for ten of eighteen developing coun- tries between 1976 and 1978 (Table 9). Therefore, it seems worthwhile for the IMF to consider a broader definition of food. One possible approach would be to define the basic food basket of the poor in each country and to determine the food import bill in relation to the items required to meet that minimum demand. Such a basket would differ from country to country. Justification for the Alternative Selected In examining the two alternatives considered by the IMF, a good argument can be made in favor of the first, in which excess cereal imports are treated as negative exports. Although the second provides a larger absolute level of benefits with the same quota limits, it is not clear that countries really need help with their balance of payments when the excess cost of cereal imports is offset by an increase in the value of merchandise exports. From the standpoint of food security, the food financing facility can make its most important contribution in years of extreme hardship when cereal prices are high, export revenues are low, and import requirements are above normal. In normal years the first approach gives considerably less compensation than the second, but in crisis years, such as 1974 and 1975, both provide comparable levels of assistance. Thus, AN IMF FACILITY TO FINANCE EXCESS FOOD IMPORT BILLS 65 the integrated facility would seem to be a better use of IMF resources and would provide good coverage with relatively lib- eral quota limits. With quota limits of 100 percent for both the export and the cereal components of an integrated scheme and a combined limit of 125 percent, credit availability to low-income countries could increase by up to 15 percent above outstanding IMF drawings for this group in mid-1979. This takes into account the repayment provisions that woulcl call for repurchase of draw- ing rights in equal installments in the fourth and fifth years following the drawing. This alternative was adopted by the IMF Board of Directors in May 1981. IMF members can choose whether to participate in the new food financing facility. If a country believes it will benefit more from continuing to participate in the CFF under the old rules, it may choose to do so. Once an option has been selected, however, it may not be changed for three years. Payments for a member's excess cereal imports are to be based on data for the latest twelve-month period, but if tirnely data cannot be obtained during a food crisis, cereal imports may be estimated for up to twelve months ahead and provision made for rapid repayment of excess drawings, if any. As with the export earnings facility, a member must demonstrate that it has a balance of payments need, that the excess in cereal imports results from circumstances largely beyond its control, and that it will cooperate with the IMF to seek appropriate solutions for its balance of payments difficulties. The facility will operate for an initial period of four years with an evaluation after two years. The facility adopted is compatible with other forms of food security assistance, particularly food aid. Because food aid is relatively more available in normal years and is considerably more concessional than the food financing facility, it would pre- sumably be the preferred form of assistance. In a crisis situation, however, food aid supplies are likely to decrease as prices climb, and countries' needs for alternative balance of payments support would increase proportionately. 66 AN IMF FACILITY TO FINANCE EXCESS FOOD IMPORT BILLS The question of accounting for food aid flows in establishing the amount to be compensated by the food financing facility will be a potentially troublesome issue. It need not be difficult, how- ever, if the cost of food aid is counted as zero in the year of the drawing, in the same way that the cost of the IMF drawing itself entails no foreign exchange expenditure in the year it is obtained. Thus, excess cereal import costs would be calculated against the actual foreign-exchange expenditure for commercial cereal im- ports plus the transport costs for food aid. On the one hand, if food aid in that year dropped sharply and the cost of commercial cereal imports rose steeply as a consequence, that increase could be fully compensated by the food financing facility. On the other hand, if food aid rose significantly in response to an unusual need, the country could not receive help from this plan for the amount covered by food aid. In the longer term, adjustments to variations in food aid flows and food import requirements would have to be worked out through normal adjustments in the country's balance of payments. Determination of Need The overall objectives of the facility are very clear. It will provide assistance only when import bills are above normal and are not offset by favorable export earnings; they must be tem- porary and largely attributable to circumstances beyond the con- trol of the government requesting assistance. However, dis- tinguishing what is a situation attributable to circumstances be- yond government control from one that is the result of domestic policies is sometimes a complex task. The method for calculating drawings-based on deviations from a five-year moving average-implies a built-in adjustment mechanism to reflect changes in the trend level of imports. If the excess food import bill lasts for a single year, aside from quota limits on drawings, the effect of the excess cereal import AN IMF FACILITY TO FINANCE EXCESS FOOD IMPORT BILLS 67 bill on the net import capacity (net of export earnings) can be largely offset by the drawings from the facility. If the increase in the food import bill were to last for several years, the "excess" as defined to measure import requirements would progressively disappear. Under such circumstances, the country might apply for financial assistance from the IM4F for balance of payments difficulties, usually under requirements of cooperation with the IMF in implementing policy adjustments.' Thlus, over the years, governments requesting financial assistance from the facility to cover additional imports to offset production shortfalls resulting from inappropriate domestic policies would not benefit relatively more than others. During the first and perhaps the second year of a production shortfall, however, to the extent that it is difficult for the IMF to determine in practice whether the shortfall is attributable to circumstances beyond the government's control, there is a possibility of some abuse of the scheme, but it would be short-lived. A different case arises when additional imports are the result of increases in the levels of domestic stocks rather than a pro- duction shortfall. Holding stocks is in itself an expensive op- eration, and it is hard to find cases of"overinvestment" in buffer stocks of cereals in developing countries with food shortages. For countries that have a buffer stocks policy, it is a positive feature of the facility that it allows for more flexibility to expand imports in years when import prices are low. But it is not an- ticipated that the creation of the facility will induce larger levels of stocks in countries with food shortages.2 l. A similar situation has been well illustrated for the facility on export revenues in Louis Goreux, Compensatory Financing Facility, Pamphlet Series no. 34 (Wash- ington, D.C.: International Monetary Fund, 1980). 2. For an empirical analysis showing the relation between domestically held stocks and financial assistance, seeJohn McIntire, Food Security in the Sahel: Variable Import Levy, Grain Reserves, and Foreign Exchange Assistance, Research Report no. 26 (Washington, D.C.: International Food Policy Research Institute, 1981). 68 AN IMF FACILITY TO FINANCE EXCESS FOOD IMPORT BILLS A question has arisen as to whether the IMF should apply "conditionality" to the use of the facility: that is, should it impose conditions regarding domestic distribution and consumption sta- bilization policies which governments must follow. Imposing conditionality in the form of policies to stabilize domestic con- sumption raises difficult practical and normative issues. Is it practical? How could the IMF expect to know what should and could realistically be done in each case without hiring an army of experts to continually examine each country's program of action? It would be very expensive, the political environment in many developing countries would make it difficult to respond positively to the conditions, and it would have dubious results, even if implemented. The dialogue about correct policies is better carried on in a larger context of relations with the World Bank, the Food and Agriculture Organization, and bilateral aid donors. The facility can, however, help to implement such pol- icies. Years of food shortages are not the appropriate time to delay assistance while domestic policies are investigated. The test of cooperation for borrowing under the facility should not go beyond the determination of need in the sense of an actual or expected increase in the food import bill from causes beyond the government's control. Effect on the Use of Forward Contracts There is no explicit restriction in the IMF rules that prevents the use of futures markets. In fact, the facility gives more financial flexibility to take advantage of the cheapest way to buy. There are, however, two considerations, which in some cases could limit the use of futures markets. First, countries expecting to receive food aid-primarily low-income countries-might be re- luctant to buy contracts in the futures markets if these operations affect the donor's allocations of food aid to them. But it is the uncertainty about the food aid allocations rather than the rules AN IMF FACILITY TO FINANCE EXCESS FOOD IMPORT BILLS 69 of the financial facility which may actually prevent buying in futures markets. Second, to determine the compensation for commercial im- ports under the facility, the IMF uses customs data and thus would not count imports until the grain actually arrives in the country. The IMF could, however, give a loan to cover the food import deficit up to six months into the future, with the understanding that the account would be settled between the importing country and the IMF when the grain is delivered. Thus, funds from the facility could be used to finance forward contracts. For some countries, forward contracts offer an attractive alternative to acquiring and storing reserve stocks, which access to a facility may make easier to implement. However, the circumstances of other countries may dictate other choices.3 3. For elaboration of this point, see Richard Gilmore ancl Barbara Huddleston, "The Food Security Challenge," Food PolicY. vol. 8, no. 1 (February 1983), pp. 33-46. 5 Effects of the IMF Facility on International Market Prices THE DEBATE ON CREATION of the food financing facility raised several questions about its possible effect on world grain prices. Some think that the financial facility could create a significant increase in demand by developing countries for grain in some years and thus could put strong pressure on prices to rise. But if the facility covers only excess foreign exchange costs of cereal imports without affecting domestic consumption, the increase in demand generated by the facility would not be very great. In its early deliberations, the IMF made it clear that it did not have an institutional mandate to subsidize programs to increase consumption or to improve nutrition in developing countries. All it could appropriately offer would be financing to countries whose foreign exchange positions were strained by the extra pressure of high cereal import bills in certain years when normal consumption could not be maintained without additional cost. Nor do the IMF calculations based on the historical period pro- vide for any increase in demand above the amounts of grain actually imported during the period. Thus in 1966, 1967, and 1968, and again in 1974 and 1975, years in which the real value of cereal imports for developing countries as a whole exceeded the five-year moving average, a food facility would have been liable for that excess value, but would not have created additional demand beyond what was actually satisfied in those years. Using 70 EFFECTS OF THE FACILITY ON INTERNATIONAL MARKET PRICES 71 Table 10. Actual and Trend Volumes and Values of Cereal Imports in Developing Countries, 1965 to 1976 Value Wheat (billions of 1977) Volume price (in U.S. dollars) (millions of real terms, metric tons) 1977 Trend Actual U. S. Year (deflated) (deflated) Trend Actual dollars) 1965 3.16 2.98 37.3 36.0 144.5 1966 3.33 3.41 38.8 41.4 153.3 1967 3.31 3.55 38.5 40.3 149.3 1968 3.40 3.57 39.5 40.2 151.3 1969 3.43 3.03 39.6 34.6 144.1 1970 3.44 3.44 40.2 41.0 131.6 1971 4.13 3.54 43.2 42.2 132.4 1972 6.06 3.62 413.3 42.9 133.1 1973 7.80 7.02 52.0 55.1 218.9 1974 9.19 12.69 55.0 60.5 227.6 1975 10.54 12.11 59.0 59.1 153.3 1976 11.73 10.53 62.9 57.2 134.5 Source: FAO trade tapes and IFPRI calculations. data from the Food and Agriculture Organization on imports for 101 developing countries, the International Food Policy Re- search Institute has calculated that the sum of the aggregate excess cereal import bill for 1965 to 1976 would be US$5.5 billion in 1977 dollars (see Table 10). The INIF staff arrived at a figure of about 1 billion SDRs for an integrated scheme with quota limits and from 2 to 4 billion SDRS for a separate scheme with quota limits. Likelihood of Increased Demand Though empirical results still are preliminary, work done by Philip Abbott and Grant Scobie indicates that in most developing countries demand for cereal imports is quite inelastic with re- 72 EFFECTS OF THE FACILITY ON INTERNATIONAL MARKET PRICES spect to world price.' Certainly this seems to be true for countries that primarily import wheat for urban consumption. As pointed out above, relieving the strain of high cereal import costs on their overall balance of payments position is a more likely effect than increasing planned levels of cereal imports. What cannot be predicted, however, is how much countries will modify domestic consumption policies and import practices, knowing that the IMF facility will provide a safety net against unexpected cost increases. Some countries may decide, for ex- ample, that they can afford to institute programs to subsidize consumption for a broader spectrum of low-income groups. With assured financing they can sustain such programs with imports in the event of domestic production shortfalls. Others may decide to take advantage of the facility to increase their cereal imports in years of relatively low prices, using the financial assistance provided by the IMF to create domestic food reserves on which they could draw when world market prices are high. Either of these policies could result in increases in demand for cereal imports by developing countries beyond what could be predicted from past trends. In the simulation reported in Chapter 3, the six countries in Table 12 (in the Appendix) account for more than 40 percent of the population of all developing market economies. Assuming that these six countries would also account for 40 percent of additional import demand created by the facility and taking the integrated scheme as the one most comparable with the IMF facility actually adopted, expected additional demand for all de- veloping countries in a given year would be from 6 million to 13 million tons. The assumption that demand increases will be highly correlated with population is perhaps too strong, how- 1. Philip C. Abbott, "Modeling International Grain Trade with Government Controlled Markets," American journal of Agricultural Economics, vol. 61 (February 1979), pp. 22-31; Grant M Scobie, Government Policy and Food Imports: The Case of W"heat in Egypt, Research Report no. 29 (Washington, D.C.: International Food Policy Research Institute, 1981). EFFECTS OF THE FACILITY ON INTERNATIONAL MARKET PRICES 73 ever, because unmet consumption requirements vary from one country to another. Another calculation estimates the amount of additional de- mand that would be created if all developing-country importers used the facility to help pursue policies to stabilize consumption. If countries had imported the additional quantities required to maintain per capita consumption at 1961-65 levels through 1978, the aggregate amount would have ranged from a low of 5 million tons in 1969 to a high of 19.2 million tons in 1973 (see Table 11). As pointed out, quota limits do not place a significant constraint on drawings for cereal irnports, although they do affect the export componenit of an integrated facility. Thus these numbers could represent realistic limits for the ad- ditional demand that the facility could generate under both nor- mal and abnormal world market conditions. A first step in understanding the possible effect of such in- creases in demand is to compare them with the level of world grain imports without the facility. As Table 1 [indicates, addi- tional imports by developing countries would have exceeded 10 percent of total world imports in only four out of eighteen years from 1961 through 1978 had the facility been operating then. Thus, the hypothesis that the facility would have a greater effect on price in some years than in others seems borne out by the variability observed in the hypothetical demand requirements for the recent past. Imports might not increase in the year when additional import requirements are greatest from the standpoint of stabilizing con- sumption. Countries with variable import requirements and some domestic storage capacity could take advantage of the facility by importing above-trend amounts in years when prices were low and storing the extra grain for use in years when greater domestic need coincided with higher world prices. Whether us- ing the facility for forward buying is more efficient than waiting until a combination of unmet internal demand and high world prices force reliance on IMF credit depends on whether the ex- 74 EFFECTS OF THE FACILITY ON INTERNATIONAL MARKET PRICES Table I1. Actual Volumes of World and Developing-Country Cereal Imports and Unmet Developing-Country Requirements, 1961 to 1978 Additional imports required to maintain per capita Ratio of Developing consumption additional import country at 1961-65 level requirement World total total in developing in developing (million (million countries' countries to metric metric (million world total Year tons) tons) metric tons) (percent) 1961 72.5 29.0 8.5 12 1962 83.5 30.3 5.3 6 1963 83.2 32.8 5.1 6 1964 97.8 36.1 5.2 5 1965 95.0 36.0 17.3 18 1966 110.7 41.1 17.0 15 1967 103.4 40.3 5.1 5 1968 96.9 40.2 8.1 8 1969 89.1 34.6 5.0 6 1970 96.9 41.0 7.9 8 1971 109.2 42.2 8.7 8 1972 109.7 42.9 9.5 9 1973 134.3 55.1 19.2 14 1974 142.4 60.5 9.2 6 1975 135.8 59.1 8.6 6 1976 152.5 57.2 7.4 5 1977 156.0 63.1 12.3 8 1978 166.5 74.6 7.3 4 a. Importing countries with an additional import requirement greater than one million tons were: 1961 Egypt, Morocco, and Turkey; 1962 Bangladesh; 1963 none; 1964 Brazil; 1965 India and Indonesia; 1966 Bangladesh, Brazil, India, Morocco, and Nigeria; 1967 Indonesia and Nigeria; 1968 Nigeria and Turkey; 1969 Egypt; 1970 Egypt and Nigeria; 1971 Bangladesh and Nigeria; 1972 Bangladesh and Nigeria; 1973 Egypt, Nigeria, and Turkey; 1974 Kam- puchea and Nigeria; 1975 Nigeria; 1976 Nigeria; 1977 Bangladesh and Nigeria; and 1978 Ethiopia and Vietnam. China is not included in the calculation of the additional import requirement because of the lack of reliable consumption data. Chinese imports, ranging from four to twelve million metric tons per year, are included in the totals for the world and the developing countries. Source: Fv.O trade tapes and IFPRI calculations. EFFECTS OF THE FACILITY ON INTERNATIONAL MARKET PRICES 75 pected storage cost is higher than the expected price increase in a crisis year. Although the facility would allow developing countries to obtain desired imports no matter what the price, higher prices would reduce the balance of payments benefits and raise the cost. If the facility created significant additional demand for grain imports in a year when the world balance of supply and demand was already precarious, and if this demand triggered sharp price increases such as those experienced in 1973 to 1975, much of the intended benefit could be negated. Determinants of the World Price of Wheat No model of world wheat trade has been developed that can be effectively used to test the possible effect of the food facility on prices. However, a number of experiments have been made with different specifications of the world wheat economy that enable us to make some remarks about the determinants of world price with a fair degree of confidence. Other authors have shown a nonlinear relation between wheat price, cereal stocks, and demand, in which sufficiently large stocks can absorb the shocks of production shortfalls and insulate prices. T. N. Barr has developed a model of demand and price de- termination for the United States fo:r 1960 to 1971.2 Price is a function of the ratio of normal dernand to ending stocks, in which stocks increase in importance as they are drawn down. For the United States, minimum working or pipeline stocks for 1973 were considered to be about 200 million bushels, which amounts to a little less than 40 percent of the domestic demand of 520 million bushels. The estimated effects of exogenous stock changes on prices are small, provided that the stocks amount to more than 400 million to 500 million bushels. Barr's specification 2. T. N. Barr, "Demand and Price Relationships for the U.S. Wheat Economy," Wheat Situation, No. WS225 (August 1973). 76 EFFECTS OF THE FACILITY ON INTERNATIONAL MARKET PRICES tracked prices well for 1960-71. For 1972-73 the estimated farm wheat price was US$1.78 per bushel compared with the actual price of US$1.80. Simulations from 1964 forward track well for 1964-72, and alternate assumptions permit high prices for 1973. For 1973-74 the model estimated a price of US$3.90, whereas simulations under alternate assumptions regarding exports yielded prices ranging from US$3.07 to $4.01 per bushel. The model also predicted a price of US$2.10 in 1974-75. Warner points out that Barr failed to consider explicitly the relation of market prices to government loan rates, which might affect the model during other periods.3 Appendix 3 to IFPRI Research Report no. 4 relates world wheat prices to world cereal demand for 1960-75.4 Two approaches are used. In the first, a simple linear equation explains price as a function of the imports of developing countries with food deficits, which is a variable demand element. This estimate yields an R' of 0.67. A second approach detrends variations in prices and imports and adds the effects of carry-over stocks as repre- sented by lagged price. This is estimated as a nonlinear relation: P,IP*=f (M,12Q,), In (P_,-,P*) where P* is the average world price before 1972, and M, is trend imports of cereals in the food-deficit developing countries. A random variable is introduced to account for shocks such as sudden changes in Soviet demand. The R2 is increased to 0.75, and the equation is used to simulate prices based on exogenous demand determinations. For 1980 and thereafter, the equation generates an expected price of $155.80, per metric ton, ex- pressed in 1977 dollars. 3. Dennis Lee Warner, "An Econometric Model of the World Wheat Econ- omy," Ph.D. dissertation, Princeton University, Princeton, NJ., 1979. 4. Panos Konandreas, Barbara Huddleston, and Virabongsa Ramangkura, Food Security: An Insurance Approach, Research Report no. 4 (Washington, D.C.: Inter- national Food Policy Research Institute, 1978). EFFECTS OF THE FACILITY ON INTERNATIONAL MARKET PRICES 77 Gerard Adams and Jere Behrman developed a general com- modity model to determine supply, demand, and prices for the world for 1955-71.5 Wheat price is a function of the ratio of stocks to demand, time, and the imports of centrally planned economies. This specification provides a high R2. Simulations on price are fair to good. This model supports the hypothesis that stocks, not price, absorb shocks. Simulations based on 5 percent changes in supply or demancl in 1956 show little change in simulated prices after one year, and the immediate effect also is not great. The model does not consider consumer and pro- ducer responses to price expectations. Estimated long-run elas- ticity of price to the ratio of stock to demancl is -0.16 with a one-year lag in response to exogenous supply or demand shifts, and an average decline in world prices of 4.0 percent a year. Warner's model of the world wheat economy provides esti- mates for 1948-74 and simulations for 1960-'3.6 Regional sup- ply and demand submodels are generated, and closure comes through the simultaneous solution of exports, imports, and world prices; import prices essentially follow regional export prices. The basic simulation misses the 1973 price jump, in part because the model does not adequately explain stocking behav- ior in North America. Basic price simulations for earlier years are strong, and in alternative simulations prices for 1973 are forced up by modifying the model with exogenous drawdowns of stocks. Predicted Effect of the FacilitY For this study we specified a nonlinear function in which the world export price for wheat is a function of the ratio of world stocks to world demand and a time trend (see the Appendix for 5. Gerard F. Adams and Jere R. Behrman, Econometric Models of World Agricul- tural Commodity Markets (Cambridge, Mass.: Ballinger Publishing Company, 1976). 6. Warner, "An Econometric Model of the World Wheat Economy." 78 EFFECTS OF THE FACILITY ON INTERNATIONAL MARKET PRICES the methodology). The results show both a significant declining trend in real world wheat prices and a significant relation be- tween changes in the ratio of stocks to demand and the export price for wheat, particularly as the ratio dropped below 19 per- cent (see Figure 16). The equation estimates only the short-term effect on prices of a change in the relation between supply and demand in a given year. With production fixed in the short term, elasticity of supply will depend on stocks. The specification is ex post and uses ending stocks and demand for a specific year to determine the price for that year. However, ending stocks serve here as a proxy for supply; that is, beginning stocks plus production. By using known data for beginning stocks and estimated data for Figure 16. Relation of Ratio of Stocks to Demand and It'orld Price of Htheat, Net of Time Trend 400 o o 0 U 350 - nJ 300 - ao 250 ° \ r- 200 I_ 150 I I I 1 14 16 18 20 22 24 26 Ending stocks as a percentage of demand Note: 0 actual values, net of estimated time trend; - estimated values, net of time trend. Sour-ces. Real prices, World Bank, Corimmodity Trade an1d Price Trends (Washington, D.C., 1979): stocks and demand, tT.S. Department of Agriculture, Foreign Ag- ricultural Service, Foreign Agricultural Citcular-Grainis. FG-28-81 Julv 1981); stock data are adjusted using IFPRI data for ending stocks of' China. EFFECTS OF THE FACILITY ON INTERNATIONAL MARKET PRICES 79 production and subtracting estimated demancl from their sum, an estimated ratio of ending stocks to demand can be obtained for each following year, and with it an indicaition of price be- havior, assuming demand behaves normally. The equation tells us that an increase in world demand for imports triggered by shortfalls in domestic production will not have much effect on price if beginning stocks are high and can be drawn down without causing the ratio of ending stocks to demand to fall below 17 to 20 percent. This was so throughout the 1960s, when the ratio always exceeded 20 percent, and real price was remarkably stable. The ratio of stocks to demand was less than 19 percent throughout the 1970s, however, and fell below 16 percent in 1973-76 with concomitant variability in world price. Except when the ratio exceeds 19 to 20 percent, the financial facility will be operating in a time of potentially unstable prices, and the effect of aclditional (demand on prices will depend importantly on the size of stocks when the demand increase occurs. As long as world grain reserves are adequate, any additional demand that might be created by the food facility could be expected to have little effect on worldi price. The effect could be considerably greater when a bad harvest reduces world stocks and a second bad production year forces, developing coun- tries to import greater quantities than normal. It is interesting that prices increased sharply in 1973 when the ratio of stocks to demand dropped from 18.5 to 15.5 percent in one year. The price increase was considerably larger than our estimation, and this indicates that speculative buying and hoard- ing probably drive prices higher than their equilibrium level because of uncertainty about future market behavior. With the facility in place, some of that uncertainty may be reduced. To the extent that the facility reduces uncertainty about access to supplies by assuring access to cheap credit to finance cereal imports no matter what the price, some of the iincentive to corner large supplies in a high-price market may be reduced. Over the longer term, market reactions to the existence of the food fi- 80 EFFECTS OF THE FACILITY ON INTERNATIONAL MARKET PRICES nancing facility could reduce price shocks, despite the pressure of additional demand. Even a small increase in average grain prices would have some lagged effect on production. Similarly, an increase in price variability would increase the profitability of holding stocks of grain and so increase grain stocks. Thus, the expected responses of production and stocks could offset some of the effect of the food financing facility on the level and variability of international market prices. Appendix. Methodological Notes THIS STUDY WAS CONCEIVED by the authors as a follow-up to research on the costs and benefits of an insurance scheme to provide food security for developing countries with food short- ages. The objective was to consider the merits and probable effect of alternative schemes that might actually be adopted by an international organization. In addition to drawing on previous work, three new activities were planned. First, it seemed worthwhile to visit a few representative de- veloping countries to find out how much government officials felt lack of foreign exchange constrained food imports and to obtain their views about the usefulness of a food financing fa- cility. Countries visited included Bangladesh, Egypt, Senegal, Brazil, and Peru. Even though its cereal imports have declined sharply since 1976, India was included as a sixth sample country because of the size of its food requirements a2nd the high prob- ability that a substantial shortfall in domestic production would occasionally necessitate large-scale imports. The selection of countries was to some extent arbitrary. The countries included, however, represent a diversity of situations with respect to geographical region, population size, income level, degree of self-sufficiency, degree of financial constraint, and variation of domestic policy. Table 12 gives data for the indicators used to show the diversity of the countries chosen. The brief country visits and interviews were not meant to provide a definitive analysis of the food security position of any 81 Table 12. Indicators of Food Import Vulnerability for Six Countries, 1976 to 1978 Average per capital calorie Share of Share of intake as Cereal import cereal imports cereal imports Coefficient of percentage of Population, GNP per capita, volume, in staple in export variation for FAO/WHO 1976-78 1976-78 1976-78 consumption, earnings, total staple standard Region and (millions of (1977 U.S. (thousands 1976-78 1976-78' consumption, 1977-79 country people) dollars) of metnic tons) (percent) (percent) 1961-77 (percent) Asia Bangladesh 77.8 86 1,355 9 12 7.6 77 India 641.3 155 2,852 2 6 5.3 90 Africa Egypt 39.4 326 5,079 44 13 12.6 111 Senegal 5.2 365 425 31 13 15.7 94 Latin America Brazil 116.2 1,419 3,978 9 4 5.8 105 Peru 16.2 726 1,037 30 7 3.9 90 a. Excluding value of food aid. Source: Barbara Huddleston, Closing the Cereals Gap with Trade and Food Aid (Washington, D.C.: International Food Policy Research Institute forthcoming); and A. Valdes and P. Konandreas, "Assessing Food Insecurity in Developing Countries," in Alberto Valdes, ed., Food Security for Developing Countries (Boulder, Col.: Westview Press, 1981), chapter 2. METHODOLOGICAL NOTES 83 of the sample countries. Nor was there any attempt to develop a common questionnaire or interview technique. Each author visited the assigned countries to meet and talk informally with individuals in key ministries and public agencies. In general, each sought to determine the degree of comm[itment by various countries to consumption stabilization policies that could involve considerable fluctuation in the level and cost of cereal imports from one year to another; to discuss with senior officials their perceptions of the usefulness of a food financing facility com- pared with other food security measures their governments were already pursuing or planning; to learn what specific responses their governments had made to past situations requiring an un- usual import response; to ascertain whether these officials con- sidered other problems, such as logistical or managerial constraints, more serious than the availability of foreign ex- change; and to verify the production, consumnption, and trade data available from official sources. From these discussions, the authors formed some impressions about the variety of country situations that aL financing facility might have to address and the hypoitheses most appropriate for testing in the analysis of the potential effect of a food financing facility. Thus, the visits helped in formulating the design of the study and the subsequent analysis reported here. Second, a simulation model was created to test the effect on food security of a financing facility operating under three dif- ferent specifications and to cornpare the results with the situation if there were no facility. The model estimatedl a country's food consumption based on plausible determining factors, including probability distributions of domestic food production, the price of imported food, and the availability of foreign exchange. Food security with and without a financing facility operating under various rules was simulated on the basis of 3,000 randomly drawn observations from these probability distributions. The experi- ments were repeated with different sets of parameters reflecting 84 METHODOLOGICAL NOTES different government policies, and the results were used to assess the effects of alternative financing arrangements. Finally, it was decided to perform some simulations with price- forming equations for wheat to ascertain the order of magnitude for price changes that could be anticipated from shifts in demand created by a financing facility. The Simulation Model This section briefly describes the equations and decision rules of the simulation model used to test the effect on food security of a financial facility under three different specifications and to compare these results with the situation if there was no facility. The demand for food in any year is Cl = aO - a, P where Cf = food consumption and Pf = price of food. Initially the model estimates consumption, C2, and price, P2, without imports, that is, (,i= Qf where Q= domestic food production. Solving these two equa- tions yields the closed economy price of food Pf = a0/a, - (l/a,) Cf. Next it is determined whether and how much imports are desired: O if P; < Pj - C; - Q 'if Pi > Pf and, if desired imports are positive, desired consumption is C, = ao-a1 Pj METHODOLOGICAL NOTES 85 where P' is the import price, and I' and C; are the desired amounts of imports and consumption. Whether these or differ- ent amounts are actually obtained depends on the amount of foreign exchange allocated to food imports. How the model handles the foreign exchange constraint is discussed in a sub- sequent section. The desired levels of imports and consumption are not nec- essarily determined by private demand. The government can decide to import more or less than what consumers are willing to buy at the prevailing border price, or, alternatively, it can tax or subsidize imports. In the policy stipulated to analyze the effectiveness of different financial facilities, the government in- tervenes to make the domestic prices more stable than the world price. This is equivalent to imposing a more inelastic demand through subsidizing imports when the world price is high and taxing imports when world price is low. Figure 17 illustrates the difference between the two policies. DD represents private market demand, and dd represents the demand with government intervention. If in a given year pro- duction is Qf, there is a positive demand for imports (Pf < Pf). Desired consumption is c; without government intervention and CT with government intervention. Similarly, the respective de- sired import levels are C - Qf and C,(f - Qf. The corresponding domestic prices are P) and Pf. If in another year production, Qf, is between Q;4 and C;s, there woulcl be no imports if the gov- ernment does not intervene to stabilize consumption, whereas with intervention a positive import demand exists. To analyze whether desired imports will be actually obtained, foreign exchange becomes constraining when M, + MAif > X where M; and M1 are the desired amount of foreign exchange for food and nonfood imports, res]pectively, (at the prevailing exchange rate), and X is total foreign exchange availability. Un- der a completely flexible rate of exchange regime and without 86 METHODOLOGICAL NOTES Figure 17. Consumption with and without Government Intervention to Stabilize Price and Consumption Price d D P~ P1 d D Q1 C;' C;' Quantitv import controls, an excess demand would bring about a change in the exchange rate and proportional reductions in the desired demand for both food and nonfood imports. More realistically, however, allocations of foreign exchange to food and nonfood imports in scarcity situations is often made on the basis of po- litical and other considerations, that is, different implicit ex- change rates for food and nonfood imports prevail. To highlight how different propensities of governments to restrict foreign exchange to food imports (m) will influence the effect of a financial facility on food security, the allocation model is specified as follows _U; .if MV; < Xl IM; + m (Xf-M;) if M; > Xl METHODOLOGICAL NOTES 87 where XJ = X - Mf. Similarly, 12 I;if M; < X, ll; +1 m(Xf- M;)/Pf if M; > Xl and C; if M; < X, Cf C; + m(Xf - m;)I)/Pf if M; > X,. The government's propensity to r estrict foreign exchange for food imports when foreign exchange is scarce, m, can take any value between 0 and 1. A country that accords high priority to food security will give first priority to food imports, that is, m = 0, and actual imports will always be desired imports. At the other extreme, m = 1 implies that ithe food import bill will not be allowed to exceed Xf and that food imports will be restricted to Xi /Pf The three financial facilities analyzed provide for drawings of foreign exchange as follows: 0 ifX2X SAFE = l XX) if X M 88 METHODOLOGICAL NOTES 0 O if X - IV; < -Mf INTEG = (X -X) + (M'; - If) ifx X-M; >X- . In summary, whether and to what extent food consumption in a particular year falls below a critical limit depends on (a) domestic production and the world price of food, (b) the avail- ability of foreign exchange, (c) the country's policy for insulating the domestic price from world price fluctuations, (d) the coun- try's propensity to allocate foreign exchange to food and non- food imports, and (e) the type of financial facility available for reducing the foreign exchange constraint. Depending on the combination of circumstances, the effect of a financial facility on food security can range from practically zero to very large. The contribution of a financial facility to nonfood imports is inversely related to its effect on food security. The assumptions concerning the propensity for allocating for- eign exchange to food and nonfood imports are highly arbitrary. At best, they indicate only generally the range of likely behavior of governments. For simplicity, the model also does not capture changes in real income attributable to changes in the external account. Data and Parameters Used in Simulation Experiments The simulated import price of cereals is a skewed distribution with a mean of US$225 per ton, a median of US$205 per ton, and a standard deviation of US$50 per ton. Ninety-five percent of the time, the price is between US$155 and US$345 per ton. The parameters assumed for the level and stability of cereal production, the availability of foreign exchange for cereal im- ports, cereal consumption, and cereal imports when production METHODOLOGICAL NOTES 89 Table 13. Basic Country Data Used inl the Simulation Cereal Cereal Foreign exchange consump- Cereal production available for cereal twon at imports at (millions importsa (millions mean mean of tons) of U.S. dollars) import import ________ ____ - pnce pnice Standard Standard (millions (millions Country MVean deviation Mean deviation of tons) of tons) Bangladesh 19.0 1.4 425 70 21.0 2.0 India 130.0 8.0 400 450 130.0 0.0 Egypt 8.0 0.37 800 200 12.0 4.0 Senegal 0.75 0.19 77 50 1.1 0.35 Brazil 28.0 2.7 660 500 31.0 3.0 Peru 1.5 0.12 250 100 2.6 1.1 a. Assumed to equal the mean value of cereal imports. Source: FAO data for production, consumption, and imports; IMF data for foreign exchange. and the food import prices are at their mean values are shown for the six sample countries in Table 13. Private market demand for all countries is assumed to be a linear function of price, with an elasticity of -0.3 at the mean world price and mean country consumption. Stabilized consumption is correspondingly ana- lyzed through a linear demand function with price elasticity of -0.1 at the mean import price and mean consumption. Changes in the World Price of Wheat Caused by the Facility The effects of a financial facility on the world price of wheat were determined in two steps. First, two estimates of the addi- tional demand that might be created by a facility were derived: one based on the deviation of actual imports from a five-year moving trend for 1961 to 1968, and the other based on the 90 METHODOLOGICAL NOTES deviation of actual imports from the level of imports required to maintain per capita cereals consumption at its 1961-65 av- erage level. Second, a price-forming equation was derived. Following ear- lier work by Barr, Adams and Behrman, and Konandreas and others, described in Chapter 5, it was assumed that world stock levels and variation in import demand would have an important influence on price variability. Although substantial explanatory power is lost in attempting to capture price behavior through a single equation, it was felt that as long as the equation tracked historical price variation well and coincided with other theoret- ical work on wheat prices, it could indicate the likely effect of demand changes, such as those resulting from a food facility. Several specifications for a price-forming equation were in- vestigated, and the conclusions of Barr, and Adams and Behrman that the level of stocks in relation to demand is an important variable in predicting cereal prices proved useful in estimating wheat prices. The relation of stocks and demand to price is expected to be nonlinear and downward sloping so that response of prices to the ratio of stock to demand is greater when the ratio is lower; demand shocks, given high relative stock levels, are thought to have a negligible effect on prices. Drawing further on the work of Adams and Behrman, time trends in price as well as the current period response of prices to expected demand or actual forward contracting were explored. Using ordinary least squares and the Corchran-Orcutt pro- cedure for correcting for serial correlation, the following price- forming equation was selected to track wheat prices: In P;, = 9.08 + 7.17 ln (ES/D),+ 1.55 /(ES/D), - 0.039 T (2.16) (2.48) (-2.95) with p = 0.42, R2 = 0.60, R2 = 0.62, DW = 1.71, and F(3, 13) = 9.55 for 1961 to 1978 where P, -world price of wheat in 1977 U.S. dollars in calendar year t; ES, _ total world cereal ending stocks for the crop year ending in year t; D, _ total world cereal demand for the crop year ending in year t; T _ time METHODOLOGICAL NOTES 91 Table 14. Estimated and Actual Wheat Prices Associated with Different Ratios of World Cereal Stocks to World Demand (1977 U.S. dollars per ton) Estimated pricea Actual distributionb Ratio (percent) 1962 1970 1978 Years Price 14.0 394 288 211 14.5 346 253 185 15.0 309 226 166 1975 153 15.5 280 205 150 1973, 1976 219, 135 16.0 257 188 138 1974 228 16.5 239 175 128 17.0 224 164 120 17.5 213 156 114 1978 108 18.0 204 149 109 18.5 197 144 105 1971, 1972, 1977 132, 133, 96 19.0 191 140 102 1966 153 19.5 187 137 100 20.0 184 134 98 20.5 181 133 97 21.0 180 132 96 21.5 179 131 96 1962 157 22.0 180 132 96 1963, 1967 154, 150 22.5 180 132 97 1965, 1968, 1970 145, 151, 132 23.0 182 133 97 1964 163 23.5 184 135 99 24.0 186 136 100 1961, 1969 154, 144 24.5 189 139 101 25.0 193 141 103 25.5 197 144 105 26.0 201 147 108 26.5 206 151 111 a. Estimated prices are based on the finction: In P, = 9.08 + 7.17 In (ESID) 4- 1.55/(ES/D), - 0.039 T. Each value of the ratio of stocks to demand yields a corresponding price estimate. Because a time trend element (T) is included, the set of price estimates shifts downward over time. Ratios of stocks to demand are presented with the cor- responding price estimates for 1962, 1970, and 1978. b. Actual prices and the years in which thiey were observed are tabulated to correspond with observed ratios of stocks to demand. Source: Estimated prices from IFPRI calculations; actual prices from World Bank data. 92 METHODOLOGICAL NOTES (1961) = 1; and figures in parentheses represent the t values. This estimation supports the hypothesis that wheat prices have had a declining secular trend and that the price response to import demand shifts becomes increasingly important as the ratio of stock to demand drops (see Table 14). Various attempts to constrain the ES/D variable by using a threshold above which an increase in ES/D would yield no mar- ginal increase in p were inconclusive. Threshold values were defined alternatively as 0.17, 0.20, and 0.22 percent for ES/D. Several different values were tried for the variable M, but none yielded satisfactory results. References Abbott, Philip C. "Modeling International Grain Trade with Government Controlled Markets." American Journal of Agricul- tural Economics, vol. 61 (February 1979), pp. 22-31. Adams, Gerard F., and Jere R. Behrman. Econometric Models of World Agricultural Commodity Markets. Cambridge, Mass.: Ballin- ger Publishing Company, 1976. Barr, T. N. "Demand and Price Relationships for the U.S. Wheat Economy." Wheat Situation, No. WS225 (August 1973). Behrman, J. R. Foreign Trade Regimes and Economic Development: Chile. Foreign Trade Regimes and Economic Development. Vol. 8. New York: National Bureau of Economic Research, 1976. Bigman, D., and Shlomo Reutlinger. "Food Price Stabilization: National Buffer Stocks and Trade Policies." American Journal of Agricultural Economics, vol. 61, no. 4 (November 1979). Food and Agriculture Organization of the United Nations and UN World Food Council. "World Food Security: The Need for Balance of Payments Support to Meet Exceptional Vari- ations in Food Import Bills." Rome, January 1980. Gilmore, Richard, and Barbara Huddleston. "The Food Security Challenge." Food Policy, vol. 8, no. 1 (February 1983), pp. 31- 46. Green, Christopher, and Colin Kirkpatrick. "Insecurity, Food Financing and the IMF." Food Policy, vol. 6 (August 1981). Goreux, Louis. Compensatory Financing Facility, Pamphlet Series no. 34, Washington, D.C.: International Monetary Fund, 1980. 93 94 REFERENCES Huddleston, Barbara. Closing the Cereals Gap with Trade and Food Aid. Washington, D.C.: International Food Policy Research Institute. Forthcoming. International Monetary Fund. "Decision on Compensatory Fi- nancing of Fluctuations in the Cost of Cereal Imports and Consequential Amendments." Washington, D.C., May 14, 1981. Johnson, D. Gale. "The World Food Situation: Developments during the 1970s and Prospects for the 1980s." In AEI Studies on Contempory Economic Problems. Washington, D.C.: American Enterprise Institute, 1980. Konandreas, Panos, Barbara Huddleston, and Virabongsa Ra- mangkura. Food Security: An Insurance Approach. Research Report no. 4. Washington, D.C.: International Food Policy Research Institute, 1978. Lofchiie, M. F. "Agrarian Crisis and Economic Liberalization in Tanzania." Journal of Modern African Studies, vol. 16, no. 3 (1978), pp. 451-75. McIntire, John. Food Security in the Sahel: Variable Import Levy, Grain Reserves, and Foreign Exchange Assistance. Research Report no. 26. Washington, D.C.: International Food Policy Research In- stitute, 1981. Morrow, Daniel T. The Economics of the International Stockholding of Wheat. Research Report no. 18. Washington, D.C.: Inter- national Food Policy Research Institute, 1980. Pitt, M. M. "Alternative Trade Strategies and Employment: In- donesia." In A. 0. Krueger, H. B. Lary, T. Monson, and N. Akrasanee, eds., Trade and Employment in Developing Countries. Vol. 1. Individual Studies. Chicago: University of Chicago Press, 1981. Reutlinger, Shlomo. "Food Security: Magnitude and Remedies." World Development, vol. 6 (1978), pp. 797-811. . "Policies for Food Security in Food-Importing Devel- oping Countries." In A. H. Chisholm and R. Tyeis, eds., Food Security: Theory, Policy, and Perspectives from Asia and the Pacific Rim. Chapter 2. Lexington, Mass.: D.C. Heath, 1982. REFERENCES 95 Scobie, Grant M. Food Subsidies: Theitr Impact on Foreign Exchange and Trade in Egypt. Research Report no. 40. Washington, D.C. International Food Policy Research Institute, August 1983. . Government Policy and Food Imports: The Case of Wheat in Egypt. Research Report no. 29. Washington, D.C.: Interna- tional Food Policy Research Institute, 1981. Siamwalla, Ammar, and Alberto Valdes. "Food Insecurity in De- veloping Countries." Food Policy, vol. 5, no. 4 (November 1980). U.S. Department of Agriculture, Foreign Agricultural Service. Foreign Agricultural Circular-Grains. FG-28-81. Washington, D.C., July 1981. Valdes, Alberto, ed. Food SecurityJfor Developing Countries. Boulder, Col.: Westview Press, 1981. Warner, Dennis Lee. "An Econornetric Model of the World Wheat Economy." Ph.D. dissertation, Princeton University, Princeton, NJ., 1979. World Bank. Commodity Trade and Price Trends. Washington, D.C., 1979. Index (Page numbers in italics indicate tabular veloping-country imports of, 74. material.) See also Corn; Food; Grains; Rice; Wheat Abbott, P., 71-72 CFF. See Compensatory financing fa- Adams, G., 77 cility (CFF) Algeria, food consumption in, 5 Chile, food imports in, 34 Columbia, food imports in, 29 Balance of payments, 4; financial Commodity Credit Corporation, 17 support for, 36, 57, 60-61, 67. See Commodity model, general, 77 also Foreign exchange Compensatory financing facility Bangladesh, 81; country data on, 89; (CFF), 9; amended to cover cereal effects of different assistance import bills, 10-11; cereal com- schemes on, 39 ponent in, 56-57; quota limits and Barr, T. N., 75-76 drawings from, 59; SAFE and, 42 Behrman, J. R., 34, 77 Consumption: cereal production Brazil, 81; country data on, 89; dif- and, 21, 23; desired, 39; govern- ferent assistance schemes for, 39; ment policy on, 39, 68; INTEG and, stabilization measures for, 54 46; international financial arrange- Buffer stocks of food, 5, 67; finance ments to avoid reductions in, 7; for, 52-55 lack of mandate for IMF to increase, 70; stability of, 37, 38, 39, 68; of Cereal import bills, 23; facility to staple foods, 22; variability in, 4, 5 cover excess, 10, 42; relieving the Corn, export prices of, 14-15 strain of high, 72; variability of, 26-27. Cereals: buffer stocks of, 67; cause Demand: equation for, 84; increase of variable import bills for, 23; ef- in, 70, 73; inelasticity of, 71-72 fect of unstable production of, 21, Developing countries: borrowing 22; food production and variable problems of, 61; costs of food im- consumption of, 21; imports to ports for, 35; lack of foreign ex- meet shortfalls of, 24, 25; imports change in, 7; price stabilization as negative exports, 56, 64; in- policies of, 18-19; stabilization in creased demand for, 71-73, 75; food supply by, 6-7 price of imported, 3; world de- I)omestic policies. See Government mand for, 76, 91; world and de- policies 97 98 INDEX Egypt, 81; country data on, 89; ef- for chosen alternative for, 64-66; fects of different assistance level of food aid and, 66; limita- schemes on, 39; import demand in, tions and benefits of, 60-61, 64; 34 objectives of, 11, 66; participation European Economic Community in, 65; predicted effect of, 77-80; (EEC), price stabilization policies proposal for joint quota limit for in, 18 CFF and, 57-58; rejection of sec- Export earnings: compensation ond, 11; SAFE and, 42 shortfalls in, 56; food import bill Food financing schemes: alternative, and, 28-29, 31, 32-35; quota lim- 41-43; balance of payments relief its for, 57-58, 59-60. See also For- and, 51; drawings from, 46-47; ef- eign exchange fects of different, 39-41, 88; im- ports and, 50-51; investment in nonfood enterprises and infra- Food and Agriculture Organization, structure and, 61; potential effect 71 of, 83 Food aid: accounting for flows of, 66; Food import bills, 38; cause of vari- compensation by food financing ability of, 23; composition of, 62- facility and, 66; lack of responsive- 63; facility to stabilize (SFIB), 42, ness of (to production shortfalls), 46, 50, 87; foreign exchange and, 35; supplies for (in crises), 65. See 28-29, 30-35; high-income coun- also Compensatory financing facil- tries' ability to vary, 35; implica- ity (CFF); Food financing facility, tions of large and unexpected, 29, IMF; Food import bills, facility to 30, 31, noncereal items in, 61, 64. stabilize (SFIB); Foreign exchange, See also Cereal import bills facility to stabilize available (SAFE); Food insecurity: causes of, 3, 5; fa- Integrated stabilization facility (IN- cilities for dealing with, 41-43; TEG) policies producing, 3, 10. See also Food financing facility: concept of, Food security 7-8; food security and, 9-10. See Food security: causes of differences also Compensatory financing facil- in, 9; effect of financial facility on, ity (CFF); Food financing facility, 41; government policies influenc- IMF; Food import bills, facility to ing, 9, 36, 41; integrated CFF access stabilize (SFIB); Foreign exchange, and increase in, 9-10; internal facility to stabilize available (SAFE) prices and, 44, 45-46; policy A Food financing facility, IMF: alterna- with low concern for, 43-44, 48, tive proposals for, 56-60; availa- 50, 53; policy B with high concern bility of assistance from, 11-12; for, 44, 48, 50, 54, 55; stabilization compatibility with other assistance, of food supplies and, 52-54 65; creation of, 10, 56; demand Food supply. See Supply generated by, 70; determination of Foreign currency reserves, 7, 35 need through, 66-68; emergency Foreign exchange, 7; assistance price increases and, 12; futures schemes for, 43, 48; compensation markets and, 68-69; hypothetical for shortfalls in, 9; constraints, 7, withdrawals from, 58; justification 50, 54-55, 60, 85; excess food im- INDEX 99 port bills and, 29, 30, 32; facility nerability of, 82. See also Nonfood to stabilize available (SAFE), 42, 45, imports 46, 50, 87; from food financing fa- Income, programs to maintain, 6, 8 cility of IMF, 1 1; food imports and, Ind[ia, 81; country data on, 89; effects 38,40-41, 81, 86, 87; nonfood im- of different assistance schemes on, ports and, 38; restructuring use of, 39; effects of stabilization meas- 40-41, 86, 87; scarcity of, 38, 41. ures in, 53, 54; food consumption See also Export earnings lievels in, 5; networks for channel- Forward contracts, 68-69, 73 ing food imports within, 40 Futures markets. See Forward con- Inclonesia, food imports crowding tracts out other imports in, 34 INTEG. See Integrated stabilization fa- cility Government control, identifying a Integrated stabilization facility (IN- situation beyond, 66, 67 TEG), 43; average annual imports Government policies: on domestic and, 48; consumption instability distribution, 68; on farm prices, under, 46, 55; food security and, 16; food security and, 9, 36; inter- 44; foreign exchange and, 50, 88 national price instability and, 19; International food financing facility. stability of consumption and, 39, See Compensatory financing facil- 68; to stabilize domestic prices of ity (CFF); Food financing facility, grains, 18; unpredictability of, 72. IMF; Food import bills, facility to See also Food insecurity; Food se- stabilize (SFIB); Foreign exchange, curity iFacility to stabilize available (SAFE); Grain: exporters, major, 19; prices, [ntegrated stabilization facility (IN- 14-19; production and stocks of, TEG) 16-17; U.S. storage policies for, International Food Policy Research 17. See also Cereals; Corn; Rice; Institute, 71, 76 Wheat International Monetary Fund (IMF), 9; creation of integrated financial Hoarding, 79 facility by, 10; lack of mandate to increase consumption or improve IMF. See International Monetary Fund nutrition, 70. See also Compensa- Imports: compensation for commer- tory financing facility (CFF); Food cial, 69; compensation for exces- financing facility, IMF sive bills for, 9, 11; desired, 38; of Irrigation, fluctuations in food pro- essential raw materials and capital duction and, 6 goods, 34; export revenues and, 31; financing facility and, 11, 50- Libya, ratio of food imports to total 51; food insecurity and, 9-10; for- exports in, 29 eign exchange and, 38; inelastic Lofchie, M. F., 34 demand for, 34; pressure on ca- pacity for, 29; for rural consumers, Marketing, programs for rural, 61 61; variability in costs of, 19, 23, Model: to assess effects of interna- 28, volume of, 6-7, 73, 74, 75; vul- tional assistance schemes, 37; to 0oo INDEX estimate a country's food con- SAFE. See Foreign exchange, facility sumption, 37; food security and, to stabilize available (sAFE) 9-10; general commodity, 77; to Scobie, G., 34, 71-72 insure against high food import SDRs (special drawing rights), 57 bills, 8-10; simulation, 37-41, 83, Senegal, 81; country data on, 89; ef- 84-88. See also Simulation experi- fects of different assistance ments schemes on, 39 Morocco, food consumption in, 5 SFIB, 42, 46, 50, 87 Shipping rates, variability in ocean, National policies. See Government 15 policies Simulation experiments, data and Need, determination of, 68 parameters used in, 88-89. See also Nigeria: food consumption in, 5; ra- Model tio of food imports to total exports Soviet Union, price policies for food in, 29 and agriculture in, 18 Noncereal items, 61, 64 Special drawing rights (SDRS), 57 Nonfood imports: adjustments in, Speculative buying, 79 34; lack offlexibility to reduce, 39; Supply, 52-55; food stocks and, 6; preventing harmful reductions in, instability in, 3-4; policies on, 3, 51 5; ways to stabilize, 6-7 Output, efforts to limit, 16 Tanzania, food imports crowding out other imports in, 34 Peru, 81; country data on, 89; effects Trade: free, 6; managed and varying of different assistance schemes on, net, 18 39 United States: legislation to increase Pitt, M. M., 34 price instability and, 17; stocking Price: equation, 90; increases, 12, 70; behavior in, 77; support policies instability, 17, 18; insulating do- in, 16-17 mestic from world, 8-10, 36; mar- ket,40-41;stabilization,8, 16,18- Warner, D. L., 76 19, 41; supports (U.S.), 16-17; Wheat: elasticity of supply of, 78; variability, 6, 17-18. See also Ce- model of demand and price deter- reals; Rice; Wheat prices mination for, 75; stocks of, 75-76, Procurement, programs for rural, 61 77, 78, 79 Production: fluctuations in domestic, Wheat prices: determinants of world, 3, 5, 19-21, 23; fluctuations in 75-77, 77-80; effect of food fi- global, 19; of staple foods, 22, 24 nancing facility on, 75, 89-92; highest average export, 14; relative Quota limits for export earnings and to world cereal demand, 76; sim- CFF cereal component, 57-58, 59- ulations with equations for, 84; 60 trends in world, 77, 78; U.S. ex- port, 14 Repayment provisions, 65 Wheat production, efforts to reduce, Rice, variability in export prices for, 16-17 15 Wheat Trade Convention, 12 The full range of World Bank publications, both free and for sale, is described in the Catalog of Publications, the continuing research pro- gram is outlined in Abstracts of Current Studies. Both booklets are updated annually; the most recent edition of each is available without charge from the Publications Sales Unit, Departrnent B, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A. Barbara Huddleston is a research fellow and Alberto Valdes is di- rector of the International Trade and Food Security Program at the International Food Policy Research Institute. D. GaleJohnson is Eliakim Hastings Moore Distinguished Service Professor and chairman of the Department of Economics at the University of Chicago. Shlomo Reut- linger is senior economist in the Agriculture and Rural Development Department of the World Bank. 0 8018 3070 2