A .V2/S JIL KR K1- WCiiLC-d rT Poverty and Hunger Issues and Options for Food Security in Developing Countries 9275 The worl_ February 1986 food. Theg'rowthlof global food production has been faser thian the unprecedented popuaion growth of the past forty years. . ... Yet many poor counties and hundreds of miSons of poor people do not share in this abun- dance. They suffer from a lack of food secuity, caused mainly by a lack of purchasing power. Poverty and Hunger A World Bank Policy Study Why is the food of the people so scarce? . . . Where does the blame lie? . . . I have been unable to attain a proper balance between impor- tant and unimportant affairs. Let this matter be debated. . . . Let all exhaust their efforts and ponder deeply whether there is some way to aid the people. Edict of Emperor Wen on the Primacy of Agriculture (163 B.C.) It is no longer the simple field, however big, but the whole world which is required to nourish each one of us. Teilhard de Chardin The Phenomenon of Man Poverty and Hunger Issues and Options for Food Security in Developing Countries The World Bank Washington, D.C., U.S.A. Copyright C 1986 by The International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W., Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing February 1986 The map used in this document is solely for the convenience of the reader and does not imply the expression of any opinion whatsoever on the part of the World Bank or its affilates concerning the legal status of any country, territory, city, area, or of its authorities, or concerning the delimitation of its boundaries or national affiliation. Library of Congress Cataloging-in-Publication Data Poverty and hunger. (A World Bank policy study) Bibliography: p. 1. Food supply-Developing countries. 1. International Bank for Reconstruction and Development. II. Series. HD9018.D44P68 1986 363.8'09172'4 86-1583 ISBN 0-8213-0678-2 Foreword Food security has to do with access by all people only slowly. This report deals with how the second at all times to enough food for an active and healthy difficulty can be addressed, at least in part, through life. Available data suggest that more than 700 mil- food and agricultural policies. lion people in the developing world lack the food The issue of food security has long been a cen- necessary for such a life. No problem of underde- terpiece of discussions of food and agricultural velopment may be more serious than, or have such policy. Many governments feel strongly about this important implications for, the long-term growth issue and will continue to commit significant of low-income countries. amounts of their resources to help alleviate food Attempting to ensure food security can be seen insecurity. Some of the programs that result are as an investment in human capital that will make effective and contribute to general economic for a more productive society. A properly fed, growth while at the same time providing the poor healthy, active, and alert population contributes and disadvantaged with greater access to food. more effectively to economic development than one Other programs sacrifice too many of the resources which is physically and mentally weakened by in- needed for economic growth or fail to reach those adequate diet and poor health. whose food security is at stake. Problems in food security do not necessarily re- Programs must be tailored to the needs of each sult from inadequate food supplies, as is widely country, and, to avoid waste, the cost-effectiveness believed, but from a lack of purchasing power on of aiternatives must be carefully evaluated before the part of nations and of households. Economic resources are committed. Ultimately, a careful bal- growth will ultimately provide most households ancing of measures for trade, production, and pov- with enough income to acquire enough food. Sup- erty alleviation are required in most countries. porting economic growth with an equitable distri- There is no one optimal solution to the problem bution of income is therefore the first priority, and of food security, any more than there is one solution should continue as the main goal, of economic pol- to the problem of poverty. But to help developing icy. But there are two well-known difficulties: eco- countries improve their food security, the international nomic growth takes time, and, even when it is community should: achieved, the present distribution of assets and op- Intensify efforts to accelerate growth, through portunities means that large numbers of poor peo- adjustment assistance, policy reform, and produc- ple are likely to increase their purchasing power tive investment v * Further increase the attention given to poverty used in the name of food security should be used alleviation in cost-effective ways. Each country has to decide * Help design, for the short and medium term, how much food security it wants and how many cost-effective programs to alleviate chronic hunger resources it can dedicate to that purpose. This re- and prevent famines, and pay special attention to port provides insights and tools for ana- the needs of the very young, among whom malnu- lyzing problems of food security and for designing trition can cause irreversible damage programs to increase it. * Help countries coordinate food aid with other The World Bank stands ready to help countries forms of economic assistance. address their food security problems. It is also pre- This report outlines the nature and extent of food pared to contribute resources. The alleviation of security problems in developing countries, ex- poverty and hunger are, after all, the primary pur- plores the policy options available to these coun- poses of economic development. tries in addressing these problems, and indicates what international institutions such as the World Bank can and should do to help countries solve their food security problems. It suggests policies to achieve the desired goal in cost-effective ways. It also identifies policies that waste economic re- sources and fail to reach the target groups. It is in that sense as much about what should not be done as about what should be done. A. W. Clausen Probably no nation can be 100 percent food se- President cure. That is all the more reason why resources February 1986 The World Bank The report was prepared by Shlomo Reutlinger and Jack van Holst Pellekaan with the assistance of Craig Lissner, Claudia Pendred, and Colleen Roberts. The authors acknowledge the in- valuable assistance of Herman van der Tak, Marcelo Selowsky, Hans Binswanger, and G. Edward Schuh in the drafting of the final version. Thanks go to the many readers inside and outside the Bank who reviewed various drafts, to Helen Claverie and Morissa Young for typing, and to Bruce Ross-Larson for edit- ing. The work was carried out under the general direction of S. Shahid Husain. vi Contents Definitions ix Glossary xi 1. Overview 1 Chronic Food Insecurity 1 Transitory Food Insecurity 4 Options for National Policy 6 The Role of External Assistance 10 Concluding Comments 12 2. A Problem of Supply or of Purchasing Power? 13 Recent-Levels of Supplies and Prices 13 Future Levels of Supplies and Prices 15 Chronic Food Insecurity 16 Transitory Food Insecurity 21 3. National Measures to Reduce Chronic Food Insecurity 28 Overview of Policy Interventions 28 Increasing the Food Supply 30 Subsidizing Food Prices 34 Augmenting Incomes 37 Interventions in Different Conditions 41 4. National Measures to Reduce Transitory Food Insecurity 42 Stabilizing the Domestic Food Supply 43 Stabilizing Domestic Demand 47 Protecting Vulnerable Population Groups 47 Regional Schemes for Cooperation 48 vii 5. International Support for Food Security 49 Analysis and Advice 50 External Finance 50 International Trade 53 Annex A. Methodologies 55 Estimating the Share of the Population with Chronically Inadequate Diets 55 Calculating the Approximate Efficiency of a Price Subsidy on Selected Foods 56 Annex B. Data 58 The Global Food Supply 58 Terms of Trade 61 Annex C. Econometric Analysis of the Determinants 63 of Food Consumption References 67 viii Definitions Billion is 1,000 million. The country groups used in this report are defined Cereals include wheat, rice, maize, rye, sorghum, as follows: millet, barley, oats, and mixed grains. The terms * Developing countries are divided into: low- "cereal" and "grain" are used interchangeably in income economies, with 1983 gross national this report. product (GNP) per person of less than $400; and middle-income economies, with 1983 GNP per Coefficients of variation are defined as the standard m income econtries deviation around the same time trend used for person of $400 or more. Middle-income countries are also divided into oil exporters and oil importers, ras dentified below. Dollars are U.S. dollars unless otherwise specified. * Middle-income oil exporters comprise Algeria, Growth rates used in this report were computed Angola, Cameroon, People's Republic of the usin the least-squares method. The least-sq s Congo, Ecuador, Arab Republic of Egypt, Gabon, using te least-squares method. Tne least-squares Indonesia, Islamic Republic of Iran, Iraq, Malaysia, growth rate, r, is estimated by fitting a least-squares Mexico, Nigeria, Peru, Syrian Arab Republic, linear trend line to the logarithmic annual values Trico, nigeria, Tu,isiand Republa. Of the variable in the relevant period. More Trinidad and Tobago, Tunisia, and Venezuela. of th varible i the elevat perod. Mre 9Middle-income oil importers comprise all other specifically, the regression equation takes the form Middle-income developing countries not classified of log X, = a + bt + et, where this is equivalent as oil exporters. to the logarithmic transformation of the compound * High-income oil exporters (not included in growth rate equation, X, = XO (1 + r)'. In these developing countries) comprise Bahrain, Brunei, equations Xis the variable, t is time, and a - log lopmg Omn, Qatar, Sau rabi, and Xo and b = log (1 + r) are the parameters to be Uwit, Liba,man a d, estimated; e, is the error term. if b* is the least- Unted Arab Emirates. squares estimate of b, then the annual average *Industrial market economies are the members sures rtes,timat of b, then th .annualg verag of the Organisation of Economic Co-operation and Development, apart from Greece, Portugal, and Tons are metric tons. Turkey, which are included among the middle- ix income developing economies. This group is Iraq, Israel, Jordan, Kuwait, Lebanon, Libya, commonly referred to in the text as industrial Morocco, Oman, Saudi Arabia, Syrian Arab economies or industrial countries. Republic, Tunisia, Turkey, Yemen Arab Republic, * East European Nonmarket Economies include People's Democratic Republic of Yemen, and the following countries: Albania, Bulgaria, United Arab Emirates. Czechoslovakia, German Democratic Republic, * East Asia and Pacific comprises all low- and Hungary, Poland, Romania, and U.S.S.R. This middle-income countries of East and Southeast Asia group is sometimes referred to as nonmarket and the Pacific, east of, and including, Burma, economies. China, and Mongolia. * Sub-Sabaran Africa comprises all thirty-nine * South Asia includes Bangladesh, Bhutan, India, developing African countries south of the Sahara, Nepal, Pakistan, and Sri Lanka. excluding South Africa. * Latin America and the Caribbean comprises * Middle East and North Africa includes all American and Caribbean countries south of the Afghanistan, Algeria, Arab Republic of Egypt, Iran, United States. x Glossary ASEAN Association of South East Asian GDP Gross domestic product Nations GNP Gross national product CARE Cooperative for American Relief IBRD International Bank for Reconstruction Everywhere and Development CFF Compensatory Finance Facility of the IFPRI International Food Policy Research IMF Institute CGIAR Consultative Group on International IMF International Monetary Fund Agricultural Research SDR Special Drawing Rights CIAT Centro Internacional de Agricultura UNCTAD United Nations Conference on Trade Tropical and Development CIF Cost, insurance, and freight UNDP United Nations Development EC European Communities Programme FAO Food and Agriculture Organization of LJNICEF United Nations Children's Fund the United Nations USDA United States Department of FOB Free on board Agriculture GATT General Agreement on Tariffs and WFP World Food Programme Trade WHO World Health Organization xi Overview The world has ample food. The growth of global Chronic Food Insecurity food production has been faster than the unprece- dented population growth of the past forty years. Prices of cereals on world markets have even been How many people do not have enough to eat? falling. Enough food is available so that countries Where do they live? How have their numbers and that do not produce all the food they want can geographical distribution changed? import it if they can afford to. Yet many poor countries and hundreds of millions of poor people Data for 1970 and 1980 do not share in this abundance. They suffer from a lack of food security, caused mainly by a lack of Between 340 million and 730 million people in the purchasing power. developing countries did not have enough income The term, "food security," although interpreted to obtain enough energy from their diet in 1980. in many ways, is defined here as access by all peo- (These estimates exclude China because data are ple at all times to enough food for an active, healthy not available.) The estimate of 340 million is based life. Its essential elements are the availability of on a minimum calorie standard that would prevent food and the ability to acquire it. Food insecurity, serious health risks and stunted growth in children. in turn, is the lack of access to enough food. There If the standard is increased to levels that allow an are two kinds of food insecurity: chronic and tran- active working life, however, the estimate rises to sitory. Chronic food insecurity is a continuously 730 million (Figure 1-1). About two-thirds of the inadequate diet caused by the inability to acquire undernourished live in South Asia and a fifth in food. It affects households that persistently lack Sub-Saharan Africa (see the map on the following the ability either to buy enough food or to produce page). In all, four-fifths of the undernourished live their own. Transitory food insecurity is a tempo- in countries with very low average incomes. rary decline in a household's access to enough food. If it is assumed that income distributions did not It results from instability in food prices, food pro- change during the 1970s, the share of people with duction, or household incomes-and in its worst inadequate diets declined between 1970 and 1980 form it produces famine. (Figure 1-2). But this assumption is optimistic and 1 The Prevalence of Energy-Deficient Diets, 1970 and 1980 *High 1980 ! Low Not included IBRD A19229/A19230 August 1985 2 Figure 1-1. ThePrevalenceofEnergy-Deficient probably overstates the improvement. In any event, Diets in Eighty-seven Developing Countries, 1980 because of population growth, the number of peo- ple with inadequate diets appears to have increased Not enough calories for Not enough calories to under both calorie standards. The largest declines an active working life prevent stunted growth in shares and numbers were in East Asia and the (below 90 percent of and serious health risks Middle East, regions that enjoyed rapid economic FAO/WHO requirement) (below 80 percent of FAO/WHO requirement) growth during this period. In South Asia and Sub- Saharan Africa, however, the share of the popula- (87 countries, 2.1 billion people) tion with deficient diets increased slightly, and the (87cn countries,12.16billi people) absolute numbers increased markedly. 730 million peoe e) (340 million people) Income growth was the largest single influence on (730 million people)(340millionpeop the dietary improvement between 1970 and 1980. Because average incomes are expected to grow less rapidly, less progress is expected in the 1980s than the 1970s. Dietary Deficits and Food Supplies Many people have too little food to sustain an active, healthy life. But the deficiency in their diets represents only a small fraction of the food supply 3 ow-income countries Figure 1-2. Changes in the Size and Share of Population with Energy-Deficient Diets 51 percent 23 percent in Eighty-seven Developing Countries, 1970-80 (590 million people) (270 million people) Population (millions) 2,500 .: M n ~~~~~~2,000 < _ _ g / ~~~~~~ ~~~~1,500__ _ Middle-income countries 1,000 (57 countries, 0.9 billion people) 14 percent 7 percent 500 X (140 million people) (70 million people) 40 34 7 perc:m eent iS 0 i 1970 1980 1970 1980 Not enough calories for Not enough calories to an active working life prevent stunted growth below 90 percent of and serious health risks FAO/WHO requirement) (below 80 percent of FAO/WHO requirement) El Population with energy-deficient diet l Population without energy-deficient diet Source: World Bank data. Source: World Bank data. 3 in most countries-typically less than 5 percent of markets. The coefficient of variation in prices- the national food supply (or possibly 10 percent one measure of instability and the one used in this under the higher standard). This does not mean, report-ranged from about 20 percent for maize however, that a 5 percent increase in food supplies to 35 percent for rice. This instability was much would eliminate malnutrition. It means merely that greater than that of global production during the in many countries the supply of food is not the same period (Figure 1-3) or of prices in the pre- only obstacle to food security. vious decade. In some of the poorest countries, however, the What then caused this price instability during supply of food does need to be greatly increased to 1968-78? On the supply side, after years of sup- reduce chronic food insecurity. When the amount port programs that created large grain reserves, the of food consumed by malnourished people is large exporters deliberately reduced stocks. This increased through redistributive measures, this action lowered the proportion of stocks to total increased demand must be added to the already consumption and thus caused prices to be more increasing demand caused by population and eco- sensitive to fluctuations in production. In addition, nomic growth. In this situation, the food needs of demand became more unstable as a result of the many nations could not be supplied without signif- sharp rise and subsequent fall in the growth of per icant international assistance for the foreseeable capita income in developed and developing market future. Even with unprecedented growth in agri- economies, volatile exchange rates, the increasing cultural production and in export earnings, these pursuit of policies that stabilized domestic prices countries would still need other sources of finance irrespective of international price fluctuations, and to import the food they need for food security. new policies in centrally planned economies to use imports to offset sharp fluctuations in their food The Costs of Chronic Food Insecurity production. Although price instability in world grain markets is unlikely to be as great as in 1968- The costs of inadequate diets to families and nations 78, it continues because many of the same forces are considerable. Inadequate diets increase vulner- that caused past price upheavals still operate. ability to disease and parasites. They reduce The supply of food in a country is often strongly strength for tasks requiring physical effort. They influenced by variations in domestic production. curtail the benefit from schooling and training pro- With few exceptions, the output of major staple grams. And they result in a general lack of vigor, foods has been unstable in developing countries. alertness, and vitality. These outcomes reduce the Consider the differences in variation of the global productivity of people in the short and long terms, production of major staple foods and of domestic sacrifice output and income, and make it more dif- production by developing countries during 1968- ficult for families and nations to escape the cycle 78. On average, the coefficients of variation in of poverty. eighteen developing countries were 18 percent for wheat production, 14 percent for maize, and 8 per- Transitory Food Insecurity cent for rice; globally, the respective coefficients were 5, 4, and 3 percent. Because of the lack of data on short-term fluctua- Although developing countries might offset tions in food consumption, transitory food inse- shortfalls in domestic output by importing food, curity has to be assessed by looking at variables they are often constrained by a shortage of foreign that typically influence food consumption. The most exchange. Their capability to import food has been important variables are world food prices, domes- highly volatile. During 1968-78, the average coef- tic food prices, and household purchasing power. ficient of variation in export earnings was about 15 percent for developing countries, compared with Unstable World and Domestic Prices about 10 percent for industrial countries. In addi- tion, prior claims for debt repayment and fuel Large fluctuations in the international prices of major imports and tight foreign exchange reserves limited cereals in 1968-78 illustrate the vulnerability of their ability to offset fluctuations in earnings. As a developing countries to changes in world food result, few developing countries have been able to 4 Figure 1-3. World Prices and Production of Wheat, Rice, and Maize, 1960-84 Price Production 1,000 _ 5 , 800 Rice 4 600 3 3aize C~~~~~~~~~~~~~~~~~~~ < 400 0 2 Rice eat 200 1 Maize 0 I _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 1960 1970 1980 1960 1970 1980 Year Year Souirce: World Bank data. stabilize domestic food prices as much as have households, and-sometimes but not always-high industrial countries that import food. Taking the food prices. A decline in the availability of food is producer price of wheat in 1968-78 as an exam- not necessarily a primary cause of famines, as ple, the average coefficient of variation for eleven research on four particularly disastrous famines has industrial countries was 5 percent; for eighteen clearly shown. Indeed, by paying too much atten- developing countries it was 12 percent. tion to changes in food availability, governments and relief organizations have sometimes failed to Household Purchasing Power recognize other causes of famines. As a result, some relief has been misdirected. Incomes and food production vary more within The loss of real income better explains why fam- households and regions than within countries. In ines occur and who is hurt by them. The victims India in 1968-70, for example, aggregate income, typically belong to one or several groups: small- food production, and food consumption were fairly scale farmers or tenants whose crops have failed stable. Yet survey data of the per capita income and who cannot find other employment, landless and spending of 4,000 rural households showed agricultural workers who lose their jobs when agri- considerable instability from year to year. Nearly cultural production declines or who face a rapid half the households had at least one year in which rise in food prices when their wages are stagnant their per capita income fell below 70 percent of or falling, other rural workers who are affected by their three-year average. In about two-fifths of the a drop in real income in the famine regions, and households, total spending and spending on food pastoralists who get most of their food by selling showed similar deviations from the average. their animals. Problems of supply, such as those in wartime, can aggravate a famine. But famines strike Famines even when foodgrain markets are working well. In several famines, local food prices barely rose, and Famines, the worst form of transitory food inse- food was continuously available at those prices. curity, can have several causes: wars, floods, crop But the victims could not buy it, for they did not failures, the loss of purchasing power by groups of have the purchasing power. This underlines the 5 need to focus relief work on those who suffer a fall nomic growth. Others involve some tradeoff. Cost- in real income. effective measures can minimize this tradeoff. Many national governments place high priority on Options for National Policy addressing both chronic and transitory food inse- curity. But they often use measures that work The causes of food insecurity suggest that it can be against economic growth and food security in the tackled in the long term only by raising the real long run. Such measures include persistently over- income of households so that they can afford to valuing currencies, spending large sums on con- acquire enough food. This must be done in two sumer food subsidies, and building costly storage ways: by giving the people who face chronic food facilities to hold excessive stocks of foodgrain. insecurity the opportunity to earn an adequate These measures often result in economic waste, income and by ensuring an adequate food supply draw resources away from more productive activ- through domestic production or imports. ities, slow economic growth, and thus aggravate In some countries in which large numbers of rather than alleviate long-term food insecurity. people suffer from food insecurity and in which agriculture is the main component of the economy Reducing Chronic Food Insecurity (such as many African countries, India, Bangla- desh, and Pakistan), accelerating the growth of Beyond the normal array of measures to accelerate agriculture will contribute to both parts of the economic growth and employment, chronic food solution. In fact, any neglect of agriculture in these insecurity can be addressed by special interven- cases will jeopardize overall economic develop- tions. These interventions include increasing the ment-and with it the possibility of providing food supply (through production or imports), sub- enough work for the growing population and sidizing consumer prices, and targeting income enough food for all households. transfers. Many present macroeconomic adjust- Any policies that raise the incomes of the poor ment programs increase the price of traded goods and increase economic growth should obviously be (including food) and reduce government expendi- given high priority, since they can reduce or even ture (particularly subsidies). This makes it espe- eradicate chronic food insecurity at no added cost cially important that interventions to increase food to the economy. Such policies may involve shifting security be cost-effective. resources from large to small farms, from industry to agriculture, or from capital-intensive activities INCREASING THE FOOD SUPPLY. The fundamental to labor-intensive activities. They may also involve question about policies to increase the national removing price and trade interventions that reduce supply of food (production plus imports) is this: incentives to farmers. do they increase the real income and food con- But economies cannot be expected to grow sumption of the people facing food insecurity? The quickly enough to eliminate the chronic food inse- answer turns on how these policies affect food curity of some groups in the near future, even under prices and nominal incomes. the best of circumstances. Moreover, long-run eco- In discussing this question, it is important to dis- nomic growth is often slowed by widespread tinguish between foods that are internationally chronic food insecurity. People who lack energy traded and foods that are not. The domestic prices are ill-equipped to take advantage of opportunities of traded foods are determined by world prices, for increasing their productivity and output. That independent of the amounts produced domesti- is why policymakers in some countries may want cally (except insofar as the production of any trad- to consider interventions that can speed up food able commodity affects the exchange rate). The security for the groups worst affected without prices of nontraded foods are determined by the waiting for the general effect of long-run growth. amounts produced domestically and the effective The objective of this report is to identify cost- demand. effective ways to increase food security in the short The supply of traded foods can be increased only and medium term. Some measures in some circum by deliberate measures to increase imports or stances are fully compatible with efficient eco- restrict exports. Increasing the supply of a traded 6 food will tend to lower its domestic price and Policies directed toward national self-sufficiency decrease its domestic production. The effect of the in food do not necessarily reduce chronic food change in the price on the real income of poor insecurity. Self-sufficiency may be associated with people hinges on whether such people are net buy- a high or low level of food security, depending on ers or net sellers of the food. For net buyers, a fall its effect on food prices, on the incomes of the in the price is unambiguously beneficial: they can poor, and on the availability of food to disadvan- afford more of the food and are encouraged to buy taged groups. This conclusion is supported by a more because it is cheaper. These benefits can be statistical analysis of the relation between self-suf- considerable, since the poor typically spend one- ficiency in cereals and per capita food consumption half to three-quarters of their income on food and in fifty countries. In general, the choice between only slightly less on basic staples. For net sellers of getting food through added domestic production food, however, a fall in the price of food means a or through imports should depend on the reliabil- fall in their real income. If the net sellers of food ity of alternative food supplies, the comparative are low-income farmers, they will suffer more food advantage of the country in international trade, insecurity. If the poor are primarily subsistence and the cost of foreign exchange. farmers-that is, they are neither net buyers nor net sellers-an increase in the supply of a traded SUBSIDIZING FOOD PRICES TO CONSUMERS. A gen- food will have little immediate effect on them. eral price subsidy that reduces the consumer price An increase in the domestic production of a of staple foods while maintaining higher producer traded food will not affect its price. The increases prices can sustain the incentives to produce these in production will be offset by fewer imports or foods. Such subsidies usually are feasible only if more exports. Consumers do not gain. The sole the foods are consumed in processed form and can beneficiaries will be the farmers producing more be centrally processed (such as flour, bread, or tor- food, who may or may not be poor and food tillas). Otherwise, food will be bought at the low insecure. subsidized price and resold at the higher producer The supply of nontraded foods obviously can be price. increased only by increasing domestic production. Because general price subsidies transfer income An increase in the production of a nontraded food to everyone, however, their fiscal cost is high. will therefore increase its supply and reduce its Although food subsidies usually tend to benefit the price. The net buyers of the food will definitely poor more than the rich in a relative sense, the gain. Net sellers of the food, however, might gain income transfers to upper-income groups are or lose, depending on how much the lower price is sometimes greater than those to disadvantaged offset by increased sales and lower production groups. Therefore, food price subsidies should be costs. restricted to isolated areas in which the poor are Thus, the choice of food supply policies should concentrated and to foods that figure heavily in be guided by the characteristics of each commodity their spending. Moreover, the price should not be and the circumstances of each country. If the poor so low that the processed food is used to feed live- are primarily net buyers of food-as in countries stock or distill alcohol-or is transferred to unin- with widespread urban poverty or a high propor- tended consumers. tion of rural landless-increasing national food supplies and lowering food prices is an effective TRANSFERRING INCOME IN CASH AND IN policy option. If the poor are primarily net sellers KIND. Transfer payments in cash or in kind tend to of food-as in Bangladesh and many Sub-Saharan be the most efficient way of increasing the real African countries-food security can be increased incomes of the poor and of giving them the means by reducing the supply of some traded foods and to increase their food consumption. A well-known raising their prices and by substituting domestic way to transfer income is to ration food to a target production for imports. Subsistence farmers will group at below-market prices through "fair-price be unaffected. Whatever supply policy is chosen to shops" or at no charge through health centers. Such improve food security is likely to leave some groups interventions aid mostly urban dwellers and house- of poor people worse off as a result. holds that normally buy their food or have easy 7 access to such centers. But income could also be Figure 1-4. The Elements of Cost-Effectiveness transferred to rural households-either in cash or in subsidized rations of consumption goods or farm inputs. The difficulty with this form of intervention lies in avoiding leakage of the transfer payments to Efficiency losses unintended beneficiaries. Efforts to target the ben- efits more sharply to the poor will reduce the fiscal Economic cost of the transfers but will raise-often steeply costs { _ _ _ _ _ _ -the administrative costs and skills required. Some governments have tried to raise the incomes Budget of the poorest people through public employment dcosts programs. These programs seem attractive because only the poorest and the mnost frequently unem- ployed are targeted for jobs. But such programs Income are sometimes inefficient at transferring income to transfers { the poor for two reasons. First, they may convey little additional income to workers if the wage offered is only slightly higher than the opportunity wage (whatever they could obtain elsewhere) and if the workers have extra travel costs. Second, the projects may create assets whose value is much cost of buying and selling the food. Calculating the lower than the cost of producing them. Together, efficiency loss of a policy that depresses food prices these two factors can make it expensive to transfer paid to producers is more difficult. It is the cost of income to the intended beneficiaries. diverting resources from the production of food to Employment programs might be more cost- the production of commodities of lower value. In effective if they are used during periods of seasonal each case, calculations are needed to determine food insecurity, when the opportunity wage is very whether the delivery costs are higher for targeted low. It may often be more cost-effective to subsi- or for marketwide interventions and how that cost dize private employment or inputs to poor farmers compares with the efficiency loss from a policy that than to initiate public employment schemes. Tax distorts producer prices. Although the delivery costs concessions to an industry that uses labor-intensive come directly from the government budget, the technologies to produce goods with an elastic efficiency losses usually are not visible. demand can go far to improve employment oppor- Income transfers are not true economic costs. tunities and thus to increase incomes and food But when directly paid by the government (rather security. A range of options for interventions not than shifted, say, to farmers through lower input necessarily related directly to the food sector is prices), they comprise a large part of the budget usually available. cost. Budget costs alone do not measure the size of transfers or the economic costs. Yet they are cru- COMPARING COST-EFFECTIVENESS. When analyz- cial factors in choosing policy packages. Reducing ing the cost-effectiveness of interventions, it is nec- the budgetary burden of food security interven- essary to distinguish between economic costs, bud- tions can be done in two ways. One is to reduce get costs, and income transfers (Figure 1-4). The the income transfer by reducing the number of economic cost consists primarily of the value of beneficiaries or the size of the transfer. The second resources lost in transferring the income. These is to shift the cost of the transfer to food producers. involve either delivery costs or efficiency losses. The problem with the second way is that it results Calculating the delivery costs for explicit income in efficiency losses. transfers is straightforward. In a food ration pro- The choice of interventions needs to be based on gram, it is the cost of identifying and certifying the a balanced concern for budgetary and economic eligible beneficiaries and delivering the subsidized costs, the administrative and political feasibility of rations. In a program of marketwide subsidies on different interventions, and the expected benefits consumer food prices, it is the excess marketing from the interventions. For instance, targeted 8 income transfers are surely less burdensome for plex policy package is needed. It might combine governments' budgets than marketwide food sub- employment programs for the landless, subsidized sidies, mainly because they do not transfer income inputs and infrastructure for subsistence farmers, to the whole population. But they may be a and targeted food programs for the urban poor. If bureaucratic nightmare and elicit little political adequate targeting is not feasible or if budgetary support. Similarly, excessive concern with the costs are too high, the prices of some staple foods budgetary implications of interventions may lead might be lowered through import or production governments to favor supply interventions that shift subsidies. But the costs to the budget and to the burden to food producers. But the invisible resource efficiency should be kept in mind. economic costs of such interventions can be high. In summary, the choice of measures to address chronic food insecurity involves considerations of CHOOSING POLICY PACKAGES. Cost-effectiveness is the tradeoffs between: not the only criterion for selecting appropriate * Sharply targeted interventions, which have interventions. Attention must also be given to the high delivery costs but relatively lower budgetary size and makeup of the target population, the gov- costs ernment's ability to administer and finance inter- * General consumer subsidies, which are easier ventions from its budget, and the political feasibil- to implement and do not distort producer prices ity of different interventions. but have high budgetary costs Most countries have a variety of target groups * Import subsidies, which are easy to implement and thus need a package of measures to improve and have low budgetary costs, but distort producer food security. The appropriateness of the package prices and therefore may have large efficiency depends on which groups are facing food insecur- losses. ity in a country. In several countries in Sub-Saharan The appropriate mix has to be determined for each Africa, those facing food insecurity are largely food country because the tradeoffs differ in different cir- sellers, but they also include some urban poor and cumstances. For marketwide interventions, there is rural landless. In these cases, an effective policy always the danger that political pressures for sub- package to increase food security is likely to focus sidies will increase aftcr they are introduced. The on raising producer prices (especially when they government then gets locked into burgeoning bud- are below border prices), on providing subsidies get allocations that are difficult to control or phase for the urban poor, and possibly on introducing out. employment programs for the landless. The choice of compensatory policies to protect the net buyers Reducing Transitory Food Insecurity of food will depend on the budgetary cost and administrative requirements of these different Governments that do not prevent transitory food options. insecurity and food price instability may run sig- If, however, most of those facing food insecurity nificant risks. They may prolong human suffering, are urban poor, as in some Latin American coun- lose the human energies essential for development, tries, measures should be directed mainly at low- and disrupt the political fabric of their society. They ering food prices for this target group. Consumer can reduce such risks by implementing policies that subsidy programs that do not lower prices to pro- promote stability in the domestic supply and price ducers, such as those that are targeted or that sub- of staple foods and that provide vulnerable groups sidize centrally processed foods, can be cost-effec- with the means to buy enough food as the need tive if they have low budgetary and administrative arises. These policies-like those for reducing costs. Otherwise, import prices on selected foods chronic food insecurity-must be determined by heavily consumed by the poor can be lowered. This the expected costs and benefits. would reduce the income of the domestic produc- ers, however, unless they are given compensatory STABILIZING SUPPLIES AND PRICES. In most coun- input subsidies. tries, the surest and probably the cheapest way to If those facing food insecurity include many rural stabilize prices is through international trade. landless, subsistence farmers, and urban poor-as Imports or exports can offset instability in domes- is typical in India and Bangladesh--a more com- tic production almost automatically. By using var- 9 iable levies on externally traded foods, for exam- ASSISTING VULNERABLE GROUPS DIRECTLY. Stable ple, domestic food prices can be insulated from food prices and efficient domestic marketing are changing international prices. But such policies can not always enough to prevent transitory food inse- wreak havoc with the budget and the balance of curity and certainly cannot prevent famines. Even payments unless governments hold larger reserves if markets are working perfectly, the most severely of foreign exchange and rely more on food aid or affected groups can still lack the purchasing power international insurance schemes for financing var- to buy food. Such groups include the rural landless iable food imports, such as the expanded Compen- whose employment and wages are severely satory Financing Facility of the International Mon- depressed, the small-scale farmer or herder whose etary Fund. marketable surplus has been destroyed or who faces Governments trying to stabilize food supplies and adverse terms of trade, and the artisan or urban prices often do so in costly and ineffective ways. worker whose opportunities for work have col- They use quantitative controls on imports and lapsed. These groups must be identified quickly and exports of foods-controls that require substan- provided with income in cash or in kind. If a coun- tially more information on supply and demand than try can administer cost-effective work programs, public marketing agencies typically have. Even these groups must be given temporary employ- when this information is available, the agencies sel- ment. And if resources are severely limited, special dom have the managerial resources or the political programs should at least preserve the nutrition of support to act promptly and effectively. As a result, children under five and of pregnant and lactating government interventions often aggravate rather women. than reduce the instability of supply and prices. Some governments keep excessive buffer stocks. The Role of External Assistance Large stocks are seldom cost-effective because of high storage losses, low capacity utilization of stor- So far, much of the foreign assistance for food age facilities, and high interest charges on capital security has sought to accelerate agricultural devel- tied up in inventory. Countries with access to for- opment and increase food production. These are eign exchange usually find it cheaper to stabilize important aspects of the problem, especially when prices by varying their exports and imports rather they affect the real incomes of vulnerable con- than by using buffer stocks, even if world market sumers and producers. But there has been only prices are unstable. Storage costs are usually far modest progress in diminishing worldwide food greater than the cost of borrowing or of carrying insecurity, partly because of the widely held mis- extra reserves so as to be able to import more when perception that food shortages are the root of the the occasion demands. When food supplies cannot problem. The disturbing fact is that food security be stabilized through trade, investing in measures problems have become more serious in many coun- to stabilize production, such as irrigation or pest tries. Food insecurity remains even in countries that control, or to improve the efficiency of domestic have high per capita food production. marketing is sometimes less costly than relying on The international community, in supporting food buffer stocks. security, should take into account the four most important conclusions of this report: PREPARING FOR PRECIPITATE DECLINES IN SUP- * The lack of food security is a lack of purchas- PLY. Countries might stabilize supplies by adjust- ing power of people and nations. Thus, there is a ing their food trade but might lack the capability strong convergence between the objectives of alle- to Inake the food available where and when it is viating poverty and ensuring food security. needed. To overcome this, countries should invest * Food security does not necessarily come from in speeding their response time by means such as achieving food self-sufficiency in a country, nor better early warning systems, faster orders for from a rapid increase in food production. additional shipments, and speedier internal move- * Food security in the long run is a matter of ments of supplies. If response times cannot be achieving economic growth and alleviating pov- improved, small emergency food stocks may be erty. But food security in the shorter run means needed in strategic places to bridge the gap until achieving a redistribution of purchasing power and added deliveries can be made. resources. By choosing redistributive policies that 10 are also cost-effective, governments can do much nerable, particularly lactating women and children to reduce the costs of improving the food security under five. of their people. * Projects should continue to emphasize invest- * Transitory food insecurity linked to fluctua- ments that benefit the poorest people. This includes, tions in domestic harvests, international prices, and where appropriate, financing investments that foreign exchange earnings can best be reduced increase the supply and reduce the price of basic through measures that facilitate trade and provide staple foods. External finance should also continue income relief to afflicted people. to be allocated for cost-effective programs to International donors can help nations apply these improve the nutrition of vulnerable groups through conclusions to their food security strategies in three income transfers, nutrition education, drinking ways: by helping to identify and formulate appro- water supply, preventive health services, and means priate policies to alleviate food insecurity, by pro- for family planning. viding finance to support these policies, and by * External finance should help countries improve improving the external trading environment. their capacity to handle sharp, sporadic increases in food imports and intraregional food transfers Policy Advice that may be needed from time to time to ensure food security. It should also help the countries to National policies have the greatest influence on the establish small, strategically located emergency food security of countries. The main objective of stocks to tide them over while awaiting additional external advice should be to help countries pursue imports. cost-effective food security interventions, which * In climatically unstable regions, external take into account budgetary, political, and admin- finance should support projects that include com- istrative constraints. ponents-such as crop diversification, public The international community should help coun- employment, and transport facilities-to help sus- tries to assess food security problems and to develop tam minimally adequate food consumption in years the analytical capability to formulate remedial pol- of adverse weather conditions. icies. The assessments should always clearly differ- * Donors of food aid should coordinate their entiate between chronic and transitory food inse- curity and should identify the size and characteris- effove the development istance. tics of the vulnerable groups and the sources of Improve the effectiveness of external assistance. instability in food production, prices, and incomes. * To improve the cost-effectiveness of food aid provided directly to poor households, foods should have a high value to the recipients relative to the delivered costs. Exchanging donated food for cash Efforts to support food security in line with the and buying local foods may often be more advan- remedies advocated in this report should proceed tageous to recipients. in three directions. The first is to continue to * The cost-effectiveness of famine relief should emphasize lending operations that benefit the be improved by detecting emergencies earlier, poorest people. The second is to increase the use reducing the response time, and providing aid in of trade finance and other international financing cash as well as food. arrangements to provide resources for alleviating * Food aid should be administered so that it can transitory food insecurity. The third is to integrate effectively alleviate transitory food insecurity of all food aid with financial aid. kinds, including that caused by high international Several recommendations for providing external food prices. Food aid budgets in 1985 were quite financing can be made. responsive to the crisis in Africa. In general, how- * Lending operations that support policy ever, food aid budgets tend to be fixed from one adjustments needed for faster economic growth year to the next; as a result, when prices rise, less should consider the implications of the reforms for food can be bought. Even worse, food aid budgets food security. When necessary, adjustment pro- are often curtailed when prospects for commercial grams should include cost-effective programs to exports improve. Multiyear commitments by food safeguard the food security of the people most vul- aid donors, with flexible drawings against those 11 commitments in accord with the needs of recipient assistance to help developing countries use the same countries, are required. tools that developed countries use to manage trade risks. International Trade The terms of trade for commodities greatly affect Concluding Comments the real income of vulnerable groups, primarily through effects on food prices and on the earnings Many countries have intervened in their econo- of these groups. International concern for food mies in the name of food security. In some cases, of tesegrous. ntenatinalconern or ood however, these interventions have incurred high security should therefore be translated into a con- cern for how the policies of developed countries costs in terms of sacrificed economic growth. Poli- limit international trading opportunities. Multilat- cymakers need to be sensitive to these costs and to eral and bilateral actions to reduce trade restric- the difficulties of reducing or eliminating interven- tions on the exports of developing countries would tions once they are in place. Three key issues need help most, since the poor often work in the labor- to be kept in mind: intensive production of export goods. Moreover, * Interventions to improve food security have trade restrictions cause international markets to be both costs and benefits more unstable than they would be otherwise, and * Some forms of intervention are more cost- thus contribute to transitory food insecurity. effective than others, and these cost-effective meas- Actions that stabilize food prices and major ures should be chosen exchange rates would help as well. The interna- * Costs and benefits should be calculated in the tional community could also provide technical context of each individual economy. 12 A Problem of Supply or of Purchasing Power? Food is abundant worldwide, and nations with the whether it is the outcome of inadequate food sup- means to buy it have no problem acquiring all they plies or inadequate purchasing power. Evidence on need. But anxiety about food supplies remains high food supplies and prices in the 1970s shows that in many countries-and this worry has led some there was enough food for all the world's people. governments to devote substantial resources to Projections to 2000 indicate that world food pro- increasing food production. Preoccupation with duction is likely to keep pace with effective global food production is misguided, however, when it demand, perhaps even at a lower price (although takes priority over more immediate concerns. One debt-ridden countries may find that domestic food such concern is that many poor countries-and prices will go up because of scarce foreign hundreds of millions of poor people-cannot get exchange). Why, then, do both chronic and tran- even a modest share of the world's abundant food sitory food insecurity persist? An analysis of the supply. In response to this concern, this report causes of food insecurity concludes that the main focuses on the food insecurity that arises from the cause is not lack of supply or even high prices. The inadequate purchasing power of households and main cause is the weak purchasing power of some nations. This emphasis does not mean that pro- households and nations. This conclusion, along moting food production and improving food trade with the sharp distinction between chronic and are unimportant; it means that an adequate income transitory food insecurity, sets the stage for the is essential for ensuring food security. analysis of national and international policy issues The term "food security" came into use about in later chapters. ten years ago to describe a broad range of devel- opment issues. It is perhaps natural to put new Recent Levels of Supplies and Prices labels on old problems that defy easy solutions, but the price of combining loosely related subjects The growth in global grain production has more under a new name can be high. For example, rede- than kept pace with the growth in population fining problems of agricultural development and (Table 2-1). In fact, grain output increased faster general development as problems of food security than population during the 1970s for developing does little to increase understanding of either issue. countries as a group, but not in Eastern Europe, This chapter assesses whether food insecurity is the Middle East, or Sub-Saharan Africa. Data on an issue of food supply or an issue of poverty- grains are used to indicate the total supply of food 13 Table 2-1. Grain Production and Population Growth, by Country Group, 1970-82 Average annual percentage change Cereal Cereal production Country group or region production Population per capita World 2.3 1.8 0.5 Industrial market economies 2.3 0.7 1.6 East European nonmarket economies 0.6 0.8 - 0.2 Developing economies 3.0 2.1 0.9 East Africa 0.8 3.0 - 2.2 West Africa 1.9 2.7 -0.8 East Asia and Pacific 3.5 1.7 1.8 South Asia 2.7 2.4 0.3 Middle East and North Africa 1.7 2.9 - 1.2 Latin America and the Caribbean 3.2 2.4 0.8 Note: The term "grain" includes wheat, rice, maize, rye, sorghum, millet, barlev, oats, and mixed grains. All growth rates in all tables have been computed using the least-squares method. Source: World Bank calculations based on USDA data. because data for other foods are less reliable and exporting countries. Most technological improve- because grain is the most commonly consumed food ments have been focused on increasing yields and in developing countries. In Sub-Saharan Africa, account, during the past twenty years, for about however, root crops are often just as important as 65 percent of the increase in global rice produc- grains in people's diets. Even when production tion, 80 percent of that in wheat, and 90 percent estimates for root crops are included with grains, of that in coarse grains. Investment in infrastruc- per capita food output is still shown to have ture (such as irrigation and transport facilities), declined (see Table B-1 in Annex B). remunerative farmgate prices for domestic output, Trends in per capita food production are not, greater access to credit, and improvements in agri- however, a good measure of food scarcity in coun- cultural extension all have made it possible to adopt tries. Because the global and domestic food mar- Green Revolution technology and to boost produc- kets are increasingly integrated, the trend in world tion in many parts of the world. grain prices emerges as a better measure for judg- The revolution in technology has not, however, ing the availalility of food. Together with trends affected all crops and regions equally. (Annex B in the scarcity of foreign exchange, it determines shows the differences in yield increases of the five the true costs of food to nations and consumers. most important cereals between 1960 and 1982.) For decades, the real dollar prices for grains in For example, the remarkable increases in India, the world market-defined as crop-year average Southeast Asia, and parts of Latin America have prices deflated by the 1980 U.S. consumer price not materialized in Africa. Nevertheless, because index-have declined (Figure 2-1). Prices did rise of the greater integration of international com- sharply above the long-run trend in the early 1970s, modity markets in recent years, the technological but, in real terms, wheat prices now are about 15 advances in the major grain-producing countries percent below the level of the early 1970s and 20 have kept world grain prices down. These coun- percent below the level of the 1960s. Maize prices tries can now adjust supplies quickly to shifts in declined even more. Although rice prices rose in world demand. The shifts in supply have tended to the early 1970s, they also have been falling during be larger than shifts in demand, which has resulted the past six years and are projected to continue to in the trend of declining prices. do so (World Bank 1984). Of the many factors This declining trend in real prices also reflects that influence these price trends, two stand out: rapid recent sluggish growth in the demand for cereal- improvement in farm production technology and a function of the slower growth of population and slower growth in effective demand. per capita incomes. In most countries, the gross Technological improvements have reduced pro- domestic product (GDP) grew much slower in the duction costs and increased the output of world early 1980s than in the 1960s and 1970s (Table 2- agriculture, particularly that of the largest cereal- 2). Annual GDP growth in the middle-income devel- 14 Figure 2-1. Trends in World Prices of Wheat, percent in 1982-84. Similarly, the growth of GDP Rice, and Maize, 1960-84 in developed countries dropped from 2.8 percent a (1980 constant dollars) year during the middle and late 1970s to 1 percent _________________________________________ in 1981 and 1982. Simultaneously, purchases by the developed countries of wheat and coarse grains Wheat from world markets stagnated and in some cases 500 - fell. Dollar prices in world markets may reliably reflect 400- supply and demand in international markets, but c o they are not an accurate measure of the price of 'L; 300- * imported cereals in domestic currency once adjust- 200 _ 0.4percent t ments are made for exchange rates, domestic infla- 0 ' _ _ _ _ _ _ tion, and transport costs. International grain prices 100 _ have fallen steadily, but the price of imported grain in domestic currencies after adjusting for the 0 I l E exchange rate has not always moved in the same 1960 1970 1980 direction. For example, between 1970 and 1983 Year the real rupee price for Sri Lanka's wheat and rice Rice imports grew more than 7 percent a year while real 1,000 dollar prices fell about 1.5 pcrcent a year (see Annex B, Table B-2). 800 - 600- Future Levels of Supplies and Prices X 400 _ 0 Og o> 400 - 1.0 percent * * Most projections show that global food produc- 200 - * * 1 tion will continue to increase fast enough to meet the effective demand generated by growth in pop- 0 I l ulation and incomes. This added production will 1960 1970 1980 come primarily from better use of resources and improved technology, not from cultivation of more Maize land. 500- Biogenetic research is under way to produce new 400 - crop varieties that require fewer inputs and that tolerate pests, drought, and disease. Years of work in developing hardier and more productive strains 300 of subsistence tropical food crops-yams, beans, X 200 - * * millet, sorghum, and cassava-are also beginning 200 -0.6percent _ __* ~ > to pay off. Current research on the placement and 100 _** timing of fertilizer applications will increase the efficiency of inputs. Investments in lower-cost, 0 higher-efficiency devices for water control and 1960 1970 1980 management (such as small tubewells and irriga- Year tion-ditch lining) as well as investments in trans- port, marketing, and support services are also Source: World Bank. expected to improve agricultural productivity. As a result, the cost of increasing food production throughout the world should decline, and real oping countries, which previously had imported world food prices should continue to fall. large amounts of grain for people and livestock, Africa's problems and remedies are more com- fell from almost 6 percent in 1973-80 to about 1 plex and controversial than those of other regions, 15 Table 2-2. Gross Domestic Product, 1980, and Growth Rates, 1965-84, by Country Group 1980 GDP per Average annual percentage change in GDP capita Country group (dollars) 1965-73 1973-80 1981 1982 1983' 1984b Developing countries 670 6.6 5.5 3.3 1.9 2.0 4.1 Low-income countries 260 5.5 4.9 4.0 5.0 7.2 6.6 Middle-income oil importers 1,690 7.0 5.6 2.0 0.8 0.7 3.3 Middle-income oil exporters 1,270 7.1 5.8 4.6 0.9 -1.0 2.7 High-income oil exporters 10,650 9.2 7.7 0.1 -1.7 -7.0 0.6 Industrial market economies 10,420 4.7 2.8 1.4 - 0.3 2.3 4.8 a. Estimated. b. Projected. Source: World Bank (1985), p. 149. including Asia in the 1960s. Yet, because of prog- persist. The needed policy changes, institutional ress in agricultural research, technological con- adjustments, and technological developments may straints can be addressed with more understanding not occur or have the expected effect. In addition, than before. There appears to be a great potential other complicating factors cannot be ignored. For for improving yields in Sub-Saharan Africa, example, a major economic disturbance such as although fragile soils and a variable climate in many that of the 1930s, with serious distortions of inter- areas make it difficult to sustain increased food national trade and associated economic collapse, production, whether under extensive or intensive could have a serious effect on both the supply of farming. The costs of increasing food production food and the ability to acquire it. there will be high unless technology improves con- Another uncertainty is the future price of foreign siderably. Achieving such improvement will require exchange for individual countries. If it increases, many years of research. domestic prices could rise and thus reduce the food A few analysts acknowledge that future world available to households. Constrained export supplies of food will meet effective demand but opportunities, declining terms of trade, and high contend that the relative price of food will rise external debt repayments could combine to make sharply (Brown 1984). They assume that the dete- foreign exchange scarcer in developing countries rioration of the agricultural resource base will out- -and increase the cost of food imports in domes- pace technical change. No doubt, the agricultural tic currency, even if world prices continue to fall. base has deteriorated in many parts of the world, Countries short of foreign exchange will then find and, if unchecked, the deterioration will increase it advantageous to produce fewer nontradables and the cost of agricultural production in these areas. more tradables, including many foods. Deforestation and soil erosion already are cata- strophic in many regions. In the Sahel, for exam- Chronic Food Insecurity ple, an estimated 1 percent of the natural forest cover is lost each year. Despite these losses, pessi- Just as national per capita income is a poor indi- mism about world food prices is not warranted. cator of the prevalence of poverty, so per capita Technological developments can help farmers food consumption is a poor indicator of the num- adjust their farming systems to fragile environ- ber of people consuming inadequate amounts of ments. To be sure, changes in institutions, in price food. With this in mind, two questions are and credit policies, and in the distribution of land- addressed here: how many people in the develop- holdings will be needed in many countries and ing world do not get enough food to eat, and is regions to support technological advances. Many their number and share of the total declining or such changes are already taking place and provide increasing? solid ground for optimism about the continuing ability of global food supplies to meet demand and What Is an Adequate Diet? hold prices down. Of course, some uncertainties about supplies The amount of food (or energy) a person needs is 16 not easy to determine: social activities, employ- tors that impairs their performance, clearly pre- ment, genetically determined differences in the effi- vents them from leading productive lives. ciency of food conversion, environmental factors, and individual preferences all come into play. Peo- The Prevalence of Deficient Diets ple can survive with barely minimal diets. If the physical effort needed to earn an income is consid- Estimates of the prevalence of energy-deficient diets ered, however, the diets of many people are inade- are deduced from data on the energy content of quate. And if the energy to grow and take part in average diets in eighty-seven developing countries social life and economic development is also taken in 1980 and data on income distribution patterns into account, even more people have inadequate in thirty-five countries (Table 2-3). The estimates diets (Beaton 1983). Even if there were a consensus show, for two standards of energy requirements, about what performance criterion should be used, the share and number of people with energy-defi- the adequacy of diets is not easily determined. Only cient diets in various regions-the people who face if both the energy intake and the energy require- chronic food insecurity. If the energy standard ments were known-person by person-could adopted is merely enough calories to prevent researchers estimate more precisely the prevalence stunted growth and serious health risks, an esti- of energy-deficient diets. Such data have never been mated 340 million people, or a sixth of the people compiled. in eighty-seven developing countries, had energy- Concern for chronic food insecurity must there- deficient diets. If the standard is enough calories fore focus on the adequacy of average food con- for an active working life, some 730 million peo- sumption for specific socioeconomic groups. It is ple, or a third of the people in the same countries, possible to estimate the energy deficiency for these lived with dietary deficits. Most of these people- groups from the relation between food consump- four out of every five-were in low-income coun- tion and income without knowing how intake and tries. This higher figure is the better guide to requirements vary within the group. For instance, the harm that inadequate diets impose on the poorest groups consume less food, contract development. more diseases, have higher mortality rates, and are How did chronic food insecurity change during more likely to be physically stunted or exert less the 1970s? The choice of a standard is not an energy in their daily tasks than the rich. Inadequate important issue in this case. Moreover, imprecise food consumption, although only one of many fac- estimates of the level of chronic food insecurity do Table 2-3. Prevalence of Energy-Deficient Diets in Eighty-seven Developing Countries, 1980 Not enough calories to prevent Not enough calories for an active stunted growth and serious working life (below 90 percent health risks (below 80 percent of FAO/WHO requirement)' of FAO/WHO requirement] Share in Share in population Population population Population Country group or region, (percent) (millions) (percent) (millions) Developing countries (87) 34 730 16 340 Low-income (30)W 51 590 23 270 Middle-income (57) 14 140 7 70 Sub-Saharan Africa (37) 44 150 25 90 East Asia and Pacific (8) 14 40 7 20 South Asia (7) 50 470 21 200 Middle East and North Africa (11) 10 20 4 10 Latin America and the Caribbean (24) 13 50 6 20 a. The eighty-seven countries had 92 percent of the population in developing countries in 1980, excluding China. See Annex A, Table A-1 for regional classification of countries. Numbers in parentheses are the number of countries in the sample. b. See Annex A for an explanation of FAO/WHO requirements. Intake at this standard is sufficient for a person to function at full capacity in all daily activities. c. Intake at this standard is sufficient to prevent high health risks and growth retardation in children. d. The low-income countries had a per capita income below $400 in 1983: the middle-income countries had a per capita income above $400 in 1983. Source: World Bank estimates. 17 Table 2-4. Changes in the Prevalence of Energy-Deficient Diets in Eighty-seven Developing Countries, 1970-80 Not enough calories for an active Not enough calories to prevent stunted working life (below 90 percent growth and serious health risks (below 80 of FAO/WHO requirement) percent of FAO/WHO requirement) Change in share Percentage change in Change in share Percentage change in Country group or region of population number of people of population number of people Developing countries (87) - 0.06 + 10 -0.02 + 14 Low-income (30) + 0.04 +41 + 0.03 +54 Middle-income (57) - 0.18 -43 - 0.09 - 44 Sub-Saharan Africa (37) + 0.01 +30 + 0.04 +49 East Asia and Pacific (8) -0.27 -57 - 0.14 - 57 South Asia (7) + 0.03 +38 + 0.02 +47 Middle East and North Africa (11) - 0.25 - 62 - 0.14 - 68 Latin America and the Caribbean (24) - 0.07 - 15 - 0.04 - 21 Note: See the footnotes to Table 2-3- Source: World Bank estimates. not necessarily invalidate estimates of the changes diets under the higher standard in 1980-90 mil- in food insecurity. lion more under the lower standard. The share of people with deficient diets in the Another way to infer changes in chronic food eighty-seven developing countries declined between insecurity is to observe changes in the energy con- 1970 and 1980 under both standards of adequacy tent of national average diets-again assuming that (Table 2-4). The estimates rest on the rather opti- income distribution remained constant. About 770 mistic assumption that income distribution did not million people live in countries in which the energy change during the decade; this probably overstates content of the national food consumption was low the improvement, as explained in Box 2-1. The in 1970 and declined further by 1980 (Figure 2-2). sharpest declines were in East Asia and the Middle East, where per capita incomes rose rapidly. The Prognosis for the Future share declined much less in low-income countries than in middle-income countries. In South Asia and Chronic food insecurity, if measured by the share Sub-Saharan Africa, however, the share of people of people, is likely to continue to diminish, but if with energy-deficient diets increased somewhat. it is measured by the number of people, it is likely The total number of people with energy-deficient to continue to increase. Some evidence also sug- diets has increased under both standards (see Fig- gests that the progress achieved in the 1970s will ure 1-2). Compared with 1970, 170 million more not be repeated in the 1980s. For instance, the people in the low-income countries had deficient energy content of the average diet in all countries Box 2-1. Income Distribution and Diet The data necessary to directly measure changes in need more energy in their daily diet. In Brazil, for income distribution generally are not available. The instance, per capita income more than doubled while literature suggests, however, that increases in in- per capita calorie consumption changed little dur- come tend not to be distributed equally during the ing one decade. To the extent that income distribu- early stages of development (Kuznets 1975; Ahlu- tion becomes more unequal, the estimated size of walia 1976; and Chenery and Syrquin 1975). A the population with energy-deficient diets in 1980, comparison of rates of change in the energy content shown in Tables 2-3 and 2-4, has been under- of national diets and in per capita income also sug- estimated. Actual changes may have been less gests that the growth of income during the 1970s favorable. was not fully shared by low-income people, who U 18 Figure 2-2. Changes in the Population and in the Energy Content of the Food Consumed in Eighty-seven Developing Countries, 1970-80 Initial energy in the Population (millions) diet as a percentage of the requirement, 1970 High 7 113 g M r7 38 0> 105 percent) Medium . ( 158 44 63 (95-105 percent) a y rb 1 ieclining consumption Low growth Moderate growth High growth (< 0 percent) (0-0 .5 percent) (e.r-1 percent) i >t percent) Annual rate of change in per capita consumption, 1970-80 Source: Data from FAO (1i977, 1980b). during the 1970s, weighted for the population, National Food Deficits grew 0.4 percent a year-about 100 calories dur- ing the decade. This growth was associated with Although many people face food insecurity, the growth in per capita income of 2.6 percent a year. aggregate energy lacking in their diets is only a But the growth of per capita income in the 1980s small part of the energy in the food supply of most is projected to be less than in the 1970s, and per nations (Table 2-5). Even under the higher calorie capita income is the most important determinant standard, the aggregate energy deficit in most of growth in the energy content of national diets. countries is far less than 10 percent of the food So the decline in the share of people with energy- supply. For a sample of thirty-five countries, the deficient diets will probably be less in the 1980s deficit is 3.5 percent of supply under the higher than in the 1970s, and, if the population continues standard and 0.9 percent under the lower stan- to grow at present rates, the number of people dard. Increasing the food supply by 3.5 percent facing chronic food insecurity will increase further. would not eliminate the energy deficiency, since it Even a small increase in the energy content of would not necessarily improve the incomes and the average national diet-say, 2 percent during purchasing power of the poor. For most develop- the current decade-could reduce the share of ing countries, the supply of food-even the cost of people with energy-deficient diets by 10 percent. that food-is not the greatest barrier to achiev- But this would occur only if the poor shared pro- ing national food security. In many countries, the portionately in the additional food consumption. national food supply (domestic production plus The poor, however, tend to bear the brunt of slower imports) is now sufficient or could easily expand or negative growth in per capita income. More- to provide the entire population with enough energy over, food price subsidies and other government if it were distributed more evenly. Nonetheless some programs to benefit the poor tend to be sharply low-income countries, particularly in Africa, do reduced during hard times. have a national supply problem. Even assuming 19 Table 2-5. Energy Deficit of Energy-Deficient Diets as a Percentage of Energy in the Total Food Supply in Selected Low-Income Countries, 1980 (percent) Not enough calories for an active Not enough calories to prevent stunted working life (below 90 percent growth and serious health risks (below 80 Region and country of FAO/WHO requirement) percent of FAOIWHO requirement) Sub-Saharan Africa Kenya 8.9 3.3 Senegal 2.5 0.9 Sudan 2.5 0.9 Tanzania 7.8 2.8 East Asia and Pacific Indonesia 2.0 0.5 Philippines 1.9 0.6 South Asia India 5.6 1.4 Pakistan 1.2 0.4 Middle East and North Africa Iraq 2.3 0.9 Tunisia 1.0 0.4 Turkey 0.7 0.3 Latin America and the Caribbean Brazil 1.3 0.4 Colombia 0.9 0.3 Dominica Republic 3.5 0.9 Ecuador 8.8 3.1 Sample of thirty-five countriest 3.5 0.9 a. Countries listed in Annex A, Table A-1. Source: World Bank estimates. rapid growth in domestic production, large (Table 2-6). The food imports required to improve increases in food imports will be required to meet their food security by 1990 have been projected on the needs in those countries. the following assumptions: Given the present food supply in six selected low- * The energy content of the national food sup- income countries, energy consumption would be ply is set at 100 percent of the FAO/WHO require- insufficient to satisfy requirements even if the sup- ment to ensure an adequate diet for all groups in plies were evenly distributed among the population the population. This takes into account the wide Table 2-6. Volume of Cereal Imports and Their Value as a Share of Export Earnings for Selected Low-Income Countries Cost of cereal imports as a Food production Cereal imports proportion of export (thousands of tons)' (thousands of tons) earnings' Countrya 1978-80w 1990 1978-80' 1990 1978-80,' 1990 Burkina Faso (85) 1,221 1,690 52 250 0.06 0.19 Ethiopia (74) 5,307 7,346 214 3,061 0.08 0.86 Mali (85) 1,293 1,790 65 377 0.06 0.20 Nepal (87) 2,872 3,976 -23 367 -0.02 0.22 Tanzania (87) 3,646 5,048 157 1,306 0.05 0.49 Uganda (80) 2,348 3,251 43 1,073 0.02 0.34 Note: The assumptions needed to derive the numbers in this table are given in the text. a. Numbers in parentheses are the energy content of average daily diets in 1978-80 as a percentage of the FAO/WHO requirement. b. Food production figures were derived from estimated food consumption data (FAO 19806). c. The FOB cost of cereals is estimated at $200 per metric ton. d. Annual average. Sources: Cereal data for 1978-80 from USDA. Export earnings data for 1978-80 and projections for 1990 are from the World Bank. Cereal projections and country export earnings for 1990 are World Bank estimates. 20 differences in food consumption among people in Food import requirements were projected also each country. for thirty-one developing countries in which con- * Population will expand at the currently proj- ditions are better, and the results are quite different ected rate. (Figure 2-4). For instance, the six countries will * Food production will grow at the optimistic need to import 53 kilograms of cereal per capita in rate of 3 percent a year (past growth for these 1990, a ninefold increase over their imports in countries ranged from - 0.1 to 3.2 percent a year). 1978-80. The thirty-one countries will need barely * Export earnings will grow at rates now proj- 12 kilograms per capita in 1990-only twice the ected by the World Bank. imports in 1978-80. The difference in the cost of The projections illustrate that high volumes of food imports as a proportion of export earnings is food imports would be required in 1990 to meet also large between the two groups. If anything, per capita energy requirements-imports these these disparities are understated, since the group countries cannot afford. In Ethiopia, for instance, of thirty-one countries is more likely than the group the bill for required cereal imports is almost the of six to realize the assumed 3 percent annual same as the projected export earnings. If cereal growth in food production. imports are held to 10 percent of their export earn- Three main factors explain the differences in the ings, international transfer payments such as food food import burden between these two groups of aid would have to expand greatly (Figure 2-3). For countries. First, the average energy content of the many of the countries, the transfer would need to national food supply in 1978-80 was 83 percent grow more than tenfold. Achieving food security of the FAO/WHO requirement in the six countries, in these countries will therefore require a joint and 91 percent in the thirty-one countries. Second, commitment of governments and the international the average export earnings were $25.20 per capita community. in the six countries in 1978-80, and $63.10 in the thirty-one countries. Third, annual population growth is projected to be 2.9 percent through 1990 Figure 2-3. Average Annual Food Aid, 1978-80, in the six countries and 2.3 percent in the thirty- and Projected Food Aid Requirements, 1990, one countries. for Six Low-Income Countries To reduce chronic food insecurity substantially (thousands of tons) in, say, ten years will take significant efforts to expand agricultural production and the incomes of Burkina low-income groups, emphasis on strategies to gen- Faso erate employment, more efficient use of resources, Ethiopia l - land improved government programs. It will also Ethiopia . .wz .} 5.es>. < --,-.'<.w-vl<'s z#_< ~XS2