A strategy to reduce malnutrition: There is a need for the development of nutrition-oriented food policies to meet one of the developing world’s most basic needs
Author: Alan Berg

Contributor Notes

Nutrition is a basic human need, one that continues, and will continue, to remain unfilled for vast numbers of people in the developing world. This article discusses the need for nutrition-oriented food policies—a promising vehicle for a major attack on the problem.

This paper highlights that the flow of IMF-related resources to member countries was maintained at a high level during 1979, amounting to the equivalent of SDR 6,917 million, compared with SDR 4,955 million in 1978. Some SDR 3.77 billion became available to non-oil developing countries in 1979. Repurchases in the General Resources Account by all members—at SDR 4.2 billion—exceeded their purchases of SDR 1.8 billion by an unprecedented SDR 2.4 billion. These large repurchases reflected the substantial improvement in the balance of payments of some industrial member countries that had large outstanding drawings.


This paper highlights that the flow of IMF-related resources to member countries was maintained at a high level during 1979, amounting to the equivalent of SDR 6,917 million, compared with SDR 4,955 million in 1978. Some SDR 3.77 billion became available to non-oil developing countries in 1979. Repurchases in the General Resources Account by all members—at SDR 4.2 billion—exceeded their purchases of SDR 1.8 billion by an unprecedented SDR 2.4 billion. These large repurchases reflected the substantial improvement in the balance of payments of some industrial member countries that had large outstanding drawings.

Alan Berg

Accepting nutrition as a basic human need has major implications for the way in which the problem of malnutrition is perceived and addressed. Such an approach emphasizes the importance of adequate nutrition, both as an objective and as a means of economic development. Moreover, it calls attention to the nutritional needs of people of all ages and both sexes—in contrast to the tendency of most national nutrition planning to focus almost exclusively on the needs of children and pregnant and lactating women. This article addresses these implications in the light of recent nutrition-related research and experience.

This article is bused upon a longer paper, produced by the author for the World Bank.

Among all human needs, the need for food is perhaps the most basic. Adequate nutrition is central to survival and is a critical factor in an individual’s growth and capacity to function in society. Even in normal, nonfamine conditions, inadequate food consumption significantly affects the death rate. If adults fail to meet their food requirements, they lose weight. This can lead to diminished ability to cope with infection and other environmental stresses, to work, to enjoy the normal satisfactions of life, and to raise and educate healthy children. Maternal malnutrition during pregnancy results in low birth weight of infants (the most important cause of infant mortality), and malnutrition during lactation affects directly and indirectly the infant’s health, as well as the mother’s.

Undernourished children are less active than they should be. When a child’s intake of food falls below 70 per cent of standard allowances, growth slows; if low levels of feeding persist, adult stature is reduced. Whether small size per se is disadvantageous is arguable. However, severe malnutrition associated with decreased growth also results in decreased brain size and cell numbers and altered brain chemistry. Even when malnutrition is not severe, decreased growth is associated with low scores on tests of cognitive and sensory ability. Small stature caused by early malnutrition has been directly related to poorer productivity. Recent studies of work performance of sugarcane cutters concluded that taller workers were more productive at all ages than shorter workers. Work output in industry has also been linked to body size.

Improved nutrition, and the survival and well-being it implies, is a sufficient justification for a country to invest in better nutrition for its population. In addition, reducing or preventing malnutrition and its deleterious effect on work capacity and cognitive ability adds to human capital and potentially increases output. Programs that improve the nutritional status of the poor also help to achieve other social goals, including the distribution of income, and, through their impact on nutrition-related mortality, help reduce desired family size.

Strategy needed

The common official response to the problem of food and nutrition has been to grow more food. More recent analysis, taking account of low demand or purchasing power among the poor and malnourished, has also emphasized the need to increase their incomes. The relative importance of food production and income, and the effect that improvement in either would have on the level of nutrition in a country, have been projected by Bank staff from data available on Bangladesh, Brazil, India, Morocco, and Pakistan. With the most optimistic (and perhaps least likely) set of assumptions—high income growth and stable food prices—the projections of the magnitude of malnutrition in the five countries studied show a sharp reduction by 1995 in both the proportion of the population that is undernourished and in the magnitude of its deprivation. But even then, the absolute number of the undernourished remains over 150 million in these five countries. With high income growth and, as is more likely, slightly increased food prices, the number of malnourished would fall by only about 1 per cent per annum from current levels. And with more pessimistic (and, unfortunately, not implausible) assumptions, the number of malnourished would increase significantly. Projections based on the actual recent experience of these countries show that only Pakistan could expect to have calorie adequacy within the next 30 years.

One key conclusion of this analysis is that increases in income and in food production are likely to fall far short of what is required to meet basic needs in nutrition. A specific strategy to improve nutrition is thus required to complement what a policy focused only on growth would be too slow to accomplish.

Policy considerations

Emphasis here is on the relatively neglected, deliberate use of public policy to influence the character of production, processing, and distribution of food within a country to increase the amount consumed by the poor. This can occur in two ways. First, by changing the strategy of agricultural production to put more emphasis on the nutritional needs of the poor; and second, by adjusting consumer prices in ways to assure that the poor have access to what they need.

Every government now influences, in a variety of ways, the quantities and kinds of foods being produced, traded, and consumed. Yet the nutritional effects of agricultural and food policies generally are inadequately planned or anticipated. Few countries have systematically collected—or used in their planning—data on the nutritional consequences of changes in incomes and in food prices. Fortunately, several countries (for instance, Brazil, India, Indonesia, and the Philippines) have begun disaggregating food consumption data by income group. These data, along with recent analyses and project experience, suggest several key considerations for the design of food policies aimed at improving nutrition.

First, families with very low incomes spend most of their incomes on food, but still often do not consume enough for nutritional adequacy. In many countries, more than 40 per cent of the population have calorie-deficient diets, and upward of 15 per cent have gross deficiencies. Second, the foods they buy differ markedly from those bought by the rest of the population. In Indonesia, for instance, the lowest three income deciles obtain about 40 per cent of their calories from cassava and corn; by contrast, the upper three deciles obtain about 14 per cent of their calories from these foods. Third, and contrary to common assumptions, poor people tend to have an adequate balance between protein and calories in their diets, even when an important share of their calories comes from low-protein starchy staples. Regrettably, they are not always so rational when it comes to the distribution of food to the very young of their families.

Fourth, while many of the malnourished are small farmers and their families, the majority are landless or near-landless agricultural laborers, those in other low-paid rural nonfarm or urban jobs, or the unemployed. They thus benefit only indirectly from programs designed to boost the production of small farmers. Fifth, within any country, the extent of deficiencies tends to vary substantially between regions and between rural and urban areas. The variations that occur partly reflect differences in incomes and tastes. They also reflect differences in prices resulting from differences in transport, production and storage costs, marketing margins, and government pricing policies. Often nutritional adequacy of diets also varies a great deal by season and by year. Seasonal deficiencies tend to be severe in countries with only one major harvest, where price rises generally coincide with seasonally low earnings for agricultural laborers. Recent studies confirm that the highest incidence of malnutrition often occurs at this time.

Introducing a concern with nutrition into food policy calls for a broad review of agricultural production strategies. Most important is the need for more emphasis on the production of those low-cost foods consumed primarily by the poor. In Colombia, a 10 per cent increase in the supply of beef could add three times as many calories to the daily diets of the already adequately nourished group as to the diets of the calorie-deficient group. In contrast, the benefits of a 10 per cent increase in the production of cassava would be received entirely by the calorie-deficient group. Emphasis on foods favored by the poor should extend to every aspect of agricultural strategy, including research, extension services, supply of inputs, credit facilities, and marketing.

“Nutrition-oriented food production programs, in conjunction with steps to increase incomes of the poor, are the key elements in meeting nutritional needs in the longer run.”

Effective programs aimed at improving the production of small farms will generally improve the diets of small-farm households, whether or not the programs directly increase the production of the food crops they consume. However, changes in agricultural cropping patterns or policies can sometimes have unexpected deleterious effects on nutrition. When a farm shifts production from subsistence to cash crops, total income may increase but heads of households may not provide to their families as much food as before, since it must now be bought at retail prices. This particularly affects young children and women, who generally come at the end of the family food line. Intensive nutrition education or plans to continue some subsistence production may be needed to avoid the dangers inherent in such shifts. Similarly, when agricultural prices are increased as an incentive to increase production, wages for unskilled laborers, which are often at bare subsistence levels, are generally slow to rise in step, and the transitional effect on nutrition may be severe. It can be alleviated through wage policies, short-term subsidies, or other food demand programs.

Food demand programs

Nutrition-oriented food production programs, in conjunction with steps to increase incomes of the poor, are the key elements in meeting nutritional needs in the longer run. In the intervening (and often lengthy) period, measures are required to reduce the prices the malnourished pay for their food. This generally involves the often contentious issue of food ration and subsidy programs—or food demand programs. Such programs raise budgetary, balance of payments, and agricultural pricing problems. Rationing programs are also difficult to administer, and they may build up political constituencies that make them difficult to cut back. It is not surprising that many development analysts have believed that ration and subsidy programs are much more a part of the development problem than a part of the solution. These programs are, however, one of the few ways to meet the basic nutritional needs of very large groups of people. Experience in several countries shows that such programs can effectively reach the poor and substantially reduce severe malnutrition and malnutrition-related deaths.

The nutritionally most successful (and financially one of the most costly) food demand program has been carried out in Sri Lanka. A 1970 survey shows that the rice ration program (which then provided rice to nearly the whole population) provided about 20 per cent of total caloric intake and had a value equivalent to 14 per cent of income for the lowest income group. Only 5 per cent of the (adult-equivalent) population consumed less than 1,900 calories per day. In 1975, life expectancy was 66 years, higher than in richer countries such as Brazil, Korea, or Malaysia, and the highest in relation to income level of all countries for which data are available. It was 39 per cent higher than would have been predicted for Sri Lanka’s income level; infant mortality was 67 per cent lower.

How far is nutrition responsible for Sri Lanka’s long life expectancy and low infant mortality? Sri Lanka also has a tradition of assisting the poor in meeting other basic needs—its literacy rate is 78 per cent, health services are good, and the water supply is better than that of most other poor countries. Clearly such achievements in meeting basic needs complement one another. But the direct relation between food supply, nutritional adequacy, and the low death rate in Sri Lanka has also been confirmed statistically. It was seen most dramatically in 1974, when ration supplies were sharply reduced and food prices on the open market sharply increased because the cost of imported food grains more than doubled. In that year the death rate increased significantly. The literacy rate had not changed, nor had the proximity to health services or the quality of the water supply. There was no plausible reason for the increased deaths other than the shortage of food. Programs to provide the poorest with subsidized food grains in Kerala (India) and Pakistan also appear to have had a positive impact on nutritional status.

In most developing and developed countries, food is, politically, a special commodity. There is ample evidence that redistributing incomes is politically difficult. However, many countries that are unwilling to transfer income to the poor are willing to transfer large sums to subsidize staple foods.

Target food programs

Food subsidies can be expensive, however. For example, in 1975 their share of total budgetary expenditure was 21 per cent in Egypt, 19 per cent in Korea, 12 per cent in Morocco, and 16 per cent in Sri Lanka. To keep costs down, subsidies must be specifically aimed—or “targeted”—at the nutritionally needy. Otherwise, the programs are likely to run short of funds and supplies just when they are needed most, as Sri Lanka’s program did in 1974. Had the subsidies been directed more at the poor, as they are now, the increase in mortality could have been substantially reduced.

Target groups of malnourished can be identified in a number of ways—by income, by region, by season, by age, and by the staples in their diet. Income is probably the best means of indicating where nutritional problems on a large scale are concentrated. However, imposing a means test to define the target group is difficult, both politically and administratively. Another way is by geographic area—a part of Colombia’s nutrition program, for instance, provides foods to specific age groups living in specified needy geographical regions, regardless of income level. A promising but generally unexplored means of targeting is by season. Countries with subsidy programs during the months when food prices normally are highest will cover the periods of severest nutritional vulnerability and of highest demand for expenditures of human energy.

Most target programs require rationing to avoid large-scale misuse. But when poor households constitute the target group, the need for rationing, with all the administrative work it entails, can be reduced, or even sometimes eliminated, by a careful choice of the commodities to be subsidized. The best foods to subsidize are those with a low cost per calorie or gram of protein and with a positive income elasticity for the poor and a low or negative income elasticity for others. Coarse grains like sorghum and millet, processed cassava flour, and certain food legumes favored by the poor (for example, macassar beans in Brazil, lentils in central India, black gram in south India) meet these criteria. They are generally consumed in lower quantities by middle-income and upper-income people, so that a subsidy primarily would benefit the poor. Such subsidies have rarely been tried. Most subsidies of staples have, in fact, been for high-status grains like wheat and rice (which have been widely available in concessional foreign food aid programs) with the aim of reaching the urban lower-middle and middle classes.

“… malnutrition is not just a poverty problem …. countries committed to eliminating most overt malnutrition appear capable of doing so.”

The malnourished can also be reached by subsidizing inferior qualities of a given staple. For example, a subsidy on low-quality rice (short-grain, with a high percentage of broken grains) will have more effect on the diets of the poor than a subsidy on higher quality rice. So-called composite flours (such as a mixture of cassava with wheat and soy flours) could have nutritional benefits as well. Subsidizing a particular commodity—for instance, processed weaning foods or special fortified foods for pregnant women—can tend to limit beneficiaries of the subsidy to vulnerable groups.


Even food demand programs aimed at specific target groups will have substantial leakages to people outside the groups and through substitution when intended beneficiaries reduce their food expenditures or sell subsidized food. Such substitution may cause the program to become, at least partly, a disguised income transfer. There is little empirical information on the exent to which this happens. Only one study has rigorously compared the effect of a ration program on caloric intake with what would have occurred with an equivalent income transfer; it found the nutritional impact of the ration program in Kerala to be substantially greater. But even if a program turns out to be primarily a disguised income transfer to the poor, it cannot be assumed that the funds not spent on food are wasted. People poor enough to have serious nutritional problems in their families tend to spend more than half of their incomes on food, and most of the rest is spent for other basic needs like shelter and clothing.

Target programs involving ration shops or food stamps raise formidable institutional problems. There is bound to be a good deal of bureaucratic inefficiency, and the power of rationing officials to grant or withhold ration cards is an invitation to widespread corruption. But if administrative costs do not wipe out food savings and if they appear to be proportionately lower than those of other nutrition or poverty-oriented efforts, such a program will be better than one that is not targeted at all.

The potential disincentive to domestic agricultural production that ration and subsidy programs pose is also a widespread concern. Steps governments take to reduce the balance of payments and fiscal costs of the programs can lead to a net reduction in the prices farmers receive for their products and, hence, to a reduction in their output. The need to provide incentives for domestic agricultural production is increasingly accepted in developing countries, although the incentives are frequently inadequate and remain a serious problem. Food demand programs should be designed to take adequate account of disincentive risks.


Several high-priority areas for country nutrition strategies emerge from this analysis. Accelerated growth in the incomes of the poor and, with very few exceptions, in food production continue to play fundamental roles in efforts to meet nutritional needs. While growth in income and in food production is not a sufficient condition for meeting basic needs in nutrition, it is a necessary condition. Without this growth, the poor will afford less food and government revenues will not increase enough to finance nutrition and other programs. Similarly, there must be rapid increases in food supply to meet increases in population and in per capita incomes. Otherwise, the relative price of food will go up, at least partly offsetting any increases in incomes of the poor. In addition, many of the poor are dependent on food production, as farmers or farm laborers, for increases in their incomes.

New attention is needed on developing nutrition-oriented agricultural production policies and programs. And to ensure that food reaches those in need, increased emphasis is required on food demand programs, including the strong possibility in some cases of food subsidies aimed at specific target groups. Such programs should not be dismissed a priori because of past fiscal and other problems associated with untargeted programs. (Nor should they be seen as a panacea for solving nutrition problems; initial emphasis should be on experimentation and careful monitoring.)

These newer areas of policy emphasis complement more commonly used, but limited, ways of reducing the malnutrition—nutrition education, the fortification of staples with micronutrients, institutional feeding programs, and the use of health services to support nutrition programs. Different countries will have different program priorities, depending on the distribution of malnutrition between rural and urban areas, the extent to which the rural malnourished are small farm families, the particular nutritional problems and their causes, the likely cost-effectiveness of feasible interventions, institutional and funding capacity, and political constraints.

Children under three with problems of low birth weights (largely addressed through the mother) and those in the first year of life deserve priority attention, because of the effect of early malnutrition on subsequent life. But nutrition needs also exist among older children, the aged, and working adults, who are able to produce more income and food for their families if they are better nourished. Even if planners were concerned only about reaching children, it often would be more cost-effective to reach them through programs affecting malnourished households as a whole; in many countries, a large percentage of malnourished children at vulnerable ages cannot be reached effectively in any other way.

The whole problem of the malnourished can be viewed as part of a complex tangle of poverty; clearly, the fundamental causes of poverty loom large in any analysis of the problem. But malnutrition is not just a poverty problem. While virtually all people suffering from calorie deficiencies are poor, not all poor people suffer from calorie or other nutritional deficiencies. Some countries with high per capita incomes have considerable malnutrition, and certain low-income countries have little. In short, countries committed to eliminating most overt malnutrition appear capable of doing so.

Finance & Development, March 1980
Author: International Monetary Fund. External Relations Dept.