Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

IFPRI publications




The food stamp scheme in Sri Lanka: Costs, benefits, and options for modification. Neville Edirisinghe. Research Report 58. International Food Policy Research Institute, Washington, D.C., 1987.

In 1979 Sri Lanka replaced a four-decade-old food subsidy scheme, characterized by price subsidies and the rationing of rice, the main staple, with a food stamp scheme. Under the new scheme, households whose declared incomes were less than a specified amount received food stamps, which they could use to buy basic foods at non-subsidized prices. The change in the food subsidy programme was intended to minimize government intervention in food marketing and consumption, increase domestic savings, and protect low-income households from the removal of food and non-food price subsidies. The new food subsidy policy was an important element of a package of economic policy reforms made during the late 1970s. These reforms attempted to reduce government intervention in the economy and give the market a larger role in determining the prices of inputs and outputs and the allocation of resources. The economic liberalization policies were expected to provide the necessary impetus for greater growth in the economy. Indeed, economic growth was remarkable. GDP grew at an annual rate of 6.5% between 1978/79 and 1981/82.

In The Food Stamp Scheme in Sri Lanka, Neville Edirisinghe examines the costs and benefits of the changes in the food subsidy policy and suggests changes to increase the efficacy of the new income transfer programme.

The fiscal and household effects of the new scheme

Until the last year of the old subsidy scheme, rice and some other foods were made available at subsidized prices to almost the whole population. These price subsidies accounted for about 15% of government expenditures during the mid- 1970s. in 1979 the new scheme issued food stamps to only half of the population, removed all price subsidies on food, and kept the nominal value of food stamps constant By 1984 the share of government expenditures allocated to food subsidies had dropped to 3% and food subsidies as a share of GNP had dropped from a high of 6% to about 1 %. However, the wages of public sector employees were increased to compensate for the inflationary pressures that followed removal of price subsidies. This probably reduced the proportion of the savings from curtailment of the food subsidies that could be diverted to investment.

When the food stamp scheme began in 1979, the average recipient household received only 83% of the benefits it had received under the old scheme. Erosion of the real value of food stamps by inflation reduced this to 43% by 1981/82. The share of subsidies in the household budget of the average rice ration recipient was nearly 18% in 1978/79, but the share of food stamps in the household budget of the average recipients was only 9.7% in 1981/82. The smallest share after the change in scheme, 7.1 %, was for households in the plantation sector; this reduction was caused mainly by the elimination of wheat price subsidies.

The per capita calorie consumption of nearly 75% of all households either stayed the same or increased between 1978/79 and 1981/82, reflecting the effects of overall economic growth on average household income. For most households, the effects of inflation were largely offset by increased incomes and through substitutions. By substituting less expensive calories, the average household was able to reduce the food expenditures necessary to obtain the same amount of calories by about 7%.

The calorie consumption of the poorest 20%, however, declined about 8% per capita, from an already low 1,490 calories during 1978/79 to 1,368 calories during 1981/82, which suggests that the new food stamp scheme was not effective enough in helping the most vulnerable households. These households seemed to be unable to take advantage of the income-earning opportunities created by the economic reforms.

The effect of food stamps on calorie consumption was estimated with the assumption-confirmed by a statistical test - that households treat food stamp incomes just like any other income when allocating any additional income for food consumption. According to these estimates, the additional income received through food stamps enabled the poorest 20% of households to increase calorie consumption by 12%. Those in the next quintile increased their consumption by 6%. As expenditures increased, the effect of food stamps on total calories consumed declined significantly.

Effect on child nutrition depends on intra-household food distribution

Estimates of the effects that the change in system had on the nutritional welfare of preschool children revealed that the benefits they received depended on how food is allocated within the household. A special survey of 480 households showed that food stamps were associated with an increase of 5% in calorie consumption by preschool children in the lowest quintile, but with an increase in consumption by all other members of the same households of more than 10%.

However, consumption by children appeared to increase significantly after the more productive members in the household received about 80% of their recommended calorie allowance. It appears that income transfers have to be large to ensure that preschool children are affected. The results imply that other child-related intervention programmes, including health services and supplementary feeding programmes, can be an important complement to income transfers.

Targeting and cost-efficiency

The food stamp scheme restricted the stamps to only half of all households, but not all of these were in the lower half of the expenditure range. As a result, the poorest 20% -the quintile that would form the target group if income were the criterion - received only 38% of the total food stamp outlay. The 40% of households with the smallest expenditures-this includes most households consuming less than the recommended calorie allowance- received two-thirds of the food stamp budget. This is an improvement over the old scheme, under which this segment received only half of the total outlays on subsidies.

The beneficiaries not in the lowest 40%, who allocate a smaller share of additional income to food consumption, reduced the cost-effectiveness of this income transfer. Assuming that the primary objective of the food-stamp scheme is to improve the nutrition of the households in the lowest quintile, leakages to beneficiaries in higher quintiles have raised the cost to the government of a calorie to 21/2 times the actual cost incurred by the households in the lowest quintile.

The food stamp scheme needs modification

The evidence that the benefits and consumption of the poorest 20% of households are reduced under the present food-stamp scheme strongly suggests that a modification of the present food stamp programme is required if low-income households are to be helped effectively. If the Rs 1.7 billion spent on food stamps in 1982 had been transferred only to households in the bottom quintile, their per capita calorie consumption might have increased to about 1,540 calories - about 70% of the recommended allowance. If the objective of the programme were to ensure consumption of the recommended allowance of 2,200 calories per capita, about a fourfold increase in the subsidy bill would have been necessary during 1981/82.

The real problem with the programme is how to make targeting more effective. Traditional targeting mechanisms, such as those based on child malnutrition, regional targeting, or subsidies on ''inferior" foods that are self-targeting, face many problems in Sri Lanka. The available measurements of child malnutrition show that it is not confined to low-income households. An emphasis on child malnutrition would leave out many households that need assistance but do not have a child malnutrition problem. Regional targeting may not be desirable because calorie deficits and other forms of malnutrition are widespread. Subsidizing "inferior" foods such as cassava will not solve the calorie problem because these foods contribute only a small amount to the diets of low-income households.

In Sri Lanka, participation by the people in representative government is widespread, literacy rates are high, and the country has a multi-party political system and a comprehensive and competent public administrative structure. These factors augur well for devising a targeting scheme in which the administrators and the community together can be effective in screening applications for the income transfer. However, a serious policy commitment is a prerequisite for successful implementation of such a scheme.

In a broader perspective, the constraints on effective participation by the vulnerable sections of the population in the development process must be properly identified and remedial measures adopted so that the dependence on government transfers for nutritional welfare can be minimized.


Global food surpluses and priorities for African development. Ojetunji Aboyade.

Two seminars were held in conjunction with the meeting of the board of trustees of the International Food Policy Research Institute (IFPPI) in Amsterdam in February 1987: "The Emerging World Food Situation and Challenges for Development Policy'' and ''Food Surpluses and the Developing Countries: Implications for Policy Research.'' (Papers from both seminars are available from IFPRI.) The following commentary by Ojetunji Aboyade (from IFPRI Report, vol. 9, no. 2, May 1987) summarizes the policy implications of the second seminar in the context of sub-Saharan Africa. Professor Aboyade, former chairman of the Department of Economics at the University of Ibadan, currently serves as chairman of the Nigerian government's Commission on Economic Reform and as an international consultant.

The basic policy and research problem in sub-Saharan Africa remains one of continued and persistent food shortage in conditions of poverty, underdevelopment, and lagging agriculture, which are more pervasive than in other developing regions of the world. In this context, the current surge in food surpluses in the OECD countries, especially in the European Economic Community, with its potentially adverse consequences for LDC food exporters and non exporters alike, is disconcerting: it renders African food exports vulnerable and is a mixed blessing for the structural adjustment process of African food and agricultural economies.

The key to poverty reduction in Africa is to increase per capita food production, lower food prices by reducing the cost of production through technological change, and stimulate sufficient employment growth in the non-agricultural sector, especially in the rural areas, in order to increase demand for food. To implement such a strategy, the long-run policy priority areas are clear: production-focused research, institution-building and human capital, rural infrastructure, and linkages with non-agricultural rural activities. Yet there are grey areas of policy in the short to medium run that require attention. First, while smallholder agriculture should have priority over large plantations, the priorities are not clear when it comes to food versus export crops. Second, after correcting exchange rates on a continuing basis, itself a painful process, issues of producer prices and pricing administration-even under free-market conditions with the abolition of commodity boards-are important, especially for food and agricultural products that are not traded. Finally, production subsidies not only for seed and fertilizer but for land clearance, water development, and light farm machinery are a particularly important issue in the African context. It is important to recognize that a strategy of cost-decreasing technological change would in Africa be pursued under conditions of structural adjustment of the economy. As such, costly import-dependent technology would pose a substantial problem, as would the risk of an incipient decline in producer prices resulting from expanded supplies of food.

There are thus no quick fixes, domestically or externally, for Africa's food problem. Other than for famine relief, food aid within or outside the European food surplus setting is not really a solution: it could help stabilize food supplies of developing countries, but in practice has caused further destabilization. Similarly, the IMP cereal facility has not demonstrated a potential for stabilizing food availability in developing countries and is negatively perceived in key areas of Africa. Further, it may be shortsighted for the region to rely on the possible availability of OECD food surpluses if the process of structural adjustment is to endure.

One alternative would be for Europe to withdraw some of its resources from its own agriculture, adapt some of its awesome technological capability (farm machinery, irrigation, agricultural chemicals, storage facilities), and then transfer it through joint ventures to the production of those crops that are of long term relevance for food consumption in Africa.

Furthermore, Europe's food surpluses can be used for many other things in Africa besides paying labour in the form of wagegoods for building rural infrastructure, such as creation of employment opportunities for school-leavers and for skilled artisans in the urban sector. This would upgrade their technological capability and raise their productivity.

Ultimately, however, donor agencies and international research and policy institutions can only serve as catalysts in relation to the policy reforms and actions that are called for. The greatest need is for African policy-makers to develop their own capacity to absorb, synthesize, and extend the knowledge and profound understanding of the development process in their countries and then to motivate and galvanize their societies to operate in the realities of today's world.

Notice to readers

Because of space limitations, lengthy tabular material was omitted from "Seasonal Variations in Nutritional Status in Rural Areas of Developing Countries: A Review of the Literature," by W. Teokul, P. Payne, and A. Dugdale (Food and Nutrition Bulletin, 8 [4]: 7-10 [1986]). Interested readers may obtain copies of these tables by contacting the Food and Nutrition Bulletin editorial office in Cambridge, Mass., USA.

Contents - Previous - Next