Poorest are hurt by Sri Lanka's food stamp scheme

At one time or another governments of many developing countries have faced the need to make food available to low-income groups at less than market prices or even production costs. One of the most common measures adopted for this purpose has been the introduction of price subsidies on a range of staple foods.

But while food subsidies may be regarded as one of the few feasible ways of making income transfers for the benefit of the poor, the financing of such programmes can place a heavy burden on national budgets. In both Tanzania and Zambia, for example, costs of subsidizing food to consumers has far exceeded the budgets of the ministries of agriculture.

In recent years as the economic crisis has squeezed national budgets a number of developing countries have tried to curtail or eliminate food subsidy programmes. One notable example has been Sri Lanka, where, in 1979, a 44-year-old general price subsidy programme was swept aside and replaced with a direct income transfer scheme, which used food stamps. Considering that earlier attempts to cut back on the general subsidy scheme had lad to massive protests by labour unions and even to rioting, the Sri Lankan decision, part of a general package of reforms intended to liberalize the economy, and to give the market a larger role in determining prices and the allocation of resources, could only be regarded as politically courageous. And considering that the new scheme reduced the cost of food subsidies in Sri Lanka from 17.1 to 9.1 per cent of overall government expenditures, the political gamble would appear to have paid off.

However, a research report recently published by the International Food Policy Research Institute (IFPRI) has found that the well-intended income transfer scheme has resulted in a deterioration of the nutritional welfare of households in the lowest segment of income distribution and has not been able to protect them against the effects of inflation. When the new food stamp scheme was introduced in 1979, the average recipient household received only 83 per cent of the benefits previously enjoyed under the food subsidy scheme. Subsequently inflation in food prices further reduced the real value of food stamps to 43 per cent of the previous benefits. The value of food subsidies in 1978 represented nearly 18 per cent of the average household budget, compared with only 9.7 per cent for the food stamps in 1981-82. And although the per caput calorie consumption in three out of four households remained the same or increased during this period, a reflection of the effects of overall economic growth on average household income, calorie consumption in the poorest 20 per cent of households declined about 8 per cent per caput from an already low 1 490 calories to 1 368 calories. Clearly, the new food stamp scheme was not effective in helping the most vulnerable households.

Among the difficulties encountered in introducing the new scheme was that of carrying out a valid means of identifying the low-income groups. To be eligible for food stamps, household income had to be less than Rs 3 600 per year with marginal adjustments for larger families. However, the IFPRI study found that while the food stamp scheme was restricted to only about half of all households, compared with the near universal coverage of the food subsidy scheme, not all of these households were in the lower half of the income range. The poorest 20 per cent - the quintile that would form the target group if income were the real criterion - received only 38 per cent of the total food stamp outlay. "Leakages to beneficiaries in higher quintiles," the study noted, "have raised the cost to government to two and one half times the actual costs incurred by households in the lowest quintile."

The IFPRI study suggests that a nutritional goal, such as ensuring the consumption of a given amount of energy would be a more satisfactory way of measuring the effectiveness of an income transfer programme than simply seeking the enhancement of the general welfare of more disadvantaged sectors of the population. The study estimates that if all of the Rs 1.7 billion spent on food stamps in 1982 had been transferred only to households in the bottom quintile, their per caput calorie consumption could have been increased to about 1 540 calories, or about 70 per cent of the recommended allowance. To ensure consumption of the recommended allowances, 220 calories per caput per day, would require a fourfold increase in the food stamp programme funding.

Peter Hendry