Delivering Food Subsidy - The Oxford Topic of Food, Politics, and Society - page 319

stamps scheme was nonindexed real benefits were reduced to 43% by 1981-82 of what
they were in 1979. Food price subsidies formed 18% of the budget of an average house-
hold; under the food stamp scheme it dropped to 9.6%. However, targeting improved
under the food stamp scheme. Under the rationing system, only 50% of the total outlays
in subsidy went to the bottom 40% of the households that included most of the house-
holds consuming less than the recommended energy allowance; under the food stamps
scheme this number went up to 66%. But since there was a net reduction in the total real
subsidy, the nutritional status of the poor worsened. Per capita calorie consumption of
the bottom 20% declined about 8% from an already low 1490 calories during 1978-79 to
1368 calories during 1981-82.
It is clear from the previous account that the change from an in-kind transfer system
to a restricted cash transfer system (food stamps) took place during a period when the
overriding consideration of the government was budget cutting. The change in the sys-
tem was undertaken with perhaps an explicit intention of reducing the net subsidy as
evidenced by the issuance of nonindexed food stamps in an environment where price
rises were inevitable. The Sri Lankan experience raises a red flag in the minds of skep-
tics that a proposal to change a food transfer system to a cash transfer system may be a
Trojan horse to reduce the level of support to the poor.
Jamaica is another example of a country where a general price subsidy program was
replaced by a food stamp program. In the aftermath of the worldwide financial turmoil
in 1980, the Jamaican government was compelled to undertake austerity measures. One
consequence was the total elimination of general price subsidy and its replacement by a
targeted food stamp program and an expanded school-feeding program. Grosh (1992)
gives an appraisal of the food stamps program. Food stamps were issued to 142,000 ben-
eficiaries out of a population of 2.2 million within seven months of the announcements
of the program. The administrative costs were just 9% of the total cost of the program.
The gains in targeting were impressive. With food stamps, 57% of the benefits accrued
to the bottom 40%, while only 8% accrued to the wealthiest quintile. With a general
price subsidy, these numbers were 34% and 26%, respectively. What about the impact
on nutrition? No systematic study that we know of exists that could quantify what part
of the nutritional impact over the next few years could be attributed to the change from
general price subsidy to the food stamps in Jamaica. All we have is circumstantial evi-
dence. For example, according to Grosh, malnutrition among children below an age of
five years declined from 14.6% in 1985 to 7.3% in 1989, and the food stamp program was
launched in 1984.
A poster child for a cash transfer program of recent vintage is Mexico's PROGRESA
program, launched in 1998. It provided cash transfers to families conditional on the reg-
ular attendance of their children in schools and health clinics. The idea was to provide a
safety net while ensuring human capital formation. According to a study by Hoddinott
and Skoufias (2003) of 24,000 households from 506 communities, the households
receiving PROGRESA benefits had a caloric intake that was 7.1% higher than those that
were not. More importantly, the quality of their nutrition, as measured by the caloric
value coming from vegetables and animal products, was higher. The program now