Evaluates the impact of two food subsidy programs targeting poor primary schoolers in Honduras: (1) CARE”s Title II School Feeding Program, which provides children a mid-day merienda (snack); and (2) the Programa de Asignacion Familiar (PRAF) program, which provides children with a food coupon (bono) until sixth grade or they drop out of school. Both programs have had a significant effect on the key indicator, rate of academic progress (“years ahead”), which compares children”s actual rate of school progress with the ideal rate of starting at age 7 and progressing one grade per year. The effect of the Bono Program is considerably larger, however, and is also significantly associated with increased attendance, on which the Merienda shows no effect. On the other hand, the Merienda Program has a slightly larger effect of reducing repetition, while the Bono Program has only a marginal effect in this regard. Results suggest that the two programs have distinct effects, so that eliminating one in favor of the other would probably lose some program benefits. Further, evidence from the Health Centers Study suggests that the Merienda Program, but not the Bono, is associated with a greater likelihood of achieving dietary adequacy. The Bono Program appears to be significantly more cost-effective than the Merienda. But the effect of having a mid-day meal on concentration and ability to learn were not directly measured in this study, nor was dietary impact. If the Merienda Program is to be phased out in favor of the Bono Program, this should be done gradually, with efforts to promote a community-based alternative school meal and to monitor the transition on school performance and dietary indicators. A further serious concern is the long-term stability and viability of the Bono Program, which was originally intended as a short-term response to the negative effects of structural adjustment. The program, which started only in 1990, has already been suspended once; its effectiveness over time is probably dependent on its reliability. For its effects on educational progress and achievement to be felt in the larger economy, it needs to be available continuously for a generation of children. Given the positive results of this study, consideration should be given to ways of achieving the program”s long-term continuation. The effect of the Bono Program on years gained is greatest in the upper grades. Since enrollments are already high in the early grades, the program”s cost-effectiveness might be enhanced by making the coupon available starting only in grade 3, as an incentive for parents to keep their children in school. Because coverage of both programs is higher in the rural areas, there is a concern that needy households in urban areas may not have access to these benefits. There is effective targeting to the poor in both programs; the targeting is more pronounced in the Bono Program, because there is no within-school targeting for the Merienda Program. This should not be changed, as it is not reasonable to try to target the merienda within the school. In terms of national coverage, about 10% of all households receive benefits from the Bono Program; almost 30% receive school merienda. The effectiveness of offering both programs appears consistently lower than that of offering either program alone. This is due to the policy decision to make the combined program available in the neediest areas, where other disincentives to schooling are strong. This suggests that the bonos, originally intended to improve socioeconomic conditions, are not enough, by themselves, to compensate in the short run for the longer-term effects of being in a disadvantaged community. Despite the provision of the bono, policies to promote economic development of communities through other means are still needed. Some concerns regarding both programs appear to be unfounded. Mothers receiving the bono express a strong preference for benefits in this cash-like form rather than in the form of food; the issue does not appear to be control, but rather convenience and choice. The women report that they, not their husbands or families, decide how to use the bono. Women receiving the merienda were not asked their preference for receiving food versus cash. Regarding the merienda, there is no apparent substitution of the school food for home-provided meals. No direct information exists on whether children in non-merienda schools bring any food to eat in the middle of the school day; based on the sample of children in the Health Centers Study, the practice does not appear widespread. The merienda may also be a substitute for a home-provided snack, which the lowest-income families may not be able to afford. Most parents report making both a financial and a labor contribution to the Merienda Program, but presumably this contribution is below the value of the daily school meal. (Author abstract, modified)
