Search

In this paper, we present data from an evaluation of the educational impact of a school breakfast program implemented in rural schools in Peru. The results showed positive effects on school attendance and dropout rates, and a differential effect of the breakfast program on multiple-grade and full-grade schools. Particularly in multiple-grade schools the program shows a significant and positive effect on short-term memory, arithmetic and reading comprehension. The evaluation also showed an unexpected negative consequence: students in the breakfast group reduced their time in the classroom with their teachers and increased the time in recess (when they consumed the breakfast).

This paper goes further in the discussion on the determinants of school attainment in developing countries. To properly estimate the effects of school resources on academic achievement, we need to take into account the large geographical inequalities in the distribution of school resources and the supply constraints faced by students living in poor areas. We do so by implementing a two-step correction that accounts for the constraints in school choice. Our findings suggest that failing to account for these constraints leads to an underestimation of the effect of school resources on school achievement of about 100%. This underestimation is particularly important for girls and in Math. Additionally, the contribution of school resources in explaining the gap in test scores between rich and poor students is doubled once we account for the constrained choices. [ABSTRACT FROM AUTHOR]