The Educational Impact of Large-Scale School Feeding Programs

Patrick McEwan

Schools across the world devote enormous resources to the provision of subsidized meals. In Chile, a government agency provides meals to poor children attending publicly-funded schools. In 2000, it provided free meals to 785,065 students in grades 1 to 8, comprising 43 percent of total enrollments. The widespread support for school feeding programs suggests that impacts on student outcomes such as academic achievement are understood. In fact, there is virtually no convincing research in the United States, and only a few small-scale randomized experiments in several developing countries. This project assesses the impact of Chile’s large-scale feeding program on students’ academic achievement and school attendance. Free meals are provided to students attending schools with higher levels of poverty, as measured by an annual “vulnerability index.” Schools are eligible to participate if their index score is above a cut-off value. To estimate the program’s impact, I compare student outcomes in the vicinity of eligibility cut-offs. Ineligible schools just below the cut-off serve as the counterfactual for eligible schools just above the cut-off. Any difference in student outcomes can be credibly attributed to the school meals program, with several caveats. The research, an application of the classic regression-discontinuity design, will provide some of the first credible evidence on the educational impact of a large-scale school feeding program.

About Patrick McEwan

N/A