There is a general consensus that Food for Education (FFE) programs increase primary school participation. Although this view is widely held, there is limited causal evidence to support it, making it difficult to anticipate the size of expected impacts. Moreover, little is known about how the design of FFE programs affects schooling outcomes. This paper presents evidence of the impacts of alternative methods of FFE delivery on schooling in Northern Uganda using a prospective, randomized controlled evaluation conducted from 2005-2007. We compare the impacts of the World Food Programme’s in-school feeding program (SFP) with an experimental take-home rations (THR) program conditional on school attendance to examine how outcomes are affected by the timing of meals and the placement of incentives with children versus parents. Results show that the in-school meals program increased enrollment for those children who were not enrolled at baseline, but who had reached the recommended age of school entry. For many outcomes we cannot reject that the THR impact is equivalent to that of the SFP. Both programs had large impacts on school attendance, with impacts varying by grade and gender. Moreover, both the SFP and THR programs reduced grade repetition, with larger impacts coming from the SFP program. The SFP program also reduced girls’ age at entry to primary school. Neither program affected progression to secondary school. In fact, children in grades 6 and 7 in SFP schools in 2005 were significantly more likely to remain in primary school in 2007, suggesting that school meals induce hungry children to delay completing primary school.
Note
Electronic ed. With financial support for the data collection and analysis from the World Food Programme, the World Bank, and UNICEF.
