The Effect of a Free Lunch on the Academic Performance of Impoverished Children: Evidence from a Quasi-Randomized Intervention in Western Kenya

Abstract

I consider a feeding program that is targeted at orphans and impoverished children living in a rural village in the Kakamega County of western Kenya and investigate whether receiving a supplementary meal has an effect on the student’s academic performance. There are children in the primary school who are eligible for the program in the village, but for whom there are not enough provisions. I manually collected the class rank and raw test scores of a subset of students at the local primary school during my stay in the village over the summer of 2013. For students in the program I have academic data from before and after they joined the program. To analyze the data, I use an entity fixed effect regression model to determine the impact of the nutritional intervention and test whether this impact is uniform across students or is dependent on student-specific factors and find that being in the feeding program has a positive and statistically significant effect for students who joined the program before the age of 12. This significance is driven by the program’s efficacy during the “hungry season”, the period in between harvests during which many families have depleted their crops from the previous growing season.