Primary school size and learning achievement in Senegal: Testing the quantity-quality trade-off

The literature on the effects of primary school size on learning achievement in sub-Saharan Africa is rather limited. To address this gap, we examine the causal relationship between primary school size and learning achievement of two cohorts of students, followed with collection of three rounds of data in Senegal. These two cohorts attended grades 2 and 4 at the beginning of the 2009-2010 school year. We implement an empirical strategy that hinges on value-added models estimated with a doubly robust approach; we find that fourth/fifth graders (and not second/third graders) attending mid-sized schools perform significantly better than their counterparts attending smaller or larger schools. These results are robust to a number of sensitivity analyses. The potential mechanisms through which school size operates include student's gender, family socioeconomic status, class size and the urban/rural location of the schools. Importantly, we suggest optimal sizes of primary schools for both policymakers and parents. These optimal sizes are contingent upon schools' urban or rural location.