Although years of schooling and enrolment rates can be used as measures of educational progress, they are inadequate in explaining children's learning process. This is because the benefits of education are ultimately determined by the skills acquired in school, and not just the number of years spent in education. This paper investigates the determinants of learning among primary school children in Ethiopia using Round 2 and 3 of the unique Young Lives survey data. Using the framework of the education production function, the inherent endogeneity problem is discussed and addressed as much as data permits. Empirical evidence suggests that children in private schools do not learn better than their counterparts in public schools. In addition, household income has little effect on children's learning. This suggests that government-backed income transfer programmes to low-income households may not achieve substantial improvement in learning outcomes. Rather, the number of hours spent studying at school may be more relevant in improving children's learning outcomes. Thus, full-day teaching as against the erstwhile shift teaching system in Ethiopia should be fully implemented and sustained across regions and districts.
African Development Review