The objective of this paper is to evaluate the learning gap between private and public school children in primary school, and ascertain the part of the private–public school learning gap that is due to differences in observables and the part that can be attributed to private school effect.
The paper utilized a household survey data from Ghana that assessed children on numeracy and literacy in both English and local languages. The primary methodology for this study is non-linear Oaxaca decomposition. The study also utilized Welch's t-statistics to test if there are any differences in the private–public school learning gaps across several sub-groups.
Findings from this study show a substantial gain for private school attendance on both numeracy and literacy. The results show that a little over 60% of the total learning gap in numeracy and literacy in English is explained by observable characteristics. However, observable characteristics almost explain all the learning gaps in the reading and writing of local languages. Evidence from the study suggests that the private school effect is homogeneous across several sub-groups. The results reveal years of education, expenditure on extra classes, religion and urbanicity as the most important variables explaining the gap that is caused by differences in observables.
Despite the belief that private school children in Ghana have better learning outcomes, there has not been any study to quantify this learning gap in the country and this study fills this gap. While there is literature on the differences in the learning outcomes between public and private schools, those studies have focused on the differences that are attributable to the private school effect. This article does not only present the differences in the learning outcomes but also shows the proportion that is due to observable characteristics and the part that can be attributed to the private school effect.
