The “true” private school effect across countries using PISA-2012 Mathematics

It is known that in most countries, students of private schools perform better in international assessments compared to students in the public school system. However, when one controls for observable socioeconomic background characteristics at the individual and school level, public school students perform equally well. Furthermore, sorting to private vs. public schools based on unobservable characteristics takes place, which biases econometric estimates. I account for selection on unobservables using an approach based on the idea that the amount of selection on the observed explanatory variables in a model provides a guide to the amount of selection on the unobservables (Altonjie et al. 2005; Oster 2013). I use PISA-2012 data for Mathematics to derive bias-corrected estimates of the “true” private-dependent and independent school effect for 40 countries. With few exceptions, public schools outperform private schools (especially independent schools). Accounting for both peer effects and selection is necessary when evaluating school effectiveness.