Webbink, D. (2005). Causal effects in education Journal of Economic Surveys, 19(4):535--560.

In recent years, a wave of new studies on the effects of educational interventions on student performance has emerged. The realization that inputs in the education process are endogenous is important for the validity of traditional findings. Because of ignoring endogeneity bias, all traditional estimates might be wrong. Recent studies exploit exogenous variation in interventions in education produced by controlled or natural experiments. Results generated by this methodological innovation differ substantially from the traditional findings. This article reviews this new literature, illustrates new methods for identifying causal effects of interventions in education and compares the findings with the traditional literature.