Parametric and Semi-parametric Estimations of the Return to Schooling in South Africa

This paper estimates return to schooling for african and coloured women in South Africa. It compares parametric and semiparametric estimates of the sample selection model for the case of return to schooling. The parametric estimator is the one proposed by Heckman (1979) and the semiparametric estimator proposed by Newey (1991) and Klein and Spady (1993). It also attempts to correct endogeneity and mesurement error by using instruments of schooling. Following recent literature, the paper uses community variables primary and secondary school proximity and availability as instruments. Using instrumental variables increases the return to schooling substantially. Parametric corrections does not change the results but semiparametric corrections increases the return even more