The Private Return on Education and How to Solve the Endogeneity Problem: Case Indonesia

摘要

This paper discusses the return on education using the Mincer model. The Mincer equation is a log natural income associated with years of school completion and work experience. One of the problems arising from the Mincer equation using the OLS (Ordinary Least Square) method is biased estimation result. The magnitude of the bias collected by previous researcher's ranges from 0.7% -9.4%. Bias can occur due to sample selection and endogeneity problems. Problem of sample selection will be overcome with mills ratio or Heckit method while the endogeneity problem in this paper will be overcome by IV (Instrumental Variable) method. Instrumental variables must have a provision: the variables used as the instrument have no relation with the dependent variable, and the variable has correlated with the independent variable that is considered endogenous. Variable used as instrument is parent education. However, in this paper, I use pooling data from 1993-2014 with Indonesia Family Live Survey. Empirical results show the difference between return on education value using OLS and IV methods. Results with IV give a larger return value than OLS. The return value bias for the Indonesian case is still in tolerance.