Estimating the Returns to Schooling : Implications from a Dynamic Discrete Choice Model

This paper assesses the applicability of a dynamic discrete choice model in accounting for the observed OLS and IV estimates of the Mincer equation parameter on returns to education. I estimate a dynamic model of schooling and employment choices using an extended version of Keane and Wolpin (1997) to simulate educational attainment, employment history, and wages. Estimation of the Mincer wage equation using simulated data appears to validate the model, showing that both OLS and IV estimates are greater than the population average returns to schooling. Consistent with the dynamic decision model, ability selection is the major source of bias in the OLS estimates of schooling returns, and the dynamic selection caused by endogenous experience is important to account for biases in the IV estimates