Semiparametric estimation of a censored regression model with endogeneity

Censoring and endogeneity are common in empirical applications. However, the existing semiparametric estimation methods for the censored regression model with endogeneity under an independence restriction are associated with some drawbacks. In this paper we propose a new semiparametric estimator that overcomes these drawbacks. We derive conditional quantile moment conditions for all the conditional quantiles and propose a moment-based estimator. In particular, we construct two types of moment conditions and develop a computationally attractive estimator. We show that our estimator is consistent and asymptotic normal. A Monte Carlo study indicates that our estimator performs well in finite samples and compares favorably with existing methods.