ECONOMETRICS: Nonparametric Identification and Estimation of Panel Quantile Models with Sample Selection; Dr Sungwon Lee
Abstract
This paper develops nonparametric panel quantile regression models with sample selection. The class of models allows the unobserved heterogeneity to be correlated with time-varying regressors in a time-invariant manner. I adopt the correlated random effects approach proposed by Mundlak (1978) and Chamberlain (1980), and the control function approach to correct the sample selection bias. The class of models is general and flexible enough to incorporate many empirical issues, such as endogeneity of regressors and censoring. I consider a dynamic extension of the models and provide identification conditions. Identification of the static model requires that T ≥ 3, where T is the number of time periods, and that there is an excluded variable that affects the selection probability. The condition on T in dynamic models is stronger than that in static models as it is needed that T ≥ 4 for identification of dynamic models. I also propose semiparametric models for practical implementation of estimation. Based on the identification result, this paper proposes to use penalized sieve minimum distance estimation to estimate the parameters and establishes the asymptotic theory. A small Monte-Carlo simulation study confirms that the estimators perform well in finite samples.
Click here to view paper.