Identification and Estimation of Sample Selection Models without Additivity
2014/06/10
研討會日期 : 2014-06-10
時間 : 14:30
主講人 : Professor Jen-Che Liao
地點 : Conference Room B110
演講者簡介 : Professor Liao received his Ph.D. in Economics from University of Wisconsin-Madison in 2013. He is currently an Assistant Research Fellow at Institute of Economics, Academia Sinica. His research field is Econometrics.
演講摘要 : In this paper, we extend the literature on sample selection models to allow for non-additively separable errors. Without imposing additive separability, we discuss what features of sample selection models can be identified under various restrictions, with a particular focus on a nonparametric nonseparable sample selection model with possibly endogenous continuous regressors. Using a control function approach, we provide identification results for the average structural function given selection and develop a resulting three-step nonparametric series estimator. Convergence rates are derived. A simulation study compares the finite sample performance of our proposed estimator with Das, Newey, and Vella (2003) estimator and Heckman's two-step estimator, which demonstrates the utility of our approach.
