Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/73121 
Year of Publication: 
2012
Series/Report no.: 
Diskussionsbeitrag No. 504
Publisher: 
Leibniz Universität Hannover, Wirtschaftswissenschaftliche Fakultät, Hannover
Abstract: 
In this paper, we derive a semiparametric estimation procedure for the sample selection model when some covariates are endogenous. Our approach is to augment the main equation of interest with a control function which accounts for sample selectivity as well as endogeneity of covariates. In contrast to existing methods proposed in the literature, our approach allows that the same endogenous covariates may enter the main and the selection equation. We show that our proposed estimator is \sqrtn-consistent and derive its asymptotic distribution. We provide Monte Carlo evidence on the small sample behavior of our estimator and present an empirical application. Finally, we brie y consider an extension of our model to quantile regression settings and provide guidelines for estimation.
Subjects: 
Sample selection model
semiparametric estimation
endogenous covariates
control function approach
quantile regression
JEL: 
C21
C24
C26
Document Type: 
Working Paper

Files in This Item:
File
Size
302.66 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.