This paper has benefited from comments by a co-editor, two anonymous referees, Andrew Chesher, Han Hong, Jim Powell, and seminar participants at Berkeley, Duke, Harvard/MIT, Purdue, Vanderbilt, and the Midwest Econometrics conference. The first and third authors acknowledge financial support from the National Science Foundation.
Testing for Causal Effects in a Generalized Regression Model With Endogenous Regressors
Version of Record online: 3 DEC 2010
© 2010 The Econometric Society
Volume 78, Issue 6, pages 2043–2061, November 2010
How to Cite
Abrevaya, J., Hausman, J. A. and Khan, S. (2010), Testing for Causal Effects in a Generalized Regression Model With Endogenous Regressors. Econometrica, 78: 2043–2061. doi: 10.3982/ECTA7133
- Issue online: 3 DEC 2010
- Version of Record online: 3 DEC 2010
- Manuscript received April, 2007; final revision received November, 2009.
- causal effects;
- semiparametric estimation
A unifying framework to test for causal effects in nonlinear models is proposed. We consider a generalized linear-index regression model with endogenous regressors and no parametric assumptions on the error disturbances. To test the significance of the effect of an endogenous regressor, we propose a statistic that is a kernel-weighted version of the rank correlation statistic (tau) of Kendall (1938). The semiparametric model encompasses previous cases considered in the literature (continuous endogenous regressors (Blundell and Powell (2003)) and a single binary endogenous regressor (Vytlacil and Yildiz (2007))), but the testing approach is the first to allow for (i) multiple discrete endogenous regressors, (ii) endogenous regressors that are neither discrete nor continuous (e.g., a censored variable), and (iii) an arbitrary “mix” of endogenous regressors (e.g., one binary regressor and one continuous regressor).