Identification of Treatment Effects Using Control Functions in Models with Continuous, Endogenous Treatment and Heterogeneous Effects
We use the control function approach to identify the average treatment effect and the effect of treatment on the treated in models with a continuous endogenous regressor whose impact is heterogeneous. We assume a stochastic polynomial restriction on the form of the heterogeneity but, unlike alternative nonparametric control function approaches, our approach does not require large support assumptions.
We thank two anonymous referees and Whitney Newey for their comments. We also thank participants at the Berkeley-Stanford (March 2001) workshop on non-parametric models with endogenous regressors as well as participants at the UCL empirical microeconomics workshop for useful comments. J. Heckman would like to thank NIH R01-HD043411 and NSF SES-024158 for research support. C. Meghir would like to thank the Centre for Economics of Education and the ESRC for funding through the Centre for Fiscal policy at the IFS and his ESRC Professorial Fellowship. E. Vytlacil would like to thank the NSF for financial support (NSF SES-05-51089). The views expressed in this paper are those of the authors and not necessarily those of the funders listed here nor of the national Bureau of Economic Research. All errors are our own.
J. P. Florens & J. J. Heckman & C. Meghir & E. Vytlacil, 2008. "Identification of Treatment Effects Using Control Functions in Models With Continuous, Endogenous Treatment and Heterogeneous Effects," Econometrica, Econometric Society, vol. 76(5), pages 1191-1206, 09. citation courtesy of