Content area
Full text
Empir Econ (2015) 48:253281
DOI 10.1007/s00181-014-0856-0
Received: 18 July 2013 / Accepted: 8 June 2014 / Published online: 13 August 2014 Springer-Verlag Berlin Heidelberg 2014
Abstract An interesting puzzle in estimating the effect of education on labor market earnings (Card in Econometrica 69:11271160, 2001) is that the 2SLS estimate for the return to schooling typically exceeds the OLS estimate, but the 2SLS estimate is fairly imprecise. We provide a new explanation that it could be due to the restrictive linear functional form specication on the covariates and the reduced form. For the parameters of endogenous regressors, we propose two kernel-based semiparametric IV estimators that relax the tight functional form assumption on the covariates and the reduced form. They have explicit algebraic structures and are easily implemented without numerical optimizations. We show that they are consistent, asymptotically normally distributed, and reach the semiparametric efciency bound. A Monte Carlo study demonstrates that our estimators perform well in nite samples. We apply the proposed estimators to estimate the return to schooling in Card (Aspects of labour market behavior: essays in honour of John Vanderkamp. University of Toronto Press, Toronto, pp. 201222, 1995). We nd that the semiparametric estimates of the return to schooling are much smaller and more precise than the 2SLS estimate, and the difference largely comes from the misspecication in the linear reduced form.
We thank the editor, Professor Subal Kumbhakar, and three anonymous referee for comments that improved the paper substantially. Any remaining errors are the authors responsibility.
F. Yao (B)
Department of Economics, West Virginia University, Morgantown, WV 26505, USA e-mail: [email protected]
J. Zhang
Department of Economics, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong e-mail: [email protected]
Efcient kernel-based semiparametric IV estimation with an application to resolving a puzzleon the estimates of the return to schooling
Feng Yao Junsen Zhang
123
254 F. Yao, J. Zhang
Keywords Instrumental variables Semiparametric regression
Efcient estimation Return to schooling
JEL Classication C14 C21
1 Introduction
Many studies have been devoted to uncover the causal effect of education on labor market earnings, where endogeneity in education calls for the use of instrumental variables (IV) to estimate the return to schooling (see Card 2001 for an excellent summary). Let us consider the framework studied by Card (1995),...