Local GMM Estimation of Semiparametric Panel Data with Smooth Coefficient Models

In this article, we consider the estimation of semiparametric panel data smooth coefficient models. We propose a class of local generalized method of moments (LGMM) estimators that are simple and easy to implement in practice. We show that the proposed LGMM estimators are consistent and asymptotically normal. Monte Carlo simulations suggest that our proposed estimator performs quite well in finite samples. An empirical application using a large panel of U.K. firms is also presented.