Estimating Semiparametric Econometrics Models by Local Linear Method: With an Application to Cross-Country Growth

It is well established that local linear method dominates the conventional local constant method in estimating nonparametric regression models by kernel method. In this paper we consider the problem of estimating semiparametric econometric models by local linear method. We provide a simple proof of establishing the joint asymptotic normality of the local linear estimator. We then show that our results can be used to easily derive the asymptotic distributions of local linear estimators for several semiparametric econometric models. An empirical application of using a semiparametric local linear estimator to cross country growth data is examined.