Trending time-varying coefficient spatial panel data models

Abstract

This paper investigates the estimation and inference of spatial panel data models in which the regression coefficient vector is a trending function. We use time differences to eliminate the individual effects and employ GMM estimations for regression coefficients with both linear and quadratic moments. A two-step estimation method is used to improve the estimation efficiency. In the first step, GMM estimations with linear and quadratic moments are obtained. In the second step, the best linear and quadratic moments are employed to obtain the asymptotic efficiency. Time trend estimates based on these GMM estimates are also proposed. The results of Monte Carlo experiments show that the finite sample performance of these estimates is satisfactory.

Keywords: Spatial econometrics; Spatial autoregressive models; Nonlinear time trend; Generalized method of moments

JEL Classification: C13; C14; C21; C23