The Performance of Alternative General Method of Moment Estimators with Application to Employment Panel Data Model

Abstract

Dynamic panel data models make it possible to address dynamic economic relationships through the inclusion of a lagged dependent variable among the explanatory variables. In the presence of a lagged dependent variable as a regressor, however, the least squares-based estimators may not be consistent. The generalized method of moments (GMM) estimators are the most popular in dynamic panel data estimation. Two issues concerning the GMM estimator are instruments proliferation and the choice of an initial weighting matrix. In this study, we propose alternative estimators that make use of sub-optimal weight matrices in combination with reduced instruments. The small sample performance of these estimators was investigated using a simulation study and an empirical example of panel data of employment for manufacturing firms. The results from Monte Carlo experiments and real data analysis show that the performance of sub-optimally weighted system GMM estimators that utilize collapsed as well as untransformed instruments seems promising.