GQL estimation in linear dynamic models for panel data

Abstract

In the econometrics literature, it is standard practice to use the existing instrumental variables as well as generalized method of moments approaches for the estimation of the parameters of a linear dynamic mixed model for panel data. In this paper, we introduce a generalized quasi-likelihood estimation approach that produces estimates with smaller mean squared errors when compared with the aforementioned and other existing approaches.

Keywords:

• efficiency
• generalized quasi-likelihood (GQL)
• linear dynamic model
• normality-based weight matrix
• unknown error distribution

Acknowledgements

This research was supported by a grant from the Natural Sciences and Engineering Research Council of Canada. The authors thank the referees for valuable comments on the earlier version.