Econometric analysis of dynamic panel data models: A growth theory example

Recent work on the econometrics of panel data with fixed effects has emphasized the value of seeking an orthogonal reparametrization of those effects. This is a redefinition of the fixed effects to try to separate them, in the likelihood, from the parameters of interest. The aim of this paper is to show how to do this, and to carry out an actual, if rather stylized, analysis of real data using a model of some interest to the profession. We wish to show that the method works and how it works and to make the approach accessible to applied economists.