The factor analytical method for interactive effects dynamic panel models with moving average errors

The estimation of dynamic panel data models with interactive effects and moving average errors is considered. This is accomplished by making an extension to the factor analytical (FA) estimator which was originally designed for dynamic panels with fixed effects only and serially uncorrelated errors. The results show that the additional allowances have no effect on the asymptotic properties of the FA estimator. In particular, the asymptotic distribution of the estimator is free of the otherwise so common bias problem, a result that is verified in small samples using Monte Carlo simulation.