I show that a class of fixed effects estimators is reasonably robust for estimating the population-averaged slope coefficients in panel data models with individual-specific slopes, where the slopes are allowed to be correlated with the covariates. In addition to including the usual fixed effects estimator, the results apply to estimators that eliminate individual-specific trends. Further, asymptotic variance matrices are straightforward to estimate. I apply the results, and propose alternative estimators, to estimation of average treatment in a general class of unobserved effects models.
On the robustness of fixed effects and related estimators in correlated random coefficient panel data models
1 May 2005