Across many disciplines, the fixed effects estimator of linear panel data models is the default method to estimate causal effects with nonexperimental data that are not confounded by time-invariant, unit-specific heterogeneity. One feature of the fixed effects estimator, however, is often...

Across many disciplines, the fixed effects estimator of linear panel data models is the default method to estimate causal effects with nonexperimental data that are not confounded by time-invariant, unit-specific heterogeneity. One feature of the fixed effects estimator, however, is often...

We investigate the strict exogeneity assumption, a necessary condition for estimator consistency in many finance panel settings. We outline tests for strict exogeneity in both traditional (non-IV) and IV settings. When we apply these tests in common traditional finance panel regressions, we find...