Hi,

Considering one has panel data and an endogenous regressor (correlated with the time-varying error term), what is the advantage of using a fixed-effects estimator in addition to an instrumental variable approach?

Thanks in advance!

Joep