Hi all,        I have a small qualm. In my understanding, panel fixed effect regressions take account of  endogeneity in assuming that variables are correlated with the error term. Am I right in this assertion?. More so, where do 2SLS and GMM come in to out-perform panel regressions by fixed effects?.       Thank you