Hi everyone,
I am estimating a country-level gravity model with one endogenous regressor using ivpoisson command (ivpoisson gmm). From what I understand, ivpoisson gmm command doesn't compute a first stage. My question is, is it possible to simply regress the endogenous variable on the instrument and other exogenous variables to mimic a first stage regression and see if the instrument is relevant or not? If yes, should I use ppml or ols estimator? Also, should the error terms be clustered by country for the first stage (as the second stage)?
Thanks!