Instrumental Variable Gravity Model

[Jorge Vlachopoulos]

Hi there,

I am trying to run an instrumental variable approach with a gravity model. I am running a regression using PPMLHDFE (since its a panel and I use fixed effects) but now I want to tackle the endogeneity concern using IV. I have not found any command that helps me doing this (ivpoisson was not very helpful and then ivppml didnt really work). I would appreciate any advice!

[Maxence Morlet]

Could you please expand on "ivppml didn't really work"?

Otherwise, you may do 2SLS "manually". I am not sure if the so-called forbidden regression applies to Poisson though.