xtpoisson fixed effects when both endogenous variable and IV are binary

[NamJi Eum]

Dear all,

I am using Stata 15.1 and I would like to conduct panel poisson iv regressions for 67 countries across 28 quarters.
Specifically, I would like to conduct xtpoisson with country-quarter fixed effects and cluster robust standard errors within each country.
I found a solution to this when x variable is continuous on this thread (https://www.statalist.org/forums/for...-fixed-effects).

However, in my case, both my x variable and instrument variables are binary.
In this case, what changes should I make?

I would greatly appreciate any advice on this matter.

Thanks,

Nam Ji