do we have user-command like reghdfe for logit or probit models?

[Fred Lee]

Thanks for your infomation!

[Hong Il Yoo]

If you're interested in estimating -logit- and -probit- with standard errors adjusted for multi-way clustering (with usual caveats about inconsistency in mind), you can use -vcemway-. See -ssc install vcemway- and the background paper in this [external link] to the Stata Journal.

If you're interested in estimating extensions of -logit- and -probit- that account for high-dimensional multi-way fixed effects, you can look up "incidental parameter problem" in your preferred learning resources.

[Joro Kolev]

There is no fixed effects Probit, so no.

There is fixed effects Logit, but I am not aware of two way fixed effects version, so I think the answer is again no.

[Fred Lee]

If you're interested in estimating -logit- and -probit- with standard errors adjusted for multi-way clustering (with usual caveats about inconsistency in mind), you can use -vcemway-. See -ssc install vcemway- and the background paper in this [external link] to the Stata Journal.

If you're interested in estimating extensions of -logit- and -probit- that account for high-dimensional multi-way fixed effects, you can look up "incidental parameter problem" in your preferred learning resources.

Thanks for your reply!
I wnat to find out a user-written command to account for high-dimensional multi-way fixed effects for logit models.

[Fred Lee]

There is no fixed effects Probit, so no.

There is fixed effects Logit, but I am not aware of two way fixed effects version, so I think the answer is again no.

I see, thanks for your reply!