Do you have any suggestions? ivprobit does not allow for the inclusion of random terms, and meglm / gllamm will not do the trick either as instrumental variables cannot be incorporated.
What I want to estimate is similar to these 3 options, where the bold part before the hyphen indicate what is (from my understanding) impossible:
Alternatively, I could imagine that the the dichotomous-outcome-equivalent of xtivreg, re
What I want to estimate is similar to these 3 options, where the bold part before the hyphen indicate what is (from my understanding) impossible:
Code:
XT-ivprobit / IV-meglm dependentvariable level1controls level2controls (endogenouslevel2variable = level2instrument) || Country:
Code:
xtset Country Dichotomous-xtivreg dependentvariable level1controls level2controls (endogenouslevel2variable = level2instrument) , re
Comment