for a research project I am looking at a model to estimate a fixed effects logit model of the type

$$y_{it} = \Lambda(x_{it}'\beta +w_{it}'\theta +a_i) .$$

Now I know that I would usually do that in a conditional logit model using MLE. However I now have an binary instrument $z_{it}$ for the endogenous and also binary regressor $w_{it}$. I suspect I have to use GMM and a version of the moment condition $E[z'(\Delta(x_{it}'\beta +w_{it}'\theta +a_i))]$, but I am not sure how to transform the logistic function into a conditional logit one. I hope the problem is clear. Can anybody help me out?

  • $\begingroup$ Should that delta be a lambda? $\endgroup$ – Dimitriy V. Masterov Jan 30 '18 at 1:44
  • $\begingroup$ Yes, edited.... $\endgroup$ – Leon F. Jan 30 '18 at 20:45
  • $\begingroup$ What about the one in the moment condition? $\endgroup$ – Dimitriy V. Masterov Jan 30 '18 at 22:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.