Page 1 of 1

Use of IV in logit regression

Posted: Wed Oct 02, 2013 5:00 am
by Mathieu
I want to do something similar to a 2SLS with instrumental variables, the only difference is that my second stage regression is not an OLS but a logit (i.e., binary) regression (first stage is still OLS). I use Eviews8 but I do not see how I can estimate this.

Thanks for you answer!
Mathieu

Re: Use of IV in logit regression

Posted: Wed Oct 02, 2013 7:52 am
by EViews Gareth
There is nothing built in that will do it. Of course you can always estimate the two stages manually.

Re: Use of IV in logit regression

Posted: Wed Oct 02, 2013 8:21 am
by startz
If you do this in two stages, you are going to have to figure out how to correct the standard errors.