[Limdep Nlogit List] 2SLS

[feinf at umich.edu]

Here is a paper on that topic:

J Bhattacharya, D Goldman, D McCaffrey (2006), "Estimating probit 
models with self-selected treatments", Statistics in Medicine, Volume 
25, 3, pp. 389-413.

Abstract at: 
http://healthpolicy.stanford.edu/publications/estimating_probit_models_with_selfselected_treatments/

It suggests that the estimates arising from 2SLS may not be consistent, 
and suggests alternative approaches.

FF

Quoting Gail Pacheco <gail.pacheco at aut.ac.nz>:

>
> Hi
> this may be more of an econometric question, rather than a limdep
> question - but is it possible to use limdep to run 2SLS if the dependent
> variable is binary in nature?
> Any help, in terms of appropriateness of doing this and the necessary
> command would be appreciated??
>
> thanks
>
> Gail Pacheco
> Senior Lecturer - Economics
> Faculty of Business
> AUT University
> gail.pacheco at aut.ac.nz
> 64-9-9219999 ext 5708
> _______________________________________________
> Limdep site list
> Limdep at limdep.itls.usyd.edu.au
> http://limdep.itls.usyd.edu.au
>
>
>