Hannah Kewsberry wrote:
> I would like to know if anyone could kindly advise me whether we have
> commands in Stata to work out discrete choice selection for multinomial
> logit? I welcome all suggestions, Thanks!
>
If you are talking about a sample seletion model where the selection
equation is multinomial logit (i.e. more than two choices) then I have
estimated such a model using Stata.
We estimate a residential choice equation using multinomial logit (clogit)
and generate the sample selection correction terms to include in a housing
demand equation. Since there are 11 choices in the first stage, there are
eleven "Heckman" terms in the housing demand equation. We obtained the
functional form for the "Heckman" terms from Dubin
(1985) "Conditional Moments in the Generalized Extreme Value Family."
Dubin and
McFadden (1984) in Econometrica is also relevant. We programmed the
"Heckman" terms ourselves. I am not aware of a Stata program that does
this directly.
******************************************************************
Professor Jeffrey Zabel
Department of Economics
Tufts University
Medford, MA 02155
Phone: 617-627-2318
Fax: 617-627-3917
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
