Belated reply: when unusual combinations of the types/natures of the variable arise in IV setting, -cmp- from SSC is the solution in 95% or more of the cases. Such a recommendation comes out about once a month from me!
More: -mtreatnb- from SJ 6-4 works with fem_work count and other_inc multinomial (treatment-effects model, not really IV)
-selmlog- from http://www.delta.ens.fr/gurgand/ works with fem_work continuos and other_inc multinomial (sample selection, i.e. a la -heckman-, not really IV)
Nicola
P.S. I'll NOT receive/read any email but the Digest.
At 02.33 24/03/2009 -0400, Nirina F wrote:
>Dear all,
>I would like to run mlogit with exclusion restrictions (like ivprobit)
>but I don't see it in the stata help.
>Like in ivprobit, we would use the following:
>webuse laborsup
>ivprobit fem_work fem_educ kids (other_inc = male_educ)
>
>Now if the dependent variable is a categorical variable not a binary
>then how do I instrument other_inc with mlogit?
>
>Also, what is the advantage of using biprobit over mlogit if say my
>endogenous variable other_inc was actually a binary variable?
>
>Thank you in advance,
>Nirina
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
