Dear Bidisha,
I don't think that you can decompose the estimates of mlogit because you
would have to calculate the conditional expectation, i.e.
E(y|x) = Pr(y=1)*E(y|y=1) + Pr(y=2)*E(y|y=2) + .
Since the values of the dependent variable (y=1,2,.) are chosen arbitrarily,
you will not find an unambiguous solution for that. You should think about
using models for binary choice variables (see help logistic estimation
commands). Let me know if you have further questions or comments.
Mathias
------------------------------
Date: Thu, 26 Apr 2007 14:36:54 +0100
From: "Bidisha Sayema" <[email protected]>
Subject: st: Decomposition of Multinomial logit
Dear Statalisters,
I am using multinomial logit model of occupation choices (professional,
skilled, non skilled) while using various dummies of race (white, black,
asian etc) as explanatory variables. I want to do a decomposition
(Oaxaca) to see what percentage of the difference in probabilities of
occupational choice (eg. unskilled) of a minority group from that of a
white can be explained by the endowment and what percentage by
coefficients (due to discrimination). I read that in case of non-linear
model there is different methodology for decomposition and I have seen
the one by Bauer & Sinning (Bauer T. & Sinning M. (2005):
"Blinder-Oaxaca Decomposition for Tobit Models", IZA Discussion Paper
no. 1795) but I think it does not include Mlogit option. It would be
greatly appreciated if someone can suggest me about the decomposition of
Mlogit output.
Many Thanks
Bidisha
This message has been checked for viruses but the contents of an attachment
may still contain software viruses, which could damage your computer system:
you are advised to perform your own checks. Email communications with the
University of Nottingham may be monitored as permitted by UK legislation.
*
* 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/
------------------------------
*
* 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/