Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

RE st: two-step estimation with mlogit


From   ACavallo@lexecon.com
To   statalist@hsphsun2.harvard.edu
Subject   RE st: two-step estimation with mlogit
Date   Fri, 5 Dec 2003 10:02:56 -0600




If you are trying to do selection bias correction as in Lee (1983) then you
may want to look at the SELMLOG command.  Here's the description:

selmlog estimates linear regression models on a selected subset of
observations, where selectivity is modelled as a multinomial logit (as
opposed to univariate probit as in the Heckman model). It applies the
two-step method proposed by Bourguignon, Fournier and Gurgand (CREST, 2001,
www.crest.fr).

See http://www.crest.fr/pageperso/lmi/gurgand/selmlog.htm



Regards,

--Alex Cavallo
Lexecon
(312) 322-0208  voice
(312) 322-0218  fax

---------------------------------------------------------------------------------------------------------------------------------------------------------------
Fatma Bircan
Middle East Technical University
Department of Economics
210-2056


Hello,

I have been trying to do two-step estimation using mlogit.  After
mlogit estimation, I calculate the lambda values for each choice as
follows;

pj= exp(v'b)   where v is the set of explanatory variables in the
selection    equationand b is the set of coefficients. j=0,1,...

Hj=invnorm(pj)

lambdaj=normd(Hj)/normprob(Hj)

and then including the lambda values, I estimate the equation of
primary interest which is a wage eqution in my case.

However, these estimation results do not give the efficient standard
errors. My question is that is there a stata rutin to obtain efficient
standard errors? How can I do the Heckman-correction to obtain
efficient standard errors.

Tahnk you in advance


*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index