Using the control function approach makes IV logit easier (i.e. save
residuals from first stage, include as regressor in second stage, fix SE).
You might want to check out Wooldridge (2002) _Econometric Analysis of Cross
Section and Panel Data_ (see: 15.7.2 Continuous Endogenous Explanatory
Variables; 15.7.3 A Binary Endogenous Explanatory Variable) from your local
library for details.
Table of Contents at
http://www.stata.com/bookstore/cspd.html#contents
N.B. The IV estimator in Stata is not technically a two-stage estimator,
though the result is equivalent to 2SLS.
-----Original Message-----
From: Herve STOLOWY [mailto:[email protected]]
Sent: Thursday, September 02, 2004 11:29 AM
To: [email protected]
Subject: st: Two-stage logit regression
Dear All:
I would like to perform a two-stage logit (or logistic) (and not
least-squares) regression.
Do you know if there is in Stata an equivalent to the ivreg procedure?
Thanks a lot.
Herve Stolowy
Dept of Accounting and Management Control
HEC Paris
*
* 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/
