Ron,
Can you describe the two-step estimation procedure you're using and
what you're doing with -treatreg-?
-treatreg- is a system estimator for a 2-equation system in which
there is an equation for an endogeneous 1/0 variable, and an equation
for a continuous dep var in which the endog 1/0 variable appears on
the RHS.
If you run -treatreg- with the -twostep- option, it first estimates
the equation for the 1/0 variable using a probit, and then estimates
the equation for the continuous dep var a la Heckman. This sounds
like what you're doing by hand, so I'm not sure why there should be a
difference. Perhaps you ran -treatreg- without the -twostep- option?
This would mean -treatreg- is doing a maximum likelihood estimation
of the system, and so the results would indeed be different from a
two-step estimation.
--Mark
Date sent: Fri, 8 Aug 2003 11:54:12 -0400 (EDT)
From: Ronald Rutherford <[email protected]>
To: "'[email protected]'" <[email protected]>
Subject: st: Date: Fri, 8 Aug 2003 10:54:10 -0500
Send reply to: [email protected]
> Can anyone tell me why I get significantly different results using a two
> step probit/linear model for endogenity corrections versus the results
> obtained using the same set of regressors in a treatreg model?
>
> Ron Rutherford
>
> *
> * 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/
Prof. Mark E. Schaffer
Director
Centre for Economic Reform and Transformation
Department of Economics
School of Management & Languages
Heriot-Watt University, Edinburgh EH14 4AS UK
44-131-451-3494 direct
44-131-451-3008 fax
44-131-451-3485 CERT administrator
http://www.som.hw.ac.uk/cert
*
* 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/
