The Heckman model assumes that
e1 ~ N(0, sigma)
e2 ~ N(0, 1)
See, [R]A-J Heckman, page 452.
Scott
> -----Original Message-----
> From: [email protected] [mailto:owner-
> [email protected]] On Behalf Of Jelena Zurovac
> Sent: Wednesday, January 18, 2006 9:35 PM
> To: [email protected]
> Subject: st: covariance of error terms in heckman selection model
>
> Hi, I would greatly appreciate any input on my question.
>
> Consider a simple selection model:
>
> Q = aX + e1
> M* = bx + cz + e2
>
> It is estimated in Stata using the 'heckman' command. The parameters that
> are
> estimated are:
>
> sigma=st.deviation(e1)
> rho=corr(e1,e2)
> and
> lambda=rho*sigma.
>
> I am interested in the cov(e1,e2), which can be computed only if we have
> the
> standard deviation of e2. Stata does not report it and it isn't saved in
> e(). Is there a good way around it?
>
> Thank you very much,
>
> Jelena Zurovac
> Graduate Student
> University of Washington, Seattle
>
>
>
>
>
> *
> * 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/
