Hi Tinna,
If I understand your problem correctly and you are willing to use the
Heckman Two-Step procedure (instead of ML), you could split the
Heckman into the individual steps, and manually compute the inverse
Mills ratio.
The first step is the selection model, from which you can compute the
inverse Mills, and include that in your IV regression. Code follows:
probit WorkedFullTime XVector
predict Z, xb
gen InvMills = normden(Z) / norm(Z)
ivreg2 HoursWorked XVector InvMills
--Ketan Patel
On 10/31/05, Tinna <[email protected]> wrote:
> Dear statalisters,
>
> I need to control for selection in the model I am estimating (e.g.
> -heckman-), but I also suspect endogeneity (-ivreg2-). (see further in
> P.S. if needed)
>
> Is there a command that allows me to do both at the same time?
>
> Any suggestions or thoughts are appreciated.
>
> Tinna
>
> P.S. Basically I am estimating an hours worked equation, where I only
> have hours worked for those who's main activity is working (as opposed
> to being a student, disabled, retired, homemaker...). It is the
> coefficient of interest in this equation that is likely to be biased
> if not instrumented.
>
> *
> * 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/
