st: heckman model

[Shehzad Ali <sia500@york.ac.uk>]

Hi listers,

I am trying to estimate a heckman sample selection model using - heckman - 
command in stata. Here is the model:

heckman y1 x1 x2 x3, sel(y2=x1 x2 x3) twostep

Here x1 is an endogenous variable for both my outcome and selection steps 
and I would like to account for its endogeneity. I was wondering if you may 
be able to suggest a one part or two part model whereby I could loop a 
probit for the endogenous variable in both outcome and selection steps. I 
know that one possible solution is to do a two part model with first part 
being a bivariate probit model for selection, and second part being a 
treatreg model for the outcome variable y1. This way I could include x1 as 
a treatment variable in both processes, but I was wondering if there is a 
better solution to this problem.

Any help would be highly appreciated.

Thank you,

Shehzad

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/