st: Heckman selection w mprobit in first stage

[Ada Ma <heu034@gmail.com>]

Dear Statalist

I would like to estimate a Heckman selection model where the first
stage is a multinomial probit selection and the second stage is just a
continuous variable.

More specifically the first stage is whether a person is choosing to
work or retire.  The reference group is "out of labour force".

Second stage dependent variable is either hourly earnings or number of
hours worked per week.

The group of people I am looking at are aged 50+ so I think it would
be important to model the retirement decision as well as the
employment decision.

Is there a ready made Stata command for this model?  Or a
demonstration of how to use existing commands to estimate this model?

Many thanks for your help.

Best regards,
Ada

*
*   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/