Hi statalisters,
I need to estimate a model with two probit equations and "partial
observability": i.e. I observe only the product of the two binary
dependent variables.
However, I have a complication: in each equation I have a continuous
endogenous regressor (for which I have instruments available).
How can I correctly estimate this model, mixing partial observability
with ivprobit features? My first guess would be a two-step procedure
where I include in the partial biprobit the fitted values of the
endogenous regressors derived from a first-stage regression. However,
I guess there are additional SSE corrections that I would omit,
proceeding this way.
Anyone has an idea? I have been struggling with this problem for a few
days, any hint would be enormously appreciated.
Best
Margherita
--
Margherita Comola
Paris School of Economics
Email: [email protected]
http://www.pse.ens.fr/comola/index.html
*
* 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/