st: Recursive biprobit not converging

["Bush, Susan" <Susan.Bush@occ.treas.gov>]

I've got a question regarding discrete choice modeling. I am using a
recursive bivariate probit model for a paper that I'm working on. This
model follows the form:

First equation: Y1 = Y2 X3
Second equation: Y2 = X1 X2
(where Xi are independent variables and Yi are dependent variables)

I'm using the STATA command:
biprobit(Y2=X1 X2)(Y1=Y2 X2)

When I do this, I am finding that the model does not converge. Is this a
fatal problem in my modeling? Do I have to now move on to some
estimation technique that doesn't rely on maximum likelihood? STATA is
still able to generate estimates and the model fits the data fairly well
(predicts correctly nearly 75% of the time).

Thanks!
Susan

P.S. I am unable to use the ivprobit model because in all cases where
Y2=0, Y1=0



*
*   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/