Some of these questions have been asked before,so
please accept my apologies. But I was hoping for some
added insights, and I also wanted to make sure my
understanding was correct.
I have two equations:
Y1= a*Y2+b*X1
Y2= c*Y1+d*X2
where Y1 and Y2 are binary variables.
First, since these are non recursive equations, does
this mean biprobit cannot be used? Or can biprobit
still be used akin to sure to take into account
possible correlation in error terms across the two
equations? To do this do I need to use the robust
option along with biprobit?
Second, is ivprob the more appropriate way to estimate
these equations to take into account endogeneity? Or
is ivprob only suitable when one endogenous variable
is binary and the other continuous? If the latter is
the case (and biprobit cannot be used), what is the
appropriate method/ command to use to estimate non
recursive equations with endogenous binary variables?
Any help would be much appreciated!
Shyam
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
*
* 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/
