Dear Statalist,
I have the following question. I have a simultaneous
equations system of the form:
1) D= b1*Xd + b2*W + e for i=1....N
(D) equation for whole sample
2) W= a1*Xw + a2*D + u for i=1...n < N
(W) equation for part of sample
I observe D over the whole sample, where D can be
positive or zero. But I observe W only for a part of
the sample, W is missing for the rest of observations.
I want to estimate eqn (1) over the whole sample, but
if I use ivtobit then stata uses only observations for
which both D and W are observed.
Thus, I have first estimated the reduced form W
equation and predicted W for the whole sample, and
then used the predicted W in the second stage and done
tobit. Does any one have any suggestion-reference on
whether this might be correct or any suggestion on how
to do this in stata?
thank you
Sara
___________________________________
L'email della prossima generazione? Puoi averla con la nuova Yahoo! Mail:
http://it.docs.yahoo.com/nowyoucan.html
*
* 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/
