Dear all,
I am trying to estimate a selection model with endogenous regressor.
Following Wooldridge(2002), I include the mills ratio in the reduced form of my endogenous variable. Suppose Y2 is my endogenous variable, then I estimate the following reduced form:
reg Y2 on X, Z, IMR
Where X are the exogenous regressors appearing in the system, Z my identifying instruments, and IMR the inverse mills ratio.
I would like to test whether my instruments are weak or not. Thus I was thinking to apply the rule of thumb that the first stage F-statistics on the identifying instruments (Z) is greater than 10 [Staiger and Stock (1997)]. However, I noticed that including or not the mills ratio in the reduced form crucially affects the result. I would like to know:
First, does anyone know if the standard approach of Staiger and Stock (1997) is appropriate in this context?
Second, the F-test has to be performed on a reduced form including or excluding the IMR?
Thank you in advance for any help,
Sara Borelli
__________________________________________________
Do You Yahoo!?
Poco spazio e tanto spam? Yahoo! Mail ti protegge dallo spam e ti da tanto spazio gratuito per i tuoi file e i messaggi
http://mail.yahoo.it
*
* 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/
