Dear listers,
When using ivreg or ivreg2 to do a 2SLS estimation, all the RHS
variables except those explicitly specified as endogenous are assumed
to be exogenous and valid IV. Call those the "included exogenous
variables". However, I happen to have a case in which not all the
"included exogenous variables" are valid IV, and I am asking whether
users can have better control over the list of IV to be used in the
2SLS estimation.
The model I am dealing with is from the precautionary saving
literature, in which a commonly used regression model is something
like:
ln(W/P) = a0 + a1*ln(P) + a2*y + B*X,
where W is wealth, P is permanent income, y is the endogenous
variable, and X are all other exogenous variables. Note that ln(P) is
on the LHS as well as on the RHS; the literature has it on the RHS to
"control for heterogenous preference".
The task is to control for the endogeneity of y and leave ln(P)
alone, as most of the studies in this literature claimed to have
done. With ivreg or ivreg2, all the RHS variables except y are
automatically included in the IV list. However, ln(P) is clearly not
a valid IV. Therefore, I would like to have a IV list that does not
include ln(P). This problem can be solved if users can have the
freedom in specifying the IV list, but apparently ivreg and iverg2
are not written based on this idea.
I have been thinking of specifying both y and ln(P) as the endogenous
variables in order to treat the program, but I am afraid the result
is NOT what I (and the literature) set out for: control the
endogeneity of y and leave ln(P) alone.
Any thought?
Eddy
__________________________________
Do you Yahoo!?
Yahoo! Photos: High-quality 4x6 digital prints for 25�
http://photos.yahoo.com/ph/print_splash
*
* 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/
