Unfortunately, you can't instrument X*Y with X_hat*Y. It will not
generally produce consistent estimates or valid t-statistics. Wooldridge
refers to it as the "forbidden regression." Instead, you have to
instrument for (X*Y), which can--unfortunately--be a real pain in terms
of finding enough valid instruments. This is separate from the issue of
correcting s.e.'s after doing 2SLS "manually".
Wooldridge [ECONOMETRIC ANALYSIS OF CROSS SECTION AND PANEL DATA, 2002,
page 236-237] is a helpful reference on this point.
Glenn
Glenn Hoetker
Assistant professor of strategic management
University of Illinois at Urbana-Champaign
[email protected]
217/265-4081
> -----Original Message-----
> From: [email protected] [mailto:owner-
> [email protected]] On Behalf Of Silje Aslaksen
> Sent: Tuesday, November 22, 2005 7:50 AM
> To: [email protected]
> Subject: st: Date: Tue, 22 Nov 2005 14:39:05 +0100
>
> Hello!
>
> I am trying to estimate the following model using instrumental
> variables.
> The final stage has two independent variables of interest, plus an
> interaction between them:
>
> Z = aX + bY + c(X*Y) + more variables
>
> I want to instrument variable X in a first stage regression using a
set
> of
> instruments that contain Y and some other variables that are not in
the
> final stage. Of course the variable X*Y will have to reflect this
> instrumenting as well.
>
> I want to instrument for X with just M (excluding MY), and
> then substitute Xhat*Y for XY in the second stage. I know that I have
to
> correct standard errors "manually".
>
> I have done this and I have corrected standard errors "manually". My
> question is how I can perform a over-identification test (test that my
> instrument only affects Y through X) when I don`t use the ivreg
command?
>
> Silje
>
> [email protected]
>
>
> *
> * 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/
*
* 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/
