On Mon, 09 Apr 2007 12:07:05 -0500, Thi Minh Ngo <[email protected]> wrote:
I am using ivreg2 to estimate an equation with a single endogenous regressor
and 6 excluded instrumental variables. I encounter the problem of weak
instrumental variable as explained in Andrews and Stock (Inference with Weak
Instrument, 2005).
To compute confidence intervals that are robust to weak instruments, I am
trying to use the used-written programme called condivreg. However, when I
do, I obtain the error message "Multicollinearity!".
The multicollinearity seems to come from the fact that the instruments are
constructed as the interaction between two variables (A*B), and A and B are
also included in the second-stage regression as exogenous regressors. The
estimation is therefore written as follows:
xi: condivreg Y1 (Y2=i.A*B) X i.A B
Would someone know why condivreg cannot handle such a specification?
Thank you very much for your help,
Thi Minh
Thi Minh,
If you could email your dataset to me ([email protected]), I'd be happy to
look into this for you.
-- Brian Poi
-- [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/
