Hi, Mark,
Your answers are very helpful. I do use the "robust" option in the
ivreg2. According to the Vince Wiggins' discussion, I think I can just
ignore the error message. I figure out the the error comes from the
fact that one exogenous dummy variable has only one "1".
Thank you very much.
Binzhen Wu
----- Original Message -----
From: Mark Schaffer <[email protected]>
Date: Thursday, May 5, 2005 5:27 am
Subject: Re: st: a error message in the first stage regression in the
ivreg
> Binzhen,
>
> Date sent: Thu, 05 May 2005 00:35:10 -0500
> From: BINZHEN WU <[email protected]>
> Subject: st: a error message in the first stage
regression
> in the ivreg
> To: [email protected]
> Send reply to: [email protected]
>
> > Dear Statalist,
> >
> > I got a error message and cannot figure out the reason. Could you
> > help? I am runing the following model: Y=aT+bZ+cTZ+dX, Here,
TZ=T*Z
> > is the interaction term, T is the time dummy (=1 after the policy
> > change), and Z are the continuous endogenous treatments, X are
> > exogenous variables.
> >
> > The command I use is: ivreg2 Y T X (TZ Z=W V TW TV)
> >
> > I got an error message for the first stage regression:
> >
> > Error: covariance matrix of moment conditions not of full rank;
> > specification
> > test stats not reported and standard errors above may be
> > meaningless.
>
> A couple of ideas:
>
> - Were you using robust or cluster as options? Some phenomena,
> such
> as singleton dummies, can generate this. See Vince Wiggins' very
> nice discussion at
>
> http://www.stata.com/statalist/archive/2003-06/msg00646.html
>
> - Do you get the same error if you drop TZ or Z as endogenous
> regressors? With or without dropping W/V TW/TV at the same time?
>
> - Not all the collinearities may have been dropped up front. My
> recollection is that ivreg2 looks separately for collinearities in
> (T
> X), (TZ Z) and (W V TW TV), but collinearities may subsequently
> emerge that involve variables across these sets.
>
> Hope this helps.
>
> --Mark
>
> > This should be caused by collinearity, because collinearity
> variables> have been automatically dropped by stata.
> >
> > Thank you very much.
> >
> > Binzhen
> >
> > *
> > * 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/
>
> Prof. Mark E. Schaffer
> Director
> Centre for Economic Reform and Transformation
> Department of Economics
> School of Management & Languages
> Heriot-Watt University, Edinburgh EH14 4AS UK
> 44-131-451-3494 direct
> 44-131-451-3296 fax
> 44-131-451-3485 CERT administrator
> http://www.sml.hw.ac.uk/cert
>
> *
> * 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/
