I cannot say anything about inference, but note that
-xtivreg- and -xtlsdvc- implement different versions
of the Anderson-Hsiao estimator.
The basic AH model in -xtivreg- is:
ivreg d.y d.x (ld.y = l2d.y)
(i.e. -xtivreg y x (l.y =l2.y), fd-).
In contrast, the basic AH model in -xtlsdvc- is:
ivreg d.y d.x (ld.y = l2.y)
The latter may be preferred since it saves one wave
of data.
ben
> -----Original Message-----
> From: [email protected] [mailto:owner-
> [email protected]] On Behalf Of Marcello Pagano
> Sent: Thursday, June 08, 2006 4:20 AM
> To: [email protected]
> Subject: st: Anderson-Hsiao
>
> > Subject: Anderson-Hsiao
> > To: [email protected]
> >
> > I would like to implement the Anderson-Hsiao
> > estimator > in stata for my panel data set. I had though of
> > doing this using xtivreg2, fd, but according to a previous
> > posting
> >
>
> (http://newwww.stata.com/statalist/archive/2004-10/msg00522.html)
>
> >> this command cannot be used for inference and
> >> hypothesis testing. The only thing I found in the
> >> help
> >> files was the xtlsdvc command which has a
> >> Anderson-Hsiao option, but this takes forever with
> >> large datasets (I still haven't been able to get any
> >> results out) and also doesn't give much flexibility.
> >>
> >> Is xtivreg2, fd an appropriate way of implementing
> >> Anderson-Hsiao? Is there any particular way of
> >> specifying it that is necessary? Are the results
> >> useful at all (given the points made in the posting
> >> I
> >> linked to above)? If not, is there any other way of
> >> implementing Anderson-Hsiao in STATA?
> >>
> >> Thank you very much for any help!
> >>
> >> Debbie
> >
> *
> * 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/