| ![](/includes/contimages/spacer.gif) |
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: RE: Doubt about xtivreg,first differences
With -xtivreg-, everything is in first differences, that is,
y means d.y, l.y means ld.y and l2.y means l2d.y. Therefore,
twice-lagged differences are used as instrument for the
lagged differences and 3 waves are lost. If you want to use
twice-lagged y as an instrument, use -ivreg-. For example:
. ivreg d.y d.x (ld.y = l2.y)
Note that
. ivreg d.y d.x (ld.y = l2d.y)
estimates the exact same model as
. xtivreg y x (l.y=l2.y), fd
ben
> -----Original Message-----
> From: [email protected] [mailto:owner-
> [email protected]] On Behalf Of Rafael Terra de Menezes
> Sent: Monday, February 20, 2006 8:10 PM
> To: [email protected]
> Subject: st: Doubt about xtivreg,first differences
>
> Dear Stata users,
>
>
> I'm usin the command xtivreg, fd and some doubts arose while working with
> it. First, I ran the following:
>
> xtivreg y w y z (l.y=l(2).y), fd
>
> My panel has T=10 and the output shows T=7. I know first differencing the
> equation sweep out one period, and because there is a lag dependent
> variable, two years are swept out. I would like to know if the fact that
> I'm
> using yt-2 as instrument makes me lose another year.
>
> Another question: Is there any test to run after xtivreg,fd (test for
> validity of instruments for example)? And how can I deal
> heteroskedasticity
> in xtivreg?
>
> Thanks,
> Rafael Terra de Menezes
> University of S�o Paulo - Brazil
> Ms Student
>
>
> *
> * 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/