Re: st: re: robust est of between estimator

[Kit Baum <baum@bc.edu>]

No; random effects estimation, as an textbook description will show, 
involves both the within and between estimators, so that the full 
dataset is required (not merely time averages). But as help xtgee 
shows, the random effects model can be estimated by this command, which 
does have a robust option: at least the output produces "semi-robust" 
standard errors. Suggest reading [xt] xtgee for the gory details.

webuse grunfeld
xtreg invest mvalue kstock, re
xtgee invest mvalue kstock, family(gauss) link(id) corr(exchangeable) 
i(company)
xtgee invest mvalue kstock, family(gauss) link(id) corr(exchangeable) 
i(company) robust

Kit Baum, Boston College Economics
http://ideas.repec.org/e/pba1.html

On Jan 22, 2005, at 3:56 AM, Dawit Lidia wrote:


> Dear Kit Baum,
>
> Can we also apply the collapse command for random effect model to get 
> robust standard errors?
>
> Thanks
> Dawit
>
> > From: Kit Baum <baum@bc.edu>
> > Reply-To: statalist@hsphsun2.harvard.edu
> > To: statalist@hsphsun2.harvard.edu
> > Subject: st: re: robust est of between estimator
> > Date: Fri, 21 Jan 2005 17:11:30 -0500
> >
> > Sandy wrote
> >
> > I am using the between effects model in xtreg, but would like to have 
> > a
> > robust estimation of standard errors .
> > The robust option is not available for xtreg.
> > I have searched extensively for the option, to no avail.
> > Has anyone programmed this procedure?
> >
> > The between estimator is OLS on the time averages of the panel units. 
> > If you collapse your data on the time variable, you may use regress, 
> > robust.
> >
> > e.g.
> >
> > webuse grunfeld
> > xtreg invest mvalue kstock, be
> > collapse invest mvalue kstock, by(company)
> > reg invest mvalue kstock
> >
> > Now you can do the latter reg with the robust option.
> >
> > Kit Baum, Boston College Economics
> > http://ideas.repec.org/e/pba1.html
> >
> > *
> > *   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/
> _________________________________________________________________
> Last ned MSN Messenger gratis http://www.msn.no/computing/messenger - 
> Den korteste veien mellom deg og dine venner
*
*   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/