RE: st: No F-stat reported in xtreg with clustered standard errors

["Schaffer, Mark E" <M.E.Schaffer@hw.ac.uk>]

Doug,

> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-
> statalist@hsphsun2.harvard.edu] On Behalf Of Douglas Barthold
> Sent: 03 January 2013 18:47
> To: statalist@hsphsun2.harvard.edu
> Subject: RE: st: No F-stat reported in xtreg with clustered standard
errors
> 
> Hi Austin, thanks for your help.  I have two follow-up questions.
> 
> I understand the problem with the degrees of freedom, but I why does
Stata
> report an F-stat with region fixed effects included, but not when I
also
> include year fixed effects?  (Clustering at the region level, and
including a
> number of other linear constraints).
> 
> The partial option seems helpful.  As noted in the -ivreg2- help
section, "The
> partial option is most useful when using cluster and #clusters <
(#exogenous
> regressors + #excluded instruments)".  The problem is, there is no
partial
> option with xtivreg2, which I need for my fixed effects. 

xtivreg2 supports the partial option, or at least the version I have on
my machine at home does:

. which xtivreg2, all

c:\ado\personal\xtivreg2.ado
*! xtivreg2 1.0.13 28Aug2011
*! author mes

--Mark

> Any ideas for this?
> 
> Thanks,
> Doug
> 
> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-
> statalist@hsphsun2.harvard.edu] On Behalf Of Austin Nichols
> Sent: Wednesday, January 02, 2013 7:42 PM
> To: statalist@hsphsun2.harvard.edu
> Subject: Re: st: No F-stat reported in xtreg with clustered standard
errors
> 
> Douglas Barthold <douglas.barthold@mail.mcgill.ca>:
> The VCE does not have enough degrees of freedom; see
> http://ideas.repec.org/p/boc/usug07/07.html
> which means even if it were feasible to jointly test all those coefs,
it would
> be a very unreliable test (F stat overstated).
> You can partial out the region and year fixed effects using -ivreg2-
or -
> xtivreg2- (SSC); see help on the partial() option.
> That assumes that you don't care about those coefs; assume they are
all
> nonzero so you are not interested in testing their joint
> signifcance: instead of comparing to an intercept-only model as in
regular
> OLS, you are compariing to a many-intercept model.
> 
> On Wed, Jan 2, 2013 at 4:30 PM, Douglas Barthold
> <douglas.barthold@mail.mcgill.ca> wrote:
> > Dear Statalist,
> >
> > I'm trying to run panel regressions using xtreg with clustered
standard
> errors.  I'm using xtreg because I want to include region and year
fixed
> effects.
> >
> > My problem is that when I include year fixed effects, Stata won't
report an
> F-Stat, which I need.  With just region fixed effects, it works.  Any
ideas why
> this is happening?
> >
> > Thanks!
> > Doug
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/
> 
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/


----- 
Sunday Times Scottish University of the Year 2011-2013
Top in the UK for student experience
Fourth university in the UK and top in Scotland (National Student Survey 2012)

We invite research leaders and ambitious early career researchers to 
join us in leading and driving research in key inter-disciplinary themes. 
Please see www.hw.ac.uk/researchleaders for further information and how
to apply.

Heriot-Watt University is a Scottish charity
registered under charity number SC000278.


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/