st: RE: Hausman test (fe vs. re), use robust SE or not?

[DE SOUZA Eric <eric.de_souza@coleurope.eu>]

The standard Hausman test supposes that RE is efficient under the null. Robustifying violates that assumption. 
On the other hand, the fact that robustifying gives different results means that the RE assumptions are not satisfied.
Consequently, the use of the test is invalid in your case, robustification or not

Wooldridge in his 2002 textbook, Econometric Analysis of Cross Section and Panel Data, has a discussion of this and also of a robust version of the Hausman test (page 288ff). But this would have to be programmed.

An alternative is to use xtoverid: -findit xtoverid- programmed by Mark Schaffer and Steven Stillman.


Eric de Souza
College of Europe
Brugge (Bruges)
Belgium

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Hobst
Sent: 17 August 2010 14:13
To: statalist@hsphsun2.harvard.edu
Subject: st: Hausman test (fe vs. re), use robust SE or not?


Hello

I want to decide between using the fixed effects or the ramdom-effects model.

I did a hausman test xtreg y x1 x2 x3, fe estimates store FE xtreg y x1 x2 x3, re estimates store RE hausman FE RE

i get p =0.000, so i should use FE model.

But if add the vce(robust) option to the two xtreg lines above and test it again i get p=0.91

So my question: Do i need to use the vce(robust) option for the hausman test or not?

Thank you very much for your help!

Regards
--
View this message in context: http://statalist.1588530.n2.nabble.com/Hausman-test-fe-vs-re-use-robust-SE-or-not-tp5431784p5431784.html
Sent from the Statalist mailing list archive at Nabble.com.
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/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/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/