Re: st: Fixed effects panel in presence of heteroscedasticity and autocorrelation

[Volker Ludwig <Volker.Ludwig@mzes.uni-mannheim.de>]

Hi Natalie,

xtreg ... , fe vce(robust) and xtreg ... , fe vce(cluster idvar) give 
identical results.
The s.e. are robust against both heteroscedasticity and autocorrelation.
Stock / Watson (2008, Econometrica) propose alternatives that give more 
efficient s.e. in case there is no serial correlation or the correlation 
can be modelled by a moving average.

Best
Volker


Am 21.06.2011 17:13, schrieb Natalie Trapp:
> Dear Statalist,
>
> I would like to estimate a panel model in presence of 
> heteroscedasticity and autocorrelation with individual effects.
>
> However, I am unsure whether the estimator "xtreg..., fe vce(cluster 
> panelvar)" is consistent in presence of heteroscedasticity and 
> autocorrelation? In Cameron and Trivedi (2005, p. 335) I found that 
> "xtscc" could give consistent estimates of the VCE.
>
> I would appreciate any help!
>
> Best regards,
> Natalie
> *
> *   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/


--
________________________

Volker Ludwig

Universität Mannheim
Mannheimer Zentrum für Europäische Sozialforschung (MZES)
D-68131 Mannheim
phone +49-621-181-2849
fax +49-621-181-2803
vludwig@mail.uni-mannheim.de  

*
*   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/