The best solution seems -xtscc- (from ssc), which produces Driscoll and Kraay standard errors. The error structure is assumed to be heteroskedastic, autocorrelated up to some lag, and possibly correlated between the groups (panels). There also seems to be no problem if t>n. And, by using the fixed effects, you allow for the endogeneity.
Nicola
P.S. By multiplying the messages you probably *decrease* the chance to get answered (see the FAQ)
At 02.33 04/07/2007 -0400, "Aleksandra Riedl" wrote:
>Hi,
>
>I want to estimate a panel model (fe) with t>n and want to control for
>contemporaneous correlation via GLS (and for serial correlation within
>groups). Moreover, I have two endogenous regressors. If I make 2sls by hand
>via xtgls with the option panel(correlated), I will get biased standard
>errors. If I make 2sls with ivreg or ivreg2 I cannot control for
>contemporaneous correlation. Does anyone know how to handle this proplem in
>Stata?
>
>Thanks!
>Aleksandra Riedl
*
* 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/
