Testing significance of Swamy–Arora estimator of variance component?

[Peter Schilling]

Dear Statalisters,

if I run a regression of the following form:

xtreg y x, re sa

Is there any way to to test the significance of sigma_u, the point estimate for the variance of the random effects?

Thank you very much for your help and kind regards
Peter

[Fei Wang]

Peter, as below:

xtreg y x, re sa
xttest0

[Peter Schilling]

So, a typical Breusch Pagan Test - that's what I was thinking of as well.
Thanks for your reply.
Kind regards
Pierre

[Peter Schilling]

Maybe a quick follow-up question: Is it possible to get a standard error for the point estimate of the variance component (sigma_u)?
Thanks,
Peter

[Fei Wang]

Peter, I don't think Stata stores the SE of the variance estimator as it's useless for constructing common test statistics for variances. You may bootstrap the whole process to get an empirical SE if necessary.

[Peter Schilling]

That's a good idea. Thanks!