autocorrelation for random effects

[Nermine Medhioub]

Hello friends
I want to know how to correct for autocorrelation for the random effects model

[Carlo Lazzaro]

Nermine:
if you're using -xtreg,re- simply add -robust- or -vce(cluster clusterid)- standard errors.
That said (and as per FAQ), a tad more detailed post is no harm ! Thanks.

[Nermine Medhioub]

Thank you @Carlo Lazzaro.
To the best of my knowledge, employing xtreg, robust is solely intended for addressing heteroskedasticity. Similarly, my understanding is that using xtreg, vce(cluster clusterid) is only for handling both heteroskedasticity and autocorrelation. Could you please confirm if my understanding is accurate?

[Carlo Lazzaro]

Nermine:
not quite.
Unlike -regress- (where the -robust- option actually addresses heteroskedasticity only), under -xtreg- both options do the very same job: they take care of heteroskedasticity and/or autocorrelation.

[Nermine Medhioub]

thank you for the clarification @Carlo Lazzaro; this has been very useful!