[Gretl-users] Gretl-users Digest, Vol 100, Issue 22

Sven Schreiber svetosch at gmx.net
Tue May 26 12:36:13 EDT 2015


Am 26.05.2015 um 18:15 schrieb Liliya Netskar:
> Hello Sven,
>
> thank you very much for your answer.
>
> I see this. But I still don't know - taking this example - how I can
> control for individual fixed effects for /Joe, Sue, Mark/ with Gretl?
>

If your dataset has a panel structure (and recognized as such by gretl), 
and your cross-sectional units are called Joe, Sue, Mark (plus one other 
unit that is unnamed and apparently taken to be the reference unit), 
then you already did! Namely by choosing fixed-effects model, which is 
equivalent to least-squares dummy variable (LSDV) estimation.

(If you're unaware of the equivalence between unit dummies and a 
fixed-effects estimator then you would need to revisit the classic panel 
econometric theory.)

If you really want the names "Joe" etc. to appear in the output, then 
you must not choose fixed effects, but run a (pseudo-) pooled panel 
regression with your dummies in it.

But it's one or the other, both together is impossible.

If you question is something else, then I don't understand it.

hth,
sven



More information about the Gretl-users mailing list