[Gretl-devel] Re: panel FE with constant dep var: what does gretl do?
Am 13.07.2019 um 23:12 schrieb Allin Cottrell:
On Sat, 13 Jul 2019, Artur Tarassow wrote:
That's now done in git. "modtest --autocorr" (or the
"Autocorrelation"
menu item under Tests in the model window) gives
the Wooldridge test, provided the estimator is fixed or random effects
and the panel has a time-series length of at least 3.
> However, in another private chat, Sven has posed the question whether
> Wooldridge's test is really only valid for RE-/FE-type of panel
> models. Once this is clarified it may be applicable for a larger class
> of panel models.
OK, I'll stay tuned on that point.
I think it's clear that FE and RE are the most important use cases,
probably covering 95-99% of all needs here. In principle, however, I
guess that at least two more variants could be easily covered: First
there's 'panel ... --pooled', because it's just a special case when
there are no individual effects.
Secondly I was also thinking of a first-differencing approach, similar
to what the test regression itself does. Currently there doesn't exist
(AFAIK) a dedicated gretl command for such a FD estimator - that's fine,
but it means that effectively a user would do something like 'ols
diff(y) ...'. So to make the test for panel autocorrelation available it
would somehow also have to be activated for the 'ols' command, in the
context of a panel dataset. Unless I'm missing something.
thanks
sven