Dear Julia,
HT is based in three-steps: (1) run fixed-effects, get the time variant
coefficients and estimator for error variance, (2) run the "between"-effects
with instruments getting the time invariant coefficients and estimator for
the "between" error. With these generate the GLS transformation and (3) Run
a instrumental regression with the transformed (using the GLS factors)
variables. Looking the formulas of paper it seems to me that you have to
change the procedure to obtain the GLS transformation, such that it includes
a heteroskedasticity and autocorrelation correction (HAC). I think that this
way is unfeasible. An alternative is stop the procedure in (2) and perform a
HAC manually following what -newey- does. Check [R] -robust- and -_robust-
to start your research and then move into -mat accum- and -mat glsaccum-. We
know that the estimator in (2) are consistent (fixed-effects are because
they are LS and "between"-effects are because we used IV) but not efficient
(this is why HT add the last step), then your HAC correction takes care of
the last part.
Rodrigo.
----- Original Message -----
From: "Julia Spies" <[email protected]>
To: <[email protected]>
Sent: Friday, April 28, 2006 7:47 AM
Subject: st: Hausman taylor
Dear all,
I'm quite a beginner with Stata and i'm trying to run a Hausman taylor
regression. However, taking some (plausible) time-invariant variables as
endogeneous results in outrageous parameter estimates for these variables.
Nevertheless, the over-identification test suggests that instrumenting these
variables has improved the model. Does anyone have an idea what the problem
could be? I understand there is no option to correct for heteroskedasticity
and autocorrelation. Does anyone know how to do it manually?
Cheers,
Julia
*
* 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/
