Wednesday, March 01, 2006

st: xthataylor and time-invariant regressors

Dear Statalisters,

I want to estimate the impact of an invariant variable in a panel data framework. Since the Hausman test rejected the use of the random effect model, I have to use the Hausman-Taylor estimator.

However, using the xthtaylor command that performs Hausman and Taylor estimation in stata9, stata "asks" me to specify not only time-varying but also time-invariant endogenous regressors, while I have only time-varying endogenous regressors. Including no time-invariant endogenous regressors implies an error message...including some time-invariant endogenous regressors (almost randomly) results in strange results (unsurprisingly).

Does anyone have any idea to solve this problem?

Many thanks Chris

* * For searches and help try: * * *


Links to this post:

Create a Link

<< Home

This page is powered by Blogger. Isn't yours?