Re: st: Estimating firm level data on regional level data using a within estimator.

[natasha agarwal <agarwana2@googlemail.com>]

Hello Everyone,

Many thanks for the reply.

It really has widened the horizon of reading and provided a better
understanding. Although I have a few questions.

Q1. If I have not understood it wrong, Random Coefficient Models are
richer models of random effect models. So if the hausman test rejects
the use of the random effect model in favour of fixed effects models,
to what extent would it still be valid to centure into Random
Coefficient Models when the use of Random Effects model itself has
been rejected in the first place.

Q2. In the paper "Halaby, C. N. (2004). Panel models in sociological
research: Theory into practice. Annual Review of Sociology, 30,
507-544.", it is shown that even when the Hausman Test rejects the use
of Random Effect models, researchers still go ahead with the use of
the same because the correlation of the unobserved unit-specific time
constant heterogeneity with the explanatory variables is arbitrary. To
what extent in this case can one rely on the powers of the Hausman
Test?

Looking forward to optimistic replies.

Thanks
Natasha
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/