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

[John Antonakis <john.antonakis@unil.ch>]

Hi Maarten:

The assumptions of the estimators must be met. If the assumptions of the 
random effects estimator are not met, and if the Hausman test shows that 
the estimator is not consistent then the researcher has to bite the 
bullet!  It is the same thing as estimating a regression model where you 
know that x correlates with the disturbance and yet you go ahead and 
estimate the model in any case. The coefficient of x could be higher, 
lower, or of a different sign. What is the use to society to report 
estimates that one knows to be inconsistent?

Again, as I said below, one can combine a model where fixed effects are 
partialled-out and where level 2 (e.g., firm-level) varying effects can 
be included. I cannot possibly understand why someone would knowingly 
not use this approach if they knew that their estimator is not 
consistent yet still wanted to include Level 2 effects.
Finally, in the medical science they are rightly concerned with knowing 
whether a treatment is efficacious or not, or whether it is toxic or 
not; lives are at stake here so we cannot take guesses as to what the 
coefficient may be. The gold standard is the randomized field experiment 
and rightly so. There are other methods available to recover causal 
estimates and they are used a lot in the medical sciences (e.g., 
propensity scores). Same with economics: Te Heckman two-step model or 
regression discontinuity can be good alternatives to randomized 
experiments. These alternative procedures are accepted and used by many 
researchers, so I would not characterize economists or medical 
researchers as being fixated by a specific method.

Given you are from sociology and if you have a moment, take a look at 
the Halaby paper:

Halaby, C. N. (2004). Panel models in sociological research: Theory into 
practice. Annual Review of Sociology, 30, 507-544.

It talks about the fixed-effect problem nicely as well as "tricks" to 
overcome endogeneity using mixed-estimator approaches.

Best,
J.

____________________________________________________

Prof. John Antonakis, Associate Dean 
Faculty of Business and Economics
Department of Organizational Behavior
University of Lausanne
Internef #618
CH-1015 Lausanne-Dorigny
Switzerland

Tel ++41 (0)21 692-3438
Fax ++41 (0)21 692-3305

Faculty page:
http://www.hec.unil.ch/people/jantonakis

Personal page:
http://www.hec.unil.ch/jantonakis
____________________________________________________



On 06.06.2010 14:45, Maarten buis wrote:
> --- On Sun, 6/6/10, John Antonakis wrote:
>   
> > Not including the fixed-effect would be a very fatal flaw
> > to make
> >     
>
> The assumption is no stronger than the one we make any time 
> we type -regress-. I see the strenghts of fixed effects 
> regression, but it is not the solution to all problems that
> some people believe it is. So to call not using fixed effects
> regression a "fatal flaw", is way too strong for my taste. 
> A similar fixation sometimes occurs within medical sciences, 
> who sometimes refuses to look at any estimate that isn't 
> generated with a randomized experiment. The problem is that
> in these cases the method becomes a more important part of
> the argument than the data, thus forgetting what empirical
> science is all about.
>
> -- Maarten
>
> --------------------------
> Maarten L. Buis
> Institut fuer Soziologie
> Universitaet Tuebingen
> Wilhelmstrasse 36
> 72074 Tuebingen
> Germany
>
> http://www.maartenbuis.nl
> -------------------------- 
>
>
>       
>
> *
> *   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/
>   
*
*   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/