Re: st: Multicollinearity in fixed effects regressions

[Maarten buis <maartenbuis@yahoo.co.uk>]

--- On Wed, 14/7/10, Bin Dong wrote:
> when I estiamte the FE model by including individual
> dummies. Some of independent variables in my fixed
> effects regressions are time-invariant and therefore
> theoretically have perfect multicollinearity with
> individual dummies. However, I always get significant
> coefficients of these variables in my fixed effects
> regressions with different controls. Are the results
> reliable because my sample size is large (1600 obs)
> enough to provide adequate information?

No, if you added dummies for each individual in your
data then the effects of all variables that are
constant within that individual are unidentified,
regardless of how large your dataset is. This thus 
means that there is a problem with your data: either 
the individual dummies aren't correctly created or 
the variables that are supposed to be constant within
the individual aren't constant.

Also note that if you are doing anything other than a
linear regression (-regress-) you should be worried
that your model is not a fixed effects regression, 
see for example: 
<http://www.stata.com/statalist/archive/2003-09/msg00103.html> 

If you are doing a this model in a linear regression
context, you should be using -xtreg-. This at least
rules out that something went wrong witht he creation
of the dummies (your variable identifying each 
individual may still be wrong though). 

Hope this helps,
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/