Fabio,
What the fixed-effects regression does, is to "wipe out" the between groups
variation, leaving just the time variation. Therefore an R^2 of 0.60 tells
you that 60% of the TIME variation in your dependent variable is explained
by TIME variations in your independent variables.
Following this argument--as Clive told you--if you regress the
first-differences, i.e.
Y(t)-Y(t-1) = a + b[X(t)-X(t-1)] then you will be effectively eliminating
the between group variation (just as including a dummy variable for each
unit of analysis "i") and no fixed effect will be present. If you are not
running equations in first differences then don't worry about this.
I hope this helps,
Rafa
----- Original Message -----
From: <[email protected]>
To: <[email protected]>
Sent: Friday, April 30, 2004 11:10 AM
Subject: st: : xtreg
> Clive Nicholas wrote:
>
> You didn't say whether your dependent variable was differenced o
> > not. If
> it has been, then - if my reading of the methodological literature is
> correct - the fixed effect is removed. The FE at time _t_ minus the FE at
> time _t_+1 equals 0! Thus, there is no FE is interpret in this context.
>
> Sorry but I didn't understand your concept. Could you give me more details
> about it?
>
> My problem is that I don't know if with my xtreg,fe (within estimator)
with
> a within R^2 about 0.6 I can say that the evolution within each group is
> described from my regressors. I have just studied the temporal
correlaction
> of my regressors with the dipendent variable and all seems to me correct.
> Many many thanks.
> Your help is very useful for me.
>
>
> *
> * 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/
>
*
* 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/
