[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

re: st: Question about options in "xtreg, re"

From   Kit Baum <>
Subject   re: st: Question about options in "xtreg, re"
Date   Wed, 26 Nov 2008 15:48:32 -0500

< >	
Maarten said

> And I am running a xtreg, re. I need to suppress the intercept.
> However Stata 10 doesn't allow this option. Is there a rationale for
> this ?

I always think of random effects a way of estimating for each group a
separate intercept in terms of a deviation from the intercept reported
in the main table by -xtreg, re-. (These intercepts aren't estimated
directly, but are represented as a distribution.) Given this way of
thinking about a random effects model, I would be very reluctant to
suppress the intercept.

The REs are random variables drawn from a distribution with a mean of zero. If the intercept was suppressed, the mean would not necessarily be zero, just as in OLS without a constant term, where the constraint that the residuals sum to zero reflects the presence of the intercept in the model.

Kit Baum, Boston College Economics and DIW Berlin
An Introduction to Modern Econometrics Using Stata:

*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index