Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: Controlling for multiple fixed effects


From   "Zoido, Pablo" <zoido_pablo@gsb.stanford.edu>
To   "'statalist@hsphsun2.harvard.edu.'" <statalist@hsphsun2.harvard.edu>
Subject   st: Controlling for multiple fixed effects
Date   Sat, 29 Nov 2003 21:53:33 -0800

Hi,

I have a large dataset that contains purchases of different firms from
different cities who buy from different kinds of providers at different
times.  Say I want to run a regressions on the order's prices on the order's
quantities and I want to control for differences across time, cities and
kinds of providers by including fixed effects for each of these categories.
I have a large number of fixed effect for each city and a kind of provider.


How can I run such a regression without necessarily asking stata to
calculate and include in the regression all of the dummy variables for each
category.

If I use areg, or xtreg, fe I can only capture one fixed effect and the rest
have to be included by brute force.

Is there anyway around this? Can I create a variable say fe=group(city time
provider) and then use this for my xtreg, fe i(fe)?  Why not?

Thank you in advance for your help.

Pablo

ps: I am sure that many people must have confronted this problem. I am not
sure I am making myself very clear. Basically, I have a lot of fixed effects
to include in a regression that come from different sources, some times I
even need an interaction effect.  The problem is I cannot use xi: because I
have too many categories and dummies to create.


*
*   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index