st: Re: Random/fixed effects

["R.E. De Hoyos" <redeho@hotmail.com>]

Peter,

The Random and Fixed effects are panel data models. Basically, in these 
models, you are assuming that the residual is composed by two different 
parts; one of them is completely random and the other is a component that is 
inherent to the cross-sectional variable (either fixed or random).

Total variance in the panel data models can be divided into the variation 
due to differences between cross-sectional units and differences within 
cross-sectional units (i.e. time). Without this structure you cannot 
estimate the models you mentioned. For instance, the fixed effects (FE) 
model will capture ALL the variation between cross-sectional units 
exploiting the time variation. With a cross-section, including a dummy for 
each of your IDs (a way of estimating the FE model) will leave you with no 
degrees of freedom to estimate your variables of interest.

Rafa
PS. You might want to read Chapter 13 in Wooldridge "Introductory 
Econometrics"

----- Original Message ----- 
From: "peter harper" <pharper11@yahoo.co.uk>
To: <statalist@hsphsun2.harvard.edu>
Sent: Sunday, April 23, 2006 6:24 PM
Subject: st: Random/fixed effects



> Dear Statalist
>
> I have a cross-section data. I want to find out the
> spillover effects of x1, x2 and the impact of other
> variables on y.
> 1) Is firm random effects the most appropriate
> estimation method.
> 2) Is random effects the most appropriate method.
> 3) Is fixed effects the most appropriate method
> If 1, 2 or 3 applies, where would I find the appropriate command
> please in stata.
>
> Thanks in advance
> Peter
>
> *
> *   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/