st: RE: Random effects and pooled models

["Schaffer, Mark E" <M.E.Schaffer@hw.ac.uk>]

Elena,

> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu 
> [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of 
> Elena Giarda
> Sent: 27 July 2006 16:47
> To: statalist@hsphsun2.harvard.edu
> Subject: st: Random effects and pooled models
> 
> Hi,
> 
> I have panel data on individual wages and I'm trying to 
> estimate age profiles taking into account birth cohorts (9 of them).
> 
> I ran the following regression and obtained significant 
> coefficients on all
> variables:
> 
> reg ln_wage age age_squared  control_variables  cohort_dummies, robust
> 
> Then I wanted to check whether a random effects model was 
> appropriate and therefore run:
> 
> iis cohort
> xtreg ln_wage age age_squared  control_variables, re
> 
> but in the output I obtained:
> 
> Random-effects GLS regression                   Number of obs      = 
> 411913
> Group variable (i): cohort                      Number of groups   = 
> 9
> 
> R-sq:  within  = 0.4447                         Obs per group: min = 
> 17118
>        between = 0.9953                                        avg = 
> 45768.1
>        overall = 0.5088                                        max = 
> 66702
> 
> Random effects u_i ~ Gaussian                   Wald chi2(14)      = 
> 426627.54
> corr(u_i, X)       = 0 (assumed)                Prob > chi2        = 
> 0.0000
> 
> sigma_u  =  0
> sigma_e =  0.30766012
> rho = 0   (fraction of variance due to u_i)
> 
> What does rho=0 mean exactly?

This comes up on the list from time to time, e.g.,

http://www.stata.com/statalist/archive/2004-02/msg00644.html

The intuition isn't hard: sigma_u=0 means no variance in the cohort
effect, which means they all have the same intercept, which means ...
OLS.

Different random-effects estimators will give you different estimates of
sigma_u.  Try the others (sa and ml options) and see what you get.

Cheers,
Mark

Prof. Mark Schaffer
Director, CERT
Department of Economics
School of Management & Languages
Heriot-Watt University, Edinburgh EH14 4AS
tel +44-131-451-3494 / fax +44-131-451-3296
email: m.e.schaffer@hw.ac.uk
web: http://www.sml.hw.ac.uk/ecomes
 


> That there are no cohort 
> effects? I made a check and found out that I got the same 
> results of running a pooled
> regression:
> reg ln_wage age age_squared  control_variables, robust
> 
> Can anybody give me hand?
> 
> Thanks a lot!
> Elena
> 
> *
> *   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/