Re: st: how to access coefficients for each observation in panel data with fixed effects

[David Greenberg <dg4@nyu.edu>]

The assumption of the model you are estimating is that the same coefficient is applicable to all cases in the data set, so you are asking for something that you cannot have from the model you are estimating. David Greenberg, Sociology Department, New York University

----- Original Message -----
From: Amy Hieter <amy.hieter@gmail.com>
Date: Monday, April 11, 2011 2:34 pm
Subject: Re: st: how to access coefficients for each observation in panel data with fixed effects
To: statalist@hsphsun2.harvard.edu


> My apologies I was not clear.
> 
> I am trying to get the coefficient of   var1    for each observation
> in the dataset, not just for the average observation across the data
> set that I obtain from the -xtreg- output.
> 
> For example, I believe I can access the fixed effects for each county 
> by:
> 
> predict u, u
> replace u = u + _b[_cons]
> 
> I want to be able to do something similar for each variable in the model.
> 
> I took a look at he help on post estimation for -xtreg-   but I did
> not see anything that would provide these values. Though perhaps
> (likely) I am misunderstanding the help.
> 
> Hopefully this made more sense.
> 
> -amy
> 
> 
> 
> On Mon, Apr 11, 2011 at 2:06 PM, Nick Cox <njcoxstata@gmail.com> wrote:
> > Quite what you mean by the coefficients is not clear to me, but see
> > the help on post estimation for -xtreg-.
> >
> > Nick
> >
> > On Mon, Apr 11, 2011 at 6:44 PM, Amy Hieter <amy.hieter@gmail.com> wrote:
> >
> >> I am trying to get the individual coefficients for each observation 
> in
> >> my panel data set after running xtreg. There are over 3000
> >> observations (ie, counties) over 3 time periods
> >>
> >> For example, my regression is:
> >> xtreg depvap var1 var2 var3 +....+ varN, fe robust
> >>
> >> I though I might be able to get the coefficients by using:
> >> xi: regress depvar var1 var2 var3 +...+ varN, i.fips      where fips
> >> is the ID for each county.
> >>
> >> And I get:
> >> no room to add more variables because of width
> >>
> >> However, I do not think I can add more memory based on the output of
> >> "query memory"
> >>
> >> settable:                 current value      description
> >>             memory usage
> >> set memory           10M                      max. data space
> >>       10.000M
> >>
> >> Therefore,
> >> a) is there another way to get the coefficients per observation?
> >> b) am I misunderstanding the memory output and therefore can still 
> use
> >> the xi command?
> >
> > *
> > *   For searches and help try:
> > *   http://www.stata.com/help.cgi?search
> > *   http://www.stata.com/support/statalist/faq
> > *   http://www.ats.ucla.edu/stat/stata/
> >
> 
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/