st: RE: Issue with vce command

["Wooldridge, Jeffrey" <wooldri1@msu.edu>]

If you are clustering you effectively only have five cross section
observations because the clustering allows arbitrary correlation within
a country. I'm not sure you can learn much of anything with these kind
of data; you need to increase the size of the data set in at least one
dimension. Can you get more countries?

With the current data, your best hope is to include country dummies and
a couple of explanatory variables, and hope that finite-sample analysis
works. You cannot afford to cluster.

JW

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Davia Downey
Sent: Thursday, March 03, 2011 9:32 PM
To: statalist@hsphsun2.harvard.edu
Subject: st: Issue with vce command

Hello All,

I have a panel type dataset with 43 variables and 25 observations (i.e.,
5 years worth of data for 5 counties). I have tried running the panel
regression using the xtreg, re command as well as regress with the vce
(cluster county) command. The problem is that neither command is
producing standard errors (there's just a . in the standard error
column). Essentially the model is useless. For some reason when I run
either type of model (regress of xtreg), STATA assumes I only have 5
observations when in reality I have 25 (5 years for 5 counties). How can
I get STATA to recognize the panel nature of my data? I have also tried
adding year dummies to the model to no avail. Does anyone have
suggestions?

Thanks in advance.


Davia Cox Downey
Doctoral Candidate
Michigan State University
Department of Political Science and
The Global Urban Studies Program
447 Berkey Hall
East Lansing, MI 48824
coxdavia@msu.edu



*
*   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/