st: Predicted number of events following xtnbreg, re

[Cullen Hendrix <Cullen.Hendrix@du.edu>]

Greetings,

After using xtnbreg, I am interested in using margins to generate
predicted counts at several values for IVs of interest. When using fixed
effects, the estimation can simply be followed by:

margins,at(var1=x var2=y), predict(nu0)

which assumes the fixed effect is zero and which generates positive values
(which are consistent with prior expectations given the literature and
model specification). When doing the same following the use of random
effects (xtnbreg, re), however, I get:

Warning: prediction constant over observations.
missing predicted values encountered within the estimation sample
r(322);

I am trying to estimate predicted counts assuming the random effects are
zero. Any help would be greatly appreciated.

Best wishes,


Cullen Hendrix
Assistant Professor
Korbel School for International Studies
University of Denver
http://cshendrix.wordpress.com



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