I sent the query below to the list over the weekend, but did not receive any
response. I am sending it again. My apologies for the resubmission.t
I have panel count data in which two right-hand-side variables (of which
there are many) are potentially endogenous. One covariate is continuous and
roughly normally distributed while the other is a dichotomous indicator
variable. I am looking for suggestions on how to estimate such a model in
Stata. While XTIVREG(2) would be appropriate if I had a second stage
continuously-distributed DV (and of course valid first stage instruments),
my second stage DV is a count and is overdispersed. Thus, a negative
binomial specification is more appropriate for the second stage. I have yet
to find a two-stage approach to estimate this model. I would appreciate
feedback from the list on relevant literature and possible approaches to
estimate such a model in Stata. Thanks in advance.
Corey Phelps, PhD
Asst. Professor, Management & Organization
University of Washington Business School
Box 353200
Seattle, WA 98195
(206) 543-6579
*
* 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/
