Re: st: Zero inflated count data estimation in panel

[David Jacobs <jacobs.184@sociology.osu.edu>]

Tim and Julius

There are no panel zero-inflated count models in 
Stata.  It would be nice if Stata gave us one in 
version 12 but failing that, one work around is 
to cluster correct using the case identification 
variable.  This would give you a the next best 
thing to a random effects model.

The (difficult to work with) econometrics package 
named LIMDEP offers dedicated panel versions of 
zero inflated models, but I'm uncertain whether 
this program offers fixed-effects for such models.

David Jacobs

At 11:49 AM 8/12/2010, you wrote:
> Dear All,
>
> We currently face the following problem:
> We have panel data with the dimensions 350x37. 
> The dependent variable is a patent count, and 
> many (70%) of the observations are zero, the 
> remaining 30 % are highly skewed from 1 to 140. 
> We thought about using zero-inflated negative 
> binomial or zero inflated poisson estimators, 
> but we did not find any stata command for these 
> estimators with fixed effects. Is there any way 
> to run zero inflated count data regressions with 
> fixed effects in Stata? If not, is there any 
> suitable way of dealing with the problem of many 
> 0-values in panel data using stata?
>
>
> Thank you in advance
>
> Tim and Justus
> --
> GMX DSL SOMMER-SPECIAL: Surf & Phone Flat 16.000 für nur 19,99 ¿/mtl.!*
> http://portal.gmx.net/de/go/dsl
> *
> *   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/