Count panel models for endogenous regressors

Question

I have a count variable as the dependent variable and several continuous regressors on the RHS. The dependent variable and some regressors are supposed to be endogenous to unobserved factors. I have a set of IVs for the endogenous regressors.

I have two questions:

1. Is there any model for count panel data (especially for fixed effect) that addresses endogenous regressors?

I have been looking into the literature and have not found anything. It is either count models for panel data, or count models for endogenous regressors, not both.

2. Some books mention that it is difficult to address endogeneity in count panel data, but they do not really explain why. Anyone knows why?

Thanks!

Answer 1

Windmeijer (2000, Economics Letters) presents a treatment of the estimation of count-data models with fixed effects and endogeneity.

http://www.sciencedirect.com/science/article/pii/S0165176500002287

See this slideshow by Wooldridge for a pedagogic and progressive presentation of panel data models with endogeneity. Count-data models are introduced slide 54. He clearly explains the possible solutions and relate the count-data case to other nonlinear models.

http://www.ifs.org.uk/docs/wooldridge%20session%204.pdf