“…For common types of count DGPs, the ML estimatorθ ML behaves in a unique way, see [2,4], for example.θ ML is consistent, and √ T (θ ML − θ ) is asymptotically normally distributed according to N(0, I −1 (θ)), where 0 denotes the zero vector, and I(θ ) the expected Fisher information per observation. The mean observed Fisher information 1 T J(θ ), in turn, where J(θ ) is the Hessian of the log-likelihood function, approximates I(θ ). This asymptotic distribution can now be used to assess the variability of the parameter estimates, in analogy to [5].…”