Simulated Maximum Likelihood Estimation of Dynamic Discrete Choice Statistical Models--Some Monte Carlo Results

Abstract

This article reports Monte Carlo results on the simulated maximum likelihood estimation of discrete panel statistical models. Among them are Markov, Generalized Poly, Renewal, and Havit Persistence Models with or without unobserved heterogeneity and serially correlated disturbances. We investigate statistical properties and computational performance of simulated maximum likelihood methods and a bias-correction procedure. With a moderate number of simulation draws for the construction of simulator and the bias adjusted procedure, most of these complex dynamic models can be adequately estimated for panels with length up to 30. The Polya model and the Renewal model can be accurately estimated for panels with 50 periods.