Identification and Estimation of Continuous Time Dynamic Discrete Choice Games

Jason R. Blevins.
The Ohio State University, Department of Economics
Working Paper.

Versions and Availability:

Abstract. We consider the theoretical, computational, and econometric properties of a class of continuous time dynamic discrete choice games with stochastically sequential moves, recently introduced by Arcidiacono, Bayer, Blevins, and Ellickson (2016, \emph{Review of Economic Studies}). We first establish existence of a Markov perfect equilibrium in a model with more general forms of heterogeneity across firms and states. Then we establish nonparametric identification when only discrete-time observations are available under slightly weaker conditions in the more general model with heterogeneity. Our conditions include cases where the overall decision rates may be unknown. We illustrate our results using three canonical models that provide the foundation for many applications in applied microeconomics: a single agent renewal model, a dynamic model of entry and exit, and a quality ladder model of oligopoly dynamics. Using these example models, we also examine the computational properties of the model and the statistical properties of estimators for the model through a series of small- and large-scale Monte Carlo experiments. Computing and estimating the model remains computationally tractable even in our largest experiment, which has over 58 million states, reflecting the large scale of empirical models with many heterogeneous firms.

Keywords: Continuous time, Markov decision processes, dynamic discrete choice, dynamic games, identification.

JEL Classification: C13, C35, C62, C73.

BibTeX Record:

  author       = {Jason R. Blevins},
  title        = {Identification and Estimation of Continuous Time
                  Dynamic Discrete Choice Games},
  institution  = {Ohio State University},
  year         = 2014,
  type         = {Working Paper}