On the Non-Identification of Counterfactuals in Dynamic Discrete Games

This article examines the identification of counterfactual outcomes in dynamic discrete games. Building on prior results from single-agent dynamic discrete choice models, the authors argue that analogous counterfactuals in dynamic games are generally not identified because strategic interactions among agents undermine identification. Using a duopoly entry game as an example, the paper demonstrates how equilibrium multiplicity and strategic behavior complicate inference about counterfactual equilibria, highlighting important limitations for empirical industrial organization and policy analysis.