Counterfactual analysis under partial identification using locally robust refinement

Structural models that admit multiple reduced forms, such as game-theoretic with equilibria, pose challenges in practice, especially when parameters are set identified and the is large. In cases, researchers often choose to focus on a particular subset of equilibria for counterfactual analysis, but this choice can be hard justify. This paper shows some parameter values more “desirable” than others even if they empirically equivalent given data. particular, within set, predictions exhibit robustness against local perturbations forms (e.g., equilibrium selection rule). We provide representation subset, which used simplify implementation. illustrate our message using moment inequality an empirical application based model top coded