Title: Prices, Profits, and Production: Identification and Counterfactuals

Abstract: This paper studies nonparametric identification and counterfactual bounds for heterogeneous firms that can be ranked in terms of productivity. We require observation of profits or other optimizing-values such as costs or revenues, and either prices or attributes that determine prices. We extend classical duality results for price-taking firms to a setup with rich heterogeneity, and with limited variation in prices. We characterize the identified set for production sets, and provide conditions that ensure point identification. We present a general computationally-feasible framework for sharp counterfactual bounds, such as bounds on quantities at a counterfactual price. We show that existing convergence results for quantile estimators may be directly converted to convergence results for production sets, which facilitates statistical inference.