Robust Econometrics of Games in IO

forthcoming

REGIO develops advanced econometric methods to identify and estimate economic fundamentals in games played by firms, addressing a broad spectrum of real-world industrial organization scenarios. These methods bridge critical gaps where existing techniques fall short and provide a disciplined framework to relax less credible assumptions about firm behavior and the game environment, enabling researchers to draw robust empirical conclusions. Their effectiveness is demonstrated through applications, supporting strategies to ensure fair competition, consumer protection, and a clearer understanding of distributional impacts and inequality. To achieve these goals, REGIO is structured around four pillars. Pillars 1 through 3 focus on designing identification and inference methods that remain robust to misspecification by imposing minimal restrictions on firm behavior and the game environment. By crafting new tools based on partial identification, these pillars tackle theoretical challenges, such as addressing multiple equilibria and firms’ potentially limited awareness of their competitive landscapes, as well as statistical challenges. Pillar 4 pioneers the use of the novel methods to real-world scenarios, demonstrating their practical applicability and bolstering informed market interventions. Through these pillars, REGIO not only advances econometric research but also significantly enhances the interface between economics and policymaking, improving our understanding of dynamics in modern markets.