Bridging Quasi-Experimental and Structural Approaches for Robust Evaluation of US Airline Mergers

We bridge quasi-experimental and structural approaches for robust merger evaluation. First, we show that the difference-in-differences (DiD) equation is the "reduced form" of a structural model, where demand and cost parameters identify price effects of mergers even when the DiD approach faces identification challenges. Second, we propose a $\textit{synthetic GMM}$ approach by applying synthetic DiD weights to structural moment conditions to improve estimates when only a few treated markets are available. Applying this methodology to three airline mergers, we find modest efficiency gains entirely offset by increased coordination. The synthetic GMM refinement sharpens findings, uncovering anti-competitive effects standard approaches miss.