Network Competition in the Airline Industry : A Framework for Empirical Policy Analysis

This paper proposes a three-stage model of airline network competition, in which an airline's strategic decisions include its network structure (first stage), flight frequency (second stage), and pricing of all nonstop and one-stop services (third stage). The model links direct and indirect flight frequencies to capture network externality across market and is estimated without computing a network equilibrium. The empirical finding shows that ignoring airline network externality leads to an underestimation of the benefits of operating an additional flight by 12.6%. I propose a novel method to evaluate the consequences of a hypothetical merger between two airline networks (Alaska Air Group and Virgin America). The counterfactual results show that the post-merger airline would re-optimize its network structure, entering 8 markets and exiting 3 markets, resulting in a greater consumer surplus in large markets. Furthermore, the paper illustrates how network externalities shape airline network structure. Airlines will schedule 5% less of their flight frequency if they make entry/exit decisions for each city-pair independently without considering its externalities to other markets