Optimal Dynamic Nonlinear Pricing for Airline Networks

Abstract

Airfares vary across the booking horizon according to intertemporal price discrimination and adjustment of fares in response to stochastic demand shocks. Previous works studying these economic forces abstract away from the ramifications of pricing an itinerary has on revenue and consumer welfare in a larger network containing itineraries subjected to common interconnected capacity constraints. I estimate a dynamic structural model of airline network pricing with a novel high frequency data set on flight prices and seat availability while relaxing the assumption of optimality. By a method of simulated moments, I resolve the classic econometric issues of endogeneity, censoring, and truncation in order to recover the airline’s beliefs on its stochastic demand process. I perform several counterfactual experiments to compare revenue and consumer welfare under several different pricing strategies and network configurations to elucidate the channels through which pricing externalities are transmitted in the network. I show that when the airline strategically prices itineraries jointly in their network according to a network-perfect pricing policy, it can increase revenue by a lower bound of 2.3% relative to the existing industry standard of network-oblivious pricing policies that independently set seemingly optimal prices for individual itineraries, a significant gain given the industry.