Revenue sharing in airline alliances

Abstract

Airline alliances as means of collaboration among independent carriers are a growing trend in the industry. From a revenue management perspective, one of the most significant features of the alliances are codeshare itineraries by which independent airlines can collaboratively market and operate flights. Different from traditional, monopolistic airline revenue management, alliance members control a decentralized network of resources through independent reservation and information systems. To study the revenue management problem of such decentralized network environment, we propose a two-stage hierarchical approach. In the first stage, airlines agree on how to share the revenues generated by these interline products. In the second stage, airlines operate independent inventory control systems in order to maximize their own expected revenues. Through both analytical and numerical studies, we find that the choice of the revenue sharing rule has a great impact on the performance of the alliance. In particular, in the static setting where each airline uses partitioned booking limits, there exists a revenue sharing rule under which the decentralized system can do as well as the centralized system. We further construct an asymptotic regime in which airlines' capacities and demands grow proportionally large, and prove that under our proposed revenue sharing rule, the performance of the alliance under dynamic inventory control converges to the performance under static booking limit control. The numerical comparisons between several dynamic heuristic policies and the static booking limit control confirm the quality of the approximation. Nevertheless, given that the revenue sharing rule that is provably optimal in our model requires to disclose private demand information, we propose a simple revenue sharing rule that is based on public fares. This simple heuristic performs noticeably well in our numerical experiments, becoming an interesting candidate to be pursued in practice.