Air Ticket Sales as Bids from Airline Alliances

Motivated by the higher price sensitivity and service homogenisation in the airline industry in recent years, we propose a new methodology to deal with transaction prices and to estimate the effect of alliances in the US domestic market. The assumption that airlines compete on price allows us to take advantage of the observational equivalence between Bertrand competition and the reverse English auction. We then apply an MLE method, developed by Paarsch (1997) for estimating auctions, to recover the distributional characteristics of air fares using a sample of airline tickets from the US domestic market. This procedure allows us to benefit from the heterogeneity of individual prices while most studies have used average prices, which would have involved a loss of information and a potential bias. We find that an alliance operating in a market is associated with prices on average 18.9 percent higher. Additionally, we find the standard deviation of ticket prices to be 4.3 percent higher, which is likely related to more effcient revenue management practice by alliance partners operating together in the same market.