Decision and Team Problems in Airline Reservations

Reservations systems charged with selling a fixed quantity of a highly perishable commodity through many agents present an instructive example of decision making and communication in teams. Section 2 discusses the optimal limit of sales when the probability distributions of late cancellations and no-shows and of the number of standby passengers are known. An example was calculated on the basis of Γ distributions, which had been found to fit these distributions adequately. (The statistical analysis of these empirical distributions is not included here.) Section 3 describes the existing airlines reservations systems, Section 4 discusses their effect on the sales-limit problem. Section 5 considers the rationale of these communication rules. In terms of a simplified model the conditions are studied for which a decentralized quota system is as efficient as the prevailing centralized ``sell and record'' system.