تبعیض قیمت از طریق قرارداد بازپرداخت در خطوط هوایی

This paper shows how an airline monopoly uses refundable and non-refundable tickets to screen consumers who are uncertain about their travel. Our theoretical model predicts that the difference between these two fares diminishes as individual demand uncertainty is resolved. Using an original data set from U.S. airline markets, we find strong evidence supporting our model. Price discrimination opportunities through refund contracts decline as the departure date nears and individuals learn about their demand.

Consider a potential traveler who is planning to buy a plane ticket in advance. However, at this moment, he is not certain whether he will travel. Let his valuation for traveling be v. If the airline offers him a refundable ticket now, he will be willing to pay v for that ticket. But if the airline offers him a non-refundable ticket, his willingness to pay should be less than v. 1 Once he knows with certainty whether he wants to travel, there is no reason why he should be willing to pay more for a refundable ticket than for a non-refundable ticket. This paper presents a theory that explains how a monopolist can use refundable and non-refundable tickets to screen consumers and extract more surplus when consumers have to select a contract before knowing their demand with certainty.2 In the simple two-period model, the airline will offer both ticket types in advance to consumers who are uncertain about their demand and have different willingness to pay. We derive optimal refundable and non-refundable fares that depend on each consumer’s willingness to pay and the probability of travel. Consumers with high willingness to pay buy refundable tickets, and consumers with low willingness to pay buy non-refundable tickets. Furthermore, we find that the difference between these two fares consists of a quality component—the refundability value—and a price-discrimination component. A comparative-statics analysis provides an empirical implication of the model: the gap between the two fares diminishes as the date of departure approaches and consumers become more certain about their individual demand. Therefore, the airline’s ability to separate consumer types and to price discriminate vanishes. In the empirical section we test the main empirical implication of the theory. We collected from the online travel agency Expedia.com an original panel data set of refundable fares, non-refundable fares, and seat inventories across 96 U.S. domestic monopoly routes at various days prior to the departure date. The data collection focuses on posted one-way economy-class fares to control for price differentials associated with other ticket restrictions, such as, Saturday-night stayover, minimum and maximum stay, first-class travel, and connecting legs. These restrictions are commonly used as ‘fences’ to implement other forms of price discrimination and congestion pricing and to deal with aggregate demand uncertainty. The panel structure of the data controls for unobservable time-invariant carrier-, route-, and flight-specific characteristics. Moreover, the fact that both fares were collected at the same time for the same seat allows us to control for unobserved time-variant seat-specific characteristics. The estimation method, which takes into account the dynamic adjustment between the difference in fares and seat inventories, shows strong evidence that price discrimination through refund contracts vanishes as the departure date nears. In addition, a nonparametric specification indicates that most of the individual demand uncertainty, as implied by the carriers’ pricing strategy, is resolved during the last two weeks before departure.3 In the literature, [16] and [11] use advanced-purchase discounts as a means of price discrimination to improve capacity utilization in monopolistic and competitive markets respectively. In contrast, we show that an airline monopoly can use a refundability option to screen consumers and increase the airline’s expected profit. [10] suggest a theoretical model for a monopolist that price discriminates via refund contracts consisting of a price a buyer has to pay in advance and a refund the buyer can receive after he learns his valuation of the good. While Courty and Li’s purpose is to find an optimal refund contract consisting of an advance payment and a refundable amount, our goal is to find an optimal contract consisting of a refundable price and a non-refundable price for each type. In a related work, [2] present a generalization of [10] with consumers who learn their valuations gradually and with a seller that can vary the length of time during which the tickets are refundable. [5] presents a model explaining refund contracts under costly capacity and demand uncertainty. The rest of the paper is structured as follows. Section 2 presents the theoretical model, including the consumer’s problem and the airline’s problem, and describes the airline’s price menu in equilibrium. Section 3 presents the empirical analysis by first describing the data, then setting the empirical model, and closing with the results. Section 4 concludes.

This paper shows the importance to airlines of offering a menu of prices namely refundable and non-refundable fares. We show that a monopolistic airline can separate consumers who are uncertain about their demand for travel and have different willingness to pay. The fact that individual demand uncertainty is not fully resolved by the time the individual buys a ticket is used by the seller to price discriminate and extract more surplus. In our model, buyers can use refund contracts to insure against uncertainty in consumption. One implication from the theoretical model is that the gap between refundable and non-refundable fares is a function of the individuals’ travel uncertainty. If there is no uncertainty in individual demand, there is no difference in buying a refundable or a non-refundable ticket, and hence there should be no difference between these two fares. The empirical section looks at the dynamics of the gap between refundable and non-refundable prices in 96 monopoly routes and tests whether the individual demand uncertainty implied by the carrier’s pricing strategy is resolved as the departure dateapproaches. After controlling for unobserved time-invariant flight-, carrier-, and route-specific characteristics, unobserved time-variant seat-specific characteristics, and potential sources of endogeneity, the results show that the theoretical predictions are empirically supported. Second-degree price discrimination in the form of refund contracts shrinks as the departure date nears. Nonlinear parametric specifications and a nonparametric regression show that most of the individual demand uncertainty is resolved during the last two weeks, when the opportunity for price discrimination through refund contracts declines.