| Summary: | Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Civil and Environmental Engineering, 2008. === Includes bibliographical references (p. 191-196). === Over the last ten years, the rapid growth of low-cost airlines and the development of web-based distribution of airline tickets have transformed the competitive environment in the airline industry worldwide. The relaxation of fares rules by low-cost airlines has disrupted the pricing and revenue management models of large network airlines. A better understanding of passenger choice behavior is now required to support the development of new strategies to compete more effectively in the current marketplace. In order to avoid the risk of bias associated with stated preference data, we focus in this research on how to develop a model of airline passenger choice based on booking data. Previous studies based on booking data have been limited to the sole choice of an airline itinerary and did not account for heterogeneity of behavior, a major characteristic of airline markets. This is due to the properties of booking data. For instance, only the chosen alternative is recorded in airline bookings and no information is available on other travel alternatives available at the time of the booking. Similarly, booking records contain no information on trip purpose that is traditionally used to segment airline markets. In this dissertation, we develop a modeling framework to overcome these limitations and extend booking-based passenger choice models to the joint choice of an airline itinerary and fare product. Booking data was combined with fare rules and seat availability data to incorporate the impact of pricing and revenue management and reconstruct the choice set of each booking. Characteristics of the traveler and the trip were retrieved from the booking records and used to replace trip purpose. === (cont.) They were included as explanatory variables of a latent class choice model in which several factors can be used simultaneously to segment the demand without necessarily dividing the bookings into many small sub-segments. In addition, a new formulation of a continuous function of time was proposed to model the time-of day preferences of airline travelers in short-haul markets. Instead of being set to a full 24 hours, the duration of the daily cycle was estimated to account for the low attractiveness of some periods of the day such as nighttime. Estimation results over a sample of 2000 bookings from three European short haul markets show that the latent class structure of the model and a continuous function of time led to a significant improvement in the fit of the model compared to previous specifications based on a deterministic segmentation of the demand or time-period dummies. In addition, the latent class model provides a more intuitive segmentation of the market between a core of time-sensitive business travelers and a mixed class of price-conscious business and leisure travelers. This research extends the scope of potential applications of passenger choice models to additional airline planning decisions such as pricing and revenue management. In particular, parameter estimates of the model were applied to forecast the sell-up behavior of airline passengers, a major input required by the newly proposed revenue management models designed to maximize revenues under less restricted fare structures. === by Emmanuel Carrier. === Ph.D.
|
|---|
