Beta 1

Title Modeling air travel behavior
Author Warburg, Valdemar (Centre for Traffic and Transport, Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark)
Supervisor Nielsen, Otto Anker (Traffic Modelling Group, Centre for Traffic and Transport, Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark)
Bhat, Chandra (University of Texas, Austin)
Institution Technical University of Denmark, DTU, DK-2800 Kgs. Lyngby, Denmark
Thesis level Bachelor thesis
Year 2005
Abstract Modeling passengers’ flight choice behavior is valuable to understanding the increasingly competitive airline market and predicting air travel demands. This report estimates standard and mixed multinomial logit models of itinerary choice for business travel, based on a stated preference survey conducted in 2001. Previous work on air travel behavior modeling hasalmost exclusively been confined to studying either airport or airline choice. However, two recent papers have expanded the study area to cover air itinerary choice, recognizing the fact that the average air traveler focuses on all the attributes of getting from A to B. This report models itinerary choice as well as accommodating observed and unobserved variations in parameters, due to demographic and trip-related differences. This combination is new to the literature. In this report, the random parameters in the mixed logit model take a Normal distribution. Although this distribution can cause counter-intuitive signs of the coefficient, it was preferred to the lognormal distribution, because of better convergence properties, comparisons with earlier work, and disappointing results with the lognormal distribution. <p<The dataset consists of 119 business passengers and 521 non-business passengers, who each report their most recent domestic flight. A computer then generates 10 itinerary alternatives with the same origin-destination, such that the respondent has to make 10 binary choices between the actual flight and the hypothetical flights. Many demo-graphic variables are included in the dataset. The multinomial logit results show that there is significant differences the preferences of business and non-business passengers. In general, business passengers are more time sensitive, and non-business passengers more fare sensitive. In addition, the business passengers can be split up into many demographic groups, each with its own preferences. Among demographics, gender and income level have the most noticeable effects on sensitivity to service attributes in itinerary choice behavior, but frequent flyer membership, employment status, travel frequency, and group travel also emerge as important determinants. Some of these variables are also affecting choice in the non-business segment. The mixed logit models show in addition that there is significant residual heterogeneity due to unobserved factors even after accommodating sensitivity variations due to demographic and trip-related factors. Consequently, substitution rates for each service attribute show substantial variations in the willingness-to-pay among observationally identical business passengers. The results suggest that observed demographic and trip related differences get incorrectly manifested as unobserved heterogeneity in a random coefficients mixed logit model that ignores demographic and trip-related characteristics of travelers. Willingness-to-pay values are somewhat close to the values in earlier work. However, there is significant random variation around the mean values of on time per-formance, access time, and number of connections. When weighting the substitution rates by demographic shares to obtain a “typical” business traveler, the variation becomes narrower, and thus the WTP values more accurate. Overall, the results from the study should aid in targeting and adjusting service products to specific passenger groups. An example is presented on how the results can be applied in an airline’s pricing strategy. The study also facilitates a clear understanding of the air travel market, which should help air carriers design appropriate pricing schemes and better predict passenger demands when introducing new routes and services. Finally, the analysis can also assist in airport terminal planning and ground access through more accurate prediction of air travel demand.
Pages 64
Original PDF Valdemar_Warburg.pdf (0.51 MB)
Admin Creation date: 2006-06-22    Update date: 2007-03-28    Source: dtu    ID: 186147    Original MXD