Analysis of Aggregate Passenger Routes in Air Travel: An Atlanta-Based Study

Airline hub-and-spoke networks have had numerous impacts on the domestic air transport system and its passengers. This paper studies this important phenomenon in air transportation from the perspective of passenger choice, by focusing on the relationship between aggregate travelers' route choice and the factors that influence them on observed air travel demand. A multinomial logit aggregate passenger route model is developed to examine the impacts of the following route and node attributes in the network: air fare, travel time en route, waiting time at hubs, activity level of hub airport, and number of stops en route. Route elasticity in response to air fare and travel time, as well as values of travel time and waiting time are also derived. Atlanta-based passenger data are used to calibrate the model. The results are intuitive and generally consistent over routes for different city-pair markets. While the general preference of passengers for direct routes of low fare and short distance is demonstrated by the consistent estimates of air fare, travel time and number of stops, the impacts of waiting time and hub activity level are varied.