Route-level passenger variation within three multi-airport regions in the USA

Many large conurbations have more than one airport that serves the air transport needs of their region. This paper models the relative route level within three multiple-airport regions (MARs) in the United States: Boston, Washington, and San Francisco. Although passenger-level surveys provide the researcher with the most detailed information on how travelers choose between airports and airlines, as well as geographic information about both the start and end of a journey, such data are often difficult or expensive to obtain. Here, we use freely available aggregate data from the U.S. Department of Transportation and linear models to assess whether passenger levels on routes involving airports within the three cities vary in relation to known correlates of airport choice. The results show that they do, and that variables such as the “Southwest effect,” fare, and market size, among others, help explain the complexity of air-travel patterns within MARs. Importantly, the results point to the difficulty in easily classifying cities or airport pairs into simple groups. We conclude that aggregate data can be used to establish initial evidence of the strength of a functional multi-airport region and, as such, may prove useful in informing individual traveler-based research and as inputs to policy decisions.