Integrating airline operational practices into passenger airline hub definition

Airlines operate over hub and spoke networks to take advantage of economies of scope, scale, and density as well as higher hub fares. The system of airline hubs, namely hub classification and spatial allocation, has been necessarily affected by the era of airline mergers in the first 13years of the 2000s. However, existing two-to-three-tier hub classifications and the set of metrics that define these tiers may not be sufficient to capture the changes in airline hub practices due to mergers. In this study, we seek to broaden the existing method of establishing a hub hierarchy by including operational metrics along with the classic accessibility and passenger metrics, and to utilize this re-defined hierarchy to examine airline hubbing practices before and after an era of mergers. Using k-means clustering on data from 2004 (pre-merger) and 2012 (post-merger) comprising the entire U.S. airline network, we find that hubs fall into more than three tiers and that flight frequency at a hub is a key variable in defining hub hierarchy. We find that the exclusion of frequency from hub hierarchy definition has geographic consequences, as we overstate the importance of small coastal hubs when frequency is not included as a metric. In addition, we find that hub hierarchies are well-defined post-merger, in contrast with pre-merger, such that the tier level implies a consistent, commensurate level of all metrics. This finding has implications for hub definitions across research and practice, including broadening the federal hub definition and including tiered hierarchies in the field of hub location modeling.