The Costs of Inefficient Robust Scheduling Practices in the U.S. Airline Industry

Airlines today use robust scheduling to plan ahead for unforeseeable disruptions. We empirically examine three common robust scheduling practices: flexibility to swap aircraft, flexibility to swap gates, and scheduled downtime. To do so, we characterize efficient operating practices in the airline industry by estimating a multiple-outcome, multiple-input production frontier. Estimates of how robust-scheduling inputs expand this frontier, and the tradeoffs faced by carriers along this frontier, allow us to infer how carriers view the costs of robust-scheduling inputs and the value they place on various operational outcomes. These estimates provide a rationale for the different robust scheduling practices used by carriers, point towards ways that carriers and regulators can target improvement, and demonstrate the severe consequences of even small inefficiencies. For example, we find that on average across its network JetBlue achieves only 89% of the efficient level of load factor for its observed levels of other outcomes and inputs, resulting in a 3% reduction in per-flight revenues.