How Airlines Set Scheduled Block Times

Scheduled block time (SBT) setting is a crucial part in airlines’ scheduling. Interviews with the airline and relevant work in ground transportation have shown that SBT and the historical block time distribution have a close relationship. A better understanding of this relationship is a major goal for the FAA and the airlines because the benefits includes less cost, more efficient operations and better performance in the National Airspace System. This paper investigates this relationship with empirical data and multiple regression models. The historical block time information is aggregated to individual flight level to keep track of the performance of the flight over a time period. The distribution of the flight time for a flight is depicted by the difference between every 10 th percentiles. We found that departure delay plays a minor role in setting scheduled blocktime, and that SBTs have decreasing sensitivity to historical flight times towards the right tail of the distribution. Airlines tend to act “optimistically” and are willing to experience delays in trade of a shorter SBT. We also include OD pair information in the model and found longer SBT is set for larger airports, as padding for busy traffic. Taking the heterogeneity in the behavior across airlines into consideration, we further decompose the dataset to study specific airlines. The historical flight time distribution has similar effect on SBT for different airlines, and low cost carriers tend to set a shorter SBT than legacy carriers. For legacy carriers, we found that American Airlines values level of service provided the most, whereas United Airlines has a really aggressive behavior to cut SBT. Legacy carriers also set a shorter SBT for flights between their own hubs, to avoid the disruption of