"(When) Do Hub Airlines Internalize Their Self-Imposed Congestion Delays?"

We develop theoretical models of airport congestion with non-atomistic traffic and implement them empirically using data from twenty-seven major US airports to determine whether dominant airlines internalize or ignore self-imposed congestion. Estimates of minute-by-minute delay patterns at each airport calibrate structural models of landing and takeoff queues as dynamic functions of traffic rates and airport capacities. These functions determine the internal and external congestion that aircraft impose on one another. Specification tests largely reject the internalization model. Optimal pricing values all time using non-dominant aircraft cost coefficients and treats all delays as external—i.e., fees equal opportunity costs of allocating peak capacity to dominant airlines.