Comparison of Three Empirical Models of Airport Congestion Pricing

This paper compares results obtained from three empirical models of airport congestion pricing. Morrison (1983) and Morrison and Winston (1989) implement standard peak-load pricing models using econometrically estimated demand and delay functions. Vickrey (1969) and Arnott, et al. (1991, 1993) develop bottleneck models with deterministic queues to provide structural models of intertemporal traffic adjustments in response to congestion pricing and changes in capacity. Daniel (1995) combines a bottleneck model with a time-dependent stochastic queuing model. This paper applies each of these models to Daniel's airport traffic data. While equilibria from all three models approximate the existing traffic patterns in the absence congestion fees, their different assumptions about delay functions and intertemporal shifting of demand lead to significantly different fee schedules. We conclude that modeling intertemporal substitution and stochastic traffic and delay produces more realistic responses of traffic patterns to congestion pricing and causes significant differences in the optimal fee structures