Difference between Area-Based and Cordon-Based Congestion Pricing

Abstract

Modeling the effects of congestion pricing has become the focus of many researchers, and several have analyzed the cordon-based congestion pricing problem. However, few have attempted to investigate area-based congestion pricing because it is considered difficult to analyze precisely. Cordon-based pricing can be expressed easily in a traffic assignment procedure by the addition of a charge to the inbound links of a cordon area, but the impedance of area-based pricing cannot be expressed with a link-based formulation. The objectives of this study are to propose a sound network model for evaluating area-based pricing and to compare the effects of area-based pricing and cordon-based pricing. First, it is pointed out that an exact treatment of area-based congestion pricing requires consideration of the nonadditive trip-chain–based path cost. Such consideration appears complicated, but a simple trip-chain–based network equilibrium model with nonadditive path cost is formulated here, and a convex minimization problem that is equivalent to the model is presented. This formulation enables the evaluation of the effect of area-based congestion charging more exactly than the traditional trip-based model, even for a large network. Finally, the model is applied to a real urban area (Okinawa, Japan) and area-based pricing is compared with cordon-based pricing. In this example application, the optimal toll level for area-based pricing is found to be higher than that for cordon-based pricing.