“Congestion Pricing on an Urban Road Network: A study using the dynamic traffic simulator METROPOLIS” by Robin LINDSEY
University of Alberta
This paper studies road congestion pricing using the dynamic network equilibrium simulator METROPOLIS which treats endogenously mode, route and departure time choices at the level of individual travelers. Simulations are conducted for a hypothetical circular city with a road network that consists of radial arterial routes and ring roads at varying distances from the city centre. Trip origins and destinations are distributed throughout the city. A range of tolling schemes are considered: single and double cordons, area charging systems, ring road tolls, destination/parking pricing, and several comprehensive schemes. Tolling schemes that encompass a large fraction of the road network are found to yield higher total benefits than do schemes with a more limited coverage, but to offer lower benefits per dollar of revenue collected. Crude schemes that impose uniform charges on trips, links, or distance traveled do not perform very well in terms of either efficiency or equity. Time-varying tolls outperform flat tolls in terms of efficiency gains, while generating smaller toll revenue transfers. Consequently, time-varying tolls have more favorable distributional impacts on travelers.