Long-Term Consequences of Congestion Pricing: A Small Cordon in the Hand Is Worth Two in the Bush

We evaluate and compare the long-term economic effects of three cordon-based road pricing schemes applied to the Washington, DC, metropolitan area. To conduct this analysis, we employ a spatially disaggregated general equilibrium model of a regional economy that incorporates the decisions of residents, firms, and developers, integrated with a spatially disaggregated strategic transportation planning model that features mode, time period, and route choice.We find that all cordon pricing schemes increase welfare of the residents, as well as lead to GDP growth. At the optimum, the larger cordon and a double cordon lead to higher benefits than the small cordon encompassing downtown core. Nevertheless, the small cordon seems to be a safer bet because when the toll charge is set suboptimally, the net benefits from the small cordon compared to the optimum change negligibly, while the net benefits from the larger cordon decline sharply as the charge deviates from the optimal level.

In a June 22 Washington Post feature, research by RFF researcher Elena Safirova, Winston Harrington, and Sébastien Houde illustrates the economic impact and commuter time savings if traffic congestion fees were imposed on motorists in downtown Washington, DC. Similar systems have been established in London and Stockholm, and New York City is considering a per-vehicle charge to reduce congestion in busy areas.

The Post presentation outlined how such a congestion pricing scheme would work in a cordoned area around downtown Washington and estimated the travel time savings on popular commuter routes, summarizing findings from Safirova, Harrington, and Houde's discussion paper, "Long-Term Consequences of Congestion Pricing: A Small Cordon in the Hand Is Worth Two in the Bush."

In that discussion paper, the authors evaluate and compare the long-term economic effects of three cordon-based road pricing schemes - a large, double, and small cordon - applied to the Washington, DC, metropolitan area. In a cordon toll scheme, travelers who want to drive into the charged area have to pay a toll at specific points of the road network located on the cordon.

They found that all cordon pricing schemes increase welfare of the residents and lead to GDP growth. When toll charges are set at an optimal level, a large cordon and a double cordon both lead to higher benefits than a small cordon encompassing downtown core. Nevertheless, the small cordon seems to be a safer bet because when the toll charge is set suboptimally, the net benefits from the small cordon compared to the optimum change negligibly, while the net benefits from the larger cordon decline sharply as the charge deviates from the optimal level.

The work uses a model called LUSTRE, a new, integrated model developed at Resources for the Future that simulates land-use, transportation, and economic activity in the nation's capital and was the topic of a daylong workshop at RFF in March entitled "Modeling Growth for the Nation's Capital: A Work in Transit."

LUSTRE (Land Use, Strategic Transport and Regional Economy) is an integrated and spatially disaggregated land-use and transportation model calibrated for the Washington, DC, metropolitan area. It combines two smaller models: Regional Economy and Land Use (RELU), which represents economic and spatial behavior of consumers, firms, and developers in a metropolitan area and was developed by Alex Anas and Safirova; and Strategic Transport (START), which provides details on transportation choices made by economic agents such as mode, time period, and parking. START was developed by Tony May and is now maintained by MVA Consultancy of the UK. Unlike in other land-use/transportation models, the integration in LUSTRE takes place at the level of individual agents, who make trade-offs in housing, transport, and other goods based on their idiosyncratic preferences and the unique prices they face. The model also incorporates unemployment, taxes, and alternate transportation modes.