In 2021, New York City plans to join Singapore, London, Stockholm, Milan, and Gothenburg in tolling cars entering its central business district. While the purpose of these systems is to limit traffic, their deployment has also moved authorities to collect immense amounts of traffic data. Some of these data are gleaned from the toll-collection system, but cities typically go further by conducting studies and installing additional traffic sensors, with sometimes startling results. The London Congestion Charging Impacts Monitoring program published annual reports on traffic patterns with a depth of analysis never available before.
New York City should use congestion pricing to deepen understanding of its own traffic—first, to determine how well the system is working. Should trucks and buses pay more? Are exempted trips creating congestion? The data collected for congestion pricing have often been used to fine-tune its design. In Milan, for example, traffic and pollution monitoring showed that the initial version of the system was not working as planned and led voters to demand an overhaul.
The data collected alongside congestion pricing also have more general uses. Manhattan below 60th Street is so congested that the city should institute policies beyond tolling, such as dynamic-traffic signal timing, which lets engineers and algorithms change traffic lights’ stop-and-go phases to reduce delays and prevent gridlock. Part of the tolled zone is already being used for an experiment in signal-timing: the Department of Transportation’s (DoT) Midtown in Motion, which works to smooth the flow of traffic into midtown and to keep lines of cars waiting at lights from spilling out into intersections.
For better or for worse, FDR Drive and the West Side Highway will be exempt from tolling. The tolled zone, then, will be limited to city streets and will require substantial new infrastructure. It’s likely that the tolled zone will be ringed by about four-dozen sites, where E-ZPass readers will tag cars upon entry (and perhaps exit). If the ultimate design includes tolling cars that travel entirely within the zone, tolling equipment will be needed inside the zone, too.
To make the most of congestion pricing, however, requires clear thinking about what data to collect. Transportation agencies often trumpet the success of downtown congestion-pricing systems by pointing to data on inflows: the number of vehicles—sometimes broken out by type—coming into the tolled zone per day or during rush hour. Inflow is a useful statistic and can reveal, for instance, whether the toll is discouraging delivery trips, but it is overemphasized as a metric for gauging the relative success of congestion-abatement plans. Inflow fails to capture how far and how long vehicles circulate after they enter the tolled zone—and recent trends in urban travel have worsened this kind of congestion. Critics have blamed the sharp increase in the number of vehicles engaged in for-hire service or delivery for slowing traffic in London in recent years, for instance, even as the number of vehicles entering the Charging Zone has remained about the same.
For a superior measure of how congestion pricing can improve the experiences of road users, the city should install microwave sensors throughout the tolled zone to count the rate at which individual vehicles pass dispersed points, and then average these readings for the zone as a whole, as well as for different neighborhoods. Travel-time metrics—such as how long it takes, at different times under different conditions, to drive from the Empire State Building to Rockefeller Center—can better describe what drivers actually experience than just inflow data. Currently, New York City gathers information on travel time using data from yellow cabs, which record when and where they pick up and drop off passengers. For more reliable and immediate data, Midtown in Motion uses E-ZPass readers to track the passage of vehicles between given points. The toll-collection system, effectively deployed, should be able to infer travel times in more places. The same data can also be used to build origin/destination matrices (tables listing trips from one area to another), which can improve high-level transportation-planning decisions.
The DoT should also collect statistics about factors including the number of cars on the road at once, or “accumulation,” which reflects what one might notice about traffic from a plane; area “density,” or accumulation divided by its lane mileage; and “queue lengths,” which measure the number of vehicles waiting at intersections. These indicators are essential for dynamic signal-timing, which works largely by ensuring that blocks don’t “fill up”—that is, crowd so many vehicles on a block that vehicles can’t cross intersections during one green-light cycle. More broadly, these measures describe how crowded a road is, and thus how stressful it is for drivers, cyclists, and pedestrians to navigate.
As the launch of New York City’s congestion-pricing system approaches, much attention will focus on the size of the toll and who gets an exemption from paying it. But political leaders should start looking at congestion pricing as an opportunity to build a world-class data-collection and traffic-management system. The resulting troves of information can help make the streets safer and getting around more convenient.