In new research, Yulia Chikish, Gregory J. Colman, Dhaval M. Dave, Brad R. Humphreys, Zachary Santamaria, and Zachary Winship find that New York City congestion pricing has reduced emergency medical services response times.
Congestion pricing is usually intended to make drivers pay for the delays they impose on others. A driver entering a crowded business district considers their own travel costs, but not the additional delay they impose on everyone else using the road. A congestion charge raises the cost of entering a crowded area, encouraging some drivers to take public transit, travel at a different time, choose a different route, or not make the trip by car. Cities including Singapore, London, Stockholm, Milan, Oslo, and Rome have adopted versions of this policy, and prior research generally finds that congestion pricing reduces traffic, improves travel speeds, reduces accidents, and improves air quality.
Yet, most discussions of congestion pricing focus on ordinary travelers. Less attention has been paid to how traffic affects emergency responders and how congestion pricing can help. When traffic falls, ambulances can move more quickly to patients and hospitals, improving the performance of a time-sensitive public service. Prior work by Elizabeth Wilde finds that an additional minute in the response time of emergency medical services increases 90-day mortality by about one percentage point, while Daniel Brent and Louis-Philippe Beland show that traffic congestion significantly slows down first responders. If congestion pricing speeds up ambulances, then standard cost-benefit analyses may miss an important source of social value.
Our paper studies this question in the context of New York City, which implemented the first comprehensive cordon-based congestion pricing program in the United States on January 5, 2025. The program charges vehicles to enter the congestion relief zone in Manhattan, defined as the area south of 60th Street. This sharp boundary creates a useful quasi-experimental research design. We compare changes in traffic and emergency medical service (EMS) outcomes just inside the priced zone to changes just outside it, before and after the policy began. In other words, we ask whether any differences in congestion and EMS response times between north and south of 60th Street changed after congestion pricing went into effect.
This approach is a difference-in-discontinuities design. The intuition is simple. Neighborhoods just north and south of 60th Street may differ in some persistent ways, including road networks, hospital locations, or baseline traffic patterns. But if those differences were stable before the policy, then a change in the boundary difference after January 5, 2025 can be attributed to congestion pricing, provided no other policy changed discontinuously at the same boundary. This makes the 60th Street cutoff especially valuable for studying localized effects of the toll.
We use two main data sources. First, we use incident dispatch data from the Fire Department of New York City (FDNY), which record the timeline of each EMS response from call creation through dispatch, arrival at the incident scene, departure from the scene, and arrival at the hospital. Because the data do not include exact health incident coordinates, we use five administrative geographies reported in the data to construct smaller “intersection polygons,” giving us a much more refined estimate of where each incident occurred. Second, we use traffic camera data processed by the C2SMART Center at New York University. These cameras take snapshots every 15 minutes, and a machine-learning algorithm classifies visible road users as passenger vehicles, trucks, pedestrians, or bicycles.
We focus on three EMS outcomes. The first is travel time to the incident, measured from when an EMS unit is assigned to when it arrives at the scene. The second is travel time from the incident to the hospital, constructed from the time the ambulance leaves the scene to the time it arrives at the receiving hospital. The third is total EMS travel time, the sum of the first two components for incidents that involve hospital transport. These outcomes capture different parts of the emergency response cycle, and they matter for different reasons. Getting to a patient quickly is crucial, but so is moving that patient to a hospital once transport is needed.
The traffic data show that congestion pricing changed street activity near the boundary in the expected direction. Passenger vehicle density fell by about 21 percent and truck density fell by about 18 percent near 60th Street and inside the priced zone. At the same time, pedestrian density rose by about 14 percent and bicycle activity rose by about 20 percent. These results are important because they provide a first-stage mechanism: the toll reduced the number of vehicles competing for road space in the priced area, while also encouraging some substitution toward non-car transportation.
Those traffic changes translated into measurable EMS improvements. In our preferred boundary-based estimates, travel time to the incident scene fell by approximately seven to nine seconds, although those estimates are less precise. The larger and more precisely estimated effects are for hospital transport. Travel time from the incident to the hospital fell by roughly 54 to 59 seconds, and total EMS travel time declined by about 63 to 70 seconds. Relative to pre-policy baselines, this is a reduction of approximately five to six percent in total EMS travel time.
These magnitudes are plausible given what congestion pricing did to traffic conditions. Ambulances already have advantages that ordinary vehicles do not, including lights, sirens, and legal priority. For that reason, a 20 percent reduction in vehicle density need not translate into a 20 percent reduction in ambulance travel time. But even small reductions in EMS travel time can matter. A minute saved on an ambulance trip may have little significance in a commuter-time calculation, but it may be meaningful when the patient is experiencing a stroke, cardiac event, severe trauma, or another time-sensitive emergency.
The improvements were not uniform across all settings. Effects were somewhat larger on weekdays than weekends for total EMS travel time, even though vehicle reductions were larger on weekends. This pattern may reflect the fact that weekday congestion is more of an issue for emergency vehicles, so reducing vehicle density during those periods yields larger operational benefits. We also find that gains emerged quickly after the policy began and then partially stabilized over time, consistent with travelers adjusting to the new pricing regime.
An important concern with congestion pricing is whether it simply moves traffic elsewhere. If drivers avoid the toll by shifting to nearby untreated streets, then the congestion relief inside the zone could come at the expense of surrounding neighborhoods. We examine this directly by studying spatial patterns around the 60th Street boundary. We find some limited evidence of small increases in vehicle traffic immediately north of the boundary, but little evidence that these changes produced meaningful worsening in EMS response times outside the congestion zone. The EMS improvements are concentrated near the boundary inside the priced area and attenuate with distance, which is what one would expect if the policy primarily reduced marginal vehicle inflows into the congestion zone rather than simply displacing congestion nearby.
Our analysis also highlights the importance of accounting for other policy changes. In March 2025, the FDNY issued a directive requiring ambulances to transport patients to the nearest hospital rather than to a preferred facility. This citywide operational change substantially affected hospital transport patterns and confounds a simple difference-in-differences comparison of EMS outcomes inside and outside the congestion zone. Our boundary-based design is less vulnerable to this problem because it focuses on localized changes at 60th Street and accounts for pre-existing discontinuities. When we account for the directive, the evidence is consistent with congestion pricing improving EMS travel times.
The policy implication is broader than New York City. Evaluations of congestion pricing usually emphasize commuter travel time, pollution, crashes, and revenue for public transit. These are important outcomes, but they do not exhaust the policy’s social benefits. Roads are also inputs into public services. When traffic slows, it affects not only private drivers but also ambulances, fire trucks, buses, and other services that depend on road capacity. Our findings suggest that emergency response performance should be treated as a distinct welfare channel in transportation policy analysis.
This does not mean congestion pricing is a substitute for investments in emergency medical services, hospital capacity, or dispatch technology. Nor do our estimates measure downstream health outcomes directly. Instead, our results show that a policy designed to reduce congestion can also improve the speed of emergency medical transport. Given the large volume of EMS incidents in dense cities, even modest improvements in travel time may have meaningful welfare consequences. Future work should quantify those health effects directly and examine whether similar patterns appear in other cities that adopt cordon pricing or related road-pricing policies.
Congestion pricing is often described as a way to make cities move faster. Our study suggests a more expansive interpretation. By reducing the number of vehicles competing for scarce road space, congestion pricing may also help emergency responders reach people faster when minutes matter most. That benefit is easy to overlook because it does not show up on a commuter’s toll statement. But it belongs in the policy ledger.
Author’s Disclosures: The authors report no conflicts of interest. You can read our disclosure policy here.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
Subscribe here for ProMarket’s weekly newsletter, Special Interest, to stay up to date on ProMarket’s coverage of the political economy and other content from the Stigler Center.
