Despite opposition to tolling and fears that cordon pricing may push traffic onto neighborhood streets, New York City’s pricing program shows that charging for road space can reduce traffic congestion at scale and provide funding for public transit. While the cordon pricing program has produced some uneven effects across locations and travel modes, traffic speeds have increased within the congestion zone.
New York City’s cordon pricing program went into effect on Jan. 5, 2025, with two central goals: Reduce the number of vehicles entering Manhattan’s central business district (CBD) and generate a new, statutorily dedicated revenue stream for capital investment in the city’s transit system.
After more than a year of operation, evidence is emerging about how well the city’s cordon pricing program is meeting those goals.
Before Jan. 2025, drivers did not pay to enter Manhattan’s central business district. Today, passenger vehicles with E-ZPass pay $9 to enter the zone during peak periods, with the toll scheduled to rise to $12 in 2028 and $15 in 2031.
Drivers entering Manhattan through the Lincoln or Holland tunnels receive a crossing credit of up to $3, reducing the $9 congestion charge to $6 during peak hours. This charge is in addition to the existing Port Authority tunnel toll.
As of 2026, passenger vehicles with a New York or New Jersey E-ZPass pay $16.79 during peak hours and $14.79 during off-peak hours to enter Manhattan through the Lincoln or Holland tunnels.
Before congestion pricing, the Port Authority toll was the only charge for drivers entering Manhattan from New Jersey through these tunnels.
Between Jan. 5, 2025, and Jan. 4, 2026, roughly 27 million fewer vehicles entered the Congestion Relief Zone south of 60th Street compared with the previous year. A revised National Bureau of Economic Research (NBER) analysis found that congestion pricing increased travel speeds within the cordon by approximately 11 percent. The researchers also found speed improvements outside the cordon, particularly on roads frequently used by vehicles traveling toward Manhattan.
In 2026, traffic volumes continued to decline. In May 2026, entries into the Congestion Relief Zone and its excluded roadways were 6.5 percent lower than in May 2025, and year-to-date traffic in 2026 was down 3.5 percent.
But the Metropolitan Transportation Authority (MTA) cautioned against attributing that additional decline entirely to congestion pricing, since gasoline prices rose nearly 19 percent over the same period. Reduced congestion has also coincided with faster bus trips, while toll revenues support transit capital investment.
In April 2025, local bus routes operating within the Congestion Relief Zone were 2.2 percent faster than a year earlier, while express routes operating within the zone were 1.7 percent faster.
Effects outside the zone have varied by location. So far, monitoring has not shown a broad shift of traffic onto major roads outside the zone. New York Gov. Hochul’s office reported that in the cordon pricing program’s first year, total traffic fell 5 percent on the Brooklyn-Queens Expressway (BQE), 7.4 percent on the Cross Bronx Expressway, and 7.1 percent on the Major Deegan Expressway. Roads approaching the zone also generally moved faster, with some of the largest speed improvements on segments used by drivers entering the zone.
Cordon pricing also changed the rideshare market. Uber and Lyft passengers pay a $1.50 congestion toll for trips that begin, end, or pass through the zone, compared with 75 cents for yellow and green taxis. Crucially, that $1.50 charge applies to each passenger even when two or more riders share the same vehicle. A 2026 study estimated that Uber trips fell by up to 6 percent after the charge took effect, while Lyft trips increased 2–5 percent during a temporary rider subsidy. Overall ride-hailing volume declined by only about 0.5–1.5 percent.
The yellow-taxi market, however, has followed a different path. Yellow-taxi trips crossing the congestion zone increased 17 percent from January through September 2025 compared with the same period in 2024. The increase may partly reflect taxis’ lower $0.75 congestion fee and arrangements through which platforms such as Uber dispatch trips to yellow taxis via e-hail services including Curb and Arro. Under these partnerships, a rider can request an UberX trip through Uber and be matched with a licensed yellow taxi, while the request is sent to the taxi through its e-hail provider.
The difference between the taxi and ride-hailing charges raises a broader policy question about whether similar uses of road space should face similar prices. A taxi and an Uber or Lyft vehicle occupy approximately the same amount of scarce street capacity, yet their passengers are subject to different congestion charges. Riders have been shown to respond to these differences. Uber trips have declined since the implementation of cordon pricing, and Lyft trips have seen a similar decline after they removed their subsidy to offset the $1.50 charge.
The early results also illustrate why congestion pricing can be preferable to blunt restrictions on driving. Congestion is fundamentally a scarcity problem: during the busiest periods, demand for limited road space exceeds available capacity. Pricing lets travelers decide whether a trip is worth the added cost, rather than relying on mandates or administrative restrictions to reduce traffic. NBER has found that the relationship between traffic volume and congestion is steep enough that modest reductions in vehicle entries produced comparatively large gains in travel speeds.
Thus, New York should distinguish between using congestion pricing to manage scarce road capacity and using it primarily as a revenue-raising mechanism. Those objectives are related but not identical. If policymakers set toll rates largely based on how much revenue they want to generate for transit, the charge can become less connected to the actual congestion cost imposed by an additional vehicle.
A more market-oriented approach would increasingly link toll charges to traffic conditions, allowing prices to rise when road space is especially scarce and fall when congestion eases. If the purpose of congestion pricing is to make travelers account for the congestion costs they impose, the strongest policy would generally apply comparable charges to comparable road use rather than favoring one transportation provider over another. New York’s treatment of taxis and app-based rideshare services illustrates why that principle matters in practice.
The evidence to date suggests that congestion pricing has reduced vehicle entries and improved travel speeds in and around Manhattan’s central business district, while providing a dedicated source of transit capital funding. Those gains are meaningful, but they should be weighed against higher charges for some travelers and the uneven effects across roadways and travel modes.
New York City’s next step should be to treat congestion pricing primarily as a tool for managing scarce road capacity: toll levels should increasingly respond to actual traffic congestion conditions, similar vehicles should face similar charges, and policymakers should clearly disclose how toll revenue is spent. Those reforms would preserve congestion pricing’s strongest feature, a market-based price signal that lets travelers make their own decisions, while reducing the risk that the program mainly becomes a way to generate revenue for transit.