Case study — 2026 Transportation policy

Junction Thinking: what NYC's congestion pricing experiment is really teaching us

A year of data is in. The headline story is that congestion pricing worked — the more important story is where the effects went.

NYC congestion pricing zone

In January 2025, New York did something no U.S. city had done before: it started charging drivers a fee just to enter Lower and Midtown Manhattan. A year of data is now in, and the headline numbers are striking — fewer cars, faster streets, cleaner air, a flush transit budget.

But the more interesting story isn't whether congestion pricing "worked." It's where the effects went — and that's the part every planner watching this experiment should be paying attention to.

Here's what we're seeing, and what it means for how cities should think about pricing scarce urban space.


The idea in one sentence

Road space at rush hour is scarce, but until now it's been free — so people over-use it, and everyone sits in traffic. Congestion pricing puts a price tag on that scarcity, the same way we price parking spots or airline seats during peak demand. Economists call this a market-based demand management tool: instead of banning cars or building more lanes, you let price do the work of matching demand to the road's actual capacity.

The mechanics in NYC: drivers pay once per day to cross into Manhattan below 60th Street (the "Congestion Relief Zone"), with rates varying by vehicle type, time of day, and payment method. Nearly all of the revenue is legally required to fund the subway and bus system.

What actually happened

The topline results are hard to argue with:

If the only question is "did it reduce traffic and fund transit," the answer is a clear yes.

The insight that matters more: pricing doesn't delete demand, it moves it

This is the part that gets lost in the celebratory headlines. A toll on one geography doesn't make the underlying trips disappear — it reroutes them. Some drivers switch to transit. Some shift their trip timing. And some just find a way around the tolled zone entirely.

That's exactly what's showing up in the data. Traffic and truck volumes appear to be climbing on routes that bypass the toll — the Cross Bronx Expressway, corridors through the South Bronx, and the Staten Island Expressway. These are neighborhoods that were already carrying more than their share of pollution and traffic burden before the toll existed.

A zone-based price doesn't solve a regional problem. It relocates it — and if you're not watching for where it lands, you can accidentally make things worse for the people with the least power to object.

Who wins, who pays

Congestion pricing is, functionally, a transfer. It takes money mostly from car owners entering Manhattan and converts it into service improvements that mostly benefit transit riders. In a city where roughly 85% of Manhattan commuters already take transit, that transfer benefits the majority.

One recent academic estimate (Chow et al., 2025) puts a number on this: toll revenue of about $523 million against roughly $397 million in cost borne by drivers who lost time, money, or route flexibility — a net gain overall. But "net gain" is a citywide average, and averages hide distribution. The same researchers found that making sure no one ends up worse off would require real, targeted spending — extra bus service, fare discounts, subsidies aimed at people who don't have a good transit alternative — not just a general transit budget increase.

In plain terms: the people who benefit most are already transit-dependent urbanites. The people who pay most are car-dependent commuters from outer boroughs and suburbs, who often have fewer transit options to begin with. Good policy design has to actively close that gap, not assume it closes itself.

Three system-level takeaways for planners

  1. Price the road, but plan the whole network. A congestion charge is a single lever pulling on a much bigger system — curb space, delivery schedules, bus lanes, regional highways. If delivery trucks respond to the toll by shifting to evenings, curb demand shifts too. If commuters detour around the zone, arterial roads in other boroughs absorb the load. Treating the toll as an isolated policy, rather than one input into a networked system, is how you miss the second-order effects until they show up in someone's asthma rate.
  2. Spillover isn't a side effect — it's the mechanism. The traffic reduction in Manhattan and the traffic increase in the South Bronx aren't two separate stories. They're the same story. Any pricing policy that changes behavior in one place will change behavior somewhere else; the only question is whether you designed for that redistribution or got surprised by it. Regional coordination — with New Jersey transit agencies, with outer-borough highway planning — isn't a nice-to-have, it's part of the actual policy.
  3. "Net positive" is a necessary result, not a sufficient one. Aggregate welfare gains are the easy bar to clear. The harder, more important bar is whether the households already carrying the most environmental and economic burden are being protected or exposed. That requires tracking outcomes by neighborhood and income, not just citywide averages — and being willing to fund targeted mitigation even when the topline numbers already look good.

What we'd want to see next

A short checklist, from a systems-planning lens rather than a purely fiscal one:

Congestion pricing in New York is, on balance, working the way its designers hoped: fewer cars, faster streets, cleaner air, a funded transit system. But the most useful thing this experiment offers other cities isn't the proof that pricing works — it's the reminder that pricing redistributes, and that redistribution has to be planned for with the same rigor as the toll itself.

← All work Get in touch →