Are Free Buses Safer? The Data Don’t Show What Mamdani Claims

Mayoral frontrunner Zohran Mamdani has promised to make New York City cheaper—freezing the rent, creating city-run grocery stores to fuel competition, and making buses free. These are ambitious programs, though many people are skeptical.

If elected mayor, Mamdani will have a chance to enact these plans and see if they work. But turning progressive ideas into functioning city policy is rarely simple. Indeed, the evidence from Mamdani’s past experimentation indicates that the results may be more mixed than he has suggested.

As a state assemblyman, Mamdani sponsored a pilot program to make buses fare-free, an initiative he often points to. Five bus lines, one in each borough, were free to board from September 2023 through September 2024.

The final report on this pilot came out in June 2025. Most of the results were unsurprising. Free fares meant that more people used the buses: nearly 50,000 more rides were recorded than the previous year. But it also meant that the buses collected less money—an estimated $16.5 million in forgone revenue.

But one result was surprising: free buses also seemed to make buses safer. Mamdani has highlighted this finding throughout the campaign, writing in The Nation that assaults on bus drivers dropped by 38.9 percent. Because drivers no longer had to challenge people attempting to board without paying, the argument goes, they could focus on driving the bus without worry.

This would be an impressive result, if it were accurate, but it’s an overestimate of the pilot program’s true effect. We can’t be confident that the pilot had any impact on driver safety.

First, the 38.9 percent figure was an early estimate of the program’s effect after the first nine months of free buses. When the report for the full year came out last June (including data from summer 2024, when teens were out of school), the reduction was smaller—31.9 percent.

Mamdani is running for mayor and so perhaps should not be expected to read updates to evaluation reports. But a second problem with the analysis is that he has focused on the reduction in assaults on the pilot routes without comparing that reduction to the change in assaults throughout the overall bus system. A reduction in assaults on the free buses can’t necessarily be attributed to the pilot program if assaults were falling on non-free buses, too.

In fact, the final report shows, assaults across the system during this period fell by 15.4 percent. In order to assess the impact of the free-bus system, we should look not at the change in assaults on the pilot routes but at the change in the pilot routes relative to the system as a whole. A simple comparison (31.9 percent minus 15.4 percent) suggests an actual impact estimate of 16.4 percent—less than half of the 38.9 percent Mamdani previously claimed.

Finally, to some extent we should not be thinking about these effects from a quantitative perspective. This was a small pilot program (despite its $16.5 million price tag), not set up to gauge rigorously the effects on relatively rare incidents like driver assaults.

The five bus lines in the pilot experienced 47 assaults in the year before the pilot and 32 assaults the year after. While stopping even one assault is laudatory, this reduction does not represent a large enough impact to be confident that it is actually caused by the fare-free policy.

In fact, one of the bus lines, the Q4, saw an 80 percent increase in assaults during the pilot period. With only five bus lines in the pilot, there is simply not enough statistical power to say whether the fare-free pilot actually had any effect. It’s quite possible that the reduction Mamdani touts was the result of chance.

Mamdani is a prohibitive favorite to win the election, meaning he will likely get the chance to implement his policy ideas. He should remember that the free-fare pilot was just a test, a chance to figure out the mechanics of a fare-free system, not whether it works. As he contemplates scaling the bus pilot (and his other campaign ideas), he needs to be attentive to what the data say—and be ready to pivot if the numbers don’t add up.

Photo by Gary Hershorn/Getty Images