The Subway’s Crime Number Went Down. The Reasons to Be Scared Went Up.
On July 2, Mayor Zohran Mamdani (D) and NYPD Commissioner Jessica Tisch stood before the midyear crime numbers and called it “the safest start to any year on record.” The subway line item let City Hall report that major transit crime is essentially flat — down about 1 percent.
The topline is true. There were 1,085 major transit crimes from January through June 2026, against 1,093 in the same stretch of 2025 — a decline of seven-tenths of a percent. But that number is carried almost entirely by one category: grand larceny, the pickpocketing and bag-grabbing that fell 7.2 percent and is running at historic lows.
Strip the property theft out, and the categories New Yorkers actually fear all rose. Subway murders doubled. Robbery climbed. Felony assault climbed. The number the mayor cites is technically accurate and substantively misleading — a composition trick in which falling theft masks rising violence, and the two roughly cancel in the headline.
- −0.7%the toplinetotal major transit crime, 1,085 vs 1,093 — the “down 1%” City Hall cites — NYPD CompStat
- +100%murdersubway murders doubled, 4 in 2026 vs 2 in 2025 — NYPD midyear data
- +9.6%robberysubway robbery rose even as citywide robbery fell 11.9% — NYPD CompStat
- +4.4%felony assaultfelony assault underground climbed year-over-year — NYPD midyear data
- −7.2%grand larcenythe one category dragging the total down — running at historic lows — NYPD CompStat
The framing was set months before the midyear release. Back in April, describing the first-quarter numbers, Commissioner Jessica Tisch told reporters transit crime “ended down, actually 1 percent.” By the July 2 briefing, the mayor had folded the subway into a much larger boast: citywide, major crime was down 5.8 percent, and the city recorded just 122 murders through midyear — the fewest in its recorded history — alongside record lows for shootings.
Those citywide numbers are real, and they are good. That is exactly why the subway line deserves its own reading rather than being swallowed by the aggregate. “Down 1 percent” is the number a rider hears and files under “fine.” It is not fine. It is the arithmetic average of a property-crime category falling fast and a set of violent categories rising — and on a train platform, those two are not interchangeable.
Total major transit crime: 1,085 (2026) vs 1,093 (2025) — down 0.7%.
Grand larceny alone fell 7.2% and is at historic lows. Remove it, and murder, robbery and felony assault all moved the other way.
The headline is the net of a good trend and a bad one canceling out. City Hall quotes the net. Riders live the parts.
Here is the same NYPD data, broken into the categories the topline blends together. Two violent categories rose; the property category fell hard enough to drag the sum back to roughly flat. This is not a mystery or a spin claim — it is the department’s own midyear breakdown, drawn on a single axis.
The percentage chart understates one category because percentages hide small base numbers. Subway murder is a count, not a rate: there were 4 killings in the transit system through midyear 2026, against 2 in the same period of 2025 — a doubling. Earlier in the year the gap was starker still. In May, with the toll at 4 against 1 at that point in 2025, the NY Post reported subway murders “up 300 percent.” Rape, the other low-count category, moved the good way: 3 reports versus 6.
Murder: 4 in 2026 vs 2 in 2025 — a 100% increase, and NYC’s first fatal subway shooting of the year came in January.
Rape: 3 in 2026 vs 6 in 2025 — down, and worth stating plainly; not every violent category rose.
The point: on a system carrying more than 1.2 billion rides so far this year, a category can double while the percentage math elsewhere keeps the total looking calm.
The single cleanest way to see the subway’s problem is to hold it against the city above it. Citywide, robbery fell 11.9 percent in the same midyear window — part of the broad, genuine decline City Hall is touting. In the subway, robbery went the other way, up 9.6 percent. Same crime. Same six months. Opposite direction.
That divergence is the argument for treating the subway as its own accountability question rather than a rounding error inside a citywide win. When a rider is more likely to be robbed underground this year even as street robbery falls across all five boroughs, the “safest start on record” framing is doing work the raw data does not support for the space where a large share of New Yorkers spend their commute.
Major transit crime so far in 2026 is at roughly the same level as at this point last year, according to NYPD data.
The flat midyear line also smooths over a genuinely bad start. Through February 8, major transit crime was up 17.1 percent year-over-year — 246 incidents against 210. Robbery over that early window was up 58 percent (60 vs 38); assault up 9 percent. The stretch included the city’s first fatal subway shooting of 2026. NYPD attributed part of the jump to an unusually cold snap that drove more people, including homeless New Yorkers, underground.
NYPD Chief of Transit Joseph Gulotta put the department’s answer on visibility: “You can see our officers on the trains inspecting trains, riding the trains. That’s the key.” The enforcement footprint is not small. Gov. Kathy Hochul (D-NY) put 750 officers into the system under her “Safer Subways” program, alongside National Guard and State Police. NYPD added 150 more officers in late February, then another 175, and moved 200-plus to specialty train patrols. By March, transit crime was falling across categories again — a real improvement off a bad January.
One governance choice sits inside this timeline. Early in the year, Mayor Mamdani (D) ordered NYPD to limit police ejections of homeless riders during weather emergencies — a humane instinct that critics tie to the same cold-weather window in which the transit spike landed. After a fatal subway shove in May, he ordered NYC Health + Hospitals to review its psychiatric-evaluation and discharge protocols. Both are defensible policies; both are also part of the record voters are entitled to weigh against the numbers.
At the July 2 briefing, Mayor Mamdani (D) framed the data in human terms: “Each of these numbers represents a life untouched by violence,” he said, describing a New Yorker “who no longer dreads riding the subway home at night.” That is a fair way to talk about the citywide murder record. Set against a subway line where killings doubled and robbery and assault both rose, it also reads as exactly the kind of averaging this page is about.
Commissioner Tisch — appointed under Mayor Adams and retained by Mamdani, and speaking here in a nonpartisan department role — credits “precision policing and the extraordinary work of the men and women of the NYPD.” The critics do not dispute the citywide gains; they dispute the subway spin layered on top of them. NY Post columnist Nicole Gelinas of the Manhattan Institute has been bluntest, noting that the system has logged some 47 murders since 2020 — roughly 8 a year, against about 2 a year from 1997 to 2019.
“This is not normal. Mamdani and Hochul, you're on notice: This is not normal.”
Nicole Gelinas · Manhattan Institute / NY Post · May 2026
The scrutiny is not only conservative commentary. At the FY2026 budget hearings, City Council Committee on Public Safety Chair Yusef Salaam (D) — himself an exoneree of the Central Park Five case — pressed Tisch and Gulotta on subway safety. The demand from every direction is the same: stop quoting the average, and account for the parts of the system where the trend is moving the wrong way.
NYC straphanger stabbed in the stomach during an attempted robbery on a moving subway.
NYPD itself concedes that “the perception of safety is just as important as the reality.” Two facts explain why the felt risk outruns the average. First, subway violence is highly concentrated: roughly half of it occurs at just 30 of the system’s 472 stations — about 6 percent of stations carrying half the violence. For a rider whose commute runs through one of those 30, the citywide average is cold comfort. Second, the incidents that go viral — a stabbing at Grand Central, a shove onto the tracks — are exactly the ones that shape whether people ride at all.
That perception is not a soft concern; it is the ballgame. The system has carried more than 1.2 billion rides so far this year, and the recovery of ridership — along with the congestion-pricing and fare revenue that ride on it — depends on whether straphangers trust the train. Charlton D’Souza of the advocacy group Passengers United, who has toured the system late at night, puts the ask simply: “The subways need constant security.” The NY Post has documented riders “canceling travel plans” after chaotic scenes underground.
Chaos in a notorious NYC subway station leaves straphangers crying in confusion — and canceling travel plans.
City Hall’s subway number — “down about 1 percent” — is arithmetically true and built to reassure. It is also the net of a falling property-crime category masking rising violence.
Underground, murders doubled (4 vs 2), robbery rose 9.6 percent while citywide robbery fell 11.9 percent, and felony assault rose 4.4 percent. Grand larceny, down 7.2 percent, did the work of holding the total flat.
The safest way to read the data is the honest one: the headline went down, and the reasons to be scared went up. Both belong in the same sentence — and only one of them made it into the press release.
