The appropriately named How Many Stops Act recently revealed just how many stops the NYPD conducted last year since the law went into effect: The department documented 1,185,728 investigative police encounters between July and December 2024. That’s an average of 6,310 a day.
Under How Many Stops, officers must record level 1 and 2 encounters for the first time, which provides a fuller picture for researchers like Stephen Koppel and Michael Rempel of John Jay College’s Data Collaborative for Justice. Police also have to report the stopped individual’s race, age, and gender to the best of their ability, as well as any use of force.
Level 1 stops simply request information from civilians without requiring suspicion of criminal activity in making a stop. Level 2 stops allow officers to approach individuals with accusatory questions and ask for consent to conduct a search under “founded suspicion” of criminal activity. People can walk away and ignore questioning at both levels.
Level 3 stops, better known as Terry stops or stop-and-frisks, were already documented before the legislation. They allow the NYPD to detain an individual for questioning and deploy “reasonable force” under reasonable suspicion that a crime was or will be committed.
Koppel said the researchers found that 98% of encounters were level 1 stops. Level 2 stops accounted for another 1%. In other words, the NYPD would be required to report only 1% of recorded investigative street encounters without the How Many Stops Act.
“We only had that [data] on level 3 stops, and it was totally unclear what was going on with the other types of investigative encounters,” said Koppel. “Finally, when they pull[ed] back the curtain here, we saw that there were roughly 1.2 million stops, [99%] of which we had no accounting of [and] no information about.”
Rempel, who directs the data institute, pointed to just 1% of level 1 stops leading to arrests and another 1% leading to summons (how much overlap between the two is unknown). However, they led to use of force 575 times, despite their more casual initial nature.
“The definition of level 1 is the police had some question; the basis of it was more than a hunch, but it doesn’t really mean that there’s any type of accusatory interaction,” Rempel said.
He added that the 7% arrest rate for level 2 stops was surprisingly low since those require “founded” suspicion, often for criminal possession of a weapon. Unsurprisingly, these encounters are often uncomfortable and hostile, so they require more than a hunch or a whim to justify them.
While a majority of level 1 encounters stemmed from radio dispatches in response to a victim or a witness, police officers self-initiated most level 2 encounters.
The researchers found that 39% of people stopped for level 1 encounters were Black. The rate jumped to 59% for level 2 encounters. In Manhattan, Harlem, and East Harlem precincts like the 23rd, 25th, and 28th registered the most level 2 encounters despite totaling fewer index crime complaints than precincts like Midtown North and South, as well as the Upper East Side’s 19th. Overall, East New York’s 75th tallied the most level 2 encounters — the only precinct to surpass 1,000 or more such stops.
NYCLU Assistant Policy Director Michael Sisitzky, a spokesperson for Communities for Police Reform, said the new data unsurprisingly confirms racial disparities in investigative encounters among all three stop levels.
“The majority of people who are stopped by NYPD, whether it’s a level 3 reasonable suspicion stop or all the way down to level 1 stops, [are] people of color,” said Sisitzky. “That is something that’s not surprising, just based on what we’ve seen from NYPD enforcement and deployment trends, but it’s good to have the data there to show what that impact is and how those interactions look differently, depending on the communities that New Yorkers live in and the demographic profile of New Yorkers as they’re experiencing NYPD activity.”
While not new or unique data, reviewing How Many Stops numbers also confirmed increased stop-and-frisks under the current administration. Level 3 stops rose by 184%, from 8,947 to 25,386, since 2021 when Mayor Eric Adams took office. However, they pale in comparison to 2011, when the city recorded 685,724 level 3 stops, largely of Black and Brown New Yorkers. As a result, the city faced racial profiling allegations over such practices in Floyd v City of New York, a class action lawsuit that ultimately succeeded.
“We were able to win policy and legal victories in 2013 to lead to stop-and-frisk declining because [of] having that data on level 3 stops,” said Samy Feliz, a member of the Justice Committee. “Now that we have more of that data, we can see what the NYPD is doing in our city and we can use that data to continue to build campaigns and do more to protect our communities from police violence.”
Feliz came across the How Many Stops Act through the Justice Committee, which championed passage of the legislation and works with families of those killed by police. A then-sergeant fatally shot his brother during a 2019 traffic stop and NYPD commissioner Jessica Tisch is currently deciding whether to fire the alleged killer cop after a department judge found him guilty last month in a Civilian Complaint Review Board prosecution. The officer hailed from the Bronx’s 52nd Precinct, a command top five in both level 1 and level 2 stops based on the data.
The mayor famously vetoed the How Many Stops Act last year, but could not prevent the bill’s passage due to supermajority support from the City Council.
The law’s proponents, such as lead sponsor Public Advocate Jumaane Williams, maintain the recording process takes officers half a minute or less through an easy-to-access smartphone drop-down menu. They also say the NYPD Patrol Guide already mandates accounting for investigative encounters when tagging body-worn camera footage.
However, the current data remains crude for the researchers, who cannot account for overlap between categories due to the aggregate numbers.
“We don’t have individual rows of information, so in a sense, it limits our ability to break down the data,” said Koppel. “Ideally, we’d do more granular analysis, where, in a particular precinct, we’d be able to break down the race [and] ethnicity, or we break down the types of stops that occurred.”