Skip to content

How can New York City reduce crime on the subway?

In the New York Times today, Nicole Gelinas notes that subway crime is up recently and compares that to the big reductions of earlier years:

How did New York keep its subways safe through 2019? From 1990, the peak year for subway crime, until the late 2010s, the city’s police department, as well as prosecutors and judges used the criminal justice system to deter small crimes, such as turnstile-jumping, before they could escalate to violence.

Ha ha ha. No. Crime plunged in the subway for the same reason it plunged in the rest of the city: less lead poisoning of children in the '70s led to less crime among adults in the '90s. Turnstile jumping had nothing to do with it.

But lead is no longer a factor. So what does subway crime look like recently?

Crime jumped during the 2020 pandemic year and has been declining steadily ever since. Why? It could be because of more policing and more fare compliance and more work to get the mentally ill out of subway stations. It could be. But probably not. The most likely reason is that, for some reason, the pandemic had a big effect on crime which then waned in the four years since. No one knows why.

Needless to say, that isn't very satisfying. We all want answers. But in this case we just don't know. Crime surged everywhere during the pandemic and we don't know why. It then started to decline everywhere, and we don't know the reason for that either. But policing didn't suddenly change in 2020 nor in the two or three years after that. So it almost certainly has little to nothing to do with that. Nor did policing have much if anything to do with the 1990-2019 crime decline. Anyone who nevertheless insists that it must be the answer to the 2024 crime rate is relying on vibes, not evidence.

33 thoughts on “How can New York City reduce crime on the subway?

  1. ucgoldenbears

    You don't even mention denominator effects. Crime per million rides jumped in 2020, but the number of rides fell. Number of rides has been increasing since then.

    1. cmayo

      Yeah, it's this.

      Ridership cratered in 2020 for obvious reasons and was still only at about 67% of 2019 levels in 2023.

      That's still an increase in crime incidents (by about 20%) compared to 2019, but the spike in Kevin's chart for 2020 is barely an increase at all (about 2%).

      And when you think about it, this makes sense: those committing crimes on the subway are probably less likely to have stopped using it during the pandemic, no?

    2. Justin

      Adjust for post-Covid ridership, and violent felony numbers are worse. In the first six months of 2024, trains carried nearly 585 million passengers; that’s 69 percent of the ridership in the same period in 2019, which was nearly 844 million. Violence per million rides has risen to 0.83, lower than the peaks reached during the worst of the pandemic, but 67 percent higher than in 2019.

      1. FrankM

        Adjusting for ridership is not as straightforward as it might seem. Fewer riders mean fewer victims, but also fewer criminals. But larger crowds deter crime - most crime is in isolated areas - so fewer riders could increase the crime rate per rider.

        Don't make this more complicated than it is. As Kevin points out, crime increased everywhere during the pandemic and it's been declining everywhere since. If there's a reason to expect crime in subways should follow a different trend than everywhere else, I sure can't see it.

  2. akapneogy

    As mayor, Giuliani sought to prevent bigger crimes by discouraging broken windows and squeegee stops. So, the purported benefits of prosecuting turnstile jumping have a long pedigree. Ironically, Hizzoner has since been disbarred and has turned delinquent on his court ordered compensation for defamation.

    1. ColBatGuano

      Plus, there's no evidence that Giuliani's policies were any more effective than other cities attempts to reduce crime. They did get to harass a lot of Black folks though.

  3. Brett

    Making turnstile jumping harder through gates that can't be easily circumvented would probably help even further, along with pushing people to use monthly passes/etc instead of daily passes for tracking.

  4. Justin

    Quality of life New York City is so far gone that recovery isn’t really possible. Just read the N.Y. Post. It’s very entertaining.

    https://nypost.com/2024/08/14/us-news/11-year-old-migrant-boy-arrested-over-violent-nyc-phone-robbery-sources/

    ““At this point in time, we’re ready to call it: This is a migrant robbery pattern,” NYPD Chief of Patrol John Chell told reporters from Seventh Avenue and West 59th Street. Police have linked roughly 10 robberies that have taken place inside or along the southern end of the green space to a group of up to 12 migrant boys or young men.”

    The pic of the small boy in cuffs is priceless. Anyway… it’s not the rate that matters. People just know it’s bad because it’s so visible.

    1. Citizen99

      Ah, the New York Post!

      From Wikipedia:

      In 2017, the New York Post was reported to be the preferred newspaper of U.S. president Donald Trump,[49][50] who maintains frequent contact with its owner Murdoch.[50] The Post promoted Trump's celebrity since at least the 1980s.[51]

      1. jte21

        And the cops know they can feed the Post whatever bullshit they want and they'll print it, no questions asked. 9 times out of 10, whatever the Post prints about minorities (esp. refugees) and crime is a complete fucking lie. It's a Murdoch rag and not even worthy of being used as birdcage lining.

  5. joey5slice

    Kevin, I'm all in with you on the lead-crime connection. And the pandemic and change in the social order obviously had huge multifaceted societal impacts that we will likely never fully understand.

    But it is not true that "policing didn't suddenly change in 2020 nor in the two or three years after that."

    After George Floyd and the ensuing protests, police started "working to rule" much more. Conservatives talk about the "George Floyd" effect and how it is an indictment of protestors; I think liberals tend to push back on this, but the clear answer in my view is to acknowledge the impact and instead see it as an indictment of police departments, who allowed (and possibly encouraged) their rank and file to do a worse job because they didn't like being criticized.

    By the way, turnstile jumping may not lead to violent crime, but you know what it does lead to? More turnstile jumping. Turnstile jumping at my local station is way, way up, and it's routinely committed by people who (by appearances, anyway) are on their way to their professional office jobs in midtown. I recently saw 12 people (I counted) jump (or crawl under, or pull back and then push through) the turnstile in the space of 5 minutes, and 7 or 8 of them looked like office-working professionals. Once you realize that there are no consequences for not paying, and you see other people regularly not paying, why would you pay? I pay because it's the right thing to do, but lots of people don't seem to feel the same way.

    You know when everyone pays? When there's a cop standing by.

    1. lsanderson

      'You know when everyone pays? When there's a cop standing by.'

      And how many fares are saved compared to the money made by the cop standing there? Would a nice cardboard cop standup work half as well for a much larger return?

      On of the things Australia's big cities seems to have figured out is that if lightrail in the CBD is free, it cuts down on fare jumpers.

    2. iamr4man

      It amuses me to think that the turnstile jumpers who appear to be office workers are likely the same people who are calling for tougher penalties for shoplifting.

      1. emjayay

        I've maybe seen office worker types going in through the emergency exit gate in NYC, but never jumping a turnstile.

        Yesterday I walked by two MTA people wearing safety yellow vests that said Customer Service on them in big letters in the Coney Island subway station (a big one with a bunch of subway lines, most terminating there). They were probably some of the people hired to stop fare evasion, although they have no police power. They were standing next to each other leaning against a railing in a location where no one would walk by and ask them any questions, chatting or looking at their phones, about 30 feet from the emergency entrance gate facing 90 degrees from the gate - while a long steady stream of people were entering the subways through the held-open gate.

    1. Five Parrots in a Shoe

      If you are making the point that police are, for the most part, selfish vindictive a**holes, you are preaching to the choir here.

      But handwaving "could be" arguments are not the same as evidence.

  6. KJK

    Less people on the Subway makes it less safe to ride, and in 2019, there were almost 1.7 billion rides, which dropped by about 62% in 2020 to 639 million rides. Ridership in 2023 recovered to about 1.15 billion. In 2024 ridership is up over 14% over 2023 but still about 25% below pre pandemic levels.

    1. golack

      That does count for most of the increase/rider. And as you said, fewer riders means more opportunity (i.e. a rider by themself) for crime.

  7. emilruebe

    "Crime surged everywhere during the pandemic and we don't know why. "
    Actually it didnt surge "everywhere". In Germany crime did a huge dip during 2020 and 2021, exactly going along with lockdown times. It went up again even above normal after the COVID crisis. What I definitely know: This has nothing to do whatsoever with lead.

  8. cephalopod

    I ride public transit in a different city, where crime went up a lot due to the pandemic, and has been dropping steadily. Increased ridership certainly helped. Having lots of middle-aged office workers around is certainly a deterrent. Those are the people who call the cops at the drop of a hat.

    There was also a lot of enforcement that happened over the last year. Not a ton of transit police, but a lot of lower level enforcement people who checked fares, told people to stop smoking, and could call in the police quickly if necessary.

    The thing about that kind of enforcement is that it makes transit less appealing as a hangout location for young, bored, jobless people. With nothing to do all day there were a lot of fights, running across the tracks right in front of trains, and various petty crimes. They have largely moved elsewhere.

  9. gdanning

    A big chunk of this is almost certainly Simpson's paradox. Ridership is down the most in places/times where the crime rate per passenger was always lowest, such as commuters to Manhattan's white collar business districts. Crime per passenger is up because the average passenger is less likely to be one of those commuters. In contrast, late night ridership by people coming home drunk is probably a higher pct of the total than before.

    eg, as of 2023 "At 77%, weekend recovery is higher than the weekday recovery of 66%." https://new.mta.info/agency/new-york-city-transit/subway-bus-ridership-2023

  10. cld

    Did crime surge everywhere, or just in the US?

    Or was it mostly in areas where the fuck you population was in hysteria about wearing masks?

  11. Citizen99

    It shouldn't be a surprise that a civilization-wide disruption of the social fabric by, say, a global pandemic would lead to a rise in crime. There are probably many reasons for this, particularly if you try to imagine how the criminal mind works. Everything that happens is viewed through the lens of "How can I take advantage of this to get myself more stuff?"

    That undoubtedly happens inside the brains of all sociopathic individuals, from the lowliest shoplifter to the loftiest con man sitting in the Oval Office.

  12. lawnorder

    The covid pandemic was a world wide phenomenon. Did other countries see similar jumps in their crime rate in 2020? If not, then look for something unique to the United States as a cause, perhaps together with covid.

  13. Justin

    "Swedish gang members have been hired by Danish gangs for violent attacks on rivals and Danish Justice Minister Peter Hummelgaard said there were 25 such instances just since April. He labelled them "child soldiers" and said Sweden had a "sick" culture of violence not wanted in Denmark."

    https://www.reuters.com/world/europe/nordic-countries-join-forces-combat-spread-swedish-gang-crime-2024-08-14/

    I'm guessing that crime in NYC or just about any US city is far worse than in "sick" Sweden. And yet, see how they are appalled? Americans have lost their senses.

    1. cephalopod

      Crime in Sweden can be pretty bad. There are criminal gangs from the Baltic states that have moved in. Bombings were a preferred method of violence for a while.

      1. Justin

        I remember reading about the bombs... which is crazy. I don't know how accurate this is for today, (compared 2022 to 21?)

        With 62 people shot dead last year, up from 45 in 2021, Sweden's overall homicide rate is about one-sixth of the U.S.'s. But in a European context, it is extraordinary. Stockholm's gun-murder rate was roughly 30 times higher per capita than London's.

  14. rick_jones

    If you run a large enough number of sheep through a pass with wolves you will get a lower rate of attacks per million sheep.

  15. dydnyc

    Love this blog and I’m a very long time Kevin fan. that’s said, I doubt he reads these comments. IMHO, on crime stats he is really not helped by his mostly secluded life in Orange County. I’ve lived in NYC for twenty years, all as a regular rider of the subway. Since COVID there is essentially no enforcement of minor crimes like stall jumping, smoking on a train, setting small fires for fun, etc. And with no enforcement there is no way for these incidents to end up in crime stats. In the case of turnstile jumping there’s also no victim apart from the tax payer so it won’t show in a victim’s survey. The point is that there is a serious “vibe” shift in NYC and frankly I don’t like it. And these stats he references (plus Lead, Lead, Lead) simply don’t capture the way life has changed here. But anyone actually living here understands this. I have a doctorate in Sociology and I’m by no means dismissive of data driven arguments about crime. But a lot of this population crime data can’t be trusted when the conditions and actions it measures are processed in materially different ways by enforcement agencies pre/post COVID. Wish someone with a bigger platform would challenge Kevin on this.

  16. spatrick

    Crime surged everywhere during the pandemic and we don't know why. It then started to decline everywhere, and we don't know the reason for that either.

    And it's a good chance we never really will. There are good guesses as to why and given the amount of guns in the U.S. it's no surprise but as for definite answers, I'm afraid you'll just have to go with hypotheses.

    1). Pandemic shuts down whole of society. Cooped up population (especially certain more restrictive areas than others) with people dying around them more willing to lash out, solve problems with violence, take advance of certain situations due to boredom (school, after school, summer jobs/activities shut down for a couple of years, now just coming back online), lack of opportunities, etc. Police forces also affected by COVID with officers/jail staff getting sick, reducing numbers in the field, crippling communication and emergency planning (which is what happened in Minnesota in 2020).

    2). Police drawback, especially after what happened with George Floyd and riots that followed. Some of it was petulant "blue fluism" or featherbedding with their pensions (claiming PTSD disability) but in many cases officers (especially older ones and those nearing retirement) basically said "I'm outta' here! I'm not going to spend my time getting crap thrown at me or get sued or arrested trying to do my job!" even if they do it badly. (Actually it's probably a necessary culling of the ranks to get of bad or cops who just-don't-get-it) but it's going to take a long time to rebuild police forces and thus, crime goes up because police forces are understaffed.

    3). Here's something people are going to have to come to grips with: Violence/crime, for a lot of people, is fun! Sure you're talking about getting some instant wealth but in reality it's not that much and for some people a lot of it goes to their drug habits. Rarely does one nab a big score. For most people, it's just the thrill of doing and getting away with it or being part of a larger group action (anarchists for example). Coming out of a global pandemic with all sorts of restrictions, loss of life, shutdown businesses and economy, it was the only fun they had,

    By the way, for the fellow from NYC, all Kevin can go on is the data entered. He does not have a "vibes" tri-corder to get that more accurate assessment of the situation if that's what you're implying. My sympathy to you for living in a city where any weird incident is blown out of all proportion by the local media driving by ratings and subscriptions. However, one statistic that can be measured with certainty (unless more and more people are hiding dead bodies lately) is homicide and that number has been declining.

Comments are closed.