Whether or not an increased police presence deters crime is largely inconclusive. Photo: Robinson Greig on Unsplash

A couple of weeks ago, Gothamist published a story that elicited some head-scratching and condemnation from various reporters, advocates, and other observers. It detailed how a man suspected of shooting a bystander during an attempted robbery in Times Square had gone into a subway station and was able to evade the cops, one of several circumstances where someone wanted by police was able to use subway stations or trains themselves to their advantage.

The piece was panned for a simple reason: crowded subway stations and the trains they hold can obviously be used for getaways, just as any other crowded area or method of transport can. If the point of the thing is to get you from A to B, then it’s natural that sometimes A can be a crime scene, yet it strains credulity to imagine a similar piece written about how criminals are taking advantage of nefarious sneakers to get away, or how crowded parks are a crime tool. Gothamist is no New York Post, and I can chalk this up to a sort of human-interest piece getting out of hand, but it does seem to reflect a general anxiety over disorder in the subways that has been bubbling ever since Covid emptied them out, hitting a fever pitch after people returned.

Which brings me to two significant and relevant recent figures. One is 3,875%; that’s the increase in the NYPD’s overtime budget between 2022 and 2023, from $4 million to $155 million. At the exact same time, assaults on subways hit a 27-year high, continuing a trend that started around 2020 and puts the figure at 52% over the 2019 number, though the increase from 2022 to 2023 was small. (You might notice both those links are from Gothamist, which I want to emphasize does great work; the subway getaway story just had an especially strange framing.)

There are two ways to read this fact: one, that there simply isn’t enough of a police presence in the subways, that things would be even worse without them, and that what’s needed is an even heavier deployment. Or, you could argue that the presence of the cops was essentially ineffective, and the spikes in assaults would have happened whether or not their presence increased, which would mean that a bigger police presence would have no effect. (I suppose you could also argue that it is not the incidence but the reporting of assaults that are going up, as more cops are around to intervene in them; this seems unlikely given that, unlike other types of crimes, subway assaults aren’t likely to go unreported.)

Let’s take a look at what evidence exists for either hypothesis. There have, unsurprisingly, been a few studies attempting to document the extent to which mere police presence deters crime, and the results have by and large been inconclusive, pointing to generally weak correlations, with some limited utility only if very targeted. Put another way, while certain analyses have found that police presence alone can deter serious crimes like homicides in certain so-called “hotspots,” having more cops around throughout an entire city (or subway system) seems to do relatively little for the incidence of broader crime. There are, of course, externalities to it anyway: higher costs, as exemplified by the overtime numbers, and the ever-present risk of racially biased policing and harassment, which at worst can result in the escalation of what would have otherwise been non-issues. As I’ve pointed out before, the NYPD transit overtime costs alone could cover the total cuts to NYC libraries four times over.

The NYPD has been cracking down on fare evasion, but the amount paid in fines is just a fraction of what is being spent on its overtime. Photo: Billy Williams on Unsplash

One problem here is that the direct cause-and-effect is hard to measure because enforcement is so circumstantial: what are police-community relations like in this place and at this time? What is the police’s specific mandate in each situation? When this cop surge on subways was announced by Adams and Governor Kathy Hochul about a year and a half ago, what they claimed to want was “omnipresence” as pure deterrence. There was some talk of combating fare evasion, going after guns, and assisting with efforts to target severely mentally ill people on the trains, but by and large the objective seemed to be to have cops simply be seen. They’ve certainly been seen, though often standing around on their phones, a phenomenon so widespread it’s become a joke acknowledged by Adams himself.

This ends up being kind of a stunted policy conversation given the all-or-nothing tendencies baked into the debate, from both perspectives but particularly the more pro-enforcement one. Asking basic questions about the effectiveness, return on investment, clearance rates (the rates of crimes actually being solved), arrest disparities, and legal immunities around police can get you easily branded as anti-police or even pro-criminal, which is counterproductive to the result we’re all supposed to be trying to get to: public safety.

I consider myself a pretty adamantly pro-housing, YIMBY type, with an emphasis on affordable housing (I wrote all of last week’s edition on how building more housing in NYC doesn’t have a silver bullet). If I’m earnest about that, it means considering the benefits and drawbacks of, for example, the 421a tax abatement. It means taking into account some evidence that rent stabilization can have a distortive impact on the market as a whole.

I don’t particularly want that to be true; it seems antithetical to my general assumption that more market regulation is a positive force to combat some of the inherently predatory aspects of residential properties as investments or assets. Yet it would be unserious and shortsighted to wave it off on purely ideological or knee-jerk grounds, especially if thoughtfully engaging with the substance could actually get me closer to the objective I claim to want.

In an ideal world, we could all do the same with public safety and policing. One could point out that a scenario that fundamentally removes police from the picture is deeply unpopular and logistically suspect, while also noting that the police as currently recruited, trained, equipped, and legally insulated are often ineffective at best and harmful at worst. If the transit surge and the robots (the patrol robot derisively known as the “crime roomba” was recently retired from service in the Times Square station without seeming to have accomplished much of anything) and everything else just isn’t really working, why would we just double down?

Even on the points that are considered successes, the results are a little dubious; the NYPD touted that the surge had targeted the problem of fare evasion, and it had massively increased fare evasion summonses to over 100,000 last year. Let’s assume, very generously,  that every single person issued a summons pays the fine at an average of $75 a pop — in reality a huge chunk don’t; I personally know someone whose ticket never appears to have even been registered by the cops — and that each summons dissuades thirty additional fare evasions.That gets us to 3,000,000 fares now paid and $7.5 million in fines collected, which sounds pretty good until you do some back of the envelop math and realize that at $2.90 a fare, that represents a bit over $16 million in a best-case scenario, or about 10% of what the NYPD spent on transit overtime, without even getting into regular duty.

So what was the point? What money was saved, what crimes prevented? I am not asking rhetorically. I want these questions to be answered seriously by Adams, by NYPD Chief of Transit Michael Kemper, by our elected officials. I think you should be asking them, too.

Felipe De La Hoz is an immigration-focused journalist who has written investigative and analytic articles, explainers, essays, and columns for the New Republic, The Washington Post, New York Mag, Slate,...

Leave a comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.