New York City mayor-elect Eric Adams recently wrote an op-ed flirting with the idea of bringing back the “stop and frisk” preventative policing policy.

This is a sign of how far the politics of crime have shifted under the pressure of rising shootings, and in general, I think Democrats could learn a lot from Adams’ success. But this particular gambit is a bridge too far — anti-police activists say plenty that’s wrong, but they are fundamentally correct about stop and frisk.

Many of their talking points underrate potentially promising crime-control interventions by invoking a fundamentally false tradeoff between promoting racial justice and effectively catching people who break the law. But “stop and frisk” made the opposite error, inflicting clear racial justice harm for dubious crime control benefits.

The practice is fundamentally a form of collective punishment exacted on young male residents of low-income neighbors plus some non-residents who happen to look like the residents. And while the practice of surging cops into a high-crime area to frisk a bunch of people does seem to reduce crime, there is basically no evidence that the actual frisking plays a useful role. All the causal work in crime reduction is likely achieved simply through the officers’ presence.

Stop and frisk is unfair, ineffective, and unleashes a toxic political dynamic in which the police delegitimize themselves in the eyes of many citizens by harassing innocent people. By intimately linking “there are a lot of cops around” with “a lot of innocent Black and Hispanic men are getting hassled,” it makes people think that police presence per se is bad for racial justice. It shouldn’t be! Policing is a critical public service and vulnerable communities suffer when it’s under-provided, just as they do when under-provided with parks or schools.

But they also suffer when it’s provided in inappropriate ways, and aggressive stops are very inappropriate.

Innocent people don’t like to be frisked

I got stopped-and-frisked once back when I was a young male person walking around New York City; it’s not pleasant. This is different from traffic cameras or other forms of passive surveillance where nothing bad happens to you if you’re given the all-clear. Like everyone else, I am annoyed when a speed camera catches me speeding. But driving past a camera that says you’re fine is fine. Having your day interrupted so cops can yell at you and feel you up is not fine, and it’s not like the cops are apologetic when you turn out not to have a gun on you.

In my personal circumstance, getting frisked was a hassle but a little bit of a funny story. For young Black and Latino men — especially ones living in low-income neighborhoods where they’d be stopped a lot — it was humiliation on top of an inconvenience, one that would be visited on them and their friends with some frequency.

This is a real cost, and importantly, it’s not just a statistical disparity — it’s a genuine form of statistical discrimination.

Racial profiling, not just disproportionality

What’s the difference? If you scanned the entire city with a Magic Gun-Detector Ray, the odds are you’d find more men than women, more young people than old people, and more Black people than white people. So in a sense, the burden of this would fall disproportionately on young Black men. But critically, it would fall specifically on people carrying illegal guns, leaving those without illegal guns unharmed. But the NYPD does not have a Magic Gun-Detector Ray; they have the stressful and violating stop and frisk procedure.

You could just stop people totally at random and frisk them, sort of like how the TSA makes everyone go through the metal detector.

But precisely because being frisked is bad, that wouldn’t be politically tenable. So the NYPD focused very specifically on young men living in high-crime, lower-income neighborhoods.

As an abstract statistical optimization, that’s not incorrect. But it’s also not fair. There used to be a tendency for some academics to make a big deal out of the difference between “taste-based discrimination” (cabbie won’t pick up the Black passenger because he hates Black people) and “statistical discrimination” (cabbie won’t pick up the Black passenger because he realizes he’s likely to live in a peripheral neighborhood where it’s hard to pick up new fares), but from the standpoint of the victim of discrimination, who cares? At the end of the day, stop and frisk as a policy punished a subset of the city’s young men who had not committed a particular crime for superficially resembling a group of people who may have. And the discriminatory enforcement was even out of proportion to the statistical disparities in underlying offenses. It was unfair, it generated tremendous resentments, and it was unnecessary.

New York didn’t need stop and frisk

As I found myself more on the “law and order” side of various policy arguments in the summer and fall of 2020, I was always struck by how unreflective many of my fellow travelers were about how they were wrong about stop and frisk.

Bill de Blasio got elected mayor on a platform of ending the practice, and shortly before his election, a federal judge ruled that it was unconstitutional. Conservatives warned that crime would surge, but instead it continued to fall for several years. That, I think, discredited the police and their allies in many people’s minds and contributed to the growth of an empirically implausible view that cities could do without policing altogether.

But here’s the deal. We have many studies (Chalfin & McCrary; Chalfin, Hansen, Weisburst, and Williams; Mello) showing that all else being equal, cities with higher police staffing levels have less crime. We also have studies (Klick & Tabarrok; Rosenfeld) showing specifically that putting additional police officers in a specific geographic area reduces crime there. Yet it’s precisely because “have more cops around” is such an effective anti-crime strategy that it’s mistaken to reach superficial conclusions based on the apparent success of Bloomberg-era stop and frisk initiatives. When John MacDonald, Jeffrey Fagan, and Amanda Geller looked in detail at Operation Impact, they found that 100% of the local-level crime reduction was attributable to the presence of additional officers and zero percent to the practice of doing stop and frisks without probable cause.

A complementary study by Rosenfeld and Fornango finds no impact of stop and frisk on robberies and burglaries.

This was a really bad policy that subjected lots of people to discriminatory treatment for years for no good reason and contributed to the unraveling of the urban civic compact.

It’s not clear what Adams is actually proposing

Back to mayor-elect Adams: the headline of his op-ed in the Daily News says “How we make New York City safe: Mayor-elect Eric Adams explains why we need stop and frisk and proactive policing.”

But if you actually read the piece, he appears to be defending not the specific policy known as “stop and frisk” but the general principle that sometimes police officers should stop a person and frisk him.

In fact, as American courts have affirmed over many years, stop, question and frisk is a perfectly legal, appropriate and constitutional tool, when used smartly, as opposed to indiscriminately against hundreds of thousands of young Black and Brown men, as it was for years in New York City. Not only that, but it is a necessary tool, whereby police approach someone who fits a witness description or otherwise appears to be carrying an illegal weapon.

It’s true that police abolitionists are a real thing in the United States, but the judicial precedent that ended stop and frisk in New York didn’t say anything like that, and when de Blasio ran for mayor on a platform of ending stop and frisk, he did not adopt a policy preventing the NYPD from ever stopping a suspect.

So if you take Adams literally, he’s not proposing any policy change at all. But he clearly wants it to seem like he is proposing some kind of middle ground between de Blasio’s policy and Michael Bloomberg’s. And he’s likely signaling this because the NYPD unions hated de Blasio, and Adams wants them to like him.

But this is an extremely difficult thing to flim-flam around. If Adams goes with the headline “we need stop and frisk,” then the risk is that some police supervisors (cops appear to be highly sensitive to direction from middle managers) and rank-and-file officers will interpret that as a green light or perhaps even a mandate to return to the Bloomberg era approach. If that’s not what he wants, he needs to be a lot clearer and not play this game — both because the old policy was bad, but also because it sets off a cycle of delegitimation. Most American cities have experienced a huge surge in shootings since the George Floyd protests, a surge that is likely related to some form of de-policing. You can interpret that de-policing in a few different ways,

The lesson of the Operation Impact study is that there is absolutely no need for that — a visible presence on the street and bona fide arrests based on probable cause do all the work.