It’s July 2023. The chief of police of the city of Los Angeles, Michel Moore, addresses the monthly graduating class of the Los Angeles Police Department’s training academy. “This is a profession that many will say, and I will say, is the hardest profession in America,” Moore says.
Moore is the picture of modern American law enforcement: close‑cropped hair, immaculate black uniform, commanding voice, an air of quiet professionalism.
In front of him are the recruits, fresh‑faced and eager, sitting on rows of white folding chairs in the academy courtyard. Behind them are their families, ready to celebrate. A bagpiper serenades the crowd. The police band plays. The sun shines. It is a perfect Los Angeles day.
“Far too many believe they know how to do it,” Moore continues. “But few actually step forward. And [of] those few that do step forward, even fewer have the qualifications to meet the demands of what it means to be a police officer.”
[Reihan Salam and Charles Fain Lehman: We’re underfunding the police]
At that, the young men and women in the graduating class leap to their feet and throw their caps into the air.
But there is something profoundly strange about this and all the other graduation ceremonies that happen every month in the academy courtyard—something that would leave almost any police officer in the developed world shaking their head in bafflement. And that peculiarity is a clue to one way in which Americans misunderstand their lethal-violence problem.
Los Angeles is a city of 3.9 million people. The Western city most comparable to it is probably the German capital of Berlin. Los Angeles covers a much larger area, but the populations of the two cities are virtually identical. Both are centers of government, culture, and industry—wealthy, world‑class cities at the heart of vibrant metropolitan regions. To patrol the 469 square miles of Los Angeles, the city’s police department employs roughly 8,500 police officers. How many police officers do you think Berlin has?
As I was writing this chapter, I asked dozens of well‑educated people that question. Almost every answer I got put the Berlin police force somewhere between a quarter and a half the size of the LAPD.
I suspect that if I conducted an opinion poll of many millions of Americans, most would answer the same. Americans believe themselves to be overpoliced. Many conservatives believe that to be a good thing. Many liberals believe it to be a troubling fact. But few question that America is a place that spends more money on law enforcement than do its peers around the world.
And that belief is completely and totally wrong: Berlin has about 18,000 police officers, making its police department twice the size of Los Angeles’s.
American police departments, compared with their counterparts around the developed world, are tiny. Phoenix has 2,500 police officers. Munich—which has a population of roughly the same size—has nearly two and a half times as many. Dallas has 3,100 officers. The Brussels metro area, comparable to Dallas in population, has 6,700.
Had anyone from Europe happened upon the graduation ceremony that day in July, their first thought would have been, Where is everyone?
Aurora, Colorado, sits half an hour east of Denver, in the foothills of the Rocky Mountains. It’s the third-largest city in the state, with more than 400,000 residents, and is more diverse, economically and culturally, than its larger neighbor. It’s a very typical modern American city, and for close to a decade its police chief was Daniel Oates, a longtime veteran of the New York City Police Department. Oates is part of the most recent generation of law-enforcement leadership: better educated than his predecessors, more comfortable with data and analytics, and more aware of the importance of the police department’s relationship to the public.
I called him and asked a simple question. Aurora is the same size as Zurich, but Zurich has 1,700 police officers and Aurora had 650 when Oates was running the police department. What would have happened, I asked, if I could have given you another 650 officers, putting you in the same range as Zurich? He burst out laughing at the thought. “It would’ve been glorious,” Oates replied.
He had clearly never been asked this question in an interview before. But he had certainly thought about the subject long and hard, because he then launched into a compelling, granular 30-minute description of exactly what he would have done in Aurora with twice as many officers.
Half of those extra officers, he began, would be assigned immediately to patrol—that is, responding to radio calls. He continued:
You’re always looking at response times. You’re looking at the 24 hours of the day; you’re looking at the seven days of the week. You’re looking at calls for service, and your ability to get to the higher‑priority calls as quickly as possible.
So if you gave me, in Aurora, another 600 cops, all right, probably around 250 or 300 of them, I’d put [them] right on patrol. And half of those would go to filling in the holes in the 24‑hour, seven‑day‑a‑week chart where I felt the response times were not sufficient, because you can really, really protect people, reduce victimization, if you can get there faster. Period.
Police departments keep careful records of exactly how long it takes a squad car to respond to a radio call, so Oates could identify exactly when and why response times got unacceptably long. Suppose you noticed that week in, week out, people in a certain part of the city were waiting 45 minutes for an officer to arrive between 10 p.m. and midnight. In Aurora, Oates had lots of those holes in the schedule—so many, in fact, that he was hundreds of officers short of a fully staffed patrol team. With more officers, he could have plugged those holes.
Then he began to talk about what he would do with the other 300 new officers. He had a long list. At the top was a greatly beefed‑up community-policing division:
The person who doesn’t answer the radio, but builds relationships in the community and is the day‑to‑day representative of the organization dealing with, you know, not urgent crime problems, but problems in a community for which police intervention could make a difference. Okay, so things like graffiti or kids hanging out in the park at night and doing stupid stuff.
A police department relies on the goodwill of the population it is serving in order to be effective. And Oates’s point was that if your force is composed mostly of a small number of overworked officers rushing from one crime scene to the next, you never get the chance to build that goodwill. With more officers, the department could do that. As he told me:
If the relationships are close, where someone trusts the cop and the cops trust the community and they’re welcome into their home, they’re welcome into their meetings. They’re being invited to barbecues, whatever, and christenings and all that kind of stuff.
If that’s the relationship that the cop is having with the community and the community leaders, then when a problem emerges for which the police or government intervention could make a difference, it emerges early. We hear about it before it’s a big problem, and not only do we hear about it from the community when it’s a little problem, we hear about it in a form of dialogue and a relationship, where ideas about how to address the problem also emerge from the community because there’s a level of trust.
As Oates was talking about his experiences in Aurora, I was reminded of something the criminologist Jens Ludwig said about what he learned from riding around at night with Chicago police officers: Once you realize that most murder is expressive, you understand the value of having lots of police officers. An officer cannot easily disrupt a carefully planned act of instrumental violence. But they can interrupt the kind of argument that builds into a lethal encounter. As Ludwig put it:
So let me give you, like, the clearest-cut example, right? So I’m driving around one night. Super late at night on the South Side of Chicago with a sergeant at the time on the Chicago police force. We’re in an unmarked car. We turn the corner, and there’s this gigantic guy. He’s got this woman pinned against the fence, a chain‑link fence on the South Side. She’s holding a little kid’s hand standing next to her, and he is right in her face, screaming, swearing. Like, you can’t imagine how heated this thing is.
The overwhelming majority of female murder victims are killed by someone they know, and this had all the hallmarks of something that could spiral into exactly that situation. Who knows what trivial matter had set the man off? But he was angry. Ludwig continued:
And she’s just, like, cowering, pushed all the way up against the fence, nowhere to go. So the cop pulls [the] car over and says, “What’s going on here?” And the guy says, “Everything’s fine.” And the sergeant says, “Buddy, I think you should go take a walk.” And the guy says, “I think everything’s fine.” And the cop says, “Go take a fucking walk.”
And then the guy, like, very grudgingly walks away, and we then wind up giving the lady and her kid a drive home.
The gigantic guy didn’t commit a crime. And the officer didn’t make an arrest. What the officer did do, however, was change the chemistry of the situation—interrupt the kind of encounter that could lead to an act of real violence.
When I told Daniel Oates that story, he agreed completely. He said that for every arrest an officer makes, he does 10 or 15 or even 20 of those spontaneous interruptions. And if Oates had had twice as many officers, they could have interrupted twice as many of those arguments, and defused them before they turned into something far worse.
This is what is happening in Western European countries. They have much bigger police departments because they understand exactly what all those extra police officers can do. They believe in police. And that is the other thing that would have baffled any visitor from Europe who happened upon graduation at the LAPD academy. You Americans write best-selling novels about cops. You glorify them in movies and television shows. So why don’t you believe in them?
I asked Oates whether a meaningful expansion could ever happen. “If you went before the town council or a community meeting in any city, and you were like, I’m here to argue for doubling the size of your police force, and you gave the exact same speech you just gave me, what are the odds they give you what you wanted?” Oates’s answer: “Zero. Zero. Zero. It’s not gonna happen.”
In 2020, an African American man in Minneapolis named George Floyd was killed by a rogue police officer, setting off an extraordinary wave of protests around the country. The motto of many of the protesters was “Defund the police.”
If the police were badly behaved, the thinking went, it was because there were too many of them.
As the prominent criminal-justice reformer Mariame Kaba wrote in The New York Times at the height of the protests:
Regardless of your view on police power—whether you want to get rid of the police or simply to make them less violent—here’s an immediate demand we can all make: Cut the number of police in half and cut their budget in half. Fewer police officers equals fewer opportunities for them to brutalize and kill people.
Following the protests, city after city saw their police force shrink. Of the 15 largest police departments in the United States, 12 had a markedly smaller police force in 2025 than in 2019. Philadelphia lost 23.1 percent of its police force during that period; Washington, D.C., lost 21.4 percent; San Francisco, 17.5 percent; Phoenix, 15 percent; and on and on. The reformers won the day. But what they did not understand is that there is a trade-off between certainty and severity: When criminals think they can get away with crimes, policing and punishments need to be more severe to achieve the same deterrent effect. Severity is not something that comes from lots of police. It’s something that comes from having too few police.
In a brilliant study titled “The Injustice of Under-Policing in America,” the scholars Christopher Lewis and Adaner Usmani show that this trade‑off can be found in police departments across the globe. They took a broad sample of countries and ranked them by their levels of police violence: How many civilians, in the course of a given year, were killed by law enforcement? Then they compared that number with the capacity of a country’s police department. (The statistic they used was police per homicide, which is a more accurate indicator of the amount of resources available to a law-enforcement community than the standard measure of police per capita.)
[Listen: The evidence on policing and crime]
Lewis and Usmani found that in the countries with the most police capacity—such as Switzerland, Norway, and Japan—it’s almost unheard of for a civilian to die in an encounter with law enforcement. But the fewer police per homicide a country has, the more violent the police are in their dealings with the public. As Usmani told me:
If you think about a country like Brazil or a country like South Africa or a country where I’m from, Pakistan, these are countries where, actually, the police are quite scarce. Everyone knows they’re quite scarce, and they also know that they behave incredibly heavy—heavily handedly there.
If a government doesn’t have lots of police, then it has only one way to demonstrate its authority: with spectacular shows of force.
Where is the United States in Usmani and Lewis’s chart? Right next to Pakistan, Iran, and Brazil.
“There’s a Netflix show, Flint Town, which follows the Flint, Michigan, police department before and after the election of Trump,” Usmani said. Flint is a city of 80,000 people—with a police force, at the time, of 93 full‑time officers. Usmani continued:
What you see the Flint Police Department struggle with is precisely the fact that they only have a few officers to police a place that’s very violent, and so they invent counterinsurgency. You observe them inventing counterinsurgency, where they’ll go out in a few heavily armed jeeps and basically break some knees and return to base. And that’s a very different style of policing than the developed world, but very familiar in a country like Brazil.
In recent years, many Americans have decided that their country’s problem was that it had too many police officers, when in fact its problem was that it had too few.
This article has been adapted from Malcolm Gladwell’s new book, The American Way of Killing: The Invention of an Epidemic.