On Monday, Mayor Zohran Mamdani of New York announced that the city was disbanding the N.Y.P.D.’s Strategic Response Group, a high-profile specialized police unit that had been tasked with counterterrorism operations and monitoring protests, and which had become a particularly contentious flashpoint for the mayor’s left-wing supporters when it raided Columbia’s campus and violently dispersed pro-Palestine protests in the spring of 2024.
Disbanding the S.R.G. was a campaign promise the mayor sat on for almost a year as part of a détente on the role of the police that also included an ongoing partnership with the police commissioner, Jessica Tisch. The new policy is a sign of the changing political tides: In the wake of a big blue wave nationwide, one of the country’s most popular politicians feels secure enough in his standing to tackle police reforms that just a year ago seemed radioactive. But it is also a sign that we are all living in a very different world when it comes to crime than we were just a few years ago, at the peak of the pandemic crime surge. Back then, you heard a lot about social chaos and urban doom loops, the bad old days of the 1970s and the possibility that flight to the suburbs could wreck American cities once again. Instead, what followed was historic drops in violent crime almost across the board. By some measures, American cities are actually safer now than their European counterparts, which many Americans have long envied as walkable, high-trust utopias.
Between 2021 and 2025, the number of murders in New York City fell by more than 36 percent. Though this is miraculous, it isn’t exceptional: In those same years, the national murder rate fell by 40 percent. And it has continued to decline, according to estimates, by between 15 percent and 20 percent this year alone. Half of the number of people who would have been killed given crime rates in 2021 are now likely to survive given the rates of 2026.
When murder declined this much over the course of the 1990s, it was as if the clouds had opened and a new age of urban possibility dawned, with decades of white flight reversing and a future of information economy gentrification coming into view. We stopped talking about superpredators and the crack epidemic and started talking about police violence and eventually the new Jim Crow. We stopped worrying about child abduction and started worrying about excessive safetyism. We began hearing a lot less about burning neighborhoods and cities going bankrupt and a lot more about housing shortages and the unmanageable cost of rent.
That drop in crime was momentous, marking the end of the 20th century in American life and the beginning of the 21st, perhaps as decisively as the arrival of the internet or the end of the Cold War. In percentage terms, this drop already appears to be larger. It has also come faster.
Some large shares of these declines are returns to prepandemic trends, after a horrific Covid-era surge. Although the pattern may seem predictable and intuitive now, it was hardly conventional wisdom five years ago that all would quickly return to normal, or better than normal. Instead, spikes in violence and disorder looked to many less like temporary blips than a new semi-permanent reality. And yet the past five years haven’t just brought the country back to prepandemic levels of crime; murder rates are now well below the prepandemic trough, achieved in 2014. Last year’s murder rate was tied with 1955 and 1956 for the lowest on record, according to the F.B.I. This year’s rate will be even lower — perhaps the safest the country has been at any point in its entire history.
What makes the decline doubly miraculous is that no one can explain exactly why it is happening. In September the Council on Criminal Justice convened a summit, bringing four dozen experts together to hash out the various hypotheses and come to some sort of consensus. They came up with almost 50 possible explanations, from the end of pandemic disorder to the targeted use of violence intervention programs, the aging of the population and the decline of drinking, expanded surveillance systems and the loneliness epidemic.
Sometimes you hear conservatives talk about the effect of immigration enforcement or progressive retreat on bail reform. Liberals may talk about what is sometimes called community violence intervention — directly and proactively engaging with gang members and others at high risk of committing violent crime — or the benefits of social spending contained in President Joe Biden’s American Rescue Plan. Maybe we’re finally leaving the lead paint generation behind; maybe we’re getting better at treating gunshot wounds; maybe the spike in murders in 2022 left fewer would-be killers alive in subsequent years; maybe calls to defund the police really did demoralize officers even though few politicians echoed them, fewer localities implemented them and, nationally, the level of police employment recovered from a brief drop in 2021.
This is the kind of exhaustive list you come up with in the absence of any obvious answer, giving everyone an excuse to ride a hobbyhorse while making each of those explanations seem partial at best. And in the end, those four dozen experts at the Council on Criminal Justice couldn’t really agree on much. Only one factor was described as a major contributor by more than half the participants: the easing of social disruption from Covid, which isn’t even the kind of policy change that might be easily implemented or built upon.
That we remain so unsure what is driving such a consequential shift seems, at first blush, pretty weird. But it shouldn’t be, because crime isn’t unique in this respect. Many of the forces most profoundly changing the shape of contemporary life remain pretty inscrutable when it comes to causes and stubborn when it comes to policy response.
It’s uncomfortable to admit that we don’t really know what lies behind a rise in suicide rates or school shootings, for instance, in part because we want to believe we have a solution we can lift off a shelf and implement right away, and in part because epistemological humility is psychologically uncomfortable and politically unpalatable. In an age of social media factionalism, we also tend to weaponize crises to score partisan points, even if doing so requires us to flatten what are pretty deep social pathologies into problems we insist can be solved easily — if only those on the other side of the argument would let us. More often than not, the opposite is true: We don’t fully understand the challenge, or exactly what to do to address it, and the pathology persists in large part because it is complicated and hard to solve.
Positive surprises often remain mysterious, too, however eager you’d imagine leaders might be to claim credit. Why did suicide rates decline in the 1990s, for instance, and why have school shootings fallen over the last few years? In other cases, the mystery is why improvements are so limited or circumscribed: GLP-1 use has grown sixfold since early 2022, and more than one in 10 Americans reports using them for weight loss, yet the American obesity rate has fallen from 42 percent to 38 percent over that time, according to a running database maintained by Epic Research. In the world of education reform, there has been an awful lot of excitement about the so-called Mississippi Miracle, in which a new emphasis on phonics-based reading has generated some striking improvements in test scores in the South. But as even some of its most ardent proponents will acknowledge — and as classroom teachers might’ve predicted — not everyone implementing the new pedagogy has seen improvement. Few enduring social crises can be solved by silver-bullet policies, or even explained straightforwardly. Murder has been declining for decades now, and if anything, we know less about what has driven it now than we thought we did 10 or 20 years ago.
One thing we can know, however, is where it has left us — with the country less murderous than it has ever been in our lifetimes. Maybe than it has ever been. This is amazing progress, but it is also pretty disorienting.
For a very long time, Americans have understood that theirs was a violent country, prone to killing, perhaps especially because of the widespread presence of guns. In many parts of the country, this pattern still holds: Murder rates are much higher in American cities than in European ones, even with the enormous recent declines, and perhaps all of the difference is attributable to guns.
But if the pace of New York’s murder rate through October 2026 holds through the end of the year, the city will be safer than Berlin, Barcelona and Madrid, though still more deadly than many other European cities.
Perhaps even more striking: When looking past murder rates to lesser crimes, like assault and residential burglaries, it is no longer clear that U.S. cities should be regarded as more violent or crime-ridden than some of their counterparts in Europe. The best data we have shows that by several measures, U.S. cities are at least as safe, if not safer.
Comparisons like these are inherently tricky, which is one reason those studying crime prefer to focus on homicide rates: Dead bodies typically get reported to the police, and though there are some differences between what is counted as a murder in Lisbon versus Los Angeles, the differences are relatively small and the comparisons pretty easy to make.
But complicated is not the same thing as impossible, and for several years now the criminologist Matt Ashby at University College London has been working to standardize the relevant data to produce what he calls the “least worst comparison” between global cities — lining up jurisdictions and standardizing definitions and data to make it easier to compare the prevalence of car robberies in, say, Los Angeles and Rome. And the results are pretty eye-opening, especially for anyone who might assume that American cities are significantly more dangerous than those elsewhere in the rich world.
Comparing rates of assault with rates of injury, Ashby finds that Berlin is considerably more dangerous than New York, London is more dangerous than Chicago and Los Angeles is much safer than any of those cities. Rates of personal robbery are higher in Berlin and Paris than in New York and Los Angeles, he says. Rates in Chicago are almost twice as high as they are in New York — but rates in London are higher still. Rates of rape — which are especially difficult to compare — are almost exactly the same in Paris and Chicago, and comparable between Barcelona, Los Angeles and Seoul. New York is far better. And London and Sydney are much worse.
This pattern is so contrary to conventional thinking about American crime that it can be hard to believe. And the difficulty of comparison means each data point should be taken with a grain of salt. But the country has also always been slow to acknowledge good news.
The last few years of crime rate trends have been especially vertiginous, but the story of America getting safer is a long one, stretching back four decades now. And yet for 33 of the past 36 years, a majority of Americans told pollsters that they believed crime was getting worse. In 2025, only 49 percent of the country said so — the lowest number since 2001, and a sign that the country’s crime paranoia may be lessening if not disappearing entirely.
In some cases, there has been an even faster change in vibes. Forty-nine percent of voters in New York City reported crime as the most urgent issue facing the city in 2022. In 2026, only 9 percent of voters said crime was the most important issue in deciding who to vote for as governor. That is a spectacular drop in anxiety, which reflects pretty honestly the basic transformation of the risk landscape in the city. We don’t yet know what changes of this scale might mean for policy and politics, especially if the pattern replicates nationally. But one possibility is experiments with public safety reform of the kind that looked, as recently as that last mayoralty, like political albatrosses and atavistic relics of so-called Woke 1.
The post American Cities May Never Have Been Safer. We Don’t Know Why. appeared first on New York Times.




