Crime in Context, Locally and Nationally

Beginning in 2022, at the end of each year, Vital City has asked some of the country’s and New York City’s most influential policymakers and researchers to comment on the most important trends in public safety and to speculate on what the future might bring. The exercise is open-ended and serves as an invitation for some of the most thoughtful thinkers in this space to simply share what’s on their minds. Commenters come from academia as well as the policy world and hold views from across the political spectrum. What each of the contributors shares is a respect for data and a desire to understand crime and improve public safety in big cities like New York.

I have come to look forward to reading these contributions every January as I pore over the year’s final crime statistics and engage in my own reflection about what the numbers are telling us. I am therefore honored to once again have the opportunity to read these wonderful contributions, to provide a synthesis for Vital City’s readers and to share a few of my own thoughts about what the data mean for public safety in New York City and for the country.

Looking back to the dark days of the COVID-19 pandemic and the aftermath of the killing of George Floyd, murders in the city had risen by 40% year over year in 2020, and they increased again in 2021 before falling only modestly in 2022. Other crimes rose as well, and some continue to remain elevated. Today, we have several years of postpandemic data and therefore the benefit of a deeper sense of perspective.

Much of what we’re seeing in New York follows national trends. Across large U.S. cities, murders are down by approximately 17% since last year, with cities like Philadelphia and Baltimore registering particularly impressive declines. The data are consistent with the idea that, as the pandemic has waned, life has returned to normalcy. Many, albeit not all, workers have returned to the office, schools and community centers are open for business and law enforcement has returned to public spaces, having stepped back for public health reasons — and possibly in response to negative community sentiment — in 2020.

However, a deeper understanding of violence in our cities demands a more detailed look at longer-run trends. As readers of Vital City will know, New York City experienced a dramatic decline in violence during the 1990s. While murders declined nationally by 50%, they declined by over 70% in Gotham. The causes of the so-called Great Homicide Decline are still debated and are far from completely understood, but most researchers believe that the waning of the crack cocaine epidemic, investments in police manpower and management and the rising prison population probably had something to do with the generational decline in lethal violence. Demographic and economic explanations like changes in wages and poverty rates do not explain the decline in violence well.

What is less well known is that New York City experienced a second great homicide decline between 2010 and the COVID-19 pandemic, with murders once again falling by around 50% — this time, during a period when murder trends were flat nationally. New York City’s decline in murders during this period was overwhelmingly concentrated among gang and group violence and came on the heels of a remaking of policing that prioritized gang enforcement over the less well-tailored policy of making very large numbers of police stops and misdemeanor arrests, a policy that was ruled unconstitutional by a federal judge in 2013. The pandemic temporarily disrupted the improvement in public safety, but, with murders and shootings down once again, a return to normalcy is on the horizon.

At the same time, the nature of violence has shifted underneath our feet. Daniel Penny’s acquittal for the death of Jordan Neely reminded many New Yorkers of Bernard Goetz’s 1987 trial for shooting four would-be robbers and has invited a number of comparisons in the popular media. Superficially, the two cases might seem similar. Both events happened on the city’s subway system and both litigated a claim of self-defense against would-be attackers. But in other ways, the cases are quite different and serve as particularly salient examples of what the broader data tell us about a shift in the nature of violence in the city.

In the bad old days of the 1980s and early 1990s, many New Yorkers had a fear of being robbed, especially by kids — who were predicted to become a new generation of superpredators. There were more than 100,000 robberies in the five boroughs in 1990, which, at the time, accounted for two-thirds of violent index crimes. By contrast, there were only 17,000 robberies in 2024 — an 83% decline despite a rising population — with robberies accounting for just one-third of violent index crimes. In short, in a generation, the principal violence problem in New York City shifted from robberies to assaults.

There are even broader forces at play here. Indeed, crimes with a financial motive have, for years, been in free fall, while crimes borne out of anger have risen. Since 2001, well after the great crime decline of the 1990s was over, robberies, burglaries and grand larcenies in the city have fallen by 40%, 58% and 47%, respectively. Meanwhile, felony assaults — violent crimes which have no known financial motive — are up by 21%. The same pattern seems to hold among homicides as well, with many observers in and outside of law enforcement reporting that a considerable majority of murders are the result of petty disputes and not robberies gone wrong, drug trafficking or other forms of financially motivated gang activity. Today, the violence is seeded in social media posts, not in fights over street corners. Notably, the same pattern holds among subway crimes — a category of crime that has fueled news cycles in the city and beyond — with violent subway assaults tripling over the past 15 years amidst a sharp decline in financially motivated crimes like robberies and grand larcenies.

What’s going on? Why is this happening? We don’t know for sure, but …

A second feature of New York City’s violence problem that has shifted is the age of the perpetrators. Just 15 years ago, the average arrestee charged with a violent street crime was 23 years old, with a sizable share under the age of 18. Today, the average age is 32, and fewer than 10% of violent street crime arrestees in New York City are under the age of 18. Proportional to their numbers in the population, youthful violent offenders are only slightly more common than those aged 45 to 64. A similar pattern holds nationally, albeit to a lesser degree. Given the biological basis for youth violence, this is an astounding shift in such a short period of time and demands new theories to explain the nature of today’s victimization.

With these thoughts in mind, contributors to this volume offer a number of compelling insights into what happened in 2024 and what we should look out for in 2025.

Jeff Asher starts us off with a high-level look at what happened to violence in the United States in 2024. No one follows the data more closely than Asher, so his take on the big picture should be of great import for any interested observer. Asher begins by noting that murder was down by 17% among 309 U.S. cities with available data covered in the Real-Time Crime Index. Overall, reported violence was down by around 4% and reported property crimes were down by around 9%. Alex Piquero, Chandler Hall and Nick Wilson consider variation among cities across the U.S., noting that some cities have seen violence decline more than others. What do the most successful cities have in common? As they note, these cities seem to have focused their attention on the people and places that drive the violence and have adopted multiagency, multipronged violence prevention partnerships that adopt a public health approach to violence prevention, engaging a wide range of partners alongside law enforcement.

In keeping with the sizable declines in lethal violence the country has experienced since 2020, and especially over the longer span of time, Asher, Caterina Roman and Marcos Soler each point out in separate contributions that there is a considerable disconnect between the crime data and what members of the public say that they are experiencing. While murders and robberies are down considerably from their peak levels, Americans’ fear of these crimes is at a 30-year high. Nearly two-thirds of Americans think that crime is rising nationally and 55% of Americans believe that crime is rising in their community. While Americans seemed to notice the great crime decline of the 1990s, crime shifts since 2000 seem to be more difficult for the public to detect. One of the findings from the available survey data, noted by Soler, is that Americans simply trust crime data less than they used to and they also have less trust in the public officials charged with using those data.

While the available survey data and data on violence suggest a mind-warping disconnect, Charles Fain Lehman reminds us that disorderly behavior is also a major driver of New Yorkers’ level of fear and well-being. Lehman considers why the public has gotten tougher on crime as serious crimes like murder have started falling and concludes that disorder itself — including problems like disorderly behavior by homeless people, public drug use and shoplifting — may explain the perceptual gap. While an analysis by Soler suggests that changes in public fear do track changes in murder, Lehman’s suggestion about the salience of disorder appears to hold consistently in surveys of New Yorkers. The disconnect may also be part and parcel of a shift from more instrumental to less instrumental violence, which might feel more random to people and harder to avoid.

Peter Moskos and Megan McArdle have chosen to focus their attention on particular types of crime that have been especially salient for New Yorkers in 2024. Moskos writes about crime in the city’s subway system, providing a fascinating look back to the late 1980s and early 1990s — the “bad old days” before the city’s first great crime decline. Moskos documents policy solutions that people in charge pursued at the time to make the subways safer, which included moving homeless people out of the transit system as well as a resurgence in police activity, including an increasing number of misdemeanor arrests and summonses. McArdle’s piece tackles shoplifting. While it is a minor crime when compared to murder and assault, McArdle makes the case that, due to the enormous volume of shoplifting, shoplifting can nevertheless have a devastating effect at scale, leading to “a kind of urban murder.” If retail stores are rendered economically nonviable, she argues, the result will be blighted streetscapes, fewer eyes on the street and the diminishment of the most organic form of crime control of all — that which comes from the unencumbered use of public space by citizens, each of whom is a potential witness to criminal activity that happens in their presence. As Jane Jacobs wrote in her seminal work, “The Death and Life of Great American Cities”: “The safety of the street works best, most casually, and with least frequent taint of hostility or suspicion precisely where people are using and most enjoying the city streets voluntarily and are least conscious, normally, that they are policing.”

Subway violence and shoplifting aside, Gerard Torrats-Espionosa and Andrew Papachristos each note that, amidst the decline in lethal violence, there is still one empirical regularity that defines our understanding of violence: Violent crimes are incredibly concentrated among a small number of people in a small number of communities, all of which are poor and marked by considerable social disadvantage. In New York City, many of these communities are in the Bronx and, as John Hall notes, residents of the Bronx have yet to experience the return to normalcy that residents of other boroughs are beginning to experience. The Bronx is currently experiencing the highest number of major crimes since 2000, and Hall suggests a number of pragmatic approaches that could be employed to bring greater safety. Beyond the tactical approaches Hall has suggested — many of which involve law enforcement — Torrats-Espinosa and Papachristos call attention to social science evidence in favor of interventions that support communities — such as summer youth employment programs and neighborhood revitalization efforts — as well as emerging, albeit still speculative, evidence on the impact of community violence intervention programs, interventions that are intended to empower community members to coproduce public safety.

Recognizing the intense concentration of violence among a small number of people and places, Megan Kang describes her critically important ethnographic research on the people who carry illegal guns — and who therefore disproportionately drive the risk of violence. Kang notes that illegal and legal gun carriers have similar motives for carrying: the fear of becoming the victim of a violent crime and the desire to be able to protect oneself. In a country that guarantees its citizens the right to bear arms and in which guns are readily available, Kang’s research looms large and provides a sense for how gun violence might be reduced in a city like New York. Recognizing that gun carrying — not simply ownership — adds risk to any interaction, and that many illegal gun carriers would rather be caught with a gun than without one, it is important to recognize that making the city safer is one of the most direct ways to end the feedback loop between violence and gun carrying. Of particular note is that in 2019, when the city experienced its lowest homicide rate since the early 1960s, the share of homicides perpetrated using guns was just 55%, a considerable outlier in a country in which over 8 in 10 big-city homicides were committed with a gun. Safer cities ultimately mean fewer people feel the need to arm themselves on a daily basis.

Finally, in a particularly fitting contribution for a data-driven periodical like Vital City, John Roman reminds us how little we actually know about the nature of crime and violence and urges developing a more comprehensive framework to think about crime shifts over time. Social science methods are designed mostly to test the effects of interventions (e.g., Did a medical treatment reduce mortality? Did a court reform affect public safety?) and are less well designed to understand why the world looks as it does (e.g., Why did crime rise in 2020 and fall after 2021?). Roman argues we need to develop a more holistic understanding of crime at the macro level, explaining the cyclical nature of crime, accommodating different types of crimes as well as how crime varies across sectors and regions. While scholarly research is an important input into policymaking, the research process is slow and, as Andrew Papachristos reminds us, can privilege the status quo. Therefore, good policy design relies quite a bit on theory and developing a macro view of the crime problem.

Where does this leave us in 2025? On the one hand, the data suggest that, in many respects, public safety is beginning to turn to normalcy after an alarming rise in violence in 2020 and 2021. On the other hand, assaults in New York continue to rise, following a trend that goes back at least 15 years. Three important empirical regularities loom large in thinking about the effectiveness of the City’s response. First, crime is best deterred by increasing the certainty that an individual who offends will be punished and is less responsive to the unlikely prospect of a very severe sanction. Second, crime is heavily concentrated among a small number of repeat offenders. Third, crime is a complex problem that requires a portfolio of solutions, which will tend to include investments in both law enforcement and communities. Tailored solutions tend to be the best solutions.