Powered by 
Cities are the economic engine of America. The gravest threat to that engine is gun violence. Beyond the lives lost and families devastated, every murder that happens in an American city reduces that city’s population by seventy people—fewer people and businesses moving in, more moving out. Gun violence is an important reason why Chicago has lost a million people since 1950, why Baltimore shrank by 40 percent, and why Detroit has lost fully two-thirds of its population.
Existing approaches to the problem have led to a lot of political bickering and not a lot of progress. Gun control tends to dominate the public debate. But the government is also in the business of behavior change, trying to reduce the chances people use guns to injure or kill one another. Governments deploy criminal justice “sticks” to deter violence and bigger social policy “carrots” to help lure people away from crime. These efforts have been in various combinations politically difficult, expensive, and ineffective.
The result is that while most leading causes of death have plummeted since 1900, the murder rate in contrast is not that different today from what it was 125 years ago. Our murder rate is five times that of the United Kingdom, seven times Australia’s, and twenty-five times that of Japan. Almost all the difference is due to murders committed with firearms.
The good news is that this conventional wisdom largely misunderstands the problem of gun violence. Why is that good news? Because it implies that the inability of past policies to solve the problem of American gun violence doesn’t mean the problem can’t be solved—it just means we’ve been focusing on the wrong things. Seeing the problem in a different way points to a new set of policies that are nonpartisan and cheap (relatively speaking), and, an accumulating body of data suggests, can also be remarkably effective.
Conventional wisdom and its limits
When Pew researchers asked Americans what they think causes gun violence, the top response is the ease of getting guns illegally (86 percent), followed by “family instability” (74 percent), which many think harms children’s moral or character development, and lack of economic opportunity (65 percent).
The focus on guns and how they are regulated is not surprising given that the overwhelming majority of murders in America are committed with guns. The role of gun availability itself in contributing to the gun violence problem is ambiguous as a conceptual matter. Guns make violence between people more deadly, which criminologists call the “instrumentality effect.” But guns can also help people defend themselves against criminal attack, and even the threat of this sort of defensive gun use could potentially help deter future crimes. The data suggest that the instrumentality effect dominates, so that places with more guns wind up having more murders.
Yet for better or worse, regulating firearms in America has been politically difficult. While in surveys most Americans support more gun regulation (with the exception of an outright ban on handguns), the minority that opposes regulation is extremely energized and organized. The result is a fairly modest set of federal regulations. Cities and states can supplement federal laws, but open city and state borders make it hard for any single place to unilaterally regulate its way out of the problem. In my hometown of Chicago, for instance, almost none of the city’s crime guns were bought in Chicago; they almost all come from nearby suburbs or nearby states like Indiana.
Public policy has focused not just on the issue of gun availability but also on behavior change: reducing the rate at which people use guns to hurt one another.
The widespread perception that gun offenders are characterologically bad people is one important reason behind the country’s massive investments in expanding the criminal justice system since the 1970s, especially the prison population. While at least the initial increases (through the early 1990s) are thought to have helped reduce violent crime, most experts think that strategy has diminishing marginal returns.
The widespread perception that lack of economic opportunity drives gun violence, driven partly by the concentration of gun violence in low-income areas, has led to a series of policies that try to address the “root causes” of violence going back at least to FDR. But most cities have found it enormously difficult in practice to substantially change deep-seated social conditions like economic or racial segregation. And even when public policy has been able to ameliorate some social problem like poverty, it has typically not had the hoped-for effect on violence.
Following the conventional wisdom has, in other words, been a bit of a path to nowhere.
A different perspective
Why haven’t the policies suggested by conventional wisdom been more effective? Part of the story is that conventional wisdom relies on some assumptions about the nature of gun violence in America that don’t seem to fit the facts particularly well.
Consider for starters the traditional approaches to behavior change, to reducing the rate at which people use guns against others: depending on one’s politics, this has been mostly either bigger criminal justice “sticks” or bigger social policy “carrots.” Yet, interestingly, both sides have a shared point of agreement: that the solution to interpersonal gun violence is to change the incentives for violence. That is, the implicit assumption is that before anyone pulls a trigger there is some sort of rational comparison of benefits and costs. The mental image people presumably have of gun violence is something like organized, hierarchical gangs fighting for drug turf.
But that assumption does not seem to be correct. It is true that there are indeed some shootings that stem from gang wars for drug-selling corners, or people shot by robbers trying to get their property. These are what criminologists call “instrumental violence,” where the violence is a means to some larger end. But data from the FBI’s Supplemental Homicide Reports (SHR) suggests that type of violence accounts for only something like 1 of every 5 murders in America.
The overwhelming share of violence in America—something like 4 of every 5 murders, according to the SHR—is what criminologists call “expressive,” where the violence is not a means to some other rational end: the violence is the whole point. Someone is, to put it plainly, pissed off. That emotional reaction is so strong it turns into violence.
These are garden-variety arguments that escalate into tragedy because someone has a gun. Who gets into arguments? You. Me. Everyone.
And if you’ve ever been in a heated argument, you can attest that the last thing you are in that moment is a perfectly rational benefit–cost calculator. I think about the time I was walking our family’s pandemic puppy, Aiko, a seventy-pound mix of hound dog and German shepherd, when a neighbor’s dog came barreling down their driveway, off leash, barking, snarling, baring its teeth on the way to attack Aiko.
What to do? I could have used the sort of slow, deliberate, effortful thinking that behavioral scientists sometimes call “slow thinking” or “System 2.” That’s the little voice in our heads that we’re aware of that gives us the capacity to sometimes be rational (recognizing that the little voice in our head is not always rational). That type of thinking would have led me to realize I could have just put myself between the other dog and mine, or picked Aiko up, or, worst case, let her get into a scrap and resign myself to yet one more vet visit.
Sadly, my System 2 deliberate self was nowhere to be found at that moment.
One of the most important lessons from behavioral science over the past several decades is that the little voice in our heads doing deliberate thinking is not the only type of thinking that we do. Because System 2 deliberate thinking is effortful, our minds are designed to do as little of that as possible. We’re even less likely to use that type of thinking when we’re under stress, since the sort of adrenaline dump you get when another dog is attacking yours depletes mental bandwidth.
As this other dog was charging toward mine, my other channel of thought—the effortless, automatic responses that happen below the level of consciousness, or System 1—led me away from a rational, conciliatory response. Some of my System 1 responses that are typically so helpful for me in routine, low-stakes situations got overgeneralized and led me into trouble.
For example, the tendency for System 1 to engage in egocentric construal—to personalize and think everything’s about me—is helpful in lots of day-to-day situations. When my wife or daughter (or both) come into the kitchen in the morning with a furious look on their faces, there really is a reasonable chance it’s due to some stupid thing I did. But when the neighbor’s dog was charging down the driveway—most likely an accident, a dog slipping its leash, etc.—egocentric construal led me to feel like “Why is he letting his dog do this to me?”
Similarly, throughout the long arc of humankind, it’s been helpful for System 1 to quickly and effortlessly assess how things are going. It does that by categorizing things in a simple binary or black-and-white way, which creates two types of feelings: “fine” or “end of the world.” Since the neighbor’s dog charging down the driveway towards Aiko is not fine, the only other option was end of the world—known as “catastrophizing,” or making a bad event feel even worse than it is.
From my neighbor’s perspective, he assumed I must know what he knew––that this was just an accident, his dog had slipped the leash (a version of what behavioral scientists call the “curse of knowledge”)––so screaming the most horrible profanities imaginable at him right in front of his home, where his wife and young children were having breakfast, made me the unreasonable one in this interaction. We both believed we were the good guy here.
Here is where everyone involved got lucky: Hyde Park is chock full of what journalist Jane Jacobs, in her wonderful book The Death and Life of Great American Cities, famously called “eyes on the street”––people around the neighborhood who enforce a shared set of social norms, including stepping in to interrupt conflict before it escalates. Being near the University of Chicago, one of the private security guards that the university has hired to patrol Hyde Park happened to be nearby stepped in and broke things up.
This is what most shootings in America stem from: System 1 rather than System 2 thinking. Misconstruals of the situation, misunderstandings, faulty mindreading, quick responses before considering all options and their pros and cons. No wonder traditional policies that have focused on changing the incentives for violence (carrots and sticks) haven’t had bigger impacts.
Note this perspective also highlights a problem with a key assumption behind the American approach to gun regulation: that we can neatly divide the population into people who are versus are not at risk for using a gun to harm someone else, and that we can give most people unfettered access to guns so long as we keep guns away from the small share of high-risk people. In reality, anyone is at risk for misusing guns because everyone is prone to System 1 mistakes.
Maybe the clearest way to describe this alternative behavioral science perspective on gun violence comes from a conversation I had a few years ago with a staff supervisor at the Cook County Juvenile Temporary Detention Center. He said he tells most of the residents in there, “If I could give you back just ten minutes of your lives, none of you would be here.”
A different set of solutions
This new perspective on gun violence suggests the solution isn’t so much about changing incentives to get people to decide differently, as it is from preventing people from making life-altering mistakes in their decision making. We can either help people learn more about the sorts of mistakes they are prone to and how to avoid them, or can employ different strategies to increase the chances that some third party is around to step in and interrupt a bad decision and the conflict it has fueled before that escalates into violence—the same way that the University of Chicago security guard helped me from a decision I would have regretted.
Teaching meta-cognition: The good news is people can be taught to recognize and avoid many of the situational misconstruals, catastrophizing, and other automatic System 1 thoughts that shape (and sometimes distort) decisions in high-stakes moments. The key capacity to develop is the ability to step back from one’s thoughts and think about them more objectively, or “metacognition.” A series of randomized experiments from Chicago and elsewhere show that as little as fifteen to thirty hours of participation can have remarkably large effects: up to 50 percent reductions in violent-crime arrests and 33 percent reductions in weapons arrests.
The even better news is that these programs can be delivered at close to zero cost in ways that—assuming we can maintain program fidelity as scale increases—can have population-level effects on gun violence. For example, a large share of the shooters or victims of shootings have had some previous contact with the criminal justice system. That means infusing this type of programming into detention facilities (jails, prisons, etc.) would reach almost all the people we most need to reach. Pilot demonstrations show that such programs can be delivered in detention centers very cheaply by either existing staff (guards, etc.) or graduate student volunteers.
Data-driven interruption: Once we realize the problem is interpersonal conflict, not just broad social conditions like poverty, segregation, etc., we can understand why there is so much predictability in when and where the sort of conflict happens that leads to shootings: because there is predictability to when and where we’ll get lots of social interactions that could go sideways and turn into arguments, especially involving either young people or inebriated people.
That suggests a very low-cost way to prevent shootings: concentrating our first responders (police, security guards, violence interrupters, etc.) on the times and places where gun violence is most likely. A growing body of research shows that statistical models are much better than human intuition is at identifying those violence “hot spots,” and that using data to target first responders can substantially increase the amount of crime prevention we get from existing criminal justice spending. This strategy is cheap, relatively speaking, because building and deploying statistical models is a rounding error compared to the larger criminal justice budget.
Neighborhood vibrancy: In a city like Chicago, there might be fewer than a thousand paid eyes on the street at any point in time—police, security guards, non-profit violence interrupters, etc. This is far fewer than the 2.7 million potential unpaid eyes on the street—city residents—who could step in and help interrupt arguments before they escalate into violence. That is, it’s not just neighborhood poverty that’s a key root cause to worry about, it’s also neighborhood vibrancy.
There is a growing body of evidence showing that even fairly modest, inexpensive changes to the built environment can substantially change neighborhood vibrancy and the presence of eyes on the street. For example, cleaning up vacant lots increases the willingness of neighborhood residents to spend time outside and reduces shootings on those blocks by 30 percent. Opening or closing a neighborhood store, which affects the amount of local foot traffic, can cause swings in crime and violence by 20 percent. Improved street lighting reduces serious nighttime crimes by 36 percent.
Some evidence that these effects are due not merely to deterrence comes from a study of the deployment in the Stockholm subway of security cameras—which may deter, but do not interrupt, injury. Cameras reduced property crime but not violent crimes. That would seem to suggest that the effects of changing neighborhood vibrancy on violence specifically must involve something that people can do that cameras can’t—like interruption.
These different solutions to the gun violence problem have a number of things in common. First, they are almost completely absent from the national policy conversation. In past presidential debates, for instance, almost all the conversation focuses on the traditional solutions implied by conventional wisdom: gun control, criminal justice, and root causes. Second, they’re all relatively cheap. Third, they’re largely nonpartisan—they sidestep some of the biggest fault lines we have in society about guns, the criminal justice system, what the government should or shouldn’t do to redistribute resources, etc. And, finally, the data suggest they all can work.
Enacting these policies would, for starters, save tens of thousands of families every year from the devastating loss of a loved one, a risk that very disproportionately affects black and Hispanic Americans.
But the benefits would also extend far beyond the victims themselves, given the important collateral costs of gun violence in destroying our cities. Schoolchildren can’t learn in places where shootings are common, businesses can’t thrive, and neighborhood residents won’t stay; the data show they leave in droves. The result can be a downward spiral in which people leave and, beyond taking their tax dollars with them, create less-vibrant neighborhoods, which fuels more gun violence, which in turn fuels more people moving out, and so on. We now have a set of feasible policies that can turn these vicious cycles into virtuous cycles.
Jens Ludwig is the Edwin A. and Betty L. Bergman Distinguished Service Professor at the University of Chicago; Pritzker Director of the University of Chicago Crime Lab; Director of the Bike Shop @ UChicago, a new center dedicated to using artificial intelligence to solve pressing social problems; an elected member of the National Academy of Medicine; and author of Unforgiving Places: The Unexpected Origins of American Gun Violence (2025, University of Chicago Press), which was named one of the New Yorker’s Best Books of 2025.