Reducing Gun Violence Requires Thinking Differently

There’s a way out of this problem, however, and it starts with reframing the nature of gun violence. If you think again about the conventional wisdom, you might notice that there’s something the two sides, divided between the political right and the political left, share in their perspectives: They both frame it as an incentive problem, solved by either bigger sticks or bigger carrots.

The assumption in both cases is that people are doing some sort of cost-benefit calculation in their head before they pull a trigger. It’s treating gun violence as intentional, deliberate behavior.

Most gun violence is not so calculated. Take a more-or-less randomly selected example I pulled from the Chicago Tribune from June 2, 2024, in South Shore. It’s 3 p.m. on a Saturday. Two groups of kids are arguing with one another in the middle of a crowded street about whether someone in one of the groups had stolen a used bike from somebody in the other group. A 16-year-old pulls out a Glock semiautomatic handgun from his waistband, aims into the other group, pulls the trigger, and hits a 17-year-old kid in the chest. The victim is raced to the hospital, where he’s pronounced dead.

Economic incentives do a poor job of explaining this event. The value of a used bike on the South Side of Chicago is negligible—if you’re lucky, you might get $50 for it on Facebook. And even with the deficiencies in Chicago’s criminal justice system, shooting someone at 3 p.m. on a crowded street comes with at least a 50/50 chance of getting arrested and convicted. (In fact, the shooter is spending the rest of his natural life downstate with the Illinois Department of Corrections.) So a $50 bike traded off against a 50 percent chance of spending the rest of your life in prison—who in their right mind would rationally be willing to flip that coin?

This is what most gun violence in America actually looks like. It is not premeditated. It is not driven by economic motivations. It is the product of garden-variety arguments that end in tragedy because someone’s got a gun.

The national data back this up: The overwhelming majority of gun violence is driven by altercations. So what is happening? Is there any sort of predictable structure to these confrontations that we can use to figure out how to prevent them?

The trouble with System 1

If you’ve read the late Daniel Kahneman’s 2011 book Thinking Fast and Slow, you will know that your mind actually engages in two types of cognition, or two systems of thought, even though you may be aware of just one.

The type of cognition that we’re all conscious of, which Kahneman calls System 2, is a little voice in our head that’s deliberate and slow and gives us a chance to think rationally. It solves problems. It does cost-benefit calculations. And it’s very, very mentally effortful. It takes so much mental effort that our minds are designed to do as little of it as possible.

What we’re doing most of the time is relying on System 1: a series of automatic responses that we have developed for the routine, low-stakes situations that we see over and over again in our daily lives. That type of thinking is so low effort, and happens so quickly, we’re not even aware we’re doing it.

For example, when you see a newspaper headline, a CNN crawl, or other unfamiliar text, you don’t consciously ask yourself, Should I read that? You just read it. It’s automatic and involuntary. That is System 1 at work.

Here’s another example. Read the following word as quickly as you can:

BLUE

If you’re like most people, you experienced a little mental hiccup because of the mismatch between the appearance of the word and its meaning. This is called the Stroop effect, and it illustrates that the automatic response that normally works so well for you can lead to trouble when it’s overgeneralized. It works well usually, but not always.

So how does that relate to gun violence? One of the things we know about underresourced neighborhoods in cities across the country is that the adults are overwhelmed, leaving kids to learn quickly that they’re basically on their own to navigate daily life. (As it turns out, gun violence is disproportionately committed by and against young people—teens and young adults—late at night and on weekends.)

A young person in one of those neighborhoods might reasonably learn: If I’m walking to school and someone challenges me for my lunch money, handing it over would mark me as an easy victim. Tomorrow someone’s coming for my coat. The day after that, my phone.

So this becomes System 1 thinking for these kids: When I get challenged, I push back—not in proportion to the provocation, necessarily, but in a way that signals to everyone in the neighborhood that they’re not a good target.

This may ordinarily be an effective approach for kids in these neighborhoods. But it can lead to tragedy when someone’s got a gun.

Thinking slow

According to this perspective on the gun violence problem, it’s not caused by bad people or economic desperation. It’s normal people making bad decisions under difficult circumstances.

A few years ago, I was visiting the Juvenile Temporary Detention Center on the West Side of Chicago, and a staff leader there put it differently. He said he told 80 percent of the kids incarcerated there, “If I could give you back just 10 minutes of your lives, none of you would be here.”

How do we know if that notion is true? And if it is, what can we do about it? We may have gotten some answers from the first randomized experiment that we did at the Crime Lab. It was a study of a nonprofit in Chicago, Youth Guidance, which runs a group-counseling program called Becoming a Man. Boys in grades 7–12 get out of class once a week to talk and take part in various exercises as part of BAM.

One activity they do is the fist exercise. Student A and Student B are paired up, and Student A is given a rubber ball. Student B has 30 seconds to get the ball out of his partner’s hand. The only rule is there are no rules.

As it most often goes, for 30 seconds Student B basically tries everything in his physical power to pry the ball free from Student A. Then the counselor calls time, gives Student B the rubber ball, and Student A does the same thing to him.

The counselor debriefs with them afterward, asking about the strategies they used and inquiring about the method neither tried: asking for the ball. Youth Guidance has done this with something like 10,000 kids all over Chicago. They tell me that no more than a handful of kids asked.

Student B explains to the counselor, “Well, if I would’ve asked, he would’ve thought I was weak.” But when the counselor turns to Student A and asks, “What would you have done if your partner had asked for the ball?” Student A says he’d have been more than happy to hand it over. After all, it’s way better than getting beat up for 30 seconds.

This activity puts kids in a situation where they’re being challenged, elicits their automatic response, and in slowing things down afterward for reflection, shows them that in an out-of-the-ordinary situation, that normal response is maladaptive.

To study the effects of BAM, we worked with the Chicago Public Schools and Youth Guidance to enroll about 2,700 kids on the South and West Sides of Chicago. Most of the kids participated for 12–15 hours. Then we used Chicago Police Department data to look at what happened to them after program participation, and we found that their violent crime arrest rate was 45 percent lower than that of kids who weren’t enrolled in BAM.

To an economist like me, this makes no sense. Becoming a Man is not changing incentives. It is not miraculously erasing poverty or anything like that. Nonetheless, a few years later we redid the study, and this time we found a 50 percent reduction in the arrest rate for violent crimes.

The Crime Lab has also studied many similar programs, all of which look a little bit different and use different curricula but have one shared feature, which is helping people anticipate when their System 1 responses are going to get them in trouble. While the question of what happens at gigantic scale remains an open one, at demonstration scale we consistently see encouraging evidence of big reductions in violence. That’s inconsistent with conventional wisdom and only makes sense if there is something to this behavioral-economics explanation for gun violence.

Changing the policy debate

What are the implications? If the problem underlying gun violence is these 10-minute windows, let’s not keep arguing about whether we should be dangling bigger sticks or bigger carrots in those windows. Let’s focus instead on doing two types of things for which we have really good experimental evidence.

One is changing what kids bring with them into these 10-minute windows through programs such as BAM.

The other thing is changing the situations themselves in a particular way. The late author and activist Jane Jacobs in her 1961 book The Death and Life of Great American Cities, documented the huge variation in safety rates among equally disadvantaged neighborhoods and argued that “eyes on the street” were the key explanation for it.

Eyes on the street can be any sort of prosocial adult who’s willing to step in and defuse a situation. It could be a cop, a community violence intervention outreach worker, or simply a neighborhood adult walking to the corner store—which is why Jacobs argued for mixed-use residential development. A mountain of evidence suggests that the presence of these people is helpful in reducing violent crime.

This perspective on violence prevention helps us understand many of the key patterns with gun violence that conventional wisdom cannot. Why, for example, is gun violence concentrated among young people? That can’t be due to, for example, economic desperation, since the data suggest for most people economic concerns increase (not decrease) with age. But behavioral economics gives us an explanation: For most of us, as we get older, we get wiser. Similarly, why is gun violence disproportionately concentrated at night and on weekends? Because that’s when relatively more young people and relatively fewer prosocial adults are out in public places.

This perspective also helps us understand the mystery of Greater Grand Crossing versus South Shore. What are people bringing into the situation? There’s reason to think people in one neighborhood are leaning more on System 1 than their neighbors across Dorchester Avenue.

In the 2013 book Scarcity: Why Having Too Little Means So Much, MIT’s Sendhil Mullainathan and Princeton’s Eldar Shafir argue that living in situations of high stress depletes people’s mental bandwidth. If System 2 takes more effort, that leads people to rely more on System 1 in navigating daily interactions. When you look at the data, you can see there’s a lot more disorder, premature mortality, and other stressors in Greater Grand Crossing than South Shore.

What do we know about the situations themselves? Partly because South Shore is along Lake Michigan, it has developed many more mixed-use residential and commercial buildings than has Greater Grand Crossing. There are, consequently, more eyes on the street.

For too long, our policies for preventing gun violence have focused too narrowly on System 2. They’re not solving the problem. We need to be much more attentive to the fact that there’s another type of cognition that we engage in, one that is underappreciated in its importance to this issue. If we don’t, we have every reason to expect that America’s exceptionalism in this grim respect will remain unchallenged far into the future.

Jens Ludwig is the Edwin A. and Betty L. Bergman Distinguished Service Professor at the Harris School of Public Policy, the Pritzker Director of the University of Chicago Crime Lab, and author of the book Unforgiving Places: The Unexpected Origins of American Gun Violence, published in April by the University of Chicago Press. This essay is adapted from a talk he gave in February as part of the Think Better series hosted by Chicago Booth’s Roman Family Center for Decision Research.