Americans are once again looking for answers after the deaths of at least 19 children and two adults in last week’s mass shooting at an elementary school in Uvalde, Texas. Beyond questions around the police response and whether the massacre will lead to meaningful gun control, a big, fundamental concern looms: Why are guns such a problem in the United States, and what needs to happen for the situation to change?

Mass shootings are a distinctly American horror. What’s perhaps even more horrific is that, while each one is devastating, mass shootings cause only a small fraction of the gun deaths in America. The US has an unusually high rate of gun homicides among developed countries — for children 14 and under, almost eight times the rate of the next country in the ranking — and total gun deaths have only been increasing over recent years.

Legal restrictions on gun ownership, including mandatory waiting periods on handgun purchases and laws against children and youth carrying guns, could result in fewer deaths. But passing such legislation is a heavy political lift. In the absence of federal action, can anything move the needle on firearm deaths?

There is growing evidence that non-gun-control measures — including interventions to support at-risk youth and programs to improve access to mental health care — can and have been very effective, says Jennifer Doleac, associate professor of economics at Texas A&M University and the director of the Justice Tech Lab.

In a 2018 article in the Regulatory Review, Doleac described a number of these potential alternative solutions, such as summer job programs for at-risk teens, criminal justice reform, or changes to Medicaid. With meaningful and sweeping congressional action hardly a sure thing, should policymakers turn to these alternative ideas?

I spoke with Doleac on Zoom last week about her work, new evidence from the past few years, and what she sees as the best policy interventions to reduce gun fatalities in the US. A transcript of our conversation, edited for length and clarity, follows.

Miranda Dixon-Luinenburg

After the Texas shooting, I think we all want to know what the US can do to prevent future shootings. Gun control regulations are one of the first things that come to mind. But you’ve talked about how this is already such a politicized issue, with huge amounts of money and attention and resources being spent on both sides, and it can be unclear which regulations just reduce gun ownership or actually reduce shooting fatalities. Where do you see things now?

Jennifer Doleac

So the best research evidence we have suggests that gun control — restricting people’s ability to carry concealed weapons, requiring background checks — does reduce suicides and homicides at least some amount.

But economists or social scientists like myself look for natural experiments to try to understand what the effect of a policy is on some outcome. And gun laws are not great natural experiments. It’s hard to argue that they’re random, right? They’re the focus of so much political pressure and so much attention from the public that they’re not going to sneak through a legislature or community unnoticed.

So it’s very likely that changes in gun laws are correlated with changes in other things, like preferences, sentiment, or other local priorities that themselves could lead to changes in gun violence. Combined with how difficult meaningful gun reform is to pass, I think our attention and time and energy could be better spent on alternatives that could bring us the same goal but perhaps be easier to achieve.

Miranda Dixon-Luinenburg

Let’s move on to the other programs and policy interventions you’ve talked about. I’d like to hear whether your thoughts have evolved over the last four years.

You’d mentioned the summer job programs for teens, and there being some data that this reduces mortality from gun homicide and suicide, especially for young men. And also the cognitive behavioral therapy for at-risk young men seems like it falls into a similar category. I’d like to hear a bit about those.

Jennifer Doleac

When policymakers and practitioners ask me what they can do, what’s a reliable way to reduce crime in general and violent crime in particular, summer jobs are always the first thing I bring up.

There’s so much evidence from several studies now, and it really is the gold standard. We always love to see the randomized controlled trials, which are hard to run in a lot of these contexts, but we have them with summer jobs, and so we just have really great evidence that we should be pouring money into these programs and fully funding them in the cities where they’re not yet fully funded.

With cognitive behavioral therapy, I think our understanding of it has evolved, and it’s clear it’s not necessarily super easy to scale. As programs get bigger, you might have to hire people who are slightly less motivated [to do the work], and so figuring out how to scale some of these really effective programs, as is common in a lot of settings, is a challenge.

There’s a recent study that just came out from researchers at the University of Chicago Crime Lab looking at the effects of a program focused on youth at very high risk of violent crime. They’ve been doing this incredible randomized controlled trial that I believe combines cognitive behavioral therapy with jobs, I think, and just general outreach and violence interruption. They just released preliminary results on that, and it seems like things are moving in the right direction, but not quite statistically significant. It signals just what a hard problem this is.

Miranda Dixon-Luinenburg

There’s also the juvenile curfew study by Patrick Kline at UC Berkeley, where Washington, DC, instituted an evening curfew for youth under 17 with the goal of reducing violent crime, and voluntarily reported crime measures made it look like it was working. And then you published a study that featured a new way of more objectively recording gunfire using audio sensors [and] found the opposite effect — that 911 calls and reported crime went down but recorded gunfire went up.

Jennifer Doleac

We use data from Washington, DC, where the juvenile curfew changes time a couple of times a year, so we’re able to use that as a natural experiment. The basic intuition here is you’re encouraging a lot of kids to go home instead of hanging out on the streets with their friends, and you are giving the police a reason to go and check in with whatever kids are still out and maybe keep them from getting into trouble.

On the flip side, you’re removing lots of young people from the streets who might be very well-behaved but could be potential witnesses, and we know just having lots of people around deters crime. It also switches what the police are doing, where now they’re focused on enforcing this curfew rather than whatever they were doing instead, which might have been more useful.

And so you’ve got all this stuff happening and it’s unclear what the net effect would be, and it is a kind of policy that’s really tough to study the effects of because when you have fewer people around, crime is less likely to be noticed. In that paper, we had data from ShotSpotter, which measured gunshots. That allowed us a more objective measure of what crime was going on, and so we’re able to say that when the curfew was in effect, we actually saw gunfire go up rather than down, which is in line with it being counterproductive.

Miranda Dixon-Luinenburg

You also mentioned Medicaid coverage in early childhood, and Medicaid being required to cover mental health benefits, both being things where there’s data to show that access to Medicaid-covered mental health care reduces deaths.

Jennifer Doleac

The evidence on health care has definitely been booming — this has been a really hot research area. Expanding access to health care, including through Medicaid, reduces crime, particularly violent crime. The mechanisms are likely access to mental health care and substance abuse treatment. Since a lot of these programs do both, it’s hard to disentangle which one is really the driver, but they’re obviously related.

There’s an amazing paper by Elisa Jácome from a couple of years ago where she looks at what happens to kids when they’re kicked off of Medicaid at age 19, and she finds that with the kids that get kicked off, you can see it in the graph — suddenly they’re much more likely to be arrested and go to prison. The effect is entirely driven by kids that were getting mental health care and medication for various mental illnesses through Medicaid. So really simple changes, like increasing access for kids that were getting this kind of treatment, could be extremely effective.

There’s a neat study by researchers at Notre Dame, looking at a program that was a really light-touch intervention. Anyone whose screening when they came into jail scored high enough for history of mental illness had outreach workers call them after they were released and try to connect them with a local health care service, basically just make an appointment for them. Just with that, they saw recidivism drop substantially.

Miranda Dixon-Luinenburg

Are there any other programs or approaches to reducing gun deaths that have come up more recently that you think are promising?

Jennifer Doleac

There’s more and more evidence that air pollution increases violent crime in particular — not just changing your brain development when you’re a child, which is what we think of with lead exposure, but just day-to-day changes in pollution. There have been studies where you’ve got neighborhoods on either side of a highway, and if the wind is blowing the car exhaust fumes from the highway over into this neighborhood, then violent crime goes up over here. And the next day, if it blows the other way, the violent crime goes up over there.

Miranda Dixon-Luinenburg

So one possible concern is that some of these proposals might also be politically or financially costly. Health care is a pretty fraught and partisan issue right now. Which proposals do you think would get the most broad support and are realistic to implement soon?

Jennifer Doleac

For the health care question in particular, I think expanding Medicaid has become a big political issue, and is linked to Obamacare in unhelpful, political ways. But it feels like there’s an opportunity to talk about expanding health care in other ways, like the low-touch intervention that connected people with local health care services.

And even in the gun control conversation, a lot of times it’s the right that brings up that mental illness is the problem, right? So I feel like there must be some common ground there, if we take away the labels of what the programs are called.

Miranda Dixon-Luinenburg

Coming back to the Texas mass shooting. These major public incidents are the tip of the iceberg for total gun-related deaths, right? There are also huge numbers of other gun homicides associated with violent crime, and gun suicides, all of which are rising. How do you think about that in terms of policy interventions and which measures would be effective? Because for measuring effectiveness, mass shootings have a very small sample size.

Jennifer Doleac

That’s a great question. The bottom line is we don’t really know. There are far more mass shootings than we want there to be, obviously, but not enough to have good empirical evidence on what changes the number or severity or deaths of those sorts of shootings. And I think the official definition for various government agencies is any incident that had at least four victims in it, which it turns out is much more likely to be just normal “street crime.” I think we just don’t know yet if the kinds of interventions we’ve been talking about would be effective for these big, seemingly random public incidents.

I’ve become really interested in the extent to which crime experts and terrorism experts should be talking more. It feels like there’s enough overlap with what we would think of as terrorist incidents, even if they don’t think it fully fits in that bucket — and perhaps enough overlap with what makes a normal crime, even though we don’t think it quite fits in that bucket, either — that we should both be at the table. Clearly neither group has all the answers, and maybe it requires a bit more creative thinking and pulling from both sets of expertise.

We don’t have solutions yet. So it’s time to start thinking outside the box.