What can we do to reduce the epidemic of mass shootings in the U.S?
In this newsletter I'll talk about data on the most widely discussed strategies. Statistics, as you'll see, play an important role in what we've learned.
I have my political biases, but out of respect for the victims and their loved ones I'll keep this as nonpartisan as possible. My focus will be on legislative, psychological, logistical, and communication strategies for reducing mass shootings. At the end I'll describe what you can do to help address the problem.
Legislative strategies
What sorts of laws might reduce mass shootings?
Evidence is available from two kinds of studies.
1. Case studies.
On balance, case studies suggest that stricter gun control laws reduce mass shootings and other forms of gun violence. However, case study findings can be hard to interpret. Here are two examples that illustrate the point:
(a) Australia's 1996 National Firearms Agreement (NFA) is often cited as evidence for the effectiveness of gun control legislation, as gun violence in Australia declined following implementation of this law. For example, from 1979 through 1996, Australia experienced 13 mass shootings (five or more victims, excluding the shooter). From 1997 through 2016, the number dropped to zero.
One limitation of case studies is their messiness. Researchers can't set up each "case" with the precision of a lab study. Although that 13-to-zero stat is impressive, some declines in gun violence had already been occurring in Australia prior to 1996, and most states and territories were not fully compliant with the NFA after it was implemented. To put it in statistical terms, the NFA was effective, but the strength of the effect is hard to gauge.
There's also a key interpretive issue: What was it about the NFA that made it effective? This particular law included, among other things, bans on most civilians owning semi-automatic and automatic firearms, buyback guarantees for newly prohibited guns, stricter licensing requirements for private ownership, and restriction of gun sales to licensed dealers. Were all these changes necessary to impact mass shooting rates, or did just one or a few of them do the trick?
This is an important question, because if you're a gun control advocate, you can't just advocate for "gun control". You have to be more specific.
In sum, although most experts cite Australia as evidence that stricter gun control laws reduce mass shootings and other forms of gun violence (and I do agree with them), the fact that case study methods were used calls for some caution in which laws we choose to support.
(b) People who oppose tougher gun control laws often say that Chicago has some the strictest gun control laws in the U.S. and one of the highest rates of gun violence. One day after the Uvalde massacre, for example, Texas Governor Gregg Abbott cited Chicago several times as proof that tougher gun laws won't stop mass shootings.
A closer look at Chicago points to a different conclusion. Here are three key considerations:
–Most experts consider Chicago's gun laws strict but not among the strictest in the U.S. In conjunction with Illinois state law, Chicago ranks near the middle in this respect.
–Chicago's mass shooting statistics are frequently misrepresented. For example, CBS reported two weeks ago that"Chicago has the dubious distinction of having the most mass shootings in the U.S. since 2018". In fact, during this time period, Chicago had the most mass shooting victims (the 4th highest rate per capita). It did not have the the highest number of mass shootings per capita. Here too, it ranks near the middle among larger American cities.
–60% of guns used in Chicago-area crimes, and more than half of those used in mass shootings there, were obtained in other states, where gun control laws are much more permissive (e.g., Indiana). If guns obtained outside of Chicago were removed from the data, the city would have one of the lowest rates of mass shootings in the U.S. since 2018 (although one expects that at least some of those shooters would've obtained guns within city limits instead.)
In sum, the Chicago case does not illustrate that gun control laws fail to deter mass shootings.
2. Comparative studies.
Comparative studies can yield stronger generalizations than case studies do.
A prominent example is a 2021 study by Emma Fridel published in Justice Quarterly. Among other things, Fridel examined whether the state-level incidence of mass shootings could be predicted by gun ownership and concealed carry legislation. (Mass shootings were defined as four or more fatalities, excluding the shooter. This definition yields lower rates than often reported, because it excludes victims who survive.) Data from 1991 through 2016 were analyzed after controlling for key variables such as SES, crime rates, and percentages of males. Fridel found no relationship between the strictness of concealed carry laws and mass shootings. However, she did find that states with higher rates of civilian gun ownership had more frequent mass shootings. This relationship was remarkably consistent and is reproduced below:
On the x-axis of this graph, "Average" is the mean number of guns per household in a state. From "Lowest" to "Highest" we proceed from two standard deviations below the mean to two standard deviations above the mean.
As you can see, among states with the lowest numbers of guns per household, an average of 4.67 years passed between mass shootings. Among states with the highest numbers of guns per household, that figure drops to 1.5 years.
These data undermine the idea, articulated by the NRA and others, that a more heavily armed populace is a safer one, at least with respect to mass shootings.
However, we need to be careful about teasing out the legislative implications. There's more than one way to reduce the overall number of guns, and some strategies may have a greater impact on mass shootings than others. For example, stricter licensing requirements could, in theory, lead to the same reduction in numbers of civilian guns as restrictions on how many guns individual civilians can legally own, but only the former might impact mass shootings. All we can say for sure from studies like Fridel's is that fewer guns predict fewer mass shootings.
Gun rights advocates often claim that there's no data linking gun control legislation to mass shootings, or that the data are inconclusive. I found these statements puzzling at first, because most studies show that more restrictive gun control laws do predict fewer mass shootings, while a few studies show no relationship. If 8 studies show that X lowers cancer rates, while 3 studies show no relationship between X and cancer (and zero studies show that X increases cancer), you should probably conclude that X is beneficial. So, why do some people continue to say there's no clear evidence that stricter gun control laws reduce mass shootings?
Apart from the fact that personal opinions may be more informed by political ideology than by data, there's something else going on here – specifically, a series of reports and literature reviews disseminated by the Rand Corporation, a prominent nonpartisan organization. These widely-cited reports and reviews repeatedly argue that links drawn between gun control legislation and mass shootings are "inconclusive".
I took a close look at the Rand work and was startled to discover that for most topics of research, their standards are rigorous and reasonable, but for gun violence studies, the standards are absurdly high and, in some instances, literally impossible to attain.
Most disturbing to me is that, again and again, Rand authors raise hypothetical rather than actual problems with the studies they review. Here's an example drawn from a 2021 report. See if you can pick out the one key word in the following sentence (which will otherwise seem like gibberish unless you've taken a lot of stats classes):
"The distributional characteristics of mass shooting incidents are such that standard assumptions of asymptotic consistency and normality for the parameter estimates may not hold, thus threatening the validity of effect estimates and associated statistical inferences."
The key word here is "may". The authors are saying – with more jargon than actually needed – that because mass shootings are relatively uncommon, when you look at any particular sample of them, they may not be representative of what's truly going on. That's true on general principle – the smaller the sample, the more likely there's something quirky about it – but is it true of the actual studies? The Rand authors don't say. They could've e-mailed the researchers to ask about those distributional characteristics (or looked at the fine print of the studies, which in some cases tell us that the distributions were fine). Instead they just speculated.
I was so disturbed by Rand's approach that I reached out to Dr. David Hemenway, a professor of public health at Harvard and, for several decades, one of the leading gun violence researchers. I asked Dr. Hemenway whether he found Rand's standards for conclusive evidence to be inordinately high, at least for gun violence. He agreed. Here's part of his June 7 e-mail to me:
"Almost everything they [Rand researchers] conclude is: "the evidence is inconclusive" which means "we don't know for sure" which is unhelpful. Policies which you and I, after looking at the evidence, might both be willing to bet 10-1 that they reduce or increase the problem are all lumped together as "we don't know if they are good or bad...Unfortunately, too often their "evidence is inconclusive" means to the media and the public, that the policy probably had no effect."
I suggest keeping this in mind next time you hear someone say that data linking gun control laws to mass shootings is non-existent, or inconclusive.
In sum, here's what I would conclude from data on legislative strategies:
1. Stricter gun control laws predict reduced rates of mass shootings.
2. Claims that gun control laws have no impact on mass shootings, or that the research is inconclusive, are not supported by the actual evidence.
3. It's difficult to translate the research evidence into highly specific legislative recommendations.
My third conclusion is one that I only briefly touched on in this section. Consider this: Bans or increased restrictions are currently being discussed in Congress and elsewhere with respect to numbers and types of gun (e.g., semi-automatic), types of magazine (e.g., high-capacity), types of bullet (e.g., 9mm), and key accessories (e.g., bump stocks). Stricter legislation is also being discussed with respect to licensing, age limits, background checks, waiting periods, red flag obligations, restrictions on where guns are sold and who can sell them, etc. It's not feasible that comparative studies could measure the impact of each of these variables while controlling for the rest. Some archival studies look back in time and say, for example, that if 21% of mass shootings in a certain time and place were carried out with assault weapons, then laws banning assault weapons would've prevented 21% of those shootings. The logic here is shaky though, because the shooters might've switched to different weapons, or obtained assault weapons illegally.
In short, the research literature tells us that more restrictive gun control laws are likely to reduce mass shootings, but there's not necessarily guidance as to which versions of those laws would be best.
Psychological strategies
As I discussed last week, mass shooters are experiencing emotional crises at the time of their attack and, in some cases, have shown signs of mental illness. At least two characteristics of shooters illustrate how psychological strategies may help prevent mass shootings.
1. Many or most shooters exhibit warning signs. The signs may be specific (e.g., texts to friends indiciating plans for attack) or general (e.g., recent acts of violence). What constitutes "warning signs" is a subjective call, so experts don't agree on how many shooters exhibit them, but the estimated percentages are high (i.e., from 56% to over 90%). This tells us that one way to prevent mass shootings would be to increase public awareness of warning signs, and to encourage the sharing of information with mental health professionals. Extreme risk laws (so-called "red-flag laws") might also be helpful, as they empower courts to restrict access to guns when someone exhibits violent intent, and they've been shown to reduce the incidence of other kinds of gun violence.
2. Suicidal intent is especially prevalent among mass shooters – studies suggest that anywhere from about half to 78% of shooters express suicidal intent before and/or during their attacks Dr. James Hensley, a leading expert, refers to mass shootings as "angry suicides". This may be an oversimplification, but it does suggest that supporting highly aggressive people who are suicidal would help prevent mass shootings. To this end, Catherine Barber and colleagues at the Harvard Injury Control Research Center have been working with gun dealers, trainers, etc. on recognizing suicidality in clients and finding constructive ways of providing support rather than simply selling them firearms or training them on their use.
Clearly mental health support would help with many problems, including mass shootings, though it wouldn't be feasible to single out potential shooters. (There would be too many false positives, and we might not catch everyone who ultimately resorts to gun violence.) Fortunately, a rising tide lifts all boats. Mental health support to K-12 students is widely considered to be inadequate, for example, and thus strategies like increasing the number of counselors available to students in public schools are likely to have widespread benefits and, in the rare case, might even prevent a mass shooting. Another example where support for all K-12 students might deter extreme violence by individuals would be programs that focus on helping students cope with stress and anger – see, for example, KRS Edstrom's Student Stress and Anger Management Program.
Logistical strategies
I'm using "logistical strategies" as an umbrella term for changes involving facilities as well as people that are meant to increase public safety.
Some logistical strategies, like ensuring that every school has a working metal detector at all entry points, or providing supplemental training to police officers, may or may not diminish the severity of mass shootings, but they're unlikely to cause harm. The only downside is the time and money needed for implementation.
Some logistical changes, even if desirable, do have adverse side effects. For example, since Columbine, K-12 schools have increasingly restricted the number of entry points to school buildings. Although experts agree that this strategy hasn't reduced the incidence of school shootings (in part because because any entry point is vulnerable, and because not all shootings occur indoors), one could argue that with a single entry point, the interior of a school at least will be safer. However, educators report that restricted entry points cause traffic jams significant enough to require adjustments to student schedules (and, in theory, impair rapid egress during emergencies).
Finally, some logistical strategies might actually increase mass shootings. I'm thinking in particular of the proposal that we increase the number of armed citizens (either specific people, such as teachers, or the general public, so that everyone is prepared to respond to a threat). This idea has been proposed informally (as when Governor Abbott tweeted that Texans should buy more guns) as well as formally (via legislation or policies designed to increase legal access to guns).
I don't think the data support this proposal. America has the most guns in the world per capita, the highest per capita rates of gun violence among industrialized countries, and the 2nd highest rates of mass shootings among all countries. (See here for details.) International as well as regional comparisons show that gun violence is directly correlated with the number of guns per civilians. Moreover, studies show that over time, increases in the number of guns per capita predict increases in rates of gun violence per capita. I think these stats all provide direct evidence against the idea that arming more people will deter mass shootings.
To be clear, the statistics aren't saying that guns alone are the problem. Rather, what the they're saying is that simply increasing the number of people with guns won't help.
The usual rebuttal at this point is that, in retrospect, if such-and-such a person had only had a gun, a mass shooting could've been averted. I wouldn't disagree, but these are specific instances. They're not inconsistent with the overall relationship between numbers of guns and rates of mass shootings. Moreover, simply putting guns in peoples' hands is not, in itself, a solution. Here's how Harvard professor David Hemenway puts it:
"One huge problem is that so many people in the U.S. are armed who really aren’t well trained. Going to a gun range and shooting a few times does not make you well equipped to deal with violent situations where your adrenaline is going like crazy...and you have seconds to make the right decision... You could shoot the wrong person, or you could get in the way of the police or others who are well trained... Most people, unless they are with the armed services or a member of the police force, never encounter such violent scenarios... Do we really want continuously to train millions of people for an event that virtually almost none of them will ever encounter?... For mass shootings you would have to keep training over and over for the training to be at all effective."
In sum, many logistical strategies for increasing public safety are available, but the data suggest that arming more people is unlikely to reduce the incidence of mass shootings.
Communication strategies
Here my focus is on how mass shootings are reported in the news and social media, and, more broadly, how we talk to each other about guns and gun control.
1. Media reportage.
Copycat behavior contributes to at least some mass shootings (see here for discussion). As a result, news organizations have increasingly suppressed images, biographical details, and even the names of shooters, under the assumption that this will reduce copycatting. At the same time, some journalists argue that graphic images are needed, because they remind us of the carnage and could increase public calls for stricter gun control laws.
I think it's unclear whether suppression would help. Not all shooters articulate their motives, and among those who do, it's hard to say how much they're motivated by copying prior shooters. Also, the internet never forgets. Some manifestos, images, and videos from the past will always be accessible.
I also think it's unclear whether publicizing graphic images would help. On the one hand, decades of research in social psychology suggest that when you scare or anger people, and you provide them with simple remedies, they will take action. On the other hand, changing gun control laws is complicated and takes time; making people more upset than we already are isn't necessarily helpful. Meanwhile, publicizing the images may inspire future shooters.
2. Conversation.
You know already how much shouting there is around gun control, and why the issues are so emotionally charged. Having more civil conversations might prevent mass shootings by reducing gridlock around the creation of new laws and policies that would be effective.
Studies show that people become less polarized when they allow nuance into a discussion, consider the perspectives of those who disagree with them, and seek common ground. These are useful strategies. For example, if you're a gun control advocate, don't just say that guns are the problem, because they're clearly only part of the problem (i.e., they're necessary rather than sufficient conditions, as I mentioned in an earlier newsletter). At the same time, if you're a gun rights advocate, don't say that guns aren't the problem, because it's clear that they're part of it, and that better legislation would help. Meanwhile, try to remember that whichever side you're on, neither one of you wants another shooting.
Much of what we want to avoid is illustrated by a statement Texas Governor Greg Abbott made following the Uvalde massacre. (I'm not picking on the governor for partisan reasons; I'm picking on him because this is the clearest, most succinct example I could find of several ways we should not be discussing gun control, regardless of our political beliefs.)
"I hate to say this, but there are more people who are shot every weekend in Chicago than there are in schools in Texas... And we need to realize that people who think that, ‘Well, maybe we could just implement tougher gun laws, it’s going to solve it.’ Chicago and L.A. and New York disprove that thesis. And so if you're looking for a real solution, Chicago teaches that what you're talking about, it's not a real solution... The ability of an 18-year-old to buy a long gun has been in place in the state of Texas for more than 60 years...and over the course of that 60 years, we have not had episodes like this. And why is it that for the majority of those 60 years, we did not have school shootings? And why is it that we do now?"
From a nonpartisan, communication-focused perspective, here are four things wrong with what the governor said:
1. Meaningless comparisons.
The number of people shot in a particular city should not be compared to the number of schools in a particular state. The comparison is meaningless. It's not apples and oranges. It's apples and motorcycles.
2. Cherrypicked examples.
As I mentioned earlier in this newsletter, the Chicago data do not show that gun control laws are ineffective. (Likewise for data on L.A. and New York.) But even if we assumed that the governor was right about the data, these are individual cities. Individual cases, in other words. Surely you can find a city, somewhere in the U.S., with strict gun control laws and a lot of gun violence. If you look hard enough, you can find a city that illustrates anything. In short, if you're seeking a real solution for mass shootings, you need to consider all of the evidence, not cherrypick cases.
3. Irrelevant data.
Although "long guns" are deadlier and result in more casualties, handguns are by far the most commonly used weapon in mass shootings. For example, one analysis showed that 81% of guns used in mass shootings between 2009 and 2021 were handguns. Thus, whether or not Americans have access to long guns is mostly irrelevant to our problem with mass shootings. In fact, even if you only counted mass shootings in the U.S. that were carried out by means of handguns, we'd still rank #2 in the world in mass shootings per capita, and we'd still outrank all other industrialized nations. Meanwhile, a lot has happened here over the past 60 years, Our population has nearly doubled (186.5 million people in 1962 vs. about 335 million today), the number of guns has nearly quadrupled (about .35 per citizen in 1962 vs. about 1.2 per citizen today), and access to guns, including more efficient guns, has increased immeasurably. Mass shootings have been increasing for roughly two decades, with a spike in recent years. The governor's reference to the 60-year time period is irrelevant.
4. Lack of alternatives.
The governor did not answer his own questions (though elsewhere he implied that mental illness is the problem). He rejected calls for stricter gun control legislation, but offered no proposals for dealing with mass shootings and other forms of gun violence. Clearly we need to do something, and doing something begins with proposing something.
In sum, better communication might indirectly help reduce mass shootings by enabling multipartisan agreement on revised gun control laws, increased spending on mental health, and other potentially effective strategies.
Conclusion: What can we do to reduce mass shootings?
1. Acknowledge the need for multiple strategies. According to the available evidence, here's what may reduce mass shootings:
–fewer guns, and more restrictive gun control laws
–greater mental health support, including more public awareness of warning signs
–greater logistical preparedness for mass shootings
–better communication about guns and gun control
2. Learn more. Here some resources:
–For data on mass shootings, see here and here.
–For data on causes and remedies for mass shootings, see here, here, and here.
3. Reach out. Here are some options:
–Talk to people, especially those whose politics differ from your own. Forward them this newsletter if you wish.
–Check out this Everytown For Gun Safety web page, which describes strategies for outreach, volunteering, financial support, and communicating elected officials.
–Contact your senators by e-mail or phone (you can find their contact information here).
–Text ACT to 644-33 to join Everytown for Gun Safety, which will send you local resources (if you provide a zip code) and e-mail updates (if you provide an address).
Thank you for reading. Next week I'll be discussing love, which can surely help diminish gun violence but (fortunately) can't be legislated.
I did not see in the discussion the Constitutionality of these restrictions. I think theoretical is good but it needs to hold up in court and based on the past rulings, "Shall not be infringed" means exactly that. I did not see the discussion of removal of social media, restrictions to propaganda, restrictions of any media as those are rights covered under the 1st Amendment. Statistics alone cannot answer this as we have God given rights that are protected. Those rights have been and will continue to have great cost. You also forgot to add all the mass shootings by governments.