Mass Shootings, Rampages, And Cascades of Failure

From the minigolf course at the 2009 FIGMENT participatory arts event, Governors Island, New York. From Wikimedia Commons

Airplane crashes are incredibly rare. The odds of dying in one are approximately one per 14 billion miles travelled. It is the safest form of travel and nothing else is even close. You are more likely to die driving to the airport than you are to die in an airplane crash.

But it didn’t get that way on its own. Airplane travel has been the safest form of travel for a while but it has only improved over time. The reason is twofold. First, airplanes are built with a lot of redundancy and pilots are exquisitely trained to deal with emergencies. Indeed, when an airplane crashes, it is rarely a result of one thing going wrong. It’s usually a cascade of multiple things going wrong: a series of mistakes, errors and mechanical failures that managed to weave its way through the layers of safety checks, redundancy and training. But every day, airplane crashes are prevented by the regulations, innovation and training we have poured into preventing them.

The second reason is that we refuse to accept that crashes are just going to happen. Whenever an airplane crashes, the NTSB does a thorough systematic investigation. They figure out precisely what went wrong and figure out if it’s a problem that’s likely to recur. If so, they make recommendations or issue mandates on what needs to change to prevent it from happening again. As a direct result, airplane travel, which was already safe, has become even safer over the decades.

————————————

Maybe I’m being too much of a scientist, trying to apply the rules of logic, reason and data to a situation those tools are unsuited for. And perhaps comparing airplane crashes — which are usually unintentional — to the horror that unfolded in Texas may seem like a stretch. But I think the comparison, while imperfect, is illuminating. Like airplane crashes, rampage shootings are rare and rampage shootings at schools are incredibly rare, a literal one-in-a-million chance. The majority of gun deaths are suicides and the vast majority of the remaining are the day-to-day individual shootings, mostly connected to the drug war. Like airplanes, schools are incredibly safe — a child is more likely to die on the drive to school than in school. Like airplane crashes, there is usually not any one thing that caused a mass shooting. They are usually the end result of multiple failures — someone got a gun who shouldn’t have, someone didn’t get help who needed it, a cop on the scene made a bad decision. Like airplane crashes, I’m convinced that we prevent far more than happen. Most would-be mass shooters are diverted somewhere along the way by getting psychological help, by being denied a gun, by finding the safety procedures too intimidating, etc. And, like plane crashes, their rarity does not diminish the visceral horror we feel when one happens.

However, unlike airplane crashes, rampage shootings are increasing. And unlike airplane crashes, we seem to have a political establishment that, in response to each bloodbath, either retreats to familiar ideological corners or throws up its hands and says, essentially, “mass shooters gonna mass shooter!”

A lot of people want to believe this paralysis is because of nefarious forces like the NRA. But I think it’s something much more mundane: the causes of mass shootings don’t fit neatly into ideological priors and there is no one clear-cut obvious solution to them. Oh, there are plenty of people who think they have the answers. But while those answers aren’t necessarily wrong, the solutions are, at best, only going to help on the margins. Democrats always want more gun control but there is little supporting evidence for policies beyond what we are already doing.¹ Securing schools and arming teachers is the Republicans’ favored issue. But it’s hard to imagine we can make our schools more secure without turning them into literal prisons. The most recent school shootings have taken place at schools that already had extensive security measures.

Mental health has also been raised. And with caveat that the mentally ill are more likely to be victims than perpetrators of crime, I think there is some merit to this. The US not only has one of the highest rates of mental illness in the developed world but, because we are the only developed country without universal healthcare, has a massive problem of untreated mental illness. But that may only help at the margins as well; the mass shooters with mental illness had already been diagnosed with them.

Moreover, there is clearly something more to this than guns, cops, psychologists and security theater. There is clearly something deeply diseased in the way some people are navigating through the world that we aren’t talking about; that we may not even know how to talk about. Seven years ago, Malcolm Gladwell wrote about the recent tendency of would-be shooters to congregate online, praise past shooters, fantasize about doing the same and, in the rare cases where they act on it, build on past shooters’ methods to up their body counts. It’s not clear, what, if anything we can do about that.² But it’s revealing of something much deeper, much darker and much more dangerous than our current approach is capable of dealing with.

I don’t necessarily have an answer here. But I think we need to change the way we look at these things: not as single-point failures with one solution but as systematic cascade failures with a set of solutions; not as isolated individual events, but a symptom of something wrong in society as a whole. Any time one of these shootings happens, it is thoroughly investigated. But mostly as an individual crime. We study them as isolated tragic incidents rather than a reflection of something much larger. The reports about the Sandy Hook massacre, for example, can tell you everything you want to know about the shooter’s methods, mental health and the failures around him. But it’s still fundamentally approached an isolated incident, not as a manifestation of a growing problem.

If we want to make these tragedies less common, we need to approach it the way the NTSB approaches crashes. Study each incident in detail and make a list of specific things that went wrong. Stop thinking of these as single-point failures (“This wouldn’t have happened if he didn’t have a gun!”) and start thinking of the cascade of failures that led to it. Don’t ignore an idea just because it wouldn’t have prevented this particular shooting. Look for patterns in the noise. Stop focusing only on the decisions the pilots made after the engine fell off the wing and think about why the engine fell off the wing in this first place.

This isn’t either-or. No single technology or piece of training prevents plane crashes; it’s the combination of things that has reduced them to such low levels. And no single policy is going to prevent all mass shootings; but a combination of approaches might make them far less common. A better background check system may not have prevented Uvalde, but it might have prevented Sutherland Springs. Red flag laws might not have prevented Virginia Tech, but might have stopped Parkland. Better mental health resources might not have stopped Las Vegas, but they might have stopped Sandy Hook. More attention to open threats might not have stopped Buffalo, but that might have stopped Columbine. If mass shootings, like airplane crashes, are a result of a cascade of failure, put more layers in to make the cascade less likely. A single strand of twine won’t hold up a tennis ball; but strand after strand wrapped together will hold up a piano.

The good news is that we have the beginnings of this in the NIJ database. It has identified some broad trends we’ve all seen — mental health issues, suicidality, threats. Some of the policy implications are clear. But we need to go deeper. We need to think harder.

There are limitations in this approach that the NTSB does not have — hydraulic pistons don’t have rights the way humans do. And the political reality is that some obvious things that can be done — like enhanced background checks or a national red flag system — will be hard to do. But I refuse to believe that we have to just accept that these things are going to happen. I refuse to believe there is nothing we can do except drag out the same old cliches (“Assault weapons! Video games!”) and then wait for the next one to happen. If you combine a bunch of little things that chew at the margins of this problem, eventually you might make some actual progress.

Calling for a “study” of a problem is the most inside-the-beltway cliche imaginable. But what other options are there? Clearly what we are doing now — proposing never-never gun control measures, shoving more cops into schools and pontificating on mental health — isn’t working. What else can we do but break each shooting down to it components, sift through the rubble of wrecked lives and hope that something emerges — or more likely a host of somethings — that is doable? What if we looked at gun control and mental health and social isolation and open threats? And then try to combine those approaches to make the cascade less likely?

Shouting at the darkness will not help you find your way out of it. You have to start walking. You have to feel your way around. You have to use what limited means you have. Maybe approaching this issue as a systematic failure rather than a series of isolated incidents won’t let us completely out of the darkness. But it might get us to a place with a little more light. And a little less heat. And maybe a little less death.

1. Research on gun control is, as a general rule, of very poor quality. RAND’s meta-study had to throw out the vast majority of it for obvious methodological flaws. But the NRC reached the same conclusion in their 2004 meta-study: that gun control only has small effects on gun violence.
2. Although there are some hints that denying them fame might help at the margins.