The Sandy Hook elementary school shooting was a tragic ending to a year that was filled with shocking gun crime. Before that, the Oregon mall shooting was the most recent reminder of how violent 2012 was, and before that was the Batman shooting in Aurora, Colorado. There have of course been more.
In the wake of these tragedies, there is an emotional response. National leaders express their sorrow and solidarity and local community members grieve. This is necessary and right.
Others feel called to respond to tragedy by judging and analyzing it with an eye toward preventing similar events in the future. This way of looking at violence inevitably requires statistics. A whole horde of them, and the sheer mass can become overwhelming, especially since the average person has almost no context with which to evaluate them. Statistics are often thought to be the final word on something, but usually they are just the entry point, the beginning of being able to think seriously about something.
When I learned about the latest shooting and went back to some of the previous acts of violence this year, I wanted to know what is unique about what we're going through, mainly because I feel like it is unique. Something seems dangerously out of sorts. Of course, the whole point of statistics is to make some phenomena stable enough to investigate across time, by many people, and with rigor. Statistics are supposed to be the antidote to intuition, feeling, and emotion.
But when I looked at some of the research and data concerning acts of violence, I found out that criminologists, sociologists, and police officers don't really have a grip on incidents like Sandy Hook yet. I couldn't even really get started in understanding school shootings, because I didn't have the right terminology or the right vocabulary, and it took me a while to just get to that point. My goal is to catapult you over that barrier and then give you a taste of how hard it is to think about these things.
First, what are the relevant statistics to bring to bear when thinking about 2012's gun violence? There is no statistic called "school shootings." One could look at homicides committed by young people, but of course that would include a bunch of convenience store robberies and other things, and one could look at homicides with young kids as victims, but this would be futile as well, because a bunch of public shootings beyond school shootings would be included.
A word that is thrown around casually by people is "spree." In it's informal use, it's clearly a correct label. Holmes and Lanza went on shooting sprees. But using "shooting spree" as a statistical category to understand if incidents like Sandy Hook are becoming a greater or smaller problem —-- to supplement our emotional reaction —-- would be grossly deceptive. The reason is that a shooting spree, for the purpose of the FBI's crime statistics, involves the death of more than one person in more than one location. Even this definition is contested. It seems that a spree was first introduced to distinguish a spree killer from a serial killer. The serial killer strings out his violence over time, but a spree killer, the idea is, somehow indulges in violence in bursts. The most authoritative source I could find on this was an encyclopedia of crime (p. 1567), which confirms that location is the main distinguishing aspect of a shooting spree.
School shootings might be shooting sprees, but they might not. Sandy Hook for example was not a shooting spree.
The most common statistic that is quoted in reference to school shootings is a "mass killing." This isn't really rigorously defined either, though the FBI does write (just in 2005, it seems) that mass murders involve a single location, and crucially, that they involve four or more casualties.
Generally, mass murder was described as a number of murders (four or more) occurring during the same incident, with no distinctive time period between the murders. These events typically involved a single location, where the killer murdered a number of victims in an ongoing incident (e.g. the 1984 San Ysidro McDonalds incident in San Diego, California; the 1991 Luby’s Restaurant massacre in Killeen, Texas; and the 2007 Virginia Tech murders in Blacksburg, Virginia).
Most things you see on the internet loosely follow this, but introduce their own (often hidden) qualifications and adjustments. For example, this well-known criminologist cites the four-person criterion, and argues that mass shootings have fallen since their high in the 1980s. Another well known criminologist cites that same criterion but then gives data that shows that the 2000s had 198 mass shooting incidents compared to 194 in the 1990s and only 163 in the 1980s. Something must be up because the first person says that there were only 24 incidents in the 2000s.
Some news outlets are up on this. Mother Jones conducted their own study, and wisely prefaced it with the claim, that ``"As we delved into the research, we realized that robust data on this subject was hard to come by, in part due to the lack of clear criteria.'' Other places are sloppier. My example is the Brady Campaign's website, but I'm not implying anything about their position. It is a fact though that their list of mass shootings in 2010 includes incidents in which less than 4 people died. I'm not saying that it doesn't matter when less than four people die, but only that it won't be comparable if we have a certain statistical category in mind. Their site though does show how common mass shootings, of one type, actually are.
Setting confusions aside about what is a "mass" murder, there is still the question of how useful even that statistic is. Most of the incidents that we classify as school shootings would count as mass shootings, but even mass shootings include incidents that happen outside of school, between adults, and with various motivations (including terrorism). Even a shooting that happens at a school might not count as a school shooting as we intuitively think of it. For instance, if terrorists tried to shoot school children, we would be rightly horrified, but such an incident would have immediate consequences for our diplomatic relations with other countries, with military policy, etc.
On the other hand, a school shooter, in the ordinary sense, is a certain type of psychological profile, and this may make us think that a body count or a location criterion will not allow us to think carefully about whether these incidents are becoming more common or more deadly. Thus, some researchers think that it is worth to introduce the idea of a "rampage killer" or a "pseudocommando" where such killings display a pattern of pre-meditation, heavy weaponry, and are caused by a history of social ostracism and abuse. James Holmes, "the Batman shooter," would be a classic pseudocommando perpetrator. I can't say whether the same is true in Sandy Hook.
It may seem like semantics, but in this case it is not. In a really good article, Robert and John Harris argue that social scientists are at a loss for how to think about incidents like Sandy Hook. Part of making progress on these shootings will depend on how we characterize the thing we want to study. That way, we can gather data about it and think about it more carefully.
This is not just a task for researcher. Part of coming to grips with school shootings involves being a careful consumer of information and helping us, as a society, refine our concepts about violent offenders.
Consider: there did not always use to be such a thing as a serial killer. This term had to be invented, refined, and then used in studies and ordinary thinking in order to produce our knowledge about serial killers — a body of knowledge that has influenced criminology, police work, and psychology. I think we are facing a similar moment with regard to "school shooters," and being clear about who they are and why they do what they do might be the beginning of a reevaluation of the way we think about schools, bullying, firearms, and any number of other topics.