In the wake of Monday’s Navy Yard shooting, there has been much lamentation that mass shootings are on the rise in America. “If you have been thinking that we live in an era that is more marked by this type of mass bloodshed than any era before,” remarked Rachel Maddow, “I am sad to tell you you are right. It did not used to be this way, but more and more, this is part of how we live.”
(MORE: The Troubled Mind of the Navy Yard Shooter)
The problem with this claim is that it isn’t true – or to be more charitable, it’s “true” in such a limited way as to be meaningless.
Maddow is defining “this type of mass bloodshed” as mass shootings in which 12 or more victims were killed. There have been 12 such shootings in the United States since 1949, and half of them have taken place in the last six years, which on its face sounds, as Maddow suggests, like a very ominous trend.
But anyone familiar with statistics should be made immediately suspicious by what statisticians refer to as the “cut point” for Maddow’s analysis. Why did she choose 12 victims? The answer is because it created the appearance of a statistically significant trend, where no such trend exists.
Suppose Maddow had defined “this type” of mass shooting as one in which at least 14 victims died. Using that definition, it turns out that the rate of this type of mass shooting in America was nearly twice as high in the 25 years between 1966 and 1991 as it has been in the 22 years since (there were four such shootings in the former period, and two in the latter).
Or we could use the FBI’s definition of a mass shooting: one in which at least four people, not including the perpetrator, are killed. This is a vastly larger category than the one Maddow employs: there were about 600 such incidents in the United States between 1980 and 2010. As James Alan Fox, a professor of criminology at Northeastern University points out, the rate of such mass shootings does not appear to be rising.
Speaking of statistics, here are a few others that ought to be relevant to policy makers:
*Homicide rates in the United States are lower now than they have been at almost any time in the last century, having fallen by more than 50% since 1991.
*Mass shootings, even using the very broad definition employed by the FBI, make up a tiny fraction of homicides – usually less than one percent.
*No one knows why homicide rates doubled between 1960 and 1980, or why they’ve declined just as sharply in the years since. Many theories have been suggested, including the aging of the population, high rates of incarceration, legalized abortion, and even declines in the percentage of lead in the environment (the latter, improbable-sounding, hypothesis has some surprisingly strong statistical support). But criminology is very far from an exact science, and these various theories remain little more than educated guesses.
Given all this, it’s very difficult to say what, if anything, can be done about either gun violence in general, or mass shootings in particular. It seems unlikely that the kind of weak gun control measures that are politically conceivable in America today, such as limits on certain sorts of weapons, and tighter background checks, will make much if any difference in regard to social problems that we, at bottom, do not really understand.
That is a frustrating and somewhat depressing conclusion, especially since even today’s far lower homicide rate is still vastly higher than that in almost all other wealthy industrialized nations. But it is better to admit to not knowing what to do than to do something for the sake of pretending otherwise.