The Corner

Politics & Policy

Throw Another Gun Study on the Pile

The Washington Post writes up a new one from a group of researchers centered around Boston University. It finds that “shall-issue” concealed-carry laws, which allow all citizens to carry guns provided they don’t have significant criminal records and meet any training requirements, “were significantly associated with 6.5% higher total homicide rates, 8.6% higher firearm homicide rates, and 10.6% higher handgun homicide rates, but were not significantly associated with long-gun or nonfirearm homicide.”

The study’s methods seem basically sound, and it differs in interesting ways from previous work, including using new statistical techniques and an additional year of data, as well as disaggregating homicides according to what type of gun they were committed with. Perhaps the biggest nit I can pick is that the overall and firearm-homicide results include justifiable homicides, but Michael Siegel, the study’s lead author, tells me those were removed for the analyses of handgun and long-gun homicides. (They rely on different data sets, one of which includes information on whether a homicide was justifiable and the other of which does not.) Some might also worry that it doesn’t account for homicide trends within states that had begun before right-to-carry laws were enacted, though that technical debate is a bit beyond my ken as a journalism major. And with these studies one can always argue about which variables the authors chose to “control” for and which were left out.

This is worth taking seriously, but it doesn’t push me away from my general skepticism of claims that concealed carry either increases or reduces crime. Here’s a brief explanation as to why.

I’ve been following this debate for more than a decade, during which time there’s been a fundamental shift in the literature. Back in the day, in general, some studies found that concealed carry reduced crime while others were unable to detect a difference. In 2005, the National Academies put out a report coming down with the agnostics, yet one member of the panel dissented, preferring the more-guns-less-crime view. But more recently, some concealed-carry states have seen increases in crime relative to other states, and the results in later work have thus shifted. Some studies now find an increase in violent crime, though until now, increases in murder have been a bit more elusive.

Here is an often-cited recent analysis, for example. Dig through the tables to see that, bizarrely, for the full period of 1977 to 2014, the results for overall violent crime are pretty consistent while the results for murder are usually statistically insignificant. (The results limited to 2000 to 2014 are if anything more chaotic and uncertain.) That study actually includes many more years of data than does the current one (which runs from 1991 to 2015).

The violent-crime trend might be real . . . but does that mean concealed carry becomes harmful over time, as more people acquire permits, or is something else going on today in the disproportionately red and rural states that enacted these laws years ago? The new study’s finding of a homicide increase concentrated among handgun homicides certainly buttresses the former possibility. But why is it able to find a result (increased homicide) that didn’t show up consistently in previous research using more data?

For the record, I generally respect Siegel’s work — and he’s proven willing to defy the consensus of his public-health field on vaping, another issue dear to us conservatarian types. But my jaded belief after spending far too much of my life reading these kinds of studies is that (A) there are many subjective decisions researchers have to make when they do this work and (B) the true effects of concealed carry are small enough that they will appear or disappear based on these subjective judgments, whether or not researchers are intentionally biased.


The Latest