Some on the left are pretty taken with this thing, which purports to show a link between gun ownership and gun laws on the one hand, and mass-shooting rates on the other, among U.S. states. Like most gun studies, it provides some interesting data and analysis but doesn’t really end any debates.
A few notes:
- They don’t actually have data on gun ownership. They’re estimating it with the percentage of suicides committed with guns, which correlates with actual gun ownership decently but far from perfectly, and may be picking up other factors as well. There have been numerous attempts in recent years to develop a better proxy.
- While the study includes data from 1998 to 2015 and deploys a blizzard of complicated statistical techniques, it’s essentially cross-sectional in nature — i.e., the authors’ claim is that “states with more permissive gun laws and greater gun ownership have higher rates of mass shootings,” rather than that, within states, trends in gun laws or gun ownership correlate with trends in shootings. The latter would be a far stronger finding because it would suggest that a given state, with its given attributes, has more or less violence depending on the gun laws/ownership at the time.
- The study controls for “median household income, percent high school graduation, percent female headed households, percent in poverty, percent unemployment, incarceration rate, and percent white.” As always with these studies, one wonders what the results would look like with a different set of controls.
- Just so everyone is clear on what the study is measuring, it includes all mass shootings in which four or more people die, including domestic and gang incidents, not just the public spree shootings we often associate with that term. (A separate analysis divides them into domestic and non-domestic.) Further, these data come from the Uniform Crime Reports, which police departments report into voluntarily, and which necessitates that they exclude Florida completely.
- It seems, er, slightly hard to believe that increasing gun ownership in a state by 10 percent would increase mass shootings by 35 percent. And that number is so imprecisely estimated that the “confidence interval” runs all the way from 13 percent to 63 percent.
For a good summary on previous work on the link between gun laws and mass shootings — specifically how inconclusive such work is — see the RAND Corporation.