The Pump Handle is on a holiday break. The following, which was originally published on Jan. 29, is one of our favorite posts from 2016.

by Kim Krisberg

In the midst of another national debate over gun safety regulations, some argue that higher rates of gun ownership will protect people from dangerous strangers with deadly intentions. Physician and public health researcher Michael Siegel set out to study that argument. He ultimately found no relationship between gun ownership and stranger-related firearm homicides. But he did find that gun ownership levels translated into higher homicide risks for one group in particular — women.

“Our data refute the hypothesis that the more guns out there, the more of a deterrent there is to strangers committing firearm crimes,” Siegel, a professor in the Boston University School of Public Health, told me. “However, we found ownership rates were significantly correlated with non-stranger homicide rates. It’s the worst of both worlds because on the one hand, a higher prevalence of guns doesn’t seem to protect anyone…but it does increase the risk that non-strangers will kill each other with guns.”

To conduct the study, which was published in the journal Violence and Gender, Siegel and study co-author Emily Rothman, an associate public health professor at Boston University, examined state-specific homicide data from the Supplemental Homicide Reports of the FBI’s Uniform Crime Reports. Siegel noted that his study is the first to examine the relationship between state-level gun ownership and homicide rates in which the data is broken down by gender and by the nature of the relationship between the victim and perpetrator (i.e. stranger versus non-stranger). The study found that women living in states with higher rates of gun ownership faced a greater risk of being killed by someone they knew when compared to women in states with a lower rate of gun ownership.

But here’s the really interesting part: Among men, the difference in state-based firearm ownership rates explained about 1.5 percent of the variation in firearm-related homicide rates. For women, firearm ownership rates explained 40 percent of the variation in homicide rates with a firearm. In other words, while multiple factors must be considered in predicting the male homicide rate, just knowing a state’s gun ownership rate would allow a researcher to fairly accurately predict the female homicide rate.

“That finding was pretty striking for us,” Siegel said.

Simply put, more guns in a state means more women are killed. Siegel and Rothman write:

The magnitude of the association between firearm ownership and femicide committed by firearm was substantial. For each 10 percentage point increase in firearm ownership, our model predicts that the female firearm-related homicide rate increases by 10.2 percent. Thus, if the proportion of firearm ownership in Wyoming were 40 percent (the mean for all states) instead of 73 percent, our model predicts that its homicide rate would be approximately 33.8 percent lower (2.2 per 100,000 instead of 3.3 per 100,000).

In studying data that spanned from 1981 to 2013, researchers found that firearms were used in about 62 percent of homicides. More than 78 percent of all homicides involved an offender who was an intimate partner, family member or other acquaintance. In all, a non-stranger perpetrated more than three-quarters of the firearm-related homicides studied. But while men were more likely to be killed with a firearm, women were more likely to be killed with a gun by someone they knew.

The study found no evidence that greater availability of guns protects women from being killed. On the contrary, the study suggests that greater availability of guns translates into a higher risk that a woman will be killed by someone she knows.

“The data is just not consistent with the argument that having a high prevalence of firearms is protective against stranger homicide,” Siegel told me. “But for women, the level of gun ownership in a state is one of the primary factors in predicting her risk of being killed by a firearm.”

One of the obvious policy implications of Siegel and Rothman’s study is underscoring the importance of rules that restrict gun ownership among domestic violence offenders or those who’ve violated restraining orders. However, Siegel said the study also speaks to the critical importance of studying gun violence from a public health perspective — something the Centers for Disease Control and Prevention has been prohibited from doing for years. In 2013, President Obama signed an executive order allowing CDC to resume gun violence research and has recommended new funding for such research, though such funding proposals haven’t made it through Congress.

“This is a national problem — a societal problem — and everyone needs to be involved in tackling it,” Siegel said. “It really will take national resolve.”

To request a full copy of the new study, visit Violence and Gender.

Kim Krisberg is a freelance public health writer living in Austin, Texas, and has been writing about public health for nearly 15 years.