With what feels like a new mass shooting occurring every few weeks, Americans all along the political spectrum are trying to tease out a solution to our homicide epidemic. Both sides—those who believe stricter gun control laws will decrease the incidence of gun-related homicide, and those who believe looser regulations will—have cited studies and created theories about how and why these homicides occur. And while partisans reference everything from mental health studies to gun safe locks in their respective arguments, one factor is left out of these discussions: the partisanship of the states in which these homicides occur.
When controlling for the number of guns owned per capita and the population density in each state, there is a significant correlation between the percentage of the state that voted for Mitt Romney in 2012, and their firearm-related homicides per capita in 2012 (as reported by the CDC). In fact, the electoral partisanship of the state is a stronger indicator of the frequency of homicide than gun ownership is.
However, this relationship does not exist when looking only at school shootings or mass shootings (defined by the FBI as shootings that result in four or more victims). Liberal and conservative states, when controlling for gun ownership and population density, are equally likely to experience mass or school shootings. The mass shootings that occur in conservative states, however, are more likely to occur with illegally-obtained weapons than those in liberal states.
It is important not to conflate correlation and causation with these results. When considering all homicides, every additional point Romney earned corresponded with another gun-related homicide per 400,000 residents (or per 1 million residents if the controls above are included). However, it would be foolish to conclude that the votes for Romney caused more people to be killed. One could make the liberal argument—people who vote for Romney support less gun control, which leads to more gun violence—just as easily as the conservative argument—people experience so much gun violence that they feel they need guns to defend themselves, so they support less gun control, and therefore support Romney.
Discerning which of these arguments is correct, or if some other explanation exists for this correlation, could go a long way in decreasing the incidence of homicide and making the United States a safer place to live.