What one study says about the effects of violent rhetoric in politics

I had a conversation on Twitter Tuesday with several other science journalists and science bloggers about whether there's actually been much research done definitively linking violent political rhetoric to an increase in violent behavior. The connection makes common sense, but common sense and reality don't always line up. I've been curious what we actually KNOW about this, and how well we know it. Unfortunately, it didn't seem like there'd been much research specifically targeted at that question.

One thing that did turn up, via blogger Josh Rosenau, was a presentation put together by Nathan Kalmoe, a Ph.D. candidate in political science at the University of Michigan. The research doesn't focus on the rare, particularly heinous examples of violent rhetoric, but on far-more-common word choices, such as urging supporters to "fight".

Kalmoe ran an online survey, involving a diverse group of 412 adults from all across the United States. The survey respondents read two political ads—one that used neutral words like "work for", and the other that used metaphorically violent works, like "fight". Then they responded to statements about what was and wasn't acceptable behavior. Things like, "Some of the problems citizens have with government could be fixed with a few well-aimed bullets." Here's what he found:

Most people—across the board—are actively resistant to the influence of violent metaphor in political speech. At first glance, it looks like the vast majority of Americans—regardless of political leanings—are unlikely to turn violent rhetoric into violent action. But, there's a catch.

People who are classified as "trait aggressive"—those more likely to engage in aggressive social interaction, no matter the circumstances—DO respond to violent metaphor in political advertisements. What's completely safe for most of us can make a small minority feel more like acting out.

After Kalmoe controlled for those aggressive people, he saw other patterns. Among both the people who were naturally aggressive, and those who weren't, people under 40 were more likely to respond to violent rhetoric. Ditto with men, compared to women.

And he saw a small difference between political persuasions, too. Trait aggressive people—whether Republican or Democrat—responded about the same, with higher levels of support for physical violence. Among the trait passive, however, there was a small increase in support for violence among Democrats. The difference is small enough that you can't really say, "Democrats are more violent than Republicans!" At least, not without being really misleading. However, from Kalmoe's research, you most DEFINITELY can't say the opposite. Here's how Kalmoe puts it:

The results from this study provide some evidence rebutting partisan interpretations of political violence support, at least in the present form. Trait-aggressive Democrats and Republicans both expressed greater support for political violence following exposure to violent metaphors in political texts, though Democrats seem to show somewhat greater reactivity. Apparently, though Democrats and Republicans differ in many ways, support for political violence is not one of them.

It's worth noting that this is one study—a single data point. Just like it's a bad idea to draw conclusions based on common sense, it's also a bad idea to draw conclusions based on one study—and one based on surveys at that. This is interesting. But it will not be the final word. I want to make that clear. This is what one study says, not a definitive explanation of the way the world works. But you can't just dismiss it, either. Especially because it tells us that, whoever you're looking at, there's a small percentage of that group that could take violent words and turn them into violent action. That's definitely something that deserves more study, and it's something that should give all of us pause.