Political Violence: Causes and Cures
Regardless of your political views, you can't miss how much joy Democrats are feeling around Kamala Harris's run for president.
Outside this growing bubble of enthusiasm, the political atmosphere remains far less joyful. Within hours of President Biden's endorsement, more than 11% of tweets on X that referenced Vice President Harris included attacks on her race or gender. Not her political record. Not her political views. Simply her identity as a woman of color.
Should we blame Republicans for the toxicity of our political discourse? Although I'll be describing a few, narrowly-defined senses in which Republicans tend to be more politically hostile than Democrats, this kind of hostility is multi-partisan. It reflects partisanship without being overwhelmingly concentrated within any one party. (I'll be voting enthusiastically for Kamala in November, but I have to admit that her unofficial campaign slogan – "When we fight, we win" – is one of the more aggressive ones used by a presidential candidate in at least a century.)
At the same time, actual incidents of political violence have become almost exclusively a far-right phenomenon. In this newsletter, I'll touch on why. More broadly, I want to discuss some of the causes of political violence as well as potential remedies that have been studied. Scholars have grappled with these issues for millennia; modern statistics allow political scientists relatively granular insights related to time, place, and ideology.
As I noted last week, since 2016 the U.S. has experienced sharp increases in hate crimes, armed demonstrations, white supremacist activity, and threats against members of Congress, federal judges, and election workers. Support for political violence has also grown during this time period. Although it may be that only around 3 to 4% of Americans support overt, specific forms of political violence, that amounts to millions of people. The starting point for addressing problems like these is to understand what causes them.
A stereotype
Take a moment to imagine the person who sends a threatening email to a federal judge, calls for the assassination of a political leader on social media, or assaults a political opponent at a protest outside a state capital. What do you see?
To be honest, I see a poor, uneducated white guy with a hostile, slightly paranoid look in his eyes. He's a MAGA Republican who thinks the 2020 election was stolen and calls the January 6 riots heroic. He lives in a red state, owns guns, hates liberals, and dismisses Kamala Harris as a "DEI hire." I'll refer to this guy as "J.D."
My stereotype of J.D. is inaccurate in several ways, each illustrating something important about what actually does contribute to political violence.
1. Not all correlates are causes.
J.D., as I mentioned, owns guns. In fact, two large survey studies published this year show that support for political violence is greater among gun owners compared to non-owners.
One of these studies, led by Garen Wintemute at UC Davis, looked at personal willingness to engage in political violence. Who's most willing to injure or kill someone to achieve a political objective? The answer is gun owners, particularly if they've purchased guns recently and/or carry their weapons in public all or nearly all the time. Wintemute and colleagues found that between 10 and 20% of these individuals report being "somewhat", "very, or "completely" willing to injure or kill someone else for political purposes.
This is chilling, vitally important data. The folks most willing to engage in extreme political violence are the ones who are most capable of doing so, because they're armed in public. (As a recent Armed with Reason post observes, lax gun ownership laws enable political violence – you can't shoot an ex-president, for instance, if you don't carry a gun.)
Although public carry may be correlated with a willingness to be politically violent, correlates aren't necessarily causes. The reason that J.D. threatens politicians or attacks a liberal protestor isn't gun ownership per se. Rather, something else makes him more inclined to purchase a gun and to be more politically violent. That "something else"might lead people who don't own guns to be politically violent too. More on the topic of firearms later.
2. Not all causes are equally influential.
J.D. hates liberals, Democrats, people of color, the LGBTQ community, and anyone else he views as his political opponents, or at least supported by them. Political scientists would say that J.D. is affectively polarized.
"Affective polarization" refers to strong personal dislike of your political opponents. J.D. doesn't just disagree with liberals about political ideology; he despises them.
Rachel Kleinfeld, a Senior Fellow at the Carnegie Endowment for International Peace, and one of the leading scholars in this area, observes that affective polarization, by itself, doesn't account for political violence. Studies don't show simple correlations between the two. In addition, affective polarization in the U.S. has been increasing for decades (see American National Election Studies data below), but we only see overall spikes in political violence beginning in 2016.
This is not to say that affective polarization is unrelated to political violence. People who attack their political opponents don't generally feel warm and fuzzy toward them. But not everyone who hates their adversaries ends up attacking them. Affective polarization is merely one of many contributors to political violence, and so, as you'll see, describing J.D. as simply hating his political opponents leaves out a few things.
3. Multiple-cause explanations can be a rabbit hole.
J.D. is a MAGA Republican. Indeed, a study led by Garen Wintemute, published earlier this year, found that support for political violence is stronger among MAGA Republicans than others.
J.D. has a paranoid look in his eyes. Indeed, another study published this year showed that trait victimization – the ongoing feeling of being victimized by others – is associated with support for political violence.
J.D. supports the January 6 insurrection. Indeed, support for political violence can be a predictor of actual violent behavior; certainly it's a precursor.
You can see the pattern emerging here. Lots of variables, by themselves, contribute to support for political violence. But the more contributors we identify, the more we find ourselves in a rabbit hole, in the sense that we haven't explained how these variables relate to each other.
Is political violence like a dam breaking, where each contributor just sort of accumulates, like water from different sources, until a breaking point is reached?
That doesn't seem to be how it works. Our deepest insights about the causes of political violence come from more complex statistical models, and from the narratives that incorporate statistics and well as other kinds of evidence about the root of the problem.
Trait aggression
I described J.D. as having a hostile look in his eyes. Not surprisingly, studies show that trait aggression – i.e., being aggressive by nature – predicts support for political violence as well as actual violent behavior more strongly than other variables.
At the same time, simply being aggressive doesn't guarantee that a person will be violent, much less express support for it. In their influential 2022 book Radical American Partisanship, which I discussed a bit last week, Nathan Kalmoe and Lilliana Mason identified a number of contributors that interact with trait aggression. For instance, one of their most persuasive findings is that moral disengagement leads to support for political violence, but only among the most aggressive people. Let me break that down.
Kalmoe and Mason refer to "moral disengagement" as a more extreme version of affective polarization. It's not just a sort of personal distaste for your political opponents. It's the view that they're evil, threatening, and/or subhuman. This is illustrated by Donald Trump's references to immigrants crossing the Mexican border as "animals", or critics calling Trump a "monster".
What Kalmoe and Mason found is that moral disengagement is linked to support for political violence, but only among the most aggressive people. If you're aggressive to begin with, viewing your political opponents as evil and so forth gives you justification for attacking them. If you have poor self-control, and the leaders you admire are encouraging you to behave violently, the likelihood of your actually doing so increases.
Social context
Part of what's missing from my stereotype of J.D. is social context. Political violence is more likely in a society where citizens are politically polarized and public figures are promoting moral disengagement as well as normalizing violent behavior. This is a point stressed again and again by leading experts: Political violence has gone mainstream.
Here I want to quote at length a remarkable passage from a 2023 paper by Rachel Kleinfeld. In this passage she draws together much of the data on psychological as well as social contributors:
"In moments of low political polarization, aggressive people whose impulse control is low are likely to focus most of their rage on personal interactions: domestic violence, road rage, violent crime, or even a school or workplace shooting.... In other words, in a low-polarization environment, aggressive, impulsive individuals who don’t fear consequences may commit violence, but the targets will largely be apolitical..
However, as partisan leaders and media personalities demonize the other party, they can create feelings of rage among followers... Dehumanizing and denigrating rhetoric [makes] certain groups appear to be both threatening and, at the same time, vulnerable. Finally, the normalization of violence by political leaders, in particular, may provide a sense that acting violently against those groups will be permitted, may not be punished, or could be lauded and turn one into a hero..."
Thus, the individuals committing political violence...may find leaders who make violence seem normal or even laudable, build followers’ rage, and suggest a target for that anger in a political figure, government official, or minority scapegoat. Affective polarization within society gives them a way to cloak violent impulses in a greater cause that may even allow them to imagine themselves socially embraced or heroized (a potent draw for individuals who may feel ostracized because of their irritable, aggrieved personalities). And hints that political leaders accept such violence reduces concerns about the consequences of their violence, which might otherwise stop them from taking part in a rally or other event where their aggression could manifest."
In short, J.D.'s taste for political violence is not just the result of his inherent aggressiveness, but also because he views his political opponents as subhuman and threatening, a view supported by political leaders and others he admires who are also normalizing – if not rewarding – violence against a shared adversary.
This brings me back to the topic of guns. One of the influences that promotes gun ownership as well as the proclivity for political violence is the influence of the firearms industry. Last month, Everytown posted a sophisticated analysis of how the industry and its lobbyists wield their influence. The gist of their messaging to consumers is that crime is an existential threat that can only be countered by gun ownership. Gun-control advocates are thus treated as part of the problem and demonized, thereby fanning the flames of armed extremism. As the Everytown writers put it
"The firearms industry and its representatives in the gun lobby feed into the conspiratorial mindset of extremists, potentially leading to an escalating feedback loop in which the most radical extremists see any effort to regulate firearms as an existential threat deserving of a violent response."
In other words, part of the reason someone like J.D. owns guns is that he's convinced they're the only way to protect himself against threats that include criminals as well as anyone who seeks to limit his access to guns. Through a process of moral disengagement, gun-control advocates are lumped together with criminals as existential threats, thereby increasing the likelihood that J.D. would act out against someone simply because their political views are more liberal than his.
More on demographics
I described J.D. as poor, uneducated, and white. Some data links these characteristics to political violence, some does not. Most notably, Robert Pape at the University of Chicago looked at 654 people arrested for participation in the January 6, 2021 riot and found that the majority of them were educated professionals – doctors, lawyers, CEOs, etc. The fact that J.D. is poor and uneducated may not be directly relevant to his violent behavior.
I also described J.D. as living in a red state. That's definitely misguided. Pape, for example, showed that the January 6 insurrectionists who'd been arrested came from 44 states and a variety of counties, including a nearly even mix of those that Trump won in 2020 and those that Biden won.
However, J.D.'s whiteness does seem to matter. After accounting for the size of each county, Pape found that counties producing the most insurrectionists were the ones who'd lost the most white members of their population since 2010.
You may have already guessed where this is going. Pape argues that these data demonstrate the influence of "the great replacement" theory. Here's how he put it in a recent Slate interview:
"There is a right-wing conspiracy theory called the great replacement, which says that white people are being overtaken by minorities and that this is going to cause a loss of rights for white people. It used to be on the fringe. It’s been around a long time, but what’s special now is that that theory is embraced in full-throated fashion by major political leaders and also by major media figures. If you live in an area that’s losing white population, you can start yourself to connect the dots to the spinning that’s going around with these narratives."
So perhaps it wasn't so inaccurate to describe J.D. as white. He's not living in a red state though, but rather in a county or some other geographical region with a growing racial/ethnic minority population. (Consistent with Pape's findings is data from Kalmoe and Mason, and from a study published this June by James Piazza, suggesting links between support for political violence and white Republican resentment toward people of color.)
Finally, I described J.D. more specifically as a white male. The data doesn't show whether males are proportionally more inclined to political violence than females, in part because males are more prone to all forms of violence. On general principle, J.D. is more likely to be male. At the same time, Kalmoe and Mason and others have shown that under certain conditions, those who are more supportive of political violence (if not actually more violent) are white male Republicans who view women and people of color as both Democrats and threats to their position of social dominance.
Causes of political violence: Bottom line
As I noted last week, support for political violence in the U.S. is comparable across both major parties, though higher among Republicans in some specific, narrowly defined respects (e.g., in support for the January 6 insurrection). However, actual political violence, though rare, has become almost exclusively a far-right phenomenon in recent years. In this newsletter, you can see some of the reasons why: Incitement from right-wing political leaders and media figures. The influence of the firearms industry. White male fear of replacement.
So, here's a slightly more accurate depiction of J.D., our perpetrator of political violence:
J.D. is a person (probably a white male) with far-right views (sympathetic at least to MAGA Republicanism) who despises his political opponents (whom he would categorize as liberals and/or Democrats). He's highly aggressive by nature. He not only supports political violence (such as the January 6 riots), but also engages in it. Political violence is one way his innate aggressiveness is channeled against those who make him feel threatened (including racial/ethnic/gender minorities and their supporters) because the political, industry, and media voices he respects have been telling him that these are his political opponents, that they are indeed threats, and that they should be dealt with violently.
I don't write many newsletters on politics, but when I do, I lose a few readers (presumably Republicans), and occasionally I get an angry email (definitely from Republicans). What can I say? Four decades ago this newsletter would not have called out Republicans or the far-right for being the primary source of political violence. Historians and political scientists tell us that back then, political violence was ideologically-driven rather than linked to political party, and the far-left was just as guilty as the far-right. Times have changed. This is not to blame Republican ideology or Republicans in general. Rather, it's the more extreme versions of that ideology, and the more extreme Republican leaders and media figures that have normalized violence and encouraged the violent. I blame them more than I blame J.D.
Two remedies for political violence
1. Studies show that gathering people of different political ideologies and encouraging constructive dialogue can reduce affective polarization. This is a good thing – I support the practice – but there's no evidence that these effects are enduring or would translate into reductions in actual violent behavior. We want to find ways of talking with J.D., without expecting too much. The most effective strategy, according to contemporary political scientists as well as some social psychology classics, is to get people to not just talk, but to also recognize commonalities and shared interests that they can work together to attain. (Google "Robbers Cave Experiment", if you're not familiar with it already, and see how conflicts were resolved at the end of this 1954 study. It's a great read.)
2. Studies also show that the words of political leaders and media figures have an influence. As President-elect Biden put it, late in the day on January 6, 2021:
"The words of a president matter, no matter how good or bad that president. At their best, the words of a president can inspire. At their worst, they can incite."
I hope that quote lives on. The implication – you're probably sick of hearing this – is that Donald Trump is part of the problem (not to excuse those on both sides of the aisle whose rhetoric is violent too), and thus part of the solution is to find ways of incentivizing political leaders and other prominent citizens to tone down the rhetoric.
After the near-assassination of Trump on July 13, there was bipartisan condemnation of political violence, coupled with pleas from both parties for more civil discourse. That was a good start, but the sentiment lasted for exactly two days. Once the Republican National Convention began on July 15, the Democrats were once again "a clear and present danger" (Ron Johnson), the "Bidenvasion" of immigrants was destroying America (Kari Lake) and the unifying cry became "Fight, fight, fight!" (the crowd). As Rachel Kleinfeld points out, treating one's political opponents as a threat is likely to deepen polarization, in part by creating an arms race in which the other side keeps on responding in kind.
Regardless of your party affiliation, the best way to influence the rhetoric of political leaders is to through your vote. As for J.D., it would help if we could influence his voting behavior too, a process that ideally begins during the K-12 years and focuses on engaging him in coursework related to history, government, and media literacy. The goal should not be to make him stop voting Republican, but rather to encourage him to support more civil candidates who show respect for the democratic process and everyone it's meant to serve. When we fight confusion and intolerance, we may win.
Thanks for reading!