Some Republican voters supported the January 6th storming of the capitol, raising fears that the U.S. will continue to escalate violent extremism, moving everyday partisans toward endorsement of violence against their political opponents. Nathan Kalmoe and Lilliana Mason find that partisanship leads a sizeable minority of Americans to support violence or wish harm on the other party’s leaders and followers, especially after they lose elections. Drawing on survey experiments and history back to the American Civil War, they show the importance of messages in moving us over the brink or back from it.

Studies: Radical American Partisanship

Guests: Nathan Kalmoe, Louisiana State University; Lilliana Mason, University of Maryland


Matt Grossmann: When partisans endorse violence, this week on The Science of Politics. For the Niskanen Center, I’m Matt Grossmann. Some Republican voters supported the January 6th storming of the Capitol, raising fears that the US will continue to escalate violent extremism, moving everyday partisans toward endorsement of violence against their political opponents. On this week’s special edition, I talk with Nathan Kalmoe of Louisiana State University and Liliana Mason of the University of Maryland about their new book project, Radical American Partisanship. They find that partisanship leads a sizable minority of Americans to support violence or wish harm on the other party’s leaders and followers. Especially after they lose elections. We also draw on Kalmoe and Mason’s prior books, With Ballots and Bullets and Uncivil Agreement. Their research connects polarization to its violent form. Here’s our conversation.

So January 6th, we saw a dark chapter in American politics come to fruition all day on our television screens with insurrection at the Capitol. Liliana, what was it like to see your research on violent American extremism come to life?

Liliana Mason: Yeah. Well, personally I live in DC, so it was scary. We were watching it on the television and I had my kids with me. To see this really happening, to see real violence occurring was really frightening. Nathan and I had been talking and even posting things on Twitter about how this sort of position taking by these legislators, pretending that the election was stolen is really dangerous rhetoric. And that there’s real support for violence out there. And the leaders have a role to play in either stoking or pacifying that violence. One of my first feelings was just, I was so angry that these leaders had the opportunity to tamp this down so many times and they just intentionally did not do that. And this was the sort of obvious results.

Matt Grossmann: And Nathan, should we have been surprised by either the form that this took or the time that it took or should it have been seen as kind of an inevitable progression to this, and is there anything about what happened that’s changed your thinking?

Nathan Kalmoe: Well, I think Lily’s point is a good one. Nothing is inevitable, if you go back far enough and there were certainly opportunities for this to be headed off, particularly by Republican leaders. But the combination of risk factors that we saw going into November and December and into January, make the attack, at least in general terms, unsurprising, even though it was shocking to see it play out live on television. We know that from our research that a sizable number of Americans think that partisan violence is at least a little justified. And especially when they’re asked if their party loses the election, they’re especially likely to endorse violence, in that case. We know that Trump’s lies about the election gave them that justification and his cheerleading made the January 6th event a focal point. So something violent was likely to happen even before we saw the specific online messages by people who ended up as attackers, who are coordinating their actions in advance.

Matt Grossmann: So Lily, this project combines your previous interests and your own prior book that we’ve discussed before, Uncivil Agreement, talks about the importance of overlapping identities and increasing polarization. Is that part of the backstory of how we got to where we are? We of course saw a lot of white male conservative Republicans on display at the Capitol insurrection.

Liliana Mason: Yeah. And we also saw a lot of sort of overt white supremacy, signs, Confederate flags, these types of things. In our research, we found that, that there is some role of these either cross-cutting or aligned identities playing a role in how much people are morally disengaging from the other side. So kind of dehumanizing people on the other side. And in particular, we find that people who are higher … Republicans who are higher in education and high education for Republicans at this point is a cross-cutting identity, tend to be less dehumanizing of Democrats. And similar effects we see among Democrats, where if they’re evangelical Christians, then they’re less dehumanizing of Republican.

Cross-cutting identities do tend to actually improve or heal these really dangerous attitudes. One of the things that has been a really important cross-cutting identity or aligned identity is both racial resentment and hostile sexism for Republicans. When they’re very, very high on racial resentment and hostile sexism, they’re much, much more morally disengaged from Democrats. But when they’re very low on racial resentment or hostile sexism, they actually are sort of the least morally disengaged people that we can find. So not only do cross-cutting identities have a role or sorted identities have a role, but also sorted racial and gender based attitudes seem to be having a role in encouraging this type of behavior and attitudes.

Matt Grossmann: And Nathan, the book obviously builds on your recent book, With Ballots and Bullets about a civil war partisanship and its role in reconstruction. What can we learn from history for our politics today? Are we at a point that should be compared to civil war and its aftermath?

Nathan Kalmoe: I think there’s echoes, but we’re certainly not quite there. There’s plenty to worry about, The Civil War killed three quarters of a million Americans, and that’s not anywhere close to where we are, but it’s true that the Civil War and what came after have important insights for understanding what’s happening today. Just a little bit of the historical background, the parties back then were defined by their differences over slavery. But a lot of people forget that the rebellion started with rejection of a presidential election. When Southern Democrats refused to accept Lincoln’s election as the first Republican president. In the North, you had Northern Democrats who initially supported the war against their former Southern democratic allies, but then ultimately turned against it later in the war. And those divides between Republicans in the North and Northern Democrats had really important consequences for fighting and winning the war. Even though nobody fought explicitly for their party, ordinary partisans were more likely than Northern Democrats to enlist in the Union Army over the course of the war. And especially at times when the party leaders were most divided over the war, because partisan leaders who they trusted told them it was their patriotic duty and they interpreted what was happening through the lens that they were given by their party leaders.

Likewise, anti-war Democrats were discouraged from enlisting by leaders who told them this was a senseless war and a vindictive crusade against the South. And they also said, the war would end white supremacy including in the North, which they wanted to maintain. I think the first three points that come out of all of that are, you’ve got ordinary partisans who can get extraordinarily violent. And they do that in large numbers, when encouraged by their leaders for what seems like a noble cause and that race and other social identities are really important in fueling violence, especially when it’s aligned with the parties.

I think it’s important to recognize the violence in those terms, that it’s both racial and partisan. And that combination is serving both social and political goals at the same time. And especially in light of the fact that when dominant groups are feeling like their privileged positions are threatened, it provides a major motivation for violence when they can’t get what they want through ordinary politics. And that’s true then, and now, and around the world.

I just want to make clear there’s huge moral differences between the Republican cause then, and now. In on the one hand, violence upholding democracy in the Civil War case and undermining it in this case, but the mechanisms leading to violence psychologically and socially are very similar. One other key historical difference as a contrast, rather than a similarity is, that Civil War violence was so deadly because you had party controlled governments who are organizing that violence. It was basically industrialized killing in a war between governments and not just small militias and mobs. So I don’t think that’s a plausible outcome today. Lincoln basically got no votes in the states that rebelled, I think 27% was the worst that Biden did in any of the states. And it would have to involve Republican governors ordering national guard troops to fight the US military, which would cause some of the US military to dessert and support their partisan state governments.

I know I’ve said a lot already, but I just want to make one more historical point for comparison, I think this is most important. We’re focused on violence historically and in the present, but that’s not ultimately the main threat. Violence was just a mechanism among many to maintain control of government throughout the late 19th century and again, in the attack this year. So when they can’t win elections legitimately or through disenfranchisement even, and when they can’t secure control through corruption of the courts, violence is the mechanism that is still available. The rebellion against Lincoln’s election and the reconstruction violence after the Civil War, wasn’t the end goal. It was just that mechanism for whites to regain initial political control, so they could use government to firmly establish white supremacy and end multiracial democracy after the war.

So even though white Southerns lost the war for independence, afterwards they succeeded, they won in using violence and intimidation to get control of government and the legal apparatus at which point they entirely dismantled democracy. And basically re-established slavery under a different name. So the biggest threat today is that there’s many Republican leaders within government who are doing things to undermine government, especially at the state level. And they’re far more dangerous to American democracy than any mobs or militias and have done more damage already than anything that violence could accomplish short of a widespread civil war. And they’re still preparing to undermine democratic elections further.

Matt Grossmann: So we of course, don’t have a survey data to compare back all the way to that era, to our own. But the new project Lilly, draws a lot on recent surveys that you have conducted about partisan support for violence and some precursors to it. So why don’t you walk us through what that evidence looks like and what should we take from it and what it means to have someone say on a survey that they’re either endorsing violence, okay with harm to someone else or related indicators.

Liliana Mason: Yeah. So the first two things that we find to be very consistent predictors of pro violent attitudes. The first one is relatively obvious, which is trait aggression. So people just reporting in their daily lives that they have gotten in fights, they get angry easily. They’ve hit someone before. So just aggressive people tend to be more approving of political violence. But the other consistent predictor is a strong partisan identity. So people who identify with their party as a social identity. So when they talk about their party, they say we, rather than they. Those type of partisan identifiers are consistently more likely to say that it’s at least a little bit justified to participate in partisan violence. And those really are our most consistent predictors.

There are things that sometimes are there, and sometimes they’re not there, in terms of predicting these things, like I said before. Hostile sexism and hostile racial resentment in the presence of a Republican identity is a strong predictor and we also have seen that the strongest partisans are the most responsive to party leader rhetoric, which we’ll talk about a little bit later. There are things that we can use to predict violence and we can use to potentially change violent attitudes. Now the issue of expressive responding versus actual violence is an important one.

Matt Grossmann: Before you get to that, why don’t you just describe what these outcomes look like a little bit, in terms of people’s endorsement of violence.

Liliana Mason: Right. The questions that we use to assess endorsement of violence are generally four different questions. So the first is, how acceptable is it to threaten leaders of the other party? The second is, how acceptable is it to threaten and harass a person, a regular partisan from the other party on the internet in a way that makes them feel frightened? The third is, to what extent is violence justified to achieve political goals, violence by your party? And then the fourth one until recently was, what if your party loses the 2020 election, how much do you think violence would be justified then?

So in general, those last two questions are really getting directly at violence and the first two are more about sort of a violent environment or a sort of threatening environment. And what we generally saw was that we would see 10 to 15% of people agreeing that it would be justified for the people in their party to engage in violence for political ends. But that number often doubled when we said, “What if your party lost the 2020 election? Would it be acceptable then?” So making people imagine an electoral loss consistently increased their acceptance of violence for political ends. And so, one way to interpret what we saw on January 6th is this is the side that lost. And so we should expect more anger and more acceptance of violent behavior from people who we already could have predicted once they lose, they are more accepting of political violence.

Matt Grossmann: And Nathan, do you find that these indicators are actually pretty similar across the two political parties, but of course we are witnessing a big upsurge in right-wing violence and warnings that are more associated with that. So that could be just attributable to Trump. It could be something else about the long running differences. So I guess, how do you interpret your findings in light of the violence that we see in the real world and what signs should we look for that there was left violence forthcoming?

Nathan Kalmoe: Yeah. So we generally do see more asymmetries… or more symmetry than asymmetry between Democrats and Republicans in the range of extreme attitudes that we ask about and violent attitudes are obviously a risk factor or a precondition for engaging in violent behavior, but there are several important steps from those attitudes to actually taking actions that are clearly important and that’s what produces the asymmetries and action that we see. Leaders are the obvious explanation for a major difference there in terms of action versus attitudes in fanning those flames or discouraging them. The role of leaders is clear in US history. It’s clear around the world and it’s clear in the post-election violence that we just saw. Party leaders are indicating their approval not just for a general kind of a hostility, but for their followers to take action, either implicit or explicit approval and endorsement, and they provide opportunities and coordination to act. And those are just as important helping to focus people’s attention and their actions in specific directions and not just a diffuse sense of hostility.

In terms of left violence, the biggest danger for violence from Democrats and aligned groups would be probably a reaction against violence by Republicans or by right-wing groups, particularly if Democrats feel that police and other security forces aren’t keeping them safe. Obviously, there’s a lot of distrust of law enforcement among many Democrats already. There’s a Bright Line Watch survey this fall that found that when partisan violence is contextualized as a response to violence by the other side, then the support for that violence actually triples from the baseline level. So we repeated this question in our post-election survey as well. You see between like 40 and 50% of partisans say that violence is okay if it’s a response to violence by the other parties, so that’s a real danger of cycling violence. More broadly, we don’t think that Democrats are immune to mobilization into violence. We don’t think there’s anything inherently different about the psychology of Democrats that makes them not susceptible to this. It’s just that democratic leaders haven’t radicalized them in the same way and certainly haven’t mobilized them into violence in the same way. And Democrats haven’t insulated their core supporters from cross pressures of information and social influence in the same ways that Republicans have.

Matt Grossmann: Absolutely. I’m not sure I got the full opportunity to talk about the expressive responding issue, so why don’t we go back to that a little bit? Should we see these say regular Trump supporters who showed up at a rally and then made their way to the Capitol, but maybe we’re not intending to break in as similar to your survey respondents who are kind of halfway down the scale, but not yet in endorsing violence? How big of a connection is there between what would happen in the real world and what we see on surveys?

Liliana Mason: We’ve been worried about the idea of expressive responding for a while. And a lot of our survey experiments were designed to try to figure out, would people be willing to harm other people in their actual behavior? Now on January 6th watching this attack, I thought about that problem of expressive responding because obviously, this is far beyond expressive responding. This is an actual violent activity. And I think what we’ve been arguing for most of this time is there are people who are just saying that they approve of violence because it feels good to say that, but not everyone is doing that. And in fact, what we saw in the Capitol attacks was that it doesn’t take very many people to create a huge amount of chaos and real danger, not just to lives and property, but to the processes of democracy itself.

So often, Nathan and I will talk about how it only a couple of people, it only takes a few people to actually act out these violent attitudes. And I think now we’ve really seen the perfect example of that. And it’s possible that there were people that were at that rally as you said, that might not have entirely approved of using violence for political means. But once they saw what was happening, they either got carried away and got very excited because they were together in a group and group psychology works very differently than someone sitting at their computer. But in general, I think it’s important for us to remember that not everyone who says they approve of violence will engage in violence, but it just doesn’t take that many people to create dangerous environments.

Matt Grossmann: And Nathan, maybe we should remind people what some of these precursors that you look at are, so these things like shot in Freud. I don’t know if I’m pronouncing that right. Moral disengagement, dehumanization, these kinds of indicators, what do they look like? And how much should we be concerned about them alone, even if people are not directly endorsing violence?

Nathan Kalmoe: Yeah. So the extreme views that we are seeing go well beyond just support for partisan violence, where as we said it tends to be between like 10 and 20%. Although, when you start to add conditions like losing the election or the other side being violent, that number jumps pretty dramatically. When we ask about moral disengagement items, these are… Let me just explain the concept. The psychologists basically see these as rationalizations for violence and other kinds of harm against other people, so they include vilification, a sense of righteousness for your own group, minimizing the harms that are caused through the action and diffusing blame, those kinds of things. So we asked a number of questions that get at moral disengagement as again, precursors for actually taking action and supporting action that is violent.

One of these items is, do you see the other side not just as wrong in policy, but as a real threat to the United States and its people? And somewhere around 60% of our partisan respondents see the other party not just as wrong, but as a threat to the country. We asked, “Do you see them not just as wrong, but as evil?” And about 40% say that they see the other side as evil. And we asked, “Do you see them as lacking the traits to that are essential to being a human,” basically, are they less than human and about 20% of American partisans endorse that idea. An important set of trends here that we see is that some of the most extreme views have lower supports and I guess from a democracy perspective, that’s really encouraging. And from a methological point, if I can jump in the weeds for a second, it’s encouraging to see that partisans are responding in a modulated way that indicates that they’re not just saying all the worst things that they can imagine just because it feels good. They’re considering the specifics of what we’re asking them and they’re saying, “Yeah, that sounds good,” or, “No, that goes too far for me.”

Matt Grossmann: So Lily, we’ve all been more engaged in insights from the comparative politics literature lately in American politics. And two that jumped out at me, one is just whether there is a part of this, that’s an inevitable response or… we don’t like the word inevitable, but that should be expected as a response to diversification, to secularization, to liberalizing social trends, to the decline of white Christian America as it’s been posed in one case that this is the kind of upsurge that we should expect to follow from that. And the other is that it’s sort of important whether people see a route to victory through traditional political institutions and that if they don’t, if this is the last election or they believe that it is, then they’re more likely to go down to those routes. So how much do those apply to the American case?

Liliana Mason: Yeah, I think that’s just a really great insight. One of the ways that I tend to organize my thinking about the current kind of partisan divide is that it’s really centrally about the traditional social hierarchy in which white men are at the top and they get to have more rights and more protections than anyone else. With the Republican party essentially saying both that the traditional social hierarchy is good and that it is no longer present, that it has been destroyed. And the Democratic party saying that the traditional social hierarchy is not fair. It’s not egalitarian and it still exists and there’s still work to be done to dismantle it. And increasingly questions about this, about systemic racism for instance are increasingly dividing white Democrats and white Republicans where white Republicans have basically had the same attitudes about these issues since 2009, whereas white Democrats have become far, far, far more progressive on their belief that systemic racism exists and is a problem that still needs to be dealt with.

So really this is I think a central part of the divide that we’re seeing and the source of so much of this violence and anger is that we’re really fighting this battle over the traditional social hierarchy, which is a deeply embedded part of our society. And it’s true that the group that has moved to the most is white Democrats, right? They have become increasingly progressive, whereas white Republicans have kind of stayed where they were. But what that means is that that is sort of a threat to white Republicans who feel that this group of people, this group of white people from their own racial group is leaving them behind, is turning their backs on them.

And that’s the type of debate where there’s very little place for compromise when we’re in a debate like that because it becomes very moralized for both people who want to see social progress and people who don’t want to see any more social progress. So in order to depolarize for instance, if we need to find a place of compromise between Democrats and Republicans, it’s not normatively ideal for Democrats to become less committed to egalitarianism and multicultural democracy, for instance. So that makes a very difficult argument that’s really going on. And it’s a very deep argument and I think it does help explain why things are so fraught right now.

And then in terms of the idea of institutional power, that’s another huge part of this because generally, Republican voters due to systemic and institutional rules are overly represented, right? People in small and… sorry, people in rural areas and in particularly rural states are disproportionately represented in the Senate. The electoral college provides a lot more power to Republican, well, mostly rural, but that tends to be a correlated with Republican votes. And so there is this sense that if we make the system more egalitarian, if we have a system in which you actually need 50% of Americans to vote to get 50% of the seats in the Senate, which currently you need 17% of Americans to get 51 seats in the Senate.

If we wanted to create sort of more egalitarian democracy, that would be a direct threat to Republican’s ability to remain in power because it is these, they’re sort of old institutional rules that have allowed them to remain in power for as long as they have. They don’t win the popular vote for the president in the last few elections very often. So that does create an incentive, I think, for a lot of Republicans and even Republican legislators and officials to maintain the status quo again, really pushing Republicans to keep everything the way it was while Democrats are trying to push the nation forward in a more egalitarian direction, which is sort of an irreconcilable debate. And I think that’s really part of this tension.

Matt Grossmann: It’s an odd behavior now for Republicans though given what you’ve said. I mean, I have my colleagues that study terrorism and election violence that say, well, you need them to believe that there’s a path to winning the next election and they can work through current channels, but as you’ve said, not only is there a path to winning, there’s no loss of competitiveness for the Republican party at the moment. But on the other hand, I guess we would not want to tell people that they’re going to have disproportionate power forever. So I guess is there any way to get people off thinking that this is the last and only election that they can win without necessarily saying that they’re going to retain the same level of advantages that they have always had?

Liliana Mason: So I’m not actually convinced that the leadership right now wants to come they’re voters down. It seems actually that there’s two things happening. One is that a lot of the leaders are frightened of their own voters. People in the house, there were a number of apparently representatives in the house, Republicans, who would have voted for impeachment, except that they were afraid for the lives of their families. They’re actually physically afraid for their lives from threats from their own voters. And the second thing is that the more they rile up these voters and tell them that if the Democrats win, everything is over, democracy is over.

And we see this all the time on online rhetoric from people on the far right. This is it. They’re going to destroy our entire country. The more they rile people up with this, the more they get turnout. The more they have people paying attention to what the leaders are saying, being engaged and potentially becoming violent, which is just a side effect of them being engaged. And so a lot of these leaders, I think, are both afraid to tell people to calm down, that it’s okay right now, you’ll have another election. And there are slightly motivated to tell people to stay angry because without that, they don’t have the same type of enthusiasm and support.

Matt Grossmann: So Nathan, on the one hand, the insurrectionists at the Capitol were kind of a classic example of the story that you’re telling. They were partisans motivated by having lost an election, and mixing all of the identities that we’ve been talking about. But there were a whole bunch of other things kind of mixed in. There’s been a lot of focus on the QAnon conspiracy theorists for example, but there was older militias and everything else you can think of that coalesced there. So one way of seeing it is that it wasn’t really partisan. It was this mix of people that were just bought together by Trump himself. Is there any hope that the violence will decline now that Trump has gone and that this was a little bit more personalistic and less partisan, or was this classic partisanship that brought all these people together?

Nathan Kalmoe: Well, it wasn’t a surprise that people who did show up and behaved in the most extreme way were the ones who were the most extreme within the party coalition. Each of the groups that you mentioned has some idiosyncratic ideas, but that’s normal in a broad partisan coalition within a two party system. We shouldn’t understate their broad agreements that they share with each other and with the rest of the party, particularly around maintaining the power of dominant racial and religious groups. All of those groups see both Trump and the Republican Party as good if not necessarily ideal vehicles for achieving their goals as do Republican activists and voters more generally.

With Trump as an ex-president, that might help reduce violence somewhat, although I think the danger remains high. It might help with reducing violence because presidents are the top party leader. Trump seems unlikely to maintain his singular control as a focal point of the party now that he isn’t president. That at least slight diffusion of leadership might help to reduce some of the coordination, at least, among extremist groups within the Republican coalition. I would say that Trump’s biggest threat to the United States was during the transition period not just because he had control over the security state at that time, but also because he was his party’s empowered focal point. And I think that really mattered.

I think the concern now would be less about widespread violent uprising and more about targeted violence against individual leaders, including assassinations. But there’s new leaders who are emerging already among Republicans who could serve as a coordinating and focal role for driving potential violence and new events, including violent events, could provide focal points that cause violence to spiral, as we were talking about earlier. And overall, I think that as Lilly has mentioned, the biggest threats remain. You still have Republican leaders either because they’re true believers or because they see strategic advantages or because they’re afraid of their own voters who refuse to accept essential aspects of democracy and of the 2020 election result and in many cases are continuing their work to undermine that. And so, the overall threat environment related to our political context really isn’t changing and in some ways it’s maybe getting more concerning.

Liliana Mason: I would also just add to that that just because the insurrectionists were targeting Republicans doesn’t mean that there wasn’t partisanship behind their anger. One of the things that Nathan and I have found is that people are very supportive of violence against apostates within their own party. So if someone within the party kind of betrays the party, people support violence against that candidate almost as much as they support violence against out-group candidates or out-group elected officials. So, just because people are targeting individuals like the vice-president within their own party doesn’t actually mean that there’s no partisanship that’s driving it. In fact, partisanship might make that type of violence worse because they see the vice-president as betraying the interest of the party.

Matt Grossmann: Lilly, your book is also looking at some things that might either make these attitudes better or worse. So what can we say about what messages matter and how much who says them matter.

Liliana Mason: This is one of the things that I think is one of the more optimistic findings that we write about in the book, which is that leader messages really, really matter. And it’s not just leaders, even regular people on the internet writing things that are antiviolence or sort of superordinate identity, which means we’re all Americans. We might be in these two different groups, but we’re also all within this one group of Americans. Those types of messages do tend to reduce acceptance of violence. And particularly when we see them from leaders… We tested messages from both Trump and Biden. Biden’s messages worked better than Trump’s messages, and we think that’s partially because this was not a consistent narrative coming from Trump and from Biden. The idea of anti-violence and unity has been a consistent message.

But when we experimentally had people read antiviolence messages from Biden, what we found was that the effect of partisanship. So we know that strong partisans tend to be the most approving of violence, that affect actually reversed so that the strongest partisans were now the least approving of violence, because effectively what’s happening is that the strongest partisans are listening to this leader. They are more likely to listen to their leader and essentially obey because they’re so strongly attached to the party. So the pacifying language from people like Joe Biden really can reverse a lot of these trends among the people who are most at risk of supporting violence.

The issue is if we have consistent, as I think we saw at the January 6th rally, if we have consistent pro-violent mobilization and rhetoric, leaders can have the opposite effect. But we do see that there is a real potential for leaders if they consistently remind people that it is not acceptable to behave in violent ways. We are all Americans and we’re together as American citizens and patriots. That type of language works, and it works among the people who are the most willing to accept violence in general.

Matt Grossmann: Nathan, how much do online and offline social networks play in this process and are we stuck at just rooting for kind of the very last step in your process, the actual resort to endorsement of violence, that’s sort of where social networks are kind of drawing their line, or can we actually go back further in the process all the way to reducing people’s moral disengagement?

Nathan Kalmoe: Our experiments are optimistic in that regard too. We tested inflaming messages among ordinary people that were in our experiments that we mocked up as Twitter posts that were hostile in various ways towards opposing partisans. And we also tested the pacifying or unifying kinds of rhetoric as well. And we found maybe thankfully that the ordinary people were not as likely to move the dial on the inflaming kinds of rhetoric, but they were just as effective as Lilly said at reducing the hostility not just in terms of violent partisan attitudes, but also these broader attitudes about moral disengagement in relation to the other party.

This is consistent with interpersonal research more generally that particularly people in our immediate social networks, our families, our friends, maybe our coworkers as well, can be really effective at persuading people in politics. And it seems like that includes with extreme attitudes. And so that extends as well into other studies that find social influence more generally, for instance, prejudice reduction, or at least the reduction of racist statements on social media when confronted by somebody from the in-group. So broadly speaking, our results show that these social media interactions can play an important role, even when they’re coming from strangers, in reducing some of the most extreme forms of partisan hostility.

Matt Grossmann: How well do you see that mapping onto what’s actually happening with the social networks in the real world? On the one hand, we had de-platforming be successful to some extent after Charlottesville. On the other hand, it just reorganized in other forums and through other groups. Where are they drawing the line compared to what you find?

Nathan Kalmoe: Social media has certainly gotten more proactive in removing content that could potentially lead to violence in just the last several weeks. You mentioned Charlottesville earlier. They’ve gone much further recently. And of course, that is well within their rights. They’re private companies who can control the rules of who and how participates on their forums. I think it’s, based on what we have seen from our research and what we’ve seen other people present in their studies, whatever you feel is the trade-off for taking people off, it certainly has a benefit of reducing the potential for inflammatory statements with a broad audience. Particularly when it involves taking the president off of Twitter and Facebook, that I think is enormously important far more so than the relatively limited reach of ordinary people even when aggregated in large numbers. So in that aspect, I won’t give you a definitive normative statement about whether it’s ultimately the best course of action, but it certainly has positive benefits, and I’ll let other people decide what the cost of that are.

There’s sort of a whack a mole aspect to it to a certain extent that Twitter and Facebook cut people out and other social media websites try to take up the difference there but they’re never able… It’s true that those ideas don’t go away but the ability to communicate those ideas and to coordinate I think is effectively reduced because those other platforms, those other venues are not compensating fully for what is lost in terms of the anti-democratic and other sort of problematic rhetoric that is being taken off of social media.

Matt Grossmann: Lily you also find that some of the [inaudible 00:42:48] of endorsement of violence are maybe more ingrained having to do with personality or trait aggression. Does that mean that it’s going to be harder for these attitudes to change from any messages and does it matter if the parties increasingly sort in part based on these differences in personality?

Liliana Mason: Yeah, I mean, I’m not so worried about that effect because we’ve seen presumably trait aggression is relatively equally distributed across the electorate, across time, right? The way that I think about it is more like trait aggression is like the Tinder and the political environment and political rhetoric is the spark that’s really required to turn that aggression into approval of violence and moral disengagement and these types of attitudes. We don’t see average differences on trait aggression across the parties. Democrats and Republicans are relatively similar in terms of their average levels of trait aggression. They’re not actually sorting on that particular personality trait. We do know the Democrats and Republicans are different and on other personality traits like open versus closed, worldview type of personality traits but we don’t have evidence that those types of personality traits are correlated with aggressiveness or with approval of violence. For now, I’m not as worried about it because I think that it’s the type of thing that it doesn’t light on fire on its own.

The thing that we have control over is the political environment and how much the political environment is really activating the connection between aggression and violence and moral disengagement and these types of really harmful outcomes.

Matt Grossmann: Nathan, we were all hoping that 2020 was sort of a year unto itself with COVID and election year and we might be able to turn the page. Obviously that hasn’t been the case so far in 2021. Some of your own research shows that election year effects might not be that big. Does that pretend more of this ahead? Is there anything about the last year that should encourage us that maybe once we get into a new environment things will change?

Nathan Kalmoe: Yeah, we see some small shifts in violent attitudes over the course of our several years of surveys that seem to correspond with events or changes in the partisan media environment but we think that this general level of animosity isn’t particularly unusual at least in the 21st century. We don’t have surveys that cover that whole period but it seems more like there’s an underlying level of support for partisan aggression and violence and then how that manifests is a function of that broader environment, how it goes from attitudes to actions so the key is still what leaders do with those latent violent views and whether they encourage extremism and violence or try to stop it and it’s been asymmetric. We’ve seen Republican leaders including lots of media figures who have been encouraging that kind of extremism so the dampening, if there is a dampening is going to need to start with those leaders and it’s not clear that they have incentives to do that for a variety of reasons that we’ve mentioned already.

Even if there is a sharp leadership change, I suspect that many people will remain radicalized by what they’ve already been told and that sort of latent hostility to the point of aggression against their political opponents. I would say that there’s some encouragement from 2020 I guess that it’s not much of a glass half full but maybe a glass 5% or 10% full that we didn’t see a total collapse of the system of American elections. It was put under huge pressure and in a few key spots, sometimes in the court, sometimes among state administrators Republicans in those positions didn’t bend and break to the pressure that they were put under by other Republican leaders and by their voters. The problem going forward that makes this really half empty is that those people are now in the process of being replaced by Republican leaders and voters who are furious with them not going along with the attempt to overthrow the election and so this is I think we’re still in five alarm crisis mode as a result of that.

Matt Grossmann: Lily, what should we be looking for in the year ahead or in the research ahead and maybe talk a little bit about the methodological point of what is the role of survey research and understanding this relative to say listeners heard last week discussions of the actual armed groups that might have been leading the violence or level investigations of relationship between politicians and those folks. Where do you see survey research helping and fitting into the broader scholarly agenda?

Liliana Mason: Yeah, I think that one of the things that we’re really I think that lucky is not the right word but that allows us to understand these attitudes better is that we’ve been asking the same questions repeatedly more than a dozen surveys since 2017 and so to some degree because we have such a rich amount of data, we can actually see things like trends and election effects and the effects of individual events that are occurring. Longitudinally it’s important I think to understand how prevalent these attitudes are and whether or not this is a volatile type of attitude which really doesn’t seem to be, it doesn’t really seem to be very volatile at all. Without this survey data we wouldn’t be able to say that, right? We would just be making guesses and we also over time have seen that we have this consistent level of 10 to 20% of Americans who are generally open to partisan violence and political violence.

The survey research itself I do think provides some insight into what’s going on in partisan’s minds. We’re limited because we’re not going out and surveying militia members. We don’t have very many people in our samples that report ever having committed political violence, right or ever having done something violent on behalf of politics. I think in our last survey out of 1,000 people it was 17 people that reported ever having done that. We’re not actually surveying the militia people that are doing this violence but we do get a sense from what is the rest of the country think about this and how much support is there and how do we change these attitudes? What types of things can change these attitudes by using survey experiments and trying to think creatively about ways to measure things.

I think that’s been really helpful at least for us particularly in writing the book to have this sort of longer sense of what is the scope of this and what types of people are more likely to participate in this type of behavior and attitudes. The other thing is that one of the challenges that we’ve found is that it’s very difficult to study partisan violence in an era in which partisan violence is still occurring and it’s one of the things that we’ve had to do is just sort of say, “All right, we’re cutting off our data analysis and just finishing this book right now,” because we can’t continue to bring in new examples, as dark as that is that at this point it is sort of a difficult thing in American politics to study partisan violence because it is this changing and evolving thing. We know the underlying attitudes and so we can do research about that. The actual amount of violence in the incidents will hopefully we can explain in with future data but so far we’re just really interested in kind of the ways that Americans are thinking about this.

Nathan Kalmoe: I would just add to that that we conceptualize these broader attitude, public attitudes about partisan violence as not just a risk factor for the few individuals who are going to get mobilized into violent acts but also as a broader hostile partisan environment of people who aren’t going to take the actions themselves but who are talking to friends and family who might carry those actions out. For every one person who is attacking the Capitol, you have many more people back in their families, their friend groups, their communities who probably were encouraging them and saying, “Yeah, you’re right. I hope that you go to Washington,” and maybe even approving of the attack itself after the fact. This is not just about what the individuals in our surveys are going to carry out on their own but also the kind of encouragement and the atmosphere of partisan hostility and even legitimation of violence that they provide that will help to provoke and enable the people who do decide that they’re going to take action.

Matt Grossmann: Well, there’s a lot more to learn and thank you to Nathan Kalmoe and Lilliana Mason for joining me. Please check out Radical American Partisanship. We’ll be waiting anxiously for it and then join us again next time on The Science of Politics.

TapTheForwardAssist, CC BY-SA 4.0, via Wikimedia Commons