• As the Rhetoric Escalates: Talking to Lilliana Mason

    How does a left-leaning political constituency end up calling itself conservative? How might partisan identities constrain our supposedly rational policy preferences without us even noticing it? When I want to ask such questions, I pose them to Lilliana Mason. This present conversation focuses on Mason’s book Uncivil Agreement: How Politics Became Our Identity. Mason is an assistant professor of government and politics at the University of Maryland, College Park. Her research on partisan identity, partisan bias, social sorting, and American social polarization has been published in academic journals, and featured in mainstream media outlets including the New York Times, the Washington Post, CNN, and National Public Radio. Mason received a 2017 Emerging Scholar Award from the American Political Science Association (APSA).


    ANDY FITCH: To start with some of the depressing social context, Uncivil Agreement outlines a broad national split into opposed political camps or cultures (increasingly sorted through overlapping race / class / religion / region demarcations), fueled by partisan identity differentiation (with its negative stereotyping of the other side, its uninformed or imaginary news distortions, its emotionally volatile cycles of anger and enthusiasm, its zero-sum prioritization on short-term victory over long-term good governance) more than by substantive policy disputes. So here could you, in true nonpartisan fashion, first sketch how / when some of these most unsavory aspects of present social behavior have in fact proved evolutionarily advantageous for the species, and perhaps even for American political culture? And could you then start to trace a more problematic threshold in our recent politics at which competitive rivals become embittered enemies, and at which partisanship becomes less a source of galvanizing social organization than of dysfunctional social division?

    LILLIANA MASON: In terms of a social identity’s most basic function, these identities do help to keep human beings organized, and to keep society going. Social cooperation seems to require that we think of an “us” and a “not us.” These types of social categories help us to make sense of a complicated world. So civilization more broadly seems to require that we identify with groups, and that we privilege our own groups over others. This doesn’t necessarily mean hating other groups. It simply means liking our own group the most, and doing the most work to help our group.

    I don’t see anything inherently wrong with that type of social behavior. And political parties and partisanship likewise can serve the important function of helping us organize our understanding of the political world, and can simplify politics. Again we need that. In a democracy, we can’t expect every individual citizen to study every single Congressional bill, or to understand the inner workings of every government agency. Instead, we elect representatives who can go and deal with all of that for us. And here political parties provide what we call heuristics — mental shortcuts basically, helping us decide who we want to vote for, who we think should represent us. If you lean in one partisan direction, then you can assume that usually your party will best represent your particular interests, and behave in ways you probably agree with.

    The problem comes when you begin to identify with your party so strongly that you cannot imagine ever voting for the other party. Democracy requires accountability. If an elected official does something that we don’t like, that doesn’t serve our interest, then we should have the option and the impetus to vote against this person next time. But really strong partisanship allows bad behavior to continue in government, and allows representatives to ignore or even work against our interests with virtually no consequence.

    In terms of partisanship becoming a constrictive bind, one crucial aspect in your account of how social identity shapes today’s politics comes from the premise that we cannot just isolate one particular category of identity (either race, or religion, etcetera) and study its singular potential to override voters’ rational policy choices, but must instead consider how any number of overlapping identity categories might serve to rigidify one’s demographic mega-identity. So here, while perhaps keeping in mind a conventional folk-theory sense of how America’s democratic decision-making should work, could you outline a couple real-life scenarios in which we might mistakenly presume that a policy-driven political difference plays out (either amid well-informed and complexly calculating voters, or among single-minded issue-driven voters), and where you would instead detect the operations of a more diffusive (but also in some ways much more monolithic) ideological identity?

    Well first, I don’t ever want to make the argument that people have no real policy preferences, because obviously they do. But in terms of a folk theory of democracy, I really wanted this book to argue against what had been a prevailing political-science understanding of how voters behave. Here I like to make the analogy of bankers choosing an investment, scrutinizing very soberly which policies will work best for them and their family and community, evaluating all the different options. We used to assume that voters in a two-party democracy operated this way. If one party changed its position regarding an individual voter’s policy preference, we thought this voter might change parties accordingly. But a lot of the newer research in political behavior suggests that voters don’t really shift this way, because they have these strong identity connections to specific parties. People often will change their own policy preferences rather than change their vote.

    In his paper “Party Over Policy,” Geoffrey Cohen describes manipulating which party supposedly advocates for either a stringent or a generous welfare policy. He describes how people who identify as partisans will switch their own policy preference in order to match their party’s. And then Cohen’s findings get a bit alarming when he asks these people whether party identification has influenced their policy position. People say no. They don’t believe party ties influenced their policy position, even though Cohen literally just manipulated his presentation so that this type of influence could happen. And then he actually asks them to write a letter to the editor of a local newspaper. He emphasizes that their local representative reads these sometimes, so their letter might have an influence on changing the policy, so they should really think through why they hold this position and what good supporting arguments they can give. And people come up with reasons and write these very thoughtful letters on behalf of a policy that they just embraced on the spot.

    And so a lot of us probably consider other voters irrational and hopelessly biased, but think that we’re not. But we also should keep in mind, from Cohen’s experiment, that biased voters don’t necessarily recognize it. And even worse, some research into motivated reasoning has found that people with the most political knowledge actually are the most prone to this type of thinking. They have the intellectual ammunition. They have enough political knowledge to offer what feels like a coherent explanation, and to undermine the opposing argument. So again, if our folk theory presumes that the most informed voters will make the smartest and most sober choices, contemporary social science tells us that these engaged people might behave less like bankers, and more like sports fans.

    Could we start addressing mega-identity here also? It particularly interests me that prioritizing specific aspects of one’s intersectional identity could seem likely to produce a diffusive range of political positions — but that, in Uncivil Agreement’s account, mega-identity at present often stacks various identity aspects on top of each other, so that we end up stuck with something more like a stark partisan binary.

    So within studies of American political behavior, social psychology, and sociology, people have been writing about the concept of social identities for years. I didn’t invent the concept of “mega-identities,” otherwise known as identity alignment or complexity (or as the absence of cross-cutting identities). But essentially, the concept is that people have a huge number of different group identities, any of which might seem the most salient at any given time. I mean, you have an alma mater. You have a hometown. You might follow sports teams. All of those designations involve you (and a bunch of other people) identifying with a certain group label and considering yourselves connected to one another. And in general, we know from social-identity theory that the identity at the top of your mind at any given moment most likely will be the identity facing the most pressing threat. And partisan identity plays its own unique role, since we have regularly scheduled threats to the status of our party, right? Every two years we sense a potential threat to this particular group identity. Every time you go through one of these status contests, also known as elections, your party might lose, though you still also have these other parts of your identity, which may or may not connect to your party’s victory or loss.

    But over the past few decades, the parties have become increasingly aligned with other social identities including race, religion, rural or urban location. And when these links start connecting our parties and other parts of our social identities, then all of this gets drawn into that one particular political competition. The outcome of an election then feels so much more consequential for our own broader sense of who we are. We can’t just say: “Well one part of me lost, but the rest of me is still doing great” (or vice versa). Instead, we feel devastated when we lose and really really great when we win.

    And here Marilynn Brewer’s work on identity alignment also shows us that if most members of Group A also belong to Group B, we can consider these well-aligned identities. Brewer has found that when two or more identities strongly align, then the people within these groups feel more intolerant towards out-group members. Take the example of people possessing both Irish and Catholic identities. If you’re Irish Catholic, you probably know a lot of Irish people and a lot of Catholic people. You don’t necessarily need to know a lot of non-Irish or non-Catholic people, because they’re not necessarily part of your group. But if you’re Irish and Jewish, then you probably know some non-Irish and some non-Jewish people, since you yourself don’t fit into any well-overlapping categories. So just being a member of two non-overlapping identity categories breeds tolerance, because it exposes you to a broader range of people. At the same time, well-aligned categories breed intolerance because you can think of “outsiders” as unlike you in multiple ways. So once these mega-identities get formed, we start to think of out-group partisans as quite different from us — not just in terms of their political views, but also racially, religiously, and with any number of overlapping categories. We feel ever more socially distant from these out-group members, which makes it easier to dehumanize them, to think about them with less generosity.

    And amid the demographic sortings that Uncivil Agreement delineates, where might we also be overlooking temporarily dormant (but nonetheless foundational) tensions within supposed partisan blocs: with for example many communities of color more religious and more socially conservative than their white counterparts, or with demographic designations of gender largely absent from your analyses but quite crucial to contemporary ideological-identity formations, or with vast intra-community differences present amid any reductively categorized “Latino” or “Asian American” constituencies? I’m sure you have thought through, and have needed to answer for, such inevitable argumentative snags on countless occasions. And politicians themselves long have grasped the need to proactively define the most pressing demographic dynamics within a given election cycle. So what lived approach have you developed for applying mega-identity designations provisionally, to see what you can make of them, without internalizing these social constructs of the moment as some fixed taxonomy of the prescripted identity and / or voting positions available to us?

    I’d start by going back to James Madison’s idea that the health of a democracy depends not on a lack of conflict, but on a multiplicity of conflicts. So we can have these polarized relationships, but lining up in all different directions, and actually helping to weave society together. You might dislike your neighbor supporting the wrong sports team, and being an alumnus of a rival school to yours, but maybe you both go to the same gym and church, and so you recognize that he’s ultimately a relatable guy. A society that fosters such multiplicity of conflicts generally can remain a stable, less hostile type of society. Our own society has fostered those kinds of cross-cutting identities in the past, and could do so again. Right now, though, social identities line up more often than they get scattered around.

    And as you’ve mentioned, alongside more polarized mega-identities, Uncivil Agreement places prospects for cross-cutting citizens to stitch back together such seemingly stark divides. In an era of inflamed political passions, cross-cutters can provide essential civic ballast by being the citizens least likely to engage in angry, or excessively enthused, paranoid, or vindictive politics. Partisans might conceive of such cross-cutters as mystifyingly muddled or confused or indifferent, but in your account, cross-cutters can play a crucial role in coaxing democratic governments to respond flexibly, pragmatically, to ever-changing domestic and global circumstances. So particularly at a historical moment when we see the percentage of cross-cutting Americans increasingly diminished, how might we most proactively reformulate the positive value that even supposedly disengaged cross-cutters bring to our political culture? 

    First, to give a concrete example that actually goes a little beyond what the book does, my more recent research points to how some racial and religious identities are still cross-cutting, particularly within the Democratic Party. As you suggested, African American voters are largely Christian, a religious identity associated with the Republican Party. And so that asymmetry between the parties, and that potential for cross-cutting, really interests me right now. The Democratic Party, as sort of an umbrella, currently incorporates almost all of the non-white identities, including all of the non-white evangelical identities. The Republican Party basically can be understood as the party of white Christian rural people, and the Democratic Party contains everyone else. So if you pick two random Republicans out of the entire population of Republicans, they most likely share racial and religious identities. If you pick two random Democrats out of the entire population of Democrats, they have much lower chances of sharing those identities.

    I think the Democratic Party right now is about 56% white, for example. That helps to make the Democratic Party much more heterogeneous than the Republicans, which also has all of those other consequences I mentioned earlier in terms of well-aligned identities. Democrats have to cope with cross-cutting identities percolating under the party umbrella, as a basic political condition. You can’t say that every Democrat necessarily votes on behalf of their particular racial group or religious group. Participants in this relatively diverse coalition have to make compromises, and to understand each other as compatriots, even when they differ in a variety of social ways.

    Republicans often don’t face this same constraint. That allows them to be much more strict about adhering to the correct identities, and being the “correct” type of party member. And a Republican who does not identity as white, Christian, conservative, rural often has a much less solid partisan identification. Again, evangelical Democrats also might have some cross-cutting identities that weaken their partisanship, but on average, Republicans feel much more acutely the effect of not being part of all the right groups than Democrats do.

    And then more broadly for how political cross-cutting plays out in the book: I ran an experiment where I had people read blog postings that were very threatening to their political party. These posts would say that ordinary Americans don’t trust Republicans, or they’d call Barack Obama a radical socialist, and they’d use various kinds of threatening language. And I found that people with cross-cutting identities responded to these messages with relatively low levels of anger, whereas people with well-sorted partisan identities responded with the maximum level of anger that it was possible to report. And then I tried to figure out whether people felt a threat specifically to their party identity or to their policy preferences. I found that people with extreme issue positions got very angry when I threatened issue positions, and people with strong partisan identities got really angry when I threatened the party. But the people with well-sorted identities actually got angry at every kind of message I could give them, whereas people with cross-cutting identities typically didn’t have this same emotional response. Again here I probably should point out that people with well-aligned identities didn’t necessarily get more angry than the angriest voters used to get. But the people with cross-cutting identities, the calmest people, the people who dampen the overall emotional reaction, just keep becoming an increasingly diminishing portion of the electorate. So today’s angry person might not be any angrier than before, but we have many fewer non-angry people, making us angrier on average.

    So the book gets into this concept of “bad activism” — because in general, given the turnout problems in our national voting history, we tend to think of all political engagement as good. But when you act on behalf of these elevated anger levels, and of unquestioning partisan loyalties, you might actually just perpetuate this extreme polarized system.

    Of course, like you said, political cross-cutters also tend to engage less frequently, or less actively in our politics. So perhaps something like mandatory voting could help. Getting those cross-cutters engaged would make our political conversations a bit less intense. It could open more room for compromise. Democratic lawmaking requires finding room for compromise, for having a civil conversation, for respectfully discussing competing perspectives with your opponent. And right now, you could think of political cross-cutters as the only remaining contingent in the population who can hold government accountable, who can possibly change their votes if they don’t like a given political party’s new approach, or if they sense their representative has done something wrong.

    Well in terms more generally of how we enact (or act out) our partisan politics, I long have wondered why so many professionally successful progressives compulsively perform the roles of angry activist, of public protester — as if no other possible means of political engagement existed (say, immersing oneself in the instrumental activities of local governance), as if they somehow lacked whatever social capital their most disempowered fellow partisans lack, and as if this strident posturing wasn’t likely to coax forth its own opposition (basically neutralizing whatever positive charge such interventions were intended to “activate”). And so here again, how might your synthesis of mid-20th-century political-science literature on ideological identity (or the lack thereof) and contemporary psychological literature on self-esteem (and intra- as well as inter-group dynamics) help to explain why such seemingly self-defeating, emotionally reactive (rather than strategically policy-centric) approaches not only play out frequently even among some of American’s most well-credentialed citizens, but in fact make them (make us) feel so good?

    In terms of mid-20th-century thinking on ideological identity, I’d start with the classic 1964 work by Philip Converse, which basically starts by recognizing that the vast majority of the electorate remains relatively unaware of which policy goes with which party, or which policies speak to left or right principles. Most people just don’t have a consistent set of beliefs along these lines. And then a book by Donald Kinder and Nathan Kalmoe from a couple of years ago found that same basic pattern still playing out today.

    Again my own related work uses the term “constrained.” If you have a “constrained” set of issue positions, then all of your policy preferences consistently swing to the left or to the right. The people with well-constrained issue positions often have more extreme issue positions also. But for many voters, an identification with an ideological label means something different than holding a constrained set of issue positions. You might call yourself a conservative, and feel very, very strongly identified as a conservative, and still not actually hold consistently conservative policy positions. In fact, a number of political scientists have found that the American electorate, as a whole, prefers on average left-leaning policy positions. But on average, this same electorate calls itself conservative. And because we do have these generally left-leaning policy preferences as an electorate, conservative politicians feel the need to double down on conservative social identification. Matt Grossmann and David Hopkins published a book called Asymmetric Politics about this difference between Democrats and Republicans. So we might anachronistically think of Democrats as the “identity politics” party, but it actually makes much more sense for Republicans to campaign on identity.

    References to mid-60s precedent also get me thinking about some Western European liberal democracies emerging from the immediate postwar context more clearly sorted into major center-right and center-left political parties, of the type that mid-20th-century APSA might endorse, but then ending up with populist demands for in-group recognition not so different from what we ourselves face today. So where do you see Uncivil Agreement tracking specifically American phenomena, and where do you see your book tracing broader structural conditions characteristic of an increasingly globalized economy?

    The difference between the American two-party system and some European multi-party systems is that the populist, nativist parties emerging in Europe are still largely only minor parties. Because we in the U.S. have only two parties to choose from, the populist, nativist party now represents half of the partisan electorate (whether those partisans agree with it or not). So the American system gives what might otherwise be a minority view (that America should be a white, Christian nation) a large boost in terms of power. And the zero-sum nature of the two-party competition in American politics facilitates such powerful partisan bias that people feel compelled to vote for their own party, even if they don’t entirely agree with what the party says.

    And for one quick follow-up on these intra-group dynamics: even if, from the outside, one perceives a relatively monolithic partisan group providing participants with a stable, secure, empowering base of emotional belonging, from within these groups themselves one might feel the incessant anxieties, frictions, dislocations of what Sigmund Freud would have termed a “narcissism of small differences.” Uncivil Agreement doesn’t focus much on the frequently factionalized, fraught internal combustion propelling partisan politics forward. But if you had unlimited space and time with this book, how might you have addressed those even more intimate aspects of both personal and group partisanship?

    That gets us to an important point about where our current story of social sorting and partisan alignment begins. I’d say that starts in 1964 with the Civil Rights Act, which gives a great example of how intraparty conflict can break up a party. Of course you don’t normally see large numbers of voters switch directly from one party to the other. Usually, for a few decades, you might see an expanding number of voters calling themselves independents. So following the Civil Rights Act, many conservative Southern Democrats didn’t like the Democratic Party anymore, and maybe felt that they belonged better with the Republican Party, but still wouldn’t call themselves Republicans. It really took a couple decades, and possibly for a next generation to come along. So that basic Democratic Party rift produced large numbers of cross-cutting identities for a few decades. From the early 70s through the early 90s, plenty of people found themselves in social groups that didn’t match their political groups. People still were receiving confused signals from the parties about who belonged where. Ironically, that confusion actually played a role in reducing polarization.

    But more generally, and even amid these intraparty conflicts or tensions, I think one can make the argument that socially aligned polarization has been useful in various ways, particularly in allowing people of color to have a much more influential voice in our political system. Even as we have divided ourselves, one broader benefit has come from one of the parties finally representing the interests of non-white people. And of course part of what so viciously divides us today in terms of partisanship comes out of race remaining the great source of conflict throughout American history.

    So with past intraparty and interparty conflicts, you still had white people and men nearly always winning. But increasingly today you see the Republican Party feeling the need to defend not just conservative principles, but a whole traditional social order premised on white-male patriarchy. To some extent, all of our social sorting actually has brought us to this point of the parties dividing on any number of lines of race, gender, sexuality. Partisan polarization at least has made these social divides broadly visible — and therefore, possibly, addressable on a large scale.

    And then again in terms of intraparty conflict, within the Democratic Party you have this basic struggle over who gets represented, by whom. You don’t have a unified identity so much as a whole bunch of different people trying to speak to each other and cooperate and advocate for their own vision of the party. Within the Republican Party, you only find this to a lesser degree. We now see a Republican rift between the Trumpers and the Never Trumpers, but with the Never Trumper crowd not big enough to do real damage.

    Well amid our ongoing social sorting, I did find it hard to stay within Uncivil Agreement’s clearly delineated confines of psychological / political partisan dynamics, without addressing the daunting “external” reality of social media (and a much broader array of micro-targeted data analytics). So how might Uncivil Agreement’s psycho-political reflections scale themselves up to, say, a scenario that former Homeland Security Secretary Michael Chertoff recently sketched for me: in which GPS tracking might determine I am about to cross through a largely immigrant neighborhood, prompting a local alt-right organization to send me some racist pseudo-TV-news message (or maybe some more personalized contact from a live, empathic, “concerned fellow citizen”) about certain social groups’ criminal tendencies? How might present-day partisanship find itself further rigidified by such technological developments, which seem to hold the possibility for cementing, institutionalizing, and making structural any number of inter-group divisions — way beyond what we even can imagine at present?

    So like you suggested, I intentionally didn’t bring up social media in this book because it requires a whole other book. Jamie Settle’s new book Frenemies does focus, for example, on social media and polarization. But again I tend to describe the more gradual emergence of our present sorting starting with the Civil Rights Act, and with controversies over school prayer and abortion. The latter issues got the religious right engaged, so then another identity started lining up with partisanship. And all of this happened before widespread Internet usage. And partisan cable news also spread widely almost alongside the Internet. So you do get this perfect storm of multiple identities moving into alignment and picking up these new social megaphones, making it much more clear to audiences who is “us” and who is “them.”

    And I agree that social media still has the potential to drive us much further apart. But that whole process also might go too far, and push people to realize how insane these discussions have become. Eran Halperin, for example, did this really interesting study in Israel, essentially circulating overtly extreme propaganda in an already very conservative neighborhood, and actually having this moderate people’s attitudes. Apparently residents just did not feel that extreme, that bad. And maybe as Americans keep holding up this ugly mirror to ourselves, we could recognize something similar. Of course the question here becomes whether we stop engaging in this way before any violent conflict ensues. No doubt that type of micro-targeting you describe might just encourage violence, as social media already has done in multiple countries.

    Here still in terms of holding up a mirror to ourselves, and also of projecting an “us” or “them”: of course it’s not hard to sympathize with critical discourses that seek to redress historical oppression and contemporary inequalities (perhaps by favoring one long-suffering social group at the expense of other more privileged groups). So I can understand the motivations behind that form of identity politics on the left, while still condemning narrow nativist identity politics on the populist right. But does this very act of “understanding” one side in such polarized disputes, of giving one’s in-group the benefit of the doubt, also explicitly enact one of Uncivil Agreement’s working definitions of bias — with bias defined (as you mentioned at the start of this talk) by the positive / preferential treatment provided to one’s partisan in-group, as much as by some more malevolent assault perpetrated on one’s partisan out-group? Can characterizing bias this way (as a relatively universal, relatively benign-feeling aspect of ordinary in-group behavior) maybe deflate the emotional stakes of providing some psychologizing account of the opposing side’s pathologically unjust, uncivil mindset? And at the same time, can such a depersonalized conception of bias raise the policy stakes by prioritizing the need to proactively address disparate social outcomes produced by systemic / institutional bias, rather than to police the endless possibilities for personal / intentional bias?

    First, I think it’s important to differentiate between the implicit bias that most people hold, and the explicit bias that motivates hatred toward others. Implicit bias is simply the automatic response that your brain comes up with before you can even really think things through. These implicit biases often reflect the influence of larger social inequalities and stereotypes, but they can also be combatted. Although we all engage in implicit bias, we also have the ability to practice “un-biasing” our responses. Patricia Devine has a lot of work demonstrating that people can self-regulate implicit racial bias if they are motivated to do so, and if they practice.

    Explicit bias is the type of prejudice that is conscious and measurable simply by asking people questions about their feelings toward out-group members. This is more familiar, overt prejudice. It can also be moderated, but only if people are willing to try. Those who hold explicit out-group bias, and are unwilling to work on reducing it — that I think is where this idea of “incivility” comes in.

    I also think it’s important to note that the two parties don’t approach civility the same way. The “anti-PC” message promoted by the Trump campaign is a distinctly uncivil approach to discourse. The Trump campaign was also largely rooted in the promotion of white identity, and the exclusion of non-white racial out-groups. So I don’t believe that the universal concept of human susceptibility to in-group bias is really comparable to some of the deeply racist and intentionally uncivil rhetoric of at least the pro-Trump element of the Republican Party.

    To start then pivoting back to the most contemporary applications of this book, could we consider the 2018 midterms? What positive and less-positive-than-it-might-seem aspects do you see, say, amid historically high turnout in so many districts? Where do you see cross-cutters offering a post-2016 course correction for American politics? Where do you see angry uninformed partisans further crowding out possibilities for real debate? What might you find helpful in relatively wonkish electoral referenda (say on nonpartisan-redistricting processes, or on post-felony voter eligibility) succeeding? What most constructive and clear-eyed lessons should those advocating more civil agreement take from 2018?

    In terms of the 2018 electorate, cross-cutting identities stood out the most among college-educated, suburban, Republican women — the group already drifting away the most from the Republican Party. Some of these women voted for Democrats, and some just didn’t vote at all. When typically Republican voters don’t vote, that also causes Democrats to win elections.

    And then in terms of these 2018 results: first we should keep in mind that having unified control of government does tamp down some of the conflict within government. So I know this sounds crazy, but we actually might see much more partisanship over the next two years than we have since 2016. Now that Democrats and Republicans can fight it out at the institutional level, we may start seeing even more polarized public attitudes emerge — especially with the news media already covering politics (even legislation) as a zero-sum horse race. Instead of asking how much a bill might cost, or which Americans it might benefit or harm, or who among the population supports it, news media typically ask: “Will this be a win for Democrats or for Republicans?” So we did have a blue wave in 2018, with some Republican cross-cutters bowing out, but who knows if that will continue as the rhetoric escalates?

    Returning then to outward-directed anger as a galvanizing force in our highly sorted politics, one demographic category you rarely dwell on in Uncivil Agreement’s mega-identity demarcations is age. But as I watch my parents and their peers retire, I see how drastically isolating (and in-group insulating) retirement can be, and how such isolation might just intensify the dynamic you describe in which social homogeneity makes us all the more prone towards distorted and hostile projections onto supposed threats and purported rivals. What broadest mega-sorted ideological-identity implications might you see as our baby-boomer population bulge transitions from perhaps its most socially integrated to its most socially isolated phase of life?

    Good question. And I also picture this particular group you describe watching the American population diversify in any number of ways right before their eyes. Or I mean, I think of them sending all those absurd chain e-mails. They’re much more likely than younger people to spread fake news. But in Congress at least we won’t have so many of these boomers after a certain point. Pew just put out a report showing that for the first time the number of baby boomers in Congress has dropped, while the number of Gen X and millennial Congressional representatives can only increase. So in one sense we might be seeing the final push for this traditional, older white patriarchy. But at the same time, I guess we might watch lots of Americans now get newly radicalized in their retirement homes. Though if Republicans can’t build an electoral coalition crossing racial lines, I don’t think those boomers in the retirement homes can provide sufficient numbers to win elections. The Republican Party just won’t be able to find enough white, evangelical, rural people unless our elections are severely manipulated.

    Still, for all of these questions, we should keep in mind that nothing makes the Republican Party inherently white and Christian. Post-Trump, we definitely could see some different kind of Republican leader come along who actually embraces the 2012 GOP autopsy report urging the party to reach out to Latino voters. If the party does so, that could generate a huge number of new cross-cutting identities, which actually would help us all — because we wouldn’t have such stark racial cues about voting. We hopefully wouldn’t have interparty conflicts feeling so personally dire for so many individuals. Fewer voters might believe that the opposing party threatens their entire self. That all hopefully comes out of the Republican Party recognizing that it has transformed since the 1990s from a majority party to a shrinking, aging demographic.

    Yeah, though the political theorist David Runciman also has helped me to recognize that if we think this angry retiree vote will simply die out soon, we fool ourselves into forgetting that many more old people will live much longer than ever in the past.

    Yeah. Good point. I mean if this population you describe starts funding the more extremist racist groups, that gets scary pretty fast.

    Maybe here we could talk through prospects for (and limitations to) contact-theory approaches, as laid out in your “How to Fix It” chapter. If digital technologies are currently running an unprecedentedly far-reaching social-engineering experiment on what happens when a culture hardwires itself for sorted partisan identity, could any comparative experiment nudging us towards inter-group engagement more positively push in the opposite direction? And if our isolation from particular groups prompts us to invent rationales for why we should be isolated from such groups, could an increased diversity of social encounters prompt us to create rationales for why we in fact value these exchanges? And what does the history of prejudice-reducing efforts around school desegregation teach us about the real-life consequences of such contact-theory initiatives?

    Contact theory operates from the basic idea that if you spend time with strangers then you’ll get to know and like them. But further research has found some necessary additional conditions, such as that you need to establish a situation with no overt conflict or competition between the two groups. And research also has shown that, sometimes at least, you might only need a friend or family member to have this social contact for it to shape your own feelings.

    But more broadly, one early and important study took place during the Korean War, while the military was being desegregated pretty randomly, battalion by battalion. This study found that white soldiers in desegregated battalions had become more racially tolerant than white soldiers in the segregated battalions. So social contact did its work here — probably in part because everybody could be on the same side, with the same status.

    And sharing a common enemy.

    Right. I mean, I often joke that the easiest way to end our current dynamic would involve aliens attacking the Earth, and allowing us all to see each other as belonging to the “human team.” I’ve been brainstorming with other scholars lately about how we all could encourage contact between partisans — who really do live in different social spheres at this point. Today’s Democrats and Republicans shop at different grocery stores. They wear different clothing brands, and watch different TV shows. And the military stands out as one of our best-integrated social spheres. So maybe some type of national-service program encouraging high-school graduates to go off and do a year of service, rebuilding our infrastructure, could help. This program could do something useful for the country as a whole, even while instilling a sense of individual patriotism and of interpersonal commonality across partisan divides, and while giving participants something like reduced college tuition or some equivalent to incentivize them to do it.

    And again, since constructive “contact” can at times happen more remotely, I’ve also thought about how we’ve changed our mass media following the Civil Rights Era, and how what my father would see on television about race differs so drastically from what I saw. We developed basically a new set of rules about what type of language to use, and how to depict others. Obviously we haven’t always done a great job with this. But currently, we still have no equivalent social stigma for extreme and demeaning partisan language. We have no social sanction for engaging in this really aggressive, divisive, demonizing way with our fellow Americans. So I’d love to see a TV show that positively depicts people from mutually unfamiliar groups, or that engages them both in conversation, or puts them on some project and makes them complete some complex task together. 

    So for one probably impossible question to answer, still on the need to curb conflict or competition among contact-theory participants, your book makes the case that “Our emotional relationships with our opponents must be addressed before we can hope to make the important policy compromises required for governing.” I can totally agree with that. But until we address (presumably through substantial policy reform) the pressures that economic precarity impose on so many Americans right now, how can we ever get to the kind of sturdy, reassuring ground that would allow us to re-found our most emotionally charged inter-group relationships?

    I hear what you mean about that being a catch 22. I’d say, first of all, that work by Michael Tesler and others has pretty well convinced me that economic anxiety didn’t cause racial resentment. Racial resentment was stoked first, and then that made people feel more economically anxious. And so one optimistic account of our present social tension might present these trends as kind of lancing the boil of racism, finally — actually addressing, from a governmental perspective, an entire political structure devoted to America’s original sin. Inevitably, that lancing would feel messy and chaotic.

    It would feel pretty much how this present feels, I’d imagine. And so again, maybe we needed all these identities to line up along partisan lines in order finally to have this conversation to actually address all of these unsaid things. Maybe we must take that necessary step to get to a more culturally, racially, socially tolerant future. And maybe we never could get there without the Nazi sympathizers coming out of the woodwork to say their crazy part, and without really really uncomfortable protest and social discord. So maybe it’s not a big catch 22 after all. Maybe we do need first to deal with the underlying racism of the system, and to deal with each other as partisans, so that we can finally deal with each other as people. Maybe having the parties divided by race makes this conversation much more likely and possible. Defusing the anger we have for each other could possibly happen without legislation, and we just need to have difficult conversations about systemic racism. And maybe we’re having those now. Or maybe we’ll never stop having them. I mean, the only time we had a civil war it was over race, basically. So we haven’t proven that we can successfully have these conversations.

    And to close on one slightly different angle, in terms of your concluding detection of an “unfortunate truth…that these deep social divisions are allowing opportunities for policy compromise to go unnoticed,” in terms of such a bitterly partisan moment perhaps in fact presenting more low-hanging fruit than ever for potential policy compromises, I especially would appreciate hearing how cross-cutters (those presumably feeling less threatened, less catalyzed) might today help to stabilize American society by finding the wherewithal to “fight” their more fired-up peers on both partisan right and left. Where do you see (or most want to see) that happening? 

    Great question. Here again the catch 22 comes from cross-cutters being less likely to vote, but also the voters most likely to hold government accountable. And we have seen some anti-partisan tendencies even in the parties. Some Democrats just elected to the House do seem less linked to the party, whether they are further to the center or to the left. They ran on this need to get stuff done, and to find a way to do it. So definitely having candidates with cross-cutting identities run for Congress helps, because, for the average citizen, politics has become mostly just a spectacle at this point. That makes some people pay more attention, and some pay less attention. The people interested in winning pay more attention, and the people committed to effective policy pay less.

    Samara Klar and Yanna Krupnikov published a great book called Independent Politics, where they address all these Americans who call themselves independents but who actually just feel embarrassed to call themselves partisans, because of how nasty partisanship has become. They don’t overlap perfectly with cross-cutters, but they are part of this group who does seem the most likely to want to get stuff done, but also the least engaged as consistent voters, and increasingly more turned off each day. So that does seem to me the central problem, with no good solution, short of forcing them (and everybody else) to vote.