Donald Trump and Hillary Clinton supporters
Credit: Gage Skidmore/Flickr

In June 2012, a hopeful Barack Obama predicted that once he won reelection, the hyper-partisan “fever” that had been roiling Washington might finally “break.” Congressional Republicans had spent Obama’s first term committed to blocking whatever he proposed, sight unseen, in order to make him a “one-term president,” as Mitch McConnell tartly put it. But, Obama reasoned, once there was no second term to prevent, “we can start getting some cooperation again.”

Six years later, Obama is gone from the White House, and Republicans control all three branches of government. “The fever may break” now looks more like a sentence missing a direct object: American democracy.

Uncivil Agreement: How Politics Became Our Identity by Lilliana Mason University of Chicago Press, 192 pp. Credit:

For a solid decade, wonks and pundits have been charting America’s growing political polarization. By now, a reasonable consensus seems to be that, first, partisan polarization is at record-high levels; second, it exists not only among elites, but also among the mass public; third, it is increasingly driven by negative partisanship—dislike for the other party—rather than substantive issue-based differences; and, fourth, it is the end product of a half century of geographic and demographic sorting. Fifty years ago, both parties contained broad and overlapping coalitions. But today the Republicans are overwhelmingly the party of white, primarily rural and exurban Christians who call themselves “conservatives,” and the Democrats are the party of everyone else.

The case for optimism (“The fever may break”) is that, ultimately, there are decent people on both sides who care more about their country than they do about their party winning. And that at some point, a new equilibrium will emerge, because it has to. Or that somehow, we’ll muddle through, because we always have.

The case for pessimism (“The fever may break us”) is that once all political conflict becomes organized around a single us-versus-them dimension, something dark and destructive happens to our collective psyche. We see each other not as fellow Americans, but rather as enemies who would use their power to destroy us and betray the true promise of the country. And once that happens, it’s hard to get back to peace without some kind of breakdown or violence.

I was pessimistic before reading Lilliana Mason’s new book, Uncivil Agreement. I am even more pessimistic now. Mason, a political scientist at the University of Maryland, has written an extremely important analysis of the social psychology of what happens when the political world gets divided into two warring tribes, with no overlap. The central contribution of the book is to show that partisan sorting isn’t just a consequence of identity—it also creates identity. This is critical, because, as Mason observes, “identities themselves have psychological effects of their own.”

Mason draws on a tradition known as social-identity theory, which explains how we construct our sense of self based on group memberships. Humans need to both fit in and feel special. Affiliating with a larger identity group—a religion, a race, a profession, a class, a hometown—gives us both an us and a them. In normal times, this is fine, because we have lots of overlapping identities and, as a result, a wonderful mess of things to agree and disagree about. But when all our identities line up, and we divide into two opposed teams based on them, ugliness follows.

Group theory was a staple of an earlier tradition in mid-twentieth-century political science, when the discipline was more grounded in sociological approaches that treated citizens as members of communities—Catholics, Protestants, union members, farmers—who made sense of the world by talking to each other and deciding what was good for “people like us.” In that era, writes Mason, “social divisions between Americans over party, ideology, religion, class, race, and geography did not align neatly, so that particular social groups were friends in some circumstances and opponents in others.” As a result, many more Americans weren’t sure which party represented them better. They were what political scientists call “cross-pressured voters.”

Individually, these voters were far from ideal citizens. They rarely followed politics and were poorly informed. They switched back and forth between the parties, voting based on whims, or not at all. But, collectively, they formed a kind of pragmatic center, flexible enough to allow broad swings in response to changing conditions and candidates. “Democracy needs these voters,” writes Mason. “Not only are cross-pressured voters a source of popular responsiveness, they are also a buffer against social polarization.”

Today, there are far fewer cross-pressured voters. Mostly, Republicans live surrounded by other Republicans, and Democrats live surrounded by other Democrats. Partisanship is now a “mega-identity.” And this has big consequences.

When partisan opponents have less and less in common, politics becomes more of a zero-sum game: I see any political gain for you as a loss for me, and vice versa. Being right becomes more important than getting it right, since our sense of self-worth is on the line. Motivated reasoning—a cognitive bias in which we process information in a way that supports our preexisting beliefs—becomes especially prevalent. Group-based partisanship leads us to think that our side is the natural majority (after all, everyone I know agrees with me), which makes us more likely to get angry and confrontational. All in all, Mason writes, “We are priming the pump for a very energetic battle.”

For all those cheering the latest wave of civic engagement, then, this signals a note of caution. Partisanship, as Mason notes, is a “significantly more powerful predictor of political action” than issue-based concerns. Mobilizing partisans to engage in team-based politics might be a great way to win elections. But it also reinforces the existing dynamic of politics as being all about our side winning. 

Mason invokes the controversial conclusion of the 1954 political science classic, Voting: A Study of Opinion Formation in a Presidential Campaign, which “suggested that an electorate composed entirely of ‘deeply concerned’ voters would be unresponsive at the aggregate level,” largely because they wouldn’t be able to shift partisan loyalties in response to changing conditions. Of course, the alternative—disengaged voters who bounce between the two parties and don’t care enough to get involved—seems pretty discouraging, too. This points to a larger flaw in two-party democracy. More on that shortly.

Meanwhile, the sad news for the many Washington wonks who are still gamely searching for “bipartisan solutions” is that even where both sides might agree on a substantive compromise, their dislike for each other tends to overwhelm any potential cooperation. “All of the political arguments over taxes, welfare, abortion, compassion, responsibility and the ACA,” writes Mason, “are built on a base of automatic and primal feelings that compel partisans to believe that their group is right, regardless of the content of the discussion.” To be fair, bipartisanship does still happen sometimes, especially outside of Washington and on lower-profile matters that neither side expects to become campaign issues. But on the big stuff, the partisan win-lose calculus tends to overwhelm deal making between rival elites. 

Note, for example, that in constructing the Affordable Care Act, Democrats went out of their way to try to incorporate Republican ideas, in hopes of winning bipartisan support. They got none. For Republicans it wasn’t about the policy. It was about trying to deny Obama a win and giving themselves a campaign issue (“Repeal socialist Obamacare!”). Many members of Congress have private friendships with opposing partisans and would prefer to wheel and deal. But when their leaders—or their charged-up voters back home—tell them not to compromise, they listen. In other words, our political divide can’t be bridged by clever new policies that borrow from left and right. As soon as a policy becomes about one side getting its agenda through, there is no possibility of bipartisan lawmaking, no matter the substantive bipartisan appeal of the policy.

Mason sees Trump as an apotheosis of the logic of winning and losing. His platform was incoherent and unworkable, and it failed all the conservative litmus tests. Instead, he literally promised “winning.” And, as Mason notes, “Winning grows increasingly important as identities grow stronger.” Trump’s support was strongest among those who “shared multiple Republican-linked identities.”

Affiliating with a larger identity group gives us both an us and a them. In normal times, this is fine, because we have lots of overlapping identities. But when all our identities line up, and we divide into two opposed teams based on them, ugliness follows.

While Uncivil Agreement largely ignores the asymmetries in the party coalitions, Mason published an academic paper earlier this year (“One Tribe to Bind Them All,” with political scientist Julie Wronski) finding that because the Republican coalition rests on fewer identities (basically: white and Christian), Republican voters tend to have a stronger sense of social-identity team partisanship. Democrats are a more diverse coalition, which makes them a little less subject to the same all-or-nothing thinking. The implication is that it’s Republicans who play more identity politics than the Democrats, but because they have fewer identities, their version has the illusion of seeming more unifying.

While it’s true that partisan polarization exists at the voter level, Mason sometimes overlooks the degree to which it is crafted and exaggerated by party leaders, especially congressional Republicans. Mason cites a fascinating set of polling data from 2013, in the wake of the Sandy Hook shooting. According to the poll, 81 percent of Republican respondents said they personally supported expanding background checks—but only 57 percent said they supported the Senate passing a bill doing so. Mason concludes that “when it came to the moment of public partisan competition, party victory trumped preferred policy for many Republicans.”

She’s not wrong: it is deeply alarming that Republican support for gun control dropped off by 24 points thanks to partisan hostility. But that still leaves a comfortable majority of Republican voters supporting legislation to expand background checks. Why, then, did almost all Senate Republicans end up voting against it? The answer is not just their desire to deny Obama a win; it’s also their need to cater to the minority of hyper-engaged Republican voters who take their voting cues from the National Rifle Association. The NRA is far from alone in the coalition of existentially revanchist groups that power the GOP coalition and confine the actions of its elected leaders. The public may be more polarized than in the past, but most Americans still tend to agree on some basic liberal positions, whether it’s gun control, fighting climate change, or taxing the rich. It’s the influence of right-wing interest groups, media, and donors that keep those preferences from becoming policy. But the implication of Mason’s research is that growing partisan-based identity will make those special interests’ jobs ever easier. 

Our political divide can’t be bridged by clever new policies that borrow from left and right. As soon as a policy becomes about one side getting its agenda through, there is no possibility of bipartisan lawmaking, no matter the substantive bipartisan appeal of the policy.

Mason concludes with the requisite attempt to answer the pressing question “Can we fix it?” The most promising idea is what Mason calls an “unsorting” or a “rift in one party”—the possibility that one of the two major parties splits apart, scrambling the existing alignment that is doing so much damage. Of course, given that our single-winner plurality congressional and presidential elections all but enshrine the two-party system, and the fact that the declining Republican electoral coalition still has a few election cycles left in it, this seems highly unlikely to happen on its own. So far, even the more temperamentally moderate Republicans have mostly remained on the Trump train, especially if they plan to stay in office. Thanks to voter self-sorting and partisan gerrymandering, elected representatives are more worried about fending off more extreme primary challengers than appealing to whatever swing voters are still out there. Meanwhile, the tantalizing possibility of winning the always-potentially-but-never-actually-definitive next election wrecks any incentives for bipartisan compromise. (Political scientist Frances Lee’s 2016 book, Insecure Majorities, is the definitive read on this point.) Given that American institutions generally require considerable compromise to make significant policy changes, the current alignment is a recipe for a government mired in bitter and dysfunctional gridlock, which will only generate more frustration and anger.

But the situation is not hopeless. We could update our electoral system to drain two-party partisan competition of its worst zero-sum, winner-take-all tendencies. One promising approach is ranked-choice voting, which was just implemented in Maine. In a ranked-choice system, voters can go beyond the binary partisan choice, instead ranking several candidates in order of preference. If one candidate wins a majority of first-place votes in the initial tally, the election is over. But if no one wins a majority, the last-place candidate’s votes are distributed to their voters’ next-ranked choice. This continues, knocking off the last-place candidate in each round, until one candidate has more than half of the votes. The advantage of this is that it pushes candidates to broaden their appeal in order to pick up second- and third-choice votes, since a plurality isn’t enough, and diminishes the value of the kind of negative campaigning that reinforces partisan hostility.

America would almost certainly be better off in a proportional representation system, like most other advanced democracies, where no party expects to have a majority and therefore compromise and bargaining become the norms. Nothing in the Constitution would stop us from electing members to the U.S. House of Representatives in multimember districts with ranked-choice voting, which would create an opportunity for multiple parties to compete nationally. And state governments could take the lead in experimenting with this approach. The Senate is a tougher nut to crack, but we could use ranked-choice voting for Senate elections, too.

The social-psychology experiments that Mason draws on tend to divide groups of people into just two teams. But, as Mason notes repeatedly, when group identities are overlapping, not binary, social harmony is far more likely. The more teams, the more peace, because allegiances are always shifting. Without the winner-take-all nature of closely balanced two-way competition, leaders would have fewer incentives to demonize other parties. After all, they might need those parties to form a coalition government after the election.

Uncivil Agreement will stand as one of the most important political books of 2018. It is clear and compelling, and makes an incredibly important point: When all of our group identities are aligned into two competing political teams, two-party politics is completely unworkable. If the current realignment toward ever deeper cultural and ethnic divisions continues, our brains are just not equipped to resolve it peacefully. The fever will not break on its own.

Our ideas can save democracy... But we need your help! Donate Now!

Lee Drutman is a senior fellow in the New America Political Reform program and the author of Breaking the Two-Party Doom Loop: The Case for Multiparty Democracy in America.