What this covers

Lilliana Mason, political scientist at the University of Maryland, joins Russ Roberts on EconTalk to discuss her book Uncivil Agreement and argue that American partisanship has intensified not because citizens disagree more on policy, but because party membership has fused with racial, religious, and cultural identity. Mason contends that treating partisanship as a group identity rather than a policy disagreement explains the anger and incivility observed in contemporary politics far better than standard models of polarization. The mechanics are straightforward: when elections function as contests between mega-identities, electoral victory becomes psychologically loaded in a way that can override substantive policy preferences, and citizens will sacrifice policies they actually support—a phenomenon Mason illustrates with the example of background-check legislation that 90% of Americans favored but most Republicans opposed to deny their opponents a win.

The conversation traverses how social sorting has eliminated cross-cutting identities, making the opposing side easier to dehumanize and harder to compromise with. Mason draws on intergroup-conflict research, including Marilyn Brewer's work on in-group preference and Robbers Cave studies showing that even meaningless group assignment produces zero-sum competition. She traces how party identities became aligned with other social identities beginning in the 1960s, creating a situation where citizens can flip policy positions overnight when party leadership shifts stance, with no loss of loyalty—a shift that undermines both accountability and the compromises necessary for democratic governance. The discussion also addresses potential structural solutions, including Duverger's Law and the possibility of moving to proportional representation to reduce zero-sum dynamics, before closing with examination of what might improve political discourse.

Sharpest takeaway

Mason argues that American partisanship has intensified not because Americans disagree more about policy, but because party identity has become aligned with racial, religious, and cultural identities, so that elections become zero-sum contests between mega-identities, and the drive to win can override actual policy preferences.

  • Partisanship behaves like any group identity, so intergroup-conflict research explains political hatred better than policy-distance models
  • Social sorting has eliminated cross-cutting identities, making the other side easier to dehumanize
  • The need for partisan victory can flip individuals' policy positions overnight, undermining accountability and compromise

The argument · threads3 threads · 38 claims
0.85

Ideological bundling and identity-based group membership intensify polarization despite underlying policy convergence.

2 pointscentrality 5/5
  • On many specific issues such as gay marriage and marijuana legalization Americans have moved closer together over time, so there is less policy polarization, even as the feeling of us-versus-them and partisan identification separate from ideology has grown stronger.

    “Now, people seem to be closer together. So, there seems to be less polarization on many issues. And yet, as you point out, on the feeling of us versus them, that seems to be getting stronger.”

  • Intersectionality—the idea that if you agree with me on one issue you must be with me on every issue or you are not on the right team—is consistent with social sorting, because requiring people to check many identity boxes creates tighter in-group bonds and makes it harder to tolerate those who differ on any single dimension.

    “if you are with me on this issue--whatever it is--you've got to be with me on every issue. You are not on the right team.”

0.56

Seeking out thoughtful opposing voices and diverse in-person engagement mitigates outrage better than following unlike others online.

1 pointcentrality 4/5
  • Advice to follow people unlike yourself on social media can backfire and increase outrage because opponents seem so unfair or wrong; a better version is to seek out the quieter, more thoughtful voices on the other side and to find in-person activities and groups unlikely to simply confirm your prejudices.

    “a lot of the people aren't like me--I do, I follow on Twitter, I do increase my outrage because they are so unfair... So, when I give that advice, I now add, and try to find the quieter, more thoughtful ones.”

0.29

Landmark social psychology experiments warrant skepticism due to potential experimenter manipulation and questionable replicability.

1 pointcentrality 2/5
  • Many dramatic social psychology experiments from the 1950s, including Robbers Cave, warrant skepticism because experimenters may have tweaked conditions to get dramatic results, and it is unclear how often the same outcome would recur if the experiments were repeated many times.

    “I'm a little bit skeptical of those, that experiment... I wonder, you know, how much the experimenters tweaked the experience to get something dramatic. I wonder how many, if they'd done it 20 times, would have happened every time.”