Why Political Polarization Feels Different Now

Every election cycle, someone declares that political polarization has reached a new high, and every time, they're probably right. That's not a comment on any particular race or headline. It's a structural feature of what's actually happening in American political life, and it's been building for longer than most of the conversation around it admits.

I want to slow down and ask a more basic question than "why is everyone so angry." What exactly is polarizing? Because the honest answer is: not what most people assume.

Two Kinds Of Polarization We Keep Confusing

There's a distinction political scientists draw that almost never makes it into ordinary conversation, and once you see it, you can't unsee it. Ideological polarization is about positions. It measures how far apart people sit on actual policy questions: taxes, immigration, abortion, guns. Affective polarization is about feeling. It measures how much people distrust, dislike, or simply refuse to see fairness in the other side, independent of what that side actually believes.

The researchers Shanto Iyengar, Gaurav Sood, and Yphtach Lelkes gave this second phenomenon its name in a 2012 paper in Public Opinion Quarterly, "Affect, Not Ideology: A Social Identity Perspective on Polarization."

Their central finding, since replicated many times: partisanship increasingly functions like a social identity rather than a policy stance — closer to a sports rivalry or a religious affiliation than a set of considered opinions. You can hold moderate views on most issues and still feel visceral distrust toward people who wear the other team's jersey.

Here's the part that surprised me when I first read the research: these two forms of polarization don't move together. The American National Election Studies feeling thermometer has asked respondents since 1978 to rate their warmth toward the opposing party on a 0-100 scale. It shows a much steeper and steadier climb in cold feelings toward the other side than any comparable measure shows in actual policy disagreement. This is the single most important fact in the whole conversation about polarization, and it's the one that gets skipped past fastest. We are not primarily divided by what we believe. We are divided by how we feel about the people we imagine believe differently.

Why It Feels Sharper Right Now

If affective polarization is the real engine, then the explanations that focus purely on policy disagreement are looking in the wrong place. Something else is doing the sorting. In his 2000 book Bowling Alone, Robert Putnam documented the decline of the civic associations, unions, and mixed-membership clubs that used to put Americans with different views in the same room on a regular basis. Bill Bishop's 2008 book The Big Sort took the argument further, documenting that Americans increasingly choose to live among political neighbors, moving into communities and social circles that mirror their own views rather than communities that challenge them.

Put those two findings side by side and you get a fairly plain picture. We used to be stuck around people who disagreed with us, whether we liked it or not, in the union hall, the bowling league, the extended family gathering. Structural stuckness with people who disagreed with us. Now we can, and increasingly do, opt out of that stuckness. And the algorithmic feeds most people scroll every day are optimized to keep them engaged, not informed, which tends to reward the most emotionally charged framing of the other side rather than the most accurate one.

Is polarization the disease, or is it the symptom of something quieter, the disappearance of the ordinary, unchosen contact that used to keep us honest about each other? I lean toward the second reading. The gap in how we feel about each other isn't total, but it is wide, and it appears to be a downstream effect of losing the everyday spaces where disagreement used to happen face to face, with stakes attached to the relationship rather than to the argument.

The Sorting Nobody Voted For

There's a particular irony worth sitting with. Nobody sat down and decided, as a matter of policy, to sort the country into ideologically homogenous neighborhoods and media diets. It happened as the aggregate result of millions of individually reasonable choices: move somewhere with good schools and low crime, follow accounts that feel true, spend time with people who get you. Each choice made sense in isolation. The pattern they add up to didn't.

I think this matters because it changes what kind of problem polarization actually is. It's not primarily a problem of bad arguments or insufficient facts. If it were, better information would fix it, and there's little evidence that simply supplying people with more accurate facts about the other side reduces the temperature. It's a problem of structure: the physical, social, and digital architecture of daily life stopped putting us next to people who disagree with us, so we lost the practiced skill of disagreeing with someone we still had to see at church, at work, or at Thanksgiving.

What Structure Can Do That Willpower Can't

This is where most advice about polarization goes soft. "Be more open-minded." "Listen to the other side." "Assume good faith." All true, none of it durable, because none of it survives the first real disagreement without some kind of scaffolding underneath it. Good intentions dissolve fast under actual conflict. What holds up is structure: a format, a set of rules, a shared expectation about how the conversation will proceed before anyone's feelings are on the line.

Civility without structure tends to fail for a predictable reason, and the polarization research backs that up from a different angle entirely. The instinct to just try harder to be civil to people you already distrust runs into the same wall every time: it asks willpower to do a job that only structure can do reliably. A format that both people have agreed to in advance, before the disagreement gets hot, changes what's possible in ways that goodwill alone doesn't.

Two structured moves that research on dialogue consistently returns to:

• Restate before you respond. Before replying to someone's position, restate it in a form they would recognize as fair — not a strawman, not a summary designed to make their view look weak, but a version they'd actually nod at.
• Position switch. Argue the other side's case as if it were your own for a few minutes before you argue against it.

Both moves sound small. The evidence here is thinner than the identity research above. A handful of small studies on structured dialogue formats — restatement exercises, perspective-taking tasks — report shifts in participants' warmth toward the other side after a single session, even when the underlying policy disagreement barely moves. The effect sizes are modest and the research base is nowhere near as deep as Iyengar's core finding, but the direction fits what an identity-based account of partisanship would predict: change how people encounter each other, and the feeling can shift even when the argument doesn't.

Why do these work when "just listen better" doesn't? Because each forces a concrete, checkable task instead of a vague aspiration. You can tell whether someone restated your position accurately. You can't easily tell whether someone is being "open-minded." One of those is a skill you can practice and get better at. The other is a mood you can't reliably summon on demand, especially not with someone you've been primed by accumulated distrust to see as the opposition.

For facilitators: a workable session keeps to 45–60 minutes, pairs participants in groups of two to four, uses a neutral facilitator whose role is to enforce the format (not to judge content), and defines success as the speaker confirming their position was restated accurately — not as agreement.

A Different Kind Of Fix

I don't think the answer to affective polarization is persuading people to change their politics. That's not what the problem is, and treating it that way misses the actual mechanism. The problem is that the ordinary, low-stakes contact that used to keep partisan dislike in check has mostly disappeared, and nothing has structurally replaced it. Bowling leagues didn't fix polarization by resolving political disagreements. They fixed it, to the extent it was ever fixed, by making the people on the other side of an argument into people you also knew as neighbors, teammates, and friends.

That's a harder thing to rebuild than a policy position, because it isn't really about policy at all. It's about whether the architecture of your week, your neighborhood, your feed, ever puts you in a room with someone who disagrees with you and gives you a reason to stay in that room past the point of discomfort. I think that's the real work: not winning the argument, but building or finding the structures that make staying in the room possible in the first place.

None of this resolves quickly, and I don't think it's supposed to. The sorting took decades to build. It would be strange if the way back took anything less.

Frequently Asked Questions

What's the difference between ideological and affective polarization? Ideological polarization measures how far apart people's actual policy positions sit. Affective polarization, a term formalized by Iyengar, Sood, and Lelkes in a 2012 Public Opinion Quarterly paper, measures how much people distrust or dislike those on the other political side, independent of specific policy views.

Is polarization actually rising, or does it just feel that way because of social media? Both. Cross-party dislike, as tracked by the ANES feeling thermometer, predates most social media platforms. But feed-based media, optimized for engagement rather than accuracy, appears to amplify and accelerate a sorting process that was already underway.

Does where people choose to live actually contribute to polarization? Yes, by reducing unchosen contact. Bishop's The Big Sort (2008) and Putnam's Bowling Alone (2000) trace two sides of the same shift: people sorting into homogenous communities as the civic associations that used to mix political lines eroded.

Can structured dialogue practices actually reduce polarization, or is that wishful thinking? Structure changes what's achievable in a way that "listen more" appeals generally don't, because a shared format gives participants a concrete, checkable task. The two techniques described above — restate-before-you-respond and position switch — don't require either person to change their politics, only to agree to the format before the disagreement starts. For session-design details, see the facilitator note in the body above.

What can one person actually do about a trend this large? Not much at the national level, and I think it's honest to say so. But the same research that explains why polarization spread, the loss of ordinary cross-cutting contact, also points at what rebuilds it locally: structured spaces where people who disagree still have a reason to stay in the room together. That's a smaller, slower answer than a policy fix, but it's the one with an actual mechanism behind it.

Last updated: 2026-08-19