Braver Angels on Why America's Divide Keeps Deepening

The most striking thing about the fourth plenary of the 2026 Braver Angels National Convention was that the organization’s own founders used it to say their work has not been enough. Co-founder Bill Doherty told the room in Philadelphia that the country is more divided than when the group started in 2016. Co-founder David Blankenhorn said the same thing an hour later in different words. Braver Angels is a cross-partisan nonprofit that runs structured workshops between self-identified “reds” and “blues,” and this session, titled “Renewal Happens Together,” was built around an uncomfortable pairing: ten years of growth, and a problem that has widened underneath it.

This was not a debate, so there were no opposing sides on stage. What the plenary offered instead was a sustained argument about what has gone wrong in American political life and what might fix it. That argument is worth taking seriously on its own terms, and it is worth setting against the people who think it is mistaken.

The argument from the stage

Doherty put it bluntly: “we’ve gotten worse.” He named social media as part of the cause and described a shift in the character of American disagreement, not just its intensity. Blankenhorn framed the same point differently, saying that what the movement is up against continues to get worse and that its work has not yet broken through at what he called a nation-changing level.

The measurable evidence points the same direction. Pew polling on partisan hostility found that the share of Republicans describing Democrats as more immoral than other Americans climbed from 47 percent in 2016 to 72 percent in 2022, with the Democratic figure moving from 35 to 63 percent over the same span.

What that captures is affective polarization, meaning dislike and distrust of the other side rather than distance on policy questions. It has risen considerably faster than measured disagreement about what government should actually do. Comparative research places the United States as an outlier among wealthy democracies for the speed of that increase, which is the closest thing available to a hard version of Doherty’s claim.

Contempt as the diagnosis

The plenary’s account of the cause was consistent across speakers: the problem is not that Americans disagree but how they treat each other when they do. Braver Angels workshops teach participants to identify what the organization calls the four horsemen of polarization, which are stereotyping, dismissing, ridiculing and contempt.

Mary Thomas-Vallens presented the Dignity Index in the same frame, describing an eight-point scale where 1 marks the deepest contempt and 8 marks the belief that every person holds inherent worth. She attributed to its co-creator Tim Shriver the idea that contempt is the enemy of your own cause, a claim about strategy rather than manners. The argument is that treating opponents as beneath argument does not defeat them, it hardens them.

There is real research underneath this. More in Common’s perception gap study found that Democrats and Republicans imagine roughly twice as many of their opponents hold views they consider extreme as actually do. The people with the widest gaps were the most politically engaged and, in some measures, the most educated, and the wider a person’s gap, the more likely they were to describe opponents as hateful or brainwashed. If that is right, a significant share of American animosity is directed at a version of the other side that does not exist in the numbers claimed.

Division inside families

The plenary’s most persuasive material was not statistical. CEO Maury Giles told the room that one of his daughters had announced she was taking a break from the family, and that he and his wife had seen their grandchildren for the first time in months only the week before. A volunteer from Northfield, Minnesota, said the hardest thing for him was watching his family tear itself apart over politics while feeling he lacked the courage to intervene.

These are anecdotes, and the plenary presented them as such. But the phenomenon behind them is documented. Research on affective polarization has found measurable effects on close family relationships, including shortened holiday gatherings between relatives of opposing partisanship, and links to strained family ties and social distrust more broadly.

This is where the organization’s pitch is strongest and where its scope is clearest. It is a movement that treats political division as a relational injury and applies techniques borrowed from family therapy, which is exactly where Doherty’s professional background lies. Whether that scales to a country is a separate question the plenary raised but did not settle.

Political violence and the conspiracy problem

Doherty’s sharpest illustration was local to him. He noted that the convention fell near the one-year anniversary of the assassination of Melissa Hortman, the Minnesota House speaker emerita, and said that a conspiracy theory blaming the governor began circulating almost immediately, despite Governor Tim Walz having been a close friend and ally.

This checks out and is arguably worse than he described. Hortman and her husband Mark were killed on June 14, 2025, and claims linking Walz to the killings spread within days, pushed initially by conservative accounts with very large followings. PolitiFact found no evidence for the claim, which did not stop it from being amplified from the presidency in January 2026 or from resurfacing around the anniversary.

Doherty’s point was that this represents a change in kind, not just degree. Grief itself has become contested territory. It is difficult to read the Hortman case as anything other than support for that reading, and it is the single strongest piece of evidence the plenary offered for its central claim.

Where the gridlock argument overstates the case

The plenary’s answer to “what now” was that dialogue must lead somewhere, and the vehicle presented was Common Sense American, a program of the National Institute for Civil Discourse that briefs its members on policy options and takes the results to Congress. Its ambassador director Mary Tanner described a four-for-four record on surprise medical billing, infrastructure, the Electoral Count Reform Act and Workforce Pell. All four became law, though the claim that member briefs helped pass them is the organization’s own assessment rather than an independent finding.

One characterization needs adjusting. Workforce Pell arrived inside the One Big Beautiful Bill Act, which cleared the Senate 51 to 50 and the House 218 to 214 on party lines. The provision itself had long-standing support in both parties, so calling it bipartisan describes the policy accurately and the vote that carried it not at all.

A second claim is wrong as stated. The program’s Minnesota ambassador told the room that no significant immigration legislation has been passed and signed into law in three decades, and used that to argue both sides have chosen total victory over partial progress. In fact, President Trump signed the Laken Riley Act in January 2025, after it passed the Senate 64 to 35 and the House 263 to 156 with Democratic votes in both chambers. Its own sponsors called it the most significant immigration enforcement legislation since 1996.

The defensible version of the claim survives the correction, and is arguably more interesting. What has not happened in thirty years is a broad settlement covering legal pathways, status and enforcement together. What has happened is enforcement legislation passing with bipartisan support. That is not a Congress incapable of acting on immigration. It is a Congress capable of acting on only one half of it, which is a different diagnosis with different implications for what citizen pressure could achieve.

The dissent: sorting, not polarization

The plenary treated mass polarization as established fact. A significant strand of political science disagrees. Stanford’s Morris Fiorina has argued for two decades that the American public is not experiencing unprecedented polarization, and that what has changed is party sorting: the two parties have become internally homogeneous and sharply distinct, while ordinary voters have moved much less than the coverage suggests.

On this account the divide runs between political elites and everyone else rather than between red and blue citizens. Fiorina’s conclusion is that “the fault does not lie in a polarized citizenry” but in two sorted parties, each representing a minority, fighting to impose narrow programmes on a large and varied country.

This view is not a rebuttal of the perception gap research so much as its natural partner. If most Americans are less extreme than each side imagines, then the felt experience of a country at war with itself may be substantially a product of what people are shown rather than what their neighbours believe. That reading supports Braver Angels’ method and undercuts its framing at the same time. The workshops would be correcting a misperception rather than healing a real chasm, which is a smaller and more tractable problem than the one the plenary described.

The dissent from the left and from the right

Two sharper objections come from opposite directions and reach the same conclusion about bridging work.

From the left, the argument is that the polarization frame is itself the error. A 2026 article in Perspectives on Politics argues that treating asymmetric radicalization as ordinary two-sided polarization naturalizes a false equivalence and produces analytical blind spots. On that reading, asking people to lower the temperature is the wrong instruction if one side is moving away from democratic norms rather than simply moving rightward.

From the right, conservative writers have argued that groups like Braver Angels understate how existential conservatives believe the current moment to be, and that some depolarization organizations are distrusted for claiming neutrality while carrying assumptions from one side. The practical form of the objection is simple arithmetic: hours spent at bridging events are hours not spent winning. Both critiques agree that dialogue is a luxury good in a genuine emergency. They disagree entirely about whose emergency it is.

What the evidence says about whether this work works

The strongest available answer speaks to all of these positions at once. A 2024 megastudy published in Science tested 25 interventions on 32,059 participants and found that 23 of them reduced partisan animosity, with the largest effects coming from sympathetic portrayals of political opponents and appeals to shared identity. Those are precisely the mechanisms a Red/Blue workshop runs on.

The same study found that reducing animosity did not necessarily reduce support for undemocratic practices or for partisan violence. The researchers concluded that distinct psychologies underlie the two. What most reduced antidemocratic attitudes was not warmth but accuracy: correcting misperceptions about what the other side actually believes.

That finding cuts in an awkward direction for the plenary’s framing. It suggests the work does what it claims, and that what it claims is narrower than renewal. Getting Americans to dislike each other less is achievable and measurable. Preventing the next conspiracy theory about a murdered legislator appears to be a different job requiring different tools.

Franklin’s argument that the divisions are old

The plenary closed with an actor playing Benjamin Franklin, and the historical argument he made was the session’s most substantial. Assembling men in the hope of profiting from their collected wisdom, he said, also assembles their prejudices, their passions, their errors of opinion, their local interests and their selfish views, all of which were on daily display in 1787. The language tracks Franklin’s actual closing speech to the Constitutional Convention closely enough to be recognisable.

The point being made was that American division is not a modern deviation from a harmonious past. Contemporaries expected the convention to fail, and some wanted it to. What Franklin praised in the resulting document was not its perfection but its amendability, the fact that a people who discovered its flaws retained the power to change it. He then made the rising sun observation that James Madison recorded on September 17, 1787, as the last delegates signed.

Blankenhorn had made a compatible case earlier, borrowing Václav Havel’s distinction between optimism and hope. Optimism, on that account, is believing your work will succeed. Hope is proceeding because the work makes sense regardless of the outcome. It is the argument a movement makes when its own leaders have just told the room that the numbers are going the wrong way.

How two other democracies structured the same problem

Two foreign models make the shape of the American approach clearer, and both change how it reads.

Ireland went the opposite way on who participates. Its Citizens’ Assembly, established in 2016, seated 99 citizens chosen at random to mirror the population by age, gender, class and region, chaired by a Supreme Court judge. Over five weekends the assembly deliberated the constitutional ban on abortion, heard expert evidence and recommended change. Ireland’s Citizens’ Assembly was purely advisory, yet its recommendation led directly to the 2018 referendum that repealed the Eighth Amendment.

Germany went the opposite way on recruitment. Deutschland Spricht, launched by the newspaper Die Zeit in 2017, uses a matching algorithm to pair strangers who disagree on a set of yes-or-no questions and sends them to meet once. Harvard and Stanford economists studying the programme found that a single conversation reduced affective polarization and left participants more generous toward people from opposing parties. The model now operates across more than 30 countries as My Country Talks.

Braver Angels recruits people who already want to bridge divides, which means its reach is bounded by who volunteers. Germany reaches participants through a newspaper’s existing readership, capturing people who were not looking for a depolarization organization. Ireland conscripts by lottery and routes the output into a constitutional vote. All three shift attitudes. Only the Irish model has a formal path into policy, and the closest American equivalent, Common Sense American’s congressional briefings, runs through voluntary membership rather than random selection. That is the ceiling built into the American version, and it is the reason a plan to reach 15 million people has to grow an opt-in list to succeed.

Discussion Questions

Do you think political division in your country has deepened over the last ten years?
Is the bigger problem what people disagree about, or how they treat each other while disagreeing?
Have you ever been surprised to learn what someone on the other side actually believed?
How much of the sense that a country is divided comes from what people are shown online?
Does treating opponents with dignity make it harder to hold them accountable?
Which description fits better: a polarized public, or two sorted parties?
Should reducing personal animosity count as success if support for political violence does not fall?
Why do groups working to bridge divides attract criticism from both the left and the right?
Would you trust a randomly selected group of citizens to recommend policy on a divisive issue?
Imagine you are organizing a conversation between neighbors who disagree about immigration. What is your first question to them?

Braver Angels on Why America’s Divide Keeps Deepening