Map the Divide: Qualitative Analysis of Polarization

According to Frayed But Not Broken: Understanding American Division in a Polarized Age, the United States faces serious polarization, but the evidence does not show that the country is inevitably descending into widespread political violence. Greg Berman's report argues that public discourse often looks more divided than everyday political attitudes and behavior.

• HFG's report is based on 18 expert conversations conducted by Greg Berman during 2024 and 2025 across history, political science, psychology, sociology, philosophy, philanthropy, and civil society.
• A Reuters/Ipsos poll cited by HFG in July 2024 found that four in five Americans feared the country was sliding into chaos, while the Polarization Research Lab found in polling conducted since 2022 that just under 3 percent supported partisan violence.
• Gallup data cited by HFG for 2025 found that 45 percent of Americans identified as independent, compared with 27 percent identifying as Democrats and 27 percent identifying as Republicans.
• HFG cites research showing that 70 percent of US social media users never or rarely posted or shared political or social issues, according to Pew Research Center data published on May 4, 2021.
• HFG's central qualitative insight is that the bonds connecting Americans may be frayed, but civic associations and ordinary cross-group relationships can still support pluralism.

The HFG report concludes that American polarization is urgent but does not by itself prove that widespread political violence is imminent. Greg Berman reached that conclusion after interviewing 18 experts during 2024 and 2025 and comparing their perspectives with polling, historical analysis, and political-violence research.

The report separates several concepts that are often blended together: elite polarization, affective polarization, ideological polarization, false polarization, and political violence. The Northwestern University Institute for Policy Research defines affective polarization as the gap between positive feelings toward one's own party and negative feelings toward the opposing party, and HFG reports that this gap increased markedly over the two decades before the report's November 2, 2026 publication.

The report also identifies a gap between political presentation and public behavior. Gallup data cited by HFG for 2025 showed that 45 percent of Americans identified as independent, while 27 percent identified as Democrats and 27 percent as Republicans. HFG interprets those figures as evidence that red-versus-blue labels can obscure the number of Americans who do not identify with either major party.

HFG cites the Polarization Research Lab's polling since 2022, which found that just under 3 percent of respondents supported partisan violence. HFG also cites Armed Conflict Location & Event Data, which reported that extremist group activity and acts of political violence involving extremist actors declined each year after 2020.

The report does not minimize threats or intimidation. HFG cites Washington Post data showing that more than 1, 000 Americans were shot and killed by police annually from 2020 through 2024, while noting that the report found little indication that those deaths were directly related to politics.

Several interviewees offered concise warnings and qualifications. Yphtach Lelkes said, "The state of America is that we have strong partisanship but partisanship is not necessarily synonymous with violence." Clionadh Raleigh added, "A lot of people are getting radicalized in their opinions without becoming radicalized in their actions."

Researchers studying political polarization should analyze perceived division, stated beliefs, online participation, and observed violence as separate qualitative and quantitative signals. HFG's report shows why a high level of fear does not automatically demonstrate high public support for violence.

A robust research workflow should compare elite statements with public interviews and open-ended survey responses. Researchers should also segment findings by party identification, independence, ideology, education, age, geography, and civic participation when those variables are available.

The report makes misperception a central analytic issue. HFG cites the 2021 study "The Ties that Blind, " which found that partisans consistently overestimated the prevalence of their opponents' extreme political attitudes. Researchers can therefore code not only what participants believe, but also what participants think the opposing group believes.

The report also points to civic association as a potential protective factor. Peter Coleman argued that societies are better able to mitigate intergroup violence when they contain "crosscutting structures, " while Robert Talisse said, "The lifeblood of democracy is that we disagree. That's why we need democracy." Qualitative research can examine where those cross-group relationships exist and what activities help sustain them.

The HFG report shows that American polarization is real, consequential, and more complex than a simple red-versus-blue narrative. The report's 18 expert interviews, polling evidence, and discussion of false polarization show why researchers should compare public perceptions with behavior, party identity, online participation, and support for violence. AI-enabled qualitative research can make that comparison faster while preserving themes, segment differences, and source-level evidence. Try Evidano for free to analyze interviews, reports, and open-ended survey data in one workspace.