# Radical American Partisanship **Nathan P. Kalmoe and Lilliana Mason · University of Chicago Press, May 2022 · 216 pp.** The book is the standard political-science attempt to measure how many ordinary American partisans endorse, excuse or wish for violence against the other side, and what makes those views rise or fall. It builds on Mason's *Uncivil Agreement* (2018), which argued that party has become a "mega-identity" stacking race, religion and region together. *Radical American Partisanship* asks what happens when that identity turns lethal. It draws on four years of national surveys and experiments, with data running through the 2020 election and the [[January 6 Capitol Attack]]. ## The findings - **Violent partisanship is a minority view, but not a tiny one.** In the authors' 2017–18 precursor surveys, **5–15%** of partisans endorsed violence. **15–20%** in each party agreed the country would be "better off" if large numbers of the other party's voters "just died." The book refines these figures, and the precursor paper is the cited source for them. - **[[Moral Disengagement]] is the gateway.** Between **40% and 60%** of partisans showed it: seeing the other side as evil, less than human, or a threat that justifies breaking norms. - **The two parties look alike on attitudes but not on action.** Rank-and-file Democrats and Republicans are "remarkably similar in partisan moral disengagement and violent views." Real-world political violence, though, comes mostly from the right. The authors' explanation is asymmetric: "one side is moving toward multiracial democracy while the other is violently opposed to it." - **Personality matters more than ideology.** Ideological extremity does **not** predict support for violence. **Trait aggression** was the most consistent predictor, and hostile sexism was a strong one among Republicans. - **Elections and leaders move the numbers.** Telling partisans their side would win in 2020 roughly *doubled* their support for violence. Anti-violence messages from party leaders reduced it. That makes leader rhetoric a lever in both directions, which is the mechanism behind [[Stochastic Terrorism]]. **The January 6 wave** (as reported by Rachel Kleinfeld in the *Journal of Democracy*, Oct 2021, citing Kalmoe and Mason): - Republican support spiked in October 2020. - By February 2021, about a fifth of Republicans and 13% of Democrats called violence justified. - About a quarter of Republicans said violently taking over state buildings was at least "a little" justified. ## The critique Westwood, Grimmer, Tyler and Nall (*PNAS*, March 2022) argued that this whole literature **overstates** support for violence. Prior estimates had a median of 18.5%. With concrete violent scenarios and filters for inattentive respondents, they found **2.9%**. They traced the gap to vague "violence" wording and to disengaged respondents clicking at random. The dispute is now a standard methods caveat. Survey "support" for political violence measures disposition, not intent, and the figure moves a lot depending on how the question is worded. ## Since 2022 Kalmoe now runs UW–Madison's Center for Communication and Civic Renewal; he was previously at LSU. Mason is SNF Agora Institute Professor at Johns Hopkins. Neither has published new violence-support figures since the [[Assassination of Charlie Kirk]] (Sept 2025); Mason's public comments since then have been interviews. Other pollsters have filled the gap. PRRI, for example, found that the share of Republicans agreeing "true American patriots may have to resort to violence in order to save our country" fell from **35% (Aug 2021) to 19% (Nov–Dec 2025)**. ## Why it is in this vault The book supplies the measurement for things the vault otherwise asserts from anecdote: - that [[Affective Polarization]] and [[Negative Partisanship]] have a violent tail - that the tail is driven by identity and personality, not by policy disagreement - that elites can lengthen it or shorten it It sits alongside [[Tribal Epistemology]], [[Social Identity Theory]], [[Realistic Conflict Theory]] and the [[Contact Hypothesis]] as the quantitative anchor of the [[Political Behavior Hub]]. ## See also - [[Political Behavior Hub]] · [[Political Science Hub]] · [[Reading Log Hub]] - [[Affective Polarization]] · [[Negative Partisanship]] · [[Polarization]] · [[Moral Disengagement]] · [[Dehumanization]] - [[Stochastic Terrorism]] · [[January 6 Capitol Attack]] · [[Assassination of Charlie Kirk]] · [[Arlie Russell Hochschild]]