Electoral democracy rests on the conferral of Loser’s Consent. The Capitol insurrection of January 6, 2021—and the widespread sympathy for it that endures among Republican citizens and elected officials—can be understood as an unprecedented denial of that Consent. We hypothesize that insurrectionist sympathies among 2020 election losers are structured in part by hypercompetitiveness—a psychological need to win at all costs.

Using original survey data collected in the spring of 2022, we find strong suggestive support for our hypothesis. Many of the Americans who fail to condemn the Capitol riot may not be simply knee-jerk partisans, well-intentioned victims of propaganda, or sycophants in thrall to a would-be authoritarian leader. Substantiating the anxieties of many observers, those people may instead be unwilling to abide by democratic rules if it means they have to lose.

Citizens’ political attitudes are more idiosyncratic than elites’. As a result, when citizens engage in issue-specific persuasion, they are more likely to advance arguments that depart from what is typical for their party. Such atypical arguments may differ in the style they adopt or the considerations they emphasize. We report results from a study identifying the persuasive consequences of such atypical arguments and distinguishing content-grounded from expectancy-grounded pathways. First, a content analysis shows that typical and atypical arguments differ in both style and the substantive considerations they raise. Then, in an experiment, we find that atypical arguments from out-partisans are perceived as more compelling, while atypical arguments from in-partisans are perceived as less compelling. Both effects emerge primarily when the author’s partisanship is hidden, pointing to a content-grounded pathway. Citizens with atypical views shape political discussion through how they engage with issues—not merely by violating partisan expectations.

Racial disparities have been documented in many aspects of the death penalty throughout the US. We document the powerful roles of race and racial resentment in explaining who supports and who opposes capital punishment. We review the literature on support for the death penalty, add new analyses at the individual level, and show the consistent and substantial impact of race and racial resentment, which are considerably more important than variables sometimes thought to drive support for the policy, such as religious tradition, ideological orientation, or residing in the South. We also develop a set of yearly state-level estimates of public opinion on the death penalty and racial resentment. Both nationally and at the state level, public opinion, racial resentment, and death sentencing are tightly connected. Thus, we show a powerful level of responsiveness to public opinion, but also that this opinion is tainted by racial animosity toward African-Americans.

Observational research has linked sympathy for the January 6 insurrection to hypercompetitiveness, an understudied dispositional urge to win by any means necessary. Doubt remains, however, about 1) the causal effect of hypercompetitive impulses on electoral democracy and 2) the features of competitiveness more broadly. In this investigation, we demonstrate the causal influence of hypercompetitiveness by randomly assigning a subset of non-Democrats to an apolitical competitive game. Exposure appears to trigger more hypercompetitiveness, which in turn predicts January 6 sympathy. Moreover, as theorized, hypercompetitiveness’ mediated effect is pronounced among those with high levels but absent among those with low levels. As a conceptual validity check, we further demonstrate that a different type of competitiveness that is not associated with a willingness to break rules has no influence. These results enhance causal leverage in support of hypercompetitiveness as a driver of January 6 support, and they invite more scholarship on the politics of personal competitiveness.

In “The Nature of Belief Systems in Mass Publics,” Philip Converse famously provided an account describing the mass public as non-ideological, exhibiting little to no constraint among issue opinions. Still, since the publication of his famous 1964 essay, a great deal has changed. The dominant political cleavages have shifted from economic to social issues. Party cues have largely become much clearer. Partisan and ideological sorting has reached an all-time high. Sources of national political information have proliferated. Has any of this changed the relative constraint of mass belief systems? To shed light on this question, we use ANES data spanning a 32-year time period to conduct a series of network analyses aimed at explicating the structure of the mass public’s belief systems. Using measures of network clustering—network density and average path length—we show that constraint remained relatively stable in the 1990s and 2000s, with a substantial increase in issue constraint occurring in 2012 that persisted through 2024. We subset these analyses by political knowledge and find that this change is largely driven by low-information rather than high-information respondents. By 2024, the differences in constraint across these groups attenuate significantly. Furthermore, using measures of network centrality, we find that racial resentment, authoritarianism, and partisanship are strikingly strong organizers of belief systems. These attitudes’ influence over mass belief systems has markedly increased from 1992 to 2024. We again subset these analyses by respondents’ knowledge and find that high-knowledge respondents are much better at using political concepts such as ideology in structuring political opinions compared to low-knowledge respondents. Additionally, this difference is present for partisanship in 1992 but not in 2024, while racial resentment and authoritarianism maintain relative hegemony by 2024 regardless of information levels.

One of the fundamental problems of survey research is how best to measure latent attitudes. Many common measures are vague, easily primed, or endogenous to outcome variables of interest—or some combination therein. Rather than ask respondents to generalize via abstract measures, many of which they have never encountered or pondered, I introduce a new approach: measuring latent attitudes via orientations toward common aphorisms, e.g., “Don’t judge a book by its cover,” that express a worldview or ethical principle in a familiar yet pithy, axiomatic statement. Such sayings are socialized during the critical period of language development and are building blocks of language akin to idioms. Drawing on meme theory, I argue that aphorisms are readily transmittable forms of cultural information—memes—that are uniquely suited to minimize common survey measurement problems such as endogeneity and ambiguity. I apply this theory to risk attitudes, finding that aphorisms that tap risk acceptance, e.g., “Nothing ventured, nothing gained,” constitute a highly coherent scale, are associated with known correlates of risk acceptance, and predict greater political participation. In doing so, I show that using attitudes toward common aphorisms presents a new method by which survey researchers can tap latent attitudes that are 1) theoretically exogenous to most political outcomes of interest and 2) concrete representations of abstract worldviews in the minds of respondents.