Introduction
With the uncertainty of the COVID-19 pandemic, the murder of unarmed Black people such as George Floyd and Breonna Taylor, and the unprecedented global protests for Black life which followed, the U.S. population experienced a range of emotions, such as fear and anger, preceding a tumultuous 2020 presidential election cycle (Dunn Reference Dunn2020). On the one hand, such emotions may provoke a sense of helplessness. On the other hand, rather than demobilizing the public to act, such emotions may motivate those who believe their actions have an impact, to take action. Few would have predicted the January 6, 2021, Capitol insurrection, just weeks prior to the inauguration of President Joe Biden. While the majority of participants in this violent uprising were white men, the media prominently highlighted the involvement of male conservative activists of color. However, we still know little about the perspectives of men in the general public regarding this event in our nation’s history, particularly across racial/ethnic and other identity groups.
Participants at the January 6th insurrection have been identified as members of The Proud Boys, The Three Percenters, Oath Keepers, QAnon, and other far-right, white supremacist organizations (Washington Post Staff 2024). Academic research has highlighted the importance of racial resentment and negative attitudes toward immigrants in explaining support of the January 6th insurrection (Barreto et al. Reference Barreto, Alegre, Isaiah Bailey, Davis, Ferrer, Nguy, Palmisano and Robertson2023; Davis and Wilson Reference Davis and Wilson2023). However, few studies have focused on examining men of color’s insurrection attitudes relative to their white male counterparts, as the majority of participants were men and news coverage often focused on men’s participation.
Rather than treating all men as a monolith, this study examines: how does racial efficacy and anger about race relations shape men’s attitudes toward the January 6th Capitol insurrection, and how do these effects vary across racial groups? Drawing from established theories of public opinion formation, we consider the roles of partisan attachment, political efficacy, and its extension to racial efficacy (Phoenix and Chan Reference Phoenix and Chan2024) in this relationship. Additionally, we investigate the role of emotions—particularly anger—in shaping male respondents’ perceptions of the Capitol insurrection. We hypothesize that white men who experience less racial efficacy are more supportive of the insurrection and its tactics. Conversely, racial efficacy will not be a predictor of insurrection attitudes for men of color. Moreover, men of color who express more anger about race relations will be less supportive of the insurrection and its tactics.
We utilize the 2020 Collaborative Multiracial Post-Election Survey (CMPS) a cooperative, multiracial, multiethnic, multilingual, post-presidential election online survey in the United States. The 2020 CMPS was the only national, post-election dataset to yield responses on the Capitol Insurrection across a large number of identity groups, such as men of color. In the next section, we examine some theories in the existing literature using the examination of public opinion formation such as political efficacy, racial efficacy, and the role of emotions in political attitudes. This section is followed by our research design and methods used to examine the extent to which racial efficacy and anger influences men of color’s attitudes to the events on January 6, 2021, relative to their white male counterparts. Our findings support the importance of anger, specifically race-related anger, in developing political attitudes on January 6th. In addition, we find that racial efficacy is more relevant for explaining the attitudes of white men compared to other racial groups of men. Our findings underscore the need for an intersectional approach in the study of public opinion formation. The failure to account for factors such as race and gender can lead to an under- or over-estimation of results related to individuals with cross-cutting identities. Using the CMPS provides a large sample size of respondents to allow researchers to disaggregate data to better understand men across race/ethnicity. We now turn to theories of public opinion formation to understand which sociodemographic and political factors may shape public perceptions of January 6, 2021.
Theories of Public Opinion
Previous theories of public opinion studied a variety of relevant factors to explain how people form and uphold political attitudes. Scholars have long studied the influence of factors such as ideology, media consumption, and sociodemographic characteristics on political attitude formation (Campbell et al. Reference Campbell, Converse, Miller and Stokes1960; Hoewe and Peacock Reference Hoewe and Peacock2020; Iyengar and Kinder Reference Iyengar and Kinder1987; Jacoby Reference Jacoby1991; Jost Reference Jost2006). For instance, we might anticipate that people with more conservative ideologies view the events of January 6th not as an insurrection, but as a protest fighting for a fair and free election. In addition to the enduring relevance of these factors, one particularly relevant framework, drawing on theories of partisan attachment and the formation of political attitudes, was the concept of political efficacy. Developed by Campbell et al. (Reference Campbell, Converse, Miller and Stokes1960), political efficacy is defined as “the feeling that individual political action does have, or can have, an impact upon the political process… It is the feeling that political and social change is possible, and that the individual citizen can play a part in bringing about this change” (187). Since its initial conception, political efficacy has been understood as a construct with two dimensions: (1) internal efficacy, an individual’s perception of their own capacity to navigate the political process (Niemi, Craig, and Mattei Reference Niemi, Craig and Mattei1991), and (2) external efficacy, an individual’s perceptions of how responsive members of the political process and political bodies are to the demands of citizens (Balch Reference Balch1974).
Political efficacy is a crucial prerequisite to political participation, as people need to believe that their actions have a capacity to make a change, in order to choose to act (Campbell et al. Reference Campbell, Converse, Miller and Stokes1960; Pranger Reference Pranger1968; Verba and Nie Reference Verba and Nie1972). Political efficacy can also serve as an indicator of how much trust an individual has in the democratic process—if an individual is confident that their political participation within the system has an impact on political outcomes, they are less likely to believe that the outcomes produced by the system are illegitimate or pursue political participation beyond what is sanctioned by the system (Easton and Dennis Reference Easton and Dennis1967; Gamson Reference Gamson1968). Shingles (Reference Shingles1981) highlights the connection between high levels of internal efficacy and low levels of external trust, arguing that this combination can lead to the usage of unconventional methods to influence the policy process. While political attitudes, such as efficacy, are crucial for our understanding of public opinion, we also take into account the role of social identities such as race and gender.
Scholars have demonstrated the existence of a gender gap in political participation in a variety of arenas. While women do vote at higher rates than men, men are more likely to perceive themselves as qualified to run for office (Fox and Lawless Reference Fox and Lawless2010; Lawless and Fox Reference Lawless and Logan Fox2010), engage in political competition (Kanthak and Woon Reference Kanthak and Woon2015; Preece and Stoddard Reference Preece and Stoddard2015), and speak at public meetings (Karpowitz and Mendelberg Reference Karpowitz and Mendelberg2014). Men are also more likely to exhibit greater levels of political efficacy (Preece Reference Preece2016; Wolak Reference Wolak2020). Men are more likely than women to believe that they are competent enough to formulate opinions and participate in politics (Gidengil, Giles, and Thomas Reference Gidengil, Giles and Thomas2008; Thomas Reference Thomas2012). Gendered differences in political participation are theorized to be attributed to fundamental differences in socialization (Preece Reference Preece2016).
Differing racial experiences also illustrate the importance of understanding political attitudes through an intersectional lens and disaggregating analyses by gender and race. In the 2016 election, the voting behavior of women of color significantly differed from their White women counterparts, who were more likely to vote for Donald Trump. These vast differences in levels of women’s support for President Trump only became visible after disaggregating by gender and race (Frasure-Yokley Reference Frasure-Yokley2018; Junn Reference Junn2017; Phillips Reference Phillips2018). The majority of work studying “people of color” as an identity and its relationship to gender primarily focuses on “women of color,” (Bejarano et al. Reference Bejarano, Brown, Allen Gershon and Montoya2020; Carey and Lizotte Reference Carey and Lizotte2023; Greene, Matos, and Sanbonmatsu Reference Greene, Matos and Sanbonmatsu2022; Nguy, Davis, and Chan Reference Nguy, Davis and Chanforthcoming) but recent work illustrates that men of color, specifically Black and Latino men, also hold linked fate attitudes (Matos and Sanbonmatsu Reference Matos and Sanbonmatsu2024).
Racial Efficacy and Political Attitudes
Building upon the widely used concept of “political efficacy,” Phoenix and Chan (Reference Phoenix and Chan2024) proposed a construct they called racial efficacy, defined as “an individual’s belief that their racial in-group possesses sufficient influence over government outcomes” (522). This theory bridges the gap between an individual’s consciousness of their racial identity and how this translates to their decision-making surrounding their political attitudes and participation. Like political efficacy, racial efficacy is measured using an index of questions to assess the extent to which the respondent believes that people elected to office help and make changes for members of the respondent’s racial group, as well as the respondent’s perception of their influence on the government (Phoenix and Chan Reference Phoenix and Chan2024).
Previous work consistently demonstrates that members of racial minority groups, particularly Black Americans and Asian Americans, have lower levels of political efficacy than their white counterparts (Cohen Reference Cohen2012; Form and Huber Reference Form and Huber1971; Hero and Tolbert Reference Hero, Tolbert, G M and Bowler2005; Tate Reference Tate1991). Given the importance of racial experiences in shaping efficacy levels, Phoenix and Chan (Reference Phoenix and Chan2024) hypothesize that individuals who perceive or experience discrimination against their own racial group will experience lower levels of racial efficacy than their counterparts. On the other hand, individuals who feel better about their group’s political incorporation or the state of racial dynamics will express greater levels of racial efficacy.
Applying theories of racial efficacy to attitudes on January 6th by racial groups, we anticipate that all groups of men who experience greater amounts of racial efficacy are less supportive of the insurrection and its tactics. The January 6th insurrection was a form of political participation that was unsanctioned by the political system, and a result of an aggrieved group believing that the outcome generated by the political system (i.e. the election of Joe Biden) was illegitimate. Measures of efficacy are useful norms for determining whether or not someone believes not only in their ability to influence outcomes as an individual, but in the legitimacy of the democratic process in taking these perspectives into account. If people believe the system works for members of their group, then they will not believe an insurrection, which uses methods outside of the formalized system, is necessary.
Participants in the January 6th insurrection believed that their actions were necessary because the outcome generated by the prevailing political system was illegitimate. This belief primarily resulted from disinformation intentionally spread by President Donald Trump and his supporters to claim that significant, unchecked voter fraud resulted in votes being uncounted or fraudulent votes being cast in key swing states, and had this not happened, Trump would have won the 2020 election. These claims included allegations of voting machine tampering, inadequate access to observe ballot counting, the counting of ballots from people who were deceased, ineligible, or non-citizens, and lack of enforcement of voter eligibility requirements (Danforth et al. Reference Danforth, Ginsberg, Griffith, Hoppe, Michael Luttig, McConnell, Olson and Smith2022). These allegations, which, after a series of investigations, were determined to be baseless, resulted in a movement of people who believed that the election outcome was “stolen” from Donald Trump. For Trump’s supporters, to have prevented this outcome, there needed to be stricter voter eligibility checks, such as voter ID laws, and less access to other modes of voting, such as absentee and mail-in voting. While Trump’s rhetoric set the stage for his possible election loss to be perceived as “stolen” in the short-term, in the long-term, he also set the stage to oppose efforts to expand access to the vote, a method that has historically been used to restrict voters from racial minority groups.
Eventual investigations of voter fraud also disproportionately targeted members of marginalized racial groups. In the six states where election integrity units were expanded after Trump’s 2020 election loss, when the race of the defendant was identifiable, 76 percent of those targeted for prosecution were Black or Hispanic, compared to white people making up the other 24 percent (Jouvenal Reference Jouvenal2023). Ellis (Reference Ellis2022) notes that over 86 lawsuits were filed to challenge election practices in Pennsylvania, Michigan, Wisconsin, Arizona, Nevada, Georgia, Minnesota, and New Mexico. These lawsuits “focused principally not on the practices of a particular state, but on the practices in cities or counties that were principally urban or otherwise contained a significant presence of people of color” (462). Through these strategies, Trump and his allies were able to make implicit racial appeals that the election outcome was being stolen, due to the practices of urban, minority-dominated counties. As implicit racial appeals typically target white audiences (Mendelberg Reference Mendelberg2001), it is unlikely they will be effective in motivating minorities’ racial efficacy and subsequently influencing their attitudes toward the insurrection.
Beyond these implicit racial appeals leading up to the insurrection, the January 6th insurrection itself quickly became a racialized event, advocating for the interests of white supremacist groups and actors, against outgroups, typically racial minority groups. Many rioters waved Betsy Ross American flags and Confederate flags, symbolizing support for times in American history where slavery was legal and women and people of color had few legal rights (Washington Post Staff 2024). They used racial slurs while yelling at Capitol police officers, and they brought nooses and gallows to symbolize the historic lynchings of Black people and other people of color throughout American history (Austin-Hillery and Strang Reference Austin-Hillery and Strang2022). Rioters present at the January 6th insurrection have since been identified as members of The Proud Boys, The Three Percenters, Oath Keepers, QAnon, and other far-right, white supremacist organizations (Washington Post Staff 2024).
Recent research has highlighted the importance of racial resentment and negative attitudes toward immigrants in explaining support of the January 6th insurrection (Barreto et al. Reference Barreto, Alegre, Isaiah Bailey, Davis, Ferrer, Nguy, Palmisano and Robertson2023; Davis and Wilson Reference Davis and Wilson2023). White backlash, or resentment from White Americans at perceived gains or advances made by other racial groups, provides one explanation for January 6th participants’ violent reaction to the election outcome (Craig and Richeson Reference Craig and Richeson2017). The outcome of the 2020 election represented a win for Democrats, including the coalition of nonwhite racial minority groups, whose high turnout rates secured the presidency for Biden. Supporters of Trump were also primed to perceive minority political power as a threat—on the campaign trail, Trump and other Republican party elites referenced the “Great Replacement” theory. This conspiracy theory suggests that elites, such as those in the Democratic party, are intentionally trying to reduce the power of the white Christian population by supporting the growth of populations of immigrants and racial minorities (Davey and Ebner Reference Davey and Ebner2019). Given this increased use of “Great Replacement” conspiratorial rhetoric, scholars argue that white supremacist and male supremacist values have fed mass sympathy for the January 6th insurrection (DiMaggio, Wahlrab, and Ochs Reference DiMaggio, Wahlrab and Ochs2024).
While higher levels of political efficacy have been established as a prerequisite to traditional political participation, lower levels of efficacy do not preclude the possibility of people engaging in political activity across the board. Non-traditional political action can be justified by a low sense of efficacy in the existing system when people feel they cannot enact their desired changes through traditional participation methods (van Stekelenburg and Klandermans Reference van Stekelenburg and Klandermans2013). Some scholarship fails to find a relationship between low political efficacy and non-traditional forms of political participation, such as conventional protest participation (Ardèvol-Abreu, Gil de Zúñiga, and Gámez Reference Ardèvol-Abreu, Gil de Zúñiga and Gámez2020). However, distinguishing between conventional and unconventional protest, the latter of which is more demanding, higher risk, and socially illegitimate reveals another dynamic. The relationship between efficacy and conventional protest participation appears similar to the relationship between efficacy and electoral participation where high levels of efficacy correlates with high levels of conventional protest participation (DiGrazia Reference DiGrazia2014; Tausch et al. Reference Tausch, Becker, Spears, Christ, Saab, Singh and Siddiqui2011). Unconventional, higher-risk protest participation, on the other hand, is related to greater feelings of alienation from the political system and lower efficacy (DiGrazia Reference DiGrazia2014; Tausch et al. Reference Tausch, Becker, Spears, Christ, Saab, Singh and Siddiqui2011). The January 6th insurrection was an unsanctioned, risky, and violent action that differs from the conventional repertoire of peaceful protest in the United States. In the case of the insurrection, if white men do not believe the political system works for their racial group, then they may believe unsanctioned and risky political action, like an insurrection, is necessary to further their group interests. This leads to our first hypothesis,
H1: White men who experience less racial efficacy are more supportive of the insurrection and its tactics.
In contrast, considering the political attitudes of men of color and the racialized motivations of the events on January 6th, we argue racial efficacy will not be a predictor of men of color’s political attitudes toward the events on January 6th. Because the January 6th insurrection relied on implicit racial appeals and depicted explicit white supremacist activism, men of color who experience less racial efficacy, doubting the system works for them, likely see the insurrection’s existence and lack of immediate shutdown as yet another example of the system failing them. But as mentioned previously, if men of color do experience greater levels of racial efficacy, they are also unlikely to support the insurrection, because it is an unsanctioned form of political participation that calls the legitimacy of the political system into question. Political participation outside of the system would be unnecessary in a system capable of using conventional methods to meet the needs of the group. We argue that racial efficacy is not a sufficient explanatory factor for whether men of color support the insurrection.
H2: Racial efficacy will not be a predictor of insurrection attitudes for people of color.
Anger and Political Attitudes
Another relevant body of work in the study of public opinion and political behavior is the study of emotions and politics. While earlier work on emotions and politics focused on emotions as a binary, such as positive or negative affect (Watson and Tellegen Reference Watson and Tellegen1985), contemporary work has a more nuanced understanding of affect. Emotions and politics scholars developed significant theoretical frameworks for exploring how emotions impact political behavior and public opinion formation. A great deal of this work on emotions comes from affective intelligence theory (AIT), which theorizes how specific emotions have distinct behavioral and attitudinal responses (Marcus, Neuman, and MacKuen Reference Marcus, Russell Neuman and MacKuen2000). This framework draws from neuropsychological work (Adolphs, Tranel, and Damasio Reference Adolphs, Tranel and Damasio1998; LeDoux Reference LeDoux1992) and applies its findings to the study of politics. The theory of affective intelligence identifies two main preconscious ways through which individuals process emotions, typically noted as the disposition and the surveillance system. According to the theory, people have preconscious appraisals to their environment, which triggers a set of responses. These responses are based on the type of emotion that is felt and which cognitive processing system it triggers.
Emotions like anxiety or fear cause a break in relying on preexisting beliefs and ideas. Rather, fear and anxiety trigger the use of the surveillance system which causes individuals to seek out information and find solutions (Marcus et al. Reference Marcus, Valentino, Vasilopoulos and Foucault2019). Unlike anxiety or fear, anger acts as a catalyst in individuals becoming more rooted in their existing beliefs and opinions, rather than propelling them to seek out or take in new information (Marcus et al. Reference Marcus, Russell Neuman and MacKuen2000).
Importantly, anger triggers the use of the disposition system which can cause people to continue to engage with the same stimuli that causes the anger. That is, when people are angry, they are more likely to be confrontational, punitive, and engaging (Zeitzoff Reference Zeitzoff2014). For example, Ryan (Reference Ryan2012) finds that when invoking anger in people through political ads and campaign websites, individuals actually increase their interaction with the political website and are more likely to express how they feel. Further, anger in the face of reactionary political threats—such as the political presence and rhetoric of Trump in 2016—causes Black Americans to mobilize at higher rates (Towler and Parker Reference Towler and Parker2018).
Scholars additionally explore how anger impacts political attitudes (Valentino et al. Reference Valentino, Brader, Groenendyk, Gregorowicz and Hutchings2011; Kushner Gadarian and Brader Reference Kushner Gadarian, Brader, Huddy, Sears, Levy and Jerit2023). Given that anger impacts people’s threat sensitivity, it ultimately leads them to be more supportive of punitive policies and critical of political and civic threats (Huddy et al. Reference Huddy, Smirnov, Snider and Perliger2021; Wayne Reference Wayne2023). For example, Fisk, Merolla, and Ramos (Reference Fisk, Merolla and Ramos2019) find that exposure to terrorist threats causes people who express anger to be more supportive of drone strikes. That is, when threats are looming, the angrier people are the more they support protective policies that not only guarantee their safety but also punish the perpetrators. Unlike anxiety, which causes people to also be supportive of protective policies (Albertson and Gadarian Reference Albertson and Kushner Gadarian2015), anger leads people to be critical and in support of vengeful policies.
Part of what is missing from much of the literature on emotions and politics is the study of how emotions differentially impact racial and ethnic minority groups. Phoenix’s (Reference Phoenix2019) work on anger and Black Americans created a framework for understanding how emotions, particularly anger, manifest differently for racial and ethnic minority groups. While anger typically has strong and mobilizing effects for the general population, for Black Americans, anger actually translates to motivations for collective action outside of the existing political system. By focusing on anger, he not only finds that anger is cognitively processed differently by Black and White Americans, but that anger has distinct mobilizing effects for each particular group. For Black Americans, anger does not have an electorally mobilizing political effect, but anger does have a positive impact on Black Americans’ engagement in organizing and protesting. Not only do racial and ethnic identities impact emotions and cognition but they also influence how these emotions are translated to political behavior and attitudes.
In sum, while anger leads White Americans to vote more, anger actually leads Black Americans to engage at higher levels in protests and other forms of collective action. Additional work on Black politics and anger finds that feeling angry about race increases Black Americans: willingness to donate to Black organizations, participate in protests, and support Black community nationalism (Banks, White, and McKenzie Reference Banks, White and McKenzie2019). Despite these essential contributions to our understanding of the racial implications of anger in politics, we know little about how these dynamics factor into attitude formation toward radical political events such as the January 6th insurrection.
Given what we know about anger and opinion formation, we focus on how anger about race relations impacts men of color’s attitudes about the January 6th insurrection. Understanding the degree of violence present during the insurrection, it can be expected that anger plays a role in the opinions people of color form about an event that poses a threat to their security. Further, we expect men of color, particularly Black men, to express higher levels of anger about race relations (Banks et al. Reference Banks, White and McKenzie2019).
Particularly, drawing from the literature on race, ethnicity, and anger, specific sources of anger create a clearer pathway for how negative emotions translate to opinion formation. Work studying the impact of anger toward racism and/or race relations finds that feeling this source of anger impacts how individuals perceive a specific policy (Banks Reference Banks2014). Given that the insurrection, as we previously discussed, is framed and appraised as a racialized issue, we can expect that anger about race relations will have a direct impact on attitudes about the insurrection for men of color. Specifically we hypothesize,
H3: Men of color who express more anger about race relations will be less supportive of the insurrection and its tactics.
Conversely, an event such as the January 6th insurrection can have distinct attitudinal effects on White men. Given that this was an event that was racially charged, the relationship between anger about race relations and the insurrection might translate differently for White men. While we know that anger causes individuals to be more critical of threatening events and supportive of vengeful policies, it would be equally necessary for White men to think of the insurrection as a threat to their own safety and security. Given the potential benefits a Trump presidency could offer White men, we do not anticipate that White men perceive the insurrection as a threat to their own wellbeing. When it comes to White Americans, race is also an important factor that impacts how anger translates to political behavior and public opinion. However, the underlying substrates of the anger for White Americans are significantly different. Unlike men of color, for whom an event like the insurrection symbolizes a real threat to their security, White men likely appraise this event differently.
Work on white grievance finds that Whites who feel as though they are the targets of discrimination are more likely to be skeptical about elections (Filindra, Kaplan, and Manning Reference Filindra, Kaplan and Manning2024). While the insurrection can be perceived by men of color as a threat to their safety as racial minorities, for White men, the insurrection could be perceived as a defense of their status. The existing work on white grievance informs our understanding of how Whites can be more likely to be skeptical of election outcomes when they feel as though their racial group is being discriminated against. For White Americans, anger related to race is more complicated to disaggregate. In fact, some White Americans, particularly white men, have reported feeling that racial progress has led to them (whites) being the primary targets of discrimination. Furthermore, previous work on White youth finds that those who experience higher levels of anti-white bias are more likely to engage in serious and violent offenses (Isom Scott and Stevens Andersen 2020).
Despite this, we do not expect that White men who express higher levels of anger about race relations will be likely to support the insurrection and its tactics. The January 6th insurrection while supporting the interests of aggrieved Whites is likely too disruptive an action to engender broad support from Whites. White men, holding an advantaged position in society, are less likely to respond favorably to such disruptive protests especially in uncertain times (Jost et al. Reference Jost, Chaikalis-Petritsis, Abrams, Sidanius, van der Toorn and Bratt2012). This leads us to assume it is unlikely that anger about race relations has a predictive relationship toward attitudes about the insurrection for White men.
H4: Anger about race relations will not be a predictor of insurrection attitudes for White men.
Data/Methods
To test our hypotheses, we utilize the 2020 Collaborative Multiracial Post-Election Survey (CMPS), which provides a unique oversample of African American, Latina/o, and Asian American respondents that is also large enough to disaggregate across both race and gender (Frasure et al. Reference Frasure, Wong, Vargas and Barreto2024). The CMPS was distributed online nationally between April 2, 2021, and August 25, 2021. Within each racial group, the data were weighted to fall within the margin of error of the adult population in the 2019 Census ACS 1-year data file for age, gender, education, nativity, and ancestry.
The 2020 CMPS is an ideal dataset for testing our hypotheses because it is the first large-scale national survey that oversamples both gender and racial minority groups, allowing us to disaggregate across both race and gender. Since the focus of our paper is to examine how predictors of January 6th support vary across men of different racial groups, we filter our analysis to a robust sample of 1,918 Black men, 1,531 Latinos, 1,667 Asian men, and 1,740 white men, with 6,856 men in total. The CMPS also includes questions regarding anger and racial efficacy, our independent variables of interest. While past studies on anger in politics focus on measuring the general feeling of anger, the CMPS contains questions that allow us to not only measure anger but also to measure what respondents are attributing their anger toward. We utilize two questions to measure anger attributed to race: “I am angry that racism exists,” with responses ranging from strongly agree to strongly disagree, and a question asking respondents’ level of anger toward race relations in the past year, with responses ranging from much less angry to much more angry. The responses to these questions were scaled to create a singular anger about race index between 0 and 1 (Cronbach’s alpha = .749). Following the creation of a validated measure from Phoenix and Chan (Reference Phoenix and Chan2024), we use responses to the following questions to create an index measure of racial efficacy (Cronbach’s alpha = .891), scaled between 0 and 1: “How often would you say public officials work hard to help [R’s racial group] people?” and “How often would you say [R’s racial group] people have a say in how the government handles important issues?”.
Our dependent variable is attitudes on the January 6th insurrection. The first question we use to measure January 6th attitudes asks how respondents felt about the people who attacked the Capitol. Respondents could choose between three response choices: “Many were white supremacists and racism was an underlying factor in their actions,” “Some may have been white supremacists, but they were mostly upset about Biden winning,” and “They were not white supremacists, racism had nothing to do with their actions.” We code those who chose the response that insurrectionists were white supremacists with racism as an underlying factor of their actions as a 1, and the other two choices as 0. As a robustness check, we also ran a model where the dependent variable was not dichotomously coded, yielding substantively the same results. We report these findings in the appendix.
The second question asks how respondents categorized the event that took place at the Capitol, choosing between two response options: “Mostly a protest that went too far” or “A coordinated act of insurrection against the United States.” We code those who chose to describe the event as an insurrection, and not a protest, as 1, and those who describe the event as a protest as 0. Given the dichotomous nature of our dependent variables, our analysis includes descriptive statistics and logit regression analysis.
Turning to covariates, we also control for well-established indicators of public opinion. We include standard controls for demographic variables such as age, education, foreign-born status, and religious attendance. We also include political variables, such as conservative ideology, political interest, trust in the government, and protest attendance. We suspect that choice of news also may impact public opinion, so we include Fox News viewership in the model. Discrimination and identity-related variables, such as linked fate, have also been shown to affect public opinion across minority groups (Cortland et al. Reference Cortland, Craig, Shapiro, Richeson, Neel and Goldstein2017; Dawson Reference Dawson1994; Sanchez and Masuoka Reference Sanchez and Masuoka2010). Thus, we include experienced discrimination and linked fate measures in our model specifications. In addition, we also control for traditional measures of external political efficacy (Campbell and Converse Reference Campbell and Converse1972; Verba and Nie Reference Verba and Nie1972). We do not control for partisanship in our models because we understand that when studying Black individuals there is often little variation in partisanship. Instead, we control for conservative ideology as a theoretical construct to capture existing political predispositions. As a robustness check, we include partisanship in the models and find no substantive difference in results. Models with partisanship included are shown in the Appendix Tables 4 and 5. Lastly, we suspect that anger attributed toward race relations will be more likely to affect attitudes toward the insurrection rather than other forms of anger. Because the CMPS contains questions that allow us to measure respondents’ attribution of anger toward the economy as well as race, we create an index measure that scales responses toward two questions that measure the attribution of anger toward the economy. All question wording is available in the Appendix.
Results
We begin our analysis by observing the descriptive statistics of men in the 2020 CMPS across our dependent variables of interest. Fig. 1 depicts the opinions of men on those who participated in the events that took place on January 6, 2021. We report that 50% of Black men identify January 6th participants as white supremacists, which is the highest identification of their male counterparts. Latino and Asian men identify January 6th participants as white supremacists, similar to all men in the sample at about 40%. White men had the lowest identification with that category at 26%. Over 40% of all men identified with the category that stated some January 6th participants were white supremacists, but many were upset about Biden winning. On the other hand, white men had the highest level of identification with the statement that identifies January 6th participants as not white supremacists at 31%, with Latino men and Asian men identifying similarly to all men (17%). Black men had the lowest level of identification with that category with only 10% identifying with January 6th participants not being white supremacists.
Opinions on Jan 6 participants among men in 2020 CMPS. n = 6,856 men; n = 1,667 Asian men, n = 1,918 Black men, n = 1,531 Latino men, n = 1,740 White men.
Fig. 2 describes how different racial groups of men describe the events of January 6th. The distribution of responses follows a similar pattern as opinions on the participants. Black men are most likely to identify January 6th being a coordinated act of insurrection against the United States at 63% followed by Asian men. Interestingly, Latino men are divided in half regarding the event with 50% identifying it as an act of insurrection and 50% identifying it as a protest that went too far. White men are the least likely to identify January 6th as being a coordinated act of insurrection against the United States at 41%. White men are also the most likely out of all racial groups to instead describe the event as a protest that went too far at 59%, fourteen percentage points higher than all men in the sample. On the other hand, Black men’s identification with January 6th being an insurrection is only eight percentage points higher than all men, which illustrates significant distinctions in public opinion on January 6th among different racial groups of men.
Opinions of Jan 6 among different racial groups of men in 2020 CMPS. n = 6,856 men; n = 1,667 Asian men, n = 1,918 Black men, n = 1,531 Latino men, n = 1,740 White men.
To directly test our hypothesis, we run the model with covariates specified in the methods section for five groups: the all-male sample, Black men, Latino men, Asian men, and white men. We first focus on the relationship between racial efficacy and opinions on January 6th, as illustrated by the predicted probability graphs shown in Fig. 3.
Predicted probabilities of attitudes towards Jan 6—racial efficacy.
We find support for H1—only for White men is racial efficacy positive and statistically significant in predicting negative attitudes toward the insurrection. From lowest to highest levels of racial efficacy among white men, the propensity to evaluate the participants of January 6th as White supremacists drastically increases from 10.9% to 37.7%. In addition, when we consider how White men evaluate the insurrection as a protest versus an insurrection shown in Fig. 4, racial efficacy is also positive and statistically significant, increasing the propensity to call the insurrection from 21.4% to 65.5%. This significant rise supports our first hypothesis, suggesting that as white men’s sense of racial group efficacy increases, their likelihood of holding unfavorable attitudes toward the insurrectionists and the event itself increases notably.
Predicted probabilities of attitudes towards Jan 6—anger.
We report notable differences across racial groups shown in Fig. 3. Our findings mostly support H2: for all men, Black men, and Latino men models, the relationship between racial efficacy and attitudes toward insurrectionists is insignificant and null. While we expected racial efficacy to have no relationship with men of color’s attitudes, we find an exception in Asian men, for which we observe an interesting negative statistically significant relationship. As racial efficacy increases, the probability of holding negative attitudes toward the insurrectionists decreases from 43.9% to 29.1%, and of calling the event an insurrection as opposed to a protest from 62.7% to 44.34%. This negative correlation may reflect a different dynamic within the Asian American community; we further consider such implications in our discussion section.
Overall, these patterns underscore the nuanced ways in which racial identity and perceived racial efficacy intersect with political attitudes, with racial efficacy having a pronounced and statistically significant impact among white and Asian men, in opposite directions. For White men, higher racial efficacy significantly increases unfavorable attitudes toward the insurrectionists and calling January 6th an insurrection, while for Asian men, higher racial efficacy decreases such unfavorable attitudes.
To continue our analysis, we run a model to test the relationship between anger about race and attitudes toward the events of January 6th. Fig. 4 illustrates the predicted probabilities, showing the effect of race-related anger on these attitudes. In line with H3, we find that men of color who express greater anger about race relations are less supportive of the insurrection and its associated tactics. Specifically, anger related to race significantly and positively predicts stronger unfavorable opinions toward those who participated in January 6th, as well as an increased likelihood of labeling the event as an insurrection rather than a protest among Black, Latino, and Asian men. These trends indicate that racialized emotions strongly influence political attitudes and perceptions of such critical events in men of color.
Contrary to our fourth hypothesis, we find that race-related anger also significantly predicts unfavorable opinions toward January 6th participants and increases the likelihood of labeling the event as an insurrection among White men, albeit less strongly. For White men, the baseline probability of unfavorable views toward the insurrectionists is 15.3% at the lowest level of race-related anger, rising to 30.2% at the highest level—an increase of 14.9 percentage points. Similarly, the probability of viewing January 6th as an insurrection increases from 29.5% to 50% as race-related anger escalates. These findings suggest that while racialized anger influences white men’s attitudes, its impact is more moderate compared to men of color.
Discussion
In this study, we sought to understand the distinctions in perceptions of the insurrection between white men and men of color. We find that White men with a strong sense of racial efficacy demonstrate negative attitudes toward the insurrection, supporting H1 wherein we anticipated a negative relationship between racial efficacy and support for the insurrection. We find partial support for H2 with racial efficacy insignificant for Black and Latino men’s perceptions of the insurrection; however, Asian men high in racial efficacy were significantly less likely to feel negatively about the insurrection. Men across racial minority groups with a strong sense of anger about race relations were less likely to support the insurrection as we anticipated in H3. Finally, White men were also less likely to support the insurrection and its tactics when demonstrating racial anger in contrast to our H4 expectations. In the following section, we explore explanations for these findings and how the implications such results hold for the discipline.
While our findings for H2 are generally in alignment with our theoretical expectations, Asian Americans who have low levels of racial efficacy are more likely to perceive the insurrection negatively, unlike men of other racial groups. We might attribute this to Asian men with low racial efficacy being demoralized about the feasibility of enacting large-scale change in the political system. Scholars find group efficacy can be associated with mobilization efforts, particularly for minority groups. Marginalized people with little hope are less likely to feel this sense of racial efficacy (Hasan-Aslih et al. Reference Hasan-Aslih, Shuman, Goldenberg, Pliskin, van Zomeren and Halperin2020). As people lose hope and their efficacy weakens they are less likely to engage in collective action (Cohen-Chen and Van Zomeren Reference Cohen-Chen and Van Zomeren2018) this disposition may extend to their attitudes worsening their perceptions of such efforts. In this case, Asian men low in racial efficacy are less likely to engage with insurrection or perceive it positively as they lack the hope or a sense of belief in such actions actually yielding systemic change.
Additionally, we might attribute Asian American men’s low racial efficacy being correlated with decreased support for the insurrection as a by-product of assimilation processes. As Asian communities assimilate into American life, they might develop preferences for social systems to remain consistent. The endorsement of assimilation ideologies among minorities is associated with increased system-justifying tendencies (Shockley, Wynn, and Ashburn-Nardo Reference Shockley, Wynn and Ashburn-Nardo2016). These processes lead to distinct outcomes for Asian American men as Asian American assimilation is distinct with notable gains in educational and socio-economic outcomes yet still vulnerable to xenophobia as exemplified during the COVID-19 pandemic (Lee and Sheng Reference Lee and Sheng2024; Reny and Barreto Reference Reny and Barreto2022). This distinct position may encourage a more cautious political leaning that perceives the insurrection negatively as Asian American men, while less privileged than White men, occupy a high social standing relative to Latino and Black men (Kim Reference Kim1999; Zou and Cheryan Reference Zou and Cheryan2017) that upheaving the system may disrupt. Furthermore, assimilation processes can undermine racial solidarity bringing groups together under a singular identity and weaken groups’ collective action tendencies (Ufkes et al. Reference Ufkes, Calcagno, Glasford and Dovidio2016). In this context, such attitudes may extend beyond decreased likelihood to participate into negative perceptions of the participants and effort.
While we anticipated this relationship to be statistically insignificant, we find White men high in racial anger are less likely to express positive attitudes toward the insurrection. Anger tends to function differently across racial groups with White people finding it mobilizing whereas Black people do not (Phoenix Reference Phoenix2019). However, this scholarship focuses on anger broadly rather than anger toward specific issues. Herein it seems that racial anger functions similarly across racial groups, likely due to a strong sense of egalitarianism. Scholars find that egalitarianism can engender empathy for minority group members (Lucas and Kteily Reference Lucas and Kteily2018) as demonstrated by this sense of racial anger. Egalitarianism is correlated with increased engagement with collective action efforts against inequality even among advantaged group members such as White men (Ho and Kteily Reference Ho and Kteily2020). We might anticipate that such egalitarian beliefs would be negatively associated with the perceptions of the insurrection, a collective action effort meant to preserve inequalities and undermine the democratic process.
Racial anger is a significant influence on White men’s perceptions of the insurrection. Though this is counter to our expectations, racial anger across race may be driven by different motivations that lead to similar results. Chudy describes a theory of racial sympathy wherein White Americans feeling distressed over Black misfortune expressed increased support for racially progressive policies (Chudy Reference Chudy2024). As White people express distress at Black misfortune, we speculate they are more likely to express a strong sense of anger about the state of race relations. Subsequently, such attitudes encourage racialized perceptions of the insurrection, Whites aware of White privilege are more likely to recognize symbols of racism and oppose them (Rios, Mischkowski, and Stephenson Reference Rios, Mischkowski and Stephenson2022). As such symbols were displayed throughout the insurrection with leaders of White supremacist groups in attendance, we might anticipate that racially sympathetic White men are more likely to feel racial anger and perceive the insurrection as a White supremacist event. Future scholars might explore the role of racial sympathy and awareness of White privilege in developing racial anger and its political consequences.
Conclusion
While we find support for some of our theoretical expectations, we recognize limitations future scholars might address, the first being our measurement of attitudes toward the insurrection. Rather than assessing how men across race participated in the events of January 6, 2021, we instead sought to understand public perception of the events and its participants. This was in part driven by concerns of social desirability as participants hesitate to express positive attitudes toward the insurrection. These concerns ring especially true for ascertaining participation or willingness to participate given ongoing police investigations and arrests during data collection. However, this conservative measurement approach limits our understanding of any positive sentiments or willingness to participate in the events.
Through utilizing an observational survey method to assess these attitudes, we find evidence for a relationship between racial anger, racial efficacy, and insurrection attitudes; however, the causal direction of this relationship is unclear. Future scholarship may consider the causal direction between anger and radical political events through an experimental approach, fielding a survey priming anger and assessing attitudes toward contemporary radical political events.
Additionally, we note that this measure of anger focuses specifically on race rather than anger more broadly. Though this attributional measurement better equipped our study to understand the underlying mechanism behind participants’ anger, there are consequences to this approach. As discussed earlier, racial anger may be perceived differently across racial groups. A more generalized measure of anger might allow for more consistency between racial groups’ understanding of the question. Future scholars might examine racial and gender differences in perceiving radical political events through both attributional and generalized anger measures. These findings offer a variety of directions for future scholars to continue this academic inquiry; we observe a few of these directions below.
In this study, we ascertain men of color’s insurrection attitudes relative to their white male counterparts as the majority of participants were men (Matfess and Margolin Reference Matfess and Margolin2022) and news coverage often focused on men’s participation. However, women did play a role in the insurrection, with over a hundred women facing charges for their participation (Matfess and Margolin Reference Matfess and Margolin2022). Comparing men’s attitudes toward the insurrection with their female counterparts might elucidate the role of gender and race in perceptions of the insurrection. Especially upon accounting for the results of Asian men in the racial efficacy hypothesis, it would be fascinating to explore whether or not this effect translates to Asian women’s perspectives on the insurrection.
Furthermore, these explorations could serve as an opportunity to explore the gendered and racial differences in anger’s effect on political attitudes. Following this line of inquiry, future scholars might compare women of color’s attitudes with white women’s on the insurrection. White women carry an important position in the American far-right movement (McRae Reference McRae2018), and as earlier mentioned, had a notable presence at the insurrection. Scholarship comparing white women and women of color’s support for Trump finds white women tend to be more likely to favor Trump than their women of color counterparts (Frasure Reference Frasure2018). Recognizing the prominent role white women hold in conservative movements, examining distinctions in the role of anger across women of differing racial groups might be enlightening as to what motivates white women’s support toward far-right political activism.
Finally, future research should consider the role of anger in political participation beyond electoral engagement. Prior scholarship on anger in politics largely focuses on anger’s influence in voting (Phoenix Reference Phoenix2019; Valentino et al. Reference Valentino, Brader, Groenendyk, Gregorowicz and Hutchings2011). However, this study demonstrates that anger plays a prominent role in understanding the public response to radical activism. Future scholarship might directly measure the role of anger in collective action tendencies given the generally low levels of protest participation in the United States. This exploration is even more crucial for understanding the effectiveness of racial efficacy in influencing non-electoral political participation as little is known about the effect of racial efficacy on participation generally.
While there are various means of further exploration, this study offers crucial insights in assessing political attitudes among Americans. These findings deepen the discipline’s understanding of race and gender distinctions in public perception of the January 6th insurrection. Furthermore, this research offers insights into public opinion formation in radical political events and their racial implications. We apply political psychology’s exploration of emotions, pivoting from assessing voting behavior toward opinion formation. Additionally, we take an attributional approach to anger, garnering a better understanding of factors motivating anger among men, which is a contribution particularly crucial for understudied populations such as men of color.
Prior studies focus on anger’s influence among samples with a majority of one single racial group, whereas we investigate its influence across a rich, multiracial sample. We leverage the substantial sample sizes of the CMPS to uncover nuanced differences in opinion formation across racial groups among men. This innovation is crucial, as the majority of scholarship on the events on January 6th focus on attitudinal shifts by partisan ties (Anderson and Coduto Reference Anderson and Coduto2024; Jost, Goya-Tocchetto, and Kay Reference Jost, Goya-Tocchetto and Kay2023) rather than exploring the importance of social identities. Furthermore, our adoption of a racial and gendered analysis furthers political science’s exploration of the influence of racial efficacy on political attitudes across racial groups.
Political scientists’ study of group efficacy often focuses on political efficacy rather than alternative motivations for efficacious attitudes. Recent scholarship recognizes the importance of racial efficacy in opinion formation but our innovation applies this mechanism in the context of contemporary political events, rather than traditional measures of political attitudes. Our focus on political events demonstrates the versatility of the measure and its broader applications in contemporary politics. Through drawing on our findings, future scholars can explore factors shaping opinion formation of political events, the role of racial efficacy in opinion formation, and distinctions in political attitudes across race and gender.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/rep.2025.8
Funding statement
This research did not receive any specific financial support.
Competing interests
None.
