White Americans who perceive themselves to be “last place” in the racial status hierarchy are most drawn to alt-right extremism

The study examines how rising economic inequality in the U.S., which is also racialized, shapes white Americans’ subjective status perceptions and support for alt-right extremism. Although white Americans, on average, possess greater wealth than racially minoritized groups, stereotypes that whites are wealthy and high status may prompt some white individuals to compare themselves upward to their racial in-group and feel they are falling behind. If these individuals also believe they are being passed in status by People of Color, such perceived intergroup status threat may increase the appeal of ideologies and political movements that purport to benefit white people. The authors hypothesize that a measurable “Last Place” subjective status profile among non-Hispanic white Americans—characterized by feeling behind most white people and close to or already passed by racial outgroups—will be associated with heightened support for alt-right events, candidates, and ideologies.

The paper situates the research within evidence of widening U.S. economic inequality and increasing political polarization. Economic inequality is racialized, with persistent and growing wealth gaps (e.g., the median white family has far more wealth than the median Black family). Stereotypes link whiteness with wealth and Blackness with poverty, and broader perceived racial economic hierarchies typically place white Americans at the top, followed by Asian Americans, Hispanic/Latinx Americans, and Black Americans. Despite objective advantages, white Americans tend to underestimate racial wealth gaps and increasingly report feeling left behind and less economically optimistic. Group position and social identity theories suggest that perceived threats to group status, especially amid salient race/class hierarchies, can motivate backlash and support for exclusionary or extremist politics. Prior work connects such despondency among white Americans to poorer health outcomes and to political radicalization under perceived status threat, setting the stage for the current investigation of subjective status profiles and alt-right support.

Design: Two studies using large, quota-based samples of non-Hispanic white Americans residing in the U.S. Pilot study focused on piloting measures and latent profile structure; Study 1 tested associations between latent profiles and alt-right outcomes. Ethics and preregistration: Both studies approved by Colgate University IRB (ID: ER-22-40). Preregistered at https://aspredicted.org/LXP_198 (with later simplification of hypotheses; moderation by white identification reported in Supplementary Table 6 and was not significant). Participants born outside the U.S. were excluded (Pilot: N = 14; Study 1: N = 54), a criterion decided prior to analyzing Pilot data but omitted from preregistration due to oversight. Sampling and participants: Pilot targeted approximately N ≈ 500 non-Hispanic white Americans without a college degree via Lucid Panels, with quotas for gender, age, and region; abstract reports Pilot N = 465. Study 1 recruited N = 1,516 via CloudResearch Managed Research with census-based quotas on education (including oversampling those without college degrees), region, gender, and age; after exclusions and handling missing status data via listwise deletion (N = 26), final N = 1,449 (M age = 50.47, SD = 17.19; 732 women, 701 men, 16 non-binary; median education: some college/associate; median income: $40,000–59,999; political IDs: 34.6% Republican, 32% Independent, 5.4% Other). Some quotas were opened late in sampling, slightly skewing regional representation toward Midwest and Northeast. Mobile device users were excluded due to measure incompatibility. Measures: Participants completed demographics and checks, then measures of social solidarity (e.g., bond with white Americans), perceived social class, social class identity solidarity, and self–other overlap with racial and class groups. The key subjective status measure asked participants to place themselves and racial groups (white, Black, Asian, Hispanic/Latinx) along a status continuum reflecting money, political power, education, and job respect. Outcomes included: (a) support/feelings toward 11 right-wing events/groups/movements (e.g., Jan 6, Proud Boys, Oath Keepers, #AllLivesMatter; 1–7 scales; α = .95), and 8 left-wing events/groups/movements (to mask hypotheses; α = .89); (b) support for unnamed right-wing and left-wing modeled political candidates (1–7); (c) two measures of alt-right ideology: a 17-item scale developed by the authors (subdomains: perceived anti-white bias, anti-government beliefs, violent anti-government beliefs, anti-immigrant beliefs; α = .93) and an adapted 5-item right-wing extremism scale (Kamenetski et al., 2021; α = .86). Exploratory measures included SDO, RWA, motivations to respond without prejudice (internal/external), belonging and identification with race/class groups, feelings of being left behind by the U.S. government, and additional demographics. Analytic strategy: Latent Profile Analysis (LPA) in Mplus 8.10 (Mixture Add-on), using participants’ placements of self, white people, Black people, Hispanic/Latinx people, and Asian people as indicators. Covariances among indicators were fixed to 0 within profiles; maximum likelihood with robust SEs was used. Competing models from 1–5 profiles were compared using log-likelihood, AIC, BIC, SABIC, entropy, Lo-Mendell Rubin test (LMR), and bootstrapped likelihood ratio test (BLRT), also considering theory fit and avoiding profiles with <10% of the sample. Study 1 additionally examined associations between profile membership and outcomes via MANOVA/ANOVAs, controlling for income, education, age, and gender; robustness checks added social and economic conservatism as controls or removed controls altogether. Entropy for the 3-profile solution in Study 1 was 0.83, supporting subsequent use of assigned profile membership. Power: Pilot aimed for ≈500 participants based on LPA standards. Study 1 used G*Power for a priori power for linear regression assuming a medium-small effect (f² = 0.30), 80% power; target N = 967; recruited N = 1,500+ to accommodate attrition and attention-check failures.

LPA solutions: Both Pilot and Study 1 converged on a 3-profile solution with favorable fit and theory coherence. Profiles were: - Second Place: self below white people but above other racial groups; awareness of racial economic inequality. Proportion: 28% (Pilot), 22% (Study 1). - Third Place: pattern reflecting stereotypes linking white/wealth and Black/Latinx/poverty; self behind groups stereotyped as poorer but ahead of others in Pilot; majority profile. Proportion: 62% (Pilot), 66% (Study 1). - Last Place: self far behind all other racial groups, ranking all groups toward the top of the status hierarchy; theorized to be most associated with alt-right support. Proportion: 11% (Pilot), 12% (Study 1). Associations with alt-right outcomes (Study 1): Profile membership significantly predicted outcomes overall (MANOVA: F(8, 281) = 5.18, p < .001, partial η² = .02). Univariate ANOVAs (Table 5): - Alt-right candidate support: F(2, 1408) = 7.73, p < .001, partial η² = .011. - Alt-right events/groups/movements support: F(2, 1408) = 16.19, p < .001, partial η² = .022. - Alt-right ideology (authors’ scale): F(2, 1408) = 10.78, p < .001, partial η² = .015. - Alt-right ideology (Kamenetski et al., 2021): F(2, 1408) = 11.76, p < .001, partial η² = .016. Estimated marginal means (Table 6) showed the “Last Place” profile highest on all alt-right outcomes. Examples: - Alt-right candidate support: Second Place M = 3.84 (0.13), Third Place M = 4.11 (0.07), Last Place M = 6.70 (0.17). - Alt-right events support: Second Place M = 2.77 (0.09), Third Place M = 3.01 (0.05), Last Place M = 3.60 (0.12). - Alt-right ideology (authors’): Second Place M = 3.62 (0.08), Third Place M = 3.68 (0.05), Last Place M = 4.21 (0.11). - Alt-right ideology (Kamenetski): Second Place M = 3.05 (0.08), Third Place M = 3.11 (0.05), Last Place M = 3.67 (0.11). Individual differences (Tables 7–8): Profiles differed on education, income, perceived social class, feelings of being left behind, conservatism, SDO, white identification and belonging, and class identification/belonging (all p < .05). Notably, the “Last Place” profile showed the highest white identification and belonging and higher SDO and economic conservatism. Importantly, the Last Place effect was not reducible to objective status: participants in the Second Place profile had the lowest income and education, yet were not highest on alt-right support.

Findings support the hypothesis that a subset of white Americans who perceive themselves as falling behind both their racial in-group and racial outgroups—the “Last Place” subjective status profile—are especially supportive of alt-right events, ideologies, and candidates. This suggests that upward in-group comparisons combined with perceived intergroup status threat can motivate alignment with extremist right-wing politics that promise to restore or protect white group advantage. The observed higher white identification, white belonging, and SDO within the Last Place profile are consistent with theories that heightened group identification and preference for group-based hierarchies can intensify reactions to perceived status threats. The fact that Second Place participants had the lowest objective status but did not show the highest alt-right support underscores the distinct contribution of subjective, relational status appraisals—beyond absolute socioeconomic position—to political extremism. These results illuminate how economic inequality and salient race/class hierarchies can shape psychological profiles that, in turn, are linked to political radicalization.

Amid widening economic and racial economic inequalities and growing polarization, this research introduces and validates a within- and between-group subjective status measure and identifies a theoretically meaningful “Last Place” profile among non-Hispanic white Americans. Membership in this profile is robustly associated with greater support for alt-right candidates, events/groups/movements, and ideologies, even when controlling for objective socioeconomic indicators and demographics. The work highlights the importance of subjective, comparative status perceptions for understanding contemporary right-wing extremism. Future research could experimentally probe causal mechanisms linking upward in-group and intergroup comparisons to extremist support, test interventions that recalibrate status perceptions or reduce perceived intergroup threat, and examine generalizability across contexts, time, and other demographic groups.

- Sampling and generalizability: Quota-based online panels; some quotas were opened late, yielding slight overrepresentation of the Midwest and Northeast. Results are intended to generalize to non-Hispanic white U.S. residents; other populations were not studied. - Exclusions and scope: Only non-Hispanic white Americans born in the U.S. were included; mobile device users were excluded due to measure incompatibility, potentially affecting representativeness. - Preregistration deviations: The foreign-born exclusion criterion was decided before Pilot data analysis but not initially documented in preregistration; key hypotheses were simplified to drop a moderation by white identification (reported as non-significant in supplements). - Measurement/analysis: Self-report measures; missing data on the status measure handled via listwise deletion. LPA-based profile assignment entails some classification uncertainty (though entropy = 0.83 in Study 1 mitigates this). Cross-sectional design limits causal inference. - Oversampling: Individuals without a college degree were oversampled, which may influence profile prevalence estimates (though analyses controlled for education and income).