Antisemitism is predicted by anti-hierarchical aggression, totalitarianism, and belief in malevolent global conspiracies

The study investigates which factors predict antisemitism in contemporary UK populations, considering both 'old' (Judeophobic) and 'new' (antizionist) forms simultaneously via the Generalised Antisemitism (GeAs) scale. Prior debates focus on whether antisemitism is primarily a phenomenon of the political right or left; the authors argue instead that it may be driven by cross-cutting ideological traits such as conspiracism, openness to totalitarian governance, and anti-hierarchical aggression (desire to overturn the social order). The research aims to identify demographic and ideological predictors of antisemitism across the political spectrum, thereby informing theoretical understanding and interventions.

The paper situates antisemitism within scholarship distinguishing 'old' Judeophobic attitudes and 'new' expressions via anti-Israel discourse, noting empirical correlations between anti-Jewish and anti-Israel attitudes (e.g., ADL 2023; Allington et al. 2022; Frindte et al. 2005; Staetsky 2017, 2020). Social Dominance Orientation (SDO) and Right-Wing Authoritarianism (RWA) have mixed associations with prejudice and antisemitism; Frindte et al. (2005) found RWA but not SDO predicted antisemitism, while Swami (2012) found both predicted belief in antisemitic conspiracies in Malaysia. Left-Wing Authoritarianism (LWA) is conceptualized as endorsement of overthrowing existing social orders (Altemeyer 1996); sympathy for violent extremism and anti-capitalist revolutionary aims correlates with antisemitism in UK samples (Staetsky 2017, 2020). Conspiracy beliefs are historically linked to antisemitism (Cohn 1967; Byford 2011; Allington et al. 2021), although evidence is mixed (Swami 2012). Support for totalitarianism is theorized to align with antisemitism given historical cases (Nazi Germany, Fascist Italy, Stalinist USSR). Ethnic nationalism and anti-immigration attitudes have been associated with antisemitism (Staetsky 2020). Demographic correlates reported in prior work include higher antisemitism with age and lower with education (mixed evidence), higher among certain ethnic/religious minorities in Western countries, and lower among women in some studies.

Design: Two cross-sectional studies in the UK. Study 1: self-selecting, quota-limited online sample via Prolific (n=809; 30–31 Oct 2020), quotas for age (18–25 vs 26+) and gender, UK residents. Study 2: representative online panel via YouGov (n=1853; 16–17 Dec 2020), demographic quotas based on 2011 UK census; non-response bias addressed via YouGov demographic weights in linear models (univariate/bivariate unweighted). Measures: Outcome—Generalised Antisemitism (GeAs) scale (Allington et al., 2022a), comprising two subscales: Judeophobic Antisemitism (JpAs; 'old' antisemitism) and Antizionist Antisemitism (AzAs; 'new' antisemitism). Predictors—Generic Conspiracist Beliefs (GCB; Brotherton et al., 2013) domains: malevolent global conspiracies, personal wellbeing, government malfeasance, extraterrestrial cover-up, control of information; Totalitarianism (McClosky & Chong, 1985); Ethnic Nationalism (adapted items on criteria for being 'truly British', incl. ancestry, religion, race); Political Cynicism, Social Cynicism, Political Trust (Pattyn et al., 2012); Social Dominance Orientation (SDO7: Dominance, Egalitarianism; Ho et al., 2015); Sympathies for Radicalisation (SyFoR; modified; Bhui et al., 2014); Right-Wing Authoritarianism (RWA: Submission, Aggression, Conventionalism; Duckitt et al., 2010); Left-Wing Authoritarianism (LWA: Anti-hierarchical Aggression, Anti-conventionalism, Top-down Censorship; Costello et al., 2021). In Study 2, due to length constraints, only the three strongest ideological predictors from Study 1 were retained: Totalitarianism, GCB Malevolent Global conspiracies, and LWA Anti-hierarchical Aggression (AHA; abbreviated to six top-loading items from Study 1). Demographics: age (continuous), gender (female=1 vs other=0), ethnicity (other-than-white=1 vs white=0), education (degree=1 vs non-degree=0). All scale variables normalized to 0.00–1.00 prior to analysis. Analytic strategy: Study 1—bivariate analyses: Welch unequal-variance t-tests for binary demographics (effect size d) and Pearson correlations for age and ideological scales (product-moment r). Study 2—bivariate analyses as above; multivariable modeling: nested linear regression models predicting GeAs (canonical), with parallel models for JpAs and AzAs for transparency. All numeric predictors standardized (z-scores) in linear models. Weighting: YouGov demographic weights applied to linear models in Study 2. Power: Study 1—80% power to detect r≈0.12 at p<0.010; gender-group comparisons well-powered; ethnicity comparisons less powered due to smaller other-than-white subsamples. Study 2—With 7 predictors and df≈1845, models with f²=0.02 detectable with 99% power. Software: R 4.2.2 with dplyr, ggplot2, knitr, kableExtra. Data and code available OSF: https://osf.io/9p2f8/.

Study 1 (n=809): • Demographics: Other-than-white ethnicity showed substantial positive associations with GeAs (d=0.65, 95% CI [0.45, 0.84]), JpAs (d=0.46 [0.27, 0.65]) and AzAs (d=0.60 [0.40, 0.80]). Female gender: lower JpAs (d=-0.43 [−0.56, −0.29]) but marginally higher AzAs (d=0.14 [0.00, 0.28]); net small reduction in GeAs (d=-0.18 [−0.32, −0.05]). Degree attainment: small negative association with GeAs (d=-0.21 [−0.35, −0.08]), stronger for JpAs (d=-0.31 [−0.45, −0.17]). Age: weak negative correlation with GeAs (r=-0.11), driven by AzAs (r=-0.18); JpAs uncorrelated with age. • Ideological bivariate correlates with GeAs: strongest were Totalitarianism (r=0.34), GCB Malevolent Global conspiracies (r=0.34), LWA Anti-hierarchical Aggression (r=0.32); other notable GCB domains: Personal Wellbeing (r=0.30), Government Malfeasance (r=0.30). Ethnic Nationalism strongly predicted JpAs (r=0.37) but not AzAs (r≈-0.08). RWA factors positively correlated with JpAs (r≈0.26–0.29) and negatively with AzAs (r≈-0.12 to -0.13). SDO factors were weak predictors. Study 2 (n=1853): • Bivariate demographic effects: Other-than-white ethnicity associated with higher GeAs (d=0.75 [0.51, 0.99]), JpAs (d=0.65 [0.40, 0.89]), AzAs (d=0.61 [0.36, 0.87]). Female gender: lower JpAs (d=-0.23 [−0.32, −0.13]) but higher AzAs (d=0.27 [0.18, 0.37]); effect on GeAs null (d=0.01). Degree: negatively associated with GeAs (d=-0.45 [−0.55, −0.35]), JpAs (d=-0.57 [−0.67, −0.48]) and modestly with AzAs (d=-0.16 [−0.26, −0.05]). Age: negatively correlated with AzAs (r=-0.15) and weakly with GeAs (r=-0.08); JpAs uncorrelated. • Bivariate ideological correlates (Table 2): LWA Anti-hierarchical Aggression (AHA) r=0.39 with GeAs (0.32 JpAs; 0.34 AzAs); GCB Malevolent Global r=0.36 (0.38 JpAs; 0.22 AzAs); Totalitarianism r=0.31 (0.40 JpAs; 0.09 AzAs). • Multivariable models (Full model, Table 7): GeAs adj. r²=0.29 (f²≈0.40). Strongest standardized coefficients: LWA AHA β=0.29 (p<0.001), GCB Malevolent Global β=0.19 (p<0.001), Totalitarianism β=0.17 (p<0.001), other-than-white ethnicity β=0.43 (p<0.001), degree β=-0.25 (p<0.001). Age and female non-significant for GeAs. JpAs adj. r²=0.32: Totalitarianism β=0.27, GCB β=0.21, LWA AHA β=0.21, other-than-white β=0.41, degree β=-0.25, female β=-0.30 (all p<0.001). AzAs adj. r²=0.16: LWA AHA β=0.29 (p<0.001) strongest; GCB β=0.11 (p<0.001); Totalitarianism non-significant (β=-0.01, p=0.800); other-than-white β=0.30 (p=0.001); degree β=-0.11 (p=0.031); female β=0.18 (p<0.001); age β=-0.08 (p=0.001). • Overall: Ethnicity, anti-hierarchical aggression, and belief in malevolent global conspiracies consistently predict both 'old' and 'new' antisemitism; totalitarianism predicts 'old' but not 'new' antisemitism.

The findings indicate that antisemitism in the UK is less about left–right self-identification and more about cross-cutting ideological traits. Anti-hierarchical aggression and conspiracist worldviews are linked to general antisemitism and to both Judeophobic and Antizionist forms, suggesting they tap a common latent factor. Totalitarian sympathies relate specifically to Judeophobic antisemitism, aligning with historical precedents. Demographically, other-than-white ethnicity consistently predicts higher antisemitism scores across forms; education is protective, particularly against Judeophobic antisemitism; age relates inversely to Antizionist antisemitism; and female gender shows opposing associations with JpAs (lower) and AzAs (higher), cancelling out for GeAs. These results refine debates about whether antisemitism is a right- or left-wing phenomenon by showing it is associated with conspiracism, revolutionary anti-hierarchy, and openness to totalitarian rule, which can occur across the political spectrum. The models explain more variance in JpAs than AzAs, consistent with omission (by design) of some AzAs-specific predictors (e.g., anti-conventionalism, government-malfeasance conspiracies) in Study 2.

Using a two-step design, the paper identifies robust predictors of antisemitism among UK adults: belief in malevolent global conspiracies, support for totalitarian government, and especially anti-hierarchical aggression, alongside demographic predictors (higher among other-than-white ethnic groups; lower among degree-holders). Totalitarianism predicts 'old' (Judeophobic) but not 'new' (Antizionist) antisemitism, while conspiracism and anti-hierarchical aggression predict both. The work contributes validated effect-size estimates and shows that focusing on specific ideological dispositions—rather than left–right self-placement—better explains antisemitism. Future research should: include a broader set of conspiracy domains and LWA/RWA facets in multivariable models; consider more complex models (e.g., path/latent variable models) that reflect the two-subscale structure; use alternative measures less susceptible to social desirability; and examine mediators (e.g., field of study, media consumption) that might explain demographic associations.

Study 2 retested only three ideological predictors due to survey length, likely limiting explained variance for Antizionist antisemitism. Linear regression may oversimplify a construct with two related subscales; structural models could be more appropriate. Self-report agreement with antisemitic statements may be affected by social desirability, especially among more educated respondents. Demographic associations (gender, education, ethnicity) persisted after controlling for the included ideological predictors, implying unmeasured mediators; lack of granularity in education (field/discipline) could mask divergent patterns.