Skip to content
Frontiers
Search

Original Research

Frontiers in Social Psychology ·

Volume 4 - 2026 | https://doi.org/10.3389/frsps.2026.1715636

Censorship for thee, but not for me: dark personality traits are associated with support for anti-democratic actions and violent extremism among US political partisans

  • Daniel W. Snook1*corresponding author
  • Nina Skubala1
  1. Department of Psychology, Florida Gulf Coast University, Fort Myers, FL, United States
Article metrics0 viewsView details
PDFCiteMetrics

Abstract

In democracies around the world, a perennial tension exists between the desire to achieve ideological goals while remaining within the confines of a legitimate democratic system. Research has yielded mixed results on the role that dark personality traits play in such processes, but a recent theory, the Dark Ego Vehicle Principle, sheds light on why some individuals engage in peaceful ideological activism, and may also help explain support for anti-democratic and violent behaviors to promote one's ideology. The Dark Ego Vehicle Principle postulates that those with higher levels of dark personality traits engage in ideological activism to satisfy their own ego needs (e.g., to get attention from others or to dominate or abuse others), rather than to support or promote their cause in good faith. The present work expanded on this base, using multiple regression to show that for both Democrats (Study 1, N = 280) and Republicans (Study 2, N = 248), Dark Personality Traits, as well as Authoritarianism and Social Dominance Orientation, were positively associated with support for anti-democratic actions and violent extremism, but commitment to prosocial party values were not. In the Republican sample, moderation analyses also showed that Needs Frustration strengthened the impact of Dark Personality Traits on support for violent extremism, but not anti-democratic actions; for Democrats, these moderations were not significant. These results strengthen the base of evidence supporting the Dark Ego Vehicle Principle.

Keywords

Introduction

The great thing about having enemies is that you can pretend that all the badness in the whole world is in your enemies and all the goodness in the whole world is in you! Attractive isn't it? So, if you… enjoy abusing people, then you can pretend that you're only doing it because these enemies of yours are such very bad persons.

– John Cleese, “The Advantages of Extremism” (Cleese, 1987)

Individuals living in democratic societies are faced with a decision when considering what means to use to promote an ideological cause. They can opt for peaceful, democratic means, such as voting, protesting, sharing information, and contacting elected officials. However, these means have historically been viewed by violent extremists for various ideologies as ineffective and slow (Rapoport, 2004), leading some to endorse or take part in anti-democratic means, such as gerrymandering, voter suppression, censorship, and targeted violence. The recent left-wing terrorist attack that left Charlie Kirk, a prominent conservative activist, dead, is one of a string of examples of ideologically motivated actions that have broken loose of democratic norms in the US thus far in 2025. Recent anti-democratic government actions include censorship violating the First Amendment (Johnson, 2025), describing criticism of the government as “really illegal” (Broadwater, 2025), and the unconstitutional, illegal arrests of lawful US residents (Bier, 2025), while recent individual actions have included the politically motivated arson of New Mexico's Republican Headquarters (Department of Justice, 2025a) and the right-wing terrorist attack that resulted in the deaths of Minnesota State Representative Melissa Hortman and her husband Mark Hortman (and their dog, Gilbert), as well the injury of Minnesota State Senator John Hoffman and his wife Yvette Hoffman (Department of Justice, 2025b). Recent data from The Polarization Research Lab, an organization that collects weekly polls in order to obtain data on levels of political party polarization, support for violence, and respect for democracy, show that while a majority of members from both parties support democratic principles like free speech and fair elections, almost 30% support at least one anti-democratic action (Polarization Research Lab, 2025).

Research has yielded mixed results on the role that dark personality traits, a cluster of socially aversive and malevolent personality traits, plays in anti-democratic behaviors and violent extremism (the endorsement or use of violence to achieve ideological goals). Recent work in this area has introduced the Dark Ego Vehicle Principle (DEVP), which postulates that those with dark personality traits do not engage in ideological activism, whether peaceful or violent, to support a cause in good faith, but rather to satisfy their own ego needs. Specifically, DEVP states those higher in dark personality traits are not driven by their commitment to the prosocial elements of their ideological cause, but rather to use their cause as a “vehicle” for boosting their ego by receiving attention for their advocacy or dominating others via anti-social rhetoric and behavior (Krispenz and Bertrams, 2023c; 2025). The present work expands on this research by assessing the relationships between variables relevant to DEVP (i.e., the dark tetrad, authoritarianism, social dominance orientation, and prosocial traits, such as altruism and commitment to ideological values) and support for anti-democratic and violent extremism in samples of US Democrats and Republicans.

Dark personality traits

The dark triad theory of personality (Paulhus and Williams, 2002) describes a personality type that consists of three overlapping traits: narcissism, psychopathy, and Machiavellianism. Rather than describing disorders with diagnoses listed in the DSM-5 (e.g., narcissistic personality disorder), the dark triad describes three non-pathological personality traits that can be measured separately and are interrelated (Furnham et al., 2013). Individuals who score high in narcissism tend to display grandiosity and seek to enhance their public image, while those who score high in psychopathic traits tend to display low empathy and have a tendency to engage in impulsive or risk taking behaviors (Paulhus and Williams, 2002). Individuals who score high in machiavellianism tend to be manipulative and seek to gain power even at the expense of others (Jahangir et al., 2024). Further research on dark personality traits has since incorporated everyday sadism, defined as the enjoyment of other people's suffering (Paulhus and Dutton, 2016). Together, these four traits have become known as the dark tetrad (Buckels et al., 2013) and each is commonly referred to as a dark personality trait.

Authoritarianism and social dominance orientation

In psychological research, authoritarianism has been characterized as a blind submissiveness to persons in authority, displays of dominance toward perceived subordinates, and hostility toward dissenters (Adorno et al., 1950). Initial studies focused on right-wing authoritarianism (RWA), which is characterized by a combination of submission, aggression, and conventionalism (Altemeyer, 1996). However, in subsequent research, it was proposed that RWA better represented the submissive aspect of authoritarianism, while social dominance orientation (SDO; Pratto et al., 1994) better explained the dominant aspect of authoritarianism (Altemeyer, 1998). SDO refers to a set of beliefs about the natural hierarchy of social groups and the legitimization of dominance and power differences in intergroup relations. SDO has been found to be related to narcissism (Cichocka et al., 2017; Zeigler-Hill et al., 2020) and other dark personality traits (Hodson et al., 2009), as well as various forms of prejudice, racism, and violent extremism (Whitley, 1999; Duckitt and Sibley, 2007; Kteily et al., 2011).

More recent literature has also described left-wing authoritarianism (LWA; Conway et al., 2017; Costello et al., 2022; Crawford and Brandt, 2020). Costello et al. (2022) have developed an empirically substantiated conceptualization of LWA, which is made up of three correlated facets: anti-conventionalism, or a strict adherence and endorsement of progressive values, as opposed to conventional or conservative values; top-down censorship, characterized by the desire for those in authority to deal with opposition and to suppress offensive and intolerant speech; and anti-hierarchical aggression, which is the desire to forcefully overthrow and undermine established authority, and punish those in power.

As described below, both authoritarianism and SDO have been measured in relation to dark personality traits and support for both peaceful and violent actions for an ideological.

Previous research on the dark ego vehicle principle

In their initial research establishing Dark Ego Vehicle Principle (DEVP), Krispenz and Bertrams (2023c) investigated the relationship between trait narcissism and left wing authoritarianism (LWA) in two studies, finding that antagonistic narcissism (Study 1) and psychopathy (Study 2) were positively associated with the anti-hierarchical aggression component of LWA, whereas the prosocial traits of dispositional altruism and commitment to social justice values were not. These findings helped produce the theory that some people do not engage in activism to advocate for ideological values (in this case, social justice), but rather to “endorse or exercise violence against others to satisfy their own ego-focused needs” (Krispenz and Bertrams, 2023c, p. 2,714) and that dark personality traits can help explain who tends to engage for more ego-focused, rather than ideological, reasons. These researchers continued to investigate DEVP in various left wing activism groups. Dark personality traits, especially narcissism, were found to be associated with activism for anti-Zionist anti-semetic ideology (Bertrams and Krispenz, 2024), anti-sexual assault activism (Bertrams and Krispenz, 2023a), feminist activism (Krispenz and Bertrams, 2023a), environmental activism (Krispenz and Bertrams, 2023b), and LGBTQ and gender identity activism (Bertrams and Krispenz, 2023b). Most recently, DEVP has been demonstrated in groups of both left- and right-wing activists (pro- and anti-abortion activists, respectively), with narcissism, LWA, and SDO (but not RWA) being positively associated with radicalism intentions (defined as behavioral intentions to do illegal and/or violent activities for one's cause; Krispenz and Bertrams, 2025). Similarly, Bertrams and Krispenz (2025) showed that narcissism helped explain forms of anti-semitism that are associated with being politically left- and right-wing (anti-Zionist antisemitism and Judeophobic antisemitism, respectively). This study provided two novel additions to DEVP literature: first, it showed evidence that DEVP operates in an explicitly right-wing context, whereas previous DEVP research had been demonstrated only in left-wing political contexts (e.g., anti-zionism, LGBTQ activism, etc.) and second, it demonstrated DEVP principle at work for radical (illegal/violent behaviors), rather than only activist (legal and non-violent), intentions.

Other research on dark personality traits and violent extremism

Other research has also explored the potential role of dark personality traits as risk factors for violent extremism. Meta-analytic data show that dark personality traits tend to have small, but noticeable associations both with justifying ideological violence and intentions to carry out ideological violence (Wolfowicz et al., 2021). In a study conducted in a Croatian sample, those who endorsed peaceful activism tended to score lower on a general measure of dark personality traits, while those who endorsed radical (illegal and/or violent) activism had higher scores (Pavlović, 2024). Further research has also shown that measures of a general dark personality trait have been associated with support for violent extremism (Pavlović and Franc, 2023), as have specific traits including psychopathy (Bélanger et al., 2023; Nai and Young, 2024), sadism (Bélanger et al., 2023), Machiavellianism (Nai and Young, 2024), and narcissism (Pavlović and Wertag, 2021). In addition, some of these dark personality traits have been shown to interact with other variables, such as obsessive passion for one's ideology (when one's motivation to advocate for one's ideology dominates one's identity and goal pursuits; Bélanger et al., 2023) and perceived group relative deprivation (when one feels one's ingroup is treated unfairly compared to outgroups; Pavlović and Franc, 2023), to predict support for violent extremism even more strongly.

The present study

Here, we seek to extend research on DEVP, as well as on dark personality traits and violent extremism more generally, by assessing the relationships between these variables and support for anti-democratic actions, including the use of violent extremism. Support for anti-democratic actions refers to support for actions, whether by individuals, groups, or governments, that undermine democratic institutions, norms, or rights (Greene, 2023; Holliday et al., 2024). Given the highly polarized sociopolitical climate in the US, research on this topic is increasingly relevant. The present study seeks to further investigate the implications of the DEVP by assessing the extent to which dark personality traits are associated with support for anti-democratic actions and violent extremism when also including other relevant variables, like commitment to prosocial ideological values, authoritarianism, and social dominance orientation. Importantly, this research will expand on existing knowledge by assessing the DEVP in two novel, but highly relevant samples: US Democrats and US Republicans.

Based on the findings of previous DEVP research, we hypothesize that multiple regressions will show that a general dark personality trait variable, as well as narcissism specifically, will be associated with greater support for anti-democratic actions (Hypothesis 1 and 2) and violent extremism (H3 and H4). In line with DEVP research, we also hypothesize that prosocial ideological variables (commitment to social justice for Democrats and commitment to family values for Republicans) will not be associated with greater support for anti-democratic actions (H5) or violent extremism (H6) in multiple regression analyses. In line with DEVP theory, we also hypothesize that a general dark personality trait variable and narcissism will remain a significant predictor of support for anti-democratic actions (H7) and violent extremism (H8), even when accounting for the influence of authoritarianism and SDO in both multiple regression and mediation analyses. In addition, expanding on other on the role of dark personality traits in violent extremism, we hypothesize that personal needs fulfillment (as described by Self-Determination Theory; Deci and Ryan, 1985) will interact with dark personality traits, such that greater needs fulfillment will reduce the intensity of the relationship between dark personality traits and support for anti-democratic actions (H9) and violent extremism (H10).

Study 1: Methods

Procedure and participants

Participants were recruited via Prolific, an online research platform that specializes in gathering high quality data from participants. Prolific provides an option to select sub-populations from among their base of participants, which they call “screeners.” These allow researchers to pre-select participants based on various self-reported demographic variables, such as nationality (e.g., US and UK), gender identity, and political preference. For Study 1, we selected two such screeners to limit our potential participant pool to US residents who identify themselves as members of the Democratic Party. Based on a power analysis for multiple regression (assuming small-to-medium effect sizes and power set at 0.80), in which a sample of 185 people was suggested (G*power; Faul et al., 2009) and a power analysis for mediation using Qin (2024) method (assuming small-to-medium effect sizes and power set at 0.80), in which a sample size of 225 was suggested, we recruited a sample of 351 participants who met our screener criteria. Data collection took place over 2 days in late August of 2025. Participants took a short survey which included a required informed consent document, demographic questions, included two attention check questions, a measure of social desirability bias, and the scales described (see Measures). Prolific's software automatically rejects participants who complete surveys exceptionally fast (more than 3 standard deviations below the mean completion time), as well as participants take too long to complete the surveys indicating inactivity (in this case, over an hour). Of the 351 participants who completed the survey in a reasonable amount of time, 9 failed one or both attention check questions included in the survey. Of the remaining 342 participants, 62 were removed due to having responded to four out of four items measuring social desirability from the Brief Social Desirability Scale (BSDS; Haghighat, 2007; e.g., “I never lie to people” and “I always smile at people when I see them”) affirmatively (response options were Yes/No), indicating a lack of honesty or candor in their responses. In survey research asking people to self-report attitudes or behaviors that are socially frowned upon, such as anti-democratic tendencies and support for violent extremism, participants with high social desirability bias are especially problematic. Thus, the final sample size of data with adequate quality was 280, with an age range of 18 to 68 (M = 38.54, SD = 12.46). Participants' racial, gender, and socio-economic characteristics for Study 1 and Study 2 are described in Table 1. All elements of this study were approved by the Florida Gulf University IRB.

Table 1. Frequency Distributions for Demographic Variables in Studies 1 and 2.

VariableLevelStudy 1 (*N* = 280 Democrats)Study 2 (*N* = 248 Republicans)
FrequencyPercentageFrequencyPercentage
GenderMale9132.511345.6
Female18566.113454.0
Other41.410.4
SESUpper class (Wealthy)00.010.4
Upper middle class227.93413.7
Middle class11842.111245.2
Lower middle class10938.97931.9
Lower class (below the poverty line)3111.1228.9
Race/EthnicityBlack3612.9156.0
White19569.621084.7
Hispanic134.662.4
Native American31.110.4
Pacific Islander10.400.0
Asian279.6104.0
Other (e.g., identify as mixed-race)51.862.4

Transparency and open science

Full survey items and data files are included in the “Files” section of the following anonymous OSF link: https://osf.io/syr53/?view_only=30309bf8503947a0a2a9e641bf19d1af.

Measures

All items for all scales were answered on a 6-point response scale with options ranging from Strongly disagree (1) to Strongly agree (6). For all scales, mean scores of items were calculated to represent the measurement of the variable. Prior to calculating a mean score for scales, any reverse-scored items were re-coded.

Dark personality variables

Dark personality variables were measured using the Short Dark Tetrad scale (SD4; Paulhus et al., 2021), which is a 28-item questionnaire with four subscales measuring Machiavellianism (7 items, e.g., “Whatever it takes, you must get the important people on your side”), narcissism (7 items, e.g., “Group activities tend to be dull without me”), psychopathy (7 items, e.g., “People who mess with me always regret it”), and sadism (7 items, e.g., “Some people deserve to suffer”). We used these items to calculate mean scores for individual traits, as well as an overall dark personality trait, following previous research (e.g., Pavlović, 2024). The overall dark personality trait variable showed excellent internal consistency (Study 1, McDonald's ω = 0.89; Study 2, McDonald's ω = 0.90), while the specific traits of Machiavellianism (Study 1, McDonald's ω = 0.75; Study 2, McDonald's ω = 0.72), narcissism (Study 1, McDonald's ω = 0.71; Study 2, McDonald's ω = 0.69), psychopathy (Study 1, McDonald's ω = 0.72; Study 2, McDonald's ω = 0.70), and sadism (Study 1, McDonald's ω = 0.74; Study 2, McDonald's ω = 0.74) showed adequate internal consistency.

Social justice values commitment

Social Justice Values Commitment was measured using the Social Justice Scale (Torres-Harding et al., 2011) with 5 items (e.g., “I believe that it is important to try to change larger social conditions that cause individual suffering and impede wellbeing”). The scale demonstrated excellent internal consistency (McDonald's ω = 0.88).

Left wing authoritarianism

Participants completed a short version of the Left-Wing Authoritarianism Index (LWAI; Costello et al., 2022; de Oliveira Santos and Jost, 2024) with 4 items (e.g., “It is important that we dismantle the US's traditional nationalist, imperialist values”). The LWAI demonstrated good internal consistency (McDonald's ω = 0.80).

Social dominance orientation

We used the short version of the Social Dominance Orientation 7 Scale (SDO7; Ho et al., 2015), which has 8 items (e.g., “It's probably a good thing that certain groups are at the top and other groups are at the bottom.”). The coding was such that higher scores represent higher social dominance orientation. The SDO7 demonstrated adequate internal consistency (Study 1, McDonald's ω = 0.82; Study 2, McDonald's ω = 0.72).

Needs Fulfillment (SDT)

We measured needs fulfillment using the Basic Psychological Need Satisfaction Scale (Deci and Ryan, 2000), which is a 21-item scale that measures one's need satisfaction in general. The items measure the extent one's feelings of volition or autonomy (e.g., “I feel like I am free to decide for myself how to live my life”), effectiveness or competence (“People I know tell me I am good at what I do”), and connection or relatedness with important others are being fulfilled (e.g., “People in my life care about me.”). The scale demonstrated good internal consistency (Study 1, McDonald's ω = 0.86; Study 2, McDonald's ω = 0.83).

Support for anti-democratic actions

We used 11 items from recent studies (Costello et al., 2022; de Oliveira Santos and Jost, 2024; Moore-Berg et al., 2020), to measure support for anti-democratic actions, including support for an authoritarian candidate (e.g., “I would support a great left-wing candidate, even if they call for their political opponents to be thrown in jail”), support for online, media, and in-person censorship (e.g., “If Democrats gain control, they should heavily restrict or shut down Fox News to stop the spread of misinformation,” “Colleges and universities should ban speakers and protests endorsing intolerant views”), and gerrymandering (“Democrats should redraw districts to maximize their potential to win more seats in upcoming elections, even if it may be technically illegal”)1. The scale demonstrated good internal consistency (McDonald's ω = 0.84).

Violent extremism

We used 5 items from several previous studies (de Oliveira Santos and Jost, 2024; Gousse-Lessard et al., 2013; Moore-Berg et al., 2020), to measure violent extremism (e.g., “If a few of the worst Republican politicians were assassinated, it wouldn't be the end of the world,” “I would personally be willing to use violence to ensure that a Democrat candidate wins the next presidential election”). The scale demonstrated good internal consistency (McDonald's ω = 0.81).

Study 1: Results

Bivariate correlations

Bivariate relationships are shown in Table 2. The results show that the overall Dark Tetrad measure, as well as each of its subcomponents, Machiavellianism, Narcissism, Psychopathy, and Sadism, were all positively correlated with both Support for Anti-Democratic Actions and Violent Extremism, though they were more strongly associated with Violent Extremism. Among the Dark Tetrad subcomponents, Machiavellianism was most strongly correlated with Support for Anti-Democratic Actions, whereas Sadism was most strongly correlated with Violent Extremism. Left Wing Authoritarianism and Social Dominance Orientation were also positively associated with Support for Anti-Democratic Actions and Violent Extremism, whereas Needs Fulfillment was negatively associated. Commitment to Social Justice Values was not correlated with Support for Anti-Democratic Actions or Violent Extremism.

Table 2. Means, Standard Deviations, and Bivariate Correlations for Study 1 (N= 280 Democrats, correlations above the diagonal) and Study 2 (N = 248 Republicans, correlations below the diagonal).

VariableMean and (SD) Study 1 and [Study 2]ValueDark Tetrad OverallMachia-vellian-ismNarci-ssismPsycho-pathySad-ismAuthori-tianismSDOValues Commit-mentNeeds Fulfill-mentAuto-nomyRelated-nessCompe-tenceAnti- Demo-craticViolent Extre-mism
Dark Tetrad Overall3.09 (0.87) [3.26 (0.93)]Pearson's r10.871**0.878**0.898**0.836**−0.0390.361**−0.163**−0.0790.007−0.155**0.0060.379**0.431**
p value0.0000.0000.0000.0000.5130.0000.0060.1860.9120.0090.9170.0000.000
Machia-vellianism3.43 (1.00) [3.67 (1.04)]Pearson's r0.899**10.705**0.707**0.627**−0.0700.334**−0.134*0.0060.076−0.0840.0450.361**0.381**
p value0.0000.0000.0000.0000.2400.0000.0250.9180.2060.1610.4500.0000.000
Narcissism3.16 (0.94) [3.31 (0.94)]Pearson's r0.889**0.741**10.789**0.592**−0.0220.332**−0.118*−0.0150.006−0.1150.0840.302**0.358**
p value0.0000.0000.0000.0000.7110.0000.0490.8000.9220.0540.1610.0000.000
Psychopathy2.83 (0.96) [3.03 (1.02)]Pearson's r0.914**0.771**0.785**10.647**0.0110.346**−0.124*−0.076−0.051−0.143*0.0370.332**0.362**
p value0.0000.0000.0000.0000.8510.0000.0380.2050.3990.0160.5350.0000.000
Sadism2.93 (1.11) [3.04 1.14]Pearson's r0.884**0.713**0.692**0.730**1−0.0510.252**−0.185**−0.176**−0.007−0.191**−0.125*0.322**0.396**
p value0.0000.0000.0000.0000.3990.0000.0020.0030.9120.0010.0370.0000.000
Authoritian-ism4.56 (1.08) [3.61 (0.84)]Pearson's r0.1100.168**0.0690.0730.0831−0.363**0.453**−0.1060.0170.000−0.208**0.305**0.267**
p value0.0830.0080.2820.2520.1920.0000.0000.0760.7811.0000.0000.0000.000
SDO1.80 (0.72) [3.38 (1.03)]Pearson's r0.182**0.165**0.147*0.176**0.166**0.245**1−0.606**−0.090−0.090−0.123*0.0180.202**0.235**
p value0.0040.0090.0210.0060.0090.0000.0000.1340.1350.0400.7580.0010.000
Values Commitment5.34 (0.67) [6.18 (0.88)]Pearson's r−0.0090.038−0.031−0.0450.0020.456**0.259**10.269**0.175**0.250**0.128*−0.054−0.022
p value0.8900.5520.6240.4800.9800.0000.0000.0000.0030.0000.0320.3680.719
Needs Fulfillment4.67 (0.66) [4.76 (0.67)]Pearson's r0.0430.1210.132*−0.008−0.070−0.0100.150*0.233**10.421**0.781**0.774**−0.290**−0.291**
p value0.4980.0560.0370.9020.2690.8800.0180.0000.0000.0000.0000.0000.000
Autonomy4.25 (0.70) [4.19 (0.75)]Pearson's r0.0180.0110.035−0.0170.034−0.0470.132*0.132*0.462**10.0780.043−0.017−0.049
p value0.7810.8620.5840.7940.5960.4620.0380.0370.0000.1910.4750.7750.412
Relatedness5.03 (1.06) [5.14 (1.03)]Pearson's r0.0210.0590.0540.013−0.040−0.0060.0810.205**0.737**0.08010.384**−0.285**−0.272**
p value0.7430.3570.3940.8430.5330.9280.2040.0010.0000.2110.0000.0000.000
Competence4.74 (1.10) [4.96 (1.13)]Pearson's r0.0430.152*0.155*−0.015−0.1110.0220.1030.140*0.778**0.0750.342**1−0.236**−0.227**
p value0.5000.0170.0150.8120.0800.7280.1060.0270.0000.2410.0000.0000.000
Anti-Democratic3.07 (0.95) [3.66 (1.42)]Pearson's r0.308**0.279**0.238**0.280**0.302**0.534**0.310**0.286**−0.0530.103−0.114−0.05410.702**
p value0.0000.0000.0000.0000.0000.0000.0000.0000.4070.1050.0730.3980.000
Violent Extremism2.46 (1.03) [2.41 (1.15)]Pearson's r0.446**0.376**0.325**0.422**0.462**0.346**0.290**0.043−0.183**0.011−0.149*−0.195**0.634**1
p value0.0000.0000.0000.0000.0000.0000.0000.5020.0040.8640.0190.0020.000

**Correlation is significant at the 0.01 level (2-tailed). *Correlation is significant at the 0.05 level (2-tailed).

Multiple regression

We used multiple regression to assess the extent to which key variables were related to Support for Anti-Democratic Actions and Violent Extremism. We tested two models assessing Support for Anti-Democratic Actions and two additional models assessing Violent Extremism; one model included the overall Dark Tetrad variable, while the other model included all four Dark Tetrad subcomponents as variables. Multiple regression results are shown in Tables 3, 4 below.

Table 3. Multiple regression analysis for support for anti-democratic actions in study 1 (N = 280 Democrats).

ModelVariable*B**Std. Error*β*t**p*95% CI
Model 1(Constant)0.440.600.730.47−0.75, 1.63
Dark tetrad overall0.340.060.315.69< 0.0010.22, 0.46
LWA0.370.050.427.35< 0.0010.27, 0.47
SDO0.260.090.202.880.000.08, 0.44
Social justice values commitment−0.110.10−0.08−1.130.26−0.30, 0.08
Model 2(Constant)0.290.610.480.63−0.91, 1.49
Machiavellianism0.220.080.232.930.000.07, 0.37
Narcissism−0.020.09−0.02−0.200.84−0.19, 0.16
Psychopathy0.010.090.010.060.95−0.17, 0.18
Sadism0.120.060.141.960.050.00, 0.24
LWA0.380.050.447.53< 0.0010.28, 0.48
SDO0.270.090.213.000.000.09, 0.45
Social Justice Values Commitment−0.100.10−0.07−1.020.31−0.29, 0.09

Outcome = Support for anti-democratic actions.

Table 4. Multiple regression analysis for violent extremism in study 1 (N = 280 Democrats).

ModelVariable*B**Std. Error*β*t**p*95% CI
Model 1(Constant)−1.280.65−1.980.05−2.56, −0.01
Dark Tetrad Overall0.430.060.366.64< 0.0010.30, 0.55
LWA0.340.050.366.34< 0.0010.24, 0.45
SDO0.360.100.253.73< 0.0010.17, 0.55
Social Justice Values Commitment0.040.100.030.390.70−0.16, 0.24
Model 2(Constant)−1.590.65−2.430.02−2.87, −0.30
Machiavellianism0.160.080.162.000.050.00, 0.32
Narcissism0.070.100.060.740.46−0.12, 0.26
Psychopathy−0.060.10−0.06−0.660.51−0.26, 0.13
Sadism0.240.070.263.66< 0.0010.11, 0.37
LWA0.360.050.376.57< 0.0010.25, 0.46
SDO0.400.100.284.14< 0.0010.21, 0.60
Social Justice Values Commitment0.080.100.050.730.47−0.13, 0.28

Outcome = Violent Extremism.

Model 1 for both Support for Anti-Democratic Actions and Violent Extremism show that Dark Tetrad, Left Wing Authoritarianism, and Social Dominance Orientation, but not Social Justice Values Commitment, were positively associated with both Support for Anti-Democratic Actions and Violent Extremism. Model 2 for Support for Anti-Democratic Actions showed that Machiavellianism was the only Dark Tetrad subcomponent that was significantly associated with Support for Anti-Democratic Actions. Model 2 for Violent Extremism showed that Machiavellianism and Sadism were both significantly associated with Violent Extremism. When comparing Models 1 and 2 for Support for Anti-Democratic Actions, there was no significant increase in variance captured by Model 2 compared to Model 1 (Δ R-Sq = 0.01, ΔF = 1.74, p = 0.159); the same was true for Models 1 and 2 for Violent Extremism (Δ R-Sq = 0.02, ΔF = 2.34, p = 0.074).

Mediation analyses

We also conducted mediation analyses to assess the extent to which the relationships between Dark Tetrad variables and our outcome variables (Support for Anti-Democratic Actions and Violent Extremism) could be explained by Social Dominance Orientation or Left Wing Authoritarianism. Neither Social Dominance Orientation or Left Wing Authoritarianism were significant mediators for the relationship between the overall Dark Tetrad variable and Support for Anti-Democratic Actions (Standardized Indirect Effect = 0.03, Bootstrapped Confidence Interval = −0.02,0.08; Standardized Indirect Effect = −0.01, Bootstrapped Confidence Interval = −0.05, 0.03 respectively) or Violent Extremism (Standardized Indirect Effect = −0.03, Bootstrapped Confidence Interval = −0.01,0.08; Standardized Indirect Effect = −0.01, Bootstrapped Confidence Interval = −0.05, 0.03 respectively).

Moderation analyses

Finally, following previous research, we tested for interactions between Dark Tetrad variables and Needs Fulfillment using a series of moderation analyses. The interaction between the overall Dark Tetrad variable and Needs Fulfillment on Support for Anti-Democratic Actions was not significant (B = 0.02, t = 0.30, p = 0.762). The same was true for Violent Extremism (B = −0.03, t = −0.20, p = 0.761). All interactions tested between individual Dark Tetrad subcomponents and Needs Fulfillment on Support for Anti-Democratic Actions and Violent Extremism were also not significant.

Study 2: Methods

Procedure and participants

Participants in Study 2 were recruited using the same means as those in Study 1, meaning they were recruited via Prolific, but in Study 2 screening questions yielded a pool of potential participants who were US residents and identified themselves as members of the Republican Party. Based on a power analysis for multiple regression (assuming small-to-medium effect sizes and power set at 0.80), in which a sample of 185 people was suggested (G*power; Faul et al., 2009) and a power analysis for mediation using Qin (2024) method (assuming small-to-medium effect sizes and power set at 0.80), in which a sample size of 225 was suggested, we recruited a sample of 349 participants who met our screener criteria. Data collection also took place over 2 days in late August of 2025. We applied the same data quality parameters described for Study 1. Of the 362 participants who completed the survey in a reasonable amount of time, 13 failed one or both attention check questions included in the survey. Of the remaining 349 participants, 101 were removed due to having responded to four out of four items measuring social desirability from the Brief Social Desirability Scale (BSDS; Haghighat, 2007; e.g., “I never lie to people” and “I always smile at people when I see them”) affirmatively (response options were Yes/No), indicating a lack of honesty or candor in their responses. In survey research asking people to self-report attitudes or behaviors that are socially frowned upon, such as anti-democratic tendencies and support for violent extremism, participants with high social desirability bias are especially problematic. Thus, the final sample size of data with adequate quality was 248, with an age range of 18 to 74 (M = 43.88, SD = 14.02). Participants' racial, gender, and socio-economic characteristics are described in Table 1 (See Study 1 Methods). All elements of this study were approved by the Florida Gulf University IRB.

Transparency and open science

Full survey items and data files are included in the “Files” section of the following anonymous OSF link: https://osf.io/syr53/?view_only=30309bf8503947a0a2a9e641bf19d1af.

Measures

Items measuring Dark Personality Variables, Social Dominance Orientation, and Needs Fulfillment were identical in Study 2 to those of Study 1. Items measuring Anti-Democratic Attitudes and Violent Extremism were structured in the same way, but used wording appropriate for Republicans, rather than Democrats. Only Authoritarianism (Left Wing Authoritarianism among Democrats in Study 1, Right Wing Authoritarianism among Republicans in Study 2) and commitment to core party values (Social Justice Commitment in Democrats among Study 1, Family Values Commitment among Republicans in Study 2) were measured using different items in Study 2.

Family values commitment

Family Values Commitment was measured using 5 items assessing support for key Republican talking points regarding “family values” (e.g., “I believe that maintaining traditional family roles, duties, and responsibilities strengthens society,” “I believe it is important to promote strong, stable families built on traditional values”). The scale demonstrated good internal consistency (McDonald's ω = 0.86).

Right wing authoritarianism

Participants completed a short version of the Right-Wing Authoritarianism Scale (RWA Scale; Altemeyer, 2006; de Oliveira Santos and Jost, 2024) with 4 items (e.g., “What our country really needs instead of more “civil rights” is a good stiff dose of law and order”). The RWA Scale demonstrated excellent internal consistency (McDonald's ω = 0.89).

Support for anti-democratic actions

We used 11 items from recent studies (Costello et al., 2022; de Oliveira Santos and Jost, 2024; Moore-Berg et al., 2020), to measure support for anti-democratic actions, including support for an authoritarian candidate (e.g., “I would support a great right-wing candidate, even if they call for their political opponents to be thrown in jail”), support for online, media, and in-person censorship (e.g., “If Republicans gain control, they should heavily restrict or shut down MSNBC and CNN to stop the spread of fake news,” “Colleges and universities should ban speakers and protests endorsing anti-American views”), and gerrymandering (“Republicans should redraw districts to maximize their potential to win more seats in upcoming elections, even if it may be technically illegal”). See the supplementary file for the full items. The scale demonstrated excellent internal consistency (McDonald's ω = 0.91).

Violent extremism

We used 5 items from several previous studies (de Oliveira Santos and Jost, 2024; Gousse-Lessard et al., 2013; Moore-Berg et al., 2020), to measure violent extremism (e.g., “If a few of the worst Democrat politicians were assassinated, it wouldn't be the end of the world,” “I would personally be willing to use violence to ensure that a Republican candidate wins the next presidential election”). The scale demonstrated good internal consistency (McDonald's ω = 0.78).

Study 2: Results

Bivariate correlations

Bivariate relationships are shown in Table 2. As in Study 1, the results show that the overall Dark Tetrad measure, as well as each of its subcomponents, Machiavellianism, Narcissism, Psychopathy, and Sadism, were all positively correlated with both Support for Anti-Democratic Actions and Violent Extremism, though they were more strongly associated with Violent Extremism. Among the Dark Tetrad subcomponents, Machiavellianism and Psychopathy (for both r = 0.28) were most strongly correlated with Support for Anti-Democratic Actions, whereas Sadism was most strongly correlated with Violent Extremism. Right Wing Authoritarianism and Social Dominance Orientation were also positively associated with Support for Anti-Democratic Actions and Violent Extremism, whereas Needs Fulfillment was not associated with Support for Anti-Democratic Actions, but was negatively associated with Violent Extremism. Commitment to Family Values was positively correlated with Anti-Democratic Actions, but was not correlated with Violent Extremism.

Multiple regression

As in Study 1, we used multiple regression to assess the extent to which key variables were related to Support for Anti-Democratic Actions and Violent Extremism. We tested two models assessing Support for Anti-Democratic Actions and two additional models assessing Violent Extremism; one model included the overall Dark Tetrad variable, while the other model included all four Dark Tetrad subcomponents as variables. Multiple regression results are shown in Tables 5, 6 below.

Table 5. Multiple regression analysis for support for anti-democratic actions in study 2 (N = 248 Republicans).

ModelVariable*B**Std. Error*β*t**p*95% CI
Model 1(Constant)−1.380.58−2.390.02−2.52, −0.24
Dark Tetrad Overall0.360.080.234.46< 0.0010.20, 0.51
RWA0.770.100.457.78< 0.0010.57, 0.96
SDO0.200.080.152.680.010.05, 0.35
Family Values Commitment0.070.090.040.760.45−0.11, 0.26
Model 2(Constant)−1.250.59−2.130.03−2.40, −0.09
Machiavellianism−0.080.12−0.06−0.640.52−0.32, 0.16
Narcissism−0.010.13−0.01−0.090.93−0.28, 0.25
Psychopathy0.190.130.141.440.15−0.07, 0.46
Sadism0.230.100.192.300.020.03, 0.43
RWA0.780.100.467.91< 0.0010.59, 0.98
SDO0.190.080.142.580.010.05, 0.34
Family Values Commitment0.080.090.050.800.42−0.11, 0.26

Outcome= support for anti-democratic actions.

Table 6. Multiple regression analysis for violent extremism in study 2 (N = 248 republicans).

ModelVariable*B**Std. Error*β*t**p*95% CI
Model 1(Constant)−0.210.49−0.440.66−1.17, 0.75
Dark tetrad overall0.470.070.386.97< 0.0010.34, 0.60
RWA0.450.080.335.47< 0.0010.29, 0.62
SDO0.200.060.183.210.000.08, 0.33
Family values commitment−0.200.08−0.15−2.500.01−0.35, −0.04
Model 2(Constant)0.070.480.140.89−0.88, 1.02
Machiavellianism−0.050.10−0.04−0.480.63−0.25, 0.15
Narcissism−0.150.11−0.12−1.330.19−0.36, 0.07
Psychopathy0.260.110.232.370.020.04, 0.48
Sadism0.350.080.354.28< 0.0010.19, 0.52
RWA0.470.080.345.77< 0.0010.31, 0.63
SDO0.190.060.173.110.000.07, 0.31
Family values commitment−0.200.08−0.15−2.530.01−0.35,−0.04

Model 1 for both Support for Anti-Democratic Actions and Violent Extremism show that Dark Tetrad, Right Wing Authoritarianism, and Social Dominance Orientation, were all positively associated with both Support for Anti-Democratic Actions and Violent Extremism, whereas Family Values Commitment was negatively associated. Model 2 for Support for Anti-Democratic Actions showed that Psychopathy and Sadism were both significantly associated with Support for Anti-Democratic Actions. Model 2 for Violent Extremism also showed that Psychopathy and Sadism were both significantly associated with Violent Extremism. When comparing Models 1 and 2 for Support for Anti-Democratic Actions, there was no significant increase in variance captured by Model 2 compared to Model 1 (Δ R-Sq = 0.01, ΔF = 1.30, p = 0.275). However, this was not the case for Models 1 and 2 for Violent Extremism, where Model 2 did capture significantly more variance than Model 1 (Δ R-Sq = 0.04, ΔF = 4.80, p = 0.003).

Mediation analyses

As in Study 1, we conducted mediation analyses to assess the extent to which the relationships between Dark Tetrad variables and our outcome variables (Support for Anti-Democratic Actions and Violent Extremism) could be explained by Social Dominance Orientation or Right Wing Authoritarianism. While Right Wing Authoritarianism was not a significant mediator for the relationship between the overall Dark Tetrad variable and Support for Anti-Democratic Actions (Standardized Indirect Effect = −0.01, Bootstrapped Confidence Interval = −0.003, 0.11) or Violent Extremism (Standardized Indirect Effect = −0.03, Bootstrapped Confidence Interval = −0.001,0.07), Social Dominance Orientation was a significant mediator for both. As shown in Table 7, the relationship between the overall Dark Tetrad variable and both Support for Anti-Democratic Actions and Violent extremism is partially explained by variance in Social Dominance Orientation.

Table 7. Results from significant mediation models (Study 2 Only, N = 248 Republicans).

PathPredictor variableOutcome variable*B*BetaSE95% CI*t**p*
aDark Tetrad OverallSDO0.200.180.070.06, 0.342.900.004
bSDOSupport for Anti-Democratic Actions0.360.260.080.20, 0.524.41< 0.001
c'Dark Tetrad OverallSupport for Anti-Democratic Actions0.390.260.090.22, 0.584.38< 0.001
Bootstrapped Indirect EffectDark tetrad overallSupport for anti-democratic actions0.070.050.030.01, 0.15
aDark tetrad overallSDO0.200.180.070.06, 0.340.290.004
bSDOViolent extremism0.240.220.060.12, 0.373.81< 0.001
c'Dark tetrad overallViolent extremism0.510.410.070.37, 0.647.20< 0.001
Bootstrapped Indirect EffectDark tetrad overallViolent extremism0.050.040.030.01, 0.11

Moderation analyses

Finally, following previous research, we tested for interactions between Dark Tetrad variables and Needs Fulfillment using a series of moderation analyses. The interaction between the overall Dark Tetrad variable and Needs Fulfillment on Support for Anti-Democratic Actions was not significant (B = −0.07, t = −0.56, p = 0.575). All interactions tested between individual Dark Tetrad subcomponents and Needs Fulfillment on Support for Anti-Democratic Actions were not significant.

However, we did find a significant interaction between the overall Dark Tetrad variable and Needs Fulfillment on Violent Extremism (B = −0.26, t = −2.64, p = 0.009), such that the more participants felt their needs were being fulfilled, the weaker the relationship between the overall Dark Tetrad variable and Violent Extremism was. We also found significant interactions between Needs Fulfillment and Machiavellianism (B = −0.28, t = −3.21, p = 0.002) and Psychopathy (B = −0.27, t = −2.82, p = 0.005) on Violent Extremism (but not for Sadism or Narcissism), with the same pattern of effects: as Needs Fulfillment increased, Machiavellianism and Psychopathy had weaker relationships with Violent Extremism. Simple slopes analyses revealed that the relationships between dark personality variables and Violent Extremism were stronger when individuals were low (-1 SD) for Needs Fulfillment (Overall Dark Personality: B = 0.73, t = 7.90, p < 0.001; 95% CI = 0.55, 0.91; Machiavellianism: B = 0.66, t = 7.21, p < 0.001, 95% CI = 0.48, 0.85; Psychopathy: B = 0.65, t = 7.39, p < 0.001, 95% CI = 0.47, 0.82) and were weaker for those who scored high (+1 SD) on Needs Fulfillment (Overall Dark Personality: B = 0.37, t = 3.82, p = 0.002, 95% CI = 0.18, 0.57; Machiavellianism: B = 0.28, t = 3.50, p = 0.001, 95% CI = 0.12, 0.44; Psychopathy: B = 0.29, t = 3.16, p = 0.002, 95% CI = 0.11, 0.47). See Tables 8 and 9 below for the full results of the moderation analyses.

Table 8. Results from significant moderation models (study 2 only, N= 248 republicans).

Predictor Variable*B*SE95% CI*t**p*
Dark tetrad overall1.810.480.87, 2.753.80< 0.001
Needs fulfillment0.490.33−0.16, 1.151.480.140
Dark tetrad overall x needs fulfillment−0.260.10−0.46, −0.07−2.640.009
Machiavellianism1.840.440.98, 2.714.20< 0.001
Needs fulfillment0.620.33−0.04, 1.271.860.065
Machiavellianism x needs fulfillment−0.280.09−0.46, −0.11−3.210.002
Psychopathy1.740.450.85, 2.633.85< 0.001
Needs fulfillment0.480.30−0.10, 1.071.600.105
Psychopathy x needs fulfillment−0.270.09−0.45, −0.08−2.820.005

Outcome= violent extremism.

Table 9. Results from non-significant moderation models.

Predictor variables*B*SE95% CI*t**p*
Outcome variable= anti-democratic in study 1 (*N* = 280 Democrats)
Dark Tetrad Overall0.280.38−0.48, 1.030.720.471
Needs Fulfillment−0.450.26−0.97, 0.06−1.730.085
Dark Tetrad Overall x Needs Fulfillment0.020.08−0.13, 0.180.30.762
Outcome variable= violent extremism in study 1 (*N* = 280 Democrats)
Dark tetrad overall0.610.4−0.19, 1.401.510.133
Needs fulfillment−0.330.28−0.87, 0.22−1.170.243
Dark Tetrad Overall x Needs Fulfillment−0.030.08−0.19, 0.14−0.30.761
Outcome variable = anti-democratic in study 2 (*N* = 248 Republicans)
Dark tetrad overall0.830.65−0.44, 2.111.290.198
Needs fulfillment0.10.45−0.79, 0.990.230.822
Dark tetrad overall x needs fulfillment−0.080.14−0.34, 0.19−0.560.575

Discussion

Implications for DEVP research

Generally, the findings presented support the DEVP as a theory. In both samples, an overall dark personality trait variable, as well as some specific dark personality traits, were associated with increased support for anti-democratic actions and violent extremism, but commitment to prosocial ideological values (social justice for Democrats and family values for Republicans) were not. Specifically, general dark personality, Machiavellianism, and sadism were each associated with increased support for anti-democratic actions and violent extremism for Democrats, while general dark personality, psychopathy, and sadism were each associated with increased support for anti-democratic actions and violent extremism for Republicans. Thus, Hypotheses 1, 3, 5, and 6, were supported, but Hypotheses 2 and 4 (those predicting narcissism would be positively correlated with support for anti-democratic actions and violent extremism) were not. Taken together, the results support the notion that activism, including activism that is anti-democratic or violent, is more attractive to those higher in dark personality traits because they seek to use it as a vehicle for satisfying dark ego-needs rather than advancing a prosocial cause.

These findings have several additional implications for related research. One finding that adds to the complexity of DEVP research (and research on the role of dark personality traits on violent extremism in general), is that sadism, rather than narcissism, was the individual dark personality trait most consistently (across both studies) and strongly associated with increased support for anti-democratic and violent forms of activism in our primary analyses (multiple regression). This is potentially important, because much of DEVP research has focused on the role of narcissism, specifically antagonistic narcissism, a sub-trait of narcissism characterized by entitlement, arrogance, lack of empathy, reactive anger, distrust, thrill seeking, exploitativeness, and manipulativeness (Miller et al., 2016), with many DEVP studies studying only antagonistic narcissism, but not other dark personality traits. Many of the features of antagonistic narcissism have explicit overlap with other dark personality traits (e.g., lack of empathy and thrill-seeking with psychopathy; exploitativeness and manipulativeness with Machiavellianism), so it is possible that this variable shares features with a general measure of dark personality or dark triad traits. It is also key to note that the majority of DEVP research has found connections between narcissism in peaceful activism (e.g., Krispenz and Bertrams, 2023b), but has not explored the connection between narcissism and support for illegal, anti-democratic, or violent behaviors for one's cause (with the sole exception of Krispenz and Bertrams, 2025). It is therefore possible that narcissism is a better predictor of peaceful activism (i.e., people higher in narcissism do peaceful activism for positive attention from others), than anti-democratic or violent activism, which may be viewed negatively by others (something narcissists naturally want to avoid). However, key DEVP contributors Krispenz and Bertrams (2025) recently conducted research in which they found that higher antagonistic narcissism was related to greater intentions to commit illegal or violent activism for a cause (in this case, radicalism intentions for pro-choice and pro-life causes), although these analyses did include other dark personality traits, such as sadism, as covariates. In any case, general narcissism, as measured in our samples, was not associated with support for anti-democratic or violent activism in multiple regression.

Instead, across our two ideologically opposed samples, general dark personality trait and sadism stood out as the consistently strongest dark personality traits predicting support for anti-democratic actions and violent extremism. Sadism has been specifically identified in previous research as a correlate of violent extremism (Bélanger et al., 2023), but has not been previously implicated in DEVP research. The only (as far as we are aware) DEVP study assessing support for violent activism (Krispenz and Bertrams, 2025) did not measure the relationship between sadism (or psychopathy, Machiavellianism, or a general dark personality trait) and illegal or violent activism for a cause. Thus, the results presented here extend previous research and tentatively suggest that while an overall dark personality trait likely predicts activism of any sort for a cause, narcissism may be a stronger predictor of peaceful activism for a cause (satisfying dark ego needs for positive attention from others for doing social acceptable activism), whereas sadism may be a stronger predictor of violent forms of activism (satisfying dark ego needs of causing pain to others, especially when using the moral neutralization that they “deserve it”). Previous work has shown that uniquely sadism contributes to aggressive behavior, over and above the effects of general aggressiveness and other dark personality traits, possibly due to the fact that sadism is associated with greater positive emotion during aggression (Chester et al., 2019). In a US political context, one might imagine how members of each party satisfy this dark ego need by rejoicing when the other party publicly blunders or suffers a setback, consistent with general research on political schadenfreude and related constructs (e.g., Combs et al., 2009). However, this is an early finding and will require much more testing.

Another interesting finding from the results with possible implications is that, in most multiple regression models tested, there was no demonstrable increase in predictive value by assessing specific dark personality variables (i.e., all four dark tetrad variables) compared to only one, overall, dark personality variable, with the exception of models assessing violent extremism among Republicans (Study 2). It may be that studying dark personality, as a general construct, is simple, but equally effective compared to assessing all dark tetrad traits (and their sub-traits) in understanding anti-democratic and violent activism. Parsimony dictates that this could be a beneficial approach, although there are likely to be both pros and cons to this method.

These findings also build on recent work by providing evidence that DEVP processes appear to be present in activism for both left-wing and right-wing ideologies. Until recently, DEVP processes had only been demonstrated in left-wing ideological groups, but in a 2025 study, Krispenz and Bertams showed evidence of DEVP processes in pro-life activism (a generally right-wing ideology). The present study shows further evidence of DEVP processes in an explicitly right-wing sample (US Republicans), which adds to the evidence base that DEVP occurs across the proverbial aisle.

Another finding with implications for DEVP theory is that the majority of mediation models tested showed that neither SDO nor authoritarianism mediated the relationships between dark personality traits and support for anti-democratic actions and violent extremism. In addition, dark personality traits retained predictive relevance for outcome variables when including SDO and authoritariansism. Generally, these results show support for Hypotheses 7 and 8 and suggest that dark personality traits make unique contributions to support for anti-democratic and violent forms of activism. However, one mediation model was significant, such that the relationship between the overall dark personality trait and support for anti-democratic actions and violent extremism were both partially explained by variance in social dominance orientation (SDO) in the Republican sample. Perhaps hierarchical beliefs about social groups is an unavoidable manifestation of higher dark personality traits among conservatives who support anti-democratic and violent forms of activism. Since SDO is predicated on the idea that some groups are better than others and deserve power over others, SDO may provide a path by which those with dark personalities can easily justify support for treating others unfairly or aggressively.

Finally, while commitment to prosocial values were not associated with greater support for anti-democratic actions and violent extremism, in the Republican sample, multiple regression analyses showed that family values commitment was associated with lower violent extremism (this was not the case for social justice commitment among the Democrat sample). Though this is a small effect and ought to be interpreted cautiously, it may suggest that commitment to the strength and stability of a traditional family unit could be protective against endorsement of violence for the Republican cause. This could imply that, while commitment to the prosocial elements of one's causes tend not to be associated with greater support for violent activism (in line with DEVP theory), some prosocial elements of a cause may sometimes serve as starting points for narratives designed to prevent violent extremism.

Implications for dark personality traits and violent extremism

Recent research has highlighted the potential importance of personality factors in violent extremism (Obaidi et al., 2025), and several previous studies have specifically shown relationships between dark personality traits and violent extremism (Bélanger et al., 2023; Nai and Young, 2024; Pavlović, 2024; Pavlović and Franc, 2023; Pavlović and Wertag, 2021; Wolfowicz et al., 2021). Most researchers who study violent extremism tend to agree that dark personality traits do not explain violent extremism, and evidence bears this idea out; there are many factors that contribute to violent extremism and there is no single personality variable, or set of personality variables, that will accurately reveal who becomes a violent extremist and who does not (Horgan, 2005; Corner et al., 2018). However, this has also led some to dismiss the idea that psychopathy or other dark personality traits play any consistent role in violent extremism, but an increasingly large evidence base is forming to suggest that perspective is also reductive and out of touch with existing evidence. The results of the present studies add more fuel to this fire— dark personality traits were consistently linked to both greater support of anti-democratic actions and violent extremism across two opposing ideological groups. Thus, one implication of this research is that, although the vast majority of people who engage in violent extremism do not have a clinically diagnosable disorder related to dark personality traits (i.e., we cannot conveniently label them all “psychopaths”), the data increasingly supports the idea that dark personality traits, measured as continuous variables, consistently correlate with violent extremism.

Some research has also identified interactions between dark personality traits and motivational variables that are known to be key in radicalization processes (Bélanger et al., 2023; Pavlović and Franc, 2023). We tested for similar interactions between dark personality traits and fulfillment of basic psychological needs as outlined by Self-Determination Theory (Deci and Ryan, 1985), revealing mixed results. Our moderation analyses produced largely null results, with no significant interactions between dark personality and needs fulfillment on support for anti-democratic actions in either Democrats or Republicans, showing that Hypothesis 9 was not supported. However, results for Hypothesis 10 were more nuanced; low needs fulfillment did intensify the relationship between dark personality traits and violent extremism for Republicans, but not Democrats.

The fact that, in the Republican sample, the effects of dark personality traits on violent extremism were compounded by the frustration of one's basic psychological needs for autonomy, competence, and relatedness is in line with previous research (Bélanger et al., 2023; Pavlović and Franc, 2023). However, like past research, our results are complex. For instance, Bélanger et al. (2023) found that obsessive passion for one's ideology, which is theorized to stem from the frustration of basic psychological needs (Bélanger, 2021), interacted only with sadism, but not other dark personality traits, in its effects on violent extremism in three samples (Republicans, Democrats, and Black Lives Matter activists). Pavlović and Franc (2023) found similar interactions, but these, too, were limited. They tested an overall dark personality variable, but no specific dark personality traits, which interacted with relative group deprivation (the attitude that a social group one belongs to does not have its needs being met compared to other social groups) to predict violent extremism. In their study, this effect was limited to a sample of Croatian college students who did not identify any specific ideological group or cause. While the pattern of evidence across studies shows some support for the idea that dark personality traits interact with motivational variables to contribute to violent extremism, more testing is necessary. In addition, our results indicate that this interaction does not take place for processes leading to anti-democratic, but not necessarily violent, actions.

Limitations

This study has several limitations that warrant discussion. First, the data collected is entirely self-reported. While we took precautions to ensure data quality (e.g., the inclusion of attention checks and social desirability measures), the response biases endemic to self-report data can never be totally controlled. In addition, the sample collected was not a random sample and was not designed to be representative of the US population, but rather to approximate demographics for each political party in the US. However, the demographics of our sample do not precisely match the demographics of the respective parties (Democratic Party in Study 1, Republican Party in Study 2). In particular, the proportion of male and female participants in both our samples had more female participants than is representative of either party (See Pew Research Center, 2024). Following previous research on gender, dark personality traits, and violent extremism (Pavlović, 2024), future research ought to assess the interplay of gender and politcal affiliation on DEVP variables. We also neglected to account for religious affiliation and religiosity, though these may also be worth exploring in relation to DEVP variables.

Another weakness of this research is that the screening employed to detect socially desirable responding resulted in dropping many potential participants—about 24% across both studies. This constitutes a limitation, as this may have skewed the nature of the remaining sample such that many participants higher in narcissism may have been selectively excluded, as they likely to have been particularly susceptible to socially desirable responding.

A potential measurement limitation is that there is some conceptual and measurement overlap between authoritarianism and support for anti-democratic behaviors. For instance, authoritarianism and its sub-facets, such as top-down censorship, are inherently anti-democratic. Although these variables have been commonly measured together in previous work, it is possible that there is overlap in the measurement of these constructs which may muddy the waters of our results and this issue should be taken into account and temper the strength of conclusions drawn.

An additional limitation is that this research is neither longitudinal nor experimental and, thus, cannot give any information as to change in key variables over time nor any causal information regarding the relations between predictor variables (e.g., dark personality traits, commitment to prosocial values) and outcome variables. Finally, it is worth mentioning that these studies have limited external validity; they are conducted only in the context of US political partisanship and applications to other democracies worldwide is limited. Likewise, the findings should not be directly applied in non-democratic contexts.

Conclusion

Recent events continue to emphasize the importance of understanding how and why people justify and commit anti-democratic actions and violent extremism. In the US, this question feels particularly urgent. With nearly a third of Americans in both parties supporting at least one anti-democratic action (Polarization Research Lab, 2025) and new examples of ideological violence seemingly every week, most recently (at the time of this writing) a sniper attacking an ICE facility in Dallas, Texas (Goodman et al., 2025), it is critical for the US to strengthen its commitment to the democratic process, perhaps by understanding what factors undermine it. The findings of these studies suggest that regardless of party affiliation, those who use ideological causes to satisfy dark ego needs are also more likely to endorse anti-democratic and violent ideological actions.

Correction note

This article has been corrected with minor changes. These changes do not impact the scientific content of the article.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Manuel Galán, Catholic University San Antonio of Murcia, Spain

Reviewed by: Alex Bertrams, University of Bern, Switzerland

Jonathan Frederick Bassett, Lander University, United States

Leonard S Newman, Syracuse University, United States

Footnotes

This item may have been particularly impactful in both studies, as data collection took place in late August 2025 and Texas's redistricting conflict and subsequent lawmaker walkout took place in early August 2025.

Statements

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding

The author(s) declared that financial support was received for this work and/or its publication. Support for this research was provided through a Florida Gulf Coast University College of Arts and Sciences Professional Development Award.

Data availability

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/supplementary material.

Ethics approval

The studies involving humans were approved by Florida Gulf Coast University IRB. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

Author contributions

DS: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Writing – original draft, Writing – review & editing. NS: Conceptualization, Investigation, Methodology, Writing – original draft, Writing – review & editing.

References

  1. Adorno T. W., Frenkel-Brunswik E., Levinson D. J., Sanford R. N. (1950). The Authoritarian Personality. New York, NY: Harper and Row.
  2. Altemeyer B. (1996). The Authoritarian Specter. Cambridge, MA: Harvard University Press.
  3. Altemeyer B. (1998). The other “Authoritarian personality.” in Advances in Experimental Social Psychology, ed. M. P. Zanna (San Diego: Academic Press), 47 – 92. doi: doi:10.1016/S0065-2601(08)60382-2
  4. Altemeyer B. (2006). The Authoritarians. Available online at: (Accessed August 30, 2025).
  5. Bélanger J. J. (2021). The sociocognitive processes of ideological obsession: review and policy implications. Philos. Transact. Royal Soc. B 376: 20200144. doi: doi:10.1098/rstb.2020.0144
  6. Bélanger J. J., Adam-Troian J., Quimpo N., AlKindi Y., Gajić M., Nisa C. F. (2023). The dark tetrad personality traits moderate the relationship between ideological passion and violent activism. Psychol. Viol. 13: 43. doi: doi:10.1037/vio0000414
  7. Bertrams A., Krispenz A. (2023a). Dark-ego-vehicle principle: Narcissism as a predictor of anti-sexual assault activism. Curr. Psychol. 43, 3585 – 3598. doi: doi:10.1007/s12144-023-04591-4
  8. Bertrams A., Krispenz A. (2023b). Involvement in LGBQ activism is related to pathological narcissistic grandiosity and virtue signaling. SSRN Electr. J. doi: doi:10.2139/ssrn.4204911
  9. Bertrams A., Krispenz A. (2024). Antisemitism as a dark-ego vehicle. Curr. Psychol. 44, 676 – 692. doi: doi:10.1007/s12144-024-07120-z
  10. Bertrams A., Krispenz A. (2025). Measuring Antisemitism in Austria, Germany, and Switzerland. SSRN doi: doi:10.2139/ssrn.5457236
  11. Bier D. J. (2025). One in Five ICE Arrests are Latinos on the Streets with No Criminal Past or Removal Order. The Cato Institute. Available online at: (Accessed August 30, 2025).
  12. Broadwater L. (2025). Trump Says Critical Coverage of Him is ‘Really Illegal.' The New York Times. Available online at: (Accessed August 30, 2025).
  13. Buckels E. E., Jones D. N., Paulhus D. L. (2013). Behavioral confirmation of everyday sadism. Psychol. Sci. 24, 2201 – 2209. doi: doi:10.1177/0956797613490749
  14. Chester D. S., DeWall C. N., Enjaian B. (2019). Sadism and aggressive behavior: inflicting pain to feel pleasure. Personal. Soc. Psychol. Bull. 45, 1252 – 1268. doi: doi:10.1177/0146167218816327
  15. Cichocka A., Dhont K., Makwana A. P., Back M. (2017). On self–love and outgroup hate: opposite effects of narcissism on prejudice via social dominance orientation and right–wing authoritarianism. Eur. J. Personal. 31, 366 – 384. doi: doi:10.1002/per.2114
  16. Cleese J. (1987). The Advantages of Extremism.” Party Political Broadcast by the Social Democratic Party and Liberal Party/Liberal Alliance. Available online at: (Accessed August 30, 2025)..
  17. Combs D. J., Powell C. A., Schurtz D. R., Smith R. H. (2009). Politics, schadenfreude, and ingroup identification: the sometimes happy thing about a poor economy and death. J. Exp. Soc. Psychol. 45, 635 – 646. doi: doi:10.1016/j.jesp.2009.02.009
  18. Conway L. G., Houck S. C., Gornick L. J., Repke M. A. (2017). Finding the loch ness monster: left-wing authoritarianism in the United States. Polit. Psychol. 39, 1049 – 1067. doi: doi:10.1111/pops.12470
  19. Corner E., Gill P., Schouten R., Farnham F. (2018). Mental disorders, personality traits, and grievance-fueled targeted violence: the evidence base and implications for research and practice. J. Personal. Assess. 100, 459 – 470. doi: doi:10.1080/00223891.2018.1475392
  20. Costello T. H., Bowes S. M., Stevens S. T., Waldman I. D., Tasimi A., Lilienfeld S. O. (2022). Clarifying the structure and nature of left-wing authoritarianism. J. Personal. Soc. Psychol. 122, 135 – 170. doi: doi:10.1037/pspp0000341
  21. Crawford J. T., Brandt M. J. (2020). Ideological (a)symmetries in prejudice and intergroup bias. Curr. Opin. Behav. Sci. 34, 40 – 45. doi: doi:10.1016/j.cobeha.2019.11.007
  22. de Oliveira Santos D., Jost J. T. (2024). Liberal-conservative asymmetries in anti-democratic tendencies are partly explained by psychological differences in a nationally representative US sample. Commun. Psychol. 2: 61. doi: doi:10.1038/s44271-024-00096-3
  23. Deci E. L., Ryan R. M. (1985). Intrinsic Motivation and Self-Determination in Human Behavior. New York, NY: Plenum. doi: doi:10.1007/978-1-4899-2271-7
  24. Deci E. L., Ryan R. M. (2000). The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychol. Inq. 11, 227 – 268. doi: doi:10.1207/S15327965PLI1104_01
  25. Department of Justice (2025a). Albuquerque Man Charged in Connection with Arson Attacks on Tesla Dealership and Republican Party of New Mexico Office. Available online at: (Accessed August 30, 2025).
  26. Department of Justice (2025b). Vance Boelter Indicted for the Murders of Melissa and Mark Hortman, the Shootings of John and Yvette Hoffman, and the Attempted Shooting of Hope Hoffman. Available online at: (Accessed August 30, 2025).
  27. Duckitt J., Sibley C. G. (2007). Right wing authoritarianism, social dominance orientation and the dimensions of generalized prejudice. Eur. J. Personal. 21, 113 – 130. doi: doi:10.1002/per.614
  28. Faul F., Erdfelder E., Buchner A., Lang A. G. (2009). Statistical power analyses using GPower 3.1: Tests for correlation and regression analyses. Behav. Res. Methods 41, 1149 – 1160. doi: doi:10.3758/BRM.41.4.1149
  29. Furnham A., Richards S. C., Paulhus D. L. (2013). The dark triad of personality: a 10 year review. Soc. Personal. Psychol. Compass 7, 199 – 216. doi: doi:10.1111/spc3.12018
  30. Goodman J. D., Aleaziz H., Fortin J., Karlamangla S. (2025). Dallas ICE Office Shooting Updates: Detainee Killed and 2 Others Wounded. New York Times. Available online at: (Accessed August 30, 2025).
  31. Gousse-Lessard A. S., Vallerand R. J., Carbonneau N., Lafrenière M. A. K. (2013). The role of passion in mainstream and radical behaviors: a look at environmental activism. J. Environ. Psychol. 35, 18 – 29. doi: doi:10.1016/j.jenvp.2013.03.003
  32. Greene J. (2023). Anti-democratic Politics in the US: Research Starters: EBSCO Research. EBSCO. Available online at: (Accessed August 30, 2025).
  33. Haghighat R. (2007). The development of the brief social desirability scale (BSDS). Eur. J. Psychol. 3, 10 – 5964. doi: doi:10.5964/ejop.v3i4.417
  34. Ho A. K., Sidanius J., Kteily N., Sheehy-Skeffington J., Pratto F., Henkel K. E.. (2015). The nature of social dominance orientation: theorizing and measuring preferences for intergroup inequality using the new SDO scale. J. Personal. Soc. Psychol. 109, 1003 – 1028. doi: doi:10.1037/pspi0000033
  35. Hodson G., Hogg S. M., MacInnis C. C. (2009). The role of “Dark personalities” (narcissism, machiavellianism, psychopathy), big five personality factors, and ideology in explaining prejudice. J. Res. Personal. 43, 686 – 690. doi: doi:10.1016/j.jrp.2009.02.005
  36. Holliday D. E., Iyengar S., Lelkes Y., Westwood S. J. (2024). Uncommon and nonpartisan: Antidemocratic attitudes in the American public. Proceed. Nat. Acad. Sci. 121: 2313013121. doi: doi:10.1073/pnas.2313013121
  37. Horgan J. (2005). The Psychology of Terrorism. Milton Park: Routledge. doi: doi:10.4324/9780203496961
  38. Jahangir M., Shah S. M., Zhou J., Lang B., Wang X.-P. (2024). Machiavellianism: Psychological, clinical, and neural correlations. J. Psychol. 159, 155 – 168. doi: doi:10.1080/00223980.2024.2382243
  39. Johnson A. (2025). “ Trump's Assault on the First Amendment.” International Bar Association. Available online at:
  40. Krispenz A., Bertrams A. (2023a). Further basic evidence for the dark-ego-vehicle principle: higher pathological narcissism is associated with greater involvement in feminist activism. Curr. Psychol. 43, 14619 – 14633. doi: doi:10.1007/s12144-023-05451-x
  41. Krispenz A., Bertrams A. (2023b). Understanding involvement in environmental activism: Relationships to pathological narcissism, virtue signaling, dominance, and sensation seeking. SSRN Electron. J. doi: doi:10.2139/ssrn.4560799
  42. Krispenz A., Bertrams A. (2023c). Understanding left-wing authoritarianism: Relations to the dark personality traits, altruism, and social justice commitment. Curr. Psychol. 43, 2714 – 2730. doi: doi:10.1007/s12144-023-04463-x
  43. Krispenz A., Bertrams A. (2025). Ideological impartiality of the dark-ego-vehicle principle: antagonistic narcissism is associated with pro-choice as well as pro-life radicalism intention. SSRN 5405946. doi: doi:10.2139/ssrn.5405946
  44. Kteily N. S., Sidanius J., Levin S. (2011). Social dominance orientation: cause or ‘mere effect'? J. Exp. Soc. Psychol. 47, 208 – 214. doi: doi:10.1016/j.jesp.2010.09.009
  45. Miller J. D., Lynam D. R., McCain J. L., Few L. R., Crego C., Widiger T. A.. (2016). Thinking structurally about narcissism: An examination of the five-factor narcissism inventory and its components. J. Personal. Disord. 30, 1 – 18. doi: doi:10.1521/pedi_2015_29_177
  46. Moore-Berg S. L., Ankori-Karlinsky L. O., Hameiri B., Bruneau E. (2020). Exaggerated meta-perceptions predict intergroup hostility between American political partisans. Proceed. Nat. Acad. Sci. 117, 14864 – 14872. doi: doi:10.1073/pnas.2001263117
  47. Nai A., Young E. L. (2024). They choose violence. Dark personality traits drive support for politically motivated violence in five democracies. Personal. Individ. Diff. 230: 112794. doi: doi:10.1016/j.paid.2024.112794
  48. Obaidi M., Bergh R., Benningstad N. C. G., Dovidio J. F. (2025). Individual differences in violent extremism. Nat. Rev. Psychol. 5 29–46. doi: doi:10.1038/s44159-025-00509-y
  49. Paulhus D. L., Buckels E. E., Trapnell P. D., Jones D. N. (2021). Screening for dark personalities the short dark tetrad (SD4). Eur. J. Psychol. Assess. 37, 208 – 222. doi: doi:10.1027/1015-5759/a000602
  50. Paulhus D. L., Dutton D. G. (2016). “Everyday sadism,” in The Dark Side of Personality: Science and Practice in Social, Personality, and Clinical Psychology, eds. Zeigler-Hill V. and Marcus D. K. (American Psychological Association), 109 – 120. doi: doi:10.1037/14854-006
  51. Paulhus D. L., Williams K. M. (2002). The dark triad of personality: Narcissism, machiavellianism, and psychopathy. J. Res. Personal. 36, 556 – 563. doi: doi:10.1016/S0092-6566(02)00505-6
  52. Pavlović T. (2024). Gender and dark personality traits as determinants of collective action participation: a view from croatia. Anali hrvatskog politološkog društva 21, 65 – 91. doi: doi:10.20901/an.21.10
  53. Pavlović T., Franc R. (2023). Antiheroes fueled by injustice: dark personality traits and perceived group relative deprivation in the prediction of violent extremism. Behav. Sci. Terror. Polit. Aggress. 15, 277 – 302. doi: doi:10.1080/19434472.2021.1930100
  54. Pavlović T., Wertag A. (2021). Pro-violence as a mediator in the relationship between the dark personality traits and support for extremism. Personal. Individ. Diff. 168: 110374. doi: doi:10.1016/j.paid.2020.110374
  55. Pew Research Center (2024). “ Partisanship by Gender, Sexual Orientation, Marital and Parental Status ”. Available online at: (Accessed August 30, 2025).
  56. Polarization Research Lab (2025). “ Why Polarization Threatens Democracy.” Available online at: (Accessed August 30, 2025).
  57. Pratto F., Sidanius J., Stallworth L. M., Malle B. F. (1994). Social dominance orientation: a personality variable predicting social and political attitudes. J. Personal. Soc. Psychol. 67, 741 – 763. doi: doi:10.1037/0022-3514.67.4.741
  58. Qin X. (2024). Sample size and power calculations for causal mediation analysis: a tutorial and shiny app. Behav. Res. Methods 56, 1738 – 1769. doi: doi:10.3758/s13428-023-02118-0
  59. Rapoport D. (2004). The four waves of modern terrorism in “ Attacking Terrorism: Elements of a Grand Strategy,” A. K. Cronin and M. L. James (Washington, DC: Georgetown University Press)
  60. Torres-Harding S. R., Siers B., Olson B. D. (2011). Development and psychometric evaluation of the Social Justice Scale (SJS). Am. J. Commun. Psychol. 50, 77 – 88. doi: doi:10.1007/s10464-011-9478-2
  61. Whitley B. E. (1999). Right-wing authoritarianism, social dominance orientation, and prejudice. J. Personal. Soc. Psychol. 77, 126 – 134. doi: doi:10.1037/0022-3514.77.1.126
  62. Wolfowicz M., Litmanovitz Y., Weisburd D., Hasisi B. (2021). Cognitive and behavioral radicalization: A systematic review of the putative risk and protective factors. Campbell Syst. Rev. 17: e1174. doi: doi:10.1002/cl2.1174
  63. Zeigler-Hill V., Sauls D., Malay P. (2020). Through the eyes of narcissus: Competitive social worldviews mediate the associations that narcissism has with ideological attitudes. Self Ident. 20, 811 – 840. doi: doi:10.1080/15298868.2020.1779118

Citation

Snook, D. W. & Skubala, N. (2026). Censorship for thee, but not for me: dark personality traits are associated with support for anti-democratic actions and violent extremism among US political partisans. Frontiers in Social Psychology, Volume 4. https://doi.org/
Received
Accepted
Published
Volume
4 - 2026

Copyright

Copyright © 2026 Snook and Skubala.

License: https://creativecommons.org/licenses/by/4.0/

*Correspondence: Daniel W. Snook

Article metrics

Views
0
Downloads
0
Citations
0
Shares
0

Related articles

Original ResearchHow does belief in conspiracy theories affect loneliness and connectedness? A longitudinal investigationMatt N. Williams, John R. Kerr, Mathew D. Marques, Haneen Abu-Obeid, Stephen R. HillOriginal Research“They got what they deserved”: situational justification and schadenfreude toward an LGBTQIA+ organizationIvan Giuseppe Cammarata, Costanza Scaffidi Abbate, Martina Macrillò, Stefano BocaOriginal ResearchAnother one bites the dust? Threatened meat-eater identity and veg * n conspiracy beliefsDominika Adamczyk, Natalia Korczak, Marta Marchlewska