The Dark Triad traits predict authoritarian political correctness and alt-right attitudes

Abstract

It is well established that mainstream personality traits are associated with moderate, traditional political attitudes. However, very little is known regarding trait predictors of extreme political attitudes. In the current study (N = 511 U.S. residents), we examined the relationships between the Dark Triad traits, Entitlement and three extreme political attitudes that are highly covered in mainstream media: White Identitarianism (‘Alt-Right’), Political Correctness-Authoritarianism, and Political Correctness-Liberalism. We found that Dark Triad traits and Entitlement had incremental validity in the prediction of these 3 attitudes over demographic factors. The Dark Triad traits and Entitlement explained a substantial portion of variance in White Identitarianism and Political Correctness-Authoritarianism, and only a small portion of variance in Political Correctness-Liberalism. Across all attitudes, Psychopathy and Entitlement were the most consistent, strongest predictors. Results indicate that, from a Dark Triad perspective, Authoritarian PC advocates have more in common with extreme right advocates than those holding PC views related to compassion.

Psychology; Applied Psychology; Individual Differences; Personality Assessment; Political Behavior; political correctness, dark triad, alt right, extremism, political attitudes, entitlement.

1. Introduction

An extensive literature has established that personality traits are associated with a range of political attitudes and behaviors. The most reliable findings relate to openness to experience and conscientiousness; these traits predict political left and political right political orientation respectively (Furnham & Fenton-O’Creevy, 2018; Sibley et al., 2012) as well as political knowledge and participation (Mondak and D Halperin, 2008). More limited research has found seemingly robust effects of other traits. For example, individuals high in trait agreeableness are more likely to identify with the political left (Furnham & Fenton-O’Creevy, 2018) and report being less inclined to engage in political protest (Brandstatte and Opp, 2014); while individuals high in extraversion are more likely to express an interest in politics (Leone et al., 2012).

The majority of research on personality traits and political constructs has focused primarily on mainstream political attitudes and behaviours. These studies often use unidimensional measures of left-right political orientation (e.g. Furnham & Fenton-O’Creevey, 2018) or simple two-dimensional measures of liberalism and conservatism (e.g. Jonason, 2014). Whilst informative, these studies tell us little about important political attitudes falling outside of the traditional continuum (see Layman and Carsey, 2002). This is particularly relevant today, with increasing reports of the alternate or ‘regressive left’, and the alternate or ‘Alt-Right” (AR; Hawley, 2017; Caldwell, 2016). Consistent with this, recent polling data suggests a population level shift from moderate political views to more extreme, non-traditional attitudes (Pew Research Center, 2014).

In the current paper, we focus on three sets of extreme political attitudes. These include two forms of Political Correctness typical of the ‘regressive left’: Political Correctness-Authoritarianism (PCA) and Political Correctness-Liberalism (PCL); and one form of White Identitarianism (WI) typical of the AR. Here, we define Political Correctness as a set of related attitudes characterised by the desire to avoid offense and avoid disadvantaging certain groups in society (Andary-Brophy, 2015). While both forms of PC attitudes are centred around compassionate motivations, this desire for compassion is most evident in PCL. Those holding PCL attitudes – or belonging to groups characterised by PCL attitudes - argue for the removable of ostensible social or emotional barriers of disadvantaged groups, whereas those holding PCA attitudes are primarily concerned with physical and psychological safety (Andary-Brophy, 2015). A core aspect of PCA attitudes is the belief that aggression and force are appropriate methods to achieve ideological goals. In contrast to PC, WI is defined here as a set of political attitudes characterised by strong feelings of white identity, solidarity and a belief in white victimisation (Hawley, 2017). Our definition of WI is based on the stated core beliefs of alt-right groups (see method section for more detail). Importantly therefore, our WI construct is different from Jardina’s (2019) conception of ‘white identity’, which primarily relates to how strongly individuals identify with being ‘white’. Our WI construct is also different from ‘right-wing authoritarianism’ (RWA) which primarily relates to attitudes characterised by the willingness to follow established authority and punish those who do not. The three sets of attitudes we focus on in this paper are broadly conceptualised as ‘political attitudes’ because recent literature and political commentary have identified them as primary drivers of political activism.

These three extreme attitudes have attracted growing media coverage based largely on their increasing presence and role in highly publicised violent incidents such as the Charlottesville rally (Keneally, 2018) as well as several university campus protests regarding free speech (e.g. Thanawala, 2017). They can be regarded as emergent attitudes, because their widespread acceptance by substantial portions of the population is relatively recent. For example, the term ‘alt-right’ emerged in various online publications beginning in 2010 (e.g. Breitbart news). Similarly, the introduction of numerous PC terms in common discourse, including “social justice warriors”, “trigger warnings” and “microaggressions” is also relatively recent. To illustrate this point, we ran a Google search for the term “social justice warrior” which returned a total of 789,000 results (as at Feb, 2020). The same search limited to the years between 2000 and 2010 returned only 108 unique results.

Given the extreme nature of these political attitudes and their possible role in aggressive protest behaviors, we explore their association with the Dark Triad (DT) personality traits and trait entitlement. The DT traits include three ‘dark’ dimensions of personality (Machiavellianism, narcissism, psychopathy) that are not well captured in mainstream personality taxonomies such as the Big Five. They have been linked to a set of outcomes characterised by aggressive and/or antisocial behavior generally not well explained by the Big Five such as workplace corner-cutting (Jonason & O'Connor, 2017), domestic violence (Carton and Egan, 2017), and the desire for attention (Connor et al., 2020) among others. In terms of political attitudes/behavior, the DT and entitlement have only been measured in a couple of studies. In one study, Jonason (2014) demonstrated that psychopathy was positively related to mainstream conservativism and Machiavellianism was negatively related to liberalism. In a later study on the Dark Tetrad, Duspara and Greitemeyer (2017) found that right leaning voters and those high in political extremism tended to score high in all dark personality traits. In measuring extremism however, these authors did not measure extreme attitudes explicitly, but rather operationalised this variable in terms of absolute deviations from the mean of the univariate (left vs right) political scale.

We hypothesise that DT traits and trait entitlement will positively predict WI (H1) and PCA (H2). Although thought to reflect values situated at opposite ends of the political spectrum, both are characterised by aggression and belief that some form of violence and/or intimidation is acceptable in certain circumstances. PCL on the other hand, although an extreme political attitude, is not characterised by aggression/intimidation but rather by beliefs in the importance of emotional welfare. We hypothesise negative associations between the DT, entitlement and PCL (H3). In testing these hypotheses we control for a range of demographic factors including age, gender, education, and ethnicity and thus tested for incremental validity of our hypothesised variables/measures.

2. Method

2.1. Participants

Participants were a sample of 511 U.S. residents (243 males, 268 females) stratified according to age, gender, ethnicity, and employment status to be consistent with the broader U.S. population. A power analysis using G∗Power (Faul et al., 2009) confirmed the sample was sufficient to detect medium effect sizes (power = 0.95). The sample was recruited using the Qualtrics panel and participants received a small reward for their participation. All participants completed an online survey from a location of their choosing. Participants who completed the survey faster than one-third of the mean time were removed from the final sample (n = 8). The study was approved by the QUT Human Research Ethics Committee (approval number: 180000054) and informed consent of participants was obtained.

2.2. Measures

2.2.1. Demographics

Participants completed a set of questions regarding their age, gender, education and ethnicity. To be appropriate for regression analyses, all demographic questions were transformed to either continuous or dichotomous variables. Age was simply reported as ‘age in years’, and gender was coded as 1 (male) or 2 (female). Education was coded from 1-6 broadly reflecting years of study (1 = year 10 or below, 2 = year 12, 3 = trade/cert III/IV, 4 = diploma/associate diploma, 5 = undergraduate degree/bachelor, 6 = postgraduate qualification (Phd/Masters). Ethnicity was recoded to 1 = Caucasian vs 2 non-Caucasian. Although this is a very broad category, it was meaningful in the current study given the nature of the variables studied.

2.2.2. Dark triad and entitlement

The three DT Traits (Machiavellianism, narcissism, psychopathy) were measured using the Short Dark Triad (SD3; Jones and Paulhus, 2014). This is a widely used 27-item measure of the DT known to produce reliable and valid scores. Participants responded to questions on a scale ranging from 1 (strongly disagree) to 5 (strongly agree). In the current study all three DT variables received adequate internal reliability (Machiavellianism, α = .85; Narcissism, α = .64; Psychopathy, α = .85). Psychological entitlement was measured using the Psychological Entitlement Scale (Campbell et al., 2004). This 9 item measure achieved adequate reliability (alpha = .77) in the current study. An example item from the scale is “I honestly feel I'm just more deserving than others”. This measure uses a 7 point response scale ranging from 1 (strong disagreement) to 7 (strong agreement).

2.2.3. Political correctness authoritarianism (PCA) and political correctness liberalism (PCL)

The two forms of PC attitudes were measured using the PC scale (short version; Andary-Brophy, 2015). This 36 item questionnaire measures PCL with 19 items and PCA with 17 items. An example PCL item is “There are no biologically based differences in personality, talent, and ability to reason, between racial groups.” and example PCA item is “When a charge of sexual assault is brought forth, the alleged perpetrator should have to prove his or her innocence”. The original study utilising this measure (Andary-Brophy, 2015) demonstrated a sound factor structure for these two dimensions and adequate internal reliability. Internal reliabilities for both scales were adequate in this study (PCA, α = .86; PCL, α = .68).

2.2.4. White Identitarianism

WI was measured using the WI scale (Moss & O'Connor, 2018). The measure was designed to reflect core beliefs of those identifying with ideology espoused by alt-right figureheads. Statements reflecting alt-right sentiment were compiled from an extensive online search of alt-right websites, social media sites and interviews with alt-right figure heads and statements and only sentiments common to multiple sources were retained. Statements were centred around the importance of ethnic identity, white solidarity and the belief that whites are being displaced in the U.S. The scale consists of 12 items (e.g. “there is a progressive conspiracy against white identity") and participants responded on a 5 point scale ranging from 1 (completely disagree) to 5 completely agree. An exploratory factor analysis of these items (principal axis factoring) revealed one clear factor and the scale achieved an adequate alpha in this study (α = .88). Given the nature of this scale, analyses were only conducted on the portion of the sample identifying as ‘Caucasian’.

3. Results

Prior to running focal analyses, data were checked to ensure assumptions of regression were met. Only two outliers were identified (z > +- 3.00 on Narcissism) however deletion of these cases resulted in no substantive changes to the results and consequent conclusions, so they were retained. Tests for linearity, heteroscedasticity (residual plot) and univariate normality (skewness, kurtosis) revealed nothing problematic. Table 1 summarises results from three sets of hierarchical multiple regressions conducted to test hypotheses 1–3. Hypothesis 1 was tested utilising only Caucasian participants (n = 314) whereas the remaining hypotheses were tested using all participants. Consistent with H1, DT traits and entitlement had incremental validity over demographics in the prediction of WI. At step two DT traits explained a significant increment in variance R² change = .236, p < .001, with Machiavellianism the dominant bivariate and unique predictor. At step three entitlement explained a further increment in variance R² change = .015, p = .009. The inclusion of entitlement at step three reduced the effects of psychopathy and narcissism suggesting shared variance among these predictors and entitlement. Consistent with H2 a similar set of results were reported in the prediction of PCA. DT traits explained a significant increment in variance at step two R² change = .177, p < .001 as did entitlement at step three, R² change = .03, p < .001. Again the dominant, unique predictors in step three were psychopathy and entitlement. Psychopathy was also the dominant bivariate predictor. Partially consistent with H3, a significant negative relationship was found between psychopathy and PCL. This effect remained significant at step three indicating the unique importance of psychopathy on PCL scores (note however that R² change was not significant at step two or step three). The only other significant predictor of PCL was education. Interestingly, education was not a significant predictor of the other two political attitudes measured here.1

Table 1.

Results from the 3 hierarchical regressions.

Predictors	Dependent Variable
	WI		PCA		PCL
	r	β	r	β	r	Β
Step 1 – Covariates
Age	.226∗∗∗	.223∗∗	.311∗∗∗	.298∗∗	-.049	-.025
Gender	.182∗∗	.189∗∗∗	.042	.034	-.046	-.087
Education
Ethnicity	.059	.101	-.098∗	-.131∗∗	-.232∗∗∗	-.241∗∗∗
	-	-	.159∗∗∗	.068	-.066	-.015
Adj R2	.082∗∗∗		.111∗∗∗		.055∗∗∗
Step 2 – Dark Triad
Machiavellianism	.528∗∗∗	.374∗∗∗	.414∗∗∗	.184∗∗∗	-.006	.115
Narcissism	.350∗∗∗	.142∗	.412∗∗∗	.199∗∗∗	-.001	.026
Psychopathy	.477∗∗∗	.135	.459∗∗∗	.196∗∗	-.124∗∗	-.222∗∗
Adj R2	.313∗∗∗		.285∗∗∗		.071∗∗
Step 3 – Entitlement
Machiavellianism	.528∗∗∗	.281∗∗∗	.414∗∗∗	.053	-.006	.062
Narcissism	.350∗∗∗	.061	.412∗∗∗	.085	-.001	-.020
Psychopathy	.477∗∗∗	.138	.459∗∗∗	.195∗∗∗	-.124∗∗	-.222∗∗
Entitlement	.475∗∗∗	.191∗∗	.490∗∗∗	.271∗∗∗	.051	.109
Adj R2	.326∗∗∗		.314∗∗∗		.089∗∗

∗p < .05; ∗∗p < .01; ∗∗∗p < .001.

4. Discussion

The purpose of this paper was to assess whether DT traits and entitlement predict three extreme, but increasingly mainstream, sets of political attitudes. The three extreme attitudes we measured have received extensive media attention over the previous few years however, have been the focus of very little academic research. Given the largely antisocial and violent outcomes these attitudes have been linked to, we investigated their relationship with the DT personality traits and entitlement. As hypothesised, we found incremental effects of the DT traits and entitlement in predicting two of the three political attitudes. Psychopathy, which is characterised by lack of empathy and antisocial behavior was a strong DT predictor of all political attitudes. It positively predicted WI and PCA and negatively predicted PCL. This was true both when assessing bivariate correlations (which can be regarded as moderate in magnitude) as well as when assessing regression coefficients in PCA and PCL. Machiavellianism was the strongest unique predictor of WI, however the meaning of this is not clear given the known high correlations between DT traits. Whilst Narcissism had significant bivariate correlations with WI and PCA, it tended to have small and non-significant effects when controlling for entitlement at step three. Indeed entitlement was a moderate, unique predictor of both WI and PCA and the strongest trait predictor of PCA.

Interestingly, the results indicate that, although these attitudes are thought to reflect opposing ends of the traditional left (PC) vs right (WI) political spectrum, those high in WI and PCA are very similar in terms of their ‘dark’ profile. It is possible therefore that DT traits do not influence left vs right political orientation in the same manner as the Big Five (i.e. openness to experience and conscientiousness predict political left and political right political orientation respectively), but rather influence the strategies that that people use to achieve their ideological goals. For example, right and left oriented individuals high in trait psychopathy might use similar aggressive means to achieve their goals, despite such methods being inconsistent with traditional, compassionate, left-oriented values.

Our finding that high PCA and WI individuals have certain similarities is consistent with a recent study by Conway et al. (2018) who found support for a left-wing authoritarianism scale and demonstrated symmetry with the RWA scale in terms of prejudice and dogmatism. We believe our findings complement these nicely, in demonstrating how these attitudes relate to underlying dark personality traits. We note however that a key difference in our study was our use of a multidimensional measure of PC including both liberal and authoritative aspects and demonstrating different associations with DT and entitlement. Our results supported the idea that PC attitudes are multidimensional and, not surprisingly, DT traits and entitlement are more important when predicting PCA rather than PCL.

We note a few limitations that partially limit the conclusions of our study. In particular, our study was cross-sectional and contained only U.S. participants meaning we cannot make strong claims regarding causation and our results strictly cannot be generalised beyond the U.S. Additionally, many of our measures are relatively new and although we provided evidence for internal reliability, it is possible they are not precise measures of the constructs we focus on in this paper. We therefore suggest that future research confirm the validity of our measures, particularly regarding how they relate to similar constructs including LWA and RWA.

In conclusion, our study indicates that an emerging set of mainstream political attitudes – most notably PCA, WI, are largely being adopted by individuals high in the DT and entitlement. Individuals high in authoritarianism – regardless of whether the hold politically correct or right-wing views – tend to score highly on DT and entitlement. Such individuals therefore are statistically more likely than average to be higher in psychopathy, narcissism, Machiavellianism and entitlement. We found both moderate bivariate effects and unique effects (regression coefficients) and conclude that the DT and entitlement have important shared and unique effects in predicting our attitudinal outcomes.

Declarations

Author contribution statement

Jordan Moss: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Peter J. O'Connor: Conceived and designed the experiments; Performed the experiments; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Funding statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Competing interest statement

The authors declare no conflict of interest.

Additional information

No additional information is available for this paper.