The current article reviews and critiques two published studies concerning the associations between socially conservative and liberal judgments and dark personality traits. Those studies presented statistically substantial findings of associations between socially conservative judgments on a Moral Intuition Survey and “dark triad” traits, i.e., narcissism, Machiavellianism, and psychopathy. Much fewer statistically significant associations between liberal judgments and dark triad traits were found in those studies, with a greater chance of false positive results for the latter associations. The current analysis identifies limitations in the earlier studies' methodology, statistical analyses, and societal considerations that place their findings in a more nuanced context. The paper concludes with a recommendation for further research since science is an open and evolving process.

Keywords: conservative, liberal, moral judgments, dark triad traits

Two studies were published a decade ago^1,2 (and were subsequently cited in numerous publications), suggesting that socially conservative people endorse moral judgments strongly associated with dark triad (DT) personality traits (Machiavellianism, Narcissism, and Psychopathy). The term dark triad was coined by Paulhus and Williams,³ who studied moderately inter-correlated but distinct antipathetic personality traits. The three DT personality constellations are negatively related to the Big Five personality dimension of agreeableness and positively associated with maladaptive outcomes such as physical and verbal aggression, exploitative short-term mating, and dominance.³ The initial study by Arvan² found 11 highly statistically significant correlations between what was categorized as conservative moral judgments on a variety of “hot button” social issues and dark triad personality traits. No significant correlations were found between liberal moral judgments and dark triad traits. Both studies included a brief demographic survey, a moral intuition survey devised by the author of the studies, and a validated self-report research measure of dark triad traits, the Short Dark Triad.⁴ The second study was intended to replicate the original results and to address several potential weaknesses in the first study. The statistical results were impressive and consistent with the first study. Therefore, the current critique will focus primarily on the second study.

Before addressing some of the study’s shortcomings, it is noted that other recent studies also have found significant associations and greater levels of psychopathy (as measured by various self-report instruments) in people who identify as Republicans or conservatives compared to those who identify as Democrats or liberals.^5,6 Gay et al.,⁶ found that Cold-heartedness, but not Fearless Dominance or Self-Centered Impulsivity was associated with conservatism as measured on the Psychopathy Personality Inventory-Revised.⁷ Choma et al.,⁸ found empirical support for a bi-dimensional model of political ideology with moderately negatively correlated liberal and conservative factors. The authors found differential relationships between political ideology and personality features; liberalism was distinctly associated with universal orientation and creativity, whereas conservatism was associated with dogmatism. Not all studies have found significant associations between social conservatism and negative personality traits. For example, associations were found between conscientiousness and conservative political ideology in large sample studies of relationships between the big five personality traits and political ideology, which also studied a variety of contextual variables.^9,10 In those studies, one based on a large, nationally representative sample (N= 14,601) and the other based on an extensive national sample of registered voters (N=14,672), openness to experience was associated with liberal ideology; conscientiousness was associated with conservative ideology. Agreeableness was not significantly associated with self-reported political ideology,^9,10 however, Agreeableness was associated more particularly with liberal economic attitudes and conservative social attitudes, an unexpected finding.⁹ This is particularly noteworthy, given that low Agreeableness, or high Antagonism is shared with each dark triad trait.

Given the mixed results concerning personality correlates of political ideology, it becomes essential to examine contextual factors that may affect study outcomes. The current critique is not intended to dispute the results of Arvan’s^1,2 studies or those of other studies that found associations between social or political conservatism and dark traits. Instead, a review of Arvan’s studies calls attention to certain shortcomings that should be considered in interpreting the findings. In Arvan’s follow-up study, the author found that all 15 conservative judgments on the Moral Intuition Scale (MIS) correlated highly significantly with dark triad traits. In contrast, only six liberal judgments correlated significantly with Dark Triad traits. Of the latter correlations, all but one had a greater possibility of a Type 1 (false-positive) error than the significance levels of the conservative judgments.

Arvan’s¹ second study reflects several strengths. For example, the author addressed potential sample bias, finding a balance between moderately or strongly socially conservative participants and those who were self-described moderately or strongly socially liberal. The author noted additional steps to avoid potential experimenter bias and bias attributed to item wording. There was a balance between items for which higher scores reflected a socially liberal opinion and items for which higher scores reflected a socially conservative opinion. In the following sections, several psychometric and methodological issues will be addressed, with strengths and limitations indicated as relevant.

Limitations in methodology

The author stated that the MIS in the second study represented 15 social issues, reflecting a broader set of social issues than in the original research. More correctly, the MIS included 15 items, five of which concerned illegal immigration (all five of those items used the terminology “illegal immigration” or “illegal immigrants”). Two survey items concerned abortion. It is more correct to say that the MIS represented nine social issues. This may have implications for the number of significant correlations that were found. Even though each item was intended to be treated as a single-item assessment, high inter-item correlations probably resulted from the five immigration items for both conservatives and liberals; people who selected liberal or conservative judgments on one of these items likely selected liberal or conservative judgments on the remaining items. The same would likely be true of the two abortion-related items. This does not negate the significant correlations between conservative judgments and dark triad traits. However, if high inter-item correlations existed for the five immigration items, it might be more prudent to treat the five items as one variable and perhaps to use the average “score” on the 1-5 scale to be correlated with the dark triad measure. This would somewhat mitigate the number of significant correlations. An inter-rater methodology of including several psychology experts to judge the assignment of MIS items to the dichotomous categories of socially liberal or conservative was not utilized. However, for the most part, the items likely have face validity for being consistent with either socially conservative or liberal viewpoints.

Arvan^1,2 devised the MIS with no known psychometric properties to support its reliability and validity. Internal consistency reliability cannot be calculated based on a single item and is thus unknown. Neither test-retest reliability nor parallel form reliability was examined. On the other hand, the author argued that items have face and external validity. Indeed, face validity has been noted as a potential advantage of single-item measures,¹⁰ especially when items are delineated and descriptive, which appears to be the case for most MIS items. Also, single-item assessments can be tested for convergent, concurrent, and predictive validity, which was the case in Arvan’s studies, respecting correlations between MIS item judgments and dark triad traits. Still, it has been widely argued that single-item assessments have a more significant potential for measurement error.

Arvan’s Moral Intuition Survey response scale may be confusing for research participants, which might have confounded the face validity of the survey. Selecting the difference between choice 2 (“Morally bad but not forbidden” and choice 4 (“Morally good but not required) might be inherently bewildering. For example, concerning item 10 (“A woman ought to have a legal right to have an abortion upon request”), if a research participant considers abortion to be morally good, why would the participant select “morally good but not required” as opposed to “morally required”? If abortion is morally good, why would it not be legally required (permitted)? When survey items or response choices necessitate the participant to consider a compound phrase (e.g., “morally bad, but not forbidden”) they may complicate or obscure the meaning. Thus, different participants might interpret survey items differently using the response scale given.

Limitations in data analysis

Unfortunately, no direct correlation was tested between political leanings (A) and dark traits (C). That precluded testing whether moral intuition/judgment (B) mediated the relationship between political leanings/identification and dark traits. That is, do political leanings (A) have a direct effect on moral judgment (B), do political leanings have a direct effect on dark traits in the absence of (B), does B have a significant unique effect on dark traits, and does the effect of (A) on (C) become diminished when (B) is added to the model? Second, other variables, such as religious beliefs or affiliation, might mediate the relationship between political leanings and dark traits or even affect political leanings.

Potential moderating variables such as gender and age were not examined. Such variables could affect the strength of the relationship between the MIS and the dark triad. It is possible, for example, that the strength of the association between political identification and MIS judgments would be higher in socially conservative older men than in socially conservative younger women. It is acknowledged that testing for mediating and moderating variables involves a labor-intensive expenditure of time and analyses and needs to be planned at the study's outset. It only is noted that more important information might have been gleaned from the available data beyond correlational analyses.

Arvan¹ recommended that a future study could better detect the possible association between certain levels of the dark triad and actual (albeit self-reported) misconduct. Presumably, such levels of the dark triad would be associated with certain moral judgments and present the possibility that conservative judgments are related to socially adverse outcomes, particularly antisocial behavior. An alternative approach would be to conduct a mediation analysis of these variables, as discussed above, in which moral judgment is the independent variable, dark traits the mediating variable, and misconduct the dependent variable.

Social and cultural considerations

Arvan¹ correctly noted that social liberals and conservatives are diverse groups, and it would be difficult to pinpoint issues that definitively separate liberals from conservatives. However, Arvan¹ also overlooked that attitudes concerning specific social issues have shifted in American society. For example, between 1996 and 2012, Barack Obama’s expressed position on same-sex marriage shifted several times; in May 2012, he became the first president to endorse same-sex marriage.¹¹ In 1992, Bill Clinton originated the catchphrase that abortion should be “safe, legal, and rare.” In 2012, that phrase was extinguished from the plank of the Democratic Party and the official position of Planned Parenthood.¹² While President Obama might be described as relatively liberal and Bill Clinton might be described as politically moderate, it can be argued that the views of moderates have shifted more to the left over the last decade. Thus, advocacy that marriage should be between one man and one woman might have been considered moderate 20 years ago; most Americans did not support same-sex marriage then.¹³ As of one year ago, a Gallup poll¹⁴ found that 71% of Americans, including a majority of Republicans, support same-sex marriage. What will we make of those who opposed same-sex marriage 15 or 20 years ago but now support same-sex marriage? Did such people also shift their positions over time on a dimension of psychopathy traits? The demographic group for which most people still opposed same-sex marriage in the Gallup poll comprised people who attend church weekly. Would one say that approximately 29 % of the U.S. population who oppose same-sex marriage, primarily composed of people who attend church weekly, are at risk of having psychopathic traits? Item 10 on the MIS Survey states, “A woman ought to have a legal right to have an abortion upon request.” Disagreement with this statement was considered by Arvan¹ to be a conservative judgment. Although most Americans support abortion in some or all situations, people are less likely to support abortion the further along in pregnancy.¹⁵ At 24 weeks of pregnancy, 42 % of those surveyed indicated that abortion should generally be illegal. Would the people who support abortion at six weeks but oppose abortion at 24 weeks be considered as having liberal or conservative judgments, and is such a dichotomy applicable to making these judgments? In all fairness, Arvan^1,2 conducted the studies a decade ago and did not have the benefit of recent surveys of Americans on such hot-button issues. However, this underscores the point that American society has shifted positions on issues such as same-sex marriage and abortion. Transgender issues, especially medical interventions for transgender children, are controversial. The views of Americans at present concerning transgender issues may shift over the next decade.¹⁶

Notwithstanding the limitations identified in this article concerning Arvan’s^1,2 publications, Arvan’s findings and the findings of others that conservativism is associated with dogmatism and dark personality traits must be given weight. At the same time, it is hoped that the review and critique presented here concerning limitations of methodology, analyses, and social/cultural considerations will provide more context to Arvan’s^1,2 findings. The Moral Intuition Survey response scale appears problematic, even though most items ostensibly have face validity. The study designs were cross-sectional and correlational, not supporting causational conclusions. The positions held by Americans on some of the contentious issues tapped by the Moral Intuition Survey, such as abortion and same-sex marriage, have shifted over time, complicating judgments and conclusions about personality that can be drawn. Fortunately, science is an open field of study; it remains to be seen whether new research is consistent with previous findings and whether there is new research that supports dissenting findings.

None.

There is no conflict of interest.

1. Arvan M. A lot more bad news for conservatives, and a little bit of bad news for liberals? Moral judgments and the dark triad personality traits: A follow–up study. Neuroethics. 2012;6:51–64.
2. Arvan M. Bad news for conservatives? Moral judgments and the dark triad personality traits: A correlational study. Neuroethics. 2013;6:307–318.
3. Paulhus DL, Williams, KM. The dark triad of personality: Narcissism, Machiavellianism, and psychopathy. Journal of Research in Personality. 2002;36:556–563.
4. Jones DN, Paulhus DL. Introducing the Short Dark Triad (SD3): A brief measure of dark personality traits. Assessment. 2014;21:28–41.
5. Preston OC, Anestis JC. Psychopathic traits and politics: Examining affiliation, support of political issues, and the role of empathy. Personality and Individual Differences. 2018;131:142–148.  
6. Gay JG, Lishner DA, Vitacco MJ, et al. Deconstructing the association between psychopathy and political orientation: Is it attributable to moral intuitions or moral competency? Personality and Individual Differences. 2019;149:128–134.
7. Lilienfeld SO, Widows MR. Psychopathy Personality Inventory–Revised (PPI–R). Professional Manual. Lutz, FL: Psychological Assessment Resources. 2005.
8. Choma BL, Hafer CL, Dywan J, et al. Political liberalism and political conservatism: Functionally independent? Personality and Individual Differences. 2012;53:431–436.
9. Gerber A, Huber G, Doherty D, et al. Personality and political attitudes: Relationships across issue domains and political contexts. American Political Science Review. 2010;104:111–150.
10. De Neve J–E. Personality, childhood experience, and political ideology. Political Psychology. 2015;36:55–73.
11. Allen MS, Iliescu D, Greiff S. Single item measures in psychological science. European Journal of Psychological Assessment. 2022;38:1–72.
12. Steinmetz K. See Obama’s 20–year evolution on LGBT rights. Time Magazine. 2015.
13. Flanagan C. Losing the rare in ‘safe, legal, and rare.’ The Atlantic. 2019.
14. Borelli G. About six–in–ten Americans say legalization of same–sex marriage is good for society. Pew Research Center. 2022.
15. McCarthy J. Same–sex marriage support inches up to new high of 71%. Gallup. 2022.
16. Molla R. What Americans think about abortion, in 3 charts. Vox. 2022.