Correlates of discriminatory behavior

In this post, I am going to go over a number of different predictors and correlates of discriminatory behavior and prejudiced behavior, such as racism, sexism, homophobia, etc. The main topics discussed will be personality correlates of discrimination, cognitive functioning of people who are more discriminatory towards others, and the heritability of discrimination.

Personality Features and Political Orientation

People who discriminate against others are more likely to score highly on social dominance orientation and are generally more authoritarian. Pratto et al. (1994) is a particularly well-known paper on this topic (cited by 4760 as of right now). Social dominance orientation is a person’s level of preference for inequality among some sort of social groups. Basically it is Jimmy’s preference for Group A to have power and control over Group B, whether groups A and B are racial, sexual, or class-based groups. So, it would make sense this has a sizeable correlation with prejudice, discrimination, racism, etc. Pratto et al. posit it doesn’t just predict discriminatory behavior, but of course, political views which would instate those prejudices. It could be argued not all aspects of this are bad things. For example they hypothesized, and bore out in the results, that supporting a meritocracy, a system which believes in hierarchy based on what you bring to the world, your merit that is, is correlated with social dominance orientation. I think most rational people support some level of meritocracy, but perhaps I am biased myself.

The authors used a total sample of 1,952 college students, which was lumped together of multiple samples. The total sample was somewhat diverse. Most of the students were European-American and Asian American, but the Hispanic and Black samples were still usable. The Arab American sample was very small. In terms of family income, the sample was mostly average, with some slight overrepresentation in people living below $20K per year, but that is fine. Unfortunately the samples were not all given the same questionnaires and items through each wave, so the sample for different variables is considerably lower than 1,952, but overall, the sample seems pretty decent and representative.

There were moderate, usually statistically significant correlations between sex and social dominance orientation, showing men were about 0.2-0.3 standard deviations higher in social dominance orientation than women. This result would be interesting for looking at sex differences in outcomes. Social dominance could be very helpful for moving up in modern, economic hierarchies, though I am unaware of any literature on this. I’m not sure the legitimacy of their myths scale, especially when most people think in generalizations and averages, and considering the actual accuracy of stereotypes (hence not myths), but I will report the results on it for now. The correlation between social dominance and anti-black racism was r=0.55, for nationalism was 0.54, for sexism was 0.47, and for cultural elitism was 0.40. They found a correlation of r=0.38 for social dominance orientation and political-economic conservatism, and correlations of -0.44 and -0.40 for racial policy and women’s rights respectively. These political orientation correlations are heavily mediated by a genetic correlation between social dominance orientation and political views (Haarklau et al., 2019).

Case et al. (2008) used a sample of 383 college students to measure the effect of social dominance and right wing authoritarianism on one’s views and discrimination against women and homosexuals. The sample was 80 percent white and 20 percent non-white, 53 percent female, and 91 percent heterosexual. They found correlations of r=0.203 and 0.319 between prejudice against women and homosexuals respectively and right-wing authoritarianism, surprisingly no significant correlation between discrimination against women and homosexuals and right-wing authoritarianism, and correlations of 0.253 and 0.291 between social dominance orientation and prejudice against women and prejudice against homosexuals, respectively. There was a significant correlation of 0.279 between social-dominance orientation and prejudice against women, but the same was not true for homosexuals. The association between discrimination against women was much larger for men than for women. The same was much less true for homosexuality, to no surprise.

Parkins et al. (2006) used a sample of 331 undergraduate students at a midwestern university to measure personality correlates of workplace bullying and discrimination. They did not collect very detailed demographic variables, but the information they provided shows that slightly over half of the sample was male, many of the students recieved Pell grants, and many worked more than 20 hours per week, though the average family income of the sample was fairly representative. Most of the sample (83%) was white, only 8 percent were black, 2 percent were Hispanic, 4 percent were Asian, and 3 percent were “Other”. 95 percent were heterosexual. All were employed. The authors gave the participants a comprehensive 197-item questionnaire in which they self reported their personality traits, attitudes, and behaviors for course credit and their responses were scaled from 1 (strongly disagree) to 4 (strongly agree). Higher scores on discriminatory measures indicated higher frequencies of this behavior. They created a higher order discrimination factor and higher order prejudice factor through factor analyses of the related items. Workplace bullying and discrimination were statistically significantly correlated at r=0.46.

Higher order discrimination significantly correlated with social dominance orientation at 0.459, perspective taking at -0.284, right wing authoritarianism at 0.230, anger at 0.179, and verbal aggression at 0.222. Whites were more likely to be prejudiced and discriminatory than non-whites, but this result should of course be interpreted with caution due to the very small sample of non-whites. Males were more likely to be prejudiced and discriminatory than females. There was no significant correlation between sexual orientation and discrimination or prejudice, but again, the sample was incredibly small. Surprisingly, self esteem did not significantly correlate with discrimination or prejudice. The same was true for anxiety. Prejudice correlated at 0.623 with social dominance orientation, -0.311 with perspective taking, 0.292 with RWA, 0.248 with anger, 0.178 with verbal aggression.

x used data from 156 non-psychology university students who received movie theater gift cards for their participation in the study (cool reward). The study was pretty evenly split, male and female. They created a generalized prejudice factor from a test which measured homophobia, modern racism, classical racism, modern sexism, classical sexism, modern ableism, and classical ableism. They ultimately found people who just seem to dislike everyone. The two statistically significant Big Five personality variables correlated to prejudice were openness to experience and agreeableness. The former was correlated at -0.44 to -0.45 and the latter was correlated at -0.45. So these results find prejudiced people are less open to experience and less agreeable on average than less prejudiced people.

Ekehammar et al. (2004) used a sample of n=183 to test relationships between Big Five Personality Traits, social dominance orientation, right wing authoritarianism, and generalized prejudice. They developed the following causal model: Big Five personality traits had no effect on its own on generalized prejudice but did impact it directly through right wing authoritarianism and social dominance orientation. Basically, this means, yes there is a relationship between your general personality and level of prejudice, but the cause of generalized prejudice is not your personality, rather other factors which are intercorrelated with the two variables.

Many of these studies use right wing authoritarianism as a correlate, rarely using left-wing authoritarianism. I think testing the relationships between left wing authoritarianism and bias would be a worthy experiment. A new, comprehensive pre-print by Costello et al. (2020) discusses this topic at length. In constructing a factor for left wing authoritarianism, they found particularly strong items on their questionnaires were ones which implied prejudice against conservatism, moral absolutism, and desire for ideological uniformity. The authors remark prejudice, social uniformity etc. are binding factors between left wing and right wing authoritarianism. Conway et al. (2018) constructed a measure of left wing authoritarianism by basically rewriting the right wing authoritarianism scale with left wing oriented questions. They found this measure showed a positive manifold of correlations for liberal-originated prejudice, dogmatism, and general strength of political attitudes.

Left-wing prejudice, however, does not manifest itself in discrimination against minorities. Rather it tends to be against white people. Additionally, conservatism appears to be becoming less of a predictor of anti-black bias of some sort, whereas the left wing version of prejudice may be becoming stronger. A somewhat small study of 88 students in a Californian university tested the effects of ideology on consequentialism, particularly when said consequentialism is racially based (Uhlmann et al., 2009). The researchers had the students report their political orientation on a scale basis. They gave the students two of the same moral dilemma, the Trolley problem, but with the race of the people in it being flipped. Some participants had to make the following decision: kill one white person or 100 black people; other students had to make this decision: kill one black person or 100 white people. Liberals were more likely to endorse consequentialism, meaning kill one person for the sake of 100 people, when the person being killed was white. Conservatism had no effect on this distinction.

Some other studies show similar results. Tetlock et al. (2000) showed liberals felt non-whites shouldn’t pay more for home insurance for living in a high-risk area but they were neutral when asked if whites should. And Goldberg (2019) shows white liberals are the only group which has an in-group distaste.

Summary of the literature I’ve cited above: Some personality traits are correlated with prejudice and discrimination against non-self groups, but these correlations are mediated through prejudice’s relationship to social dominance orientation and, primarily, right wing authoritarianism. The literature consistently finds an expected correlation between social dominance orientation and prejudice against other groups. Finally, left wing beliefs and left wing authoritarianism are overlooked predictors of bias and prejudice. Left wingers tend to dislike white people and political conservatives and would act in very harsh ways to them such as silencing their speech (eg. Winegard et al., 2019), changing their moral views when whites are killed, etc. Overall, there are some negative attributes which are common among people who are discriminatory against others.

Cognitive Ability

An old study by Kutner and Gordon (1964) followed a relatively small group of children for nine years in order to longitudinally test the association between prejudice and cognitive functioning. The authors hypothesized that prejudice against ethnic groups was associated with poor cognitive functioning in areas such as abstract reasoning and concept formation, and that the relationship was strong enough that changes in one variable would be associated with changes in the other. The mean IQ difference between the groups was significant and substantial: less prejudiced people scored an average of 119.3 whereas more prejudiced people scored an average of 110.9 with similar standard deviations (this was a population of upper class children). In the longitudinal findings, there hypotheses were mostly confirmed: people who changed in their amount of ethnic prejudice over time also changed in their cognitive functioning. Davidson (1976) also found a moderate negative relationship between ethnic prejudice and IQ in a longitudinal study.

My assumption is these associations are mediated by the correlation of IQ with libertarianism (Carl, 2014), political non-radicality (Rindermann et al., 2011), and probably even the relationship between IQ and socioeconomic status, considering people in more wealthy environments tend to be raised with more egalitarian views (in addition, some of this effect may be due to going to higher education and being exposed to the views there).


Some heritability estimates I’ve collected over time:

Heritability of Racism
White superiority0.40Martin et al. (1986)
Apartheid0.43Martin et al. (1986)
Generalized Prejudice0.38Lewis et al. (2013)
Racial Prejudice0.78-0.79Loehlin (1993)
Racial Prejudice0.11-0.49Truett et al. (1992)
Racial Preference0.28Weber et al. (2011)
Racial Marriage0.52Weber et al. (2011)
Mixed Marriages0.33Martin et al. (1986)
Ethnic Preference0.46Weber et al. (2011)
Ethnic Marriage0.08Weber et al. (2011)
Innate “inferiority” of blacks to whites0.50Eaves et al. (1989)
Prejudice towards non-EU nationalities0.32Kandler et al. (2015)
Discriminatory intent toward foreign nationals0.31Kandler et al. (2015)
Segregation0.27Alford et al. (2005)
Segregation0.37-0.50Hatemi et al. (2014)
Busing0.26Alford et al. (2005)
Busing0.17-0.34Hatemi et al. (2014)
Ethnocentrism0.18Orey and Park (2012)
Ethnic/Racial Minorities0.52Bell et al. (2009)

So, no this mantra that “racism is just learned!” isn’t really true. It seems to be moderately heritable.

Heritability of Homophobia

Homophobia0.36Verweij et al. (2008)
Gay Rights0.28Alford et al. (2005)
Same-Sex Marriage0.57Eaves et al. (1989)

Heritability of Sexism

Working Mothers0.36Martin et al. (1986)
Women’s Liberation0.33Alford et al. (2005)
Separate roles for men and women0.00Olson et al. (2001)

Zero Discrimination

So, the conclusion of this article seems to be that there are some socially-recognized negative things associated with discriminatory behavior and prejudice. These are mainly authoritarianism, conformity, unwillingness to attempt new things, lower cognitive ability and lower reasoning ability. Additionally, I have collected a number of studies showing moderate heritability of prejudice. While we may agree prejudice and discrimination are generally negative things, we also need to be cautious in a) coming to the conclusion that we can come to a non-discriminatory society in some way, and b) thinking that discrimination in all forms is an evil thing. Both of these things are clearly wrong.

The first (and the second) is wrong because a main function of the human brain is to discriminate between things. Hence why color discrimination and pitch discrimination, for example, are g-loaded tasks. Humans need to be able to assess risks and make probability-based judgements based on recognition of differences between objects, humans, and groups. This leads us into the second issue. Discrimination is not necessarily an evil. If you are walking down the street at night and you see a large, bulky man, hands in his pockets, hoodie on, and maybe approaching you, it is simply obvious that you have a greater probability of living tomorrow if you avoid this man. Humans developed the process of rationality and discrimination in order to increase their lifespans and then give birth to more children.

What happens when you pretend there is no statistical association between certain things and certain outcomes eg. tall, mysterious men at night and felony against you? Well, then you will probably be diagnosed with a disorder called Williams Syndrome. This is a “rare genetic disorder characterized by growth delays before and after birth (prenatal and postnatal growth retardation), short stature, a varying degree of mental deficiency, and distinctive facial features that typically become more pronounced with age.” The main mental feature of this disorder is trusting everyone. You do not realize some people are more dangerous, rude, etc. than others and so you go to everyone wanting to engage in conversation with them. There are no boundaries and this is obviously a dangerous thing to do. So, the overall conclusion: bigotry is predicted by negative things ergo bigots are typically not great people, but we should not draw radical conclusions from this about bigots or the relevance of discrimination.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s