This is an open access article distributed under the terms of the Creative Commons
Attribution License (Racial prejudice can affect individuals and groups in a variety of negative ways including the promotion of intergroup conflict, restrictions of human rights, and psychological hardships (Fekete, 2004; Sanchez-Hucles, 1999). Two individual difference factors often used to study prejudice are right-wing authoritarianism (RWA) and social dominance orientation (SDO). While these two factors often predict prejudice in similar ways, they have also been found to diverge substantially and to even operate in opposing directions (e.g., Bilewicz, Soral, Marchlewska, & Winiewski, 2017; Thomsen, Green, & Sidanius, 2008). One reason why dissimilarities occur may be the context in which they are studied. Authoritarianism, with its adherence to conventional norms and authority, may be particularly variable in how it predicts prejudice because of its dependence on the surrounding social and political context (Duckitt & Sibley, 2010). For example, should the surrounding context include authoritative social norms that are opposed to prejudice, then authoritarianism may be obediently against those prejudices. Social dominance orientation, on the other hand, has a competitive-jungle worldview that leads to support for group hierarchies (Duckitt & Sibley, 2010). As such, SDO might be a more consistent predictor of prejudiced attitudes. Continued research on how these two factors predict prejudice within different international settings is therefore warranted. In line with this reasoning, the present study took place in the interracial context of post-apartheid South Africa.
Right-wing authoritarianism is described as a set of ideological attitudes involving a proneness to obey and respect authorities and a tendency to conform to conventional norms and values (Altemeyer, 1981). It has been argued that RWA originates from concerns with maintaining collective stability, order, and security, and that these concerns feel particularly important because the social world can be unpredictable, threatening, and even dangerous (Duckitt & Sibley, 2010). It has also been proposed that the conformity and obedience tendencies of RWA may have been adaptive from an evolutionary perspective. The argument is that RWA may have enabled coordination and cooperation in large-scale groups, which became too large to be managed by kinship and reciprocity dynamics alone (Kessler & Cohrs, 2008).
Social dominance orientation encompasses a set of beliefs that groups should be organized into social hierarchies, and that such hierarchies are both natural and preferable (Pratto, Sidanius, Stallworth, & Malle, 1994). This orientation can include desires for intergroup inequalities to be maintained through subtle social processes as well as desires for the overt and aggressive subordination of some groups by other groups (Ho et al., 2012). Moreover, although SDO has sometimes been understood as depicting a preference for dominance by one’s ingroup, some of the literature has clarified, both theoretically and empirically, that it depicts a general preference for group-based hierarchies irrespective of the position of one’s ingroup (Kteily, Ho, & Sidanius, 2012; Pratto, Sidanius, & Levin, 2006).
Both RWA and SDO have been found to predict racial prejudice in western cultures (e.g., Duriez & Soenens, 2009; Hiel & Mervielde, 2005; Poteat & Spanierman, 2012). However, considering the normative threat motives of RWA, and the hierarchy motives of SDO, it is understandable that they would predict prejudice for different reasons. In addition, it has even been found that their relationships with prejudice have sometimes been in opposing directions, in certain circumstances.
Considering the different motives for prejudice that RWA and SDO provide, Duckitt and Sibley (2007) found that RWA predicted prejudice towards groups that were perceived to be dangerous; such as violent criminals, terrorists, and Satanists. On the other hand, SDO predicted prejudice towards groups that were viewed in a derogatory manner such as the unemployed, mentally handicapped, and the obese. Similarly, in a study of anti-immigrant attitudes in 17 countries, RWA predicted prejudice when immigrants were seen as increasing the crime rate, while SDO predicted prejudice when immigrants were seen as having a higher unemployment rate (Cohrs & Stelzl, 2010). In addition, Crawford and Pilanski (2014) found that RWA predicted political intolerance of groups that had cohesion-attenuating as well as hierarchy-attenuating political objectives, while SDO predicted political intolerance of groups that only had hierarchy-attenuating political objectives.
The different motives for prejudice that stem from RWA and SDO have at times led them to be at odds with each other when predicting prejudice. For example, Thomsen and colleagues (2008) conducted a study in which immigrants who had different values and traditions were a target for persecution. It was found that RWA predicted support for persecution when the immigrants were described as not assimilating into the dominant culture, thereby threatening conformity concerns. In contrast, SDO predicted support for persecution when immigrants were described as successfully assimilating into the dominant culture, thereby blurring status boundaries between ethnic groups. In addition, Bilewicz and colleagues (2017) investigated support for the prohibition of hate speech against stigmatized minorities. The researchers argued that acts of hate speech were a violation of accepted societal norms, and would therefore be opposed by the motives of RWA. Consistent with this argument, they found that RWA was a positive predictor of hate speech prohibition while SDO was a negative predictor. The present study extends this line of research by investigating how RWA and SDO predicted two types of racial prejudice within the context of South Africa.
South Africa is a racially diverse society with a legacy of legislated racial segregation during apartheid. However, in the early 1990s significant political and social change occurred, which led to the first fully democratic election in 1994. During this period of increased democracy, an effort was made by various authorities to create new societal norms and values that embraced racial equality and harmony. For example, the term “rainbow nation” was promoted by the Archbishop of South Africa, Desmond Tutu, as well as newly elected President, Nelson Mandela, and has since become a symbol of the country and a part of the nation’s parlance. Moreover, a new constitution was adopted that embodied new norms regarding racial equality, and a new national anthem was composed using five of South Africa’s 11 official languages (Allen, 2013; Baines, 1998; “South Africa’s new constitution,” 2011).
Authoritarians have ideological attitudes about submitting to authorities and conforming to conventional norms and values. Considering the ideological beliefs of authoritarians, therefore, it was possible that RWA would be negatively related to certain types of racial prejudice. At the same time, however, SDO, and its racial-hierarchy beliefs, might remain positively related to racial prejudice.
Two types of racial prejudice were included in the study. The first was beliefs in racial inequality regarding civil rights. This form of prejudice included views that race groups should have different rights, freedoms and opportunities. The second type of prejudice looked at desires for racial separation. This category consisted of views that multi-racial, or “melting pot”, societies are less desirable and that South Africa should be broken up into smaller nations based on race and ethnicity.
It was anticipated that RWA would have different relationships with the two types of prejudice. For beliefs about inequality regarding civil rights, it was expected that a negative relationship would be found; RWA would be opposed to this type of inequality. This prediction was based on the rainbow nation narrative coupled with the conventionalism concerns of RWA.
Although the rainbow nation narrative emphasized equality regarding civil rights, it did not remove or negate the cultural differences that exist between South Africa’s ethnic and race groups, which are extensive. The various ethnic and race groups produce a cultural diversity that includes 11 official languages, variations in food, music and clothing, and differences in religion, ancestry, and history (“Race and Ethnicity in South Africa,” 2015; “South Africa’s Diverse Culture,” 2014). Considering RWA’s concerns with protecting social norms, customs and traditions, it was expected that a positive relationship would occur between RWA and desires for intergroup separation.
The study also included SDO to serve as a contrast to RWA. The ideological beliefs of SDO include views that group hierarchies are both natural and preferable (Pratto et al., 1994). As such, it was expected that SDO would have positive relationships with both types of prejudice: inequality regarding civil rights as well as separation.
Finally, it was expected that race would moderate one of the proposed relationships. It was anticipated that different race groups might view South Africa’s rainbow nation identity differently regarding its authoritative status. Racial differences might exist because the rainbow nation narrative is typically associated with notable Black leaders such as Nelson Mandela and Desmond Tutu. As such, the relationships between RWA and prejudice might vary across race groups. Although a variety of races exist in South Africa, the study focused on Black participants and White participants. With the rainbow nation message being associated with Black authorities more so than White authorities, it was expected that the negative relationship between RWA and inequality regarding civil rights would be stronger for Black participants compared to White participants. On the other hand, the rainbow nation message did not attempt to reduce the importance of group-based traditions, cultures, and ancestry. Consequently, it was not expected that race would moderate the relationship between RWA and intergroup separation. In addition, the relationships between SDO and the two types of prejudice were not expected to vary across race.
It should be noted that the rainbow nation ideal has been criticized, mainly by Black leaders, as not depicting the true situation in South Africa (Anwar, 2017). This criticism is particularly directed at the economic inequalities that continue to exist along racial lines, in areas such as employment, household wealth, and land and home ownership (Anwar, 2017; Orthofer, 2016). However, the term rainbow nation can refer to different dimensions of racial equality. It can refer to the idea of racial equality regarding civil rights, and it can refer to a reality on the ground regarding the distribution of economic resources and opportunities. The current study focused on the idea of equality regarding civil rights; i.e., the prejudice of inequality beliefs regarding civil rights. It was anticipated that RWA would be a negative predictor of inequality beliefs regarding civil rights.
In addition, although Black figures in particular have criticized the rainbow nation ideal, due to ongoing economic inequalities across race groups (Anwar, 2017), the notion of the rainbow nation would still be more strongly associated with Black leaders than with White leaders. The reason being is that past White leaders, particularly those during the transformative 1990s, were often active participants in the previous apartheid government. Therefore, it was expected that the relationship between RWA and inequality beliefs regarding civil rights would be stronger for Black participants than for White participants.
One final issue concerning the moderated relationship involving RWA and civil rights is that equal civil rights might be viewed as leading to an increase in economic benefits for Blacks and a decrease in economic benefits for Whites. This change in economic norms might be particularly threatening to White South Africans with high RWA, which could lead White participants with high RWA to support inequality regarding civil rights. However, we argue that equality regarding civil rights is a substantive component of South Africa’s governing bodies and national symbols, which would lead White authoritarians to oppose inequality regarding civil rights. Of note, a study with White South Africans found RWA to be a positive predictor of support for acts of political intolerance by a majority rule, Black South African government (e.g., the government should detain enemies of the state; the government should have the right to silence those who oppose the people; Duckitt & Farre, 1994). Further, it is not an absolute that holding beliefs in equal civil rights is necessarily connected to particular economic views: authoritarians could support equal civil rights as well as existing economic inequities. The present study focused on beliefs regarding civil rights. It was expected that RWA among White participants, as an individual difference variable, would negatively predict inequality beliefs regarding civil rights.
Method
Participants and Procedure
Undergraduate students at the University of South Africa participated in the study during the second semester of the 2017 school year. The study used an online survey and 2,390 students read an online description of the study. The study was part of a larger research project on prejudice, which formed the basis for the study’s description to participants. From that group, 1,449 students completed the survey. Students who were not born in South Africa were removed, which numbered 111. Students who did not identify as Black or White were also removed. These included 112 Coloured and 62 Indian/Asian students (using South African racial classifications). The final sample was 1,164 and consisted of 898 Black participants and 266 White participants. The Black participants were 60% female and ranged in age from 18-years-old to 60-years-old (M = 32.00, SD = 8.36). The White participants were 72% female and ranged in age from 18-years-old to 73-years-old (M = 34.29, SD = 11.09). Ethics approval for the study was granted by a university institutional review board, and informed consent was obtained from participants. No incentives for participation were provided.
Materials
Right-Wing Authoritarianism
Authoritarianism was measured with the six positively worded items from Funke’s (2005) scale. Example items are “The withdrawal from tradition will turn out to be a fatal fault one day” and “Obedience and respect for authority are the most important values children should learn.” Responses were provided on a 7-point scale that ranged from “strongly disagree” to “strongly agree.” A factor analysis (principal axis factoring) extracted one factor with an eigenvalue greater than one. The items were therefore used as a single measure of authoritarianism and were not divided into the three theoretical sub-factors (submission, conventionalism and aggression). Regarding the usefulness of using authoritarianism in studies of social attitudes within the South African context, prior research with White South Africans found RWA correlated with the acceptance of parental religious beliefs and with support for restricting civil liberties (e.g., government censorship of political material; Duckitt, 1993). (overall α = .68; for Black participants, α = .65; for White participants, α = .78).
Social Dominance Orientation
The SDO7 version of the social dominance orientation scale was used (Ho et al., 2015). The scale consists of 16 items asking about views on group inequality and dominance. Example items are “Some groups of people are simply inferior to other groups” and “Some groups of people must be kept in their place.” Responses were given on a 7-point scale from “strongly oppose” to “strongly favour.” (overall α = .81; for Black participants, α = .78; for White participants, α = .88).
Inequality Regarding Civil Rights
Views about inequality regarding civil rights were measured using the following five items: “Everyone in South Africa should have the same rights and freedoms regardless of their race,” “All races in South Africa should have the same opportunities in life,” “All racial groups have an equal right to live in South Africa,” “All races in South Africa are equal in terms of their natural qualities and capabilities,” and “Each racial group in South Africa (Black, White, Coloured, or Indian) is as South African as the next.” Responses were measured with a 7-point scale ranging from “strongly disagree” to “strongly agree.” Responses were reversed coded so that higher scores reflected stronger views of inequality regarding civil rights. (overall α = .76; for Black participants, α = .75; for White participants, α = .70).
Group Separation
Desires for the race groups to be separated were measured using the following four items: “The presence of so many races and ethnicities threatens the unique traditions and customs of the different groups,” “Being a racially and ethnically diverse society makes it harder to maintain law and order,” “South Africa should be broken up into distinctly different nations based on race,” and “South Africa should be broken up into distinctly different nations based on ethnicity.” Responses were provided on a 7-point scale from “strongly disagree” to “strongly agree.” (overall α = .75; for Black participants, α = .74; for White participants, α = .71).
Results
To test the relationships between RWA, SDO and the two types of prejudice, and whether these relationships were moderated by race, a series of SEM analyses were conducted. First, an unconstrained path model, in which path coefficients were free to vary between Black and White participants, was conducted. This model was then compared against a nested model in which the path coefficients of the four relationships were constrained to be equal across race groups. Finally, tests of the individual paths were conducted to see which of the four relationships were moderated by race.
The SEM analyses included a measurement model and a path model. The measurement model included six items for RWA, five items for inequality regarding civil rights, and four items for group separation. To represent the SDO construct, 16 items were used but were grouped into four different SDO factors. According to Ho and colleagues (2015), the SDO7 measures two sub-dimensions (inequality and dominance) each of which has eight items. In addition, the eight items of each sub-dimension consist of four pro-trait items and four con-trait items. The measurement model for SDO was therefore a two-level structure, with four factors each having four items.
The unconstrained model, in which path coefficients were free to vary between Black and White participants, was conducted first. The model had adequate fit, χ2(848) = 2019.23, CFI = .90, RMSEA = .03, SRMR = .06 (see Figure 1; see Table 1 for descriptive statistics).
Figure 1
Structural Equation Model
Note. Path coefficients are standardized regression weights. Coefficients for Black participants are presented first, followed by coefficients for White participants. For clarity, the measurement part of the model is only partly presented.
*p < .05.
Table 1
Means, Standard Deviations, and Correlations
| Variable | M | SD | 1 | 2 | 3 | 4 | M | SD |
|---|---|---|---|---|---|---|---|---|
| 1. RWA | 5.58 | .99 | – | .24* | .06 | .27* | 5.07 | 1.24 |
| 2. SDO | 2.45 | .97 | .04 | – | .33* | .47* | 2.20 | 1.00 |
| 3. Inequality | 1.85 | 1.18 | -.24* | .18* | – | .39* | 1.40 | .76 |
| 4. Group Separation | 2.79 | 1.57 | .19* | .42* | .15* | – | 2.47 | 1.32 |
Note. Values for Black participants are on the left of the diagonal. Values for White participants are on the right of the diagonal.
*p < .05.
Next, the paths between RWA, SDO and the two types of prejudice were constrained across race groups. A chi-square difference test showed that the unconstrained and constrained models had significantly different fit, χ2diff(4) = 50.37, p < .001. This outcome indicates that the unconstrained model had better fit and that race acted as a moderator of the path model.
In the final step, each of the four paths were examined individually to see which specific coefficients varied across race. Four chi-square difference tests were conducted. In each test, the unconstrained model was tested against a constrained model in which one path coefficient was set as equal between race groups. Bonferroni correction was used for the four tests, which meant p values below .0125 were needed for significance. Only one test returned a significant p value: the path between RWA and inequality regarding civil rights was significantly moderated by race, χ2diff(1) = 34.79, p < .001. The other three paths were not moderated by race: for RWA and group separation, χ2diff(1) = 2.95, p = .086; for SDO and racial inequality, χ2diff(1) = 3.31, p = .069; and for SDO and group separation, χ2diff(1) = 1.34, p = .246.
As expected, SDO was a significant and positive predictor of both types of prejudice, and this was the case for both Black and White participants. The hypotheses regarding RWA were only partly confirmed. For Black participants, RWA was negatively related to inequality regarding civil rights and positively related to group separation, which was expected. However, for White participants, neither relationship was significant; although RWA and group separation were positively related in a bivariate correlation analysis. Finally, race was a significant moderator of the relationship between RWA and inequality regarding civil rights. The negative relationship between RWA and inequality beliefs was significantly stronger for Black participants compared to White participants. None of the other relationships were significantly moderated by race.
Discussion
Prejudice and racism can be a critical problem in many societies including South Africa. However, within the South African context, following the fall of the apartheid regime, political, social, and religious authorities promoted a narrative espousing racial equality (Baines, 1998). Therefore, considering the ideological attitudes of authoritarianism, it was speculated, and confirmed for Black participants, that RWA would be a negative predictor of beliefs in inequality regarding civil rights.
The ideological attitudes of authoritarianism are not only concerned with following authoritative doctrines and conventional values but also with protecting the social norms, customs, and traditions that define different groups. In the South African context, the various race and ethnic groups provide an array of cultural diversity. It was therefore possible that RWA would be related to desires for wanting to have separation between race groups in order to safeguard social conventions and group histories. For this reason, it was expected, and confirmed for Black participants, that RWA would be a positive predictor of desires for group separation. The same relationship was found for White participants when a bivariate correlation test was conducted.
Therefore, the ideological and political motivations of RWA led to two different views about race (for Black participants): the races are equal in terms of rights, freedoms, and opportunities, but race groups should maintain their own physical spaces.
Social dominance orientation behaved in a different manner and was a positive predictor of both types of prejudice. These relationships are consistent with ideological beliefs about group hierarchies. Believing that race groups should exist in hierarchies, and that such hierarchies are both natural and preferable, was related to stronger beliefs in inequality regarding civil rights and stronger beliefs that race groups should not integrate. These results show that prejudicial views from SDO are sometimes in line with those from RWA, but sometimes they are not. These results are consistent with prior research, which found that RWA and SDO sometimes operate in opposing directions when it comes to prejudicial views, with SDO being a more consistent predictor of prejudice than RWA (Bilewicz et al., 2017; Thomsen et al., 2008). The present study extends this body of research by analyzing views on race in post-apartheid South Africa.
One factor that can influence the views of authoritarians are the messages that come from relevant authorities. In the present study, the negative relationship between RWA and inequality beliefs regarding civil rights was stronger for Black participants than for White participants. One explanation for this moderation effect is that messages about racial equality in South Africa were more strongly associated with Black leaders than with White leaders. Therefore, the relationship between authoritarian tendencies and inequality beliefs regarding civil rights may have been weaker among White participants. It should be noted that racial equality is an ongoing debate in South Africa (Anwar, 2017). However, the debate is primarily directed at issues regarding economic equality and not whether equal civil rights should exist.
In addition, regarding the nonsignificant relationship between RWA and inequality beliefs regarding civil rights for White South Africans, it could be argued that this outcome was affected by White concerns with losing economic advantages, a change in economic norms that might be particularly threatening to White authoritarians. This possibility would have led to a positive relationship between RWA and inequality beliefs for White South Africans, which did not occur. It is possible, however, that the effect of racial equality norms was balanced by an effect of preserving norms regarding economic stratification. Future research could explore the connection between beliefs regarding civil rights and concerns with economic changes, particularly for authoritarians in a civilly liberal society with a history of racial inequality.
Considering that different types of prejudice exist, the effects of RWA and SDO could lead to different types of individual profiles regarding prejudiced attitudes. For example, while many racial segregationists believe in both racial separation and civil inequality, some racial segregationists may actually believe in racial equality regarding civil rights. As illustrated by the current study, this latter view could occur for people with high RWA and low SDO who live in a multi-cultural society that has normative views supporting racial equality regarding civil rights. Consequently, efforts at attenuating prejudicial views would be better served by considering the different profiles that exist and the different ideologies that drive them.
Regarding sources of authority, the characteristics of a leader’s message could mitigate its effect on the beliefs of authoritarians. For example, it has been found that contact between race groups can reduce prejudice, even for people with high RWA and high SDO (Dhont & Van Hiel, 2009). However, characteristics of the contact makes a difference, with positive contact reducing prejudice while negative contact increases prejudice (Barlow et al., 2012). Likewise, characteristics of a leader’s message, besides ingroup status, could affect its impact on authoritarians. For example, the consistency or genuineness of the message and the fidelity of its application could affect the perceived authority or legitimacy of the message. Further research could explore the different factors that moderate the relationships between authoritarianism and adherence to authoritative messages about racial equality.
The present study provides evidence that an influx of authoritative messages about racial equality, which becomes part of the normative social ethos, can lead to changes in the views of people with higher authoritarian tendencies. In other words, the views of authoritarians may be somewhat malleable and dependent on the authoritative institutions of the social context. However, it should be noted that this malleability could lead to different outcomes: Authoritative messages that promote beliefs in inequality regarding civil rights could lead to increases in prejudiced attitudes (Goehrung, 2017). In addition, beliefs in racial equality do not necessarily preclude desires for culturally and ethnically homogenous geographical areas. Moreover, a tendency toward racial equality beliefs, through RWA ideologies, could still be diminished if paired with high SDO.
Another aspect of authoritarianism, besides adherence to authority and social conventions, is a willingness to use aggression against those who do not conform (Altemeyer, 1981). Therefore, when authorities and social norms advocate for nonprejudicial attitudes, high authoritarians may have less prejudicial views but may also be more willing to act violently against those who maintain prejudiced views or even those who critique aspects of their beliefs. Future research could explore how authoritarian aggression might accompany humanitarian beliefs.
One limitation of the study is that the premises for the predicted relationships between RWA and the two types of prejudice were not tested directly. Although the rainbow nation narrative is often discussed within the context of South Africa, a measure of its normative strength was not included in the study. Similarly, although different ethnic groups exist in South Africa, with different historical migration patterns, different languages, and different traditions regarding local governance, a measure of concerns regarding the preservation of ethnic norms was not included in the study. At the same time, there was a negative relationship between RWA and inequality beliefs regarding civil rights, for Black participants, but a positive relationship between RWA and racial separation. Future research could specify what aspects of authoritarianism and the South African context led to the two different relationships between RWA and prejudice.
Another area for future research would be to test the same relationships from the present study in other countries. While prior research suggests that the constructs of RWA and SDO are consistent across countries (e.g., Henry, Sidanius, Levin, & Pratto, 2005; Thomsen et al., 2008), it has also been found that the relationship between RWA and SDO itself varies across countries (Duckitt & Sibley, 2010). Future research could explore the present relationships in other contexts, which might shed additional light on what aspects of the South African context contributed to the observed relationships between RWA, SDO and the different types of prejudice.
In conclusion, RWA and SDO both predicted support for ethnic and racial separation. However, for Black participants RWA predicted support for racial equality regarding civil rights. It was argued that messages regarding racial equality that come from relevant leaders can influence the beliefs of people who are prone to follow authorities, that is, people with high RWA. While RWA is often associated with prejudiced views, it might also depend on the social context, and the messages that relevant authorities provide.