Understanding semantic differential measures in modern South Africa: attitudes of Black Africans towards White South Africans

Abstract

The future success of South Africa’s unique democracy depends on the development of harmonious race relations. Understanding the factors underlying the country’s interracial attitudes is, consequently, important. Social identity theory suggests that Black African attitudes towards White people are connected to their evaluations of South Africa’s other racial minorities. This thesis seems counterintuitive given that White people are associated with a long history of political, economic, and social oppression in the collective memory of many Black African communities. Nationally representative data from the South African Social Attitudes Survey were used to validate the thesis that Black Africans’ evaluations of White people correlated with their assessments of other racial groups. Pairwise correlation analysis was employed to test the article’s hypothesis. The results presented in this article showed that Black Africans’ evaluations towards the White minority correlated with their evaluations of other racial minorities in South Africa. Multivariate analysis, specifically a standard (ordinary least squares) linear regression, was used to confirm the bivariate analysis. Black Africans’ attitudes towards White people were strongly correlated with attitudes towards the country’s two other major racial minorities. This finding held even controlling for contact with White people as well as a range of socio-economic characteristics. The outcomes of this article invite closer examination of the factors that underlie the generality of outgroup evaluations among South Africa’s Black African majority.

Keywords

Intergroup attitudes interracial contact semantic differential South Africa White people

South Africa represents a remarkable opportunity to test an important thesis within social identity theory. According to social identity theorists, people are less likely to favour or tolerate those who seem different from themselves (Brewer, 1999; Brown, 2011; Fishbein, 2002). In other words, feelings about one racial outgroup should be influenced by a general affection for (and tolerance of) other racial outgroups. In the case of South Africa, social identity theory would suggest that White people should be part of a ‘band of other race groups’ in the mind of modern Black Africans. However, to those familiar with South African history, the supposition made above seems counterintuitive and naive. The collective memory of colonialism and apartheid differentiates the country’s White minority from other racial groupings in the country. Moreover, substantial socio-economic power continues to be in the hands of the White minority, making them distinct from other racial minorities.

The study hypothesised that, in South Africa, Black Africans’ evaluations of White people were unconnected to their evaluations of other racial minorities. To justify the article’s central thesis, quantitative methods were used to discern whether Black Africans’ attitudes towards White people correlated with their attitudes towards other racial minority groups. Multivariate techniques were utilised, specifically a standard (ordinary least squares) linear regression analysis. Data from a nationally representative public opinion survey were employed for this purpose. The article considered attitudes towards four race groups: (1) White, (2) Black African, (3) Coloured, and (4) Indian as well as foreigners living in the country. These attitudes were measured using semantic differentials.

There is a substantial academic literature on intergroup relations that pertains to the South African case. It would be beyond the scope of this article to review this literature in its entirety, but the article will justify its thesis by reviewing the pertinent academic literature.

The ‘band of others’

Racial identities can have a powerful effect on how an individual evaluates the world and have been linked to theories of prejudice. In his classic study of prejudice, Allport (1954) argued that there seems to be an essential desire within all human beings to create and preserve systems of group-based social hierarchy. Social identity theory suggests that people have a tendency to maintain (or enhance) their own constructed racial identity by comparing other racial groups adversely to their own (Brown, 2011; Fishbein, 2002). Intergroup relations research, however, has found that ingroup pride and outgroup negativity are often quite distinct (Brewer, 1999; Gibson, 2006). Individuals can have a positive view of their own racial group, for instance, without adopting correspondingly negative views of other racial groups.

Persons who are unfavourably disposed to one particular outgroup tend to be less favourable to others. Existing research demonstrates this effect, as Duckitt, Wagner, du Plessis, and Birum (2002) remark, to be powerful and pervasive, holding across very different outgroups and even for totally fictitious outgroups. This research has been used to suggest, as Allport (1954) has done, that prejudice is essentially a trait of personality. Any group would be viewed negatively by a ‘prejudiced’ individual if that group fell outside of what that individual considered ‘mainstream’. The most important question then becomes how the individual classifies ‘mainstream’. Historical factors of racial identification play a role here. In his study of ‘whiteness’ in the United States, for example, Goldstein (2006) shows how, over the last 100 years, Jews ‘became white’ in the eyes of White Protestant America.

Race classifications in South Africa

It is important to state that ‘race’ in South Africa is a socially constructed categorisation that was imposed on the country by the dictums of the colonial (and later apartheid) state.¹ Apartheid legislation (principally the Population Registration Act of 1950) led to the classification of the South African population into four race categories: (1) Black African, (2) White, (3) Coloured, and (4) Indian. The categorisation of people into these four groups was often arbitrary and capricious as work by Christopher (2009) has attested. The current South African government considers it important to continue to collect data on racial classification in order to specify the effects of past racial discrimination as well as to monitor government policies to alleviate discrimination. But all legal bases for classifying the population by racial group were removed in 1991. Any classification of race in the modern period is now self-reported. The apartheid-era racial categorisations have become, as research by Bornman (2010) has shown, ingrained in the minds of many South Africans (also see Durrheim, Mtose, & Brown, 2011).

In the collective memory of many Black Africans, White people are associated with a long history of political, economic, and social oppression. The histories of the country’s two major non-White racial minorities (i.e., the Indian and Coloured racial minorities) are, by contrast, characterised by exploitation and dispossession. The nature of the colonial system often situated these two groups uneasily between the ruling White minority and the Black African majority. This is not a unique to South Africa – in the typical settler colonial state, a political status gap often emerges between colonising settlers (with their full legal status and rights) and the colonised. Classic middleman minority theory has been applied to the Indian minority in South Africa (see Ramsamy, 2007 for a discussion of the Indian minority). Johnson (2017) has identified South Africa’s Coloured minority as a middleman of another type (i.e., semiprivileged proletarians). These histories, no doubt, shaped how Black Africans viewed the country’s Indian and Coloured racial minorities.

Race relations in post-apartheid South Africa

South Africa has been the site of social psychological inquiries on intergroup attitudes at different times over the last 100 years. A number of scholars have used public opinion surveys to look, in particular, at how race relations in the country have changed over time. Durrheim, Tredoux, Foster, and Dixon (2011) provide a meta-analysis of existing survey research to understand changes in racial attitudes over the period 1934–2011. The evidence suggests that Black African attitudes towards the White minority have remained largely negative since the end of apartheid. Bornman (2011) and Mynhardt (2013) provide evidence of a similar trend in their own studies. Research by Talbot and Durrheim (2012) using data from 1960 and 2008 showed that the racial stereotypes about White people in the country have an extraordinary inertia, even in the face of extraordinary historical change. Increasingly, there has been a growing interest in what factors could explain the pattern of Black Africans’ attitudes towards White people.

In the post-apartheid period, the social psychological research on determinants of interracial animosity has focused on mechanisms by which interracial prejudice can be reduced. The question of whether intergroup contact can provide a remedy to the problem of interracial hostility runs through much of this work. According to Allport (1954), contact between groups generally nurtures favourable intergroup attitudes under the right circumstances. There is certainly evidence that interracial contact (particularly friendship contact) is correlated with lower levels of prejudice and a greater openness to racial reconciliation (see, for example, Dixon, Durrheim, & Tredoux, 2007; Gibson, 2004). Recent work, however, has been quite sceptical about the power of interracial contact to improve Black Africans’ pro-White sentiment (see, for instance, Gibson & Claassen, 2010; Swart, Hewstone, Christ, & Voci, 2010; Tredoux & Finchilescu, 2010). These findings are in line with existing work on interracial contact (see, for example, Shelton, 2003; Thomsen & Rafiqi, 2016) that suggests that social status of the groups involved in contact may affect the positive influence of intergroup contact.

The obvious explanation for Black Africans’ negativity towards White people in contemporary South African society centres on the historical legacy of colonialism and apartheid. This explanation would suggest that Black Africans’ attitudes towards White people should be distinct from their views of the country’s other major racial minorities (i.e., Indian and Coloured people). But few studies look at Black Africans’ opinions of the Indian and Coloured minorities and instead prefer to focus on Black–White attitudes. As far as the author is aware, social psychologists have not sought to investigate how strongly Black Africans’ evaluations of White people are correlated with their appraisals of other racial groups using nationally representative public opinion data. Many of the more in-depth studies base their findings on small student or telephone samples rather than nationally representative household samples. This limits how much we can generalise the results of these studies to the country’s adult population.

Based on the literature review, it was hypothesised that Black Africans’ assessments of White people would not correlate with how Black African individuals view other racial minorities in South Africa.

Method

Participants

This study draws on data from the South African Social Attitudes Survey (SASAS). Conducted annually by the Human Sciences Research Council (HSRC) since 2003, SASAS is a repeated cross-sectional survey and typically has a realised sample of about 3000 adults.² The sampling frame used for the survey was based on Statistics South Africa’s national census data which was used to construct a set of small area layers (SALs). In the first sampling stage, 500 SALs were drawn³ – these SALs were stratified by province, geographical subtype, and majority race group. Data were weighted to be nationally representative of the country’s adult population. This study uses data from the 2011–2015 SASAS rounds. A sample, therefore, covered a range of different socio-demographic status groups. Using the SASAS 2015 round as an example, Table 1 presents unweighted summary statistics for the Black African subset of the sample.

Table 1.

Summary statistics (unweighted) for the Black African sample in SASAS Round 2015.

	Obs.	Mean	Min	Max
Female (ref. male)	1788	0.62	0	1
Age	1787	39.88	16	93
Labour market status (ref. employed)
Part-time employed	1728	0.08	0	1
Unemployed	1728	0.44	0	1
Labour inactive	1728	0.30	0	1
Marital status (ref. married)
Married before	1764	0.15	0	1
Never married	1764	0.58	0	1
Years of education	1743	9.03	0	16
Geographic type (ref. urban formal)
Urban, informal	1788	0.05	0	1
Trad. Auth. areas	1788	0.33	0	1
Rural, formal	1788	0.05	0	1
Province (ref. Western Cape)
Eastern Cape	1788	0.15	0	1
Northern Cape	1788	0.05	0	1
Free State	1788	0.10	0	1
KwaZulu-Natal	1788	0.15	0	1
North West	1788	0.11	0	1
Gauteng	1788	0.12	0	1
Mpumalanga	1788	0.12	0	1
Limpopo	1788	0.15	0	1

SASAS: South African Social Attitudes Survey.

Instrument

Beginning in 2011, SASAS respondents were asked to evaluate five different groups of people living in the country on two 11-point (0–10) end-defined response scales, the first measuring negative/positive and the second hostile/friendly. The exact phrase for the groups evaluated were as follows: (1) ‘Black’, (2) ‘White’, (3) ‘Coloured’, (4) ‘Indian’ people as well as (5) ‘foreigners living in South Africa in general’. These semantic differential measures were then used to construct five indexes: (1) Racial Ingroup (or ‘Black’ People), (2) White People, (3) Indian People, (4) Coloured People, and (5) Foreigners. Each constructed index ranged from 0 to 10 with the highest value indicating the most positive evaluation. The small number of respondents who answered ‘don’t know’ to any of these questions was coded as missing. Cronbach’s alpha was used to assess the reliability (or internal consistency) of each index, and these psychometric performance tests demonstrated the reliability of the indexes.

A number of standard demographic dummy variables were constructed for this study. Political affiliation was obtained from a question on which political party the respondent felt most close to. There are numerous political parties in the country and the responses had to be combined into a categorical variable: (1) the African National Congress, (2) the Opposition, (3) No Party, and (4) Refused/Undeclared.⁴ An interpersonal trust measure was constructed based on two items on trusting other people.⁵ This measure was captured on a scale which ranged from 0 to 10 with the higher value representing the greater level of interpersonal trust. Using two items on acquaintanceship and friendship with White people, a categorical variable was constructed that discerned those with no contact, only acquaintanceship contact, and those with friendship contact with people.

One of the most important units of analysis in any study of intergroup relations in South Africa is economic status. There are several measures that can be used as proxies for this type of status, but this study primarily looked at economic status through subjective deprivation. Since poverty or well-being is likely a reflection of multiple forms of deprivation in an individual’s life, it was important to capture multiple dimensions of subjective poverty. The method employed here requires using five items on whether a respondent’s household has less or more than adequate access to each of the following: (1) food consumption, (2) housing, (3) clothing, (4) health care, and (5) transport. Responses to these five items were combined into a 0–10 Multiple Deprivation Index with the highest value on the index indicating the highest level of deprivation. Psychometric performance tests, via factor analysis, demonstrated the validity and reliability of this index.

The respondents’ home language was used as a proxy for their ‘ethnic’ identity. Using this method, it was possible to distinguish between five major Black African ethnic groups: (1) Basotho, (2) Bapedi, (3) Batswana, (4) Xhosa, and (5) Zulu. Those Black Africans who identified a language other than these five (such as Swazi, Ndebele, or Venda) as their home language were coded as ‘Other Groups’. Although a more nuanced subdivision by language may have been preferred, data limitations prevented further categorisation. The five groups identified here are recognised linguistic identities within the country, but it is important to be conscious of the fact that linguistic or ‘ethnic’ identity is a social construct. Furthermore, data limitations also prevented the author from constructing a variable to control for the level of attachment an individual has to this constructed ‘linguistic’ identity.

Procedure

In each SAL, seven households were randomly selected. In each household, one adult was then selected as the respondent using the Kish grid method (Kish, 1994). Interviews were completed in the respondent’s private residence. Respondents were asked to complete a questionnaire via a face-to-face interview. The questionnaire was translated into the country’s major languages and interviews were conducted in the respondents’ home language when possible. All interviewers were trained in data collection. To avoid social desirability bias, an attempt was made to match the race of the interviewee to the race of the interviewer.

Ethical considerations

Each SASAS instrument receives clearance by the HSRC’s ethic committee before the fieldwork begins. All SASAS interviews are voluntary and respondents sign consent forms, authorising their participation. The data for this study were downloaded from HSRC under the terms and conditions of the organisation’s data usage policy.

Data analysis

The study used descriptive statistics on Black Africans’ attitudes towards, and contact with, White South Africans. Pairwise correlation techniques were then employed to investigate the correlations between the different semantic differential measures created for this study. However, pairwise correlation analyses were not sufficient to validate the central thesis proposed in the article. Multivariate techniques were used to confirm bivariate findings. Given that the White People Index was constructed as a continuous variable, a standard (ordinary least squares [OLS]) linear regression approach was employed. Three models were produced using STATA 13.1 software. The first includes the socio-demographic characteristics and the second introduces the contact variable. In the third model, the ingroup and outgroup evaluation variables were introduced, and this was the fully specified model.

Results

Black Africans’ mean scores on the different outgroup and ingroup indexes are represented in Table 2 for the period 2011–2015. White people were not, on average, viewed particularly negatively by adult Black Africans, and the reported index means were located above the midpoint for the period 2011–2015. Relative to the White minority, other racial outgroups were evaluated more (albeit moderately) negatively. In all years surveyed, foreigners were evaluated, on average, the most negatively. Attitudes towards the different groups were inclined to fluctuate only within the very narrow band over the period. The close patterning of results may suggest that attitudes towards one outgroup have a statistically significant correlation with attitudes towards another. Before proceeding to the bivariate tests, it would be informative to examine the level and type of contact between the Black African majority and the White minority.

Table 2.

Mean outgroup evaluation index scores among Black Africans, 2011–2015.

	White people		Coloured people		Indian people		Foreigners		Black people
	M	SE	M	SE	M	SE	M	SE	M	SE
2011	5.64	0.08	5.28	0.07	5.13	0.08	4.79	0.08	7.23	0.06
2012	5.56	0.09	5.04	0.08	5.01	0.10	4.57	0.10	7.01	0.08
2013	5.48	0.10	5.17	0.09	5.10	0.10	4.60	0.10	6.72	0.10
2014	5.77	0.09	5.07	0.09	5.29	0.10	4.87	0.10	7.09	0.08
2015	5.63	0.08	5.28	0.07	4.92	0.09	4.85	0.09	7.35	0.08

The share of the Black African population with a White friend in 2011 was low and declined over the period. In 2015, 65% of Black Africans reported not having a White friend, a decline of 7% points since 2011. A significant share (43%) of the adult Black African population said that they had no contact with White people in 2015 and 22% reported having White acquaintances but no White friends. Intergroup scholars warn that acquaintanceship contact is less effective than friendship contact in reducing prejudice and may even be detrimental to the project of improving intergroup attitudes (Davies, Tropp, Aron, Pettigrew, & Wright, 2011; Hewstone & Swart, 2011). Further analysis lends support to that supposition. An individual with no contact had more or less the same White People Index mean score (M = 5.45, SE = 0.119) as those with only acquaintanceship contact (M = 5.546, SE = 0.163). While the larger a Black African individual’s network of White friends, the greater their score on the White People Index. Scheffe and Bonferroni multiple-comparison tests substantiate this finding.

Pairwise correlation analyses of the different semantic differential measures constructed for this article were conducted. The outcomes revealed that Black Africans’ evaluations of White people were positively and robustly related to their attitudes towards the other outgroups. Pro-White sentiment was correlated more strongly with attitudes towards Coloured people (r =0.351) and Indian people (r =0.314) than with foreigners (r =0.194). It would appear that Black Africans’ readiness to accept negative stereotypes about one set of outgroups (e.g., foreigners, Indians, or Coloureds) has a significant and negative influence on their feelings about White people. Positive racial ingroup assessments had marginal (r =0.071) but positive and statistically significant correlation with views of White people. It seems that this pairwise correlation analysis disproves the article’s hypothesis that Black Africans’ evaluations of White people are unconnected to their evaluations of other racial minorities.

Multivariate regression techniques were used to confirm the results of the bivariate analysis. The results for all three models are depicted in Table 3. A positive coefficient in the model indicates a positive assessment of the White minority. Statistical testing reveals that each model’s fit was significantly better than an empty model. Model I accounted for 8% of variance in Black Africans’ evaluations of White people while Model II accounts for just 9% of variance. In other words, the interracial contact seems to have less value when attempting to explain Black Africans’ attitudes towards White people than may have been assumed. Model III explained 27% of variance in Black Africans’ evaluations of White people, making this model better at explaining variation in the dependent than the other models in Table 3.

Table 3.

Standard (OLS) linear regression Black Africans’ attitudes towards White South Africans, 2015.

	Model I			Model II			Model III
	Coeff.	β	Sig.	Coeff.	β	Sig.	Coeff.	β	Sig.
Female (ref. male)	0.26	0.05		0.29	0.06		0.08	0.02
Age	−0.01	−0.09	*	−0.01	−0.08		−0.01	−0.07
Marital status (ref. married)
Married before	−0.42	−0.05		−0.48	−0.06		−0.42	−0.05
Never married	−0.54	−0.11	*	−0.52	−0.10	*	−0.55	−0.11	**
Ethnic group (ref. Zulu)
Xhosa	−0.30	−0.05		−0.44	−0.07		−0.30	−0.05
Basotho	−0.99	−0.12	*	−0.94	−0.12	*	−0.56	−0.07
Batswana	−0.99	−0.13	**	−0.93	−0.13	*	−0.63	−0.09
Bapedi	0.15	0.02		0.21	0.03		0.15	0.02
Other groups	0.04	0.01		−0.01	0.00		0.06	0.01
Labour market status (ref. employed)
Part-time employed	−0.07	−0.01		−0.02	0.00		0.08	0.01
Unemployed	−0.09	−0.02		0.01	0.00		−0.04	−0.01
Labour inactive	−0.04	−0.01		0.01	0.00		−0.12	−0.02
Years of education	−0.02	−0.03		−0.04	−0.05		−0.02	−0.03
Multiple deprivation index	−0.01	−0.01		0.00	0.00		0.03	0.03
Interpersonal trust	0.18	0.15	***	0.16	0.13	***	0.11	0.09	**
Intergroup contact
White contact (ref. no contact)
Acquaintances only				0.06	0.01		−0.07	−0.01
A few friends				0.27	0.05		0.19	0.03
Some friends				0.59	0.06	*	0.26	0.03
Many friends				1.44	0.13	***	0.94	0.09	*
Intergroup indexes
Indian people							0.17	0.20	***
Coloured people							0.27	0.27	***
Foreigners							0.08	0.09	*
Racial ingroup							0.01	0.01
N			1644			1628			1434
Adj R²			0.076			0.094			0.265
RMSE			2.355			2.341			2.088

RMSE: root mean square error; OLS: ordinary least squares.

(1) Data are weighted to be nationally representative; (2) standardised β coefficients are reported instead of confidence intervals; (3) coefficients indicate evaluations for White people, with a high coefficient value indicating more positive evaluations; (4) the regressions control for political affiliation, province of residence, and geographic type; and (5) Significance level at: ***p <0.001, **p < 0.01, *p < 0.05.

In Model II having contact with White people improved a Black African individual’s attitudes towards that group. Relative to having no contact (and controlling for a range of socio-demographic variables), having only acquaintance contact or a few friends were not associated with the dependent variable at the 5% level of statistical significance. Only having some friends (r = 0.59, β = 0.06) or many friends (r = 1.44, β = 0.13) was positively correlated with the dependent at a statistically significant level in Model II. This finding casts doubt on the proposition that greater intergroup contact between the Black African majority and the White minority will considerably reduce Black African animosity towards (and stereotyping of) the latter. The observed premium on intergroup contact observed in Models III was even weaker than what was observed in Model II.

All the group evaluation indexes in Model III had a positive and statistically significant relationship with the dependent. The Coloured People Index (r = 0.27, β = 0.27) had the strongest influence followed by the Indian People Index (r = 0.17, β = 0.20). The Foreigners Index had a positive correlation (r = 0.08, β = 0.09) with attitudes towards White people but only at the 5% level of statistical significance. A Black African adult’s pro-Black attitudes were not a salient correlate with the dependent variable in Model III. In other words, pro-Black attitudes do not seem to lead Black Africans to distance themselves from the White minority. It can, therefore, be inferred from Table 3 that attitudes towards other racial minorities have a greater value in explaining Black Africans’ attitudes towards the White minority than racial ingroup evaluations.

Before discussing the limits of the results reported, it would be informative to consider how the demographic and economic independent variables in Table 3 correlate with Black African attitudes towards White people. The socio-economic indicators employed were not correlated with the dependent in any of the models in Table 3. This seems to indicate that economic status was not a good predictor of Black Africans’ pro-White sentiment. An unexpected finding concerned the correlation between interpersonal trust and pro-White sentiment among the Black African adult population. The more a Black African individual trusts others, the more likely that individual will be to view White people in a favourable light. The observed coefficients on this association were stronger in Model I (r = −0.18, β = −0.15) than in Model III (r = −0.11, β = −0.09).

Discussion

There are limitations to the study’s findings that should be considered, particularly as regards the independent variables used in the multivariate analysis. The remainder of this section will discuss those limitations. It could be argued that a Black African individual’s attitudes towards the White minority may be soured by economic deprivation. Yet this study found no evidence of a relationship between socio-economic status and attitudes towards White people. On the other hand, it may be that the measures used here were not sufficient to capture how economic position influences pro-White sentiment. In an attempt to further the analysis, the study experimented with an asset accumulation index. An accurate representation of asset accumulation was provided by the Living Standards Measurement (LSM)⁶ – an index developed by the South African Advertising Research Foundation. Both bivariate and multivariate techniques were unable to locate a statistically significant correlation between this LSM measure and the White People Index. The results of this study found that economic affluence was not associated with holding pro-White attitudes. This contradicts those (such as Bornman, 2011) who consider economic status to have a negative influence on Black Africans’ attitudes towards other race groups.

In the study, Black Africans’ contact with the White minority was only beneficial to prejudice reduction if that was contact intimate and if the network of friends was wide. The noted weakness of contact as a determinant is significant if not unanticipated given existing work on Black–White friendship contact. Ongoing histories of group marginalisation and oppression obviously continue to inform Black–White contact in South Africa and constrain the positive impacts of cross-group contact on intergroup attitudes among members of minority status groups. Finally, even though it was not the aim of the study, the findings presented are the result of a cross-sectional research design and, consequently, the directionality of relationships observed here have not been discerned. Instead of contact influencing attitudes, attitudes may be influencing individuals to avoid or seek contact. Longitudinal panel studies would address this limitation.

Pro-Black attitudes among adult Black Africans do not translate into denigration of White people in South Africa. This confirms prior research which shows that individuals’ negative attitudes towards racial outgroups are not necessarily a consequence of their enhanced racial ingroup identification. It could be argued that the study’s Racial Ingroup Index did not adequately capture the complexity of positive Black African ingroup identity in South Africa. Past research (e.g., Gibson, 2006) has employed more rigorous measures of racial identification. More multidimensional measures that captured the complexities of ingroup racial identification processes among the Black African majority were, however, unable in the dataset. In addition, the data presented has found that linguistic identity was not a good predictor of Black Africans’ pro-White sentiment. But the ethnicity variable utilised for this study, as has already been stated, may not adequately capture ethnic identifications among the country’s Black African majority. More public opinion research is required on the role of ingroup identifications in Black–White attitude formation in the country.

The limitations of the bivariate and multivariate analysis have been acknowledged above. However, there is a need to acknowledge the general limitations applicable to public opinion research of this type. The public opinion respondent is constrained by the close-ended structure of survey questions, and how the question and responses options are framed. The standardisation of SASAS response options may, consequently, hide nuance and detail in a respondent’s answer. Moreover, there is no mechanism to tell how much thought and honesty a respondent in SASAS has put in when answering the questionnaire.

Conclusion

The results presented in this article show that Black Africans’ evaluations of the White minority do not differ significantly from other racial minorities in South Africa. White people were, in fact, viewed more favourably than other race groups in the country. Positive attitudes towards White people were found to correlate with attitudes towards the country’s two other major racial minorities using pairwise correlations. In a multivariate analysis, this finding held even controlling for a range of socio-demographic characteristics (such as education or deprivation) as well as Black–White contact. The outcomes of this article disproved the article’s central hypothesis. The evidence provided seems to suggest that, in the mind of South Africa’s Black African majority, racial minorities seem to be part of a ‘band of racial others’. The results of this are remarkable given the highly divergent histories of the country’s three main racial minority groups. Whether South Africa’s current multiracial democracy will be successful depends on the development of congruous race relations in the country. The success of the country’s ongoing project of interracial reconciliation is, therefore, worthy of study.

Footnotes

Acknowledgements

The author would like to thank the Coordinators of the Democracy, Governance and Service Delivery research programme, Human Sciences Research Council (HSRC), for their cooperation and support during the writing of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Allport

G. W.

(1954). The nature of prejudice. Boston, MA: Beacon Press.

Bornman

(2010). Emerging patterns of social identification in postapartheid South Africa. The Journal of Social Issues, 66, 237–254.

Bornman

(2011). Patterns of intergroup attitudes in South Africa after 1994. International Journal of Intercultural Relations, 35, 729–748.

Brewer

M. B.

(1999). The psychology of prejudice: Ingroup love or outgroup hate?. The Journal of Social Issues, 55, 429–444.

Brown

(2011). Prejudice: Its social psychology. Oxford, UK: John Wiley & Sons.

Christopher

A. J.

(2009). Delineating the nation: South African censuses 1865–2007. Political Geography, 28, 101–109.

Davies

Tropp

L. R.

Aron

Pettigrew

T. F.

Wright

S. C.

(2011). Cross-group friendships and intergroup attitudes: a meta-analytic review. Personality and Social Psychology Review, 15, 332–351.

Dixon

Durrheim

Tredoux

(2007). Intergroup contact and attitudes toward the principle and practice of racial equality. Psychological Science, 18, 867–872.

Duckitt

Wagner

du Plessis

Birum

(2002). The psychological bases of ideology and prejudice: testing a dual process model. Journal of Personality and Social Psychology, 83, 75–93.

10.

Durrheim

Mtose

Brown

(2011). Race trouble: Race, identity and inequality in post-apartheid South Africa. Pietermaritzburg, South Africa: University of KwaZulu-Natal Press.

11.

Durrheim

Tredoux

Foster

Dixon

(2011). Historical trends in South African race attitudes. South African Journal of Psychology. Suid-Afrikaanse Tydskrif vir Sielkunde, 41, 263–278.

12.

Fishbein

H. D.

(2002). Peer prejudice: The origins of prejudice (2nd ed.). Boulder, CO: Westview Press.

13.

Gibson

J. L.

(2004). Overcoming apartheid: Can truth reconcile a divided nation? Cape Town, South Africa: HSRC Press.

14.

Gibson

J. L.

(2006). Do strong group identities fuel intolerance? Evidence from the South African case. Political Psychology, 27, 665–705.

15.

Gibson

J. L.

Claassen

(2010). Racial reconciliation in South Africa: Interracial contact and changes over time. The Journal of Social Issues, 66, 255–272.

16.

Goldstein

E. L.

(2006). The price of whiteness: Jews, race, and American identity. Princeton, NJ: Princeton University Press.

17.

Hewstone

Swart

(2011). Fifty-odd years of inter-group contact: from hypothesis to integrated theory. British Journal of Social Psychology, 50, 374–386.

18.

Johnson

V. D.

(2017). Coloured South Africans: A middleman minority of another kind. Social Identities, 23, 4–28.

19.

Kish

Multipopulation survey designs: Five types with seven shared aspects. International Statistical Review/Revue Internationale de Statistique. 1994;62:167-186.

20.

Mynhardt

J. C.

(2013). Intergroup attitude change in South Africa: A thirty-seven year longitudinal study. Journal of Psychology in Africa, 23, 549–560.

21.

Ramsamy

(2007). Between non-racialism and multiculturalism: Indian identity and nation building in South Africa. Tijdschrift voor Economische en Sociale Geografie, 98, 468–481.

22.

Shelton

J. N.

(2003). Interpersonal concerns in social encounters between majority and minority group members. Group Processes & Intergroup Relations, 6, 171–185.

23.

Swart

Hewstone

Christ

Voci

(2010). The impact of crossgroup friendships in South Africa: Affective mediators and multigroup comparisons. The Journal of Social Issues, 66, 309–333.

24.

Talbot

Durrheim

(2012). The Princeton trilogy revisited: How have racial stereotypes changed in South Africa?. South African Journal of Psychology. Suid-Afrikaanse Tydskrif vir Sielkunde, 42, 476–491.

25.

Thomsen

J. P. F.

Rafiqi

(2016). The contact–prejudice relationship among ethnic minorities: Examining personal discrimination as a boundary condition. Ethnic and Racial Studies, 39, 1886–1904.

26.

Tredoux

Finchilescu

(2010). Mediators of the contact-prejudice relation among South African students on four university campuses. The Journal of Social Issues, 66, 289–308.