SEARCH

SEARCH BY CITATION

Keywords:

  • Prejudice;
  • Britain;
  • ethnic minorities;
  • attitudes;
  • cohort;
  • diversity

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Racism in decline?
  5. Is racial prejudice an attitude or a ‘value’?
  6. The nature of British prejudice: aversion to ‘blackness’ or cultural racism?
  7. Social structural factors: education, class and gender
  8. Data and methods
  9. Results
  10. Multivariate analysis
  11. Interaction effects: diverging attitudes among the young
  12. Discussion
  13. Bibliography

This article employs two previously neglected indicators of racial prejudice from the British Social Attitudes surveys to examine the social distribution of prejudices against black and Asian Britons. Three hypotheses are proposed and tested: that racial prejudice is declining in Britain; that this decline is principally generational in nature; and that greater prejudice is shown towards more culturally distinct Asian minorities than black minorities. Strong evidence is found for the first two hypotheses, with evidence of an overall decline in prejudice and of a sharp decline in prejudices among generations who have grown up since mass black and Asian immigration began in the 1950s. Little evidence is found for the third hypothesis: British reactions towards black and Asian minorities are broadly similar suggesting racial differences may still be the main factor prompting white hostility to British minorities.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Racism in decline?
  5. Is racial prejudice an attitude or a ‘value’?
  6. The nature of British prejudice: aversion to ‘blackness’ or cultural racism?
  7. Social structural factors: education, class and gender
  8. Data and methods
  9. Results
  10. Multivariate analysis
  11. Interaction effects: diverging attitudes among the young
  12. Discussion
  13. Bibliography

Racism and racial discrimination remain a part of everyday life for Britain's ethnic minorities. Black and Asian Britons, who constitute nearly a tenth of the population (Office for National Statistics 2003) are less likely to be employed and are more likely to work in worse jobs, live in worse houses and suffer worse health than white Britons (Modood et al. 1997; Blackaby et al. 2002). Although some of these differences are the result of factors like lower qualifications and residence in more depressed parts of the country, academic studies also find a consistent gap or ‘ethnic penalty’ (Heath and Cheung 2007) that cannot be explained by these. White racial prejudices are likely to be an important explanation for these disadvantages, shown in experimental studies where black and Asian job applicants with identical qualifications to white applicants were disproportionately rejected by employers (Noon 1993; Esmail and Everington 1993).

While the impact of racial prejudices on the lives on Britain's minorities has been much studied, much less academic attention has been paid to examining the distribution of such prejudices among the white population, despite their dramatic impact on the lives of black and Asian Britons. There have only been two academic studies specifically attempting to examine the patterns of white racial prejudice, and these have only undertaken limited analysis using an unreliable measure of racial prejudice (Evans 2002; Rothon and Heath 2003).2

This article utilizes two unexamined indicators of white racial prejudice to analyse the trends in and the social distribution of racial prejudice among white Britons. It makes three key contributions. Firstly, it shows how racial attitudes in Britain are structured by generation, with a large decline in expressions of prejudice among the cohorts who have grown up since immigration began. Consequently, levels of racial prejudice are falling and are likely to fall further. Secondly, it casts light on the social distribution of racial prejudice. Levels of racial prejudice are much lower among the highly educated, the professional classes and among women. The attitudes of these groups also change more rapidly between generations, resulting in a divergence in racial attitudes among the young. Thirdly, it shows that there are few differences in the levels and social distribution of white prejudice against black and against Asian Britons, suggesting the more distinctive cultural and religious practices of British Asian communities do not attract greater hostility from the white population.

Racism in decline?

  1. Top of page
  2. Abstract
  3. Introduction
  4. Racism in decline?
  5. Is racial prejudice an attitude or a ‘value’?
  6. The nature of British prejudice: aversion to ‘blackness’ or cultural racism?
  7. Social structural factors: education, class and gender
  8. Data and methods
  9. Results
  10. Multivariate analysis
  11. Interaction effects: diverging attitudes among the young
  12. Discussion
  13. Bibliography

Four major factors may be encouraging a decline in white racial prejudices in Britain. Firstly, racial prejudices now lack the intellectual and cultural supports they had in the past. Arguments about white biological, economic and cultural superiority have lost legitimacy over the past fifty years. Claims of white racial superiority have been thoroughly debunked by geneticists and biologists, and have been politically and socially marginalized as a result of their strong association with Nazi fascism and the Holocaust. The belief that white European cultures are economically superior has become harder to maintain in the face of rapid economic development in East Asia, India and elsewhere. Rudyard Kipling's ‘white man's burden’ argument that only white rule can produce stable and democratic governance in poor colonial societies, has been dealt a severe blow by the manifold abuses and subsequent collapse of the European colonial empires and the white minority regimes in South Africa, Zimbabwe and the American South.3 While these biological, economic and political arguments for white superiority are not the only bases on which racial prejudices form, they are powerful supports for such prejudices as they all make a strong and allegedly empirical case for the inferiority of non-white groups.

Secondly, racially prejudiced arguments have also lost all elite political support in Britain. The rapid increase in black and Asian immigration in the 1960s and 1970s was accompanied by a surge in support for openly discriminatory politicians, such as Enoch Powell and his supporters within the Conservative party and the extreme right National Front (Schoen 1977; Husbands 1983). However since the early 1980s, when reforms to immigration policy sharply reduced immigration rates4 and quelled public anxiety, expressions of racial prejudice have become unacceptable among Britain's political elites. The need to integrate Britain's growing ethnic minority populations and combat the damaging effects of prejudice and discrimination have lead to a political consensus built around the celebration of diversity (Solomos 1989; Layton-Henry 1992). Public figures who make even mildly prejudiced comments are now roundly condemned and swiftly despatched.

Thirdly, it has become harder to argue that Britain's ethnic minority groups are not truly ‘British’. The white British are exposed on an everyday basis to black and Asian Britons on the television, the radio and the sports field. As a result, the ethnic minorities these elite performers represent occupy a much more prominent place in the ‘imagined community’ (Anderson 1991) of the British nation. As Paul Gilroy puts it: ‘Nobody post-Lenny, Lennox and Linford, Scarey, Denise and Naomi, could doubt that there is, after all, some kind of black somewhere in or rather underneath the Union Jack.’ (Gilroy 2002: xxiv). This argument is supported by empirical analysis of British national identity beliefs, which show a sharp decline in the importance accorded to whiteness and British ancestry as markers of ‘being British’ (or English, Scottish, Welsh) among those born since mass immigration began (Tilley, Exley and Heath 2004).

Fourth, social contact with ethnic minorities has also become a much more frequent occurrence for many white Britons. Britain's ethnic minority communities have grown rapidly and have become more geographically dispersed, with large numbers migrating internally as they grow older and richer, moving from the inner city areas where they first settled to the more prosperous and previously all-white suburbs and rural areas (Simpson 2007). Ethnic minorities have also become more socially dispersed. First generation migrants, who often lacked British qualifications or English language skills, were concentrated in poorly paid manual jobs. Their British born children and grandchildren are much better qualified and many more have succeeded in securing white-collar jobs, despite continuing discrimination (Berthoud 2000; Heath and Cheung 2007). As a result of these developments most white Britons are now much more likely to regularly encounter black and Asian faces in their neighbourhoods and in their workplaces. This increases the chances of informal social contact in relaxed environments, which research has shown helps to break down racial prejudices (Pettigrew 1997; McLaren 2003). These four prominent social trends all suggest that levels of white prejudice against black and Asian minorities should now be declining.

Hypothesis 1: White prejudice against ethnic minorities is declining in Britain.

Is racial prejudice an attitude or a ‘value’?

  1. Top of page
  2. Abstract
  3. Introduction
  4. Racism in decline?
  5. Is racial prejudice an attitude or a ‘value’?
  6. The nature of British prejudice: aversion to ‘blackness’ or cultural racism?
  7. Social structural factors: education, class and gender
  8. Data and methods
  9. Results
  10. Multivariate analysis
  11. Interaction effects: diverging attitudes among the young
  12. Discussion
  13. Bibliography

An important issue that has received relatively little attention in debates about racism in Britain is the nature of racially prejudiced beliefs. Most beliefs expressed in public opinion can be characterized as either attitudes or as values. Attitudes are less stable than values, and often less informed. They are ‘off the top of the head’ opinions formed ‘on the basis of whatever considerations are most immediately salient’ (Zaller 1992: 266). Values are less responsive to changes in current conditions, tending instead to be moulded by conditions in an individuals’ youth and early adulthood, and remaining stable thereafter (Mannheim 1952; Inglehart 1997). While considerable controversy exists about where most public attitudes fit on this spectrum it is generally accepted that complex political issue debates, which tend to elicit less stable and coherent responses from the public, fall more on the ‘top of the head’ side of the spectrum (Converse 1964; Zaller 1992). Core value orientations, such as general ideological preferences for redistribution or the free market, tend to fall more on the ‘stable, formed in youth’ side (Evans, Heath and Lalljee 1996; Tilley 2002).

We may expect racial prejudice to behave more like a value for two reasons. Firstly, while policy debates about diversity and multiculturalism are complex and multifaceted, the underlying division over race and racism is clear and widely understood: an individual either holds negative views about an ethnic group or she does not. Secondly, racial attitudes involve an explicit value judgment: expressing prejudice means rejecting the right of a racially defined group to equal regard and equal treatment. Racial prejudices may therefore be more like core value orientations, with individuals retaining throughout their lives the general views about ethnic minority groups that they formed in youth.

The kind of decline we can expect in racial prejudice will be affected by the kind of belief it is. If racial prejudice is primarily an attitudinal response to current circumstances, then the shift in attitudes should be more rapid and more evenly distributed across all respondents, as they will all be sensitive to the changed social environment. On the other hand, if prejudices against minority groups resemble other core political values, we may expect them to change more slowly, and to show a generational structure. The rapid ethnic diversification of Britain since 1950 means younger generations are far more exposed to black and Asian minorities during the years when values are formed, both directly through social contact and indirectly through media and political elites. We should therefore see progressive declines in racial prejudice in generations growing up since the beginning of mass migration, as the members of each successive cohort will have had progressively more contact with minorities and minority culture during their impressionable years. However, although shifts in attitudes between generations may be large, the overall pace of attitude change will be much slower, as it will be primarily driven by generational replacement.

Hypothesis 2: The pattern of change in racial prejudice will be affected by whether prejudice behaves like an attitude or a value:

(a) If prejudice behaves like an attitude, responses will be evenly distributed across age groups and changes in responses will be rapid.

(b) If racial prejudice is a value, responses will be structured by generation, and the overall rate of change in responses will be slower.

The nature of British prejudice: aversion to ‘blackness’ or cultural racism?

  1. Top of page
  2. Abstract
  3. Introduction
  4. Racism in decline?
  5. Is racial prejudice an attitude or a ‘value’?
  6. The nature of British prejudice: aversion to ‘blackness’ or cultural racism?
  7. Social structural factors: education, class and gender
  8. Data and methods
  9. Results
  10. Multivariate analysis
  11. Interaction effects: diverging attitudes among the young
  12. Discussion
  13. Bibliography

Much of the most insightful writing on race and racism in modern Britain has been less focused on changes in the overall prevalence of racial prejudices than with understanding the origins and content of racist beliefs and the related development of anti-racist ideologies used to combat them (Centre for Contemporary Cultural Studies 1982; Gilroy 1987, 2002; Modood 1994, 2005). Early studies of British racism, while acknowledging that cultural differences also provoke hostility, were in no doubt that the most important factor provoking hostility and discrimination was skin colour (Daniel 1968; Rose and associates 1969). Reacting to these findings, anti-racist theorists of the 1970s and 1980s argued that the cultural differences between black and Asian minorities were less important than their common experience of white discrimination and hostility as ‘black’ groups. They believed that the best strategy to combat racial prejudice was for all ‘black’ minorities to mobilize against it together (Gilroy 1987, 2002; Solomos 1989).

More recently, Nira Yuval-Davis, Tariq Modood and others have taken issue with this approach (Anthias and Yuval-Davis 1992; Modood 1994, 2005). These scholars argue that in modern Britain, where biological arguments about race have lost all legitimacy, ‘colour racism’ is a declining problem. Colour matters as a means for identifying members of cultural minorities, and it is the perception of essential and irreducible differences in values and cultural practices between such racially identified groups and the majority which generates prejudice and discrimination, a phenomenon Modood calls ‘cultural racism’. Modood argues that this cultural racism particularly affects Asian minorities, whose cultures, languages and religious beliefs differ more from the majority, and attract greater white hostility:

South Asians . . . are clearly visible as a non-white group: they are a principal object of racist victimization, of negative treatment by whites on the grounds that they are an undesirable ‘Other’. They suffer, therefore, from color racism. But they also suffer from cultural racism: a certain culture is attributed to them, is vilified, and is even the ground for discrimination . . . .The perception of Asians . . . is that their defects lie deep in their culture rather than in a biology that produces their culture. This means that Asians, more than blacks, suffer a double racism. (Modood 2005: 7, emphasis added)

One of the original theorists of a unified ‘blackness’ experienced by all British minorities, Paul Gilroy, has in recent writing lamented this ‘turn away from the simple efficacy of blackness’. He claims it has made ‘anti-racism less politically focused and certainly more difficult if not impossible to organise’ (Gilroy 2002, p. xiv). Gilroy dismisses claims about the primacy of cultural markers of difference as ‘an imploded, narcissistic obsession with the minutiae of ethnicity’, although he provides no empirical evidence for his claim that British racism has been, and continues to be, primarily a dualistic black/white phenomenon. Modood, by contrast, cites evidence from documentaries and academic studies to support his contention that culture now matters as much as colour, to the disadvantage of Asians, who are victims of a ‘double racism’ (Modood 2005: 33–7) as a result. Two competing predictions can therefore be drawn from this debate about the nature of British racism:

Hypothesis 3 (a) If racial prejudice is primarily a matter of ‘blackness’, there should be little difference in the level of prejudice directed at black and Asian minorities.

(b) If racial prejudice also has a significant cultural component, white Britons will express more prejudice against Asians than blacks.

Social structural factors: education, class and gender

  1. Top of page
  2. Abstract
  3. Introduction
  4. Racism in decline?
  5. Is racial prejudice an attitude or a ‘value’?
  6. The nature of British prejudice: aversion to ‘blackness’ or cultural racism?
  7. Social structural factors: education, class and gender
  8. Data and methods
  9. Results
  10. Multivariate analysis
  11. Interaction effects: diverging attitudes among the young
  12. Discussion
  13. Bibliography

Three individual characteristics may also play an important role in attitudes towards minorities: education, social class and gender. Previous work has suggested that education, particularly higher education, encourages tolerant attitudes on a range of dimensions (Sullivan, Piereson and Marcus 1982; Sullivan and Transue 1999). More educated respondents are less likely to perceive ethnic minorities as threatening and they are also less likely to hold authoritarian values associated with racial prejudice (Hello, Scheepers and Sleegers 2006). Educated respondents also pay more attention to the views of political and media elites (Zaller 1992) and are therefore more likely to adopt the elite norm against racial prejudice. As a result, prejudice levels are likely to be lower among the highly educated. The second half of the twentieth century saw large increases in spending on education in Britain. Younger Britons are therefore much better qualified than their parents and grandparents, which may contribute to generational declines in prejudice.

Ethnic minority immigrants to Britain have tended to concentrate in lower status manual occupations. This is the result of the lower English language proficiency and qualification levels of black and Asian immigrants, and the prejudices of some British employers. If the perception that Britain's minorities are an economic threat is an important source of hostility towards minority groups (Bobo 1983) then such hostility is likely to be concentrated in the lower social classes, with whom ethnic minorities have in the past competed most directly for jobs and government welfare assistance. Hostility may also be elevated among those who are directly dependent on government assistance, such as the unemployed and those living in council housing, who may also be concerned about competition for these resources from minority groups.

There may also be a gender gap in racial prejudice. Some recent studies of American racial prejudice have found higher levels of hostility to ethnic minorities among men (Kuklinski, Cobb and Gilens 1997; Johnson and Marini 1998). Studies of the European radical right have also suggested that men are over-represented among the support base for such parties, which focus prominently on mobilizing xenophobic hostility to immigrants (Lubbers, Gijsberts and Scheepers 2002; Norris 2005). Johnson and Marini (1998) suggest that this effect could be the result of gender differences in socialized values: women are socialized to be more ‘other oriented’ than men. As a result they place greater value on cooperation, interpersonal relationships and providing assistance to those in difficulty – values which encourage more open and tolerant views towards outgroups. This gender difference in racial attitudes has never been quantitatively examined in a British context, but the socialized gender roles of women and men are similar in Britain to those in America, so we should expect the effect to work similarly.

Hypothesis 4: (a) Prejudice levels will be lower among the highly qualified. (b) The rapid rise in qualification levels between generations will contribute to generational declines in prejudice

Hypothesis 5: Racial prejudice will be higher among those from lower status manual occupations, those who are unemployed and those living in council housing as they will perceive these groups as competitors for jobs and government assistance.

Hypothesis 6: Racial prejudice will be higher among men than among women.

Data and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Racism in decline?
  5. Is racial prejudice an attitude or a ‘value’?
  6. The nature of British prejudice: aversion to ‘blackness’ or cultural racism?
  7. Social structural factors: education, class and gender
  8. Data and methods
  9. Results
  10. Multivariate analysis
  11. Interaction effects: diverging attitudes among the young
  12. Discussion
  13. Bibliography

The data employed in this analysis comes from the British Social Attitudes (BSA) surveys, a series of national surveys which have been carried out in most years since 1983.5 The surveys used in this analysis were conducted between 1983 and 1996, when a module of items on attitudes about immigration and ethnic minorities was repeated seven times. This analysis focuses on two items dealing with how comfortable respondents feel about social contact with ethnic minorities. They are worded as follows:

Ethnic minority boss:

  • (a) 
    Do you think most white people in Britain would mind or not mind if a suitably qualified person of black or West Indian/Asian origin were appointed as their boss? (If mind) A lot or a little?
  • (b) 
    And you personally? Would you mind or not mind? (If mind) A lot or a little?

Ethnic minority spouse:

  • (a) 
    Do you think that most white people in Britain would mind or not mind if one of their close relatives were to marry a person of black or West Indian/Asian6 origin? (If mind) A lot or a little?
  • (b) 
    And you personally? Would you mind or not mind? (If mind) A lot or a little?

The samples were randomly split so that half of white respondents were asked about their views on blacks/West Indians and half were asked about Asians. These racial categories –‘black’ and ‘Asian’– are rather crude and obscure a great deal of potentially significant differences, for example between African and Caribbean blacks, and between Muslim and Hindu Asians. Unfortunately, British surveys have not regularly employed detailed ethnic categorizations, so data are not available on attitudes towards these more disaggregated minorities. The black and Asian categories, while crude, still allow us to broadly gauge levels of hostility to racially distinct minorities, and to test Modood's ‘double disadvantage’ hypothesis.

These two social distance items are chosen for three reasons. Firstly, they have been asked repeatedly over a 13 year run of data, allowing more detailed analysis of the trends and distribution of racial attitudes than has hitherto been undertaken. The only other racial prejudice measure available over a long period in Britain is a question asking respondents to rate their own prejudice, which was employed in two studies examining white racial attitudes (Evans 2002; Rothon and Heath 2003). This ‘self-rated prejudice’ item is, however, seriously flawed. While respondents who consider themselves prejudiced do hold distinctly more hostile attitudes towards ethnic minorities on a range of other items, the conception of prejudice applied by respondents varies systematically with education and cohort: older cohorts and less educated respondents express much greater hostility to ethnic minorities on other measures while still rating themselves unprejudiced.7 As this analysis is concerned with the change in attitudes over generations, and the effects of education on expressed racial prejudice, these differences in the interpretation render this item an unsuitable indicator.

The social distance measures employed here have a high level of theoretical and empirical validity. Such items provide concrete examples of the kind of irrational hostility towards a group that is considered to be the essence of prejudice, but are also rooted in social situations that are easy for respondents to comprehend. American studies have employed social distance items to measure trends and patterns of racial prejudice for over seventy years and have confirmed that these measures correlate closely with other racial attitudes and measures of racial prejudice (Bogardus 1967; Firebaugh and Davis 1989; Kleg and Yamamoto 1998).

Evidence from the BSA surveys themselves provides further confirmation that these measures are valid and substantively consequential. As Table I shows, those who say they dislike social contact with ethnic minorities in the workplace or the family are significantly less supportive of a range of policies benefiting ethnic minorities. These differences in attitudes are largest when the policies in question involve special provision of resources to minority groups, such as aid to ethnic minority immigrants or special lessons for their children.

Table I. Measures of social distance and attitudes on other ethnic minority issues
 Social distance (spouse)Social distance (boss)
Mind a lotMind a littleDon't mindDiffMind a lotMind a littleDon't mindDiff
  • Notes:

  • Support present levels of immigration or favour increase in immigration

  • ** 

    Support present levels of aid or favour increase in aid

  • All differences are significant (p < 0.05)

Equal opportunities policies
Support anti-discrimination legislation (1994)58%72%81%2341%65%79%38
Equal opportunities legislation in Britain should go further (1994)14%28%45%3111%22%40%29
Levels of immigration
Asian immigration levels acceptable (1994)*11%23%48%379%12%41%32
West Indian immigration levels acceptable (1994)*16%32%53%3712%20%47%35
EEC immigration levels acceptable (1994)*43%49%65%2235%37%62%27
Australian immigration levels acceptable (1994)*63%62%72%959%58%70%11
Assistance for immigrants
View on aid for EM immigrants (1991)**36%66%79%4312%49%71%59
Extra English classes for immigrant children (1989)71%83%86%1559%78%84%25
Allow traditional dress for immigrant children (1989)28%45%52%2417%24%50%33
Allow lessons in mother-tongue for immigrant children (1989)13%16%26%136%11%23%17
N (smallest)272315665 691071095 

The analysis begins by examining the trends in these social distance items over time and the cohort structure of the attitudes, to test for generational differences. Attitudes towards black and Asian minority groups are compared, to see if there is any evidence of greater aversion to social contact with Asians as a result of their greater cultural distinctiveness.

The second section conducts multivariate statistical analysis of the social distance items. The third section introduces interaction terms to the models, allowing the pattern of generational change in attitudes to vary across social groups.

The responses to the social distance questions fall into three categories –‘mind a lot’, ‘mind a little’, and ‘don't mind at all’. Although these answers have a clear ordering, it is not obvious that the categories are equally distant from each other. As such, linear regression of the responses is inappropriate. The data are instead analysed using ordered logistic regression (Hoffman 2004: ch.4; Long and Freese 2006: ch.5). Ordered logistic regression assumes that the different response categories form part of a continuous unobserved variable, in this case ‘aversion to social contact’. When the underlying level of hostility passes a threshold, individuals’ observed responses change from one category to another. The ordered logit model estimates these threshold points from the data, and presents them as a series of ‘cut point’ coefficients, representing the probability of a particular category occurring.8 The effects of independent variables are assumed to operate equally on all categories, and are reported as coefficients representing the effect of the independent variable on the likelihood of any particular category occurring. These coefficients can then be combined with the cut points to estimate the probability of each category occurring for each value of an independent variable9. The raw coefficients generated by ordered logistic regression models are not easy to interpret, so the models are accompanied by tables and charts illustrating the effects of particular variables on the probability of intolerant responses on the social distance items.

The analysis examines period effects, the change in attitudes between survey years and cohort effects, the change in attitudes between generations. Respondents’ attitudes could also, however, be affected by the ageing process. Hostility towards outgroups is closely associated with authoritarian values (Rowatt and Franklin 2004) which Altemeyer (1995) has argued increase as people age, particularly after significant life cycle changes such as marriage and having children. Older voters may also be more concerned about perceived threats they associate with ethnic minorities, such as crime and terrorism, which may encourage greater hostility. If such age differences do exist, they should be separated from the effects of generational attitude change. Thus the models should ideally include measures of age, period and cohort simultaneously.

This task is complicated by the ‘identification problem’: age, survey period and birth cohort are logically linked, so that the effects of two are always confounded (Mason and Fienberg 1985; Tilley 2002). Restricting assumptions need to be made in order to estimate all three effects. Three different sets of restricted models are estimated to address this problem. Firstly, period and cohort effects are estimated with age effects restricted to the ‘social ageing’ factors found significant by Altemeyer in relation to authoritarian values (Altemeyer 1995): marital status and the presence of children in the household. Secondly, cohort and age effects are estimated with period effects restricted to zero. Thirdly, age and period effects are estimated with cohort effects restricted to four wide time bands. The reported effects of age, period and cohort were robust across these different specifications, so only the results from the first specification are presented.

The effect of education is tested using dummies for possession of a degree, A-levels and GCSEs/O-Levels, with the unqualified as a reference group. Class is measured using a 5 category Goldthorpe-Heath classification, with the working class as a reference group. Controls for religious observance are also included, with dummies for self-identified denomination and regularity of church attendance. Gender is measured using a dummy for men.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Racism in decline?
  5. Is racial prejudice an attitude or a ‘value’?
  6. The nature of British prejudice: aversion to ‘blackness’ or cultural racism?
  7. Social structural factors: education, class and gender
  8. Data and methods
  9. Results
  10. Multivariate analysis
  11. Interaction effects: diverging attitudes among the young
  12. Discussion
  13. Bibliography

The trends in social distance attitudes over the period from 1983–96 are plotted in Figure I. The solid lines indicate the total proportion of respondents expressing opposition to each form of social contact, while the dashed lines show the proportion who express strong opposition. Significant hostility to contact with ethnic minority exists in the white British population. Between 10 and 20 per cent of British survey respondents object to an ethnic minority boss, and between 30 and 50 per cent object to a close relative marrying someone black or Asian. Around half of these respondents register strong objections.

image

Figure I. Period trends in social distance Source: British Social Attitudes Survey 1983–1996; N = 11,729

Download figure to PowerPoint

As hypothesized, both measures of racial prejudice decline markedly over the period under examination, particularly in the 1990s. Levels of overall opposition to both forms of social contact nearly halve after 1989, with levels of strong opposition falling even more steeply. There is little evidence of higher prejudice against Asians. Respondents do express higher opposition to working for an Asian employer throughout the series, but the difference in attitudes is small. Attitudes to racial inter-marriage are almost identical for both groups in every survey.

Figure II shows how levels of racial prejudice break down by birth cohort, using data from all the survey years combined.10 There are large generational differences in prejudice levels, with the proportion of respondents giving discriminatory responses falls steeply between cohorts on all four items. The decline is linear for hostility to a black or Asian employer, but non-linear for attitudes to racial inter-marriage. Hostility to relatives taking a black or Asian spouse is fairly uniform among the oldest generations, then falls rapidly among cohorts born after the mid-1930s. This pattern of generational change suggests prejudice may be a value-type attitude, primarily influenced by conditions in youth. Generations brought up in an ethnically homogenous Britain express high levels of prejudice, while those who have come up since mass migration began express progressively more tolerant attitudes.

image

Figure II. Social distance by cohort Source: British Social Attitudes Surveys 1983–1996; N = 11,729

Download figure to PowerPoint

The generational declines in racial prejudices are very steep. Opposition to working for a black or Asian boss falls by half from 30 per cent in the pre-1920 cohorts to 15 per cent in the cohorts from the 1970s. Strong opposition falls from 15 per cent to nearly zero. Opposition to racial intermarriage falls even further, from over 60 per cent in the oldest cohorts to under 20 per cent in the youngest, with strong opposition falling from 40 to 10 per cent. There is again little evidence of elevated hostility towards Asians. The only notable divergence in attitudes actually runs in the other direction: the oldest cohorts express rather more opposition to working under a black boss and are more likely to strongly oppose white–black intermarriage than white–Asian intermarriage.

Multivariate analysis

  1. Top of page
  2. Abstract
  3. Introduction
  4. Racism in decline?
  5. Is racial prejudice an attitude or a ‘value’?
  6. The nature of British prejudice: aversion to ‘blackness’ or cultural racism?
  7. Social structural factors: education, class and gender
  8. Data and methods
  9. Results
  10. Multivariate analysis
  11. Interaction effects: diverging attitudes among the young
  12. Discussion
  13. Bibliography

Initial analysis thus suggests a declining trend in prejudice which is being driven by generational change. Table II shows results of ordered logistic regression analyses, which test the influence of a wider range of factors on racial prejudice.Figure III plots the effects of the main significant variables in the ethnic spouse models, showing how each variable changes the proportion of respondents opposing racial intermarriage. All the other variables in the model are held constant at their means. The grey bars show the proportions expressing prejudice against Asians, the black bar the proportion expressing prejudice against blacks. The figures employ different scales to make presentation of the effects as clear as possible.

Table II. Ordered logistic regression models
 Dislike social contact with
Minority bossMinority spouse
AsianBlackAsianBlack
  1. Source: British Social Attitudes Surveys 1983–1996

  2. Notes: Significant figures in bold: * =p < 0.05; ** =p < 0.01; *** =p < 0.001

Cut point 11.181.64−0.20−0.10
Cut point 22.042.700.841.00
Cohort effect−0.009**−0.018***−0.024***−0.028***
Cohort squared/1000−0.070.17−0.34***−0.27**
Period effects (reference: 1980s)
1990s−0.27***−0.39***−0.33***−0.43***
Education (reference: no qualifications)
Degree−0.79***−0.51***−0.66***−0.26*
A-level−0.53***−0.14−0.17−0.03
GCSE−0.160.08-0.030.12
Class (reference group: working class)
Salariat−0.20*−0.14−0.18*−0.24**
Routine non-manual−0.02−0.14−0.06−0.02
Petty bourgeoisie0.170.130.030.10
Manual formen/supervisors−0.090.03−0.040.06
Economic conflict
Unemployed0.010.01−0.02−0.007
Living in council housing−0.20*−0.16−0.13*−0.11
Religion (reference group: no religion)
Church of England0.16*0.030.28***0.16**
Roman Catholic−0.15−0.240.090.14
Attendance > once a month−0.34**−0.32**−0.17**−0.16*
Gender
Male0.32***0.19*0.080.05
Lifecycle events (reference group: widowed, no kids)
Single−0.160.37*−0.29*−0.14
Living as married−0.190.06−0.150.09
Married−0.210.07−0.0020.13
Divorced−0.120.15−0.13−0.06
Kids in the house−0.05−0.06−0.020.05
Pseudo R square0.0290.0350.0520.051
N6,0775,8666,0415799

There is a significant period effect in all the models. A sharp shift in attitudes occurs between the surveys from the 1980s and 1990s, with racial prejudice 5 to 10 percentage points lower in the 1990s surveys than in those conducted in the 1980s.11 There is little evidence that this change is a methodological artefact. Question wordings, question orderings, and the placement of these questions in the overall survey were not changed between the surveys from the 1980s and 1990s.12 Several factors may be contributing to this sharp decline in racial prejudices in the 1990s. The political environment may have become more tolerant, as the divisive Margaret Thatcher, who had made several controversial statements on race and immigration issues in 1990s, was replaced in 1990 by the more conciliatory John Major. The main social development of the period was a deep economic recession in 1990–1, which may have increased the salience of economic over social issues in most white respondents’ minds. We might also expect that the controversy over the Satanic Verses in 1989 and Britain's participation in the first Gulf War, fought against a Muslim nation, may have increased the salience of religious divisions over racial ones (Modood 1990). The period change in attitudes is somewhat larger for views of blacks than it is for attitudes to Asians, but unfortunately no data are available on attitudes towards Muslims specifically that might enable closer examination of this question.

While prejudice does shift significantly downwards in surveys from the 1990s, the most important predictor of hostility to minorities in all the models is birth cohort.13 Younger generations consistently show much less hostility to social contact with black and Asian Britons. This cohort effect is nonlinear for the spouse item, with a significant widening of the generation gap in attitudes for cohorts born after the 1930s.14 Predicted opposition to racial intermarriage falls from over 60 per cent in the 1910 cohort to around 25 per cent in the 1970 cohort. The proportion strongly objecting falls from over one third to around one tenth. Cohort also has the largest impact on attitudes to a black employer, with opposition levels falling from over a quarter of the 1910 cohort to under a tenth of the 1970 cohort. Generational shifts in attitudes to an Asian employer are smaller, though they remain the second largest effect in the model: opposition falls 7 percentage points between these cohorts, from 20 per cent to 13 per cent.

There is little evidence that life cycle changes or ageing processes are responsible for these cohort differences. Social ageing events such as marriage and family formation have little impact on prejudice. Only those who are single show any difference in attitudes, and these effects are only marginally significant. Models including a continuous age term and a restricted cohort term did not fit the data as well as the reported models. The size, significance and robustness of these cohort effects suggest that racial prejudices are indeed value-type attitudes: older generations have adopted the discriminatory views about race which were prevalent when they grew up and retained them despite their declining legitimacy and social unacceptability.

The level and social distribution of prejudices against black and Asian Britons are very similar, suggesting that Gilroy and his colleagues may be right to argue that visible racial differences are the most important factor driving white reactions. The models do suggest some differential discrimination among the oldest generations, but in the opposite direction to that expected by Modood. Britons born in the 1930s or earlier tend to express more hostility towards blacks than Asians on both items, with prejudice against blacks 5 to 10 percentage points higher than prejudice against Asians. This difference is perhaps the result of more hostile stereotypes about the backwardness of black cultures formed in the colonial era. Black prejudice declines more rapidly between generations, eliminating this differential discrimination completely among the post-WWII cohorts.

As hypothesized, highly educated individuals express significantly less prejudice on all items, with prejudice levels among degree holders between 5 and 15 percentage points lower than among the unqualified. The rapid rise in qualification levels does not, however, account for much of the generational shift in attitudes: the cohort effects are little altered by the introduction of controls for education. There is a significant difference in how racial intermarriage is regarded by degree holders and the uneducated. Those with degrees express greater opposition to white–black intermarriage, with opposition levels 8 percentage points higher than to white–Asian intermarriage, while the uneducated regard both groups similarly on this item. This difference is due to the greater hostility shown to black–white intermarriage by older white degree holders and will be discussed in greater detail in the next section.

Contrary to expectations, there is no evidence of elevated hostility to black and Asian minorities among the working classes or those dependent on state resources. The main effects of class are limited to one group: the salariat consistently express significantly less prejudice than the other social classes. Those in salariat occupations are more tolerant of their relatives taking a black or Asian spouse, and of working for an Asian employer. Professional workers in the salariat have the most secure and high status jobs in the economy, so they are perhaps less worried about competition from ethnic minorities who tend to be concentrated in lower status professions. Professional workers are also more likely to work for large multinational firms that operate in and recruit from many countries. Globally mobile professional workers are more likely to have experience working in different countries, or dealing with foreign colleagues or customers, which may encourage them to develop a more cosmopolitan outlook towards minority groups.

Men are more hostile to an ethnic minority employer, particularly an Asian boss, but no gender difference is found in attitudes towards ethnic minority spouses. However, looking only at the main effects of gender understates the importance of gender divides in racial attitudes, as further analysis reveals differing patterns of generational attitude change for men and women. Cohort effects also depend on education and social class. These interaction effects are discussed in the next section.

Religious identity and church attendance also have significant effects on social distance attitudes. The substantive impact of these variables is relatively small and so they are not shown in Figure III. Identification with the Church of England encourages somewhat elevated levels of hostility to minorities, particularly towards Asians who are predominantly non-Christian. Actually attending church, however, has the opposite effect: those who attend once a month or more have significantly lower social distance on both items. Those who attend church regularly may be internalizing the tolerant norms of Christian doctrine, or may at least feel a greater need to appear more tolerant towards minorities in their public attitudes, perhaps due to greater awareness of the social norm against racial prejudice.

Interaction effects: diverging attitudes among the young

  1. Top of page
  2. Abstract
  3. Introduction
  4. Racism in decline?
  5. Is racial prejudice an attitude or a ‘value’?
  6. The nature of British prejudice: aversion to ‘blackness’ or cultural racism?
  7. Social structural factors: education, class and gender
  8. Data and methods
  9. Results
  10. Multivariate analysis
  11. Interaction effects: diverging attitudes among the young
  12. Discussion
  13. Bibliography

The second set of models shown in Table III adds interaction effects between cohort and social characteristics, enabling the pattern of generational change to vary for different social groups.15 These interactions significantly improve the fit of the models, suggesting important variations in the pattern of generational change between social groups. The main effect of cohort remains significant in three of the four models, suggesting that a general generational decline in prejudice occurs in these cases. Cohort has no significant main effect on attitudes towards an Asian boss suggesting no general pattern of generational change in attitudes on this item. The main period and cohort effects are larger for attitudes towards blacks. The gender interaction term is also larger for attitudes towards blacks, suggesting more divergent patterns of generational change in attitudes towards blacks than in attitudes towards Asians. Larger generational shifts are seen among the salariat and women in all four models. In the ethnic spouse models, significantly larger generational shifts are also found among those with degrees or A-level qualifications.

Table III. Interaction effects
 Dislike social contact with
Minority bossMinority spouse
AsianBlackAsianBlack
  1. Source: British Social Attitudes Surveys 1983–1996

  2. Notes: In model but not shown: economic conflict, religion and lifecycle events variables

  3. Significant figures in bold: * =p < 0.05; ** =p < 0.01; *** =p < 0.001

Cut point 11.211.74−0.17−0.09
Cut point 22.082.810.881.02
Cohort effect−0.007−0.022***−0.020***−0.028***
Cohort squared/1000−0.060.06−0.31***−0.29***
Period effects (reference: 1980s)
1990s−0.28***−0.37***−0.33***−0.42***
Education (reference: no qualifications)
Degree−0.76***−0.48**−0.65***−0.24*
A-level−0.52***−0.11−0.14−0.008
GCSE−0.19*0.06−0.060.10
Class (reference group: working class)
Salariat−0.25*−0.21−0.21**−0.27**
Routine non-manual−0.02−0.12−0.060.0004
Petty bourgeoisie0.190.110.040.05
Manual formen/supervisors−0.050.01−0.080.05
Gender
Male0.34***0.26**0.100.09
Interaction effects
Education
Degree*cohort−0.009−0.01−0.017**−0.017**
A-level*cohort−0.01−0.007−0.017**−0.015*
GCSE*cohort−0.0050.003−0.007−0.005
Gender
Male*cohort0.010**0.020***0.013***0.018***
Class
Salariat*cohort−0.014*−0.019**−0.010*−0.010*
Rnonman*cohort−0.006−0.004−0.007−0.000
Pettybou*cohort−0.001−0.009−0.005−0.017**
Manual foremen*cohort0.0008−0.009−0.010*−0.004
Pseudo R square0.0310.0400.0560.055
N6077586657996041

The meaning of these different patterns of change is illustrated in Table IV, which shows the pattern of generational change in hostility to racial intermarriage for different social groups.16 The generational interactions fall into two different patterns. The first is divergence from similar starting points, seen in the education and class interactions. Levels of prejudice start off similarly for all groups and then begin to diverge for cohorts born in the 1930s and after who came of age after mass black and Asian immigration to Britain had commenced. This pattern of divergence fits with the idea that the educated and privileged are more responsive to the various social pressures encouraging the adopting of tolerant views. The generational shift in these groups began earlier and has progressed further than among the uneducated and the working classes, though evidence from the youngest cohorts suggests that they too are now adopting more tolerant views of minorities.

Table IV. Interaction effects illustrated: predicted total opposition by gender, education level and social class
 1910192019301940195019601970Change
  1. Source: British Social Attitudes 1983–1996

Ethnic minority spouse
Gender
Black spouse women61.959.154.849.041.833.725.3−37.6
Black spouse men50.151.551.550.047.142.837.3−12.8
Difference women–men+11.8+7.6+3.3−1.0−5.3−9.1−12.024.8
Asian spouse women53.753.150.947.242.035.528.3−25.4
Asian spouse men45.247.949.148.746.843.438.6−6.6
Difference+8.5+5.2+1.8−1.5−4.8−8.9−10.318.8
Education
Black spouse degree70.864.756.546.435.224.415.4−65.4
Black spouse unqual63.861.156.951.143.835.426.7−37.1
Difference deg–unqual+7.0+3.6−0.4−4.7−8.6−11.0−11.328.3
Asian spouse degree56.051.244.837.128.620.513.5−42.5
Asian spouse unqual57.957.455.351.746.439.832.1−25.8
Difference−1.9−6.2−10.5−14.6−17.8−19.3−18.616.7
Class
Black spouse salariat65.860.754.045.737.226.617.9−47.9
Black spouse working64.261.457.251.444.235.927.3−36.9
Difference sal–working+1.6−0.7−3.2−5.7−7.0−9.3−9.411.0
Asian spouse salariat58.855.951.345.237.329.421.6−37.2
Asian spouse working56.355.653.549.745.537.930.4−25.9
Difference+1.5+0.3−1.8−4.5−8.2−8.5−8.811.3
Ethnic minority boss
Gender
Black boss women24.420.116.513.611.39.58.0−16.4
Black boss men17.717.016.516.115.915.916.1−1.6
Difference+6.7+3.10−2.5−4.6−6.4−8.114.8
Asian boss women17.016.415.714.914.013.112.1−4.9
Asian boss men25.727.028.129.129.830.430.8+5.1
Difference−8.7−10.6−12.4−14.2−15.8−17.3−18.710.0

The interactions between generation and gender follow a different pattern. Women start with higher levels of prejudice which decline rapidly while male prejudices barely change at all. As a result, a gender gap is seen in attitudes to racial intermarriage among pre-immigration cohorts which is then reversed among the youngest cohorts. Female hostility to an Asian spouse declines by 25 percentage points between the 1910 and 1970 cohort, and opposition to a black spouse falls by 38 points. Male hostility levels fall by 7 and 13 points respectively, and show no change at all between the 1910 and 1950 cohorts. The difference between growing female tolerance and more persistent male hostility is just as pronounced in attitudes to an ethnic minority boss. This pattern suggests that women are more responsive to prevalent social norms about race. Women brought up when prejudice was socially acceptable express more prejudice than men, while the youngest generations have more rapidly adopted a tolerant view of ethnic minorities.

These interaction effects show that while racial prejudice is relatively evenly distributed among older Britons, among the young it has remained prevalent in some social groups and virtually disappeared in others. This social divergence can be most clearly illustrated using combined social categories, as is shown in Figure IV. This compares levels of prejudice among graduate women with degrees with those of men with no qualifications. These are large social groups – women with degrees constitute 5 per cent of the total sample (and 8 per cent of those born after 1960), men with no qualifications 28 per cent. There is a dramatic difference in the generational structure of the two groups’ attitudes. Opposition to racial intermarriage among degree holding women falls from over 60 per cent among the oldest cohorts to under 10 per cent among the youngest, while unqualified male opposition remains flat at around 50 per cent for most generations, only beginning to decline among the very youngest cohorts.

image

Figure IV. Diverging social groups – degree holding women and unqualified men

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Racism in decline?
  5. Is racial prejudice an attitude or a ‘value’?
  6. The nature of British prejudice: aversion to ‘blackness’ or cultural racism?
  7. Social structural factors: education, class and gender
  8. Data and methods
  9. Results
  10. Multivariate analysis
  11. Interaction effects: diverging attitudes among the young
  12. Discussion
  13. Bibliography

The evidence from this analysis suggests that a large and growing segment of Britain's white population is at ease with racial diversity. Opposition to contact with black and Asian minorities in both family and workplace was markedly lower in the mid-1990s than it was in the early 1980s. A major generational shift in white attitudes towards ethnic minorities is the most important driver of this trend. Older generations who grew up in an ethnically homogeneous Britain before mass immigration began are much more hostile about social contact with minority groups than their children and grandchildren. Racial prejudices thus resemble core political and social values (Inglehart 1997; Tilley 2005): they are strongly structured by generation and seem to be primarily influenced by the social conditions prevalent during an individual's youth.

While considerable evidence was found for a decline in discrimination based on colour, little evidence was found to support the contention that race-based discrimination is being replaced by new barriers based on perceived cultural difference. British respondents did not express greater hostility towards Asian minority groups, despite their greater cultural and religious distinctiveness. There is a larger generational shift in attitudes towards black Britons, but this is because they are regarded with greater hostility by older cohorts, not because they enjoy greater acceptance from the young. While the groups asked about here are rather simplified racial categories, they provide some evidence to support the contention from Gilroy and others that racial difference remains more salient than cultural difference as a source of prejudice. Further research is needed to investigate whether this finding is robust when more carefully defined groups, particularly groups defined by religion, are asked about.

How does Britain's growing acceptance of ethnic minorities compare with attitudes in other rapidly diversifying developed nations? Research examining trends in attitudes to immigrant minorities in Europe has been relatively limited, but recent studies have found significantly lower hostility to immigrant minorities among younger cohorts in a range of immigrant receiving West European nations (Pettigrew and Meertens 1995; Gijsberts, Hagendoorn and Scheepers 2004). These findings suggest that the generational shift towards more tolerance of diversity in Britain may be part of a general trend in diversifying European democracies, although research in Holland suggests that deteriorating economic conditions and high immigration rates can halt or reverse such shifts (Coenders and Scheepers 1998). More comparative research is needed to examine the extent of this general trend and how it is affected by conditions within individual nations.

Trends in racial attitudes have been also been extensively researched in America, where scholars have also found evidence for a similar generational shift in attitudes towards ethnic minorities. While America has been racially diverse since its foundation, black Americans only gained full acceptance to the American political community following the Civil Rights reforms of the mid-1960s. As in Britain, white American attitudes towards African Americans are strongly influenced by conditions in a respondent's youth, with sustained declines in many measures of prejudice against blacks among those coming of age after the Civil Rights era (Firebaugh and Davis 1989; Kinder and Sanders 1996). Studies examining social distance from a range of minority groups over a seventy year period have also found sustained falls in the hostility white Americans express towards social contact with minority groups (Kleg and Yamamoto 1998). The relative ranking of groups has remained the same over time, but the difference in attitudes between most and least liked groups has declined as hostility to contact with the least favoured groups has declined most rapidly.

Evidence of persistent discrimination against African Americans despite the apparent survey evidence of a sharp decline in prejudice in post-Civil Rights America has lead some American scholars to question whether the survey evidence is capturing a real shift in attitudes. Proponents of ‘new’ or ‘symbolic’ prejudice theories have contended that the apparent decline in American racial prejudice actually reflects a growing reluctance to express in public racial prejudices which are no longer considered socially acceptable, while private racial prejudices remain widespread and continue to motive discriminatory behaviour and attitudes (Bobo 2001; Sears and Henry 2003). These scholars point to evidence from more subtle survey measures which suggest continuing opposition to providing assistance to African Americans (Gilens 1999), persistent negative stereotypes about blacks (Bobo 2001) and evidence that survey respondents express elevated prejudice when they believe they can do so anonymously (Kuklinski, Cobb and Gilens 1997).

The ‘new prejudice’ arguments have been intensely debated in America (Quillian 2006 provides an excellent summary), but they provide cause for caution in interpreting the declining trend in British prejudice. The social distance items employed here are similar to the ‘blatant prejudice’ measures which new racism scholars argue no longer accurately capture racial attitudes in America. There is therefore a risk that part of the trend seen in this analysis actually reflects a generational shift in the social acceptability of admitting to racial prejudice. However, we can draw some confidence from the fact that rising British acceptance of ethnic minorities is also reflected in the preferences revealed in the housing and marriage markets: Britain's ethnic minorities are becoming less residentially segregated from the white population, (Simpson 2007) and rates of intermarriage between whites and all minority groups have risen rapidly (Coleman 1985; Muttarak 2004).17

While the indicators employed may not be perfectly reliable, the shifts in attitudes examined here thus seem to reflect a genuine increase in the social acceptance of British ethnic minorities. The generational basis to the trend mean it is also likely to have continued beyond the end of the survey series examined here, as younger and more tolerant cohorts replace the ageing and highly prejudiced pre-immigration cohorts.18 While prejudice is therefore likely to be less of a problem in the future, it is unlikely to disappear overnight. Cohort replacement is a slow process, and significant levels of hostility to ethnic minorities remain even in the youngest cohorts surveyed here. The generational decline in racial prejudices is also not evenly spread. Generational shifts in attitudes have been most rapid among highly educated Britons, those in high status occupations and among women. Younger generations of these groups seem to have been completely socialized into a tolerant set of ‘multicultural’ norms, and rarely express any discomfort about social contact with ethnic minorities. By contrast, significant pools of hostility to ethnic minorities remain among less qualified, working-class young men. Researchers and policymakers interested in understanding and combating prejudice should focus on these groups, and particularly on examining how exactly education encourages more tolerant attitudes, and on why British racial prejudice is becoming an increasingly male phenomenon among the young.

Notes
  • 1

    This research was generously supported by an Economic and Social Research Council doctoral studentship award. I would like to thank Steve Fisher, Maria Sobolewska, Anthony Heath, Geoff Evans, Jane Green, Armen Hakhverdian, participants at the American Sociological Association annual conference and the Aage Sorenson memorial conference, and the anonymous reviewers for their helpful and insightful comments on earlier drafts of this work. As usual, responsibility for remaining errors lies with the author alone.

  • 2

    The measure they employed asks respondents to rate their own racial prejudice, without providing any explanation of ‘prejudice’. Respondents interpret this abstract term in very different ways, rendering the measure unreliable.

  • 3

    In America, the demise of the ‘Jim Crow’ regime of segregation and black political disenfranchisement has been accompanied by a sharp decrease in racial prejudice, in both segregated and integrated states (Firebaugh and Davis 1989; Bobo, 2001).

  • 4

    Immigration settlements from the predominantly black and Asian ‘New Commonwealth’ fell 30% from an average of 54,000 in the three years prior to Thatcher's election to an average of 38,000 during the first Thatcher administration. Settlements remained stable thereafter at around 30–35,000 until the late 1990s, when they began to rise again.

  • 5

    The British Social Attitudes surveys are conducted by the National Centre for Social Research. I thank the UK data archive at the University of Essex for making this data available. The data can be accessed at their website: http://www.data-archive.ac.uk. Full technical details of the British Social Attitudes surveys are available at this site, or in the annually published British Social Attitudes reports.

  • 6

    For both items, the sample is randomly split and half of the respondents are asked about ‘blacks or West Indians’ and half asked about ‘Asians’.

  • 7

    For example, respondents from older cohorts who consider themselves ‘unprejudiced’ express much higher rates of opposition to race relations legislation, ethnic minority immigration and assistance to ethnic minorities than ‘unprejudiced’ people in younger cohorts. This pattern suggests a generational shift in the standards of racial tolerance, with younger generations being stricter when they rate their own prejudice.

  • 8

    Technically, the log odds of a particular category occurring, but it is straightforward to derive the probability from this figure.

  • 9

    The ordered logit model makes an important ‘proportional odds’ assumption which prevents the slopes from varying between different categories. The validity of this assumption was tested by comparing the results with a multinomial logit model, which does not assume constant slopes. The model fit and results were very similar.

  • 10

    The chart shows a seven year moving average of attitudes, to smooth out the variations between individual years.

  • 11

    Model fit was not improved by measuring the period effect for each survey year individually. The effect appears to be a shift which occurred between the surveys from the 1980s and the surveys from the 1990s.

  • 12

    There was one change made in 1994. The social distance questions always appear at the end of a battery of questions about racism in British society, immediately after a question about support for anti-discrimination legislation. In 1994, three questions were asked about anti-discrimination legislation rather than one. This may have encouraged respondents to think of discriminatory attitudes towards minorities as socially unacceptable, and reduced admissions of prejudice. However, the responses in the 1994 survey were not significantly different from those in 1996, when only one questions was asked about anti-discrimination law.

  • 13

    Cohort is measured using a centred year of birth variable, so the coefficients in the model represent the effect of changing the year of birth by one year.

  • 14

    Analysis of this nonlinearity using a non-parametric smoothed regression (Fox 2000) suggests attitudes on this item are roughly constant among generations born before 1935, and become continuously more liberal among generations born later.

  • 15

    Interactions between gender and other social characteristics, and period and other social characteristics were also tested but were not found to be significant and were dropped from the final model. Non-linear interactions between the squared cohort term and social characteristics were also tested, and found to be insignificant.

  • 16

    Calculated using the Stata programme Clarify (Tomz, Wittenberg and King 2001). Levels of all other variables are set to population means.

  • 17

    By contrast, rates of black–white intermarriage remain low in America (Qian 1997) and black and white communities remain highly geographically segregated.

  • 18

    The only social distance item asked since 1996 – an item which asked if respondents were troubled by the idea of a relative marrying a Muslim, asked in 2003 – supports this argument. The level of opposition – 27% – was considerably lower than the 1996 figure for black and Asian minorities (35%). This low level of hostility is notable given that we might expect particularly high levels of hostility towards Muslims in a survey conducted less than two years after the September 11th terrorist attacks.

Bibliography

  1. Top of page
  2. Abstract
  3. Introduction
  4. Racism in decline?
  5. Is racial prejudice an attitude or a ‘value’?
  6. The nature of British prejudice: aversion to ‘blackness’ or cultural racism?
  7. Social structural factors: education, class and gender
  8. Data and methods
  9. Results
  10. Multivariate analysis
  11. Interaction effects: diverging attitudes among the young
  12. Discussion
  13. Bibliography
  • Altemeyer, B. 1995 The Authoritarian Specter, Cambridge: Harvard University Press.
  • Anderson, B. 1991 Imagined Communities: Reflections on the Origin and Spread of Nationalism, London: Verso.
  • Anthias, F. and Yuval-Davis, N. 1992 Racialised Boundaries: Race, Nation, Gender, Colour and Class and the Anti-Racist Struggle, London: Routledge
  • Berthoud, R. 2000 Ethnic Employment Penalties in Britain’, Journal of Ethnic and Migration Studies 26(3) 389616.
  • Blackaby, D., Leslie, D., Murphy, P. and O’Leary, N. 2002 White/ethnic Minority Earnings and Employment Differentials in Britain: Evidence From the LFS’, Oxford Economic Papers 54 (2): 27097.
  • Bobo, L. 1983 Whites’ Opposition to Busing: Symbolic Racism or Realistic Group Conflict? Journal of Personality and Social Psychology 45: 11961210.
  • Bobo, L. 2001 Racial Attitudes and Relations at the Close of the 20th Century’ in NSmelser, WWilson and F.Mitchell (eds) ‘America Becoming: Racial Trends and their Consequences’, Washington: National Research Council.
  • Bogardus, E. 1967 A Forty Year Racial Distance Study, Pasadena: University of South California Press.
  • Centre for Contemporary Cultural Studies 1982 The Empire Strikes Back: Race and Racism in 1970s Britain, London: Routledge.
  • Coenders, M. and Scheepers, P. 1998 Support for Ethnic Discrimination in The Netherlands, 1979–1993: Effects of Period, Cohort, and Individual Characteristics’, European Sociological Review 14 (4): 40522.
  • Coleman, D. 1985 Ethnic Intermarriage in Great Britain’, Population Trends, 40: 410.
  • Converse, P. 1964 The Nature of Belief Systems in Mass Publics’ in D.Apter (ed.) Ideology and Discontent, New York: Free Press.
  • Daniel, W. 1968 Racial Discrimination in England, London: Penguin.
  • Esmail, A. and Everington S. 1993 Racial Discrimination Against Doctors from Ethnic Minorities’, British Medical Journal 306: 6912.
  • Evans, G. 2002 In Search of Tolerance’ in A.Park, J.Curtice, K.Thompson, L.Jarvis, and C.Bromley (eds) British Social Attitudes: the 20th Report, London: Sage.
  • Evans, G., Heath, A. and Lalljee, M. 1996 Measuing Left–Right and Libertarian–Authoritarian Values in the British Electorate’, British Journal of Sociology 47(1) 93112.
  • Firebaugh, G. and Davis, K. 1989 Trends in Antiblack Prejudice 1972–1984: Region and Cohort Effects’, American Journal of Sociology 94 (2): 25172.
  • Fox, J. 2000 Multiple and Generalised Nonparametric Regression, Thousand Oaks: Sage.
  • Gijsberts, M., Hagendoorn, L. and Scheepers, P. 2004 Nationalism and the Exclusion of Migrants: Cross-National Comparisons, Aldershot: Ashgate.
  • Gilens, M. 1999 Why Americans Hate Welfare: Race, Media and the Politics of Antipoverty Policy, Chicago: University of Chicago Press.
  • Gilroy, P. 1987 ‘There Ain't No Black in the Union Jack’: The Cultural Politics of Race and Nation, London: Routledge.
  • Gilroy, P. 2002 Introduction’ in There Ain't No Black in the Union Jack’: The Cultural Politics of Race and Nation (Routledge Classics Edition) London: Routledge.
  • Heath, A. and Cheung, S. (eds) 2007 Unequal Chances: Ethnic Minorities in Western Labour Markets, Oxford: Oxford University Press.
  • Hello, E., Scheepers, P., and Sleegers, P. 2006 Why the More Educated are Less Likely to Keep Ethnic Distance: An Empirical Test of Four Hypotheses’, Ethnic and Racial Studies 29(5): 95985.
  • Hoffman, J. 2004 Generalised Linear Models: An Applied Approach’, Harlow: Pearson.
  • Husbands, C 1983 Racial Exclusionism and the City: The Urban Support of the National Front’, London: Allen & Unwin.
  • Inglehart, R. 1997 Modernisation and Postmodernisation: Cultural, Economic and Political Change in 43 Societies, Princeton: Princeton University Press.
  • Johnson, M. and Marini, M.1998 Bridging the Racial Divide in the United States: The Effect of Gender Social Psychology Quarterly 61(3): 24758.
  • Kinder, D. and Sanders, L. 1996 Divided by Color: Racial Politics and Democratic Ideal, Chicago: University of Chicago Press.
  • Kleg, M. and Yamamoto, K. 1998 As the World Turns: Ethno-Racial Distances After 70 Years’, The Social Science Journal 35 (2), 18390.
  • Kuklinski, J., Cobb, M. and Gilens, M. 1997 Racial Attitudes and the “New South” ’, Journal of Politics 59 (2) 32349.
  • Layton-Henry, Z. 1992 The Politics of Immigration, Oxford: Blackwell.
  • Long, J. and Freese, J. 2006 Regression Models for Categorical Dependent Variables Using Stata (2nd Edition), College Station: Stata Press.
  • Lubbers, M., Gijsberts, M. and Scheepers, P. 2002 Extreme Right-wing Voting in Western Europe’, European Journal of Political Research 41: 34578.
  • Mannheim, K. 1952 The Problem of Generations’ in Essays on the Sociology of Knowledge, New York, Oxford University Press.
  • Mason, W. and Fienberg, S. 1985 Cohort Analysis in Social Research: Beyond the Identification Problem, New York: Springer-Verlag.
  • McLaren, L. 2003 Anti-immigrant Prejudice in Europe: Contact, Threat Perception and Preferences for the Exclusion of Migrants’, Social Forces 81(3): 90936.
  • Modood, T. 1990 British Asian Muslims and the Rushdie Affair Political Quarterly 62(2): 4360.
  • Modood, T. 1994 Political Blackness and British Asians’, Sociology 28(4): 85976.
  • Modood, T. 2005 Multicultural Politics: Racism, Ethnicity and Muslims in Britain, Edinburgh: Edinburgh University Press.
  • Modood, T., Berthoud, R., Lakey, J., Nazroo, J., Smith, P., Virdee, S., and Beishon, S. (1997) Ethnic Minorities in Britain: Diversity and Disadvantage, London: Policy Studies Institute.
  • Muttarak, R. 2004 Marital Assimilation: Inter-ethnic Marriage in Britain’, paper presented at the Australian Population Association Conference, 2004.
  • Noon, M. 1993 Racial Discrimination in Speculative Application: Evidence From the UK's Top 100 Firms’, Human Resource Management Journal 3(4): 3547.
  • Norris, P. 2005 Radical Right: Voters and Parties in the Electoral Market, Cambridge: Cambridge University Press.
  • Office for National Statistics 2003 Census 2001 – Ethnicity and Religion in England and Wales’, London: Office for National Statistics.
  • Pettigrew, T. 1997 Generalized Intergroup Contact Effects on Prejudice’, Personality and Social Psychology Bulletin 23: 17387.
  • Pettigrew, T. and Meertens, R. 1995 Subtle and Blatant Prejudice in Western Europe, European Journal of Social Psychology 25: 5775.
  • Qian, Z. 1997 Breaking the Racial Barriers: Variations in Interracial Marriage Between 1980 and 1990’, Demography 34(2): 26376.
  • Quillian, L. 2006 New Approaches to Understanding Racial Prejudice and Discrimination’, Annual Review of Sociology 32: 299328.
  • Rose, E., and associates 1969 Colour and Citizenship: A Report on British Race Relations, Oxford: Oxford University Press.
  • Rothon, C. and Heath, A. 2003 Trends in Racial Prejudice’ in A.Park, J.Curtice, K.Thomson, L.Jarvis and C.Bromley (eds) British Social Attitudes: The 20th Report – Continuity and Change Over Two Decades, London: Sage.
  • Rowatt, W. and Franklin, L. 2004 Christian Orthodoxy, Religious Fundamentalism, and Right-Wing Authoritarianism as Predictors of Implicit Racial Prejudice’, International Journal for the Psychology of Religion 14(2): 12538.
  • Schoen, D. 1977 Enoch Powell and the Powellites, Basingstoke: Macmillan.
  • Sears, D. and Henry, P. 2003 The Origins of Symbolic Racism’, Journal of Personality and Social Psychology 85(2): 25975.
  • Simpson, L. 2007 Ghettos of the Mind: The Empirical Behaviour of Indices of Segregation and Diversity’, Journal of the Royal Statistical Society series A 170(2):40524.
  • Solomos, J. 1989 Race and Racism in Britain, Basingstoke: Palgrave Macmillan.
  • Sullivan, J., Piereson, J. and Marcus, G. 1982 Political Tolerance and American Democracy’, Chicago: University of Chicago Press.
  • Sullivan, J. and Transue, J. 1999 The Psychological Underpinnings of Democracy: A Selective Review of Research on Political Tolerance, Interpersonal Trust and Social Capital’, Annual Review of Psychology 50: 62550.
  • Tilley, J. 2002 Is Youth a Better Predictor of Socio-political Values Than Nationality? Annals of the American Academy of Political and Social Science, 58: 22656.
  • Tilley, J. 2005 Libertarian–authoritarian Value Change in Britain, 1974–2001’, Political Studies 53: 44253.
  • Tilley, J., Exley, S. and Heath, A. 2004 Dimensions of British Identity’ in A.Park, J.Curtice, K.Thompson, C.Bromley, and M.Phillips (eds) British Social Attitudes: the 21st Report, London: Sage.
  • Tomz, M., Wittenberg, J. and King, G. 2001 CLARIFY: Software for Interpreting and Presenting Statistical Results’, Version 2.0 Cambridge, MA: Harvard University, http://gking.harvard.edu
  • Zaller, J. 1992 The Nature and Origins of Mass Opinion, Cambridge: Cambridge University Press.