Last year, the Guardian triggered a major debate over prejudice after a front page splash painted a dark picture of rising racial intolerance in Britain, writes Dr Robert Ford. But he argues that this debate was focussed on a poorly constructed measure – and that more robust measures paint a very different picture.
There is no doubt that prejudice remains a serious problem in Britain today, with racial discrimination and xenophobia still found in many social contexts. This injustice – and the deprivation and disadvantage it causes for Britain’s growing ethnic minority communities – is a serious social issue. But how widespread is prejudice against ethnic minorities among the majority white population, and how is this changing over time? This is a crucial question both for policymakers and citizens more broadly – is Britain becoming a more open and tolerant society or a more inward looking and divided one? It is, however, a difficult question to answer, as prejudice is a slippery concept and hard to measure effectively.
The problems of measuring prejudice, and the impact a poorly chosen and poorly analysed measure can have on the debate, were illustrated by a front page splash last year in the Guardian newspaper. Drawing on data from the British Social Attitudes survey in which respondents were asked to rate their own prejudice against ethnic minorities, the article argued that Britain faced a “rising tide of racial prejudice“.
The most obvious problem with this story was that the data showed no such thing. Far from “a rising tide”, a visual inspection of responses to the question revealed that reported levels of prejudice were broadly unchanged over a decade, and considerably lower now than twenty or thirty years ago. The claim of a “rising tide” was based on a selective analysis of change between only two time points, and ignored the broader trend. This underscores the crucial importance of proper analysis and reporting of social trends on a key issue – the emotive claim of a “rising tide” of intolerance, which was widely repeated by those reading the story, was not supported at all by the data.
There is also a deeper problem with the Guardian analysis: the measure of prejudice they chose to use is poorly suited to analysing how people’s attitudes change over time. To understand why, we need to look closely at the wording of the question, and how it matches up to other possible measures.
The Guardian’s chosen measure was this:
Although this looks like a straightforward question, it has a fundamental flaw. It does not provide respondents with any definition of “prejudice”, which is a complex and abstract concept. One person may consider himself prejudiced because he once laughed at a racist joke, while another may consider him or herself unprejudiced, despite holding very negative views about some minority groups, because in their understanding prejudice requires something stronger, such as support for violence against minorities.
The biggest problem this self-diagnosis approach causes is that it can mislead us about changes in attitudes over time, because the standards people apply to diagnose themselves as prejudiced are socially relative, and these standards change. In Britain, the evidence suggests two things: firstly, that younger generations are less prejudiced against minorities than older ones, resulting in a gradual decline in prejudice over time. Secondly, that the standards people apply to diagnose their own prejudice have become steadily stricter as Britain overall has become less prejudiced. In short, young people who call themselves “prejudiced” hold themselves to a higher standard than old people making the same judgement, and the standard applied on average today is stricter than it was in earlier decades.
We can show this process at work by examining how self-diagnosis of prejudice matches up with more concrete racial attitudes. If the self-diagnosis of “prejudice” is made in a consistent way, then the attitudes of a prejudiced thirty-something should be the same as the attitudes of a prejudiced pensioner. But this is not the case.
Figure 1 (above) charts the level of opposition to intermarriage with Muslims and West Indians by respondents’ self-rated prejudice levels. Crucially, we also split this out by generation, comparing those born before 1960 – who mostly grew up before mass migration to Britain began – with those born later, in a more diverse society.
Nearly half of those born before 1960 who rated themselves as “unprejudiced” do so despite opposing intermarriage between whites and Muslims. By contrast, only one in five of those born after 1960 who rate themselves “unprejudiced” express such opposition. Younger Britons are employing a stricter standard: a large majority of them accept that they cannot call themselves unprejudiced if they oppose ethnic intermarriage. By contrast, a majority of their parents and grandparents believe they can declare themselves “unprejudiced” despite opposing inter-marriage between whites and Muslims.
The same happens in reverse with those who diagnose themselves as “prejudiced”. Only 15% of those born before 1960 who rate themselves as prejudiced accept white-Muslim intermarriage, but among those born later the figure is 35%. A much larger portion of young people declare themselves to be “prejudiced” despite expressing no opposition to inter-marriage.
This effect is even more pronounced when looking at intermarriage with West Indians – a majority of the older people who call themselves prejudiced oppose such intermarriage, but a large majority (almost 70%) of young who self-diagnose as prejudiced do not oppose it. These young people do not consider accepting white-black intermarriage sufficient to declare themselves prejudice-free, while their parents and grandparents hold themselves to much less exacting standards.
We don’t know from this evidence what criterion younger respondents are using to diagnose themselves as prejudiced when responding to this question – perhaps laughing at a racist joke, feeling uncomfortable in very diverse social situations, or feeling anxious when a young Muslim man joins them on a bus or train. Indeed, one of the key weaknesses of this question is precisely that we have no idea what respondents have in mind when answering it.
Whatever this criterion used is, it is consistently different, and much stricter, among the young. We find similar differences in the standards applied by other groups: university graduates apply a tougher test than school leavers, middle class professionals are harder on themselves than manual labourers and so on.
Perhaps this shouldn’t come as such a surprise. As British people have become steadily more accepting of ethnic minorities – something I charted in an earlier blog -the level of hostility towards such groups that is considered socially acceptable has also declined, and social norms sanctioning expressions of intolerance against minority groups have strengthened. The kind of crude racial stereotyping which regularly featured in stand-up comedy routines a generation or two ago would horrify the young people of today, but was considered perfectly acceptable at the time.
Prejudice against minorities runs on a continuum from mild discomfort to violent hostility. When we ask people “are you prejudiced?”, they have to decide for themselves where the dividing line lies between those who are prejudiced and those who are not. It is natural that people will resolve this question using the most readily available information, namely the standards applied by their peers – people of a similar age and social background.
Of course, people could just lie – they know that being prejudiced is frowned upon, so they could deny holding prejudices. Yet the survey evidence suggests that many are trying to be honest: a lot of people do diagnose themselves as prejudiced, and those who do are consistently more hostile on other racial attitudes measures than otherwise similar people who say they are not. But the standards applied are not fixed across social groups or across time. Therefore, they are not very useful as a measure of the social distribution of prejudice in Britain, or of how it changes over time.
The results we get if we use self-diagnosed prejudice as a measure of long term social change are therefore deeply misleading. When we look at more concrete measures – on issues as varied as workplace relations, inter-marriage and national identity – we find consistent evidence of a rising tide of tolerance, driven in particular by generational shifts, with younger Britons more open to, and respectful of, ethnic diversity.
Ironically, it is this very trend that has led the Guardian and others astray, as the spread of more inclusive attitudes and stricter social norms sanctioning intolerance are likely to be a key reason that Britain’s young people are tougher on themselves than their parents. The apparent persistence of prejudice highlighted by the Guardian actually reflects two trends working in opposite directions: the growing acceptance of minorities in British society has been accompanied by steadily rising standards of respectful treatment. So people still diagnose themselves as prejudiced, but they are now harder on themselves, so the diagnosis means something very different now than 30 years ago. Not such an eye-catching headline, but a fairer reflection of the situation in Britain today.
- The views presented here are those of the author, and do not necessarily reflect those of other members of the Centre on Dynamics of Ethnicity (CoDE).