December 14, 2018   6 mins

Raheem Sterling is a remarkable young man. I resent him somewhat for leaving Liverpool for Man City, and I always find the way he runs rather odd – his arms out to the sides and bent at the elbow, like he’s thought about doing I’m A Little Teapot but can’t quite commit to it. But he is an extraordinary footballer, and an intelligent guy.

The other day he was abused from the stands by a Chelsea fan during Man City’s defeat at Stamford Bridge. It appears to have been racial abuse, although one of the fans in question denies that.

It was a sordid affair. Sterling, though, rose above it with aplomb, and later, on Instagram, issued a thoughtful critique of how the media treats black and white footballers. Understandably, it’s led to some soul-searching in the media (though not enough, in some quarters). And it’s raised the question, once again, of whether Britain is getting more racist.

I wanted to look at that question in a bit more depth and ask how we measure a society’s levels of racial prejudice, and what you find when you do measure it. Are we getting more racist? Since the 2016 referendum, a lot of people have wondered whether racists have become more emboldened: is that true?

The question is, it turns out, not easy to answer. It’s really hard to measure racism, or any form of socially unacceptable prejudice, because people don’t tend to admit to it. For a while people thought that the Implicit Associations Test, which (in one version) measures how quickly you associate ‘good’ and ‘bad’ words with black and white faces, was a sort of window into our souls; it led to all sorts of claims (including by Hillary Clinton) about how unconscious bias is everywhere. But it turned out not to really work. People who score highly on the IAT do not seem to behave in more racist ways than people who don’t. It’s not much use as a predictor of behaviour.

There are other ways, though. The most obvious is, simply, to ask them. But where it gets really interesting is what you ask them.

If you ask people: “Are you prejudiced against people of other races?”, or something along those lines, you get a response of about 25 to 30% saying: yes, yes I am, thanks for asking. That makes headlines most years because the National Centre for Social Research, NatCen, asks a similar question in its annual survey. And every year, the response is about the same. It has never dropped below a quarter in the 30 years they’ve been doing it. So, certainly it looks like Britain isn’t getting any less racist, right?

But, it’s a bit more complicated than that. “Are you prejudiced” is a very abstract question; there’s no objective standard to compare yourself to. If you ask something more concrete, like “would you mind if your child married someone of a different ethnic background?”, you get a picture of dramatic decline. In 1983, according to NatCen’s British Social Attitudes survey, around 50 to 60% of Britons said that they would mind; by 2013, that had dropped to a little above 20%.

That’s not to say that things are perfect – one in five people is still quite a lot – but it’s definitely an improvement. There hasn’t, as far as I can tell, been a significant survey since 2013 asking that question, but Rob Ford, a political scientist at Manchester who studies these things, told me that (for reasons we’ll come back to) it’s reasonable to assume that the trend has continued.

Also, if you break that data down by age, you find that people born before 1940 are far more likely to be against interracial marriage than people born in the 80s and 90s. You can find similar stories when you ask people if they’re comfortable with a black or Asian boss, or (in the US) if they’d be happy voting for a black president.

The problem with all of this is that there is a thing called the “social desirability bias”. That is, people tend to answer surveys – even anonymous ones – in ways that would make them look good. Being racist doesn’t make you look good, on the whole, so people are likely to understate the extent to which they’re uncomfortable with, say, interracial marriage. The downward trend may be real, but it may underestimate the true level.

There are other ways to look at prejudice. One is experimental: you treat the issue like any scientific question, and set up a randomised control trial. You test some hypothesis and change a single variable.

For instance, you could send out loads of job application forms, but change only whether the names on the forms are stereotypically black or white; or you get black and white flat-hunters to go and visit potential landlords, and see whether white ones are more likely to be accepted.

When you do that, you tend to find that there is prejudice: black people are less likely to get callbacks for jobs, for instance. But these experiments are time-consuming and expensive – especially until recently; online applications and things have made them somewhat easier now – so there haven’t been all that many of them, and it’s hard to use them to establish a trend. According to Ford, the evidence “cautiously” points to a decline, but that it’s hard to be sure.

Another way is to examine how we actually behave. For instance, regardless of whether you or not you say that you’re OK with mixed-ethnicity marriage, if you actually marry someone of another race, that’s a pretty robust indicator that you’re broadly OK with people of that race. If you beat someone up in a racially motivated attack, that’s pretty good evidence that you’re not.

If you look at mixed-ethnicity marriages, they’ve been going up for years. Nearly one couple in 10 was of mixed ethnicity in the 2011 census, up from about one in 14 in 2001. And you can see it in the ethnic makeup of the population. About 1.2% of the population self-identified as mixed heritage in 2001; that had almost doubled, to 2.2%, by 2011. Research for the BBC suggests that that is a massive undercount and that the figure might be double that again.

With things like racially motivated assault, or racial abuse, it’s not so clear whether it’s gone up or down. There has been a clear increase in the statistics – a near doubling of reported assaults in England and Wales since 2012/13 – but it’s very hard to be sure whether that’s real or not. As racial issues become more salient, the police get better at recording them and people become more likely to ascribe racial motives to an attack; and as members of ethnic minority communities come to trust the police more, they may become more willing to report. The police themselves say that they think there was a real rise, not simply a statistical artefact, around the 2016 referendum, though.

Where does this leave us? In a complex situation. There obviously are a lot of racist people out there. And a lot of the statistics are hard to make sense of, because they’re messy. But where the stats are able to show a trend, they show a declining one: Britain is less racist now than it was 10 years ago, and that was less racist than 10 years before that.

What’s more, there is a good theoretical basis to believe that we’re less racist. One of the more robust findings in social psychology is that exposure to outgroups promotes more positive views of those outgroups. Each generation of white Britons since the Second World War has had more exposure to ethnic minorities than the one before, at the shops, at school, in work.

And there’s more indirect contact, as well. Our football teams and newsreaders and actors and lawyers and doctors are more diverse. Given all this, you would expect to find that Britain is getting less racist, and in fact when you ask people whether race is important to nationality, you get “a really steep generational gradient”, according to Ford. Young people are much less likely to think that “there ain’t no black in the Union Jack”.

An important caveat: this is all about the average person. It doesn’t rule out the possibility that, while Britain as a whole is getting more inclusive, a minority has become more vocal and aggressive, emboldened by internet subcultures, Brexit, Trump, Breitbart or whatever. And that could mean that the experience of people of colour is not getting better: it only takes a small number of loud people to make a lot of noise.

And if people are yelling abuse at you in the street, or following you home at night, then it’s probably not a lot of comfort to you to learn that on average, people are more in favour of mixed-ethnicity relationships than they were 10 years ago.
Still, it’s probably fair to say that our society has improved on this issue. And to return to football for a moment, you can see this, even though the Sterling case has brought it back to our minds.

In Nick Hornby’s memoir of football fandom, Fever Pitch, published in 1992 and leading up to Arsenal’s dramatic league win in the 1988/89 season, he writes about how racism was common on the stands then. Less so, for Arsenal, as a London team with a large multiethnic following, but lots from away fans, and still sometimes from his fellow Gooners. An opposing black player misses a chance or something, he wrote, and “some neanderthal rises to his feet, points at Ince, or Wallace, or Barnes, or Walker, and you hold your breath … and he calls him a cunt, or a wanker, or something else obscene, and you are filled with an absurd sense of metropolitan sophisticate pride, because the adjectival epithet is missing … It’s not much to be grateful for, really, the fact that a man calls another man a cunt but not a black cunt.”

That wasn’t all that long ago. Compare that to now: the men who allegedly racially abused Sterling will never see the inside of a Premier League stadium again. It’s far from perfect. But it’s definitely an improvement.


Tom Chivers is a science writer. His second book, How to Read Numbers, is out now.

TomChivers