New research by Eric Kaufmann finds that local diversity does lead to more tolerant white attitudes and this is not the result of ‘white flight’. The results suggest that, as more locales become diverse, there will be more interethnic contact and more positive white attitudes to outgroups. 

Does living in a diverse area make whites more tolerant of immigration and minorities, or does it cause them to feel threatened and insecure? If whites in diverse places are more tolerant is this because the intolerant whites have left? These questions have important implications in an age of migration. Whether western societies veer to the populist right or towards cosmopolitanism hangs in large part on the answers. This discussion also bears on whether ethnic groups can rub along in divided societies such as Northern Ireland and Syria, or in segregated spots like Chicago and Cape Town.

Gareth Harris and I grapple with these issues in our recently published open access article in Comparative Political Studies. In England and Wales, we find that White British people in wards with more minorities and immigrants are more open to immigration. In wards that are almost entirely white, 90 per cent of White British people want immigration to be reduced; but where visible minorities make up half or more of the population, this drops to about 70 per cent. Half the effect arises because whites in diverse wards tend to be young single urban renters, all relatively tolerant segments. The other half of the explanation, however, seems to involve ‘contact’: mixing with minorities or becoming accustomed to their presence.

We looked at the existing literature and found 24 studies looking at attitudes in units with less than 10,000 people, such as British wards or American census tracts. In these studies, three quarters discovered positive contact effects in which whites living with larger shares of minorities and immigrants displayed more positive attitudes toward outgroups than whites in whiter areas.

Yet there is another, less sunny, side to the story. When we looked at larger geographies above 100,000 population, such as the English Local Authority (LA) or American county, the relationship reversed itself. Of 44 papers we sampled, 84 per cent found a threat effect in which whites living in more diverse places expressed more hostile attitudes to minorities and immigrants.

We confirmed this pattern for immigration opinion in England and Wales using the 2009-2011 Citizenship Surveys. Controlling for numerous properties of individuals, such as education, and of areas, such as deprivation score, white British people in diverse wards (population averaging 7,500) are ten points more accepting of immigration than those in the whitest wards. By contrast, whites in diverse Local Authorities (population around 100,000) are around five points more likely than whites in homogenous LAs to call for restriction.

Moreover, whites living in homogeneous wards nested within diverse LAs express the greatest opposition to current levels of immigration. This echoes what Rydgren and Ruth, in relation to Sweden, dub the ‘halo effect’, whereby anti-immigration party support is strongest in white working-class areas adjacent to diversity. The same pattern shows up in support for the far right in Britain. In the 1970s, East London was a hotbed of National Front activity. By the 2008, East London had become diverse and, as figure 1 shows, the red ‘halo’ of far right support shifted further out.

Figure 1: BNP vote share in 2008 Greater London Authority election

Kaufmann june fig 1

Source: Kaufmann, E. and G. Harris, Changing Places: mapping the white British response to ethnic change (London: Demos, 2014), p. 73

One explanation for the above pattern comes from contact theory: whites who live in diverse Local Authorities or counties lack opportunities to mix with minorities but are close enough to be worried. A policy solution would be to encourage more opportunities for positive inter-group contacts, such as educational, religious or sporting exchanges. A useful analogy is with attitudes to nuclear power plants. Those who live far enough away not to worry, or close enough to understand the actual risks, are less fearful than those in the middle.

But contact theory is not the only explanation for the bifurcated pattern we see. ‘White flight’, the self-selection of white British people who dislike diversity out of multicultural wards, could account for white tolerance in diverse areas and relative intolerance in adjacent districts. This question has not been properly addressed because most previous research has run into data limitations. Longitudinal surveys – those which sample the same people every year – are needed to examine movers. In Europe or the US these do not ask subjective questions so it is impossible to track the opinions of movers. Work with the US Panel Study of Income Dynamics, for example, tells us that whites move to whiter areas than minorities. There is, however, no way of discerning whether anti-immigration whites are overrepresented among those moving to whiter areas.

Luckily, the British Household Panel Survey (BHPS), now Understanding Society (UKHLS), allows us to explore the political leanings of thousands of individuals over two decades. The surveys don’t directly ask about immigration, but we know from other sources such as the British Social Attitudes Survey and British Election Study that certain kinds of voters are relatively pro-immigration while others are very much against. For instance, university-educated left-wing voters (Green, Labour, Lib Dem) tend to be pro-immigration. Conservative voters without degrees, or supporters of populist right parties such as UKIP or the BNP, tend to favour lower levels. We augment the longitudinal work with a specially commissioned YouGov survey which asks people about immigration, party support and whether they moved toward or away from diversity during the past decade.

The results are striking. Essentially, while white British people select whiter areas to move to than minorities, whites who favour and oppose immigration move to similar places. In figure 2, we see that, controlling for a wide range of individual and ward characteristics, such as education or deprivation, whites who originate in diverse wards tend to move to much whiter ones. Yet there is barely a paper clip’s difference between whites who vote for anti-immigration parties (mainly BNP and UKIP) and other whites.

Figure 2: Predicted decrease in minorities as the result of a move (white British only)

Kaufmann june fig 2

Source: Understanding Society, waves 1-3 (2009-12)

Among White British respondents in the BHPS, there is a significant difference between university-educated left-wing (i.e. pro-immigration) voters and others, but the difference is small – only two percentage points. And this only concerns white ‘avoidance’, that is, selective inflow. On the other hand, there is no difference between pro-immigration whites and anti-immigration whites in their propensity to leave a diverse area, i.e. ‘white flight’. If the population of a ward is like water in a tub, we could say that in diverse areas, the water is relatively liberal on immigration. The water entering the tub is 2 points more liberal than average while the water exiting reflects the national average. Even if the entire content of the tub changed completely to match the inflow, the most that self-selection can explain is 2 of the 10-point difference in immigration attitudes between the least and most diverse wards in England and Wales. This almost certainly overstates the effect. The ONS Longitudinal Study shows that there is a significant share of long-term white residents in diverse British wards. We know from our YouGov survey that these ‘stayers’ are more likely even than whites who leave to favour lower levels of immigration.

The bottom line: local diversity does lead to more tolerant white attitudes and this is not the result of white flight. As more locales become diverse, this should lead to interethnic contact and more positive white attitudes to outgroups. Yet at the same time, the ‘halo’ of threat will push out to new areas, awakening threat perceptions. Since more whites live in diverse LAs than in diverse wards, it is difficult to know where the balance will lie. This means policies such as refugee dispersal which reduce minority segregation may not produce more positive white attitudes to immigration. Indeed as we note elsewhere, rapid changes often produce the opposite effect, so policies designed to reduce segregation are better justified on other grounds.

Note: This article was originally on the British Politics and Policy blog, London School of Economics.

About the Author

Eric KaufmannEric Kaufmann is Professor of Politics at Birkbeck College, University of London. His latest publication is a Demos report, freely available, entitled Changing Places: the White British response to ethnic change.


As Donald Trump forms one outrageous policy after another, and as the UK government remains unclear as to what future it is pursuing for the country post-Brexit, Eric Kaufmann discusses the factors that led people to back populist rhetorics with editors Chris Gilson and Artemis Photiadou (original post on the LSE blog here)

Recent developments in Western politics – the most recent being the US travel ban – seem to come from an opposite universe, in that we used to see the West as being liberal and secular. Having researched cultural values, are these developments as shocking to you or did you see this coming?

I don’t think any of us were good at predicting developments but I do think there were factors one could have pointed to. What we see is a growing polarisation of values in Western societies. So while the political divide used to be about Left vs Right, about economic redistribution and the free market, the new emerging polarisation is what you may call culturally open vs closed, or cosmopolitan vs nationalist. It’s a cultural war but it’s really over the “who are we” question – who are we as a nation.

Is American populism (pro-Trump) the same as British populism (pro-Brexit)?

I think there are many similarities. In looking through the survey and election data I find a lot of parallels: immigration, to some extent terrorism, and the Syrian refugee issue – there is no better issue to pick up polarisation over Trump than views on Syrian refugees. And we also see with Brexit that immigration was the number one issue driving the vote. These are not the only issues but most are value-based ones.

You also have the impact of the split between those who think the world is a dangerous place and want to be safe, and those who are oriented differently and like novelty and exploration. And so that divide turns very strongly on the death penalty question – those who are pro death penalty are also pro-Brexit and pro-Trump. So we see similar attitudes. But the immigration question is important because it explains the “why now” question – we’ve always had people backing the death penalty or being against it.

So why now? The UK has had waves of immigration since the 1950s and the US has historically been a nation of immigration. And would it be fair to say that continuity sounds like a euphemism for resentment for those who are different to the majority – culturally or perhaps in terms of opportunity?

us-1767682_1920You need to look at each country and the nature of the flows. The proportion of those in European countries – of foreign-born – we haven’t seen that proportion in the past. In the US, the last time we had over 13% foreign-born was in 1900-1920, a period of quite intense, anti-immigration politics. So in a way the more surprising thing would have been if there was nothing happening.

The resentment – I think it’s largely driven by the cultural dimension you mention. I don’t think the resentment of the elite is based on the fact that people have more money or opportunity. Economic resentment is not really driving it. I think the resentment is of a perceived cosmopolitan elite that has brought these cultural changes. So it’s focused more on a liberal cultural and political elite rather than towards someone like Donald Trump, who is very elite in an economic way but not in a cultural one.

Is there a demographic divide in the distribution of personal values?

Definitely. Younger voters, people with university degrees certainly would be more liberal on all these cultural dimensions with a few exceptions. But the important point is that those demographic factors actually only explain a small share of the variation in attitudes. So you have people with degrees who are actually conservative, and people without degrees who are very cosmopolitan.

Education is one of the most important demographics. Not income, not class – education is what splits the data, more so than age. But even education is not as important as values. If you ask a specific question such as support for the death penalty, those will come out stronger than education [in predicting right-wing populist support]. Education is important because it signals a worldview, rather than because it is a marker of income, or class, or status in that way. So education is linked into that cultural worldview divide that I talked about.

How can the UK government reconcile the worldviews of these two groups, which assign opposite definitions to concepts like openness and diversity?

I think the big divide is over immigration and national identity. What the government and centrist parties need to do is to start having different messaging for different parts of the population. So when addressing a white liberal or diverse audience, you can talk about Britain becoming more diverse; but when addressing culturally conservative, mainly white audiences then that’s not a good idea – it tends to stoke fear and resentment. So what you want to do is talk about reassurance – that there is immigration but if we look historically immigrants have tended to assimilate and actually things aren’t going to change very much.

Some of the research I have done also shows that when you give a narrative of assimilation, UKIP voters, hard Brexit voters, and white working-class voters without degrees tend to respond very well. So the hard-core opposition to immigration will decline a significant amount. And part of this is to say that people aren’t all alike and you have people who just do not value diversity.

So I think we need to recognise that you actually need these different messages, because national identity is not unitary. People can identify with a country in many different ways and some people might identify with Britain through their many generations of ancestors in Britain. That’s not a problem, so long as they don’t insist that people who don’t have that aren’t British. There are many ways to be British or American and we need to allow for that.

Will cutting off the flow of immigration counter right-wing populism?

I don’t think cutting off immigration is an option given the many needs of modern societies. Granted we can talk about immigration levels and that’s an important debate and I think there has to be an accommodation of different needs – a happy medium. But I think that more important than that is the “who are we” question. I don’t think it’s enough to talk about where is France headed, where is Britain, where is America going, or what does diversity and immigration mean for France or Britain or America. The real question is [not so much what does it mean to be British but] what does it mean to be white British in an age of large-scale migration. The question is, as a member of the ethnic majority, where do you see yourself and your group moving?

Politicians have not been able to address that and that’s part of where I come up with this idea of having different messaging for different people. You need to get people reassured that we won’t see a radical change, it’s not that society will get more and more diverse and the majority will shrink and shrink and shrink – which is kind of the way people think it is. We need to counter this story of rapid transformation and replace it with what’s fairly likely: modest transformation and things staying the same.

How easy is it to change someone’s beliefs – people are now seeing that their concerns over immigration can be turned into racist policies, like the US travel ban. Would it be enough to make someone change their stance?

Social science research would suggest that it is very difficult to change people’s beliefs. That’s not to say that at the margins some people won’t be turned off by those current policies. But I think what’s likely to happen is actually a deepening of the divide and a deepening of polarisation, partly because we don’t have a centre ground that seems to be more nuanced on this question of racism.

A lot of the people who say the Muslim ban is racist – which it is – also call the wall with Mexico racist – which I don’t think it is. You can be in favour of a wall and not be racist, whereas it is not possible to be in favour of a ban and not be a racist. That’s an important distinction. And if people who support the wall say “well, whatever we support will be called racist,” they may then be desensitized and not be outraged when racist policies like the Muslim ban are put into place. That’s my concern. There should be a centre ground where we can say certain things are racist and outrageous, and other things we may not like but are not racist. Part of the problem is slinging this racism epithet around and that sharpens the divisions; each side starts to get a very one-dimensional view of the other.

Are you dealing with these issues in your forthcoming book?

The new book with Penguin will be all about the white majorities in the West in a time of ethnic transformation – how they are responding to an age of migration and ethnic transformation. And I am arguing that there are a number of responses. You get the populist anti-immigration response, trying to oppose immigration; you get a residential response in the form of white avoidance, with white majorities retreating away from diverse areas and networks; and then you also get an assimilation, an intermarriage, and contact response. And these are not mutually exclusive.

Part of what I will be arguing is that the nature of the white majority will change over time and will increasingly move to be what we would now consider a mixed-race population – most members of the “white majority” will have [an admixture of] non-white non-European background. But that doesn’t mean that they are going to stop thinking like a majority. There will be a lot more continuity than we imagine, there’s not going to be this radical shift and overhaul. But of course, the book remains to be written!


Note: Eric Kaufmann spoke at an event hosted by the LSE Institute of Public Affairs

eric_kaufmannEric Kaufmann (@epkaufm) is Professor of Politics at Birkbeck College and is writing a book about the White majority response to ethnic change in the West (Penguin).