Understanding ‘white flight’

In much of the country different ethnic groups are becoming better integrated, but in areas of minority concentration the white British population has dropped sharply in recent years. Why? In this Impact Essay, Eric Kaufmann reports back on his work for the Demos Mapping Integration project.

When census results were released in December 2012, a headline story was the dramatic change in the ethnic makeup of Britain. Nationally, the number of people from an ethnic minority background doubled in the ten years since 2001. In London, the white British dropped from 58 to 45 percent of the total. In absolute terms, 620,000 fewer white British people lived in the capital in 2011 than in 2001 despite a population boom which added a million people to London’s population. London’s dynamics echo those of diverse places up and down the country where minority growth and white British decline often march in step. In some cases minorities and whites jointly left depressed areas and flocked to growing regions, but often the two flows moved in different directions, with white British outmovers making way for incoming minorities.

Tragically, 2011 may be the last census, so this could be our final glimpse of how British society is changing at local level. The story this granular data provides is dramatic: 38 Local Authorities simultaneously made the top 50 list for highest minority growth and white British loss over the past decade. Redbridge, for example, is third on both lists, and Barking and Dagenham, which topped the charts for white British loss, came fourth for minority growth. A fine-grained analysis shows that across England and Wales, wards with larger ethnic minority shares in 2001 tended to lose white British population in both absolute and relative terms over 2001-11. Is this, as the headlines suggest, the ‘retreat’ of white Britain in the face of immigrant diversity, an attempt by white Britons to maintain ethnic boundaries through exit?


38 local authorities among top 50 for both white British population loss and minority increase, 2001–2011

Ours is a study of the ethnic majority in Britain. Studies of ethnic majorities are less common than those of minorities, whom white scholars and writers tend to find more interesting than themselves. There’s also an important class angle. Some academics suggest the white working-class leave diverse areas because their social networks are tied to kin and locality. When the locale begins to feel unfamiliar, their whole world is turned upside down. The upper middle class, by contrast, have less at stake in their neighbourhood: their identity lies in the placeless connections of profession or lifestyle. A different view is that the white working-class lack the resources to leave poor diverse areas. Trapped, they voice their discontent through support for parties of the extreme right such as the BNP. We find evidence for the first, but less so for the second, proposition.

Our project, a joint venture between Birkbeck College and Demos funded by the Economic and Social Research Council examines the white British – especially white working-class – response to ethnic change. Entitled ‘Exit, voice or accommodation’, it consists of three related strands. First, ‘exit’, the question of ‘white flight’ and residential segregation; second, ‘voice’: how white disquiet at the decline of dominant ethnicity is expressed in anti-immigration attitudes and far-right voting; and third, ‘accommodation’, the extent to which white British people come to accept shifts in the nation’s ethnic composition.

Broadly speaking, we ask how local ethnic change shapes national attitudes and voting, and vice-versa. The findings sometimes surprise. Work with the large-scale Citizenship Surveys shows that over 80 per cent of white British adults want lower levels of immigration. However, those living in wards with more ethnic minorities are relatively tolerant of current or higher levels: about 70 per cent of white British in diverse wards want less immigration but this approaches 85 per cent for those in lily-white wards.

On the face of it, this might appear to be caused by ‘white flight’, as those who dislike immigrants and minorities vote with their feet. Yet our research with the British Household Panel and Understanding Society surveys shows this not to be the case: whites who leave diverse wards for homogeneous areas are only a few percentage points more conservative than those who move the other way. They are also more liberal than whites who remain in diverse places. Age, education, income or class differences between individuals and areas don’t explain the pattern.

Surprised, we commissioned a YouGov survey in August 2013, asking people whether they moved wards in the past decade, and if so, from a more or less diverse one. Are those who are uncomfortable with a non-white Prime Minister, or want immigration reduced a lot, the same people who flee diversity? Not really. Curiously, even if we restrict our gaze to university-educated middle-class Londoners in their thirties, liberal and conservative white Britons move to ethnically similar wards. Though cultural conservatives say they would be uncomfortable in diverse areas while liberals claim the opposite, when the moving van arrives and push comes to shove, these attitudes don’t cut very deep.

Contact with minorities certainly helps explain white tolerance. But it’s important to qualify the story. First, diverse metro areas or local authorities heighten opposition to immigration in the whiter sections of town. Second, white British folk in diverse neighbourhoods are more tolerant in part because they are more transient. Diverse areas are urban places with lots of uprooted singles and renters. This churn imparts a certain rootless, liberalising effect on people – even if they are long-resident homeowners. The downside is that rootlessness may make communities less cohesive – movers are less likely to vote, for instance. But it takes the edge off ethnic tension and produces a more tolerant national outlook. In effect we encounter a small-scale version of David Goodhart’s diversity-solidarity dilemma.

The other cautionary note concerns segregation. On the one hand, ethnic segregation across the UK is slightly lower in 2011 than 2001. The number of mixed-race individuals and households nearly doubled over the past decade. In much of the country, whites and minorities grew less isolated from each other 1. On the other hand, in areas of minority concentration, minority ethnic groups became more isolated from white British people even as they increased their exposure to each other. This is because the white British population in diverse wards dropped considerably between 2001 and 2011.

The same trend can be seen in the United States during 1990-2010. Asian-white segregation, for example, declined in America over this time frame. Yet white withdrawal from Asian areas and steady Asian immigration meant groups like the Chinese or Indians had much less exposure to whites and more exposure to each other in metropolitan areas. This can’t be explained by social mobility, in which wealthy whites leave while poor Asians remain. Asian average household income in these areas was higher than white average income 2.

White British people and ethnic minorities are very unevenly distributed across the surface of England and Wales. Convergence should have resulted in more minorities in white areas and more whites in minority areas in 2011 as compared to 2001. We saw the first effect but not the second – in other words, with the exception of a few gentrifying wards, mainly in Inner London, whites moved toward whiter areas and away from diverse ones. Why are whites leaving diverse areas?

Consider London for a moment. Its white British population declined by 620,000 in ten years, a loss of 14.4 per cent. Though the census did not ask about white British ethnicity prior to 2001, we can reconstruct this from the invaluable – and endangered – ONS Longitudinal Study 3. Our preliminary analysis of a test version of the LS database incorporating 2011 Census data shows the white British decrease is largely due to a net outflow to the rest of the UK. Is this white flight? Census and survey data suggest otherwise, despite anecdotal evidence and sharp ethnic disparities in migration flows. Not since the 1860s has there been positive domestic migration to London, and the share of white British leaving London is about the same today as in 1971 or 1871. Escaping the city rather than white flight seems to be the culprit.

This does not make ethnicity irrelevant. Ethnic minorities haven’t fully caught the white British fever for the countryside. Between 1971 and 2011, London minorities lost only about a quarter as many as the white British to the rest of the UK while gaining immensely from international migration. Minority reluctance to flee seems to be linked to discomfort at being ethnic pioneers. Data shows that groups like Pakistanis and Afro-Caribbeans are leaving their ethnic concentrations, but, compared to whites moving from the same areas, are more reluctant to enter overwhelmingly white areas. Since new pockets of relative diversity such as Woking or Watford have opened up outside London, the share of minorities leaving the city has taken off, doubling in each of the past two decades.

White British, on the other hand, tread migration paths to rural and coastal idylls first romanticised in the nineteenth century. They may move in chains of family and friends, as with the Cockney migration from East London to Essex following slum clearances in London’s old East End. The challenge for integration is not ethnic self-segregation but the fact minorities are notably less keen than whites on moving to places which are over 90 per cent white – i.e. the majority of England and Wales. Meanwhile the continuing depopulation of white British people from dense, deprived, and therefore diverse, urban Britain leaves behind an ever more diverse population. This is not because they dislike minorities – whites moving out are more tolerant than stayers. However, as friends and family leave and few enter, favoured amenities disappear and a spiral may begin which is difficult to arrest.

Gentrification can reverse the process in some cases. Yet American trends suggest inner suburbs – zones of transition between cities and leafy outer suburbs – have difficulty attracting whites. Such areas may evolve into mixed-minority zones with little or no white presence 4. Areas with a low white share tend to lose white British people fastest, and white British are least likely to be comfortable living as a minority or having their children in ‘majority minority’ schools. Yet white British people living in diverse areas are the most tolerant segment of white Britain. For many reasons, policy makers should try and encourage them to stay. Positive discrimination in favour of whites to maintain integrated neighbourhoods, as in the Starrett City project in New York in the 70s and 80s, is illegal as well as inadvisable. However, some sense of the risk of white depopulation is important: housing, benefit and schooling policies in inner suburban areas such as Croydon or Barking should avoid measures that accelerate the departure of white British residents.

Our research also counsels against engineering rapid demographic change. Housebuilding should, wherever possible, be diffuse and small-scale in order to limit the impact of change on particular communities – especially those without a history of high population turnover. Dispersion of refugees and immigrants from London to heavily white provincial towns is an example of a policy which may hinder more than help the cause of integration. This is because the newcomers introduce rapid change into places without prior experience of ethnic shifts. Survey results suggest this increases hostility to immigration and fuels far-right support. Better to allow minorities to diffuse of their own accord, or if dispersion is to take place, to direct newcomers to more transient communities with previous exposure to ethnic diversity.