Is segregation on the increase in the UK?

Segregation can be an emotive subject. Claims that we are "sleepwalking into segregation" or the refusal to discuss such contentious issues at all because they give the Far Right the "oxygen of publicity," are not helpful. The debate should be a lot more nuanced and, in whatever way segregation is defined and measured, there is more than one discernible trend. Some of these may be seen as positive, while others may raise concerns.

Either way, we need to understand the trends and be able to respond with appropriate policies and practice and it is particularly important to be clear about the past successes and future challenges on the eve of publication of the Casey Review of integration^[1]. This paper deals with the question of residential segregation and it is recognised that this is only one — but perhaps the most significant —  form of divided communities.

Segregation is increasing in a number of very particular respects in the UK, especially the growing isolation of the White majority from minorities in urban zones

The question posed in the title sounds like a fairly easy question to answer, as there are many sets of statistics which can be drawn from the Census and other data. However, it is more difficult than might be supposed as segregation is a ‘slippery’ concept^[2] and can be determined in different ways and measured across larger or smaller geographic areas. This partly explains why different agencies and commentators give apparently contradictory responses. Even amongst the relatively small group of academics working in this area, there are strong disagreements over the conceptual framework, with very different views over what is measured and the assessment techniques deployed^[3].

This paper examines the different views and considers the question in both overall terms and at different geographic levels, and especially as experienced by communities on the ground. This paper comes to a clear view that whilst many areas have become more mixed, segregation is increasing in a number of very particular respects in the UK, especially the growing isolation of the White majority from minorities in urban zones. Further, the extent and pace of this change, within some communities, is very evident.

The overall position

The two current and most authoritative sources of data — the Integration Hub and the Centre on Dynamics of Ethnicity (CoDE) — present a very different picture of segregation, but do nevertheless agree that two trends are evident.

The Integration Hub states^[4]

"One of the basic indicators of ethno-cultural integration is residency: where do different ethnic groups live? Do they cluster in different areas or do we find areas becoming more mixed? The evidence suggests the answer is both".

CoDE agrees^[5]   and begins with the example of the Bangladeshi community which has spread out but with "larger clusters across the country".

And goes on to state:

"The same is true of the Indian, Caribbean, Pakistani, and African groups, and of Muslims, Sikhs, Buddhists, Christians and Hindus.  So, there are bigger clusters and more mixing at the same time."

In another paper^[6], they add: "segregation, the extent to which an ethnic group is evenly spread across neighbourhoods, has decreased within most local authority districts of England and Wales, for all ethnic minority groups... There are very few districts that have seen a large increase in segregation".

How to make sense of the different emphases? CoDE focuses almost entirely on the position of each ethnic group (including White British) in relation to each other, rather than also considering ‘minorities’ as a single group in relation to the majority community.  Most studies of segregation consider both the relationship between the majority group and minorities as a whole, as well as the relationship of individual ethnic minority groups to each other, for example, the extent of Bangladeshi segregation from Pakistanis.

Segregation has been linked to prejudice and intolerance of the ‘other’ due to the lack of contact and interaction across social and cultural boundaries

The majority-minority methodology has been adopted by the Integration Hub and is also used at the international level — see for example Johnston et al(2007)^[7] who compare five countries and provide an overview of established approaches, in which the White majority, (or ‘host’ or ‘Charter’) group is compared to that of minorities as a whole. This is alongside an analysis which examines segregation between individual ethnic groups (including the majority) in relation to each other. An approach which focuses on how the many individual ethnic groups residentially mix with each other tells a very different story to one which asks how minorities as a whole mix with the White British majority. The two accounts diverge because minorities have increased their rate of mixing with each other but not with the White British.

Consequently, the two approaches come to very different conclusions, with the Integration Hub stating:

"The index of dissimilarity, the main measurement of mixing or lack of it, shows that there is more mixing among all ethnic groups but not between minorities considered as a whole and White British.

"So while mixing between all minority ethnic groups is growing, as ethnic minorities disperse out of their historic centres of concentration, it is also the case that minorities, when grouped as a whole, are in most cases becoming more isolated from White British people in urban areas".

This stands in contrast to the CoDE Summary view^[8] that:

"The ethnic minority populations of England and Wales have grown, and live in more mixed areas in 2011 than before. This ‘spreading out’ has accelerated in the past ten years.

"Integration of Britain’s diverse communities accelerated during the 2000s, indicated by residential location, mixed ethnicity, and households with more than one ethnicity".

Mixing and segregation at different levels

It is necessary to consider trends at different geographic levels to provide a finer grained analysis of whether residential segregation is increasing or decreasing.

Firstly, as noted above there is agreement that most local authority districts have become more mixed. This is true in terms of overall diversity and the extent to which groups are more evenly spread (this is what the Index of Dissimilarity measures).

However, for most of those towns and cities which had a disproportionately low number of White British in 2001 (i.e. below the national population average of 86.8% for England) this had become even more disproportionate by 2011 (i.e. further away from the national average White British population which had fallen to 79.8% for England). This is due to outward movement of the White British population over that period in some cases and to an increase in the minority population because of natural factors or inward migration, or both. Table 1 provides examples of some of the more significant changes.

Table 1 — Local Authority districts in England: examples with below average White British population

Long term integration trends may not follow a straight line, but areas with higher minority shares in 2001 tended to become even more non-white in the 2000s. It will certainly be noticeable to communities in those areas, in terms of the population profile, but also through changes in commercial and businesses interests, such as shopping patterns, social and cultural networks and in the services of statutory agencies, such as schools, health services and housing provision. These may present challenges as well as opportunities, depending on how the changes are regarded at a local level and their propensity to adjust. However, the changes will be much more marked in ward and smaller area analysis — see below.

For most of the districts and counties that had a disproportionately high number of White British population in 2001 (i.e. above the English average of 86.8%), this had reduced by 2011 though not in proportion to the reduction of White British in the population as a whole (i.e. 79.8%) and therefore became less proportionately mixed than previously, despite being more diverse. In other words, they have moved further away from the national averages. This is due to in-migration of White British population in some cases and a smaller than average growth from the in-migration from minorities from urban areas and national inward migration. Table 2 provides some examples, across England.

Table 2 — Local Authority districts in England: examples with above average White British population

Many local authority districts are somewhere between these two types and in all cases, the future is simply unknown, given a fast-changing social and economic environment. We can only extrapolate from the trends established over the last 30 years or so and this suggests a very uneven picture. However, the ‘spreading out’ of minorities suggested by CoDE, is clearly not on a proportionate basis and seems set to continue to be focused on particular areas. This is largely because of the pattern identified by the Integration Hub:

"When whites and minorities leave inner city areas with large shares of ethnic minorities, they are generally replaced by other ethnic minorities either through population increases or by immigration".

And confirmed by Kaufmann and Harris^[9]

"For London, between 2001 and 2011, around 620,000 White British people left the city, most of whom moved to whiter areas; whites left London at three times the rate of minorities (about 100,000 of the latter left London). This was all despite London’s population increasing by approximately 1 million".

To put this another way, between 2001 and 2011 the White British population in England reduced as a percentage of the total population from 86.8% to 79.8% — a decrease of 8%. Although there was a decrease in the proportion of the population who were white in most areas, the decrease was much greater in the areas which had a low proportion of white British in 2001 than in areas which had had a high proportion. Thus for example, in Newham, which had had the lowest proportion of white British in 2001 there was a 50% decrease in the percentage of WB between 2001 and 2011; in Barrow in Furness, where 97.9% of the population were WB in 2001, this decreased by less than 1% by 2011. This does indicate support for ‘more mixing and more clustering’, but they are not equivalent trends, the clustering is noticeably more marked.

Below the district level

Analysis can also be made at the ward and Super Output Area (SOA) levels.

At ward level

The Integration Hub sets out a very clear analysis at ward level in relation to majority and minority segregation. The results show that segregation in these terms is increasing:

"Over 4.1 million (41 per cent) non-white ethnic minority people live in wards where less than half of the population is white, up from only 1 million (25 per cent) in 2001.

"In 2001, 119 wards out of 8,800 were majority non-white while in 2011, 429 were majority non-white out of 8,570. White areas have become less white while minority areas have not become less minority".

Kaufmann^[10] has prepared a separate analysis which also shows:

"There are more minorities living in the country's whitest places: a fifth of the minority population lives in 6722 wards in 2011 whereas it took fully 7554 to collect a fifth of minorities in 2001. As well, the average white share in the whitest areas is down to 94% in 2011 from 98% in 2001. So minorities are less rare in rural and provincial England. In fact there are fewer than 800 wards that remain over 98 percent white compared to more than 5000 wards in 2001".

And:

"Though white areas have become less white, minority areas have not become less minority. In fact, 4.1 million minorities (41% of the minority population) live in wards that are less than 50% white, i.e. more diverse than Yardley in Birmingham. This compares with about a million minorities (25% of the minority population) living in white minority wards in 2001. In 2001, just 119 wards were majority non-white. In 2011, 429 were. In 2001, a fifth of minorities lived in the most diverse quintile, where 33 percent of the population was white. Today, a fifth of minorities live in the most diverse quintile, which is just 21 percent white".

It is difficult to predict future trends, but given that the process of polarisation of wards has been evident since at least 1991 and has been increasing, it is likely that this will continue as the minority population continues to grow and the majority continues to re-locate.

And once the trend towards increased polarisation within wards has begun it is often dramatic. The following Table shows some fairly typical changes across the country between 1991 and 2011.

(Note: Using the most recent ward boundaries, the minority share living in majority non-white areas rose from 22 to 34% between 2001 and 2011 and the share of non-White British/Irish living in majority non-White British/Irish areas rose in the same period from 28% to 44%.)

Table 3 — Ward Level Polarisation: Typical changes 1991 – 2011

There is no accepted definition of a ‘ghetto’, a term that has largely been used in the American context, but there is some agreement that it is appropriate when an ethnic group exceeds 90% of the population in a given area, though Poulsen^[11] has suggested an 80% threshold where one minority group is dominant within the minority population.

In many of our towns and cities, the minority is increasingly isolated from the majority population and this trend appears to be continuing

In the UK, for the most part, an increasingly isolated minority population is seldom dominated by one ethnic group, especially in the major cities like London and Birmingham (though it is the case in some areas - see below). However, it can be clearly said that within a growing urban zone in many of our towns and cities, the minority is increasingly isolated from the majority population and this trend appears to be continuing.

At output area level

The Census contains finer grained analysis, with an output area generally containing around 300 people. These are too small to draw meaningful conclusions. Super Output Areas (SOAs) are an aggregation of adjacent Output Areas with similar social characteristics and Lower Layer Super Output Areas (LSOAs) typically contain 4 to 6 OAs with a population of around 1500. Middle Layer Super Output Areas (MSOAs) on average have a population of 7,200. Assessment of segregation is clearly possible at the Middle level and will be more or less intense when based upon the divisions into smaller LSOAs, tending more towards ghettoisation in some cases.

As an example of this, Table 4 sets out the LSOAs within one Bradford MLSOA and shows some higher levels of polarisation within an area which is already one of the highest in the country.

Table 4 — Bradford 042 MSOA and LSOAs:  Total Population and Ethnic Groups 2011 % 

Source: ONS Neighbourhood Statistics[12] /OD