Statistical boundaries and small area data: something worth saving?

By Nissa Finney, CCSR, University of Manchester

Statistical and small area boundaries are invisible on the ground. Yet they shape the physical nature of cities because they demarcate areas that are governed. And they are part of the construction of places because they determine a space that has political representation, or is served by a care trust, or is provided with services by a particular local authority.

Statistical boundaries are ‘territorial units’ within the UK for which data are collected and collated by the national statistical agencies (Office for National Statistics in England and Wales, General Register Office for Scotland and Northern Ireland Statistics and Research Agency). The Office for National Statistics (ONS) provides a useful guide to the geographical boundaries it works with ( There are many types of sub-national boundaries for which small area data are produced – administrative, electoral, census, health, postal. And the boundaries within each of these types change frequently. For example, census boundaries change in an attempt to provide statistics that reflect geographical areas with some social meaning and amendments to electoral boundaries may reflect demographic change. Statistical boundaries both shape and reflect society.

In the UK, statistics are produced for very small areas. For example, census data are published for ‘Output Areas’. Output Areas have a recommended size of 125 households and are generated from data after the completion of each census. Output Areas are designed to have similar population sizes to each other and to be as socially homogenous as possible based on tenure of household and dwelling type. Output Areas are small enough to sit within larger boundaries and always fit exactly within local authority districts.

What kind of data can we get for these small areas? Good examples are provided by the Neighbourhood Statistics website, ( the portal through which ONS disseminates its small area data. By selecting the area you’re interested in, you can view hundreds of data tables on all kinds of topics drawn from census and other data that ONS manages. You can find out about population, education, health, work, deprivation and more for small areas. For example, we can see the area of the University of Manchester (Lower Super Output Area Manchester 018B; Figure 1). If we’re interested, for example, in immigration and diversity we can quickly learn that:

image 1

  • 772 households live in this area
  • of the 2,802 residents over the age of 3 in 2011, 1,766 (63%) have English as their main language
  • 671 (23%) of the 2,893 residents have lived in the UK for less than 2 years
  • the three largest ethnic groups are White British (816; 28%), Chinese (478; 16%), Indian and Pakistani (215 or 7% each)

How might this type of data for small areas be used? Perhaps it is used by providers of health care or education in Manchester to tailor their services for their population. Perhaps it is used by the University to monitor how well it is engaging with the community within which it sits. Perhaps it is used by the local authority in population and economic forecasts. It is certainly used by academics interested in population change. For example, census data for small areas have been used in Dynamics of Diversity: Evidence from the 2011 Census Briefings produced by the ESRC Centre on Dynamics of Ethnicity (CoDE) in association with the Joseph Rowntree Foundation. These analyses of census small area data have revealed increases in ethnic mixing residentially   (Simpson, L (2013); Catney, G. (2013), available at Indeed, such data allow us to identify places that are superdiverse, including Moss Side, the most diverse ward in Manchester district (Figure 2). They also allow us to examine where certain population groups have grown. For example, Figure 3 shows that, between 2001 and 2011, the populations of Pakistani, African and Other White ethnic groups in Manchester and Greater Manchester grew more in areas in which these groups were less concentrated than areas in which these groups were most concentrated in 2001.

Figure 2: Superdiversity in Moss Side, as shown by 2011 Census small area data

Figure 2: Superdiversity in Moss Side, as shown by 2011 Census small area data

Figure 3: Minority populations have grown most in parts of Manchester in which they were least clustered (as shown by Census 2011 small area data)

Figure 3: Minority populations have grown most in parts of Manchester in which they were least clustered (as shown by Census 2011 small area data)

In other words, these ethnic groups have spread out residentially in Manchester over the 2000s. To the contrary, the Chinese population in Manchester district and Greater Manchester grew most over the decade in wards in which it was most concentrated in 2001, perhaps reflecting a growth in the Chinese international student population who settle in the central parts of the city where other Chinese people already reside. These patterns tell us something interesting about how Manchester’s population is changing, and allow us to speculate about and investigate what’s driving these patterns of population change.

How else are small area data being used? Perhaps you have used them. Perhaps you have used them without realising their origins.

Now is an important time to think about how these small area data are used. That is because they are under threat. The Office for National Statistics is currently assessing alternatives to a census for producing population and small area socio-demographic statistics for England and Wales. The review programme is called ‘Beyond 2011’ (—projects/beyond-2011/index.html). The impetus comes from the Treasury (Treasury Select Committee report ‘Counting the Population’, May 2008) and the UK Statistics Authority who would like to see feasible and less costly alternatives to the census that will make the 2011 Census the last of its kind. This call to find a less costly alternative to the decennial census came prior to the 2011 census. The 2011 census has been widely acclaimed as the most successful in recent times; efficiently run, cost-effective and producing a breadth and depth of data that is world-leading. ONS will have a public consultation on its Beyond 2011 proposals between September and November 2013 and will put its recommendations to government in 2014.

The Beyond 2011 proposals may mean that small area data are not produced. It is a real possibility that the future data landscape in the UK will not include the world-leading breadth and quality of small area data that we currently enjoy.

If small area data are to be included in the Beyond 2011 recommendations the case for them needs to be made. There is a danger that small area data will be lost because they’re taken for granted; because they are used by many, but their origins and the efforts to produce them, and their world-leading quality are not necessarily recognised.

It is with this concern in mind that I urge you to consider the appeal by the Beyond 2011 Independent Working Group (Members of the Beyond 2011 Working Group are Piers Elias, Tees Valley Unlimited, and co-chair of Local Authorities’ liaison with central government on population statistics (CLIP); David Martin, Professor of Geography, University of Southampton, Deputy Director ESRC UK Data Service and National Centre for Research Methods; Paul Norman, Lecturer in Human Geography, University of Leeds; Phil Rees, Emeritus Professor of Population Geography, University of Leeds; Ludi Simpson, Professor of Population Studies, University of Manchester, President of the British Society for Population Studies). to provide examples of how you have used Census statistics, particularly for small areas (local authority level and below). These can be sent to ONS at and copied to the Independent Working Group at You may also want to respond to the ONS consultation in the Autumn.

Perhaps it is helpful to think about this is terms of what we won’t have, and what we won’t be able to do, if we don’t have small area data. If small area statistical boundaries and the information about population, health, housing, education, work, migration that they contain were not to exist, what would we not know about cities, and about how cities are changing? How would our understandings of contemporary cities be different without the backdrop of the world-leading quality small area data that we currently enjoy?



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s