Postcensal small area population estimates

Update 1/10/2009

New interim mid-2008 Lower Layer Super Output Area and Middle Layer Super Output Area population estimates were released as ONS experimental statistics on 1 October 2009.

Background

The Office for National Statistics (ONS) set up a research project in 2000 to investigate the production of postcensal population estimates for geographical areas smaller than local authorities in England and Wales.

This project was set up in response to the increasing demand for small area statistics identified by initiatives such as New Deal for Communities, Best Value, the National Strategy for Neighbourhood Renewal (encompassing Neighbourhood Statistics) and a review of customer needs for population statistics carried out by ONS.

The initial aim of the project was to investigate the feasibility of producing an authoritative set of small area population estimates that would be available on a nationally consistent basis, and to make a recommendation as to the best method(s) and data sources to use.

The project has considered:

• which data sources and methodologies might be appropriate
• what level of detail could be produced (in terms of age groups, gender and geography)
• how quality of estimates might be assessed
• what frequency of production might be possible

Following on from this work, ONS has published annual Lower Layer Super Output Area (LSOA) and Middle Layer Super Output Area (MSOA) estimates as experimental statistics. These estimates have been used to produce estimates for other geographies such as wards, Parliamentary Constituencies and National Parks.

Further information about Super Output Areas

Project Progress

Research has looked at the provision of small area population estimates within the UK and abroad, and the methods and datasets used. Following on from this work, the project team identified several datasets and methods that might be used to produce small area population estimates. Evaluation of the identified methods against a set of criteria was used to shortlist these methods. Following this evaluation three methods were shortlisted - Apportionment, Cohort Component and Ratio Change.

The Ratio Change method was identified as the preferred method, though on the available evidence from the evaluation undertaken, no single method stood out. All subsequent published small area population estimates are produced using this Ratio Change methodology.

Articles have been included in a number of publications, including Horizons (ONS customer newsletter), the Local Authorities Research and Intelligence Association (LARIA) newsletter, British Urban and Regional Information Systems Association (BURISA) and Population Trends.

Release of Experimental Statistics

The following population estimates have been released as experimental statistics on the National Statistics website.

• Mid-2001 to mid-2008 LSOA estimates by broad age group and sex
• Mid-2001 to mid-2008 MSOA estimates by quinary age group and sex
• Mid-2001 to mid-2007 ward estimates by quinary age and sex
• Mid-2001 to mid-2007 National Park estimates by quinary age and sex
• Mid-2001 to mid-2007 Parliamentary Constituency estimates by quinary age and sex

Further information about Experimental Statistics

Further information about ward geography

By definition, experimental statistics are statistics undergoing evaluation. They are published to help involve users and other interested parties in their development, as a means to build in quality at the earliest stage.

User feedback on these estimates is very much welcome as this may enable us to make improvements in the future.

ONS is willing to undertake work to produce bespoke population estimates where these areas can be readily defined, either in terms of existing geographies, a postcode lookup file or customer supplied digitised boundary file. Potential customers are reminded that there are some data quality limitations with any bespoke population estimates produced and that any work undertaken may be chargeable.

Population estimates for some existing geographies eg parishes, urban areas and Travel-to-Work Areas are available on request.

LSOA and MSOA estimates

Ward Estimates

Parliamentary Constituency estimates

National Park estimates

Methodology note - LSOA and MSOA estimates (pdf, 128Kb)

Methodology note - Ward, Parliamentary Constituency and National Park estimates (pdf, 943Kb)

Publication Plans

Revisions are planned to the local authority population estimates for mid-2002 to mid-2008 in May 2010 to reflect methodological improvements, see http://www.statistics.gov.uk/about/data/methodology/specific/population/future/imps/updates/downloads/msiwpaug09.pdf. As a result, for consistency, revisions are also planned for these other population estimates to follow on from these local authority revisions:

• LSOAs and MSOAs
• Wards (for both CAS wards and 'current' wards)
• National Parks
• Primary Care Organisations
• Parliamentary constituencies

Development of a Postcode Best Fit methodology

A 'Postcode Best Fit' methodology has been developed to derive population estimates for other geographies from the LSOA estimates. In essence the Postcode Best Fit methodology is an Apportionment method, apportioning population estimates from the smallest small area geography for which population estimates are published by ONS - LSOAs, to unit postcode level based on age and sex information from patient register postcode level data.

This Postcode Best Fit methodology uses the population estimates for the 34,378 LSOAs in England and Wales by age and sex and apportions these to around 1.6m residential and communal establishment postcodes in England and Wales (with an average population of 33) based on the counts of persons by age and sex included on the patient registers.

These postcode level estimates can then be aggregated (or 'best fitted') to a range of higher geographies using a suitable postcode look-up file, for example the National Statistics Postcode Directory (NSPD). A special allowance is made for population sub-groups not included on the patient registers, covering prisoners, UK armed forces, and foreign armed forces and dependants. The LSOA counts for this special population are removed from the apportioning process and then added back in at unit postcode level, based on postcode information for the special population.

Approval was previously given by the ONS Statistical Policy Committee that this method be used for producing estimates for key geographies where these were not previously available and that they be given experimental statistics status, with user views sought on their quality. The estimates for wards, Parliamentary Constituencies and National Parks have been produced using this Postcode Best Fit methodology. These estimates are therefore consistent with both the published local authority and LSOA mid-year estimates.

Postcode Best Fit article in Autumn 2008 Population Trends journal (pages 28 to 34) (pdf, 1.9Mb)

Northern Ireland

In Northern Ireland small area population estimates for mid-1999 by broad age group are currently available at electoral ward and 1991 Census Enumeration District, produced by the Northern Ireland Statistics and Research Agency (NISRA). These are consistent with the mid-year population estimates produced by the Registrar General for Northern Ireland. These estimates were derived as part of a project to produce the Northern Ireland Multiple Deprivation Measure 2001 Download these and other statistics on deprivation measures from NISRA.

The small-area statistics from the 2001 Census of Population became available in 2003 and the Inter-Departmental Statistics Co-ordinating Group agreed that it was timely to review spatial measures of deprivation in Northern Ireland. A steering group including membership from central and local government, non-Departmental Public Bodies and the community sector are steering the review. The group issued a public tender for this research, and subsequently awarded a contract to a team led by Mike Noble at the Department of Social Policy and Social Work, University of Oxford. The final publication of the research was in May 2005. A download of the Northern Ireland Multiple Deprivation Measure is available from NISRA.

As part of this review, small area population estimates for mid-2003 were created. Population estimates are available for Output Area, Super Output Area, Local Government District and Parliamentary Constituency.

The Northern Ireland Neighbourhood Information Service (NINIS) provides free access to statistical and locational information relating to small areas across Northern Ireland. Both the NINIS main site and the newly updated NINIS site for deprivation allow users to download an area profile and use a built-in mapping facility.

The NINIS site now also incorporates an interactive mapping facility on both the main NINIS site as well as the Deprivation 2005 facility. Users have access to ready made, interactive maps for Deprivation Measures as well as Census 2001 Key Statistics.

Scotland

In Scotland, mid-2001 to mid-2007 datazone based estimates by gender and five-year age group produced by the General Register Office for Scotland (GROS) are available. These estimates are consistent with the mid-year population estimates and have been produced using the Cohort Component method and evaluated in conjunction with Scottish local authorities. The estimates and a methodology paper are available on the GROS website.

Further Information

Updates on the progress of this ONS project and availability of small area population estimates for England and Wales, Northern Ireland and Scotland will be available on this web page. If you would like further information about small area population estimates or would like to feed in your views as a user of the data please contact:

For England and Wales

tel: 01329 444664 or email SAPE@ons.gov.uk

For Northern Ireland

Dr Jos Ijpelaar
Demography and Methodology Branch
Northern Ireland Statistics and Research Agency
McAuley House
2-14 Castle Street
Belfast
BT1 1SA

tel:  028 9034 8271 or email Dr Jos Ijpelaar jos.ijpelaar@dfpni.gov.uk

For Scotland

Population and Migration Statistics Team
General Register Office for Scotland
Ladywell House
Ladywell Road
Edinburgh
EH12 7TF

tel: 0131 314 4298

top

This page last revised: Thursday, 1 October 2009