Changes to DWP statistical geographies in National and Official statistics
Updated 13 November 2018
© Crown copyright 2018
This publication is licensed under the terms of the Open Government Licence v3.0 except where otherwise stated. To view this licence, visit nationalarchives.gov.uk/doc/open-government-licence/version/3 or write to the Information Policy Team, The National Archives, Kew, London TW9 4DU, or email: psi@nationalarchives.gov.uk.
Where we have identified any third party copyright information you will need to obtain permission from the copyright holders concerned.
This publication is available at https://www.gov.uk/government/publications/dwp-statistical-geographies-in-national-and-official-statistics/changes-to-dwp-statistical-geographies-in-national-and-official-statistics
Background
Due to increasing demand from users, the Department for Work and Pensions (DWP) is making improvements to the published geography information in its National and Official statistics releases. This includes data released via Stat-Xplore.
The changes will be made throughout 2018. They will involve:
- replacing all 2001 Census Output Area (COA) based geography information with more up to date geographies based on 2011 COAs
- assigning geographies using an improved methodology, which involves using more reliable data to source address information
If you have any comments or questions regarding this, email: stats-consultation@dwp.gsi.gov.uk.
Updated 2011 COA based geographies
The benefits and data which have been updated and now use the new methodology to assign 2011 COA based geographies include:
- Children in out-of-work benefit households statistics
- Universal Credit statistics
- sanctions for Universal Credit, Employment Support Allowance, Income Support and Jobseeker’s Allowance
- Employment and Support Allowance Work Capability Assessments
- Housing Benefit
- Personal Independence Payments
- DWP legacy benefits (published as part of the quarterly DWP benefits statistics release)
The benefits and data still to be updated with the new methodology to assign 2011 COA based geographies include:
- Benefit cap
- National Insurance number allocations
- Winter Fuel Payment
- Work Programme
Effects of changing to 2011 COAs
The following information is a summary of the methodology and the effect of the changes. More details will be available within each statistical publication once the changes are made.
The Office for National Statistics (ONS) guidance states that 97% of 2001 COAs in England and Wales were unchanged against the 2011 census. This does not necessarily mean that 97% of our claimants’ geography allocations will remain the same because:
- we have changed our methodology to assign geographical information to claimants, to make it more accurate
- claimants are not necessarily spread evenly across COAs
- we publish at a range of different geographical levels, which may be differently affected by changes to COAs
The changes caused by improving the geography allocation methodology will also be minor. We believe the information will be more accurate due to:
- more accurate source address data
- a more robust method of allocating geographies when no current address information is held, by using previous address information rather than random allocation
We expect differences to be minimal for higher-level geographies, such as region and local authority.
The differences will become more visible with lower-level geography breakdowns, for example lower super output areas.
As an example of potential changes, Newcastle upon Tyne had 889 output areas in the 2001 census, which increased to 910 in the 2011 census. This was, among other reasons, to reflect the additional population (rising from 260,000 to 280,000); there should only be around 300 people living in a COA.
In some rare cases, these changes result in boundary changes to higher-level geographies, such as Local Authorities, which are built up from aggregating COAs together.
Read the guidance from ONS about census geographies, including links to further information about the changes from 2001 COAs to 2011 COAs.
Changes to Methodology
Current method
The current method of assigning geographies was developed over 15 years ago and uses the postcode from the administrative General Matching Service (GMS) data.
The methodology includes a benefit hierarchy system to obtain the most reliable postcode, and a method of random allocation to assign geographies where a postcode is missing. The method of allocating geographies to each postcode in the data uses an out of date version of the ONS Postcode Directory (ONSPD), a geography reference file. The file doesn’t contain 2011 COAs.
New method
The new method uses address information from the Customer Information System (CIS). This is thought to be the most reliable source of addresses as it links to all of the DWP benefit systems and contains the most up-to-date address for each individual.
The addresses are then put through a data cleansing procedure which makes sure postcodes are formatted correctly and the address fields are populated correctly.
2011 COAs are then assigned to claimants in this dataset using the ONSPD, starting with a direct postcode to COA lookup and then working through a logical allocation routine. These COAs are then used to merge on higher level geographies from the National Statistics Postcode Look-Up (NSPL).
We match the file onto our datasets, ensuring that our data extract date is between the address start and end of spell dates. If no known address can be matched for a period of time, a previous known address for that individual is used. This is thought to be more accurate than random allocation.