Scribd Logo
Upload

Welcome to Scribd!

  • 19 views

    DWP Information Centre: Small Area Benefit Statistics

    Uploaded by

    Victor Issac
    The DWP Information Centre provides social security statistical and management information to internal and external users. It produces small area benefit statistics by cleaning claimant data, geocoding addresses, and imputing missing geographies. The methodology aims to distribute unknown cases similarly to known ones while protecting claimant privacy through rounding. The center covers many benefits and seeks future uses of administrative data.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PPT, PDF, TXT or read online from Scribd

    DWP Information Centre: Small Area Benefit Statistics

    Uploaded by

    Victor Issac
    19 views21 pages
    The DWP Information Centre provides social security statistical and management information to internal and external users. It produces small area benefit statistics by cleaning claimant data, geocoding addresses, and imputing missing geographies. The methodology aims to distribute unknown cases similarly to known ones while protecting claimant privacy through rounding. The center covers many benefits and seeks future uses of administrative data.

    Original Title

    c2

    Copyright

    © Attribution Non-Commercial (BY-NC)

    Available Formats

    PPT, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The DWP Information Centre provides social security statistical and management information to internal and external users. It produces small area benefit statistics by cleaning claimant data, geocoding addresses, and imputing missing geographies. The methodology aims to distribute unknown cases similarly to known ones while protecting claimant privacy through rounding. The center covers many benefits and seeks future uses of administrative data.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PPT, PDF, TXT or read online from Scribd
    Download as ppt, pdf, or txt
    19 views21 pages

    DWP Information Centre: Small Area Benefit Statistics

    Uploaded by

    Victor Issac
    The DWP Information Centre provides social security statistical and management information to internal and external users. It produces small area benefit statistics by cleaning claimant data, geocoding addresses, and imputing missing geographies. The methodology aims to distribute unknown cases similarly to known ones while protecting claimant privacy through rounding. The center covers many benefits and seeks future uses of administrative data.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PPT, PDF, TXT or read online from Scribd
    Download as ppt, pdf, or txt
    You are on page 1of 21

    DWP Information Centre

    Small Area Benefit Statistics

    Methodology Conference June 2001


    Who we are

    DWP Information Centre


    Part of the Analytical Services Division
    Provides Social Security statistical and management
    information outputs to meet the needs of internal and
    external reporting requirements
    Combination of:
    statisticians
    administrative staff
    IT support staff
    Located on the Longbenton site, Newcastle-upon-Tyne
    Our role

    Data on computer ASSIST


    (5%) samples
    tapes relating to Software
    claimants of developed by
    records held on
    benefits at a Information Centre
    DWP admin
    particular to validate and
    data systems
    point in time clean data

    1999: Introduction of Small Areas Team


    Small Area Team

    Aim:
    To produce information on number of claimants for each
    DWP benefit at small area level e.g. electoral ward
    Based on data downloads of 100% of claimants for most
    benefits
    Need to balance:

    User needs Claimant’s


    for smaller right to
    geographies privacy
    Advantages of DWP data

    Main advantages of DWP administrative:

     up to date
     relatively cheap
     complete coverage of GB on that benefit
     possible to link claimants over time
     possible to link claimants across different benefits
    and build ‘client groups’
    Problem Areas

    Caveats:
     scan only contains information needed to administer
    that particular benefit
     based on a single point in time
     some claims handled clerically
     need to ensure confidentiality of claimants and
    prevent disclosure of an individual’s identity

    CURRENT POLICY TO ROUND CLAIMANT COUNTS


    DWP Information Centre
    Typically the administrative system will store the
    following information:

     Name of Claimant
     Full address
     Postcode
     National Insurance number
     Personal details e.g. gender, date of birth
     Details of partners/dependants included in claim
    Production of Small
    Area Data
    Four Main Stages:

    1. Remove non Great Britain cases GB e.g. overseas,


    Northern Ireland, Channel Islands
    2. Run data through software to improve quality of
    postcodes
    3. Match claimants with postcodes to a geographical
    area using ONS postcode directory
    4. Allocate remaining ‘difficult’ cases to an area via
    imputation methodology
    Production of Small
    Area Data: Stage 1

    Removal of non-GB cases


    Fairly easy to cut down data set via use of postcodes or
    administrative codes to identify an overseas case.
    Mainly important for benefits such as
    Retirement Pensions
    Child Benefit (armed services)
    Production of Small
    Area Data: Stage 2

    Use of Quick Address Software (QAS)


    Commercial software that looks at the address of the
    claimant.
     Corrects old/incorrect postcodes
     Corrects incomplete postcodes
     Formats postcodes with correct spaces
     Inserts postcode if missing and if it recognises the
    address
    Production of Small
    Area Data: Stage 3
    QAS processes data then matches against ONS
    Central Postcode Directory
     Allows the adding of geographic codes to each
    claimant record on the data set e.g. Local
    Authority, Electoral Ward and Parliamentary
    Constituency
     For each benefit match at full 7 character, then 6
    characters, down to 4 character postcode

     Income Support: 99.6% of cases matched to


    postcode directory ‘directly’
    Production of Small
    Area Data: Stage 4

    Allocation of ‘difficult’ cases

     Difficult cases - mid 1999 Income Support data =


    0.36%, approx. 13,600 claimants.
     Majority of these have no addresses or are ‘no
    fixed abode’ cases
     Methodology developed to further clarify results
    using ‘processing benefit office’
    Production of Small
    Area Data: Stage 4
     Split dataset into claimants that have been
    allocated and those that still need allocating.
     Sort claimant by local office, then national
    insurance number
     Merge back together so that within a local office
    an allocated case is followed by one unallocated
    case.
     Assign unallocated case the same geographic
    codes e.g. LA and ward as the case immediately
    above it. Intention is that ‘missing postcode’ cases
    are distributed around area in a similar fashion to
    those with a postcode.
    Production of Small
    Area Data: Stage 4 Example

    Allocated Cases sorted by office then NINo

    Local Office LA Count


    Unallocated cases sorted by office
    3111 40UB 1
    then NINo
    3111 40UB 2
    Local Office LA Count
    3111 40UE 3
    3111 1
    4111 40UF 1
    3111 2
    4111 41UX 2 3111 3
    5111 42AB 1 4111 1
    5111 43AB 2 5111 1
    5111 2
    7111 1
    Production of Small
    Area Data: Stage 4 Example
    Merged dataset prior to allocating LA to ‘missing’
    postcode cases
    Local Office LA Count
    3111 40UB 1
    3111 1
    3111 40UB 2
    3111 2
    3111 40UE 3
    3111 3
    4111 40UF 1
    4111 1
    4111 41UX 1
    5111 42AB 1
    5111 1
    5111 43AB 2
    5111 2
    Conclusions on
    Imputation Methodology

     Where the proportion to be imputed is low we


    believe the methodology to be robust
     Investigations have shown that the local office
    number is a very good predictor of the Local
    Authority of a claimant
     However for smaller geographical areas the local
    office is a less efficient predictor but…..
    we have very little else that we can use!
    Small Area Data

    Typical claimant information produced:


     number of claimants/partners/dependants
     age bands for claimants
     statistical groups eg: IS Lone parents, elderly etc
     gender
    But
     may not be available for all benefits and
     must not allow the identity of claimants to be
    inferred rounding
    Benefits Covered

    Currently we produce and disseminate data on:


    Income Support Attendance Allowance
    Jobseekers Allowance Incapacity Benefit
    Severe Disability Allowance Family Credit
    Disability Living Allowance

    Mid 1998 data available


    Mid 1999 on June 29th 2001
    Mid 2000 by end of 2001?
    Future uses of benefit data

    Use of DWP benefits information to model low income


    at ward level to help fill gap caused by lack of an income
    question in the 2001 Census
    Use of DWP data to assist in the development of small
    area population estimates by ONS
    Supply data to use within Indices of Local Deprivation
    Sure Start Area Monitoring and Evaluation
    Model data disclosure from administrative sources?
    Contact Details

    Nigel Brough, Penny Sinclair, Kevin Hawthorne


    DWP Information Centre
    Room BP 5201
    DSS Longbenton
    Newcastle upon Tyne
    NE98 1YX
    Contact Details

    Telephone : 0191 - 225 7972

    E-Mail : Nigel.Brough@.dss.gsi.gov.uk
    : Penny.Sinclair@.dss.gsi.gov.uk
    : Kevin.Hawthorne@.dss.gsi.gov.uk

    You might also like