Personal Independence Payment statistics - data quality and methodology

Date: 25 August 2021

Introduction

This paper outlines the processes involved, and the steps taken by the DfC analysts in publishing the statistics for PIP for Northern Ireland. It provides details on the data source, data quality management actions and the creation of data table structures for administrative use as well as limitations of the processes involved. The statistics are produced using the Code of Practice for Official Statistics. They have been assessed by the Office for Statistics Regulation and were awarded the designation of National Statistics by the UK Statistics Authority and Office for National Statistics (ONS) on 15 October 2021. Further details are available.

Data source and supplier

Data from the PIP Computer System (PIPCS; used by PIP case workers and case managers to administer and manage PIP claims) are transferred to the PIP Atomic Data Store (PIPADS) held by the Department for Work and Pensions (DWP).

A suite of data tables are then extracted and supplied to the Department for Communities (DfC) by DWP using approved high secure transfers on a monthly basis representing a 100% population of PIP claimants with a postcode in Northern Ireland at the associated dates. A Service Level Agreement is in place for this transfer.

Quality management actions

Robust quality and security processes are completed by DfC Professional Services Unit (PSU) in the development of outputs and as routine quality assurance measures in production of the statistics.

Routine checks and measures include:

• Data received from suppliers is stored by the PSU Database Management Team (DMT) in a secure server with restricted access.
• Checks are made that the count of files received are as expected, files are named as expected and that file sizes are as expected.
• Comparisons of tables received are made with those from the previous month.
• Any variations from what is expected are investigated and if necessary data suppliers are contacted to raise issues.
• Encryption of personal identifiers by Database Maintenance Team and removal of personal or sensitive variables, before loading onto a restricted database for use by internal analysts.
• The Data Suppliers (DWP) will inform DfC if there are any changes or issues with the PIPCS.
• Quality checks of the quarterly statistical outputs for publication including verification of figures, comparisons with the previous quarterly outputs, updates are made to data notes where appropriate, proof reading and accessibility checks are carried out by all team members.
• Communications with the DfC Policy and Operations staff occur regularly to ensure PSU are aware of any policy changes or operational issues that may affect the data.
• Comparisons with data from the DfC Management Information Systems as an additional quality measure when necessary.

Additional quality measures in the development of new statistical outputs:

• Checks on the reliability and completeness of variables that will be used in analysis.
• Communications with the DfC PIP Operations team who oversee processing of the data for further clarification on variables and any limitations in their use.
• Communications and regular meetings with DWP on methodology in data development.

DWP also carry out their own quality measures, and details for accessing this can be found in the ‘Further Information’ section of this document.

Limitations and other issues

Although analysts in DfC receive a large suite of data tables on a monthly basis from DWP, they do not have access to the all data tables relating to PIP held by DWP in the PIPADS due to the large volumes of data. This can limit the information that DfC analysts can create for use regarding PIP statistics. In some circumstances, DfC will request particular information to be extracted by DWP analysts where appropriate and possible.

Errors may occasionally arise due to IT system errors or human error. The following measures are in place for managing such risks:

• DWP inform PSU if there are any known issues with the PIPCS.
• If PSU identify a potential error or issue in the data that they have not been aware of, PSU will contact DWP for further information.
• In some cases, sample data may be transferred to the DfC PIP Operations team for further investigation of suspected errors. A Data Sharing Agreement is in place to cover this transfer and Data Protection Impact Assessments have been completed as part of this process.

Creation of data tables for analytical and statistical purposes

Data arrives with DfC in structures that do not provide immediate insight to PIP figures in Northern Ireland. In order to create meaningful material for analysing and reporting, information must collated from multiple variables across the various tables supplied, connected using unique linking variables.

Information within data tables are based on claims and events relating to that claim. Each claim has a unique identifier that is common across data tables so that internal analysts can create datasets that provide relevant details for claims at different stages of the claim journey. The following datasets are created by DfC:

Registrations - The earliest event for each claim ID with a postcode in Northern Ireland after the introduction of PIP in June 2016 is selected. This includes claims that do not progress through the PIP customer journey. Where claimants go on to make further claims, each claim is included in registrations figures.

Clearances - The earliest clearance event information is provided for each claim. This does not include subsequent clearances for the claim such as mandatory reconsiderations and appeals. Where claimants have made multiple claims, each claim is included in the clearance figures.

Clearance times are reported using the median measure (the middle value if all clearance times were ordered from lowest to highest value). The mean measure is not used due to the effects of extreme values (for example, due to individuals being difficult to reach or failed to attend numerous assessment appointments).

Claims in Payment - Claims listed as 'live in payment' are selected from relevant data tables and information provided is sorted to produce the latest information recorded for each claimant each month. This sorting includes the removal of records stating award rates of 'not supplied' or 'Nil' on both the daily living and mobility components in order to obtain the latest figures of awards with payment.

Mandatory Reconsiderations (MR) – Datasets are created by PSU from data tables provided by DWP on MR registrations and clearances. Similar to claims, each MR has a unique identifier. Claimants may have more than one mandatory reconsideration per claim, as well as more than one claim. MR registrations are identified at the earliest event recorded for the MR in Northern Ireland. MR clearances refer to the earliest clearance for the MR and do not include subsequent clearances. Where claimants have multiple mandatory reconsiderations, each MR is included in the MR clearance figures.

Retrospection

The figures contained within the publication and supplementary tables are subject to scheduled revisions. This is to account for retrospective actions on the Personal Independence Payment Computer System (PIPCS). These revisions are performed in accordance with T3.9 of the UK Statistics Authority Code of Practice for Statistics. To ensure that retrospective changes are captured in the published statistics, Northern Ireland PIP tables are recreated for every quarterly release, rather than appending new information to historic files.

Further information

The quarterly-released statistics bulletin and the accompanying Supplementary Tables documents are published and accessible from the dedicated DfC PIP Statistics web page.

Changes and planned changes to the Northern Ireland PIP statistics published by DfC are listed on the working document ‘Changes to Published Statistics’ document, also available at the DfC PIP statistics web page.

More information on the PIP benefit for Northern Ireland, is available.

DWP provide statistics for PIP in Great Britain, and these can be accessed online. Information on the methodology used by DWP for PIP statistics in Great Britain is also available.

If you wish to have your say on any aspect of the PIP statistics for Northern Ireland, please complete the PIP Statistics User Survey.