The Pensions Administration Standards Association (PASA) Data Working Group has published an updated version of its Dashboards Data Matching Conventions (DMC) guidance, which sets out how pension schemes must choose how they wish to compare ‘find requests’ from dashboard users against the member records they hold. The DMC guidance, originally issued in December 2021, has been updated to include a call to action, and also details the next steps being worked on, as well as providing links to other industry guidance.
In the pensions dashboards ecosystem, thousands of these ‘find requests’ will be sent from the central Pension Finder Service (PFS) every day to all connected data providers. Data providers will compare personal data elements in each incoming find request, against the personal data elements they hold across their records (see the diagram below, courtesy of Pensions Dashboard Programme – PDP).
From 2023 onwards, on their staging date, all pension schemes and providers must comply with new legislative duties for pensions dashboards. As part of this, schemes and providers need to soon decide how they want to digitally compare and match ‘find requests’ from users of dashboards against all the records they hold. The detail of the new legislative duties will be published for consultation in early 2022, but schemes can start to consider the importance of matching and data quality now. The updated DMC Guidance is intended to help schemes with making these matching decisions.
Start to consider the importance of your matching and data quality now –
Three core data elements
Based on current practice, many schemes will decide to match on Surname, Date of Birth (DOB) and National Insurance Number (NINO). Some may also add the fourth data element of Forename.
Data accuracy is key
Schemes should be working towards having a high level of confidence in the accuracy of Surnames, DOBs and NINOs for deferred and active members. This will involve continual, systematic checking of the accuracy of actual values held in these data elements.
Accurate data enables simple matching
Only where schemes believe they are successfully validating the accuracy of all Surnames, DOBs and NINOs, will ‘simple’ matching on these three core data elements produce robust positive matches (in the Guidance, this is referred to as Option 1).
Third parties control data
Administrators understand schemes and providers have never been able to fully control the accuracy of personal data. Third parties (such as employers, deferred members, previous administrators, etc.) are responsible for data accuracy, both initially on joining and over the whole lifetime of individuals’ pension entitlements.
Data assessment and improvement
Schemes work hard to continually assess and improve, as far as they can, the accuracy of data such as Surnames, DOBs and NINOs. Many schemes are on a journey to high confidence in this data, but given the reliance on third parties, full accuracy may never be possible.
Risks of simple matching
Where a scheme has yet to reach a high level of confidence in their personal data, there are risks of choosing simple digital matching on Surname, DOB and NINO. These risks will negatively impact members, administrators, trustees, and others.
Competing legislative risks
Schemes will have to assess their risks against the twin objectives of satisfying data protection requirements as well as complying with their new statutory duties to carry out matching under the forthcoming pensions dashboards legislation.
More sophisticated matching options
Based on the previous two points, it is likely schemes still on their journey to high confidence in data accuracy will wish to choose more sophisticated matching approaches. The Guidance sets out examples of more sophisticated matching: ‘maybe matching’ rules where some of Surname, DOB and NINO nearly match and using other data elements such as Address Line 1 and Postcode.
Deciding how to match depends on your scheme’s data accuracy
The key point for schemes to understand is their decision(s) on how to match must be made in the context of their specific scheme’s known data accuracy (particularly Surnames, DOBs and NINOs across all deferred and active records).
Collaboration is key
Schemes need to engage in detail on this topic soon. They should work collaboratively with their administrators, technology providers and wider data specialists.
PASA’s call to action –
This iteration of DMC Guidance is being published to support schemes’ immediate preparations. Supplemental Guidance may follow, reflecting further learnings from any testing, or further clarifications. Also, there may be other topics to consider.
In the meantime, all schemes should now be engaging on this important topic of matching. Working very collaboratively with their administrator, schemes should seek to understand the level of confidence in the accuracy of the surnames, DOBs and NINOs on all deferred and active member records. Schemes should take steps to improve the accuracy of this personal data. For more on this, see the Guidance published by the PASA Data Working Group: https://www.pasa-uk.com/guidance/data.
