Volume 2017 | Issue 31
The Pensions Regulator places great importance on good governance and administration, and ensuring trustees comply with their basic duties, such as good record keeping.
From next year, the scheme returns submitted by trustees will include additional mandatory questions, which will require the disclosure of their scheme’s latest common and conditional data scores. Common data is used to identify members, while conditional data is specific scheme data used to calculate member benefits.
Defined benefit (DB) schemes begin completing their scheme returns for 2018 in January and Defined contribution (DC) schemes in July.
The Regulator has warned that “In your scheme returns you will be obliged to tell us about the quality of your data. If you disappoint us we will take action against you”.
Trustees are accountable for any problems with the administration of their scheme, even though the day-to-day running is normally delegated to a service provider. Since it first published its “good practice” guidance for measuring data back in 2009, the Regulator has been campaigning to raise pension scheme data quality standards. Its current guidance expects trustees to measure their common and conditional data scores at least annually, and to correct errors in their common data, and put plans in place to improve the quality of conditional data. However, as this is not a statutory requirement, and there is no obligation to report breaches to the Regulator, some trustee boards have been slow to comply. The only surprise is that it has taken the Regulator so long to require data scores to be included on scheme returns.
Common data is the data used to identify members. It includes member details common to all schemes such as names, addresses, dates of birth, target retirement dates, marital status and National Insurance numbers. The Regulator expects schemes to have 95% accurate pre-June 2010 common data records and 100% accuracy for post-June 2010 records. Common data is measured on a per-member, rather than per-test, basis. Thus, if a member has one item of common data missing, such as a postcode or date of birth, the test is failed for that member.
In reality, therefore, no scheme is going to have 100% common data scores, as a member only has to change their address, marry, or die, without the trustees being informed, and the common data score will fall below the target. That’s not to say, however, that trustees can be complacent. They must check their scores at least annually (or more frequently if improvement plans require it).
Conditional data is specific scheme data used to calculate benefits. This will vary from scheme to scheme depending on its type, structure and design.
Trustees are expected to consider what data is required for the effective administration of their scheme, and to measure changes in the data over time.
Thus, the Regulator has not tried to produce a data specification covering every possible circumstance. It has however produced tables showing the items that could provide an outline of what can be regarded as conditional data. These tables should only be regarded as illustrative.
Action for Trustees
Trustees of both DB and DC pension schemes need to ensure, by the time they come to complete their scheme returns for 2018, that they have a recent set of data scores for their scheme and that those scores do not “disappoint” the Regulator. (The Regulator would deem “recent” to mean data scores from within 12 months of the scheme return.) Trustees of DB pension schemes that have not completed a data check this year should be scheduling one in as soon as possible, so they can take remedial action if necessary and run a further update before the end of 2018. Trustees of DC pension schemes should be scheduling a data check from August onwards for the same reason.
Impact of the Pensions Dashboard
Whilst the underlying need to improve data scores comes from the Regulator’s initiative, it is against the background of the government’s drive to introduce a pensions dashboard which is intended to enable members to access all their pension savings data online and from one source.
Under the management of the Association of British Insurers, an industry group has built a prototype dashboard, which when fully developed, is intended to allow members to see all their pension savings in one place and allow them to plan for their retirement more effectively. This is intended to be operational by 2019, although no formal date has yet been set for when trustees will be required to arrange for data to be included on the dashboard.
However, work undertaken now to improve scheme data should make the release of information to the pensions dashboard much simpler for trustees.
Trustees will have to provide up to date common and conditional data scores to the Regulator in their 2018 scheme returns. Those scores will need to show the scheme’s data is of sufficient good quality to satisfy the Regulator. Where there are gaps, trustees will need plans in place to improve the quality of the data, in order to answer inevitable follow up questions from the Regulator.
Inevitably, some trustee boards will leave checking their data scores to the eleventh hour. They risk finding either, that their administrators are struggling to do the inevitable surge in work that will occur towards the end of the year (DB) and again in the Spring of 2018 (DC), or that the results of the check disclose some nasty surprises.
Trustees have had since 2009 to improve the standards of their records. They have now been given a further 12 months’ warning by the Regulator that data scores will come under scrutiny in 2018. Those that now fail to take the necessary action are unlikely to find the Regulator in a sympathetic mood.