The call for input on staging dates, recently issued by the Pensions Dashboards Programme, makes the dashboard feel very real.

That's great news. The dashboard is the best opportunity in a generation to improve engagement with pension savers.

However, those savers will not thank us if the dashboard gives them unreliable information or shares their personal information unintentionally with other people. My concern is that the current approach risks doing just that - particularly for defined benefit (DB) schemes.

As a pensions industry that understands the challenges, we need to call this out. At the risk of being labelled anti-dashboard (which I'm not), we should walk before we run.

First, the concept of estimated retirement income (ERI) for DB schemes remains flawed.
Converting a detailed historic DB promise - with multiple NRAs, bridging pensions, GMP uplifts and equalisation - into a single (or small number of) ERI will result in simplification.

There will be a difference between what members see on the dashboard and what they get told by their scheme. That will undermine trust.

From a practical perspective, retirement quotes in most DB schemes are calculated only when requested and are checked for error by a human being. Including ERI on the dashboard will be a huge undertaking to automate calculations, with large cost implications for schemes and sponsors, which may be repeated once GMP equalisation is complete. As well as the cost, the lack of human oversight will inevitably result in mistakes.

Secondly, matching for pension schemes is wholly untested.

A key issue is that of false positives and false negatives – how often a member gets sent data that is not theirs, and how often a member fails to find a record that really is theirs.

The comparison with open banking is often thrown around. But with open banking you link your banks accounts yourself. You don't ask your bank to email your data to hundreds of other banks, to check whether they have an account with you. Until this is tested on real pensions data, we have no idea how often records might end up linked to the wrong people.

We also need to move away from the idea that matching algorithms should be set by individual schemes and providers. It is unrealistic and inefficient in so many ways. But added to this, if a provider or scheme is expected to take responsibility for any adverse consequences – for example, the release of data to the wrong person - they will be well advised to opt for a cautious approach to matching, which means fewer matches. A common approach to matching, or at least a common default, is essential.

All these issues could further undermine trust in pensions. So, what is the solution?

Given the uncertainties around matching and the challenge of producing a meaningful and reliable ERI, most people I speak to are supportive of mandatory dashboard requirements being initially limited to finding records and scheme contact details. ERI would be left until such time that matching has been tested and retirement incomes are properly defined and reliably available.

This would allow schemes to connect more quickly, and retirement estimates could still be offered voluntarily by schemes that are comfortable to do so.

The Pensions Administration Standards Association (Pasa) has recently issued a similar message, which they call "safe, simple, soon". But whether you explain it in three words or the 600-odd in this article, they come down to the same thing.

Of course, we could be wrong. Actuaries are cautious, it's in our DNA to point out risks. It might all be fine. The point is we don't know.

If the policymakers and regulators who make the final decision want to push through the current proposals and accept the risks, then they are entitled to do so. But what they cannot do is push them through and then turn around when things go wrong and say nobody warned them.

Share Story: