Pensions schemes and poor data – why is it and how to sort it

ImageIn July 2013, the Pensions Regulator released a report which made for disappointing reading. Despite all the efforts to date, many schemes are still falling short of the standards The Pensions Regulator has set for measuring data quality and maintaining accurate records, despite steady improvements in some areas.

In 2009, The Pensions Regulator issued good practice guidance about scheme data and subsequently set targets for schemes to meet its ‘common data’ standards by the end of 2012. These standards relate to the members name, date of birth and National Insurance number, which is a key tool in the identification of individual scheme members.

Clearly, the Pensions Regulator is underwhelmed with the findings and has threatened that if schemes have failed to meet targets for keeping accurate records then it will begin to issue improvement notices and if that does not work, financial penalties on pension schemes.  

For those smaller pension schemes, without access to sufficient resources are the ones causing the Pensions Regulator the biggest headache. They are more likely to be part of the two million or so members where common data has not been measured or significant improvements are required to meet the standard set of 95% having data stored. There is also the issue that the scheme conditional data, that is data specific to the individual scheme, is still not what it needs to be. This is particularly important if the there is a scheme valuation, or even more importantly, a buy-in or scheme wind up in the offing.  

 

Not surprisingly then, the Pensions Regulator has timetabled the end of 2013 to publish the results of its detailed record keeping review and the accompanying guidance.

 

What do Trustees need to do?

First and foremost trustees need to accept that administration needs more air time both in meetings and outside of them.

 

The data used by the administrators is a living thing, it changes (members die, re-marry, leave, move address etc) and corrections to earlier administration errors are made almost daily. As many decisions relating to the scheme are based on the member data, valuation assumptions, investment strategy for example, having the best data available is much more than a ‘nice to have’ it is essential.

 

From this, trustees need to engage with their administrators to develop a plan for both monitoring and improving scheme member data. In turn, administrators need to be open and   

honest with their clients so that they are seen as part of the solution, not the problem. In the past, it has been too often the case of trustees and their advisers, shooting the messenger when the administrator reports data problems.

 

There are tools in the box to use. For example monitoring pensioner deaths is going to bring benefits not only in having more accurate pensioner data but also as a weapon against scheme fraud. Contacting deferred members may also flush out deaths where there is an immediate benefit payable but then no further scheme liability. Asking active members to ensure that details of spouses are accurately recorded could influence the potential scheme dependents pension liability. A simple review of whether a data item is stored in a database is not sufficient – there needs to a check made to ensure that the item makes sense. The use of dummy data  was prevalent in the not so far off past, such as NI numbers based on a universal date of birth NI01010000F denoted a female and ending in M a male. Such data needs to be identified and cleaned.

 

All of this requires a budget and in these times, there are competing needs both for the scheme sponsor and the trustees. But not having a budget, both financial and in time, is a false economy and could well end up costing more money with data having to be put right when the clock is ticking.

Advertisements