Your browser is no longer supported

For the best possible experience using our website we recommend you upgrade to a newer version or another browser.

Your browser appears to have cookies disabled. For the best experience of this website, please enable cookies in your browser

We'll assume we have your consent to use cookies, for example so you won't need to log in each time you visit our site.
Learn more

To cut its costs, LGPS should fix its data

  • Comment

Error-strewn data can force actuaries to overstate a fund’s liabilities but cleaning it up is likely to prove tricky. Jimmy Nicholls reports

The Local Government Pension Scheme (LGPS) is not alone in the public sector – let alone the wider economy – in having data problems, but most data problems do not result in bloated financial liabilities and assets misallocated due to a false view of how well funded a scheme is.

The rewards for fixing bad data are therefore clearer than in many sectors. Earlier this year Catherine McFadyen, a partner and actuary at consultancy Hymans Robertson, told a pensions conference that cleaning up data at one pension fund client, including removing backlogs and leavers, had reduced the client’s liabilities by 5%.

“There is a premium for getting our data right. Ultimately, better data leads to better decisions which leads to better outcomes,” she told the Pensions and Lifetime Savings Association Local Authority Conference. “There is definitely an accuracy reward to be had on the liabilities too.”

Dodgy data bloats liabilities because once it is handed to actuaries they must make some conservatives guesses as to what it isn’t telling them. A pension fund may have fewer obligations to someone who is no longer contributing than is guessed at, for instance.

Phil Triggs, tri-borough director of treasury and pensions at Westminster City Council, says: “What normally will happen is where an actuary is working with compromised data or data of a lesser quality, they tend to take a prudent assumption on the missing or compromised data.

“Therefore, the liability normally works out higher than if accurate data had been used. Obviously, if the fund can get to work on the data and get it back to a desirable standard then the valuation can be more accurate, and the contribution rates reduced.”

Ultimately, better data leads to better decisions which leads to better outcomes

Checking the figures

It sounds like a no-brainer but maintaining accurate records of pensions members is tricky. Even something as seemingly straightforward as logging who has left and arrived at a given job across employers cannot be taken for granted. Sometimes people stop making contributions and it is not obvious why.

Data checks tend to occur in two instances. An annual check takes place when employers submit their returns, looking for starters and leavers that have not been logged, as well as inaccurate pay. Additionally, every three years when a data set is given to an actuary for calculating liabilities checks take place.

“Because of the way things used to be done, there’s a lot of historical data that hasn’t been checked year-by-year to the extent that we would do it now,” says Jeff Houston, head of pensions at the Local Government Association. “There’s potentially an awfully big job.”

The LGPS ran as a final salary scheme up until 2014, and then changed to a career-average revalued earnings (Care) system. Pensions once calculated at the end of a career must now be done every year. This, as Mr Houston puts it, “has shifted the priority in terms of getting your data right”.

Under the final salary system, resources could be thrown at working out kinks in membership numbers and pensionable pay figures at the point of calculation. “That was the attitude – however right or wrong,” Mr Houston says.

“A lot of this is about resources and sometimes you have to prioritise. Providing you could throw the resource at it at the point of calculation to get the pay right and the membership right, what happened before that was important but not vital.”

Piercing together accurate pensions data records is ‘an awfully big job’

Under the Care scheme, pensions are not only calculated every year but the figures are then combined with final pay for pre-2014 service, boosting the complexity of the sums.

“This has put most pension teams and employers under a significant amount of pressure to get it right every time, every year,” Mr Houston says.

Westminster’s Mr Triggs adds: “In terms of data submission and the requirements to have a good, accurate database, it’s far more incumbent on the authority to have this precise record.

“There has to be a quality process and there has to be a secure portal that holds this in a proper way, so that the data is extractable and provides meaningful information for those final pensions to be calculated.”

Mr Triggs says that outsourcing has contributed to the problem. “A lot of it arises from outsourced services for pensions administration. The private sector has taken on administration and it’s not worked out well, so eventually it’s been brought back in house or possibly transferred to another administering authority,” he says.

“My preference has always been to keep pensions administration in house. Some external administration providers have not done the private sector reputation any good. One or two providers are regarded across the spectrum as a disaster in terms of admin outsourcing.”

Hit reset

One fact that is emblematic of the LGPS’s data difficulties is that many of the IT systems handling payroll data relevant to pensions struggle to communicate in ways that would allow better, more automated processing.

“Being able to automate pensionable pay exactly has been the holy grail for pensions ever since I’ve been in the business, which is an awful long time,” says Mr Houston. He adds that the easiest way to improve data handling in LGPS would be for everyone to use a good-quality payroll system.

“We’ve got hundreds of payroll systems out there, and they vary enormously in their ability to deliver what we need,” he says. “You get situations where an employer will get a new payroll system and suddenly you’ve got no history.”

Understandably, most payroll procurement is intended at achieving payroll, not pensions. But it is notable, says Mr Houston, that all payroll systems do tax well. “If you get your tax stuff wrong there are serious implications, which is why your payroll systems do tax well,” he says.

There is no reason to believe that payroll systems will be made uniform any time soon, even if programmers make them better at talking to other software over time. Even so, Mr Triggs says: “I think as we’ve moved on these systems have got more reliable and generated better ways of extracting the information we require.”

Making the scheme simpler for employers is another aim for reformers. “Our choices are either we make the scheme simpler – which is a significant legislative challenge – or we make it easier for the employers to provide the information that’s needed. It’ll be probably end up being a combination of the two,” Mr Houston says.

“We’ll make it simpler where we can, but spend more time recognising most employers are not there to do pensions, they are there to do something else. All this complicated pensions stuff is not something that’s a priority for them.”

Being able to automate pensionable pay exactly has been the holy grail for pensions ever since I’ve been in the business, which is an awful long time

Priority or not, non-compliant employers can face pain. Under regulations, administering authorities can recover extra costs from employers incurred in correcting data. This is termed “fining” by some, but it is described as recovery in the rules. According to Mr Houston, recoveries have become more common since the switch to a Care scheme in 2014.

He offered some mundane solutions to improving employer compliance with the rules. While the LGA produces a “nice, long employer guide” Mr Houston says many of them would prefer a YouTube video explaining how to record that somebody is leaving a job.

More long-term is allocating more resources on both the authority and employer sides. “You then have to put in place systems to keep it as up to date as possible from that point onwards, and then ensure that you don’t start letting those backlogs build up again,” Mr Houston says.

For pensions teams resources can vary wildly, despite the often limited direct impact of austerity. Some, such as Greater Manchester Pension Fund, are run separately from the rest of the council functions giving them greater leeway to hire staff to clean up data.

“If they need to pay for a project to check all their own data because they think it’s going to save money they go to the pensions committee, they make the business case and, if it’s a good case, the committee approves,” Mr Houston says.

“At the other end of the extreme are authorities where the function can be just seen as another team within the council. If you have cutbacks or a recruitment freeze they will apply to the pensions team as much as anybody else.”

With austerity ongoing, it is unlikely that such authorities will get much reprieve to allow them to improve their processes to prevent data errors and mount efforts to clear up mistakes already embedded in the systems.

  • Comment

Have your say

You must sign in to make a comment

Please remember that the submission of any material is governed by our Terms and Conditions and by submitting material you confirm your agreement to these Terms and Conditions.

Links may be included in your comments but HTML is not permitted.