The big scary numbers in the IoD report
Today’s IoD sponsored report on public sector pensions combines pensions envy with the standard deployment of big scary numbers about the unaffordable costs of public sector pensions – each one bigger than the previous report.
The numbers in this report have been somewhat undermined by the change in the indexing of pensions in payment announced in the budget. Linking pensions to CPI rather than RPI will reduce their cost as in most years pensions will go up by less than the increase of cost of living for pensioners, but as the whole report is based on an edifice that has been dismissed by the National Audit Office it is important to understand where these numbers come from.The key figure is a somewhat arcane special interest rate known as the discount rate. This is used to calculate the cost of future commitments in today’s money.
As pensions commitments go many years into the future small changes in this can make a big difference to such a cost. Insisting that the government uses a lower discount rate in its public sector pensions calculations has been the continuing refrain of almost all the think-tank attacks on public sector pensions over the years.
Almost all of the arguments in today’s report are built on this argument. If the discount rate was lower then the future liabilities of public sector pensions go up. That makes them more valuable to staff today, hence the headlines on the release:
“Public sector pensions twice as valuable as previously thought”
“Public sector pensions worth, on average, at least 40 per cent of salary”
Notional interest rates also have an important technical role in public sector pensions. The contributions that employers and employees make to unfunded public sector pensions are calculated using a process called SCAPE—Superannuation Contributions Adjusted for Past Experience.
The public sector pensions under attack today are the so-called unfunded schemes (ie the vast majority of public sector schemes apart from the big funded exception of the Local Government Pension Scheme.)
They do not invest money in the kind of assets that private schemes (and the LGPS) do. Instead the Treasury puts contributions into the general public finance pot – and pays pensions from this too. This is good for the scheme member as they do not suffer investment risk, but it is also good for the tax payer. All these contributions are effectively lent to the tax-payer. If they were put into a standard investment pot, the public finances would have to raise taxes or borrow to fill this funding gap. (Unfunded is a bit of a misnomer therefore, but it is either that or pay-as-you-go is what they are usually called.)
SCAPE uses a notional annual 3.5 per cent interest rate on contributions to work out what the employer and employee has to pay to fund the future pension of a scheme member who retires. This report repeats the refrain that this is too generous. This is what it says about discount rates:
Firstly, the Government uses an artificial discount rate to report unfunded liabilities, based on AA-rated corporate bonds, rather than a more pertinent discount rate based on index-linked gilts (inflation-linked securities issued by the Government to finance its borrowing). Since the Government’s chosen discount rate is higher, it has the effect of lowering the value of the outstanding liabilities. Secondly, public sector employers and employees are not charged the full current service costs of the liabilities the pension schemes are taking on each year: a completely different discount rate chosen by the Government is used to compute these. This also means that employees undervalue the benefit of a public sector pension. According to government numbers, the main unfunded schemes have combined employee and employer contribution rates of around 20 per cent of salary. The true value of such schemes, when measured using a discount rate based on the current yields (0.8 per cent) on index-linked gilts, is over 40 per cent of salary. Even when measured by discounting according to a long-term real gilt return of 2 per cent in line with economic growth at that level, the true value comes out at almost 30 per cent of salary.
It is perfectly possible to justify the government’s methodology as the Chief Actuary does here.
What particularly annoys the right is that two different discount rates are used in calculating the costs of future liabilities today and the calculations used to work out contributions.
But the real world is full of different interest rates for different purposes. There is no reason why they should be the same in each of these two very different calculations. The question is whether each can be justified.
Let’s first deal with the SCAPE rate. The whole debate about public sector pensions is full of comparisons with the private sector except here. 3.5 per cent growth in investments each year is a perfectly reasonable investment return assumption to make over time. The Standard Life pensions calculator here assumes 7 per cent growth (less 1 per cent charges). Arguably 3.5 per cent is too low, but then 7 per cent is arguably too high.
Most of the big scary numbers though flow from the other discount rate – the one used to express tomorrow’s liabilities today.
The problem with this is that it is being used to calculate a fairly meaningless number. The government does not have to put aside this money to pay for future pensions liabilities. This is not how the pensions system works. What it has to work out is whether it can afford each year to make not just the pensions promises it has already built up (which is what the liability figure tries to catch) but also the pensions promises that will be built up in the future too. In other words it should want to know whether in 2025 it can meet not just the pensions promises already built up by current staff who will have retired and today’s pensioners who are still alive then, but also pensions built up by current and future staff who will have retired by 2025.
The Treasury say the best way to do this is to estimate what pension payments will be as a proportion of GDP. This takes no account of pension contributions, but simply what will the cost of meeting its commitments will be.
This approach has been endorsed by both the National Audit Office and the Office of Budget Responsibility.
The NAO is very clear that the discount rate is a distraction:
Changes in the discount rate lead to large fluctuations in the size of pension liabilities, but have no effect on projected pension payments. For example, the discount rate increased by 0.7 per cent in the year to 31 March 2009 for the four largest schemes. There were no other changes that year to key financial assumptions underlying liabilities, but the discount rate change alone reduced the total liability across all four schemes (teachers, health, civil service, armed forces) by approximately £73 billion.
This is a point worth emphasising. We currently have historically very low interest rates. You would therefore need to invest much more money today to fund something in the future if these rates held low. But of course while we hope they do until the recovery is secure, they will not stay this low for ever – and pensions work on long time-scales.
Rightly the NAO concentrate on the cost of pensions in payment because;
- projected cash payments are considered by the government to be the most relevant measure of the cost of UK public sector pay-as-you-go pension schemes over the next fifty years;
- projected annual cash payments can be related to estimated annual Gross Domestic Product as a measure of the country’s ability to pay;
- cash projections include pensions expected to be earned in the future, and are useful for decision-making about changes to schemes, whereas liabilities represent only pensions already earned that would be unaffected by scheme changes; and
- liability calculations can fluctuate substantially because of changes in one significant assumption, the discount rate, which does not affect cash payment projections.”
So what do these projections show?
You might expect them to show the cost of public sector pensions to go up as part of the inevitable costs of an aging society. They certainly will for state pensions, health and social care according to Treasury projections. But as the graph shows public sector pensions go up a little and then fall – and these estimates were made before costs were reduced through introducing the new link to CPI. (you can click on the graph for a bigger, clearer image.)