How do you decide how much to charge public sector employers for making pension promises to employees, when those promises will be largely paid from general taxation as they come due decades from now?
That’s the unsexy but surprisingly knotty question HM Treasury has sought to answer this week, in response to its consultation on the “superannuation contributions adjusted for past experience” discount rate.
(Disclosure klaxon: As a former HMT official I was involved in the consultation. The views expressed here are my own, and not those of my former colleagues or employers. I also have a public sector pension out there, somewhere).
The answer the wonks in Whitehall have landed on is to use long-term GDP forecasts to discount the future value of public service pensions in order to set the charge levied on employers. In other words, 18 months after launching the consultation, it has reconfirmed the existing methodology in place since 2011.
In most contexts, discount rates are used to determine how much needs to be invested at present to meet a future obligation, based on expected returns in the interim. Funded pensions schemes use a discount rate based predominantly on gilt yields — meaning that volatility in gilt yields can radically alter their funding position, as my FT colleague Jonathan Eley recently explained.
But public sector pensions are special. When members retire, pensions are paid out of general government spending, not a fund.
Then why make employers contribute at all?
The aim of employer contributions in this context is to assign a reasonable present-day cost to the pension promises that public sector employers are making, so that they don’t go on a crazy hiring spree and make more promises than our kids and grandkids can meet.
Long-term GDP growth, as a proxy for the future tax base, represents the income stream that the pensions being accrued by doctors, nurses, and Treasury officials will eventually be paid from.
This is a better solution than bond yields, which reflect events happening in secondary markets as much as the governments’ future ability to borrow, and are also highly volatile.
By discounting against future GDP growth, contributions reflect the claim pension promises are making on future revenues.
If the government’s future tax base is expected to grow more slowly, employer contributions will rise, all else being equal, reflecting the fact that meeting those pension promises in the future will require a greater share of future tax revenues.
The big issue is that the OBR’s forecasts for long-term GDP have nosedived since HMT first landed on this answer in 2011.
At that time, projections of average GDP growth over the next 50 years implied a discount rate of 3.0 per cent + CPI. Now, that’s dropped to 1.7 per cent + CPI.
When public sector pension schemes conclude their current round of valuations, employers are likely to see a dramatic increase in their contributions — although less than they would have if life expectancy hadn’t stopped rising.
It’s important to understand that rising employer contributions don’t affect public sector net borrowing. Outflows are based on pensions in payment (ie pension promises made in the past that we can’t change anymore).
Meanwhile, contributions effectively go from public sector employers back into HMT coffers.
Still, if those employers are operating within fixed budgets, an increase in contributions mean less to spend on other priorities.
That also isn’t happening. HMT has said it is “providing funding for increases in employer contribution rates resulting from the 2020 valuations.”
So, what’s the point of all this?
On paper, employment costs will more accurately reflect the benefits that employers are promising to employees.
But because budgets are being topped up, the rise in those costs won’t affect how many people are hired — at least until those budgets are reviewed at the next spending review.
Arguably, it’s silly to base the number of people working in the UK’s creaking public services on how hard — or easy — it will be to pay out their pensions in, like, 50 years time.
But if the idea is to make sure that we are not making more promises than we can meet, it’s arguably better for employers to adjust their workforces sooner rather than later.
Ultimately, though, the wisdom of this approach depends on how accurate the OBR’s 50-year GDP forecasts are. That’s a long time, and lots could happen. The widespread uptake of AI might turbocharge productivity. A large single-market offering frictionless trade might suddenly materialise across the channel.
Or we might have all burnt to a crisp, in which case none of this really matters.