BY RICHARD TAYLOR
On 1 July this year, the Government Actuary’s Department (GAD) told us the Teachers’ Pension Scheme (TPS) for England and Wales (TPS) had magically gotten £3.3bn a year cheaper! It hasn’t. From April 2027, the employer contribution to teachers’ pensions falls from 28.68% to 17.68% of salary, yet every teacher will still retire on exactly what they were promised before, in a form of financial ‘magic’.
In 2022 the notional for the TPS liability was £522bn whereas today it’s a more modest £308.6bn. However, unlike most proper retirement schemes, this scheme has no real assets. The ‘deduction in liability’ over the last four years came via a a single number in a Treasury spreadsheet, a financial conjuring trick that explains everything about how England accounts for public service pensions.
The fund that isn’t
GAD’s report says the scheme has £286.9bn of ‘notional assets’ against £308.6bn of liabilities. That sounds like a scheme funded to 93% of its liabilities; it is in fact funded to 0%. Unlike the pensions of most taxpayers, which are real savings invested in a portfolio, the TPS has no bonds, stocks, gilts or even a real bank account. The ‘assets’ are simply a ledger entry in a spreadsheet. Take the 2020 GAD valuation’s figure, add in contributions paid in by teachers and employers (actually paid by taxpayers), subtract the pensions paid out, then credit a fictional interest at a rate the Treasury chooses. The actual contributions were spent by the Exchequer the day they arrived.
Here is what that ledger did between 2020 and 2024:
| 2020–24 | £bn |
|---|---|
| Contributions in | £36.7 |
| Pensions paid out | £(43.8) |
| Real cash position | £(7.1) |
| Invented ‘investment returns’ | £71.8 |
| Growth in ‘assets’ | £64.7+ |
The scheme paid out £7.1bn more than it took in and yet its notional ‘fund’ grew by £64.7bn! In 2022/23 alone it ‘earned’ £30.2bn, a 12.7% return, because inflation spiked and the formula used credits inflation plus a margin. Of course, there were no assets existed to earn it. So whenever inflation spikes, because of the the SCAPE wizardry (explained below), the TPS as a ‘fund’ automatically reports a spectacular return.
In plain English: these are not assets. They are an accounting convention that lets government report a deficit of £21.7bn instead of an unfunded liability of at least £308.6bn. The real asset behind every teacher’s pension is your current and future tax bill, it’s as simple as that.
SCAPE is simply a dial the Treasury twiddles
The interest rate on that ledger is the SCAPE discount rate. It’s not a market rate; instead it’s set by HM Treasury from the Office of Budget Responsibility (OBR’s) guess at long-run GDP growth. In the last decade it has gone from CPI + 2.8% to 2.4%, down to 1.7% in 2023, and back up to 2.0% in May 2026. The OBR doesn’t have a great track record and has often overestimated productivity growth and economic output, leading to cumulative growth forecast errors exceeding £500 bn between 2010 and 2024.
Each click of the SCAPE dial moves billions, with the GAD’s own sensitivity table showing that a 0.25% cut in the rate would add 5.5% of pay to the employer bill; approximately £1.7bn a year! The May 2026 nudge from 1.7% to 2.0% alone delivered about 6.2 of the 11 points (56.3%) in this year’s notional ‘cut’.
Even GAD’s framework can’t agree with itself, with its ‘core’ cost measure saying the reformed scheme costs 8.0% of pay, below the 10.9% cap. Whereas its ‘economic’ measure says 15.3%, well above it. The system and its rules are written so that neither number triggers action and the model never has to meet any fiscal reality. The notional fund cannot underperform, cannot be audited against a market price, and can’t run out. When the answer or number is inconvenient, the rate changes. That is not actuarial science, it’s a fiscal chimera presented as reality in a government PowerPoint presentation.
TPS’s Notional ‘returns’
| Year | SCAPE margin | CPI used | Formula | GAD’s notional return |
| 2020–21 | 2.4% | 0.5% | 1.024 × 1.005 | 2.9% |
| 2021–22 | 2.4% | 3.1% | 1.024 × 1.031 | 5.6% |
| 2022–23 | 2.4% | 10.1% | 1.024 × 1.101 | 12.7% |
| 2023–24 | 1.7% | 6.7% | 1.017 × 1.067 | 8.5% |
What real money did over the same period
- Defined Benefit (DB) scheme assets: PPF data shows they fell 22% in 2022, after the mini-Budget gilt crash. The TPS is a DB scheme unlike the majority of UK pensions which are Defined Contribution (DC) schemes
- The wider market: institutional assets fell 16% between Dec. 2021 and Dec. 2022.
- Gilts: Long-dated conventional gilts returned −36.7% over the three years to December 2022, and index-linked gilts 27.4%.
Do these arcane accounting rules really matter?
Yes, very much so. A pension liability is the present value of pensions paid decades from now, and the discount rate is the interest rate used to bring them back to today. TPS liabilities are long; 26.1 years for active members and 10.2 years for pensioners, according to GAD. Over periods that long, even very small rate changes compound into massive valuation swings.
| Basis | Discount rate | What happened 2022–24 |
| Private DB schemes | Market gilt yields | Both sides were marked to market. Assets fell 22% in 2022, but PPF says liabilities fell 39%, so funding levels improved. |
| TPS accounts (IAS 19) | Market-linked | Liability fell from £532.3bn in March 2022 to £278.8bn by March 2024. |
| TPS funding valuation (SCAPE) | Fixed by Treasury | Liability rose from £262.0bn to £308.6bn, and the notional fund grew £64.7bn on formula returns. |
Teachers’ real pay
We are repeatedly told that teachers are underpaid, but published salaries leave out the most valuable part of the package, namely pensions.
GAD puts the cost of the pension a teacher builds up each year at 24.9% of salary. Teachers pay in about 9.6% which leaves roughly 15% of salary, every year, paid by the employer and ultimately the taxpayer, on GAD’s own generous assumptions.
A teacher on £45,000 is notionally paid about £52,000 including pension, but even that’s an understatement. Try buying a guaranteed, inflation-proof income for life (e.g. an annuity) on the open market and the price is far, far higher than SCAPE assumes.
The usual defence is that pensions are ‘deferred pay’, so salaries are set lower to compensate. Unfortunately history kills that argument stone dead. The employer contributions rate rose from 16% to 23.68% in 2019 , then to 28.68% in 2024, and now falls to 17.68%. Sound like a saving yet not one pay award moved to offset any of it. A cost that changes by 33% with zero effect on pay is not being traded off against pay, it’s simply hidden on top of it. Compare like with like, salary plus pension, and the ‘underpaid teacher’ story looks very different.
Why the unions are quiet
An 11-point cut in pension contributions and barely a murmur from the teaching unions. That silence is rational. Because nothing their members get has changed. GAD is explicit that benefits are unaffected. Teachers lose nothing.
The cut protects the scheme from employers who want out. The 28.68% rate had already driven many independent schools and some universities out of the TPS. Of the 1,066 independent schools in the scheme in July 2024, only 880 remained by January 2026.
State schools don’t get the money anyway. DfE has already confirmed school funding will be reduced from April 2027 to match the lower pension cost. Government cuts what it charges itself, then cuts what it gives itself. Nobody in the state sector is better or worse off, so nobody has much incentive to complain.
Opacity suits everyone. A lower, notional employer contribution rate makes the teachers’ pension look cheaper. A true market cost would make it look like what it is: one of the most valuable benefits in the British labour market. Unsurprisingly no education union wants that conversation.
Ontario: what a real teachers’ pension fund looks like
Ontario’s teachers have a Defined Benefit pension too. The difference is that theirs has real money to back it up. The Ontario Teachers’ Pension Plan (OTPP) holds C$279.4bn/£78.74bn of real, audited assets. On its January 2026 valuation it was 111% funded, with a C$31.2bn surplus, for the 13th year running. Since 1990 about 80% of its funding has come from investment returns, not taxpayers. Its tools are real ones:
- The discount rate has to be earned. OTPP assumes 2.65% based on real data (actually higher than SCAPE’s CPI + 2.0%) and it is tested every year against an actual portfolio of assets, not a theoretical one from the OBR. In 2025 the fund missed its benchmark by 5%, a C$12bn shortfall, as it states upfront on page one of its most recent annual report.
- Shared risk. Teachers and government each pay roughly 11.5–13.1% of salary, matched one for one. This means if there is a deficit both have to pay more.
- Pensioners share the pain too. Since 2008, inflation protection on newer benefits is conditional on the plan’s health. OTPP calculates that a 26% cut in indexation could absorb a C$24bn hole. The TPS has no such lever, instead pensions rose 10.1% in 2023, guaranteed and paid by taxpayers.
- Independent governance. OTPP has a Board who are answerable to the Ontario Government (Minister of Education) and The Ontario Teachers’ Federation (who may sound like a union but are very different to any in the UK education sector), not to a Treasury hell-bent on massaging notional numbers to make its books look better to an increasingly nervous market.
Ontario isn’t perfect. It is still a guaranteed promise, and the Ontario still underwrites half of any deficit, but when OTPP says its pensions are fully funded, there are assets to prove it.
Australia: stop making promises you can’t fund
Australia took a more blunt route with the advent of their Superannuation pensions scheme (known as Super). New teachers and public servants don’t get a defined benefit; instead employers pay 12% of salary into a superannuation account that the individual member owns and who decides with whom and what to invest it in. The cost to the taxpayer is known, fixed and paid on the day the work is done with nothing left for the next generation of taxpayers to pay for.
It isn’t free of history. Canberra closed its last big civilian DB scheme, the PSS, in 2005, and set up the Future Fund to pre-fund what was left. Even so, the Commonwealth balance sheet still carried a superannuation liability of about A$318bn/£166.17bn at December 2025.
The lesson for Britain? A country that closed its schemes over 20 years ago and has been saving hard will still be paying for them into the next century. By contrast England hasn’t addressed the fundamental problems in the TPS, so hasn’t saved a penny, and is still adding to the bill every school day. On the core question, who pays for a teacher’s pension, Australia’s answer is, the employer, now and in cash.
The bond market will not wait
The TPS is just one UK government pension scheme. The 2020/21 Whole of Government Accounts put the UK’s unfunded public pension obligations at £2.17 trillion, including £792bn for the NHS and £501bn for teachers (now a notionally more modest £308.6bn).
Unsurprisingly the markets are already uneasy. The 30-year gilt yield hit 6.03% on 1 October 2026, the highest since 1998. Britain has paid the highest 10-year borrowing costs in the G7, despite carrying less debt than most of its peers. Investors are charging a UK-specific premium for doubts about fiscal credibility despite Andy Burnham’s claim that, “We’ve got to get beyond this thing of being in hock to the bond markets”.
This is a problem Labour thinks it can tax its way out of, but that escape route is closing with the OBR forecasting the tax burden rising to 38.5% of GDP by 2030/31, the highest since 1948. A government already taxing at post-war highs cannot simply tax its way to growth, nor out of a massive liability that grows with inflation, longevity and every new teacher hired.
My view: the moment bond investors start pricing these off-balance-sheet pension promises the way they price gilts, the UK’s risk premium widens sharply. Britain has been here before. In 1976 a Labour government had to go cap-in-hand to the IMF for the largest loan it had then ever made. The pension bill won’t trigger that moment on its own, but it’s the structural crack the next crisis will probably run through.
It’s everyone’s problem
This isn’t an attack on teachers. They did not design this system and they deserve a secure retirement, but preferably from fully-funded schemes like those seen in Ontario and Australia. It is an attack on the UK government’s sclerotic bureaucratic system, that hides the true cost of public service pensions from the taxpayers funding it. The bill for the TPS will fall to the children now sitting in those teachers’ classrooms as well as every family paying tax as well as every young worker, who will never see a guaranteed pension of their own.
Here is what we should demand:
- Honest accounts. Report public pension liabilities on a market basis, alongside headline debt, every year
- An end to the SCAPE escape. Take the discount rate out of the Treasury’s hands and tie it to something markets can see
- Real money for new promises. Every new pound of pension promised should be backed by a pound actually invested, as Ontario does
- Shared risk. If the cost rises, members and pensioners should share it through contributions and indexation, not leave it all to taxpayers current and future
- Honest pay comparisons. Publish teachers’ total remuneration, salary plus the true value of the pension, so the pay debate starts from all the facts.
The 2027 ‘cut’ is a re-pricing, not a saving. The promises, the cash outflows and the taxpayer’s exposure are exactly what they were. Until we’re willing to say that out loud, the bill will keep growing in the dark.
For any business that sells into the education market this is very relevant as the ongoing TPS fiasco means there will be less money to spend on their products and services. That will in turn impact their bottom line, their enterprise value, their ability to invest (and to get investment) and to pay the salaries and pensions of their own employees.
Sources
- GAD – TPS 2024 Actuarial Valuation: Results
- DfE – Teachers’ pension employer contributions funding adjustment
- DPF – SCAPE rate change to CPI + 2.0%
- GOV.UK – SCAPE rate methodology
- UCU – TPS employer contributions to be decreased
- Independent School Management – Schools leaving TPS
- OTPP 2025 Annual Report
- OTPP – Contribution rate history
- Dept of Finance – Commonwealth Monthly Financial Statements, Dec 2025
- SBS – Future Fund quarantine and PBO projections
- ICAEW – Whole of Government Accounts 2020/21
- ROIC – 30-year gilt yield breaches 6%
- Russell Investments – Gilt market nerves return
- AJ Bell / Alliance News – OBR tax burden forecast
Richard Taylor is a specialist in education sector start ups, edtech, business intelligence, market, company & sector analysis.