Closing public sector DB pensions to new entrants from 2030: what’s actually proposed

What this article covers

• Does: Set out the policy proposal to close the main unfunded public sector defined-benefit schemes to new entrants from 2030, the cost and savings figures behind it, what it would mean for someone starting a public sector career in 2030, and the main lines of criticism.
• Doesn’t: Take a side. Pension Plain explains; it doesn’t recommend, endorse, or oppose. The trade-offs are real on both sides; this article presents them, not a verdict.
• If you need advice: Speak to a regulated financial adviser, or contact MoneyHelper for free, government-backed guidance.

The UK’s biggest public sector pension schemes have no investment pot behind them. The NHS, Teachers’, Civil Service, Armed Forces, Police and Firefighters’ schemes pay pensions out of current tax receipts, backed by a Treasury guarantee. By standard public accounting measures, that accumulated obligation runs to around £1.4 trillion. The kind of number that reliably finds its way into political speeches.

In January 2026 the think-tank Policy Exchange published a paper, Public Sector Pension Reform, calling for the unfunded defined-benefit (DB) schemes to close to new entrants from 2030, with new public sector workers joining defined-contribution (DC) arrangements instead. The proposal has since been adopted as policy by Reform UK and is debated in the trade press, in unions, and across other think tanks. What follows is what the proposal actually says, what the numbers behind it claim, and what it would mean for someone starting a public sector career from 2030 onwards.

What’s actually proposed

The Policy Exchange paper calls for the six main unfunded public service schemes (NHS, Teachers’, Civil Service, Armed Forces, Police, and Firefighters) to close to new entrants from 2030. Anyone joining those schemes after that date would instead be enrolled into a defined-contribution arrangement, with employer contributions set at a market rate rather than the higher rates currently paid into the DB schemes.

What the proposal leaves untouched is just as significant. Existing scheme members (roughly six million public sector workers) would keep their current DB terms. Benefits already built up, and future accrual for people already in the scheme, are explicitly outside the scope of the reform. That’s the same “grandfathering” approach private employers used when they closed their own DB schemes to new entrants over the past thirty years. It sidesteps the legal and political difficulty of changing the deal for people already in it.

The framing argument used by proponents is comparative. Private employers largely moved away from DB in the 1990s and 2000s, and the gap between public and private sector retirement provision has widened substantially since. Policy Exchange argues that gap is economically unsustainable and hard to justify to the private-sector taxpayers ultimately backing the guarantee. Critics dispute both the comparison (private sector DC contribution rates are typically much lower than public sector DB employer rates, so “the same as private sector” can mean a sharp pay cut) and the assumption that private-sector practice is a sound benchmark for public service workforce design.

The numbers

The £1.4 trillion unfunded liability figure comes from standard public accounts methodology, specifically the present-value cost of pension promises already made. It appears in the government’s own Whole of Government Accounts and isn’t disputed as a number. What is contested is what it means. How alarming it sounds depends heavily on the discount rate applied, and critics of the framing point out that the obligation falls due across many decades, will be met from future tax revenues, and is partly offset by the economic value public workers produce. Policy Exchange treats the scale of the liability as, in itself, a reason for structural change.

The paper’s central cost and savings estimates:

• Short term: additional costs of £1.1bn a year in the early years of transition. Employer contributions on new DC members, plus transitional and administrative costs, before any meaningful savings from the reduced DB obligation have accrued.
• After 20 years: savings of £6.1bn a year. As the DC cohort grows and fewer people are building up DB entitlement, the annual cost gap starts to close in favour of the new structure.
• After 50 years: savings of around £40bn a year. By this point the existing DB membership has largely retired, and the ongoing cost of the DC arrangements is substantially lower than the DB alternative would have been.

That pattern (costs first, savings later) is the defining feature of any DB-to-DC transition. Closing a scheme to new entrants doesn’t cancel the existing liability. It keeps paying out for decades. And the employer still has to contribute to the new DC scheme for everyone hired after the cut-off date. The savings are real, but they’re a multi-decade bet.

If it happened: what would change for a new starter in 2030

Picture a nurse, a secondary school teacher, and a mid-grade civil servant all starting their careers in late 2030. Under current rules, each of them joins a defined-benefit scheme: a guaranteed pension calculated by formula, paid for life, with no investment pot to manage. Under the proposal, they’d join a defined-contribution scheme instead.

Under DB, each year of pensionable service generates a guaranteed slice of annual pension. In the NHS 2015 Scheme that’s 1/54 of that year’s pensionable pay, revalued annually while you’re an active member at CPI + 1.5%. A nurse who works 30 years, with pay rising from around £33,000 to £50,000, would typically retire on something in the £18,000 to £22,000 a year range, index-linked for life, with no investment risk on her side of the equation.

Under DC, both employer and employee contribute a percentage of salary into a personal pot. What that pot is worth at retirement depends on how the investments performed and what charges were applied. The NHS currently puts in 23.7% of pensionable pay on behalf of each DB member. DC auto-enrolment employer contributions in the private sector typically run between 3% and 8%, considerably less, although Policy Exchange suggests a more competitive rate for public sector DC, without pinning down a number. The retirement income isn’t guaranteed; it depends on markets, charges, and how the member draws the money down. The risk transfers from the taxpayer to the individual.

The teacher and civil servant face the same structural shift, if not identical numbers. The Teachers’ Pension Scheme 2015 accrues at 1/57 of career-average pay; the Civil Service Alpha scheme credits 2.32% of pensionable earnings per year. Across all three, a 30-year DB career produces a pension that’s modest rather than lavish, but one you know in advance. That certainty is what DC doesn’t offer.

The criticisms

The sharpest challenge to the proposal is about framing. UNISON, representing around 1.3 million public service workers, disputes the “gold-plated” label directly: more than half of NHS pensioners receive just £5,000 a year. The typical NHS pensioner isn’t a consultant or a senior manager. Most are nurses, healthcare assistants, porters, and administrative workers, people whose pensions reflect careers at moderate pay. Five thousand pounds a year is not a gold-plated retirement.

The DB structure does carry genuine value: the guarantee removes investment and longevity risk from the individual entirely. That’s worth something real, even when the resulting pension is small. Proponents of the reform acknowledge this: their argument isn’t that the pensions are huge, but that the structure (and the open-ended taxpayer guarantee behind it) is what’s unsustainable. Critics respond that stripping that structure from new entrants doesn’t reduce the existing liability; it just ensures future workers take on the risk instead.

The gender argument is specific. Women make up the majority of NHS workers, teachers, and local government staff. Public sector employment is one of the few parts of the labour market where women’s retirement outcomes come close to men’s, largely because DB doesn’t amplify the career-break and investment disadvantages that accumulate in DC pots over decades. UNISON’s Gloria Mills warned the reforms would “further disadvantage women in regard to employment participation, and beyond that, to entitlement to a decent retirement income.” A future generation of nurses and teaching assistants on DC would see their retirement income track market performance, and see any gaps from maternity leave or part-time working carry through into the final pot.

Recruitment and retention is the third line of critique. The pension is a material part of public sector total compensation, particularly in roles (nursing, teaching, social work) where base salaries aren’t high. If DC employer contributions came in lower than the rates currently paid into DB, without a salary offset, the overall package falls. Proponents say DC contributions could be set competitively enough to hold the line; critics say fiscal pressure would reliably push the rate down in practice, amounting to a long-run pay cut for anyone hired after 2030.

Then there’s the transition cost itself. Policy Exchange puts the initial additional outlay at £1.1bn a year. Depending on the DC contribution rate chosen and how quickly the DB schemes run down, the period before the reform breaks even could stretch well beyond a decade. Any government already managing a tight fiscal position would need to weigh that upfront cost against savings that fall mostly in budget cycles far in the future.

Where the LGPS sits

The Local Government Pension Scheme is different from the six unfunded schemes in one key respect: it has money in it. Around £402 billion in assets sits across the funds in England and Wales (March 2025 figures), with tens of billions more in Scotland and Northern Ireland. Those assets are invested in equities, bonds, property and infrastructure. At the 2022 triennial valuation the scheme was in aggregate surplus, funded at 107% of liabilities.

That funded status changes the argument materially. For the unfunded schemes, closing to new entrants eventually reduces the open-ended call on the Treasury. For the LGPS (already in surplus and backed by real assets) the case for closure looks different. The existing assets cover the existing obligations, and the question would be whether future accrual is affordable on its own terms, not whether the taxpayer is carrying an unfunded promise. Policy Exchange’s paper is focused on the unfunded schemes. The LGPS is also administered by individual local authorities, not central government, which means any reform to it would need separate legislation and separate negotiation.

For LGPS members, the practical position is that the current proposal does not directly target your scheme. That could shift as any detailed legislation is drafted, but the structural and fiscal arguments are different enough from those for the unfunded schemes that it isn’t a straightforward extension. For a full account of how the LGPS works, see the LGPS guide.

The political path

This is a published proposal, not a bill before Parliament. To become law it would need a government willing to legislate it, a parliamentary majority for the bill, and the capacity to work through considerable administrative and legal complexity. The proposal has been adopted as policy by Reform UK; other parties have not made an equivalent commitment, and an actual legislative draft has not been produced.

The next general election can’t come before 2029. Any version of this policy taking effect before the proposed 2030 start date is implausible on those timelines alone. A formal commitment by any party in 2026 would still be an announcement, not implementation.

History also suggests the process would be slow even if a sympathetic government arrived. The last major public service pension reform (the 2011 changes that eventually produced the current CARE schemes) took years of negotiation, produced a judicial challenge (McCloud), and generated a remedy process that was still running in 2026. A reform of this scale would face comparable complexity, and probably comparable litigation.

For current members and people starting public sector careers now, nothing has changed. The proposal matters as a marker of where the policy debate is heading, something worth tracking over a 30 or 40-year pension horizon. But as of today, the schemes are operating exactly as they were before the paper was published.

Common questions

Pension Plain’s take

This is a proposal, not a policy in flight. The most useful framing is to separate three things that often get blurred in headlines: the size of the unfunded liability, the case for changing the structure, and the case for changing it on this particular timeline.

The £1.4 trillion liability is real, in the sense that the public accounts say it’s there, but how worrying you find it depends on the discount rate you use and how you weigh future tax receipts against future pension promises. The case for some form of structural change has been argued for years across the political spectrum; the case for this specific change (closing the unfunded schemes to new entrants from 2030) rests on Policy Exchange’s modelling of long-run savings versus near-term costs. The savings are decades out. The costs are immediate. Whether that trade is worth making is a values call as much as a financial one, and reasonable people disagree.

For members and prospective members, the practical answer is: nothing has changed today. The schemes are operating as they were a year ago. If this becomes a serious legislative proposal, it will move through years of consultation, negotiation, and probably litigation before it lands. By that point a clearer picture of the contribution rates, transitional protections, and offsetting measures (if any) will exist, and decisions about whether the package is competitive can be made on facts rather than projections.

Pension Plain will track this as it develops. For now, the practical work for current members is to understand what they already have. The scheme guides linked at the foot of this article do that.

Key official sources

• Public Sector Pension Reform (Policy Exchange, January 2026). The think-tank paper underlying the proposal.
• Pension reforms could widen gender gap (UNISON, 5 May 2026). Trade union response.
• Pension Plain: UK public sector pensions explained (cross-cutting context).
• Pension Plain: NHS Pension Scheme · Teachers’ Pension Scheme · Civil Service Pension Scheme · LGPS.
• MoneyHelper (free, impartial pension guidance).

Fact-checked 6 May 2026.