More older people is a fact. What it costs is a choice.
An ageing Britain will cost more in pensions regardless — how much more depends on the triple lock and the State Pension age, not just demography.
What would you change?
Adjust pension contributions — and follow the consequences through your own take-home pay, your retirement pot, and (switch the toggle) what happens if 20+ million people make the same change.
Open Pensions Lab →Britain is ageing — and also growing
In the ONS 2022-based projection the population above State Pension age rises from 12.0 million in 2022 to 15.1 million in 2047, an increase of 25.5%. That number is usually quoted on its own, and on its own it is misleading.
12.0m → 15.1m
Population above State Pension age
Not the same as: A forecast. A projection shows what follows if stated assumptions hold.
43.2m → 49.9m
Working-age population
Not the same as: A count of workers. It counts people of working age, employed or not.
278 → 302
People above State Pension age per 1,000 of working age
Not the same as: A bill. It is a demographic pressure indicator, not a cost.
The point most coverage misses. The working-age population is projected to grow too — from 43.2 million to 49.9 million. That is why the dependency ratio rises far more slowly than the pensioner population: from 278 to 302 people above State Pension age per 1,000 people of working age, a rise of under 9% while the pensioner population rises 25.5%.
Two kinds of life expectancy
In plain terms: periodlife expectancy is a snapshot — how long a 65-year-old would live if today’s death rates never changed again. Cohortlife expectancy instead assumes mortality keeps improving over that person’s remaining lifetime, the way it has for decades, so it is always the higher of the two numbers. Government pension-cost projections use cohort figures, because they are projecting spending on people who will live through those future decades of improvement, not through today’s mortality rates frozen in place.
18.7 / 21.2 yrs
Period life expectancy at 65, male / female
Not the same as: How long a 65-year-old will actually live. It applies today's mortality rates for the rest of their life.
19.8 / 22.5 yrs
Cohort life expectancy at 65, male / female
Not the same as: The same measure as the period figure. This one allows for expected future improvement, so it is higher.
Period life expectancy applies today’s mortality rates for the rest of a person’s life. Cohort life expectancy allows for expected future improvement, so it is higher. They answer different questions and must never be spliced into one line. By 2047 the cohort figures are projected to reach 21.8 and 24.4 years.
What a dependency ratio cannot tell you
A ratio of people to people is not a measure of affordability. All of the following matter more, and none of them are in it:
- How many people of working age are actually employed
- How productive they are, and therefore what they earn and pay
- How much tax the economy generates overall
- Whether older people keep working past State Pension age
- How the State Pension is uprated each year
- Private and workplace pension saving
- Health, and how many working-age people are too ill to work
The fiscal effect, and what drives it
State Pension spending is not new pressure appearing from nowhere — it has been rising as a share of the economy for two decades, and OBR’s own stated milestones show the shape of that climb before looking at where it goes next.
State Pension spending, share of GDP — the milestones OBR states
From a narrow 3.3%-3.7% band before the financial crisis, to just over 4% through the recession, to 5% of GDP today — before either projection below is even reached.
View data table
| Period | State Pension spending (% of GDP) |
|---|---|
| To 2007/08 | 3.5 |
| Late 2000s recession | 4 |
| 2024/25 | 5 |
| 2073/74 | 7.7 |
| 2075/76 | 9 |
United Kingdom · Source: OBR, Welfare spending: pensioner benefits · as of Milestones as stated by OBR, not a continuous annual series
The OBR’s July 2026 report projects State Pension spending rising from about 5% to about 9% of GDP by 2075/76.
What does the projected rise actually cost?
Pick a scenario. Each one is a projection the OBR published — not a setting we invented.
9%
of GDP
£261bn
at today’s GDP — illustrative only
+£116bn
versus today, at today’s GDP
The OBR’s baseline: spending reaches around 9% of GDP by the mid-2070s if the triple lock is retained.
OBR baseline projection
What is pushing it, in the OBR’s own decomposition
- Population ageing+1.6pp
- The triple lock+1.6pp
- State Pension age rises−1.0pp
Ageing and the triple lock contribute roughly equally. That is the finding that matters politically: about half the projected pressure comes from a policy choice, not from demography.
How to read the cash figures.They apply each scenario’s published share of GDP to today’s GDP of roughly £2900bn, so a percentage means something familiar. They are not forecasts: GDP in the 2070s will not be today’s, and the OBR does not publish a cash figure for these projections. The percentages are the OBR’s; the pounds are our arithmetic.
What is pushing it up — and what is holding it down
From the OBR’s July 2025 report, which contains the fuller decomposition. State Pension spending rose from 5.0% of GDP in 2024/25 to 7.7% in 2073/74 in that report’s central projection.
- Ageing+1.6 percentage points of GDP
- The triple lock+1.6 percentage points of GDP
- Planned and assumed State Pension age risesabout −1 percentage point of GDP
“Percentage points of GDP” here means share of the whole UK economy’s output: +1.6 percentage points means State Pension spending ends up 1.6% of GDP higher in 2073/74 than it would be if that one driver alone were acting. Ageing and the triple lock add the same amount — which is the finding that matters politically: a policy choice (how the pension is uprated) adds as much to future spending as demographic ageing does.
These three drivers, as the OBR states them, add up to +2.2 percentage points (1.6 + 1.6 − 1.0) — not the full +2.7 percentage points of headline change (5.0% of GDP in 2024/25 to 7.7% in 2073/74). The remaining roughly 0.5 percentage points is attributed by the OBR to other, smaller factors not broken out individually in its published decomposition.
What the triple lock itself costs
The OBR runs the same projection with one change: the State Pension uprated by average earnings instead of the triple lock. Spending then reaches around 7% of GDP rather than around 9%.
That is the cleanest available measure of the policy’s cost, because everything else is held constant. Roughly two percentage points of GDP — about half the projected rise — comes from the uprating rule, not from ageing. Whether that is money well spent is a political question; that it is a choice rather than a demographic inevitability is not.
Two reports, not one scenario. These are two different OBR reports with different horizons and assumptions. The 9%-by-2075 figure and the 7.7%-by-2073 decomposition are not the same scenario and must not be quoted as one.
What the triple lock has actually paid, year by year
The projections above are about the future. This is the record: the actual weekly rate, and how sharply it has moved some years and not others.
Full new State Pension, weekly rate — cash terms
The rate rose from £159.55 in 2017/18 to £241.30 in 2026/27 — but not smoothly: the two sharpest jumps, +10.1% and +8.5%, both happened in successive years (2023/24 and 2024/25), driven by the CPI and earnings legs of the triple lock respectively, not the 2.5% floor.
View data table
| Period | Weekly rate (£, cash terms) |
|---|---|
| 2017/18 | 159.55 |
| 2018/19 | 164.35 |
| 2019/20 | 168.6 |
| 2020/21 | 175.2 |
| 2021/22 | 179.6 |
| 2022/23 | 185.15 |
| 2023/24 | 203.85 |
| 2024/25 | 221.2 |
| 2025/26 | 230.25 |
| 2026/27 | 241.3 |
UK · Source: DWP, Benefit and pension rates (annual series) · as of 2017/18 to 2026/27
The National Insurance Fund: pay-as-you-go, not a pot
“National Insurance Fund” sounds like personal savings set aside for your pension. It is not. It is a single account that collects this year’s National Insurance contributions and pays out this year’s contributory benefits — overwhelmingly the State Pension. When contributions fall short of benefit payments in a given year, the Fund’s balance simply falls.
National Insurance Fund: receipts against payments
2024 ran a surplus of £13.9bn; 2025 ran a deficit of £7.1bn. NICs revenue reduced compared with the same period the year before, mainly due to a reduction in some employee rates. Benefits were also uprated, increasing benefit payments.
View data table
| Period | Total receipts | Total payments |
|---|---|---|
| Year to 2024 | 145.5 | 131.6 |
| Year to 2025 | 138.3 | 145.3 |
Great Britain · Source: HMRC, Great Britain National Insurance Fund Account for the year ended 31 March 2025 (HC 1360) · as of 19 November 2025
£79.3bn
Fund balance
Not the same as: Money saved up to meet future pension promises. It is a short-term working balance, targeted at a fraction of one year's spending.
HMRC, Great Britain National Insurance Fund Account for the year ended 31 March 2025 (HC 1360)
16.7%
GAD's minimum working-balance target
Not the same as: A fully-funded reserve. £24.2bn — about one-sixth of a single year's benefit expenditure, not years of it.
£55.1bn
Above the minimum target
Not the same as: A comfortable cushion for the long term. A £7.4bn contingency Treasury Grant facility was still authorised for 2025/26 in case one becomes necessary.
Government Actuary's Department targets a working balance of at least one-sixth (16.7%) of that year's projected benefit expenditure — a short-term liquidity cushion, not a savings target sized to future pension promises. The Fund can be topped up by a Treasury Grant (from general taxation) if it falls short; none was needed in 2024/25, but HM Treasury authorised a contingency facility of up to £7.4bn (5% of estimated benefit payments) for 2025/26 in case one becomes necessary.
Could UK pension funds invest more at home?
This page is about what the State Pension costs. A separate question — raised by readers of this page — is what UK private pension funds do with the money they hold, and whether more of it could be invested in UK assets rather than overseas. Restored here in brief, with the fuller build-out on the SWOT page:
Of the £207.7bn in assets that 25 of 28 UK master trusts disclosed for their 2026 asset-allocation report, only 15.5% sits in UK investment of any kind (equities, bonds, infrastructure combined) against 84.5% overseas (The Pensions Regulator (TPR), “Master trust asset allocation 2026”). The voluntary 2023 Mansion House Compact asked funds to invest 5% in unlisted UK assets specifically.
Master trusts only — this is one part of the UK DC pension market, which is itself one part of the wider UK pension system alongside defined-benefit schemes. A separate, older and more widely-cited claim (UK equity holdings falling from roughly 53% in the 1990s to around 4% by 2023) measures something narrower and is not independently re-verified by this site; the two figures point the same direction but should not be read as confirming each other.
Full context, including why UK and non-UK master trust data cannot be combined into one whole-market figure: the SWOT page’s pension fund reform item. For the fuller picture of every funded pension type — private-sector defined benefit, master trusts and the Local Government Pension Scheme — and the unfunded public-service schemes that hold no fund at all, see who holds Britain’s retirement savings →.
The other side of the story: automatic enrolment
Everything above is about the State Pension, funded pay-as-you-go. Alongside it, most employees now build a second, genuinely invested pot — the direct result of a specific policy that worked roughly as intended.
Workplace pension participation among eligible employees
Participation among eligible employees rose from 47% just before automatic enrolment began in October 2012 to 90% by 2025 — 11.4m eligible jobholders automatically enrolled cumulatively as of February 2026.
View data table
| Year | Participation rate (%) |
|---|---|
| 2012 | 47 |
| 2014 | 65 |
| 2016 | 74 |
| 2018 | 84 |
| 2021 | 88 |
| 2024 | 89 |
| 2025 | 90 |
UK · Source: DWP, Workplace pension participation and savings trends of employees, 2009 to 2025 · as of 2012 to 2025
A genuine policy success by this measure — but participation is not the same as adequacy: it says someone is saving something, not that their contribution rate or pot size will be enough to live on, and the rate varies sharply by employer size (as low as 55% at the smallest employers, over 90% at mid-sized ones). See the master-trust default-fund detail on the funded-pensions page for what these contributions are invested in.
What these contributions are invested in — and how much of it stays in the UK — is covered alongside the other funded pension types: defined contribution & master trusts →
So are pensioners actually doing OK?
Everything above is about what the system costs. A separate, equally important question is what it delivers — whether pensioners themselves are managing financially.
14%
Pensioners in relative low income, after housing costs
Not the same as: Statistically significant change from the year before — DWP states the 2pp rise is NOT statistically significant, from an estimated 12% (derived by subtraction, not separately published).
16%
Pensioners in relative low income, before housing costs
Not the same as: The after-housing-costs figure alongside it. Pensioners are disproportionately owner-occupiers with paid-off mortgages, which is part of why the two figures differ.
DWP attributes the (not statistically significant) rise to pensioners in the bottom third of the income distribution seeing their incomes grow more slowly than the median, alongside the removal of Cost of Living Payments and restrictions on Winter Fuel Payment eligibility, even as State Pension upratings provided some offsetting support.
Two secondary sources this page checked while researching this figure quoted a substantially higher (~19%) pensioner poverty rate that could not be reconciled against DWP’s own primary release — not used here. The figures above are read directly from the primary HBAI publication.
Four things this page will not say
“There will only be X workers for every pensioner”
The ratio counts people of working AGE, not workers. Employment rates are a separate question.
“Life expectancy is falling”
Meaningless without stating sex, geography, period, and whether it is period or cohort life expectancy.
“The pension system is bankrupt”
The State Pension is a policy commitment funded largely pay-as-you-go. It is not a fund with a balance that can run out.
“Immigration solves ageing” / “makes no difference”
Migration changes the age structure and tax receipts, but migrants also age. The result depends on sustained flows, earnings and family composition.
The pensions vertical
This page and the State Pension page cover the ageing/funding story in depth. Eight further pages build out the rest of the picture — per the revived Pensions Commission, the definitive account of where the answers are still being worked out.
- Where the money is investedUK equities down from 48% to 4.9% of DB schemes' holdings. Growth vs returns, both sides.
- Is there really a "pensions crisis"?Five specific claims, tested one at a time — some hold up, some don't.
- Workplace pensionsThe 8% contribution problem, the DB→DC shift, and 13m small pots.
- Inequalities in pension wealthThe top decile holds 64%. The gender gap is 48%. The self-employed were left out.
- Pension tax relief£59bn a year, how EET works, and the 2029 salary-sacrifice reform.
- Proper international comparisonsWhy "the UK state pension" alone is the wrong thing to compare.
- What future reforms could look likeA revived Pensions Commission, and five options on the table.
- Pension mythsThe pot that doesn't exist, and six more misconceptions, tested.
Next: the State Pension itself — what it pays and what it costs, and how it compares internationally. Funded pensions: who holds Britain’s retirement savings — DB, DC and LGPS assets, and the unfunded public-service schemes. The redirection question: the SWOT page’s pension fund reform item and the revenue-opportunity page’s pension war chest. Related: the benefits system and public finances.