Each year the pension jumps to the highest of three measures and never falls back. Over time it rises faster than any single measure: in a year when inflation spikes it follows prices; the next year when wages catch up it follows earnings. Each rise then becomes the base for the next one, so the gains compound.

Earnings 6 · CPI 5 · 2.5% floor 4.

Floor years: 2013, 2015, 2017, 2021.

2026/27: £241.30 × 52 = £12,547.60

2027/28: £250.71 × 52 = £13,036.92

Increase = £489.32 a year per full pensioner.

Approximate: 8.5 − 6.7 = 1.8 percentage points

Precise: (1.085 ÷ 1.067 − 1) × 100 = 1.7%

Caveat: this compares the rise with the inflation reference figure from September 2023. Prices over the year the higher pension is actually paid (April 2024 to March 2025) could rise by more or less, so this is an estimate of the real increase, not an exact measure.

(a) Pension: (241.30 − 185.15) ÷ 185.15 × 100 = 30.3%

(b) CPI reference figures: 1.101 × 1.067 × 1.017 × 1.038 = 1.240 → 24.0%
Earnings reference figures: 1.055 × 1.085 × 1.041 × 1.048 = 1.249 → 24.9%

(c) Uprating took the highest figure each year, so the pension grew more than 5 percentage points faster than either measure compounded on its own: the ratchet in action. It is approximate because the reference figures are 12-month snapshots measured in the year before each April rise, so they do not line up exactly with April 2022 to April 2026.

2024: CPI 6.7% → £203.85 × 1.067 = £217.51

2025: CPI 1.7%, so floor 2.5% → £217.51 × 1.025 = £222.95

2026: CPI 3.8% → £222.95 × 1.038 = £231.42

Actual: £241.30. Difference ≈ £9.88 a week ≈ £514 a year per full pensioner.

(15.5 − 5.2) ÷ 5.2 × 100 = 198%, roughly three times the original estimate.

(a) £15.5bn ÷ 0.005 = £3,100bn (£3.1 trillion)

(b) 7.7 − 5 = 2.7 percentage points; 2.7 ÷ 5 × 100 = 54% increase

(c) 1.6 ÷ 2.7 × 100 = 59%: the triple lock causes more of the rise than the ageing population.

It suggests the triple lock is not the main reason pensioner poverty fell. Targeted support (Pension Credit from 2003), more private and occupational pensions, and higher home ownership did most of the work. The triple lock raises the pension for every pensioner, rich or poor, so it is an expensive way to reduce poverty. (Note: on some measures, such as before housing costs, pensioner poverty kept falling after 2011, so check which measure a source uses.)

To 2030: pension spending stays on its current, rising path. Keeping the pledge does not add new cost to the existing forecasts, because the triple lock is already built into them. The cost is an opportunity cost: compared with reforming now, spending is about £15.5bn a year higher by 2029-30 than an earnings link and £22.9bn higher than CPI only (Table 2). Borrowing and debt interest are therefore higher than under an earlier reform, which leaves less headroom against the fiscal rules. None of the government's estimated £15bn saving arrives before 2030.

After 2030: pension spending should grow more slowly, but by how much is uncertain.

• It is still a ratchet, just a weaker one. The higher of CPI or 2.5% is a double lock: whenever CPI is below 2.5% the pension still rises faster than prices, and those gains become the base for the next year. Removing the earnings leg slows the ratchet; it does not break it.
• The earnings promise limits the savings. If the pension must hold its value relative to earnings over time, it cannot fall far behind wages for long. Some of the gap may have to be closed later, so the OBR's 1.6 pp of GDP is reduced, not removed. How much depends on how the promise is delivered, which is not yet clear.

Pension savings ≠ a better overall fiscal position. Lower pension spending than under current policy is not the same as lower borrowing. The savings are earmarked for a National Care Service. If the care service costs about the same as the saving, overall spending and borrowing barely change; if it costs more, or starts before the savings build up, borrowing could rise. A care service could also ease pressure elsewhere (for example NHS beds blocked by delayed discharges), but that is hard to quantify.

Opportunity cost and trade-offs. Every pound kept in the triple lock to 2030 is a pound not spent elsewhere, or a pound of extra tax or borrowing.

• Between generations: pensioner poverty (16%) is now about half the child poverty rate (31%), and the typical pensioner's real income is forecast to rise by about 5% to 2029-30, against 1% for non-pensioners and a fall for the poorer half of them (Table 3). The triple lock is paid for largely by taxes on working-age people whose living standards are barely rising. The same money could go towards child poverty, the NHS or lower taxes on workers.
• It is untargeted: every pensioner gets the same percentage rise, however well-off. Most of the fall in pensioner poverty came before 2011, through targeted support such as Pension Credit, which reaches the poorest pensioners at a much lower cost.
• The care argument: the Prime Minister argues that social care has been neglected and that care costs push pensioners into poverty. If so, moving money from across-the-board uprating towards care targets pensioners in real hardship, and could ease NHS pressure from delayed discharges. However, care costs mainly eat into wealth (savings and homes) rather than income, so a care service may do less for measured income poverty than claimed, and some of the gain would go to better-off homeowners protecting their inheritances.
• Delay has a cost too: honouring the pledge means these trade-offs are not addressed until after 2030, while child poverty and care pressures continue.

Other evaluation:

• Scale: 0.5% of a £3.1 trillion economy is significant but not a crisis on its own.
• Wages vs prices: the cost of keeping the earnings leg arises only when earnings growth beats CPI and 2.5%. Forecasts are unreliable: the OBR's estimate overshot by 198% (Q6).
• Some money comes back: with the personal allowance frozen, more pension income is taxed (fiscal drag; the full pension of £13,036.92 in 2027/28 exceeds £12,570), and pensioners' spending generates VAT.
• Ageing still adds 1.1 pp of GDP whatever the uprating rule.
• Credibility: a pre-announced change is more likely to stick than breaking a pledge, but it must survive the next election.

Judgement: honouring the pledge keeps pension spending on its existing path to 2030, at a real opportunity cost when pensioners are now less likely to be poor than children and their incomes are forecast to grow fastest. The reform should then slow pension spending, but the size of the saving depends on how the earnings promise is delivered. Whether the overall fiscal position improves depends on how much of that saving the National Care Service absorbs. The government is choosing to switch spending between priorities, not to cut borrowing.