What a ratchet!

The State Pension triple lock has delivered real gains for pensioners since its introduction in 2012. But the case for boosting pensioner incomes over and above others has now run out, while the case for the triple lock as a way to deliver such a boost was always poor.

Pensioners have seen three times as much living standards growth as non-pensioners over the last two decades, a typical pensioner household now has the same level of income as the typical working-age household, and pensioners are less likely to be in poverty than the rest of the population. There is not a strong case for continuing to increase state support for pensioners faster than the wages of typical workers – the inevitable practical effect of the triple lock. Even if you disagree, the triple lock is also a bad way to increase the relative value of the State Pension. Because it uprates every year by the highest of earnings, prices, or 2.5 per cent, the value of the State Pension depends not just on the level of earnings and inflation but also on how volatile they are. The OBR projects that State Pension spending will rise by a further £80 billion over the next 50 years, but warns it could easily be £40 billion higher or lower depending on economic conditions.

The New State Pension is already at around 30 per cent of median full-time earnings,  close to the level recommended by the original Pensions Commission in 2005. The current Pensions Commission should now give its definitive view on the appropriate level. Our view is that level has now been reached, and politicians should find the courage to replace the triple lock with a policy that is transparent, predictable, and fair across generations. The best option would be a ‘smoothed’ earnings link, which tracks earnings growth but protects the value of the State Pension from temporary price shocks. The triple lock means the State Pension bill is now £12.6 billion higher than it would have been under a smoothed earnings link since 2012, with overall spending around £9 billion higher when we take account of higher income tax receipts and lower means-tested benefit spending. Switching to such a policy from next year would be a net saving of around £650 million in 2029-30 that would continue to grow over time.

Former statesmen from Tony Blair to Jeremy Hunt have called time on the triple lock. The polite way to do this is to say that the triple lock was a good policy for a period but it is no longer justified. But the truth is that the triple lock was always a poorly designed, unfair, arbitrary ratchet that we could never afford.

The triple lock is the mechanism by which the State Pension has been uprated annually since 2012.[1] It sets the default uprating each year as the highest of 2.5 per cent, the growth in prices (inflation), and the growth in average earnings. In general, in a stable economy behaving as economists expect, that would mean the State Pension rising each year in line with earnings. In the real world, though, the triple lock means the State Pension rises by more than average earnings in the long run. This is because it will rise by more than earnings in any unusual year in which prices grow faster than earnings, or earnings grow by less than 2.5 per cent, and this outpacing of earnings growth is never unwound, creating the so-called ‘ratchet’ effect. And the UK’s economy has been particularly poorly behaved recently, with eight ‘unusual’ years since the triple lock’s introduction 15 years ago. This matters: the current value of the Basic State Pension (BSP) is already 11 per cent higher than it would have been with earnings-linked uprating since 2012.

There are, therefore, two questions when we think about the pros and cons of the triple lock. First, and most importantly, should the State Pension rise in the medium term by more than average earnings? Second, is the triple lock a good way to achieve that?

Continuing to prioritise the incomes of pensioners over working-age adults and children would be unfair

At the current juncture, it is simply unfair to continue with a policy of the State Pension rising by more than average earnings. Pensioners have seen three times as much living standards growth over the last twenty years as non-pensioners. When the first Pensions Commission was in place, the typical pensioner had a lower income than the typical non-pensioner, but that has not been the case since 2011-12, as Figure 1 shows. And pensioners are now less likely to be in poverty than the rest of the population, and half as likely as children. On what basis, then, should pensioners see their support from the government rise faster than the wages of the typical worker in the economy?

In the polite version of this tale, the triple lock is credited with this progress on pensioner living standards (even if an increasing number of politicians are now prepared to admit that the generous uprating cannot last forever). But this risks over-claiming. In fact, much of the progress in pensioner incomes happened in the early 2000s before the triple lock’s introduction: typical pensioner incomes had caught up with non-pensioner incomes by the time the triple lock was introduced in 2012 (see Figure 1).

For pensioner poverty, the more important policy change was the introduction of the Minimum Income Guarantee, which became Pension Credit in 2003, along with a commitment to index its value to earnings growth. This created a significantly more generous safety net for pensioners: in real terms, the income floor for a single pensioner aged under 75 was £139 per week in 1997, but by 2011 it was £208 per week (both in 2026-27 prices). The result was a 15.8 percentage point fall in pensioner relative poverty between 1997-98 and 2011-12, as growing incomes for poor pensioners closed the gap with the non-pensioner median. However, since 2011-12, pensioner poverty has risen by 2.3 percentage points even with the triple lock in place, driven by stronger income growth at the median than for poorer pensioners, and by a growing share of pensioners renting privately. The triple lock, then, can take no credit for the dramatic fall in pensioner poverty in the 2000s, and the record since highlights its limitations in keeping pensioner poverty flat.

Figure 1: Pensioner incomes had caught up with non-pensioner incomes by the time the triple lock was introduced

So, the case for favouring pensioners over workers was not that strong even at the triple lock’s inception. Since then, the triple lock has hugely favoured pensioners over working-age benefit recipients, with the State Pension growing at twice the rate of working-age unemployment benefits.[2]

The triple lock is fiscally unsustainable

As well as being unfair, it is also not fiscally sustainable for the State Pension to rise forever by more than the earnings of a typical worker. Those earnings determine a big chunk of tax revenue, so accelerating support from the state at a faster rate than the revenue it receives would require a continual squeeze on all other elements of public spending (or exploding public debt). What makes the fiscal arithmetic worse is that we have an ageing population, with the number of people aged 16-64 for every person aged 65+ projected to fall from 3.3 to 1.9 between 2025 and 2075.[3] We are currently spending 5 per cent of the size of the economy (or £154 billion or 11 per cent of government spending), on the State Pension. That’s 23 per cent more as a share of the economy than we spent in 1978-79, despite rises in the State Pension age in that time, and the real-terms cost is set to grow by a further 50 per cent, or £80 billion, over the next 50 years.[4] Even in the short term, just under half of the projected real terms rise in the welfare bill by 2029-30, an issue that exercises many politicians, will come from increased spending on the State Pension (see Figure 2).

Figure 2: State Pension spending accounts for nearly half of the projected increase in social security spending by 2029-30

The arbitrariness of the triple lock is bad policy

But the thing not said enough in polite company is that the triple lock is a terrible policy, even for those wanting to increase the State Pension faster than earnings. The ratchet built into the triple lock means that the value of the State Pension ends up depending not just on the level of inflation and how fast earnings are growing, but also on how volatile they are. Crucially, the ratchet pushes up the State Pension precisely when economic volatility is highest, such as during downturns (marked by low prices and weak wages), or during bouts of cost-push inflation (when inflation typically exceeds earnings growth), thereby increasing the riskiness of the public finances. It is impossible to justify why our generosity to the older population should be a function of economic volatility.

This uncertainty compounds the policy’s unaffordability: of the £80 billion increase in cost that the OBR expects over the next 50 years, it warns it could easily be £40 billion higher or lower depending on the relative performance of prices and earnings. And it is not a policy which anyone can truly base retirement plans on with such a level of uncertainty. The number of people undersaving for retirement might be 13 million or might be 19 million, a gap that exists purely because of the triple lock’s sensitivity to future economic conditions.

If politicians believe the State Pension should rise faster than earnings, then they should say so, and set a predictable policy path with that goal in mind. Uprating by, say, “average earnings plus 0.5 per cent” would be transparent, and its cost more predictable.

We are likely already at a plausible limit for how high the State Pension should go as a proportion of average earnings

If it is not sustainable for the State Pension to rise forever above earnings, then the question becomes “how high should we go?”. When to stop the triple lock is the fundamental debate that we need to have ahead of the next election, and the Pensions Commission should be bold, and say what it thinks is a sustainable answer to this question.

There is, of course, no right answer to this (just as there is no science behind the level at which most working-age benefits are paid), but in a world of competing priorities, our view is that we have reached a plausible limit already. The New State Pension is now 35 per cent of average earnings, or 30 per cent of median full-time earnings. As the interim report of the new Pensions Commission points out, the New State Pension “is now providing the firm foundation for retirement incomes as envisaged by the first Pensions Commission”, which recommended that the State Pension should reach 31 per cent of median full-time earnings. The current proportion is also close to the third that Steve Webb, the pensions minister who introduced the triple lock, has suggested might be a good aim.

A ‘smoothed’ earnings link would be the best replacement for the triple lock

If we have reached a reasonable limit for the value of the State Pension in relation to earnings, then the best replacement for the triple lock is a ‘smoothed’ earnings link. Under such a policy, in years where price inflation exceeds earnings growth, then the State Pension would rise in line with prices temporarily, protecting its real value. However, when growth in earnings rises back above the level of price inflation, earnings growth would not immediately be used to uprate the State Pension (as would happen under a ‘double lock’ policy); it would instead be uprated in line with price inflation until its value has returned to the same fraction of average earnings as it had before the real decline in average earnings. Under current economic forecasts, implementing a smoothed earnings link from 2027-28 onwards would leave the value of the BSP at £199.65 per week in 2029-30, compared to £200.95 with triple lock uprating. This would save around £650 million in 2029-30 (net of increased spending on means-tested benefits and lower income tax receipts) compared with maintaining the triple lock, a sum which would continue to grow over time, and which could even be higher in the short term if we went through a period of greater than expected earnings or inflation volatility.[5]

For those with less political courage, an alternative might be a ‘smoothed’ triple lock. This would guarantee that the State Pension will rise by at least 2.5 per cent in every year, but without the permanent ratchet effect that is hardwired into the triple lock. Under this refinement, the State Pension would rise by 2.5 per cent if that was higher than both prices and earnings in any given year, but when growth in earnings returned to exceeding 2.5 per cent, this would not be passed on in full to the State Pension, so that, as with the smoothed earnings link, the State Pension would keep track in the long run with earnings.

Figure 3 shows how the value of the BSP would have played out since the triple lock was introduced under different uprating mechanisms. Over this period a smoothed earnings link would have left the real value of the BSP £12.30 per week higher in 2026-27 than in 2011-12, essentially the same as the proportional growth in average earnings over that period. Compared to triple lock uprating, a smoothed earnings link would have left its current value around 11 per cent (£28 per week) lower (taking into account the temporary suspension of the triple lock in 2022). The OBR estimates spending on the State Pension would have been £12.6 billion lower in 2029-30 with earnings-linked uprating than it is set to be with triple lock uprating, which we estimate would be a net saving of around £9 billion when we take account of lower income tax receipts and higher means-tested benefit spending.

Figure 3: Alternative uprating mechanisms could lead to more sustainable State Pension uprating

Note that a ‘smoothed’ triple lock applied since 2012 would see the 2030 value around 5 per cent higher than earnings uprating. In theory, values under a smoothed triple lock should converge with straightforward earnings uprating over time, assuming earnings grow faster than both CPI and 2.5 per cent in the long run. But this convergence could take a very long time: as Figure 3 shows, it would not have occurred at any point since 2011-12. What the thought experiment does show, however, is that the biggest driver of the cost of the triple lock has been the ratchet effect rather than the (albeit totally arbitrary) choice to guarantee a certain nominal increase: using a smoothed triple lock rather than the regular triple lock since 2012 would have meant net spending would now be £6.5 billion per year lower.

Stepping back, it’s important to note that even a smoothed earnings link represents a significant commitment to prioritising maintaining the living standards of pensioners over children, the disabled, and others of working-age, and the cost of the State Pension will continue to rise in real terms both due to above-inflation uprating policy but also due to an ageing population. For anyone promising to make real-terms savings to welfare, a more aggressive policy would be needed, such as linking it (as are working-age benefits) to prices. This is not something we advocate, but it is important to realise that moving away from the triple lock towards a smoothed earnings link does not mean abandoning a commitment to the relative generosity of pensioner benefits; the gap between the value of benefits paid to people before and after the State Pension age would continue to rise, and we would still need to consider how to pay for a growing pensions bill.

Politicians owe us an answer on what they see as the appropriate level of the State Pension, rather than leaving it to random economic variation. Our answer is that, given the state of the public finances and the relative position of pensioners compared to non-pensioners, the time to favour pensioners over typical workers is over. But whatever the right level, the time has surely come to move away from a ratchet that leaves the State Pension a function of economic volatility.

[1] The triple lock was announced at the 2010 Budget and legislated to apply from April 2011. However, the Government chose to uprate the State Pension by RPI in that year before the switch to using CPI to determine price inflation for uprating so that the new policy would not be less generous than the existing policy of RPI indexing. The triple lock was then applied in practice from April 2012. The triple lock was temporarily suspended in April 2022; this was because the rate of growth of earnings reported in summer 2021 was artificially inflated, thanks to the level of average earnings being artificially depressed in summer 2020 when several million workers were on furlough.

[2] A different argument for moving away from price-indexing the Basic State Pension was that price-indexing the Basic State Pension while earnings-indexing Pension Credit (as was established policy) would steadily dull the incentives for workers to save into a private pension: why bother saving when you can expect to get an increasingly generous top-up to the Basic State Pension? The original Pensions Commission therefore proposed to link the Basic State Pension to earnings, so its value in relation to Pension Credit was maintained. Crucially, though, the ratchet effect of the triple lock is not necessary for this purpose, and the Pensions Commission didn’t call for a triple lock. On top of this, the introduction of the New State Pension in April 2016 for people who reached pension age after that date also solves the disincentive to save problem as its value was set just above Pension Credit’s income floor.

[3] RF analysis of ONS, National population projections, 2024-based.

[4] Source: RF analysis of DWP, Benefit expenditure and caseload tables, Spring Forecast 2026-27.

[5] Source: RF analysis of OBR, Economic and Fiscal Outlook, Spring Forecast 2026; OBR Fiscal Risks and Sustainability Report, July 2025; OBR, Long-term economic determinants, March 2025; DWP, Family Resources Survey, including use of the IPPR tax-benefit model.