What is the Triple Lock?

Since 2012, the Government has committed to increasing the UK state pension each year by whichever is highest of three measures:

• CPI inflation (the previous September reading)
• Average earnings growth (May–July over the last year)
• 2.5%.

Whichever of the three is most generous in a given year is the one pensioners get.

That "always pick the best of three" design is easy to defend politically — pensioners never do worse than 2.5%, inflation, or wages. But it leads to unintended consequences, as shown in the two charts below: the left chart shows which measure ‘wins’ (is highest) each year; the right chart shows what that leads to over time.

Inflation uses the previous September's CPI. Earnings uses the previous May–July average weekly earnings growth.

Chart: Centre for British Progress · Source: ONS (CPI, AWE) · CBP analysis

Why is this a problem?

Look at 2016, or 2021: inflation or earnings growth dipped, and was sometimes even negative — but the pension still rose by 2.5%, because growth was set by the floor. If we look at 2023, the triple lock would have paid 10.1%, matching that year's inflation spike. But the government actually suspended the earnings link in the 2022 uprating (using a "double lock" instead) because the post-furlough earnings figures were artificially inflated. This made the one year on the left chart where the triple lock didn't apply as designed because it would have represented an unsustainable increase.

The mechanism only ever pays the best of the three each year. It never compensates for when inflation and earnings are both weak, which means pensions can only increase faster. Projecting over many years shows how that effect compounds, as the right-hand chart shows: the pension permanently pulls ahead of what a single, consistent measure would have implied, even though most individual years don’t make a large difference. This is the ratchet effect.

There's a second problem: inflation and earnings aren't independent of each other. A given month's inflation reliably predicts earnings growth around nine months later. The reverse isn't true; past earnings growth doesn't predict future inflation. 2023 into 2024 is the clearest recent example of that pattern in action: a CPI spike drove a 10.1% uprating in 2023, and the earnings measure then caught up enough to drive an additional 8.4% uprating in 2024, from what was essentially the same shock. The Triple Lock can get hit twice by one shock because earnings are often set to compensate for inflation. This is the ‘double dipping’ problem: the same rise in prices is counted twice, due to the nine month lag between inflation and wage growth.