A deal undone

On 2 November 2011 Danny Alexander, then Chief Secretary to the Treasury, set out the government’s offer on public sector pensions. Most workers would move from final salary to career average schemes, pay more and work longer. But anyone within ten years of their pension age on 1 April 2012 would be left exactly where they were. It was a sweetener, intended to take the heat out of a national strike planned for 30 November. The strike went ahead anyway. Fifteen years later, the sweetener has become one of the most expensive parts of the deal.

The new schemes arrived in 2015. Two groups of members who had been moved across went to court. One was a group of judges, whose case took the name of its lead claimant, Victoria McCloud. The other was a group of firefighters, led by Ms R Sargeant. Their argument was simple: colleagues doing the same job had kept the old terms purely because they were older. That, they argued, was unlawful age discrimination. In December 2018 the Court of Appeal agreed. In June 2019 the Supreme Court refused the government permission to appeal. The government accepted that the same flaw ran through every main public service scheme: the NHS, teachers, the civil service, police, the armed forces and local government.

It’s worth explaining who is affected. Protection depended on your age on 1 April 2012, so only people who were members on that date could have been treated differently because of their age. Someone who joined in 2013 went into the new scheme in 2015 whether they were 25 or 55; nobody discriminated against them. The remedy therefore covers people who were in service on or before 31 March 2012 and still in service after 1 April 2015, with gaps of up to five years allowed. If you joined later, McCloud passes you by.

Why reform at all? The case was made by Lord Hutton’s commission, which reported in 2011. Part of it was longevity; Hutton noted that a 60-year-old could expect to live around ten years longer than in the 1970s, and a pension for life was costing the government far more than when the schemes were designed. But part of it was fairness between members. A final salary scheme values every year of service based on your pay at the end. That suits the senior officer whose salary climbs steeply late in a career far better than the nurse whose pay is flat. Early leavers are on the wrong side of it too. Their pensions do not disappear; a preserved pension is paid in full at retirement. But it is fixed at the salary they had when they left and then only uprated for inflation, while colleagues who stay see theirs rise with every promotion. In effect, the steady earner and the leaver subsidise the high-flyer. A career average scheme spreads the benefit more evenly.

On longevity, events have been less tidy than Hutton assumed. Improvements in UK mortality slowed sharply from around 2011, and the pandemic reversed some of them. Mortality then fell to record lows in 2024 and 2025, but actuarial projections of life expectancy at 65 are still around 18 months below where they stood in 2009. We are living longer, but not by as much as the planners expected, and less predictably. That is precisely the kind of risk a large scheme absorbs easily and an individual does not.

I have written before about how the private sector abandoned defined benefit pensions. A steady drip of court cases, tighter regulation after the Maxwell scandal, the Pension Protection Fund and decades of falling interest rates made each promise more certain and therefore more expensive. Companies responded rationally and stopped making them. McCloud is the same process at work in the public sector; the courts made a promise stick, and the bill went up.

The pattern is international. The United States drifted from company pensions to the 401(k) from the 1980s, Japan’s employers have largely abandoned their traditional schemes since 2002, and even the Netherlands, long held up as the model of collective provision, is moving away from guaranteed benefits. You can read this as a story about markets and class: the paternalist employer who managed your retirement for you gives way to the individual who carries the investment and longevity risk on their own account. With it comes more freedom and less security. This aligns with social changes in many countries from the 1980s onwards: a focus on personal responsibility, ownership and individual incentives. But while there is cultural logic to these changes, there is also a straightforward economic benefit to the promise-makers; governments and companies have lowered their pension costs.

Public sector defined benefit schemes survived this retreat because most of them are unfunded. There is no pot of assets; today’s pensions are paid from today’s taxes. That gives governments a freedom no company has. A business with a pension deficit sees it on its balance sheet. In the UK it must agree a plan with the Pensions Regulator to close the gap and pay a levy to the Pension Protection Fund; the consequences are immediate. Sovereign governments have the latitude to make promises without a detailed plan to pay for them. The government’s own accounts valued pension liabilities at £1.3 trillion in March 2024. This isn’t necessarily a bad thing. The UK government can alter the rate of accrual of benefits in its schemes, and it can borrow to fund pension payments. There are limits, like the cost of borrowing and inflation. But ultimately, it is socially useful that a government can carry a vast liability for decades. This is also why the design of these schemes is fought over so fiercely: the cost is a political choice that redistributes resources from one group to another over time, rather than a bill that falls due today.

So, the public sector kept its promise, and the courts made sure it kept it to everyone. Next month we will look at what that means for the people affected: a pension that sits in two schemes at once until you retire, a choice that can move your retirement date by years, and a tax correction that, for the oldest years, can only go in your favour.