The Pension Arithmetic: Why the Math of Retirement Cannot Hold

Few fiscal commitments are as politically sacred and as quietly insolvent as the promise of a state pension. Across the developed world, governments have pledged income for life to populations that are living longer and reproducing less, and they have done so without setting aside the capital to honour those pledges. The pension is not, for most of the world’s workers, a pot of money with their name on it. It is a claim on the earnings of people who, in many countries, have not yet been born. That arrangement worked beautifully while the pyramid had a wide base. The base is now narrowing, the apex is widening, and the arithmetic that once flattered every politician who signed a benefit increase is turning against them. This is not a crisis of fraud or mismanagement. It is a crisis of demography meeting a promise that was never actuarially funded, and it is unfolding in slow motion precisely because no government wants to be the one holding the bill.

The Engine: How Pay-As-You-Go Actually Works

Most public pension systems, including the United States Social Security program, are not savings schemes. They are intergenerational transfers. The payroll taxes collected from today’s workers are paid out, almost immediately, to today’s retirees. There is no vault. The Social Security trust funds hold special-issue Treasury securities, which is to say the government owes the money to itself; redeeming them requires raising taxes, cutting other spending, or borrowing. This is the pay-as-you-go model, and its solvency depends entirely on the ratio of people paying in to people drawing out.

When the model was designed in the mid-twentieth century, that ratio was extravagantly favourable. Birth rates were high, life expectancy past retirement was modest, and a large cohort of contributors supported a small cohort of pensioners for a relatively brief period. A politician could legislate a generous benefit and never see the cost, because the cost fell on a future that looked, on paper, even more populous than the present. The system was sound only so long as each generation was larger and richer than the one before it. That assumption, once a safe bet across the industrialised world, has quietly stopped being true.

The Squeeze: The Dependency Ratio Inverts

The single most important number in this story is the old-age dependency ratio: the count of people past the conventional retirement age relative to those of working age. According to the OECD, the number of people aged sixty-five and over for every hundred of working age has climbed from roughly 20 around the turn of the millennium to about 33 in the mid-2020s, and it is projected to reach roughly 50 by 2050. In plain terms, a developed economy that once had four or five workers behind every retiree is heading toward fewer than two. The trajectory is steeper still in much of continental Europe and in East Asia, where fertility has fallen furthest.

This is the engine seizing. A pay-as-you-go system funded by two workers per retiree must either tax those two workers far more heavily than their predecessors, pay each retiree far less, or push the retirement age up so that the pensioner spends more years contributing and fewer years collecting. The OECD has estimated that simply to hold the dependency ratio steady over the first half of this century would require raising the effective retirement age by more than eight years, an adjustment that dwarfs anything governments have actually legislated. There is no fourth door. Every honest reform is some combination of the same three levers, and each one inflicts a visible, immediate loss on a large and organised constituency.

The driver beneath the ratio is twofold and irreversible on any policy timescale. Fertility has fallen below the replacement rate of roughly 2.1 children per woman across nearly all advanced economies and much of the developing world. Simultaneously, longevity has risen, so the benefit, originally priced for a short retirement, now funds two or three decades. A pension designed when sixty-five was near the end of life behaves very differently when sixty-five is the beginning of a long third act.

The Quiet Retreat: Defined Benefit Gives Way

The private sector saw the arithmetic first and acted on it with a candour that governments cannot afford. The defined-benefit pension, in which an employer guarantees a fixed income for life and carries all the investment and longevity risk on its own balance sheet, has been in retreat for four decades. In the United States, the share of private-sector workers participating in a defined-benefit plan fell from around 38 percent in 1980 to roughly 15 percent today. Measured by active participants the collapse is starker still: private-sector defined-benefit plans counted some 27 million active members in the mid-1970s and roughly 11 million by the 2020s, while defined-contribution arrangements such as the 401(k) ballooned from about 11 million to nearly 100 million over the same span.

This was not an accident of fashion. It was a deliberate transfer of risk from the employer to the worker. Under a defined-contribution plan the company’s obligation ends when it deposits its contribution; the investment risk, the longevity risk, and the discipline of saving enough all migrate to the individual. Corporations shed an open-ended liability that markets and ageing could inflate without limit. The worker, in exchange for portability and a notional pot of his own, inherited the uncertainty that a pension was invented to abolish. The state has no equivalent escape. It cannot quietly close its scheme to new members, because its scheme is, in effect, compulsory and universal, and its members vote.

The Scale of the Hole

The gap between what has been promised and what has been funded is difficult to grasp because it is so large. The World Economic Forum’s landmark retirement-savings study put the shortfall across the six largest developed pension systems (the United States, the United Kingdom, Japan, the Netherlands, Canada, and Australia) on a trajectory toward roughly $224 trillion by 2050, with the figure rising to about $400 trillion once the very large systems of China and India are added. The United States alone accounts for some $137 trillion of the developed-world total, more than half of it. Set against that, pension assets across the world’s major funded markets stand in the low tens of trillions of dollars: real money, but a fraction of the eventual claim.

These are projections, and projections carry wide error bars; the precise number matters less than the order of magnitude and the direction of travel. What is not in dispute is the timeline of the funded systems that can run dry. The trustees of US Social Security have for years projected the depletion of the old-age trust fund reserves in the early-to-mid 2030s, after which incoming payroll taxes would cover only around three-quarters of scheduled benefits. That is not bankruptcy; the contributions keep flowing. But it is an automatic, legislated benefit cut of roughly a fifth or more, arriving on a date that every serving legislator can already see on the calendar and is choosing not to discuss.

The Politics of Postponement

Here the arithmetic collides with democracy. The costs of reform are concentrated, immediate, and felt by people who are politically active; the benefits are diffuse, deferred, and accrue to a future electorate. A finance minister who raises the retirement age by two years confronts mass strikes today in order to spare an abstract taxpayer two decades hence. The incentive structure points relentlessly toward delay. France’s experience is instructive: when its government moved the standard retirement age from 62 to 64, it triggered months of nationwide strikes and was ultimately forced to push the measure through without a final parliamentary vote, using a constitutional override. The reform was modest by actuarial standards and politically near-catastrophic, and the pressure to dilute or unwind it did not end with its passage.

The result is a predictable pattern across democracies. Reform is attempted, diluted, deferred, or reversed; the structural deficit is papered over with accounting devices and optimistic growth assumptions; and the genuine adjustment is left to a successor government, or to the automatic mechanism that triggers when a fund finally empties. Because the levers are unpopular in proportion to their effectiveness, the changes that do pass tend to be the gentlest available, which means the underlying gap continues to widen even after a celebrated reform. The mathematics is patient. It does not require legislation to act.

The Coming Reckoning

The strategic implication is that the pension question will not be resolved by a single dramatic event but by a long sequence of quiet defaults. Some will be explicit, in the form of higher retirement ages and trimmed benefit formulas. Others will be implicit, delivered through inflation that erodes the real value of fixed payments, through means-testing that redefines who counts as deserving, or through the gradual conversion of a guarantee into a contribution. The retirement promise will not be broken in one act; it will be renegotiated continuously, against the citizen’s interest, for as long as the demographics deteriorate. The systems that adjust early, transparently, and in small increments will preserve trust and spread the pain. Those that wait for the trust fund to empty will discover that the arithmetic, having been ignored for a generation, collects its arrears all at once.

Read our full Report Disclaimer.