This series is adapted, with permission, from a Chinese-language series by our friend Marcus Ji, who looks at historical events through an economic lens. We have condensed and adapted the original for an international audience.
On 1 December 1942, people queued outside Britain’s government publications office in London to buy a 299-page report on social insurance. It sold out before noon. The BBC broadcast its contents in more than twenty languages, and copies were later dropped over Nazi-occupied France.
The Beveridge Report offered a promise: wartime sacrifice would be followed by protection against poverty, illness and unemployment. A welfare proposal had become an instrument of wartime morale.
That transaction runs through 2,500 years of welfare history. Governments paid for participation, order and loyalty. Once those payments became established rights, withdrawing them carried a political cost. Keeping the promise while reducing its value offered another way out.
Athens: the payment stayed, its value fell
In fifth-century BC Athens, citizens serving on juries received a daily allowance. Introduced under Pericles, the Athenian statesman, it allowed poorer citizens to participate in the courts. The payment began at two obols, small units of silver currency, and rose to three in 425–424 BC.
For roughly a century afterwards, it stayed there. Building accounts show what happened to its value relative to wages:
| Period | Daily jury allowance | Unskilled daily wage | Allowance as a share of wages |
|---|---|---|---|
| Late fifth century BC | 3 obols | 6 obols | One-half |
| Late fourth century BC | 3 obols | 9 obols | One-third |
Jurors still received the promised amount, but it replaced a smaller share of a day’s earnings.
This was not simply a government forgetting to update its payments. Over a similar period, allowances for attending the citizens’ assembly increased from one obol to at least six. The assembly needed enough people to conduct business, and attendance could be difficult to secure. There was no comparable shortage of prospective jurors.
The payment depended on whether the state needed more participants, rather than whether recipients needed more money. The same government could raise one allowance while leaving another untouched.
Why welfare becomes difficult to withdraw
Rome’s grain distributions show how benefits became attached to membership of a political community. In 123 BC, the politician Gaius Gracchus introduced grain sales at a subsidised price. In 58 BC, another politician, Clodius, made the distribution free.
Eligibility depended on being a male citizen living in Rome, rather than on poverty. A prosperous citizen could qualify; an impoverished foreigner or enslaved person could not. Grain was a benefit of citizenship. Removing it therefore meant changing what citizenship entitled someone to receive.
Two thousand years later, German Chancellor Otto von Bismarck used welfare to secure workers’ allegiance. Having suppressed socialist organisations in 1878, he introduced sickness insurance in 1883, accident insurance in 1884, and old-age and disability insurance in 1889.
These schemes made contributions the basis of entitlement. Workers were no longer merely asking for charity: they had paid into a system that owed them benefits. Bismarck hoped this would weaken support for more radical socialist alternatives.
Britain’s wartime promise extended the bargain further, presenting social protection as part of the peace people were fighting for.
In 1949, British sociologist T. H. Marshall described this shift as the development of social citizenship. Civil rights protected personal freedom and property; political rights gave people a vote. Social rights added basic economic security and a decent standard of living as part of citizenship, without requiring people to prove their poverty or moral worth.
Sociologist Gøsta Esping-Andersen later called the ability to maintain a decent life without selling one’s labour decommodification. Welfare could give someone protection even when they could not earn a wage.
This changed the politics of withdrawing support. A government cutting charity could argue that recipients no longer deserved it. Cutting a social right meant taking away something citizens understood to be theirs. Each expansion gave more people a stake in preserving welfare, making the promise harder to reverse when the cost increased.
The promise depended on future workers
After the Second World War, rapid growth helped governments deliver. Between 1950 and 1973, Western European GDP per person grew by about 3.8% annually. Social spending expanded alongside incomes.
This does not establish that welfare caused that growth, or that it inevitably undermined it. Economic historian Peter Lindert’s cross-country research found no significant negative relationship between social transfers and growth within the historical range he studied. How taxes were collected and benefits designed mattered.
Pensions, however, contained a separate dependency. Many public schemes operate on a pay-as-you-go basis: contributions from today’s workers pay today’s pensioners. The money is not all put aside for each contributor’s eventual retirement. Their future pension depends on future workers and future rules.
When large working generations supported smaller retired generations, this arrangement was easier to finance. Falling birth rates and longer retirements changed the arithmetic. Each generation had made a commitment on behalf of people who had not yet entered the workforce.
In their 2025 report, the US Social Security trustees projected that the Old-Age and Survivors Insurance trust fund would exhaust its reserves in 2033. Without legislative changes, continuing income would then cover 77% of scheduled benefits. Exhaustion did not mean all payments would stop; it meant the existing promise exceeded the projected income available to fulfil it.
China faced a similar problem. A 2019 Chinese Academy of Social Sciences projection put depletion of the urban employee pension fund’s accumulated reserves in 2035. A 2024 updated calculation estimated that delaying retirement could postpone depletion by eight or nine years. It bought time without removing the underlying demographic pressure.
More contributions, smaller benefits, later retirement
Governments facing that gap have three direct options: make people contribute more, pay them less, or make them retire later.
Athens illustrates a less conspicuous alternative: maintain the stated payment while allowing its value to fall. A pension can increase in currency terms yet buy less if prices rise faster. It can keep pace with prices yet replace a smaller share of earnings if wages rise faster. Raising the pension age can also reduce the number of years a person receives it.
These are different forms of retrenchment, but none requires abolishing the pension system. To understand what has changed, examine:
- Purchasing power: what goods and services will the payment cover?
- Replacement rate: how much of pre-retirement earnings will it replace?
- Eligibility: when can someone claim, and after how many years of contributions?
- Indexation: does the payment rise with prices or wages, and who can change that formula?
The existence of a benefit tells us less than its terms.
Can governments restore the birth rate?
More future workers would ease the pressure. But generous family policies have not reliably restored fertility to the replacement level of roughly 2.1 children per woman.
Nordic countries offered extensive parental leave and childcare. Nevertheless, Norway’s fertility rate fell from 1.98 in 2009 to 1.62 in 2017, while Finland’s reached around 1.3 in 2022.
Hungary pursued large financial incentives. Its fertility rate rose from 1.23 in 2011 to about 1.6 in 2021, before falling again. These experiences do not show that support for families is useless. They show that generous support alone does not guarantee enough births to sustain existing pension arrangements.
Welfare can share the cost of raising children. It cannot ensure that people choose to have them. A retirement promise dependent on future contributors cannot simply assume that population policy will repair the shortfall.
What is the promise worth to you?
The first distinction is between money accumulated in an individual account and an entitlement financed by future contributions. Their risks differ. The latter depends directly on the number of future workers and the decisions of future governments.
The second is between the amount promised and the living standard it supports. A payment that survives unchanged on paper may cover less of retirement than expected. That is why the contribution rules, pension age and adjustment formula deserve as much attention as the headline amount.
The people queuing in London in 1942 were buying a promise that was subsequently fulfilled in substantial part. But the growth and demographic conditions supporting it changed. The question facing each generation remains the one Athens left behind: three obols are still being paid, but how much are they worth now?
Previously in this series: Why do poor people need to understand finance more than the rich?
Next in this series: What is the true business logic of the lottery?
