On 5 November Germany's health minister told the Bundestag that more than six million people now draw benefits from the country's care insurance. When its five care grades were introduced a decade ago, the forecast was 3.5 million by 2030. The fund faces a deficit of about €8 billion next year, and the bill he was introducing raises contributions and slows benefits to close it. A week earlier Britain had announced that it would pay for free personal care in England by changing how the state pension rises. Australia began charging wealthier older people more for care at home a year ago. Japan is arguing, again, about who should pay a fifth of the cost of their care.
The previous essay in this series explained why every rich country is under the same strain: more old age than its welfare state was built for, a care risk that was never pooled, and a long drift of risk from institutions to households. This one looks at what governments, and people outside government, are trying in response. The answers fall into four groups. Some countries have built a fund for the care risk. Some have put the money in people's own hands. Some have tried new ways of delivering care. And all of them are looking for new ways to raise the money. A fifth group matters as much: the attempts that were tried and abandoned.
Building a fund for the care risk
The most direct answer to an unpooled risk is to pool it. Germany did so first, in 1995, with a compulsory care insurance that every employee pays into alongside health insurance: 3.6% of wages today, split between employer and worker, with a surcharge for the childless. Japan followed in 2000. Everyone over 40 pays a premium; half the cost comes from premiums and half from taxes, and the national average premium for over-65s is now ¥6,225 a month, reset in public every three years. In March 2025, 7.21 million Japanese were certified as needing care or support, and about 6.2 million used services each month. South Korea built a similar scheme in 2008; it now covers 1.26 million people, about 11% of over-65s, funded by a contribution of just under 1% of income.
These funds have lasted, and they share three features. The contribution is visible, the benefit is defined, and the cost is argued over in the open, because members can see it rise. None has escaped the pressure: Germany's fund is in deficit, Japan has deferred again a decision to widen the band of users who pay 20% co-payments, and Korea's state subsidy runs below its legal target. But each has turned the care risk into something a country argues about every few years rather than something a family discovers alone.
Others have pooled through taxes rather than contributions. Scotland has made personal care free at home since 2002 for over-65s and since 2019 for all adults; about 61,000 people received it at home in 2024–25. The Netherlands split its old scheme in 2015 into a national insurance for intensive, lifelong care and municipal support for the rest; spending on long-term care fell from 4.8% of GDP in 2012 to 4.5% in 2017, though the official evaluation found the hope that people would care more for each other was not met. England's promised National Care Service follows the Scottish path for over-65s, at an estimated £7.5 billion by the end of the next parliament, and Singapore has gone the other way: CareShield Life, compulsory for everyone born since 1980, pays a lifelong cash sum, S$689 a month in 2026, to anyone who becomes severely disabled.
Putting the money in people's hands
The second answer changes not who pays but who decides. Australia's National Disability Insurance Scheme gives each of its 782,013 participants a funded plan; it cost A$51.2 billion in the year to June, growing at 10.5% against a government target of 8%. Some 1.52 million Americans directed their own Medicaid-funded care in 2023, up 23% in four years. In England 117,655 people received direct payments from their council in 2024–25, about 36% of working-age adults on long-term support but only 14% of older people. The Netherlands, Germany and Sweden each offer versions of the same idea.
The evidence, covered in more detail in our briefing Who decides?, is good for people who want the control and can use it, and weaker for older people and for anyone left to manage alone. It has also taught governments what individual budgets cannot do by themselves. Australia's 2023 review of the NDIS concluded that early help needed "foundational supports" funded outside the scheme, and that budgets without navigators, price-setting and checks produce thin markets in some places and fraud in others. The lesson being drawn in several countries is not to abandon the budget but to build what has to surround it.
New ways of delivering care
The third group of experiments starts from the work itself. Buurtzorg, founded in the Netherlands in 2006, replaced managed home-nursing agencies with self-governing neighbourhood teams of 10 to 12 nurses, each looking after 50 to 60 people; by 2018 it employed about 10,000 nurses. A 2015 analysis by KPMG found its home-care costs low, though total costs per patient were about average once hospital and GP care were counted, and pilots in Wales could not confirm the gains. Its reputation runs ahead of its evidence, but it has made a serious case that care workers trusted to organise themselves stay longer and waste less.
Other models move care into households and communities. Britain's Shared Lives schemes match adults who need support with approved carers who take them into their own homes; almost 10,000 people were supported this way in 2024–25, up 13% in three years. Italy's social co-operatives, owned by workers and sometimes by users, employed 509,333 people in 2023, more than half of all employment in the country's non-profit sector. In New York, Cooperative Home Care Associates has grown into one of America's largest worker co-operatives. Australia's Mable connects people holding care budgets directly with independent support workers, without an agency in between. And Japan, short of workers, is testing machines: a 2024 study of its nursing homes found that those using care robots kept staff longer and saw fewer pressure ulcers and less use of restraints.
What these share is that they change who organises the work: nurses rather than managers, households rather than institutions, workers or members rather than shareholders. Most are small, and few have been rigorously evaluated. That is the gap in the evidence, and it is also where the most interesting experiments are.
Raising the money
Every one of these answers costs money, and the fourth group of experiments is about where it comes from. Japan asks people in middle age to pay in: those aged 40 to 64 cover 27% of the cost of the care insurance, through their health insurance, long before most will use it. Germany's bill this month widens its base: a higher ceiling on the earnings that contributions are charged on, a larger surcharge for the childless, contributions from spouses now covered free and from mini-jobs. France made the loss of autonomy a fifth branch of its social security system in 2020, funded by redirecting part of a general social contribution.
Britain has chosen to pay for its promise by changing the uprating of the state pension from 2030, which it expects to save £15 billion a year by the late 2030s. Australia has gone the other way, asking those who can afford it to pay more of their own care: since November 2025 a full pensioner pays 5% of the cost of help with independence and 17.5% of everyday help at home, a self-funded retiree 50% and 80%, up to a lifetime cap. Each choice says something about who is expected to carry the cost: workers in their 40s and 50s, all contributors, the general taxpayer, future pensioners, or the person who needs the care.
What failed, and what the attempts share
The failures are as instructive as the successes. England has tried to cap what anyone pays for care since the Dilnot commission of 2011; a cap was legislated in 2014, delayed, dropped, revived in 2021 and scrapped again in July 2024, each time when the cost came into view. Scotland's plan for a national care service was cut back in January 2025. America's CLASS Act of 2010, a voluntary public care insurance, was suspended within a year because it could not show it would pay for itself over 75 years, and repealed in 2013: a voluntary pool attracts the people most likely to claim. The Netherlands' 2015 attempt to curb fraud in personal budgets by paying carers directly broke down in its first months, needing tens of thousands of emergency payments. And in China, Xiang Hu Bao, an online mutual-aid scheme run by Ant Group, signed up 100 million members in a year and closed in January 2022 as its members aged and claims rose.
Read together, the attempts point to a handful of conditions. Pools last when membership is compulsory or near-universal and the contribution is visible; voluntary pools, from CLASS to Xiang Hu Bao, unravel. Budgets work when someone helps people use them and someone checks how they are used. New models of care spread where the money can follow the person to them, and stall where it cannot. And every scheme that survived made someone pay in advance, in plain sight, rather than leaving the bill to arrive at the moment of need.
None of the countries in this series has put all of these together. The final essay sets out what that would look like: a welfare system that treats care not only as a cost to be contained but as a source of work and capacity, and what it would take to build it.
Sources: Deutscher Bundestag, first reading of the Pflegeneuordnungsgesetz (5 Nov 2026) and the minister's speech; AOK; Japan Ministry of Health, Labour and Welfare (care insurance finance; FY2024 report; 2024 fee revision); Korea's long-term care insurance (July 2026 presentation); Scottish Government, Free personal and nursing care 2024–25; Netherlands Institute for Social Research (SCP, 2018); Nuffield Trust; Singapore Ministry of Health (CareShield Life, Aug 2025); NDIA quarterly report (June 2026); NDIS Review (2023); England direct payments analysis 2024–25; 2023 Self-Direction National Inventory; KPMG (2015) on Buurtzorg, via secondary sources; Shared Lives Plus; Istat, non-profit sector 2023; PHI; NBER Working Paper 33116; My Aged Care, Support at Home contributions; Vie-publique on France's autonomy branch; Community Care; LocalGov; CLASS Act history; Netherlands parliamentary papers on the PGB (Feb 2017); TechNode and Caixin on Xiang Hu Bao.
