In July 2020, Pamela King told MPs what had happened after her sister Diana suffered a severe stroke.
Diana was left paralysed and needed residential care. Pamela described struggling through a complicated system while paying £712 a week for her sister’s care home. This was not an optional expense. Diana could not live without support.
Parliament had already passed legislation intended to protect people from unlimited care costs. The Care Act 2014 contained the framework for a lifetime cap, originally expected to begin in April 2016. It was delayed until 2020, then postponed without a new date. A revised £86,000 cap was later promised for October 2023, delayed until October 2025 and finally cancelled in July 2024. As of July 2026, that cap has still not been implemented.
Pamela and Diana did not live inside a policy paper. They lived through the years in which governments announced protection, delayed it, redesigned it and withdrew it.
That is the difference between political time and human time.
These are not sandwich choices
Policies affecting health, housing, income, disability, education, care and retirement are not casual preferences. They are not decisions about whether to put jam or cheese in a sandwich.
They determine whether a person can remain at home after becoming disabled. They determine whether someone must spend their savings on care, whether a family can obtain early help, whether an installer is paid for completed work and whether a city can plan its transport system with any confidence.
Yet British government often treats these policies as possessions of the current administration.
One government creates a programme. Another cuts its funding. A minister reorganises the bodies responsible for delivering it. A successor replaces them. Staff move, records scatter, contracts close and local organisations begin again under a new name.
The people relying on the service cannot begin their lives again each time Westminster changes direction.
This article focuses mainly on UK government decisions affecting England, because health, social care and early-years provision are devolved. The wider problem belongs to the British system of government: long-term commitments are politically easy to announce and legally easy to undo.
Democratic change is necessary. Churn is different
A democracy must allow a new government to repeal bad laws, correct failed programmes and pursue different priorities. Durability cannot mean making policy permanent regardless of evidence.
Irresponsible churn is something more specific.
It occurs when a government substantially dismantles a policy or institution before its intended effects can be measured, without an independent evaluation, without publishing the full cost of cancellation, without protecting people who relied on the previous rules, or without preserving useful staff, knowledge and infrastructure.
The question is not whether governments should be allowed to change policy. They must be.
The question is why major decisions affecting millions of lives can be abandoned with fewer safeguards than would be expected when closing a factory, ending a commercial contract or restructuring a pension fund.
The constitution permits rapid reversal
Parliamentary sovereignty gives the UK Parliament the legal power to make or unmake any law. One Parliament cannot normally prevent a later Parliament from repealing its decisions. Unlike countries with codified constitutions, Britain has no general special procedure for constitutional legislation: ordinary Acts can alter arrangements of enormous long-term importance.
That flexibility can be valuable. It also means that social guarantees established by legislation are only as durable as the political support behind them.
The legal power to reverse policy does not itself cause policy failure. The deeper problem is the combination of that power with centralised government, short ministerial tenures, annual political pressure and funding decisions that rarely match the time needed to produce results.
The Institute for Government found that further education experienced 28 major pieces of legislation and 48 relevant secretaries of state over 30 years. Since 1978, governments have introduced around 30 skills policies or programmes. Many were replaced before they could plausibly change the skills of an entire workforce.
High internal turnover in the Civil Service compounds the problem. Frequent movement between posts harms institutional memory, reduces expertise and weakens the management of major programmes.
The National Audit Office has separately warned that short-termism, siloed decisions and over-optimistic planning continue to put long-term value for money at risk.
The result is a state that is very good at authorising a fresh start and much less capable of remembering what the previous start taught it.
The care-cost cap that never arrived
The repeated postponement of the social-care cap is not merely a story about an unimplemented law.
It transferred financial risk back to people needing care.
Age UK’s analysis of official figures found that 5,190 self-funders exhausted their assets paying for care in England during 2018-19: about 14 people each day.
Government also paid for the delays. The National Audit Office found in 2023 that more than £1 billion of a £1.74 billion adult social-care reform budget had been diverted to other pressures. Planned spending on system reform had fallen by 58%. The Department of Health and Social Care also forfeited £171 million to the Treasury as a consequence of delaying the charging reforms.
Some of the diverted money helped councils and providers facing immediate pressure. That was not necessarily an irrational use of funds. Social-care vacancies, waiting lists and hospital discharge problems were serious.
But that is precisely the trap. Permanent reform is repeatedly postponed to manage the latest emergency, while failure to complete permanent reform helps create the next emergency.
The current government has established an independent commission under Baroness Casey. Its terms include building cross-party agreement and proposing a National Care Service capable of meeting current and future needs. The commission’s long-term phase is expected to report in 2028. That acknowledgement matters, but a commission is evidence of another attempt, not yet evidence of durable delivery.
Who paid? Disabled and older people exposed to uncapped costs; relatives navigating care on their behalf; councils repeatedly preparing for reforms that did not arrive; and taxpayers who funded preparation, delay and emergency stabilisation.
Sure Start: the results arrived after the cuts
Sure Start illustrates another problem: preventive policy often requires more time than politics permits.
The programme began in 1999, providing joined-up early-years services such as health support, childcare, parenting help and employment advice. At its peak in 2009-10, it had more than 3,000 local service points and annual spending equivalent to £2.7 billion in 2023-24 prices.
After the change of government in 2010, its funding was no longer protected. By 2021-22, spending was 73% below its peak and at least 1,340 centres had closed. Some services were consolidated rather than abolished, and later Family Hubs revived elements of the integrated model. A count of closed buildings therefore does not prove that every service vanished. The contraction in funding and reach, however, was substantial.
The most important evidence emerged much later.
Research published by the Institute for Fiscal Studies found that greater access to Sure Start initially increased hospital admissions among one-year-olds, probably partly because illnesses were detected and treated. By age 15, hospital admissions were 8.5% lower. At the programme’s peak, it was estimated to have prevented more than 13,150 hospital admissions each year among children aged 11 to 15. The health gains were strongest in disadvantaged areas.
This does not prove that every centre was effective or that Sure Start should have remained unchanged. Earlier evaluations identified inconsistent delivery and weak local cost information.
It does show why long-term preventive services need long-term evaluation. Some of Sure Start’s largest measurable effects appeared almost a decade after children had stopped using the centres.
By then, much of the original network had already been reduced.
Who paid? Families who lost nearby support; councils trying to meet rising statutory needs with less preventive capacity; and the NHS, which carried costs that effective early intervention could sometimes have prevented.
The NHS keeps reorganising the people who organise it
In 2012, the government legislated for one of the largest NHS reorganisations in its history. Primary care trusts were replaced in 2013 by 211 clinical commissioning groups, intended to give clinicians more control and reduce bureaucracy.
The government later reported transition costs of about £1.38 billion by March 2015, including £473 million in staff redundancies. It also claimed that reductions in administration produced larger continuing savings. Both parts of that record matter: restructuring cost money, but it was not necessarily a net financial loss under the government’s accounting.
The structural result was less convincing. The National Audit Office reported in 2018 that the number of commissioning groups was already falling through mergers, that no overall assessment had been made of whether they had achieved their original policy objectives, and that NHS commissioning needed a prolonged period of organisational stability after almost three decades of change.
Four years later, Parliament abolished the commissioning groups. On 1 July 2022, their functions passed to 42 integrated care boards.
This reversal had a defensible purpose. The 2012 arrangements had produced fragmented responsibilities and made cooperation across health and social care harder. Moving towards integrated systems was a reasonable correction.
The failure was not that the government changed course. The failure was building an expensive national structure, failing to evaluate it properly, then returning the NHS to another large transition before the previous organisations had achieved stability.
A correction should preserve useful knowledge and capacity. A cycle of replacement makes reorganisation itself a permanent NHS activity.
Who paid? Taxpayers funded closures, redundancies and new organisations. NHS staff spent time transferring responsibilities, systems and contracts. Patients bore the less visible cost when management attention was diverted from improving care.
Green Homes Grant: six months from launch to closure
Some policies should be stopped because they are failing. The Green Homes Grant was one of them.
Launched in September 2020, the voucher scheme was expected to support up to 82,500 jobs and help 600,000 households improve their energy efficiency. It closed to new applications in March 2021.
The National Audit Office found that the Treasury had given the responsible department 12 weeks to design the scheme, consult the market and procure an administrator. The department recognised that the timetable was high-risk but proceeded.
The eventual forecast was 47,500 upgraded homes and up to 5,600 jobs. An estimated £314 million of the £1.5 billion budget would be spent, including £50.5 million on administration: more than £1,000 for every home improved. More than 3,000 complaints were made, while installers faced delayed payments and households struggled with applications.
Keeping such a scheme open indefinitely would not have demonstrated admirable consistency. Closing it was defensible.
But closure did not erase the original failure. Government launched a national programme before the delivery system and installer market were ready. Businesses were encouraged to recruit and seek accreditation for a scheme lasting only months. Households delayed work or entered a slow application process. The state transferred the consequences of its rushed timetable to the people expected to deliver and use the programme.
Who paid? Installers who invested in participating, householders left waiting, taxpayers funding a costly administrative system, and future energy policy burdened by reduced public and industry confidence.
HS2: a national promise cut back after lives had moved around it
On 11 February 2020, the Prime Minister confirmed that HS2 would proceed. He told Parliament that HS2 north of Birmingham and Northern Powerhouse Rail were both needed and would both be built.
In October 2023, the government cancelled HS2 Phase 2 north and east of Birmingham, citing rising costs and changing transport priorities. It promised to redirect £36 billion into other transport projects.
Cost escalation and poor management supplied a serious case for reconsideration. Continuing a project simply because money had already been spent would have been irresponsible.
The way the project had been governed left no painless option.
The Public Accounts Committee concluded that the remaining London-to-West-Midlands scheme would provide very poor value for money because its costs would significantly outweigh the reduced benefits. It also raised questions about property acquired through compulsory purchase, effects on connected rail schemes and whether the replacement transport commitments would be delivered.
Closing Phase 2 was itself expected to take three years and cost up to £100 million.
This is the danger of presenting a long-term infrastructure commitment as settled while governance, cost control and delivery remain unstable. People sell homes. Businesses change investment plans. Councils design local transport around the promised network. A later government may have sound reasons to change direction, but it cannot return everyone to the position they occupied before the promise was made.
Who paid? Taxpayers funded planning, land, preparation and closure. Property owners and communities lived with years of uncertainty and compulsory acquisition. Northern cities lost the transport settlement around which they had been encouraged to plan.
Durability does not mean preserving failure
These examples do not support a rule that every policy must survive.
The Green Homes Grant was badly designed. The NHS commissioning structure created in 2012 had serious weaknesses. HS2 suffered severe cost and governance failures.
A responsible state must be able to stop, repair or replace policies that are unsafe, ineffective or unaffordable.
The difference lies in the method.
Correction begins with evidence. It preserves the useful parts of the previous system, protects people already committed to it and explains why amendment is insufficient.
Abandonment treats the previous policy as though it never existed. Its costs disappear into departmental accounts. Its promises are replaced with new promises. The people who relied on it become an administrative inconvenience.
Britain needs greater policy durability, not policy immortality.
A Responsible Reversal Standard
Major policies affecting health, housing, income, care, education, retirement or national infrastructure should not be dismantled through an ordinary announcement.
Every substantial policy should begin with:
-
A declared purpose
The government must state the problem being addressed and how success or failure will be measured. -
A review date
Evaluation should be built into the original decision. Long-term programmes need enough time for their intended effects to become visible. -
Independent evidence before reversal
Except where immediate safety or financial emergency requires action, government should publish an independent assessment before dismantling the policy. -
An explanation of why repair is insufficient
Ministers should show which alternatives were considered and why amendment, local variation or staged correction would not work. -
A full reversal account
This should include money already spent, cancellation costs, future liabilities, administrative disruption and costs transferred to councils, businesses or households. -
Protection for people who relied on the policy
Transitional rights should be provided where people made serious or irreversible decisions based on existing law or an explicit government commitment. -
A continuity plan
Useful staff, records, research, data, contracts and local capacity should survive even when the organisation responsible for them does not. -
Funding that matches the task
Institutions expected to produce results over ten or twenty years should not depend on short settlements that make long-term planning impossible.
Parliament would remain sovereign. Elections would still matter. Governments could still change direction.
They would simply have to show that they understood what they were undoing.
Governments can start again. People usually cannot.
A minister may experience a reversal as a new strategy, a spending decision or a change of machinery.
Pamela and Diana experienced policy through a weekly care bill.
Families experienced Sure Start through a centre that was there, and then was not. NHS workers experienced reform through another employer, another structure and another transfer. Installers experienced the Green Homes Grant through unpaid work and a market that vanished. Communities along HS2 experienced it through land, homes and plans placed in limbo.
That is why policy cannot be treated as a contest between temporary political preferences.
A government has the right to change its mind. It should have to show its working, count the damage and protect the people who cannot reorganise their lives every five years.