The quiet collapse of the administrative state
I stumbled upon an article I missed last week by Sam Thorpe that even I found shocking.
The CQC regulates roughly 30,000 health and social care services across England. Hospitals, nursing homes, domiciliary care agencies, GP practices, dental surgeries, mental health trusts. It inspects them. It rates them. It publishes the results. The ratings (Outstanding, Good, Requires Improvement, Inadequate) are supposed to tell you and your family whether you can trust the place looking after you.
In 2019, before the pandemic, the CQC carried out more than 15,700 inspections. By 2023–24, the number had fallen to roughly 6,700. A decline of 58 per cent. In 2024, up to the end of July, the figure was 1,820. At one point, an estimated one in five of the services the CQC has the power to inspect had never received a rating at all. The oldest rating for a social care provider dated to October 2015. For an NHS hospital, June 2014. Almost a decade old.
The average age of a provider’s overall rating was 3.7 years.
Think about what this means in practice. A care home rated Good in 2020 could have deteriorated beyond recognition. A GP surgery rated Requires Improvement in 2021 could have transformed or collapsed. Nobody checked. The rating sits on the CQC website like a museum label: a record of what was once observed, disconnected from what is happening now.
This is in keeping with many other trends we’ve seen just recently where the enforcement of regulation and monitoring of standards has completely collapsed. But in this case, it gets worse.
By late 2025, the CQC had begun a recovery programme under new leadership. The 500-report backlog was reduced, by the CQC’s own account, to four. The organisation set a target of publishing 9,000 assessments between April 2025 and September 2026, and by December 2025 claimed to have published 4,308.
These numbers sound like progress, and in the narrowest sense they are. But context is everything.
In 2019 the CQC carried out more than 15,700 inspections. The 9,000 target, spread over eighteen months, amounts to roughly 500 assessments a month. In 2019, the CQC was averaging 1,313 a month. The recovery target is 38 per cent of what the regulator once achieved.
An analysis by the law firm RWK Goodman in January 2026 found the CQC was already falling behind even this modest ambition, publishing around 456 reports a month. The CQC described itself as “ahead of target.” The target was a fraction of what the job requires.
And the Single Assessment Framework (the system whose botched rollout caused much of the damage) has not been replaced. It is being revised, simplified, consulted upon. New sector-specific frameworks are promised. A rebuilt provider portal is in development. Pilot programmes are underway. The full rollout of the reformed approach is expected by the end of 2026. In the meantime, inspections continue under a hybrid model nobody fully trusts.
The CQC is rebuilding itself while it is supposed to be inspecting the health service. It is redesigning the aeroplane while flying it. And the passengers (patients in hospitals, residents in care homes, families choosing where to send elderly parents) are being asked to trust a rating system the regulator itself acknowledges is broken.
The author’s conclusion will come as no surprise either…
The pattern is always the same. An institution staffed by people who knew their sector, who had relationships with the providers they oversaw, who could walk onto a ward or into a care home and recognise danger because they had spent years learning what danger looked like. The decision to replace those people with a standardised process — cheaper, scalable, data-driven. The process failing, because no framework can replicate the judgment of a human being who knows what a safe ward smells like and what an unsafe one sounds like. And the failure compounding, because the experienced people are gone and the framework does not work and now there is nothing.
The CQC had a staff survey. In 2024, just 27 per cent of employees felt the “values” and “behaviours” of the chief executive and executive team matched those of the organisation. Three years earlier, the figure was 55 per cent. The people inside the regulator knew. They could see what was happening. The framework swallowed their reports, ignored their expertise, destroyed their relationships with providers, and left them unable to do the work they had been hired to do.
The old inspectors knew their patch. They knew which homes to worry about, which hospital managers were hiding problems, which GP surgeries were struggling. That knowledge was not digital. It could not be uploaded to a platform. It lived in the heads and notebooks and professional instincts of experienced people who had spent careers learning a trade. When the CQC replaced them with generalists operating a framework, it did not modernise inspection. It abolished it, and called the abolition reform.
I didn’t want to reproduce the entire article here but it should be read because I’ve skipped some important context. In this instance the backlog was caused by an IT platform transition, but the platform itself was part of yet another efficiency drive, assuming technology could replace the expertise of the administrative tier (on the assumption that they’re useless NHS bureaucrats).
In doing so we lost a major pillar of institutional knowledge - leading to a collapse in safety surveillance. The very same people who demanded this “efficiency” will then say the NHS is broken beyond repair.
This is a dynamic we see in just about every area of enforcement where seasoned inspectors have been replaced by checklists, reducing an important layer of supervision with a bureaucratic paperwork exercise - to the extent that service providers are failing badly and nobody in government even knows.
Interestingly, Sam Thorpe’s article links to a Guardian piece (the kind of reporting that the Guardian is actually good at) which shows that the last round of efficiency measures created more problems than it solved.
David Cameron’s “bonfire of the quangos” decision to abolish England’s council spending watchdog has left a broken system that is costing taxpayers more money than it was promised to save.
In a highly critical report, academics at the University of Sheffield said the coalition government of the Conservatives and Liberal Democrats had promised savings of £100m a year by abolishing the Audit Commission.
However, replacing the public body with a private-sector model had resulted in “chaos” and soaring costs to audit councils amid the financial crisis hitting England’s town halls.
Several councils have declared effective bankruptcy linked to years of austerity, soaring costs amid pressure on services, as well as local missteps. They include Birmingham, Nottingham and Woking.
The Audit Reform Lab at Sheffield said the average cost of external auditors checking a local authority’s finances was now at least £50,000 higher in cash terms than when the Audit Commission was disbanded in 2015.
Private-sector accountancy firms took over the job of auditing local government accounts in England after the agency was abolished, in an austerity-driven push by Tory and Lib Dem ministers to find savings and efficiencies.
“Ten years on, however, it now seems clear that these reform ambitions have failed,” the authors of the report wrote.
“Only 1% of audits were delivered on time in 2022-23, with many audits delayed by several years. Audit costs have risen dramatically in response. An unwieldy, but ultimately operational centralised bureaucracy was replaced by market chaos. The £100m per annum savings heralded by the UK government in 2014 are now a distant memory.”
The report found that average audit costs in England had more than tripled – an average increase of 238% – in the year to 2023-24. It blamed most of this increase on private-sector auditors hiking their rates.
It compared the large rise in audit costs in England to much smaller increases in Scotland and Wales, where it said there was a much stronger level of central oversight of private-sector auditing.
The take-home point for me is that there is no getting away from “bureaucracy”. It is the means by which humans organise their administrative affairs. It is a force of nature much like gravity. The job of effective surveillance and monitoring is necessarily expensive because it requires a degree of expertise that doesn’t come cheap. Inspectors are not useless back office bloat. They are an essential part of the system. There is no point in regulation if there is no enforcement.
Every prime minister since Blair has pledged a bonfire of quangos and a new government efficiency drive which leads to perpetual deckchair shuffling exercises, usually making things worse as they go, because they all start from the central premise that managing headcounts is the key to efficiency rather than looking at policy costs. What’s needed instead is a policy review of each sector, and to repair the dysfunction created by statutory obligations.
The important lesson I take from my recent works in this area are that if a party is pledging a “bonfire of quangos” then it’s reliable indicator that they have zero grasp of the issues, have done no serious thinking on policy, and have nothing original to say. They shouldn’t be trusted to run anything. The last thing we need is more witless meddling.