Get a cuppa ☕ (11 mins read)

This piece draws on conversations with two senior contacts in the sector. R spent fifteen years inside Ofsted as a Senior HMI and was involved in developing inspection frameworks. J has spent thirty-five years in residential practice and now runs a large network of professionals.

I paraphrase their contributions rather than quote them directly. Both are clear that Ofsted is needed. The argument here is not that it should be abolished. The argument is that it cannot do many of the things blamed on it, and the political contract that produced today's Ofsted is the part nobody in government will touch.

I keep coming back to that line.

Most public commentary on Ofsted lands one of two ways. Either the inspectorate is the villain (the regulator that destroys good providers and protects bad ones), or the inspectorate is the system (judgements are correct because Ofsted said so). Neither is true. Both miss the same thing.

Ofsted is two organisations stitched together. The first was set up in 1992 to inspect schools. The second arrived in April 2007, when children's social care was lifted out of the Commission for Social Care Inspection and dropped into the schools regulator, with 264 staff and a different professional tradition. The cultural DNA never converted. Eighteen months later, a child died in Haringey, a politician sacked a Director of Children's Services on national television, and Ofsted hardened into the institution we have today. It did not do that to itself. It was engineered to do it, by politicians who needed a place for blame to stop.

Eighteen years on, the reforms are coming. Single-word grades for schools went in September 2024. Single-word grades for local authority children's services come out in 2026. Ofsted has published its first standalone Areas of Research Interest. The Big Listen quarterly monitoring report ticks through 132 commitments. All of it directionally right.

And ministers, in private meetings with sector leaders, have said clearly: Ofsted itself is not on the table.

This piece is about why.

The foundational year is 1992. The Education (Schools) Act 1992 creates the Office for Standards in Education. Major government. A response to inconsistent local-authority-employed school inspectors and the question of their independence from chief education officers. Chris Woodhead is the first Chief Inspector, 1992 to 2000, confrontational by design. The remit is state-funded schools. The regulator reports to Parliament, not to the Secretary of State.

That is the original Ofsted. A school inspectorate.

For the next fifteen years, children's social care is regulated separately. The Care Standards Act 2000 creates the National Care Standards Commission (2002 to 2004), which becomes the Commission for Social Care Inspection (2004 to 2009). CSCI regulates children's homes, fostering, adoption, boarding schools, and adult social care.

CSCI's culture is the one J keeps coming back to. Inspectors arrived at a home, looked at what was happening, and asked: this isn't working, have you tried this? They were a regulator, but they were also part of the team. They held independence. They also held experience.

Then on 1 April 2007, that culture stopped applying to children's social care.

264 staff transferred from CSCI to Ofsted. Children's homes registration, fostering, adoption and boarding schools moved across. Two years later, on 31 March 2009, CSCI was wound up entirely. Adult social care and health went to the brand-new Care Quality Commission. Children's social care stayed at Ofsted.

The decision to give children's social care to a school inspectorate was political, not technical. Different professional traditions (school inspection vs social work inspection), different evidence bases (event judgement vs longitudinal practice), different unit of analysis (the institution vs the child plus their placement history plus their family). These differences were named at the time. They were noted, and the merger went through anyway.

The institution that absorbed children's social care in 2007 was still, fundamentally, a schools regulator. It had education-trained inspectors. It used education-style accountability mechanisms. It produced single-word judgements that compressed complex provision into one of four labels, and it did so on the basis of one visit, one moment, against a written framework.

Eighteen months after the transfer, a child died in Haringey.

Peter Connelly died on 3 August 2007, aged seventeen months. He had had over sixty contacts with Haringey children's services. By November 2008, three adults had been convicted of causing or allowing his death.

Sharon Shoesmith was Director of Children's Services at Haringey. On 12 November 2008 she gave a press conference saying social services were not to blame. That moment, J reminds me, was the one the country could not get past. The line that played and replayed: Ofsted said we were good.

Ed Balls, then Secretary of State for Children, Schools and Families, ordered an external inquiry. The report was delivered to ministers on 1 December 2008. Same day, at a press conference of his own, Balls used special powers under section 497A of the Education Act 1996 to direct Haringey to remove Shoesmith from office. He announced the decision before she had had any opportunity to put her case.

In May 2011, the Court of Appeal ruled that Balls's intervention had been intrinsically unfair and unlawful. She received a six-figure payout in 2013. By then, the institutional changes the political moment had produced were embedded in Ofsted, and they have not been substantively unwound.

What 1 December 2008 did to Ofsted, in J's account:

The Ofsted you encounter in 2026, when an inspector arrives at a children's home, is the one engineered in the wake of that political moment. It is the one that does not advise. It is the one that judges against a written framework regardless of what the framework misses. It is the one that defaults to defensive QA over collaborative practice.

It does that because the system was redesigned to absorb political risk.

The function the post-2008 Ofsted serves is not regulatory. It is political. When something goes wrong with a child in care, the question stops at the inspectorate. The minister is insulated.

A quick detour, because the institutional architecture we ended up with was a choice. There was another version available.

The Care Quality Commission was created at exactly the same moment children's social care was absorbed into Ofsted. The Health and Social Care Act 2008 set it up. It became operational on 1 April 2009. It took over CSCI's adult social care function, the Healthcare Commission, and the Mental Health Act Commission, all in one go.

CQC has had its own problems since. The Penny Dash review of October 2024 was unsparing. But the design choices made in 2008 are what is interesting here.

CQC's sponsor is the Department of Health and Social Care, not the Department for Education. CQC regulates regulated activities, not settings. Its operational language is "assurance" and "assessment", not "inspection" and "judgement". Its enforcement style is described in the literature as having "a different feel from Ofsted", more dialogue-led, more market-oversight oriented. The Care Act 2014 gave CQC formal market oversight powers for adult social care providers.

Ofsted has none of those for children's homes.

Two organisations, both regulating care, both created or expanded in 2008-09, with completely different design philosophies. The choice was real. Children's social care got the schools-inspection model. Adult social care and health got the assurance model. The distinction was never debated in public. It was a structural decision made by civil servants and ministers, in the year a sitting DCS was sacked on television.

What you see in Ofsted today is not what regulation has to look like. It is what one specific political moment produced.

Ofsted's statutory remit for children's social care is narrower than people think.

It registers children's homes under Part 1 of the Care Standards Act 2000. The registration authority is the Chief Inspector. DfE writes the regulatory framework, Ofsted runs the registration process within it. The framework is not Ofsted's. The decisions inside the framework are.

It inspects against the Children's Homes (England) Regulations 2015 and the published Social Care Common Inspection Framework. The standards are set in regulation. The inspection methodology is Ofsted's.

It enforces. Compliance notices. Restriction-of-accommodation notices. Prosecutions (in theory!).

It also inspects local authority children's services (ILACS), supported accommodation providers (the Outcome 1/2/3 framework), independent fostering agencies, residential special schools, initial teacher training, childcare and childminding.

What it does not do covers more ground than what it does. Worth being explicit:

What Ofsted does is write the inspection framework, run the inspections, and publish the grades.

What Ofsted does not do is everything else the system needs and is not getting.

There is a specific technical critique J raised in our most recent conversation worth flagging. I’d hesitate putting this down as ‘fact’ here… but the structure of the argument matters and is worth thinking about.

Children's Homes Regulations 2015, Regulation 5 places the provider under a duty to engage with the wider system on the child's behalf: education, health, therapy, contact, transition planning.

The provider becomes responsible for what the wider system does or does not deliver to the child.

Now compare it to what the Children Act 1989 says about a child on a child protection plan. The criteria for being on a CP plan are about being at "continuing risk of significant harm." A child can live with their family while at continuing risk, with social work involvement, on a child protection plan, perfectly lawfully.

A child living in a registered children's home, cannot be at continuing risk of significant harm. The regulation says they cannot be at risk at all.

Read those two together and you have an impossible standard. The threshold for staying in a children's home is stricter than the threshold for staying with a parent. The provider who accepts the child every other home refused, holds her, builds the longest stable placement of her life, that provider is operating against a regulation that says the child cannot be at risk in the home, while every external system (the CP plan, the CAMHS waiting list, the school exclusion process, the LA care plan) treats the child as exactly that.

An anecdote from one of the (many) conversations I have with resiliant directors from the sector:

A provider takes the child everyone else refused.

Builds a year of stability. An inspector arrives, applies the regulation as written, gives a Requires Improvement, later issues Inadequate, communicated indirectly via the Social Worker. The provider is now at risk, serves notice. The child moves to unregistered (£7k per week more) for less support. Ofsted later writes a letter of apology, retracts stance.

Reform that fixes single-word grades does not fix this. The standard the inspector measures against has not.

Reg 44 visitors are paid little. They have no professional clout. They are isolated. They report when they raise a problem in a children's home, they come under pressure from the LA running it (or contracting with the home) to soften their report.

IROs report being told: if you want to continue your job in this (local authority), you need to be supportive of the social work team. Again this is anecdotal… but that instruction, is illegal. Im inclined to believe it happens.

The corporate parent role sits on the LA in statute. The LA is the parent. It places the child, pays for the placement, and is meant to do continuous quality assurance of where its children live. In practice, LA quality assurance has been hollowed out. Devon c.100 children's homes, a few commissioners. Whatever those commissioners do, it is not the role.

So, to recap my undersatnding:

There was an Ofsted-led standards expert working group last year. The group was discussing whose responsibility it is to ensure a children's home is providing good care.

Statute is clear: the LA.

But I understand there is not the appetite to move the funding from Ofsted to LAs to do the work… the institution does not advocate for redistribution of its own resources, even where statute says the LA should be doing the function it is currently catching.

Here is the line that is hardest to put down. J said it most directly, talking about why ministers since the 2024 election have refused to put Ofsted on the table for structural reform.

The political function of the post-2008 Ofsted:

The reforms that are happening are real and directionally right. The single-word grade removal for ILACS is a meaningful structural change. The April 2026 Areas of Research Interest is the first time Ofsted has publicly named what it does not yet measure. The 132 commitments tracked in the Big Listen monitoring report represent a regulator that is moving.

But the political contract that produced today's Ofsted has not been renegotiated. The institution still serves the same political function it was engineered for in late 2008. Ministers will permit reform of the framework, the language, the grading, the methodology.

They will not permit reform of the function.

The reason is structural. If the inspectorate is rebuilt to actually share regulatory responsibility with DfE and the LA, the next political crisis when a child is harmed lands at the policy owner and the corporate parent. That is the minister and the Director of Children's Services. Those are the people who currently sit one tier behind Ofsted, watching it take the blow.

Until that political contract is on the table, the architecture stays.

I have spent a lot of this piece taking Ofsted apart. Here is the thing I keep wanting to say, because it matters.

R, who built parts of the inspection framework from inside the institution, told me she got into Ofsted to make the system better for children. She still believes that is what most people in the regulator are trying to do.

J, who has spent thirty-five years in the sector and watched the cultural shift happen in real time, says clearly that without Ofsted, providers would do the bare minimum. The standards, even if imperfectly enforced, prevent worse.

The argument is not that Ofsted should go. The argument is that the public conversation, the political conversation, and the sector conversation about Ofsted are all having the wrong fight.

The questions that would change something:

These are not questions Ofsted can answer. They are questions for ministers, for DfE policy leads, for Directors of Children's Services. The reform programme inside Ofsted's existing remit is welcome. It does not reach the questions above.

J's line, again, because the rest of this piece is only legible if you start from there.

The 1 December 2008 moment produced the institution we have. The institution we have absorbs blame so the political layer above it does not have to. The reforms now under way are real, they are directionally correct, and they leave the political contract intact.

That is not a critique of the people inside Ofsted. Most of them, by the testimony of the people I trust on this, are trying to do the job. It is a critique of the political architecture above them, which has been designed for plausible deniability rather than for the children whose lives the system regulates.

An interesting piece woudld be on what the LA's corporate parent function should look like if it was actually funded and held to account… That is the conversation nobody at the political level wants. Which is exactly why it is the conversation the sector should be having!

I think you'll agree with me… Ofsted Requires Improvement. So does the political contract that produced it.

If you've worked inside Ofsted, in a local authority commissioning team, in DfE policy, or as a Reg 44 visitor, I want to hear what this piece misses, where it overreaches, and what you'd push back on.

If you're a children's social care lawyer and you can either confirm or correct the Reg 5 / continuing-risk reading (or anything else for that matter im getting wrong), that would be a very useful contribution to where this piece goes next.

Reply to this email or DM me on LinkedIn. Share with someone who might find it interesting…

Marcus (LinkedIn)