Reforming inspection was a distraction and always will be. School accountability needs rebuilding from the ground up, based on what actually matters for learners.
If I asked you to list the ten most important things you want your Year 6 leavers to have when they leave your school, would SATs results make the list? Similarly, for many, what would the answer be at the end of Year 11? Year 13?
Every year, our children leave us as confident, resilient, independent and articulate young adults, with digital literacy, a love of the outdoors, and a range of life and soft skills that will serve them long after they’ve forgotten what a subordinate conjunction is. SATs results wouldn’t make my top ten. I suspect, if you’re honest, they wouldn’t make yours either. And yet SATs are the bedrock of our accountability system. They feed into Progress 8 scores at secondary level. They generate the data that Ofsted uses to judge our schools. They determine the numbers on which careers are built and – as we now know lives have tragically been lost. So, let me ask a harder question: If the thing we’re measuring isn’t the thing we actually value, what exactly is the accountability system for?
Nobody argues against accountability
Let me be clear: I am not arguing that schools shouldn’t be held to account. Nobody in education argues this. Parents deserve to know their children are safe and well taught. The public deserves confidence that its investment in education is delivering. When Ofsted was created in 1992, it was responding to a genuine problem – significantly variable quality across local inspection regimes, a lack of transparency and no consistent way for parents to compare schools. That was a problem worth solving. But Ofsted in 2026 is not solving that problem. It is creating new ones.
One organisation. Every context. No credibility
Consider Ofsted’s remit. This single organisation inspects state schools, independent schools, children’s homes, fostering agencies, adoption services, childcare providers, apprenticeships, prison education, initial teacher training and further education colleges. It regulates early years provision and children’s social care. All of this under one Chief Inspector – Sir Martyn Oliver, whose entire leadership career was spent running Outwood Grange Academies Trust, a large secondary-focused multi-academy trust.
I lead a small rural primary school with around 100 pupils. What does a children’s home in Kent have in common with my school? What does a prison education programme share with an early years childminder in Cumbria? And what confidence should any of us have in an inspection system where the Chief Inspector has never led a school remotely like ours? This isn’t personal. It’s structural. When one organisation tries to be the expert on everything, it becomes the expert on nothing.
Regardless of what your personal opinion is, the data is in. Ofsted isn’t working
So, what does the research actually tell us about Ofsted’s impact? A University of York study found that Ofsted ratings are weak predictors of students’ educational achievement, wellbeing and school engagement. Separate research concluded there is little to suggest that parents choosing a ‘good’ school will see better outcomes than those who chose an ‘inadequate’ one: once you control for family background, the predictive power of Ofsted grades virtually disappears. University of Birmingham researchers found they could predict Ofsted grades for over 71% of schools based solely on the percentage of students with special educational needs, free school meal eligibility, and local deprivation – not on the quality of teaching or leadership. Recent Inspection reports with clear ‘copy and paste’ sections in the feedback add further criticism to the usefulness of Ofsted inspections.
Meanwhile, the evidence of damage is overwhelming. The ‘Beyond Ofsted’ research project found 76% of respondents thought Ofsted had a negative effect on retention, with 30% considering leaving after their most recent inspection. An NEU survey found 72% of teachers have considered leaving the profession because of Ofsted-related pressure – rising to 79% among primary teachers. The NASUWT’s wellbeing survey found teachers score 38.4 on the Warwick-Edinburgh Mental Wellbeing Scale, against a general population norm of 50 to 52. And 84% of teachers say Ofsted has a negative impact on teaching and learning itself – the very thing it claims to improve!
Then there is Ruth Perry. A headteacher of thirteen years. A school rated ‘good’ in every category bar one. A coroner’s verdict of “suicide, contributed to by an Ofsted inspection.” The coroner described Ofsted’s claim that inspections could be paused for staff distress as “a mythical creature.” She warned of “a risk of future deaths if there is only lip service paid to learning from tragedies like this.” Yet here we are.
The reforms aren’t helping – they’re making it worse
Following Ruth Perry’s death, Ofsted launched the Big Listen (its largest ever consultation) and promised fundamental change. But two former senior HMI, Frank Norris and Professor Colin Richards, were so unconvinced by the questions Ofsted chose to ask that they launched the Alternative Big Listen. Their concern was blunt: Ofsted would “try to obfuscate and show themselves off in the best possible light.”
The contrast in findings is damning. Ofsted’s Big Listen found just 41% of professionals trusted the inspectorate. The Alternative Big Listen found nine in ten respondents deemed Ofsted not fit for purpose. When Norris and Richards ran a follow-up (the Alternative Big Consultation) to assess Ofsted’s proposed new report card framework, only one in ten thought the reforms were an improvement. Sixty-two per cent said they were worse. Ninety per cent rejected the five-point grading scale. Eighty-eight per cent deemed the report cards unfit for purpose.
So what did Ofsted do? It went from one single-word judgement to five grades across up to eleven evaluation areas, backed by an 80-page toolkit, multiple operating guides, and a series of webinars explaining what inspectors will be looking for. I once flippantly wrote that instead of colour-coded gradings, we should just use emojis. I think that joke was more serious than I realised. The point isn’t what symbol you use. It’s whether labelling schools, by any mechanism, is the right approach at all. If your accountability framework requires webinars to explain itself, the framework is the problem.
An NEU snap poll taken after the new framework launched found 88% of members don’t think it addresses the mental health impact and 94% don’t believe inspectors can reliably assess multiple areas in a single inspection. One headteacher said the new framework “has added to the workload and stress of Heads.” This isn’t reform. It’s multiplication of the same flawed logic.
So, what should we be accountable for?
Everyone agrees on accountability. Everyone agrees on safeguarding, that’s non-negotiable and should arguably sit with a separate, specialist process (that the Chief Inspector thinks happens annually with the LA…it doesn’t). Beyond that, the profession and the system need an honest conversation about what we’re actually trying to achieve for children.
Confidence. Resilience. Independence. Curiosity. Digital literacy. The ability to articulate an argument, work in a team and recover from failure. These are the things that matter: None of them appear on a scaled score.
I won’t pretend there’s a neat, simple, one size fits all answer (and the idea of a working single approach, answer, solution…whatever, is the biggest issue in education) answer. The what needs genuine agreement with the profession, not a consultation that avoids the hard questions. The how and when need to be co-designed, not imposed. And the honest truth is that no single model will satisfy everyone. But the current approach, where an organisation that inspects everything from prisons to playgroups assigns colour-coded labels based on a two-day visit, is demonstrably failing children, failing teachers and failing the public it claims to serve.
Stop reforming Ofsted. The problem isn’t the inspection framework. It’s that inspection, as currently conceived, is the wrong mechanism entirely.
