Inspection reform: lowering the stakes

Ofsted reform is very much on the agenda. Rightly so. The impact of inspection is now one of the factors contributing to the decline in morale of the teaching profession and, unintentionally, to our inability to close the gaps in outcomes between the rich and the poor.

Much of this debate focuses on the design of Ofsted – the framework for inspection, the grades, the frequency, the language. These matter. I have views on them – particularly the grades. But, by itself, adjusting the design of Ofsted will not solve our root problems.

This is because of a peculiar feature of accountability. The consequences of a system of accountability overwhelm the design of a system of accountability. Thus, the consequences of inspection will overwhelm the design of inspection.

Imagine a poorly designed accountability system (looking at the wrong things in the wrong way). That isn’t great but, with weak consequences, there is still space in the system for people to do the right thing – to ignore the findings or work around them. Equally, even a well-designed accountability system, when matched with extreme consequences, will produce bad outcomes. This is because it will encourage gaming and compliance to what is measured. And what is measured can never reflect the nuanced reality of the work. This is Goodhart’s Law, which says that when a measure becomes a target, it ceases to be a good measure.

It really matters what happens as a result of inspection. Until this is addressed, other sensible reforms will be wasted. Imagine, for example, dropping a single overall grade and replacing it with a basket of measures – which could well be a good idea. In a high stakes system, what will happen next is that someone will take an average of that basket of measures and use it to trigger intervention. Or they will pick one of those measures as more important than the rest and use it in the same way. The resulting system will be even worse than what came before as it will be equally punitive but less transparent. 

So, we need to start with the consequences of inspection. And this creates an extra problem. For Ofsted is not wholly responsible for the consequences of its inspections. Ofsted inspects. Other people make decisions about those results – Regional Schools Commissioners, MAT directors, governing bodies, even head teachers themselves. In one sense, Ofsted has the consequences that we give it. 

I don’t think this lets Ofsted off the hook. It knows what context it operates within. It knows what people do with its findings. It could and should adapt itself to that reality. But it is clear that we have to look at accountability in the round, and all of the actors within the system, rather than inspection in isolation if we want to make real progress on our problems.

This also suggests that the relationship between design and consequences is not as simple as I initially made it seem. Yes, the consequences overwhelm other features, but there is a relationship between the two. Knowing the consequences, we might wish to make changes to the design. And knowing the design, we might want to alter the consequences.

So, if we start with the consequences of inspection, and any design adjustments that reflect that, what follows?

I believe the stakes of inspection are currently too high. This demonstrably generates high anxiety in school leaders, which can in turn lead to punitive working environments, over-compliance with perceived expectations, and a significant reduction in innovation. We are creating incentives for a short sighted, bureaucratic, conservative and overloaded system. And the next wave of school improvement needs patience, flexibility, courage and simplicity.

One response could be to remove all consequences from inspection. Either by ceasing it altogether or decoupling it entirely from accountability. The problem with this is that we do need some way to hold schools accountable and, if inspection has no role in this, we are left with exam results as the only measure. This is dangerous, as exam results do not necessarily reflect the quality of a school. There does need to be a qualitative judgement about a school’s performance somewhere in the system.

I do remember what schools were like before the current inspection regime, and all was not wonderful. To put it bluntly, there are some schools that should be worried about an inspection. And there are many who shouldn’t give it a moment’s thought. At the moment, everybody is worried. Our first task is to target the effects of inspection better.

Who shouldn’t be thinking about Ofsted? The leaders and teachers of great schools, for one. This doesn’t mean exempting outstanding schools from inspection. It means not having an outstanding grade. The event that triggers negative consequences is a change of grade. The ‘good’ grade is very wide. Without an outstanding grade, there would then be many schools at the top end of it, who are not immune to inspection but who would have very little fear of slipping a grade. 

I think there are other good reasons for not having an outstanding grade. I don’t think excellence is amenable to regulatory judgement. It is relatively straightforward and objective to identify failure: I think we’d mostly agree. Excellence is more subjective and idiosyncratic. Think of it as a sort of ‘reverse Karenina’ – all unhappy schools are the same, but all happy schools are happy in their own unique ways. 

This does not mean that schools shouldn’t aspire to excellence. It means that we need different ways to validate that – preferably a diverse range of schemes that measure and accredit excellence.

A head teacher who has just taken on a new school shouldn’t be worrying about inspection either. They need time to make a difference – and there is often disruption and disagreement when changes are made. A new leader should not be looking over their shoulder. With the exception of routine safeguarding checks (see below), I would recommend a policy of ‘space to improve’ where a school is exempt from inspection for three years when a new head teacher is appointed. 

Luck should not play a significant role in inspection. There is no way that inspections can be fully objective. There is an element of chance. Inspectors have their subjective priorities. And every school makes mistakes, gets things wrong. For some, this happens out of sight, for others it happens during an inspection. This is merely a matter of luck: a child does the wrong thing in the playground while the inspectors are present. This sense of capriciousness contributes more than anything to the anxiety of school leaders and drives massive over-compliance and gold plating of procedures, which in turn increase workload and drive turnover.

There is a way out of this too. We split inspection into two. Firstly, an infrequent, in-depth nuanced review of the educational and pastoral quality of the school. Secondly, a frequent, fast and objective review of the administrative quality of a school, including financial and safeguarding procedures. Unless students were found to be at immediate risk, this second review would operate a bit like an MOT: if mistakes are found, the school is given a modest window to implement improvements before a return visit. If the school is unable to make those changes in that timescale, or keeps repeating them on subsequent occasions, then that would be cause for more serious concern.

This change would eliminate a lot of anxiety. It would also help increase the frequency of safety-related aspects of inspection, which are too infrequent now. The education reviews could be designed to be more supportive – possibly even incorporating an element of peer review.

There are other ways to lower the stakes. For example, we can also reduce the role of luck by focusing on inspection quality. This revolves around the selection, development, support and management of the inspection teams. There are many sincere, experienced and able inspectors. And participation in inspection can be helpful professional development for serving school leaders. With lower adverse consequences, we can accentuate these features. I would also consider removing the role of parental feedback in inspections. Parents now have many ways to make their opinions known, and schools cannot be customer service organisations and still steer a steady course.

There are no doubt other ways to manage the stakes and strike the right balance between accountability and confidence. A big part of this will be the choices that other actors in the system make and what they do with inspection results.

If we get the right balance, we may discover that we don’t need to revise the inspection framework and introduce another set of radical changes for schools to cope with. If I was a policy maker right now, in a system with as great deal of stress and not a lot of resource, I’d be thinking more about what I can remove than what I can add or change.