Accountability is a circle

At the Festival of Education back in July, Chris Passey and I spoke about professional development in schools needing to have the cultivation of agency at the heart of its design. Part of this is about schools taking a healthier view of accountability, a more collective view. In this blog, I want to outline a bit more detail about precisely what I mean by that.

Accountability has a long history in education. In recent years, we have seen a significant increase in the amount of autonomy schools have (although, any form of this autonomy rarely seems to filter its way down to classroom teachers). With this autonomy comes an expectation that schools will drive system improvement (Gilbert, 2012). Since Christine Gilbert wrote that National college of school leadership report, we have had curriculum changes, three different Ofsted inspection frameworks, the introduction and removal of performance related pay, the removal of assessment levels in Primary and the introduction of the 1-9 assessment system in Secondary. Year 2 SATs have been removed, the Year 4 Multiplication Tables check has been introduced, as has the Year 1 Phonics screening check and the Reception baseline assessment has been introduced. With such significant change, it has become almost impossible to prevent accountability from slipping into regulation (National college, 2012). Accountability is almost always now about an outcome and, where an outcome is less than desired, it becomes about the extent to which the ‘agreed’ procedure was followed. Accountability, it seems, has become a case of creating procedures so that the burden of blame can be evidentially placed at one door, thereby keeping it from others.

Before we get more into the realities of today and the hope I have for tomorrow, lets think about how we got here. Prime Minister, James Callaghan spoke about the need to open the ‘secret garden’ of education in 1976. By this, he meant that much of what went on in schools was hidden from public view and he wanted those who had a vested interest in the quality of educational provision to have some kind of visibility of the standards within schools. Then, in 1988, we had the education reform act. This act brought in the national curriculum and national testing and was followed shortly after, in 1992 by the birth of Ofsted. Ultimately, the act centralized government control over assessment and curriculum systems. This was seen at the time as a mechanism that made it possible for Governing bodies and Headteachers to have more control over the running of their schools. It probably did. But it also created a competitive environment. With reported assessment data and inspection outcomes, by the early 1990s, schools were in very real competition with each other and so the obsession with those publicly available headlines began.

The trouble with outcome driven systems is that people try to create a procedure to generate the same results over and over again. On more than one occasion in my career, I have been asked to share the Year 6 ‘formula’. People were disappointed when I replied that you need to put good teachers in front of students and support them with good behavior systems and headspace to think. You could be forgiven for thinking that proceduralising success is always a good thing. Indeed, it often is a good thing. Let’s think about when that’s true. Procedures are important when something is going wrong. Gary Klein (2009) has a chapter about procedures in his excellent book Streetlights and Shadows. In it, he writes about Peter Pronovost, a professor from John Hopkins University and how he shrunk infection rates in a hospital from 11% to 0 inside a year, simply by creating a check list of procedures for staff to follow. This worked because the think that was leading to the high rate of infections was the failure to follow guidelines that were already in place. What Pronovost did was codify the guidelines. Crucially, the guidelines were around a task that was highly predictable – maintaining cleanliness. The same input (washing hands, cleaning patient skin with antiseptic, etc) would produce the same output (absence of infection) every time. How often is the same true of school systems?

I argue that school systems are highly unpredictable. I can teach the same lesson to two Year 6 classes on the same day and have wildly different outcomes. I can deliver the same Professional development session to two groups of teachers and again, find very different outcomes. There are so many unseen variables in both these cases. Prior knowledge, mood on the day, what students or staff have been doing directly before the lesson or PD session, what else they have going on in their heads at the time. These highly unpredictable beings (people) are also interacting with each other and responding to each other in unpredictable ways. The situations I describe are messy, busy and unpredictable and ultimately, complex, (Cilliers, 2000). So, procedures work in routine situations but we can’t rely on them in complex ones.

Why do we continue to rely on procedures in complex situations if there is clear evidence that they don’t do the job well in these cases? I think the simple answer to this is accountability. When systems are unpredictable in terms of their outcome – will Year 6 perform as hoped on their SATs? Will that subject leader say all the right things to the Ofsted inspector? Our instinct has become to protect the role that we have played in the process. The easiest way to do this is to hide behind the procedure. ‘I did my bit and I have evidence to show that X person didn’t do their bit very well so the accountability is with them…’ We have come to value procedures so much that celebrated leadership has come to be the ability to prove to the person one layer above me in the system that I can get the person one layer below me to do something that I have asked them to do. Essentially, I can get someone else to do their part of the procedure. By adopting a hierarchical approach to accountability as described here, we promote the blind following of procedures in complex situations that require skilled judgement to have a chance at achieving the very best outcomes.

Klein, (2009) also recounts a conversation he had with two fighter pilots where they had been tasked with writing a procedure for new pilots to follow. They believed that the procedure they had written would help pilots ‘get the job done, but not very well’ (p. 17). I think the same is true of the way we use procedures in schools. We are using them, in part, in an attempt to limit poor judgement at the expense of facilitating skilled judgement. We are also using them so that when something does go wrong with the outcome, we can very easily pinpoint the part of the procedure that was not followed closely enough and we can attribute blame to the person responsible. Our current system of accountability has everyone accountable to the person one rung above them in the system so that this person can prove to the person one rung above them that they are effective because they have managed to get the person one rung below them to complete their part of the procedure. This limits the potential for highly skilled judgement and it is a ticking time bomb. Eventually, something will not go as predicted, because we are largely dealing with complex systems in school and when it does, someone will be to blame, not as a result of poor judgement but because of the way the system has been set up.

What can we do about this? I think we need to re-think the shape of accountability. I have set out above that it currently works in a hierarchy. Healthy accountability, should, I believe, be a circle. In the middle of the circle should be the outcome we collectively want. Around the outside of the circle sits everyone who has a hand in working towards that aim. Everyone around the circle is accountable to everyone else for doing their part. Rather than me being told by the person above me that I didn’t do my part of the procedure well enough to enable them to do their part, every rung of the ladder can now see everything that everyone is doing or isn’t doing to contribute to the shared goal. This increases vulnerability and that is the key to developing trust. Making responsibility and decision making visible in this way starts to break down the need for self-preservation.

The second thing we can do is, when something goes wrong, we should treat it as a product of judgement not a product of negligence. We should assume that at some point, we will make the wrong judgement call because the systems we are dealing with in schools are so unpredictable. We should also treat those judgement calls as products of the system dynamics that we have created. Therefore, if person X has made a judgement that could have been better, the likelihood is that they made this judgement because of the influences of the system and the dynamics of the team leading that system. In effect, we aren’t interested in the fact that they made the wrong judgement call, we are interested in the dynamics and conditions that led to that judgment call. We know that judgement in complex systems is intuitive, (Nonaka, 1994). That means that the person making the judgement is likely consciously unaware of why they are making the decision that they are and so we can only understand why it was made and improve it by examining it together.

If we assume that judgements that lead to undesirable outcomes are sometimes due to team and organizational dynamics then responsibility for those judgements is shared. Accountability, in this case, must also be shared but it also becomes something that drives learning. If we examine a judgement and work out the factors that influenced it, then we can ensure that they system improves, as well as the individual. This is similar to what Edmondson, (1999) found in her study of surgical teams. She found that the most successful teams were the ones who talked about the mistakes that were made as a team. The least successful teams were the ones who didn’t talk about them at all. Crucially, the most successful and the least successful teams made similar numbers of mistakes.

Accountability can’t be about attaching blame to the individual most closely situated to the point of failure. If it is, it’s almost always a game of chance. It’s like playing pass the parcel with high stakes accountability. Mistakes are inevitable, undesirable outcomes are inevitable. What’s more is that this is true for the very best teams in the same way as it is for less well-performing teams (Edmondson, 1999). We are kidding ourselves if we think identifying the point at which the procedure wasn’t followed in a complex system is useful for anyone. Instead, we need to see accountability as collective inevitability and we need to get round the circle and learn from the mistakes when they inevitably happen. Accountability as a circle requires everyone to get comfortable with vulnerability, but ironically, it makes every individual less vulnerable to the blame game of accountability as we currently know it.

References:

Cilliers, P. (2000) ‘What can we learn from a theory of complexity?’, Emergence, 2(1), pp. 23–33. doi:10.1207/S15327000EM0201_03.

Edmondson, A. (1999) ‘Psychological safety and learning behavior in work teams’, Administrative Science Quarterly, 44(2), pp. 350–383. doi:10.2307/2666999.

Education Reform Act (1988) Legislation.gov.uk. Available at: https://www.legislation.gov.uk/ukpga/1988/40/contents (Accessed: 26 August 2025).

Gilbert, C. (2012) Towards a self-improving system: The role of school accountability. National College for School Leadership. Available at: https://assets.publishing.service.gov.uk/media/5a7d88b7ed915d3fb959457e/towards-a-self-improving-system-school-accountability-thinkpiece.pdf (Accessed: 26 August 2025).

Klein, G. (2009) Streetlights and shadows. United Kingdom: MIT Press.

National College (2012a) The new landscape for schools and school leadership: Seminar report. Nottingham: National College for School Leadership.

Nonaka, I. (1994) ‘A dynamic theory of organizational knowledge creation’, Organization Science, 5(1), pp. 14–37. doi:10.1287/orsc.5.1.14.