This is an approximate transcript of the talk that Ben White and Becky Allen gave at researchED Kent on Saturday 30th November, 2019. It is a little light on references and grammatically requires improvement – sorry… we’ll try to tidy it up soon!

---

I (Ben) am not sure if it’s possible to win teaching. For a long time I expected to be able to. This expectation was clear in the narrative around teaching into which I was initiated as a trainee. It was frequently compounded by direct advice which I’ll paraphrase as follows:

You should be doing X. We won’t tell you exactly what it involves in your lessons. But we may occasionally check if you’re doing it… Also, it has to work.

At different points I have been told that monitoring PP, embedding SEALs, careful differentiation, responsive teaching, generating a buzz, and forensically analysing attainment gaps would do it. Often it felt that these ‘solutions’ were never fully unpacked – what exactly was I needing to do? 

Teaching is inherently complex. It is harder when we are tasked with enacting abstractions or simple models that are never explicitly defined and hence often fail when translated to the chalkface. You, no doubt, already know this. You will have learnt it from experience. However, it can feel like whoever is pedalling the policy cycle does not. We teachers are often tasked with implementing solutions premised on a vision of teaching which is far simpler, less variable and more impactful than we know it to be. 

Perpetual waves of disruption rise, fail to deliver as promised, crash, recede, and are replaced by the next.

We have perpetual disruption from politicians, who can’t work out how to control the education system. Perpetual disruption from education consultants who cycle through a series of arguments as to why schools are dysfunctional. Perpetual disruption from headteachers and executive heads, who spot good ideas in distant schools then struggle to make them work in their own. Perpetual disruption from teachers who keep trying to solve the riddle of how to ensure every child succeeds.

The complexity of complexity

There are three reasons why we can’t work how to control the system we find ourselves in: measurement; complexity; and spheres beyond our control.

We can rarely agree about what it really means to learn something because we cannot see inside the minds of our students to measure the types of changes we are trying to evoke in their brains. The invisibility of learning forms part of the challenge of education and is at the heart of the reason why we cannot decide:

• Whether an instructional technique is effective;
• Which teachers and schools are worth copying (see also ‘Nobody knows which schools are good’)

If we could precisely measure learning, we would have some hope of making sense of how to teach an individual child.

But it is a lot more complex than that for the classroom teacher, for they have to teach something to to 30 brains at once, each of whom will learn different amounts from the same episode of lessons.

Why? Here is where spheres outside our control become a challenge for us.

1. These 30 brains are different from each other in ways that are currently poorly understood, though we know more than we once did about this variation in cognitive function.
2. These 30 brains will have enormous differences in prior knowledge that is relevant to the lessons and that is intrinsic to their ability to develop schema of understanding. We can try to control variation in prior knowledge through the curriculum, but we’ll almost certainly fail for all the reasons that Graham Nuthall so eloquently described in The Hidden Lives of Learners.
3. These 30 brains will go home at the end of lessons and learn a great deal that is related to class, through formal homework or other means. Remember, children spend less than a fifth of the waking hours of their childhood at school. It can be hard for teachers to appreciate we are just a bit-part in their lives.

And so, if many students appear to make good progress, almost regardless of how we choose to teach, whilst others struggle no matter how hard we try, it is understandable that we cannot make sense of what works.

There is no perfect way to reconcile the diverse needs of 30 children in one classroom, in one hour, with one teacher.  This arbitrary group size and classification by year-groups compound the difficulties we already have when seeking to educate a single individual. 

As the US sociologist, Seymour B Sarason, says:

[A graded school system,] taking a new crop of children every year at five [to six] years of age, moving them through their studies in “lock-step” fashion ‘til graduation, makes an assumption about the equality, motivation, and performance for children of similar age that the reality of individual differences rudely challenges.

in The Culture of the School and the Problem of Change

It is almost inevitable that after each hour of instruction, these thirty individuals with their differences in motivation and capabilities will show variable ‘outcomes’. Not least because the classroom is a very different place for different students (another insight conveyed in Nuthall’s meticulous analysis of the classroom experience of different learners). Social status; relationships with peers and the teacher; parental attitudes towards school and expectations of behaviour; experiences earlier that day; variations in mood, tiredness levels, hunger – there is an endless list of variables which serve to make one lesson a unique mental experience for each student in the class. (The complexity of even a single moment in a single lesson is something we will explore in a later piece.)

For now, with the above in mind, it is hardly surprising that reducing variation in outcomes is something we rarely succeed in doing.

Fixes – such as removing children from classes for interventions and catch-up, do not fix the fundamental problem. Some children – for whatever reason – will need more hours of instruction to learn new things. 

As Bart Simpson sharply surmised: 

Let me get this straight: we’re behind the rest of our class and we’re going to catch up to them by going slower than they are?

Children within classes are so different in how they behave, but so are teachers across a school. They vary in how they respond to their classes, in part, because they are trying to juggle and balance trade-offs that are intrinsic to the nature of group learning and force them to grapple with notions about the purpose of education that can only ever produce oblique or ill-defined goals.

They also vary in how they respond to classes thanks to habits of the classroom developed in bygone eras that cannot now be easily disrupted. And because they hold a diversity of belief systems – stances towards behaviour and authoritarianism, beliefs about effective pedagogy, and so on – whenever you hear someone say “but no teacher believes that…”, I can guarantee that they do…!

Substantive complexity in the nature of what schooling is and the framing of its purpose yields diversity of practice. A complex system cannot ever be understood in terms of deterministic mechanisms and pathways. Complexity is better thought of an intractable knot that nobody can undo, even if they work on it for a lifetime. The best we can hope for is to make it sit a little better. This is why the problems of complex systems – such as making schools better or closing inequalities in attainment – are often called wicked problems.

The allure of the perfect solution to a wicked problem

The significant differences in how much different students know when they leave school represent a ‘wicked problem’. Like all wicked (or complex) problems, it can be framed in a number of ways, without a consensus as to its precise nature.

It is impossible to fully solve, and the essence of the problem and plausible solutions to it appear differently to every beholder. Any single summary of a wicked problem will imply a particular solution. This solution will be more or less alluring to different actors for a range of reasons. Because we do not agree on the essence of the problem, neither are we likely to agree on how it should be tackled.

Each definition therefore smuggles within it a solution. The way in which it is framed tends to highlight a root cause. This implicitly points towards a course of action most likely to tackle said cause.  Notice however, that relatively vague definitions (and their implied solutions) leave plenty of room for disparate applications.

For example: Why is there an attainment gap in this (or any) school?

• Response 1: ‘Some students are not trying as hard as they could’. The implied solution here would centre on changing aspirations/mindset/motivation etc, perhaps via an emphasis on growth-mindset, a rewards system, closer monitoring, the use of carrots and sticks, parental engagement or a combination of these. 

• Response 2: ‘Teaching is not tailored to the needs of individual pupils’. Solutions here might involve independent project work, differentiation by task or by outcome, or an emphasis on the use of responsive teaching techniques.

• Response 3: ‘Some students have limited cognitive ability compared to their peers’. Here solutions might involve setting, streaming, increased instructional time or varying the curriculum offer for students in other ways.

As teachers, we will each frame the complex problem in a slightly different way, which in turn reveals our solution. Some solutions are particularly alluring to teachers, for a variety of reasons.

Firstly, they appeal to our own imagined range of actions. They chime with what we imagine we can do in our job. It makes sense that we would gravitate towards solutions which emphasise the efficacy and importance of that which we spend most of our working hours doing. Or, as Maslow observed, rather more pithily  ‘I suppose it is tempting, if the only tool you have is a hammer, to treat everything as if it were a nail.’  It is very likely that we exaggerate the effect (and utility) of our own particular brand of teaching hammer!

Secondly, they contribute to the meaning which many of us derive from our work. To some extent an alluring definition tells a story of a crisis or problem which we teachers can heroically solve. As Victor Frankl observed, humans are creatures who require a why as well as a how to give meaning to our daily labours. It would follow then, that stories which portray me in the role of shining knight tackling a vitally  important problem would appeal to my need for a meaningful and significant role in the world.

Thirdly, at times of frustration, we look to other factors or agents to explain away the problem. Given the intractable nature of the problems which we seek to resolve, we will often feel a little despondent. This may manifest itself in the moment of a lesson where our 30 students are not progressing as we need them to, or over the longer term when we notice (or are told) that we are not ‘impactful’ enough. We comfort ourselves by looking to explanations outside of our remit.

Teachers and school leaders also regularly interact with, and are tasked with enacting, solutions borne of the ways that policy makers and other key agents within the education system frame a complex problem. Here, it is worth noting that policy makers face a broader facet of the same complex problem which frustrates teachers. As such, the same dynamics, biases and frustrations just explored will also apply to them. However, as their vantage point and imagined actions are different, so is the way in which they diagnose and seek to resolve the problem.

Policy makers will favour solutions which propose plausible positive policy-based means of tackling the school improvement problem. From time to time, particular definitions of the problem come to dominate. They are endorsed at policy level, but also by a range of others – a kind of umbrella advocacy.  They have a limited life-cycle, for reasons which we will explore later. We call these perfect solutions.

Perfect Solutions tend to: 

• Frame the problem in a particular way, implying a general area in which solutions or changes are needed
• Stop short of saying exactly what this will look like at school or classroom level. More akin to stating ‘Brexit means Brexit’ than they are to a fully articulated multilateral trade-deal!
• They are presented as capable of achieving significant progress with, if not fully solving, the attainment problem. 

The policy-maker excels in an articulate exposition of the policy problem, but with far less sophistication in their description of the planned solution!

Perfect solutions can be reduced to a short story – the sort that can be quickly told, grasped and recounted by those under pressure to make inroads, at policy level, into a complex social problem such as the socio-economic attainment gap via school improvement. 

The problem with schools is this… …and it can be resolved (or at least significantly improved) by this widespread (policy level) change…

Perfect solutions never achieve their goals because a wider problem in a complex system is necessarily a wicked or intractable problem. Because it is both essential to solve and yet intractable, we find ourselves on an eternal search for perfect solutions to wicked problems.

Like waves, perfect solutions tend to rise just as their predecessor is is receding. 

The data wave

Let’s go back to the year 2000. We are standing on a beach watching the gentle ripples of the tide ebbing and flowing. It is peaceful, but you – like everyone else in the education system – is about to be swept up by an enormous tide that will be all consuming for a while, but will eventually drop you off onto another beach again. What is causing the wave to rise before us?

There are always people with specialised interests and expertise within complex systems and there was nothing unusual about the data crunchers who were situated in local authorities in the late 1990s. They find themselves with hundreds of thousands of data points on the students they educate thanks to the new Key Stage tests and see this as an opportunity to find out which of their schools is providing a good education through value-added analysis, rather than which simply have affluent intakes. A worthy endeavour. The Department for Education is keen to get in on this data game and so contracts someone to create and licence for use the National Pupil Database. The data revolution is about to begin!

However, it takes more than a bunch of techie data guys to create the dominant narrative in education policy. For that we need agents – individuals and institutions that represent an umbrella advocacy – who have a wider problem in need of a solution. What was the problem that the umbrella advocacy felt data could solve? Teacher accountability? School accountability? Pupil predictions? Closing in-school attainment gaps? (Of these, the latter became the one that generated the most bizarre set of policy responses.)

Who were the umbrella advocacy who pushed data analysis as the dominant narrative in education?

• Tech solutions companies, who had multiplied once DfE had licensed the data for wider use beyond those who run schools.
• Conferences, consultants, school improvement groups and other CPD providers, including the giant in the field, PiXL. They encouraged schools to create processes and roles centered on forensically tracking data on the progress of groups.
• Ofsted and DfE who saw data as a way to hold schools to account in a complex system and make them responsible for closing the attainment gap.

School leaders readily bought tech solutions and seats at conferences because the data management arguments seemed plausible, because other leaders seemed to think these were useful ideas and products (echoing the mimetic isomorphism described here), and in any case we had no other better ideas as to how to improve schools.

Before long, nearly all secondaries were downloading GCSE predictions for their students and using them in a wide variety of ways. Data-enriched managerialism became culturally established as a normal part of school leadership. However, it took the coercive forces of the Ofsted framework to really turbo-charge the data revolution by requiring all schools to comply with it, at least on the surface.

Ofsted’s 2016 inspection framework pointed towards the following as a key indicator of school quality, thus smuggling the perfect solution into the wicked problem of school improvement:

the progress of pupils currently in the school…in all year groups, not just those who have taken or are about to take examinations or national tests.

[Para 175]

It also stated what the data needed to say for a school to be deemed successful:

[Inspectors will seek to identify whether] pupils are set challenging goals, given their starting points, and are making good progress towards meeting or exceeding these.

[Para 179]

At this point, ‘peak data’, I (Ben) became an RSL, a Raising Standards Lead, for PiXL. The writing was already on the wall as people such as myself quickly became deeply frustrated with a ‘solution’ which was internally incoherent (in this case, statistically illiterate) and inadequate as a means of closing the attainment gap – as any school-based solution would be. (Ben talked about the practical/impractical aspects of this in a separate session on ‘The Uses and Abuses of Assessment‘).

Interestingly, interviews with leaders in schools at this time suggested that some were very aware of the limitations but felt largely powerless to publicly state this. This chimes with Rittle and Weber’s early work around ‘wicked problems’ in which they noted that often it is impossible for perfect solutions to ‘be wrong’ – publicly at least.

But… the system is about to start crumbling down from peak data in the mid-2010s. But why?

Well, the system HAS to collapse for the very same reason that it becomes popular. By framing the data system as a solution to a large-scale, complex problem that by definition has no solutions, the system’s days are numbered. Data doesn’t fail in terms of the objectives of the people who created the original ideas – the data crunchers. Data managerialism fails by the agenda of the umbrella advocates – Ofsted, government, commercial companies – who use it as a perfect solution to solve a wicked problem.

The voices of dissent against fixing education through data managerialism were persistent, though quiet at first. But some voices are more effective in bringing down a dominant narrative than others. Arguments that require individuals to disrupt deeply embedded belief systems don’t tend to cut through (because it’s too hard for us do this). Instead, it was those who were accepting the plausibility of the data managerialism, but demonstrating its shortcomings in implementation that most effectively chipped away at the policy:

• Right back in 2008, Warwick Mansell wrote in TES about strange GCSE predictions being generated in non-core subjects
• Daisy Christodoulou wrote about the meaninglessness of performance descriptors and other data in pupil tracking systems
• Tom Rogers was a consistent voice collecting stories of bizarre pupil premium data policies at work in schools

And so it starts to become self-evident to everyone that the system does not succeed in its own terms because of these credible explanations of how it cannot work.

It is important to note that this doesn’t mean that all data solutions were wrong, but rather that they weren’t able to solve the problems the umbrella advocates claimed they could. And they had a set of unintended consequences that hadn’t been carefully anticipated, not least because it is so hard to predict how people will respond in a complex system.

It would be nice to think we (Ben and Becky) played our part, but by the time we got involved the wave was probably already crashing. In talking about the pupil premium, I (Becky) asked how it can be that every single school’s impact statement has demonstrated they have ‘closed the gap’, yet the national attainment gap isn’t closing! And my talk on workload (now a chapter in The Teacher Gap) led to Damian Hinds asking me to convene a DfE Workload Advisory Group (which Ben sat on) to sort out data madness. The umbrella advocates who had caused this problem were systematically working to bring it down!!!

Before we knew it, suddenly the attainment gap was no longer about using data to make targeted interventions; it was just about good teaching. Who can argue with that! (Nobody… except for those who whisper that good teaching is unlikely to close the gap.)

We now find ourselves washed up on the beach, but not in the same place we started. It is quite a messy picture because whilst many conformed to the data-based perfect solution, they did so in different ways and for different reasons.  Also some didn’t – again for a range of reasons. These schools may have less rowing back to do!

Monitoring and using in-year data (an expectation under the prior Ofsted inspection framework) was never explicitly unpacked or defined. It was, however, made clear that a lack of adequate tracking was unacceptable. As a result schools started doing a range of things in order to meet this particular requirement. There remain large differences in terms of format, frequency, and intensity of the work around data capture and recording in English schools. The loose definition of this particular perfect solution, initially at least, allowed a plurality of responses to emerge.  

However, as Becky and Sam explore in The Teacher Gap, there is a tendency towards conformity (or isomorphism). Schools, or rather the teachers and leaders within them, probably conformed for different reasons. These underlying differences in motivation may well influence the rate and degree of change in coming months and years. 

External Conformity	Internal agreement	Type of Conformity
✔	✘	Compliance
✔	✔	Internalisation
✘	✘	Non-conformity

Some conformed externally, but never internally agreed that data managerialism was a good idea. We call this compliance. Those being compliant may just dispense with practices that are no longer mandated, provided they get (and believe!) the message that they can. And importantly, provided they haven’t gone native!

This leads us to another type of conformity – internalisation. Here we conform not merely to comply, but because we come to believe that it is the right thing to do. Those who internalised the principles underpinning the outgoing ‘perfect solution’ may find it harder to give up. It has become interconnected with their understanding of the core business of schools, leadership and/or teaching. 

Arguably, this internalisation could be institutional as well as individual – structures, roles, habits and routines can almost gain a life of their own in a school. If there are data managers; raising standards leaders; frequent report cycles; a culture of forensic data analysis and the various means of evidencing this; and a system built around identifying pupils in need of external interventions then it is unlikely that such behaviour will stop or even change dramatically overnight.

And then there are the non-conformists – those who never went along with it. They are often ‘market-leaders’ that can rely on established success to negate the need to comply. For example, in interviews, Ben found that selective (grammar) schools were most likely to describe data as useful but not integral to success.

Others are non-conformists simply because the message never quite reached them! Not all schools and school leaders are immersed in the national conversation. It can take a long time for common practices in some parts of the country to permeate others.

Other non-conformists simply took a risk and stuck to their deeply held values or principles. Regardless of the reasons, these schools who were not swept as far by the receding wave are differently placed than others to weather the next one.

But it’s not just about us (teachers and schools) – what about the umbrella advocates? What of all of those people outside of schools for whom data was serving a goal? (EdTech; conference providers; Ofsted; the DfE; providers of teaching and leadership qualifications; school improvement partners…)

Already this umbrella advocacy has shifted, for they need perpetual revolution to fix wicked problems. They are riding the next wave. Though they tend to do so without explicitly stating that previously endorsed practices/approaches were deeply flawed. Where do PiXL stand on RSLs vis-a-vis their later observation that ‘prediction is a mugs game’ for example? (Ofsted is an honourable exception since the change in Chief Inspector periodically allows them to ‘clear the deck’ and start afresh.)

The curriculum wave

New waves always emerge in the same way: the interests and ideas of specialist groups get taken up by the umbrella advocates as a perfect solution to a wicked problem. The curriculum wave is no different. Curriculum discussion and debate is always central to education policy-making in all countries. At any point in time there are subject specialists (as Christine Counsell and Michael Fordham are in history) and those who are making overarching philosophical or empirical arguments about curriculum (of which the take-down of generic skills was a notable one).

For some time, the American ED Hirsch and the British sociologist Michael Young had been (quite separately) making the case that educational inequalities amongst children were, fundamentally, inequalities in knowledge. This argument was like catnip to the umbrella advocacy who by the mid-2010s were in need of a new perfect solution. By re-framing the wicked problem as one of inequalities in knowledge, the solution proposed by the knowledge-led curriculum movement becomes self-evident. It is, what the US sociologist Seymour B Sarason calls an ‘untestable abstraction’.

The usual mix of coercive and intrinsic ideas make the notion of curriculum reform as the solution to attainment gaps appealing. They serve to accelerate it’s rise. 

External coercion first came via Nick Gibb and the new National Curriculum and GCSE/A level reforms. However, it has taken the arrival of Amanda Spielman at Ofsted with an inspection framework that places curriculum at its heart to really turbo-charge curriculum reform through the system.

That said, this curriculum focus also aligns with teachers’ intuitions about education. Internally, it appeals to our imagined realm of action – the classroom, and what we think proper teaching should be able to achieve.  It underpins our newly found consensus that other solutions – ill-defined interventions based on statistically dubious analytical methods for example – didn’t work.

By placing classroom instruction at the centre of the policy solution, it allows us to focus on curriculum design, resource creation and lesson planning. These are things that many teachers say they love to do and find inherently rewarding. And at present, the solution is defined broadly enough for a plurality of possible applications to emerge – everyone’s favourite version of curriculum renewal seems to fit within the remit of this perfect solution. This helps maintain appeal with a range of teachers and leaders.

This all marks curriculum out as rather different from the data wave – which generally appealed to school leaders and data managers more broadly than it did to teachers.

The popularity of curriculum reform as an instrument for school improvement today is clear. This widespread endorsement itself has contributed to it becoming the next wave, the newest ‘perfect solution’. The umbrella of those advocating for data as a perfect solution, is (or already has) largely realigned around curriculum reform, the next perfect solution.

This isn’t to say that everyone has happily jumped aboard the knowledge-led curriculum movement. Just as with the earlier data wave, quasi-mandatory coercion will pull many aboard and some won’t jump aboard at all.

It is clear that so many now feel that curriculum reform is our best hope for raising attainment or for closing the gap, so why do Ben and I feel so nervous about what we see? We are not here to criticise the idea of curriculum revision or improvement as a desirable pursuit in many or all subjects. Equally, we instinctively feel that a knowledge-rich curriculum is one that would benefit our own children. However, we are nervous to see it upheld as a unifying or perfect solution to a wicked problem that is, by its very complexity, intractable.

This is a movement that, whilst it harnesses cognitive psychology to make assertions about curriculum design, currently sits on pretty weak empirical evidence that it can achieve whatever goals it has in mind for itself. In any case, like all ‘perfect’ or unifying solutions, it doesn’t actually specify exactly what we should do or how we will know it has worked. Remember, it’s not about implementing ONE perfect solution; 20,000 different curricula could represent a perfect solution to a wicked problem. A poorly specified programme with poorly specified outcomes cannot ever be subject to systematic evaluation.

Where the Education Endowment Foundation has attempted to evaluate curriculum programmes (such as The Curriculum Centre’s Word and World Reading Programme), the trials haven’t signalled curriculum success (though this doesn’t upset the wave, of course). And the centre-piece of the empirical evidence in ED Hirsch’s book, a rough French data cross-tabulation, isn’t particularly convincing to empiricists such as Christian Bokhove (or to us).

And whilst enormous energy is expounded on creating the ideal curriculum, with aspirations for what students might learn tomorrow in relation to what they should have learnt in the past, so far less energy is invested in working out how this ideal curriculum gets enacted in reality.

For the moment it moves from a piece of paper to the classroom we will be rudely awakened by the complexity of the compromised curriculum where perfect knowledge mapping is displaced by 30 brains with utterly different prior knowledge – or at least recollection of prior knowledge. And the progression model where we hope to tick off the knowledge as it is learnt gets disrupted by students who inevitably ‘progress’ at different speeds, and make sense of identical lessons in a multiplicity of ways.

Furthermore, in setting it up as a solution to a widespread problem such as inequalities or school quality, the umbrella advocacy – Ofsted, CPD and conference providers, consultants – requires ALL teachers to attend to curriculum reform. This creates an ironic tension: that curricula expertise is necessarily subject specific, and yet we are pursuing generic solutions that fail to respect the cultural regularities and intrinsic knowledge structures of subjects.

School leaders who are required to implement curriculum reform have a tendency to homogenise, to establish conformity to a particular set of policies, across a school. Some strategies won’t work as effectively in all domains/disciplines. For example, knowledge organisers may be helpful in history – indeed they may underpin a helpful emphasis on the importance of establishing and learning core ideas to master a topic. However, in maths they may not add value and could replace a non-compliant, but more helpful strategy. The same is true for curriculum maps, retrieval practice quizzes, and other strategies being imposed on teachers.

‘Knowing’, ‘experiencing’, ‘learning’, ‘understanding’, ‘doing’ are very different things across different subjects which means that language that resonates in one subject domain (e.g. ‘entitlements to powerful knowledge’, ‘ambitious curriculum’, ‘teaching to the top’, ‘curriculum as the progression model’, ‘a curriculum that is inch deep and mile wide’) can be quite unhelpful in another. (And the distinction between those who teach sequential knowledge domains and those who teach hierarchical ones seems particularly important here.)

The place where these curriculum reforms genuinely risk seriously upsetting the status quo, for better or worse, is in ‘the afternoons’ in primary school. (I am using ‘the afternoons’ as a euphemism for the part of the day when pupils do something other than the 3Rs.) Afternoons can seem quite strange, particularly to secondary-trained teachers, and primary teachers often find it quite hard to articulate why they are the way they are. But things are as they are for a reason, even if that reason is… well… complex… and just because we cannot easily articulate the reasons for the status quo, it is critical that we try to fully make sense of it before we try to disrupt it.

Why? Well, disrupting a complex system means disrupting a set of habits, beliefs, resources and practices. If these are not fully understood you will inevitably underestimate the scale of the challenge and so curriculum reform will fail because it doesn’t change much. Or worse, it will fail and in doing so will make a stable, though very imperfect, curriculum into a highly unstable and chaotic one.

Four signs we should heed as signals that the curriculum wave is in trouble

1. A rushed curriculum as a perfect solution that looks worse than its predecessor curriculum. Doing complex things at speed and alone – e.g. overhauling whole schemes of work, lesson sequencing, etc – can lead to changes that are more dysfunctional than the status quo. This was arguably the case with some of the byzantine systems which schools implemented to replace national curriculum levels. It can also be seen in the equally elaborate systems of data tracking, recording and use which various schools and trusts developed. Could this happen with curriculum reform? And rather than accept blame, will the umbrella advocacy consequentially lead a call for a very explicit and mandatory national curriculum?
   
2. “You can have any curriculum you want, so long as you can justify it”. Different schools will fixate on different aspects of a solution and so across the system disparate responses tend to emerge. Some of these may directly oppose each other despite being implementations of the same ‘perfect’ solutions. Some curriculum reforms in schools already do. For example, one head of RE was recently asked to restructure her curriculum to focus on careers, as that was what ‘Ofsted wanted’. This entailed the creation of a scheme of work arranged thematically by different areas of employment rather than by categories or topics native to the domain of religious studies. The ‘Wedding at Cana’ miracle, where Jesus turns water into wine and which Christans read an indication of both his divinity and messianic status was moved out of ‘Christian Beliefs and Teachings’ and into an ‘Events Management’ themed unit!

3. A gulf appears between the intended, the enacted and the recollected curriculum. Much of the talk around curriculum paints me, the teacher as the shining knight at the centre of an epic – bringing powerful knowledge to my students. This may chime with my imagined range of actions and my need to be at the centre of a solution to the complex problem. But, as much as I’d like to ignore it, this doesn’t match most people’s recollection of their lesson. As US economist Bryan Caplan maps out in his, rather polemically titled, The Case Against Education: Why Education is a Waste of Time and Money, most people remember very little subject specific information from their time in school (bar basic numeracy and literacy) and, actually most of their lives are fine despite this. Even within a school career, I am skeptical that my year 11 pupils will remember as much of their year 7 subject content as would be required in order for them to gain the full intended benefit of a carefully sequenced curriculum.
   
   I may well not be achieving what I intentionally set out to do by overhauling my curriculum. The intended curriculum and the enacted curriculum may be so far apart that I cannot use the hoped for ends of the former to justify significant changes to both.  Will we end up doing things which, with hindsight seem silly? Equally, what important practices or processes will we stop because they do not fit within the narrative around this particular perfect solution?

4. We all create documents that ‘prove’ curriculum reform has worked in our schools! Accountability always causes distortions because the reforms need to be seen to work. This produces inherent tension or confusion. The Ofsted dynamic discourages nuanced self-evaluation. The confusion is worse when this tension is not directly faced. Rather, it is left to leaders and middle leaders to infer the extent to which honesty is the best policy when it comes to mandatory evaluation, analysis or data entries. Just as with school data which many of us were required to massage or straight falsify data in order to say what it needed to say, the same will likely happen with the paperwork around curriculum.
   
   This year, when departments complete an evaluation of their curricula, is this an evaluation to support improvement or an evaluation to rationalise and evidence that it is already good/ outstanding? Surely it can’t be both at once?

We think the early warning signs of the impending curriculum crash are already starting to emerge. Suppose we all recognise we have a problem before the wave crashes over the system. Can we stop it?

We don’t think we can.

Our issue is that we cannot critique any solution to a wicked problem because the problem and the intended goals are not well specified, so no outcome proves success or failure. For example, what benchmarks in 2030 will tell us that the reform programme is a failure or that the new Ofsted framework is unhelpful? Equally, the solutions are not well specified, so the failure of any one solution does not bring down the whole project.

And in any case, virtue signalling means it takes a brave soul to criticise any proposed solution to the attainment gap, and especially curriculum reform. Those who propose solutions, even un-evidenced and implausible ones, are virtuous because the problem is one we must solve. Those who suggest a solution won’t work, and who have no better solution, are vicious because, in doing so, they are denying the problem the opportunity of a solution.

School improvement in a perfect storm

So what advice can we give headteachers and teachers who find themselves in the centre of a perfect storm? Recognising ahead of time our own limitations and/or that of education as a lever to effect social change may be helpful in itself. It can be cathartic to recognise that the system itself likely induces many to feel a vague sense of not being good enough without granting clarity on what, exactly, to do about it. In writing about wicked problems, Bannick and Trommel (2019) suggest that it, to some extent, helps if we can resolve ourselves to live with the problem – accepting it as a fact of life. This need not be as defeatist as it may first appear, particularly if it helps us develop more helpful education policies. 

This more nuanced perspective may also may help us to curb the worst excesses of dysfunctional practice, recognising what is happening and resisting pressures towards unhelpful responses where we can. Placing ourselves in the position of the policy laggard, who watches others make mistakes first, can help school leaders steer a more stable ship. I think it likely that if we manage this, whilst others do not, in 2025 our future selves will surely thank us!

Where we cannot resist the drive to change practices, an understanding of the inherent complexity AND limited efficacy of changes which we might make could encourage us to look for minimally disruptive, low-workload solutions to satisfy the stronger coercive forces.  Perhaps there are some superficial, low-workload solutions – writing some knowledge organisers or drawing some links between topics, year groups and subjects. It may be that there is little difference on long term outcomes between elaborately drawn up versions and those created more quickly and pragmatically (back to evidence – nobody knows!).

Equally there may be some viable magpie solutions – finding someone else’s solution and transplanting it into your school. That said, this may be less satisfactory for curriculum changes since you cannot just maintain your normal daily business with someone else’s curriculum as you could quite happily with someone else’s data solution!

The above advice relates to ‘living with complexity’, rather than school improvement. On a more positive note, what can we actually do with a view to making real, positive changes?  Well, here it makes sense to acknowledge what people like Dylan Wiliam have been saying for years. Schools (like teachers) can only really change a very small number of things, very systematically and really quite slowly. Alternatives that promise much more, much faster, are not to be trusted.

Once we accept this, it probably makes sense to shift the majority of our time, mental energy and resources towards problems which we can solve – complicated ones.

Complicated problems can be elaborate but, unlike complex problems, we can identify and make useful changes to their constituent parts because we understand the pathways and mechanisms that are contained within them. A route to tackling complicated problems does not exist until schools have worked to carefully define and understand them. This means getting everyone in a school to agree exactly what the problem and what the goal is – writing the assessment that proves success in advance is helpful here. We then need a shared understanding of the behavioural regularities that serve the goal – where are the direct and indirect behavioural dependencies in terms of teacher practices, resources, habits and beliefs?

By creating well-understood, complicated problems, schools can then aim for small improvements in narrow areas that are still aligned with the perfect solution. For example, suppose you are a head that has long planned to improve school activity around religious studies? Use your deep understanding of existing practice in your school to make meaningful improvements in this area that you can play up at inspection time. Similarly, if you have worked out how to implement an approach to increasing time-on-task that in turns frees up time for activities that won’t increase GCSE results, now is a good time to wrap this up as an example of successful curriculum reform. (Remember, your perfect solution can be diametrically opposed to someone elses and yet both are legitimate!)

Even governments can fix problems at scale, provided they aren’t complex problems. The success in implementing phonics is a good example of what it takes to make it work. England, like many countries, had a literacy problem. For many students, illiteracy was impeding their access to the rest of the curriculum. Illiteracy is a complex problem, but it wasn’t the one that government tried to solve. Instead, government took on a moderately sized complicated problem: ensuring children in Years R and 1 are taught to phonetically decode effectively. It is a complicated problem because, whilst it involves altering resources, training, beliefs and habits, it aims to do so on a smaller scale, disrupting about 45 minutes a day in about 40,000 classrooms.

Teaching systematic synthetic phonics was not a new idea. Years of research had already been translated into several well-honed commercial products. This was essential, for schools weren’t asked to do something ill-specified (e.g. ‘sort out phonics’); schools simply needed to select and embed a chosen phonics scheme. It required the embedding a new set of classroom practices and habits, and for many involved disrupting belief systems about how learning to read works. The government funding for the new resources also supported training to help achieve this. And of course the accountability stick of Ofsted inspections and the phonics test made compliance essential!

Has it worked? Yes, as defined by the complicated problem. 4-6 year olds in state-funded schools are all taught to decode in a systematic fashion. Has it solved the wicked problem we really care about – illiteracy amongst children and adults in England? No, but it has pushed us forward considerably towards achieving a goal that is unlikely we will ever achieve, but is worth working towards nevertheless.

This may have felt like a rather depressing talk, but we want you to take away the message that school improvement IS possible! Our best hope, however, for improving this wicked problem of school improvement is not to try to fix the ill-defined problem itself. Complicated problems where the causal pathways, dependencies and rationales for the status quo are made explicit are the ones worth pursuing.

Our argument on the back of an envelope…

• School improvement and closing the attainment gap are complex problems which cannot be neatly solved. Many of us wish it were otherwise as they are important issues.
• Because they are important, yet irresolvable problems, we are always on the search for solutions to complex problems. This happens at the classroom level, the school leader level and the policy maker level. 
• When wicked problems are defined in particular ways, they present the alluring possibility of a perfect solution that, whilst ill-defined, holds the promise of success.
• By promising so much and remaining broadly defined, perfect solutions gain endorsement from an umbrella advocacy in education who, for a variety of reasons, have an interest in working towards a solution.
• Systematic endorsement of a perfect solution distorts both the solution and the system itself. In particular, when the coercive force of accountability implies what a perfect solution should do and achieve, it compounds the distortions.
• The perfect solution will not solve the wicked problem and can easily make the education system worse since causal dependencies are not well-understood.
• Sooner or later, irrational practices borne of disparate attempts to enact it, and frustration at lack of progress, will lead to a loss of momentum.
• The wave recedes, just as its successor is gathering momentum. The eternal search for a solution to a wicked problem goes on.