This is the third in a four-part series exploring why schools lose their capacity to learn. In the first post, I explored a pattern I’ve watched play out again and again over the two decades I’ve spent working alongside schools built to do things differently. However genuinely innovative their founding vision, so many gradually lose the capacity to keep learning, and drift back toward the very institutional patterns they set out to replace. It’s not a failure of nerve, and it’s rarely a failure of the original idea. Something else is going on. In the second post I examined the critical role leadership plays, both in getting these schools off the ground, and in determining whether they ever find a way of scaling and sustaining the innovation they began with. In this post I want to explore what I consider the ‘missing ingredient’ that will determine whether the change that was started will endure.
The Missing Ingredient: Trust
There’s a piece missing from everything I’ve written so far, and I think it’s the piece that actually matters most.
In some work I’ve been doing recently with my colleague Harry Mills, we’ve updated our Friction Free Transformation framework to add a new foundational element: trust. It wasn’t there originally, and looking back, I think that was an omission worth naming – because a great deal of writing about innovation in education talks confidently about what innovation looks like, and far less about the conditions that make it possible in the first place.
Trust takes time to build, but is broken quickly
Before going further, it’s worth being clear about what trust actually is. It isn’t automatic in any relationship – it has to be deliberately built, usually over a long period of time. But it’s asymmetric: while trust takes time to establish, it can be broken in a moment, and once broken, is far harder to restore than it was to build in the first place.
a. Building Trust
In Bryk and Schneider’s decade-long study of Chicago elementary schools undergoing reform they found that schools with high levels of relational trust among administrators, teachers, parents, and students were far more likely to improve student achievement, and defined that trust through four components: respect, competence, personal regard, and integrity. What should be noted here is that schools weak in trust had virtually no chance of showing improvement in either reading or mathematics, and the researchers described this as the first evidence directly linking the development of relational trust in a school community to long-term improvements in academic productivity.
The mechanism Bryk and Schneider describe is slow by nature – trust accrues through the ordinary, repeated, small-stakes exchanges of daily school life: a principal following through, a teacher being backed up in a difficult parent meeting, a promise kept about workload. Nobody builds relational trust with a single gesture. It is built over time and requires intentional effort and decision making throughout.
b. Destroying Trust
At the other end of the scale we can see what happens when trust is broken. Trust is fragile – typically created rather slowly, but destroyed in an instant by a single mishap or mistake – and once lost, it may take a long time to rebuild, or may never be regained. The classic source here is risk researcher Paul Slovic, whose 1993 paper on trust and risk management is the origin of what’s now called the “asymmetry principle.” Slovic offers a genuinely useful explanation for why, not just an observation that it happens. He notes that negative events are more visible than positive ones, and they carry more weight because they’re rarer and higher-consequence – one broken promise stands out against months of quiet reliability the way one flood stands out against years of ordinary weather. That gives you a mechanism, not just a metaphor, and it’s been replicated well beyond risk management: subsequent research across three decades confirms negative events affect trust more than equivalent positive events build it.
Consider the following examples that illustrate how these things can play out in the context of a school…
- Built slowly – a board and principal spend two or three years earning staff trust through consistent, unglamorous means: workload commitments that are actually honoured, appraisal processes that feel developmental rather than punitive, consultation that visibly changes decisions rather than rubber-stamping them already made. None of it is dramatic. It shows up as staff willingness to take risks with new practice, to surface problems early rather than hide them – exactly the “experimenting with new practice” effect Bryk and Schneider found.
- Destroyed quickly – a single non-consultative restructuring announcement, or a redundancy process handled by the letter of the law but not the spirit of it, can undo years of that work in a single staff meeting. It doesn’t matter that it’s one decision out of hundreds; it’s the one that gets remembered, retold, and treated as evidence of what the institution “really” values – precisely Slovic’s point about negative events being disproportionately weighted and disproportionately visible.
One non-consultative decision like the one above doesn’t just damage trust in the abstract – it teaches three groups three specific lessons simultaneously:
- Teachers learn that visible failure has consequences, so the rational move is to stop attempting anything that could visibly fail – hence experimentation quietly becomes compliance with what’s already known to work.
- Leaders, having just demonstrated (or witnessed) that uncertainty gets punished, learn to perform certainty even when they don’t have it – because admitting “I don’t know” now reads as weakness rather than honesty.
- Students, watching the adults around them reward the safe answer over the interesting wrong one, learn the same lesson a generation earlier – that a mistake is a deduction, not data.
None of this shows up as a single dramatic event. It shows up as a staffroom where the same three or four people always put their hand up, while everyone else waits to see how it goes. It shows up as a leadership team that has an answer for everything, because admitting otherwise has never once gone well for anyone who tried it. And it shows up in classrooms, where the safest answer beats the interesting one, because a mistake is something you lose marks for, not something you learn from. Three different rooms, three different ages – and the same underlying condition.
It would be easy to treat these as three separate problems – a staffing issue here, a leadership style there, an assessment culture somewhere else – and hand each one its own fix. That’s certainly how most school improvement plans are written. But treating them separately misses what’s actually going on, because the teacher who won’t experiment, the leader who won’t admit uncertainty, and the student who won’t risk the wrong answer are all making the same calculation, just from different desks. Each of them has looked at what happens to people who expose themselves, and decided the safer move is not to.
That’s not a motivation problem, and it’s not a culture problem in the vague, unfixable sense the word usually gets used.
It’s a trust problem.
It’s the kind of problem Bryk and Schneider described above – respect, competence, personal regard, and integrity. Not a nicety, but the precondition for everything else in this piece.
Here’s the conditions problem in its simplest form. Experimentation requires trust. Teachers won’t experiment if failure is punished. Leaders won’t admit uncertainty if they’re expected to always have the answer. Students won’t take intellectual risks if a mistake is just going to reduce them to a grade. None of this is complicated, and yet it’s routinely overlooked in how we talk about building innovative schools. We talk as though innovation were a matter of permission, granted from above, rather than something that grows out of trust, cultivated over time. It involves the ordinary, unglamorous, repeated stuff of daily practice – a principal following through, a teacher backed up in a hard conversation, a promise kept about workload. None of it looks like reform. All of it is. The illustration below captures this:
Without trust, every one of the things I’ve been describing curdles into its own hollow imitation:
- experimentation becomes compliance
- inquiry becomes implementation
- accountability becomes surveillance
- innovation becomes performance
With trust, the same ingredients produce something else entirely:
- experimentation flourishes
- professional inquiry deepens
- knowledge is shared
- successful practice is scaled
- improvement is sustained
- and new cycles of learning begin
You could even describe trust as the renewable resource of a learning organisation.
Trust in the transformation cycle
Everything else I’ve written about – experimentation, adaptation, the capacity to keep asking “what if” – draws on it. And unlike most resources, trust doesn’t get used up by being drawn on. It gets used up by being withheld.
This is where the transformation cycle I’ve been circling throughout this piece finds its natural home, as illustrated below:
Every loop through that cycle creates new organisational capability. But it’s trust that keeps the loop turning. Take it away, and the cycle doesn’t just slow down – it seizes. People stop experimenting because experimenting without trust is just exposure. And a school that’s stopped experimenting, as I’ve already said, rarely announces the moment it happened. It just quietly stops asking the question.
Which is why the question worth asking about any school isn’t whether it has stayed faithful to its founding vision. Visions age. Contexts change. Holding onto the original idea too tightly can be its own kind of failure. The better question – the one that actually tells you whether a school is still alive to its own purpose – is simpler than that: is it still learning? In the final post in this series, I want to turn to what it takes to keep answering that question well, over years and even decades, long after the founding energy has faded and the people who asked “what if” at the start have moved on. I’m calling it stewarding innovation – because sustaining trust, and the learning it makes possible, turns out to be a different kind of work from creating it in the first place.
Call to action: Where trust actually gets built
The following ideas give school leaders some actionable ways of building trust that is enduring. Each one draws on a different part of what Bryk and Schneider found trust is actually made of – respect, competence, personal regard, integrity – so think of them less as a checklist and more as four levers a leader has available, worth pulling deliberately rather than leaving to chance.
- Make uncertainty sayable. Leaders modelling “I don’t know, let’s find out” in staff meetings – not as a rhetorical device but as a genuine, repeated practice – does more to license teacher risk-taking than any innovation policy. This is the integrity lever: consistency between what’s said and what’s rewarded.
- Redesign what gets noticed. If appraisal, moderation, and staff meetings only ever surface what worked, failure stays hidden and unexamined. Building in regular, low-stakes space to talk about what didn’t work – normalised, not confessional – is the personal regard lever: it signals people are valued for the attempt, not just the result.
- Protect the first followers. The teacher who tries something new and has it not quite work needs to be visibly backed, not quietly steered back to compliance next term. This is where trust is actually spent or protected – the competence and respect levers together.
- Separate assessment of risk-taking from assessment of outcome, for students as much as staff – feedback on the attempt, not just the grade on the result. This is the respect lever again: treating someone as more than the result they just produced.
In the fourth and final post in this series I will be discussing the idea of ‘stewarding’ innovation…
