And why goodwill is not enough.
Luke Roberts works at the intersection of restorative practice, systems change, and complexity. He supports schools, justice organisations, charities, and public sector leaders to navigate conflict, harm, and culture change in complex environments. For more on Complex Systems Luke’s book – Leading Schools and Sustaining Innovation: How to Think Big and Differently in Complex Systems is available on Amazon or https://routledge.pub/Leading-Schools-and-Sustaining-Innovation
Across criminal justice, healthcare, education and other organisations, restorative practice is often embraced with genuine enthusiasm. Training is delivered, early adopters emerge, and there is a palpable sense that a better way of responding to harm is possible. And yet, just as consistently, restorative work seems to stall. It becomes marginal, optional, or quietly overridden when pressure rises.
This is rarely because people don’t care. More often, it is because the organisational default into which restorative practice is introduced is not neutral, and is rarely conducive to relational repair.
The organisational default.
When harm occurs in an institution, responses tend to follow deeply familiar patterns. Investigate. Categorise. Control. Reduce risk. Protect the organisation. These patterns feel obvious, even inevitable. But they are not simply “common sense”; they are the inheritance of a particular management tradition.
Much of modern organisational life is still shaped by Taylorisit and production line principles: standardisation, efficiency, predictability, control, and the separation of thinking from doing. This logic has travelled far beyond factories. It now quietly structures how schools manage behaviour, how hospitals respond to error, and how public institutions handle blame.
Taylorism’s most enduring legacy is not the assembly line, but the assumption that human complexity is a problem to be managed out of the system. Within this logic, harm is an interruption, variance is a threat, and relationships are secondary to throughput and compliance.
Restorative practice enters these systems carrying a very different proposition: that harm is relational, that context matters, and that repair requires time, judgment and uncertainty. Unsurprisingly, tension arises.
A disagreement that reveals the problem.
I once had a conversation with a brilliant and deeply experienced restorative trainer who argued that when a head teacher joins a restorative conference, this signals commitment, generates buy in, and allows the school to become truly restorative.
I recognised the intuition immediately. Leadership presence does matter. A head teacher in circle can model humility, flatten hierarchy, and signal that relationships matter. In many schools this is courageous and meaningful.
And yet, I disagreed.
Not because leadership doesn’t matter, but because leadership participation alone does not change the organisational default that shapes how harm is responded to before, during, and after the conference. A school does not become restorative because a head teacher joins the circle. It becomes restorative when the structures, incentives and decision making logics of the system begin to change.
In most schools, the restorative conference exists inside a wider environment still governed by escalation ladders, safeguarding thresholds, inspection pressures, and legal defensibility. The conference may be restorative in form, while the organisation around it remains non restorative in function. Outcomes are welcomed until they collide with policy, risk frameworks, or accountability regimes. At that point, the system reasserts itself.
This is not hypocrisy. It is an inheritance.
Schools as complex systems.
Schools are not simple organisations. They are complex, adaptive systems composed of students, staff, families, leaders, policies, regulators, cultures and histories, all interacting simultaneously. They emerge from these interactions are rarely predictable or linear.
In such complex systems, conflict is not a sign of failure. It is an inevitable feature of complexity.
From a systems perspective, restorative practice makes profound sense here. Conflict contains information. It signals unmet needs, misaligned expectations, structural pressures and relational fractures. To respond effectively, the system must be able to meet that complexity with sufficient range.
This is captured in Ashby’s Law of Requisite Variety (1956): for a system to regulate itself effectively, it must possess at least as much variety in its responses as exists in the environment it is responding to. Where harm arises from complex relational conditions, narrow, standardised responses will always be insufficient.
Restorative practice increases a system’s variety. It allows responses to be shaped by context, voice and relationship rather than category alone. In complex environments, this is not a luxury — it is a functional necessity.
And yet, systems struggle to learn from it.
The “Spray tan” of change — or icing on the cake?
One reason is that organisations are remarkably good at absorbing innovations without being transformed by them. Marg Thorsborne has described this as the “spray tan” effect of school change: restorative language is applied to the surface, while the underlying structure remains intact.
I agree with the insight, but I would push it further. What often happens with restorative practice is not just superficial colouring, but icing. The cake underneath remains fundamentally punitive — layered with behaviour policies, sanctions, exclusions and risk management — while restorative processes are added on top, where they fit.
In these conditions, restorative practice becomes a veneer. It is permitted, even celebrated, so long as it does not meaningfully disrupt how the system maintains order, protects itself, or measures success. Where it does, it is quietly limited or overridden.
The system does not reject restorative practice. It co-opts it.
The real challenge of becoming a restorative system.
If schools, prisons, hospitals, and other organisations are complex systems, then becoming restorative is not about introducing a programme or securing symbolic leadership buy in. It is about altering the conditions under which decisions are made: what is enabled, what is constrained, and what risks leaders feel able to carry.
This is why culture change is so difficult. Taylorist systems struggle with uncertainty. They prioritise consistency over context, predictability over learning, and control over trust. Restorative practice exposes the limits of these assumptions — and that exposure can feel threatening.
Yet this exposure is precisely where learning lives.
Complexity demands adaptation — including from restorative practice
If complexity requires adaptation, then restorative practice itself is not exempt from that demand. A relational approach can still become routinised. Values led work can still be standardised. Without attention to system dynamics, restorative practice risks reproducing the very patterns it seeks to disrupt.
Restorative practice cannot simply be implemented. It must be allowed to evolve. It must be shaped by feedback from the system, particularly where it strains against policy, accountability frameworks, safeguarding regimes or leadership anxieties. These points of friction are not failures. They are data.
In complex systems, learning does not occur through compliance but through responsiveness. Restorative practice shows its deepest value not when it is protected from organisational reality, but when it surfaces what the system is currently incapable of absorbing.
If restorative practice is to be more than icing on a punitive cake, it must remain adaptive: willing to question its own forms, resist premature standardisation, and stay in contact with the power dynamics it disrupts. Becoming restorative is not a destination to be reached, but an ongoing process of learning how the system responds to harm — and whether it is willing to change in response.
In the next blog, I will explore what innovation in restorative practice actually requires in complex networks — where change emerges not from policy or permission, but from the capacity of relationships to carry learning faster than institutions can formalise it.
i Frederick Taylor’s Principles of Scientific Management (1911) argues that work should be studied scientifically to identify the most efficient methods for each task. Taylor proposed breaking jobs into standardized steps, selecting and training workers systematically, closely supervising performance, and aligning incentives so both workers and employers benefit from higher productivity. His ideas laid the foundation for modern industrial engineering and management practice, though they’ve also been criticised for treating workers as extensions of machines.
References:
Taylor, F. W. (1911). The Principles of Scientific Management. New York: Harper & Brothers.
Ashby, W. R. (1956). An Introduction to Cybernetics. London: Chapman & Hall.
