A developing theory of what determines whether child welfare systems act on unfairness once they can see it
Fairness Learning Theory (FLT) is LBTC’s practice-grounded theory for understanding why persistent unfairness can remain visible within child welfare systems without producing sustained change.
Child welfare systems are often not blind to unfairness. Disproportionality is visible in data. Children and families describe their experiences. Practitioners recognise recurring patterns. Reviews, research, complaints and scrutiny repeatedly surface inequity.
Yet visibility does not reliably lead to action or adaptation.
FLT starts with this paradox:
A system can know about unfairness without learning from it.
Our central question is therefore:
What stops child welfare systems acting on unfairness when they already have sight of it?
And, more specifically:
What determines whether knowledge of unfairness becomes consequential within a system?
From knowing to learning
Child welfare systems generate substantial learning activity: reviews, audits, training, improvement programmes, data, inspection and action plans.
All of these can matter. But activity is not the same as learning.
FLT proposes that system learning has occurred when knowledge changes what a system is capable of recognising, deciding, doing or adapting when it encounters the issue again.
Fairness learning goes further: it occurs when knowledge of unfairness changes the system’s capacity to recognise and alter the conditions that produce or reproduce unfair outcomes.
This means that the important question is not simply:
What did the system learn?
but:
What became different or newly possible because the system knew?
Our developing core proposition
Fairness learning requires consequential feedback and sufficient adaptive capacity at the level at which the causes of unfairness are produced.
Consequential feedback is more than information being heard, recorded or reported upwards. Knowledge of unfairness must be capable of influencing priorities, decisions, resources, accountability and the assumptions through which the system operates.
Adaptive capacity means having sufficient authority, capability, resource and system reach to change the conditions producing the outcome.
This creates an important distinction between responsibility and adaptive authority. People and organisations can be held accountable for outcomes while having only partial control over the conditions that produce them.
Our work is increasingly concerned with how systems create productive accountability: accountability that strengthens inquiry and enables adaptation, rather than compliance amplification, where pressure generates more monitoring, reporting and short-term corrective activity without changing underlying conditions.
Five enabling conditions for fairness learning
Our current theory identifies five interdependent conditions that appear to enable systems to learn towards fairness:
1. SEE — Recognise unfairness as patterned and system-produced
Treat disparities and recurring unequal experiences as signals requiring explanation, rather than automatically locating the problem in individuals or isolated incidents.
2. KNOW — Legitimate plural and marginalised forms of knowledge
Treat lived and living experience, community knowledge, practitioner insight and other marginalised knowledge as system intelligence capable of influencing how problems are understood.
3. QUESTION — Enable challenge and collective sensemaking
Create conditions in which assumptions, explanations, power and established ways of working can be questioned across roles and hierarchies.
4. CONNECT — Make fairness feedback consequential
Connect what systems learn about unfairness to the places where priorities, resources, scrutiny, accountability and decisions are shaped.
5. ADAPT — Enable structural adaptation
Locate sufficient authority, capability and resource at the level required to change the structures, rules, relationships and conditions producing unfairness.
These are not five stages or a checklist. They are interacting system conditions. Strength in one does not compensate automatically for weakness in another.
Which feedback does the system respond to?
A further area of FLT development concerns dominant feedback loops.
Child welfare systems receive multiple signals at once: children’s experiences, community knowledge, disparity data, practitioner insight, inspection, finance, performance, political priorities, risk and reputation.
These signals do not carry equal power.
FLT asks what determines which feedback becomes consequential, particularly when external accountability signals compete with—or overwhelm—signals about fairness.
This matters because stronger accountability does not automatically create deeper learning. It may enable system adaptation. But it can also intensify compliance with the accountability mechanism while leaving the conditions producing unfairness largely intact.
Understanding that difference is becoming central to our work.
A theory in development
Fairness Learning Theory has been developed primarily through LBTC’s work across child welfare systems in England and Wales, including safeguarding, multi-agency partnerships and the wider systems shaping children’s lives.
Its propositions are now being explored in analogous complex public systems.
FLT remains a developing, practice-grounded theory. We are testing and refining its conditions and mechanisms through commissioned work, research, critical reflection and engagement with existing scholarship.
We are particularly interested in understanding how systems can move from sight to action, and how accountability, feedback and adaptive capacity can combine to produce sustained change rather than additional activity.
Fairness Learning Theory underpins LBTC’s research, consultancy, learning partnerships and organisational development work.