Education Otherwise Than School-Shaped

The question is not whether learning happens in a building. It is whether the child is held by a coherent system.

In my last newsletter, I asked a question that sits quietly beneath almost every disagreement in education: what is school actually for?

When one group asks how to improve attendance and another asks how to improve engagement, they appear to be discussing the same problem — but they are beginning from different ideas of school. One assumes school is fundamentally a place. The other assumes school is fundamentally a relationship.

That distinction becomes urgent when we talk about children receiving education otherwise than at school (EOTAS). Because once education leaves the building, we discover which parts of school were structural — and which were merely spatial.

A timetable can travel. A curriculum can travel. Qualified teachers can travel. Assessment, safeguarding, friendship, challenge and intellectual ambition can all travel.

But they do not travel automatically. They have to be deliberately rebuilt.

The child who cannot attend school may still be ready to learn

Across a decade of alternative provision work, I have encountered versions of the same mistaken equation: unable to attend school = unable to engage in education. It is profoundly damaging.

A young person may be unable to cross the threshold of a particular building while remaining curious, intelligent and entirely capable of learning. They may be unable to tolerate a crowded corridor but able to contribute brilliantly through chat. They may need the camera off. They may need to begin with one trusted adult. They may need to recover their identity as a learner before anyone attempts to restore their identity as a pupil.

The task is not to reduce education until the child can bear it. It is to redesign the conditions in which their intelligence can safely return.

Alternative provision is not a waiting room

Too much provision is still imagined as a temporary holding space between school breakdown and school return. That assumption shapes everything: arbitrary placement lengths; commissioners asking when the child will go back before asking whether the child has begun to recover; online learning treated as an emergency bridge rather than, in some circumstances, the most appropriate present form of education; success defined as returning to a building, even when the young person's real destination is further education, employment or a sustainable blended pathway.

Of course, reintegration can be right. But it must be readiness-led, not calendar-led.

A date is not evidence. A completed block of provision is not evidence. An administrative deadline is not evidence.

Readiness becomes visible through patterns: engagement, wellbeing, confidence, participation, progress, and the young person's own account of what is becoming possible. The system must learn to read those signals.

No cliff edges

One principle appears repeatedly in effective alternative and online provision: do not remove the thing that is working in order to test whether something else might work.

Yet this is exactly what children are often asked to do.

A functioning placement ends on Friday. A return begins on Monday. The trusted adult disappears. The timetable changes. The platform closes.

The child must enter a new environment and demonstrate that they are ready, while every support that helped them become ready has been withdrawn. That is not transition. It is rupture dressed as progress.

A humane reintegration model overlaps provision. It allows a young person to test a return without gambling their entire education upon it. The trusted adult remains present. Curriculum maps connect. Support tapers only when the new arrangement has become demonstrably stable.

And there must be a route back. Not as a threat. Not as failure. As information. A system capable of learning must be allowed to discover that a transition was premature without punishing the child for revealing it.

The named adult is not an optional extra

Platforms, curriculum, accreditation, safeguarding systems and qualified subject specialists all matter enormously. But for many learners, the load-bearing structure is still a person: one adult who remains recognisable across the moving parts. Someone who notices when silence means concentration and when it means withdrawal. Someone who can distinguish non-compliance from overwhelm. Someone who does not make the young person narrate their history again at every threshold.

This person may be called a mentor, practitioner, key adult or guide. The title matters less than the continuity. In fragmented systems, the named adult acts as relational memory, carrying knowledge between family, provider, school, local authority and learner. Without that continuity, every institution can perform its own function correctly while the child still falls through the spaces between them.

Accountability must follow the child

Education outside a conventional setting must not mean education outside public sight. In fact, the greater the flexibility of provision, the stronger its architecture of visibility should be. Commissioners, schools and families should not have to wait for an end-of-term report to discover whether a young person is attending, engaging or progressing; they should be able to see meaningful patterns as they develop. Not surveillance, and not crude red-amber-green judgements — useful shared intelligence: session-level attendance, forms of participation beyond spoken contribution, curriculum progress, changing patterns of engagement, wellbeing indicators, and agreed escalation routes. The purpose of data here is not to prove that provision happened. It is to help the adults surrounding the learner notice what is happening.

A registration on a school roll can provide an important line of accountability, but only when someone is actively holding it. A child does not become safer because their name exists in a management information system. The real test is more demanding: is there a named institution receiving the information, attending reviews, exercising curiosity and taking responsibility for what happens next? Paper accountability is not the same as relational accountability.

Regulation should raise the floor without lowering the ceiling

Children educated outside mainstream settings are often among those most failed by mainstream systems. It cannot be right for them to receive less assurance precisely when they may need more. National standards and quality assurance for non-school alternative provision are therefore necessary.

But regulation must recognise what it is regulating. Online provision should not be assessed as though it were a slightly unusual physical school — inspecting a website is not inspecting an online education. Online quality assurance requires its own literacy: how live lessons are supervised, how changes in engagement are detected, who notices when a learner disappears from the pattern of the week, how safeguarding concerns travel between organisations, whether teaching is genuinely interactive, and whether the provision creates belonging rather than simply access.

The aim should be a strong national spine with room for local and pedagogical intelligence — not duplicated bureaucracy that consumes the capacity of specialist providers, and not standards so school-shaped that only miniature replicas of school can satisfy them.

The middle tier is missing

Our education system still too often offers families a brutal binary: attend the existing setting, or reach crisis point. Between full-time mainstream attendance and intensive statutory provision lies a missing middle — flexible, part-time, transitional, blended and online education available before disengagement becomes entrenched.

This is where early intervention should live. A learner beginning to struggle might remain connected to curriculum through a few online lessons each week. A young person receiving health treatment might continue learning without choosing between recovery and education. A child experiencing emotionally based school non-attendance might begin with one adult and one subject, then gradually widen their circle of participation. A school could commission support before absence becomes a defining identity.

This is not lowering expectations. It is refusing to wait for complete breakdown before responding intelligently.

The consultation is open — and it matters

These questions are no longer theoretical. The Department for Education is currently consulting on SEND reform: education otherwise than at school, covering EOTAS commissioning, quality assurance, reintegration, and the future of placements with accredited online providers — including whether such placements should be time-limited. It closes on 18 September 2026, and it is open to parents, carers, schools, local authorities and providers alike.

The consultation identifies many of the right conditions: trusted adults, flexible structured learning, early intervention, shared oversight, transparent reporting, readiness-led transitions. What matters now is whether those principles become lived architecture rather than reassuring language. If you hold any piece of this system — respond.

Education otherwise than at school must not become education otherwise than held

Perhaps this is the deeper test. When a child leaves a school building, what continues to hold them?

Does the curriculum hold? Does a relationship hold? Does safeguarding hold? Does institutional responsibility hold? Can their learning travel without their story being lost?

A coherent system is not one in which every child occupies the same place. It is one in which responsibility does not evaporate when the child moves between places. The children currently described as exceptions are already testing whether our systems can cope with a more distributed, more relational future — revealing where belonging depends upon presence, where presence has been confused with proximity, and where school has been mistaken for its building.

Education may happen otherwise than at school.

It must never happen otherwise than held.

—

Kirstin Stevens

Building Schools in the Cloud

First published in Building Schools in the Cloud on LinkedIn, 14 July 2026.