Blog

The updated Keeping Children Safe in Education (KCSIE 2026) has been published after consultation and is due to come into effect in schools in September. This guest post was written for us by a professional with expertise in child development who must remain anonymous, but can be followed on X at Read some Piaget please!

KCSIE 2026 Gets the Caution Right and the Child Incomplete

Keeping Children Safe in Education (KCSIE 2026) is the first version of the statutory safeguarding guidance to take a full position on gender questioning children. Neither KCSIE 2024 nor 2025 contained a full dedicated section. Read against the developmental science underpinning the Cass Review, which the guidance cites repeatedly, it gets the central judgment right, gets one instruction actively wrong, and leaves the whole section’s conceptual foundation incomplete.

What It Gets Right

Paragraph 270 describes social transition as an active intervention with significant effects on a child’s psychological functioning, not a neutral accommodation, and primary schools are told to expect full transition to be agreed very rarely. This tracks Erikson and Marcia’s account of identity formation closely. Adolescent identity is meant to pass through moratorium, exploration without premature commitment, before a settled self consolidates.

A child fully transitioned at seven has skipped straight to foreclosure, a status Marcia’s research associates with reduced capacity for later revision, since each year of accumulated social scaffolding raises the cost of reconsidering. Keeping options open, as the guidance instructs, is what the literature predicts foreclosure avoidance requires.

Paragraph 266 makes parental involvement the default, departing only where involvement itself constitutes a greater risk. This is consistent with the attachment tradition Bowlby, Ainsworth and Fonagy established: a child working through gender related distress needs the secure base a trusted caregiving relationship provides, and excluding parents by default removes the most reliable source of that regulation.

Charities working with LGBT young people raised a fair objection during consultation, that a fifth of trans and non-binary young adults report not being close to their families, but that argues for the exception being applied with real judgment, which paragraph 266 already asks for, not against the default itself.

Caution about early social transition is what four decades of identity development research would recommend regardless of the underlying question.

What It Gets Wrong

Footnote 62 instructs schools that where it is unclear whether a child holds the protected characteristic of gender reassignment, they should proceed as though they do. Gender reassignment presupposes a settled, adult type identity claim, but the guidance’s own paragraph 254 says early cross sex behaviour typically does not persist into adulthood. Footnote 62 then asks schools to treat the claim as though it had already met a threshold of stability paragraph 254 says most such claims, in children, do not meet. That is not silence on the tension. It resolves it against the evidence.

A second conflict sits between paragraph 280, which requires biological sex to be recorded accurately with no social transition exception, and paragraph 276, which invites schools to accommodate a child day to day through names rather than pronouns. Winnicott’s true self and false self distinction is directly relevant: a child addressed one way while every formal record describes them another is being asked to hold two realities at once, which is exactly the condition Winnicott identified as producing compliant, false self functioning.

A third conflict was flagged by SEEN in Sport’s consultation response. Paragraph 98’s proportionality test now weighs fairness alongside safety in school sports, correcting a gap critics identified in the draft. But the operative no exceptions rule for single sex sport is still worded to apply specifically where sport was organised for safety reasons, so a school that organised it for fairness has no equivalent protection, even though paragraph 98 treats both as equally legitimate. The correction was added in one paragraph without being carried through to the rule three paragraphs earlier.

A guidance document can get its headline judgment right and still issue instructions that pull against each other, and against the evidence it cites, at the edges.

What Remains Incomplete

The deepest gap is an absence rather than an error: the non definition of gender identity. The guidance fences around this construct using the term “questioning” without defining what the root of these questions are.

The 2023 non-statutory draft that preceded this section defined its central term: gender identity, it said, is a contested belief. KCSIE 2026 never uses the term at all, shifting entirely to questioning their gender, social transition and biological sex. This is partly an improvement, since treating gender identity as a settled category in children was never evidentially supported. But it regulates twenty paragraphs of behaviour without ever stating what that behaviour is evidence of.

Piaget’s account of preoperational reasoning and Bandura and Vygotsky’s work on social transmission both offer developmental explanations the guidance never engages, leaving staff asked to be professionally curious about a phenomenon with no theory attached.

The same gap affects how the guidance treats children other than the one making the request. Paragraphs 262 and footnote 62 ask schools to weigh the impact on other children, but only as a competing interest, comfort and dignity to be balanced against one child’s wishes.

Bandura’s finding that same sex peer models produce the strongest imitation effects means a classroom containing a visibly, institutionally supported transition is exactly the high salience, vicariously reinforced environment his framework identifies as most likely to produce further imitation.

This is not hypothetical: the clustering patterns behind Littman’s rapid onset gender dysphoria research and multiple clinical services’ referral data show new presentations appearing within existing friendship groups rather than independently, precisely as the theory predicts.

SEEN in Sport raised a narrower, safeguarding version of the same concern about shared changing rooms, and the book underlying this analysis records a case where a school’s failure to communicate a transition to a class affected every other family in it, not only the one that raised the request.

A framework that treats other children only as a fixed quantity of comfort to be weighed, rather than as an environment that shapes and is shaped by what happens within it, has no way to notice a second or third similar request arriving from the same friendship group within weeks of the first, still less to ask why.

A guidance document that declines to name what it regulates has relocated the problem onto the staff member who has to exercise judgment with no theory to exercise it with, and left them unequipped to notice when one child’s transition has become the environment several more are learning from.

None of this means KCSIE 2026 has failed. Measured against the affirmative model it replaces, in which a child’s self report was sufficient grounds for full transition, it is a substantial correction, and its instinct to slow the process, keep parents inside it, and treat transition as consequential rather than costless is the right instinct.

But caution without a stated theory of what a child at each stage can understand, decide and sustain will keep producing gaps of exactly this kind. KCSIE 2026 has the caution. It still needs the theory.