Project 02 · Series MD
Manufactured Disadvantage
The architecture produces the differences it then measures.
11 monthsYounger than the oldest child in the same year group
22%Of the whole life they have lived so far
Select a birth month. Both figures are arithmetic on a September cut-off, with the oldest child turning five at entry; they are not measurements. What the evidence measures is the consequence: the youngest in a cohort are over-identified for special educational need and under-identified for talent, and schools then act on that record as though it described the child.
Two system numbers
Children and young people with an Education, Health and Care Plan at the January 2026 census, up 12.5 per cent in a year, which the Department records as the largest annual rise since plans were introduced.DfE, published 25 June 2026
Of state-school pupils in England had a first language other than English in 2025-26, up from 20.8 per cent in 2023-24. A language label is not an ability measure, and is routinely read as one.DfE school census, published 4 June 2026
Three mechanisms, graded
The claim is not that ability is uniform. It is that three specific mechanisms turn a structural choice into a recorded individual deficit.
Relative age
Language as proxy
Identification systems