Eight and a half thousand families opened their doors to interviewers between November 2023 and August 2024. Their children were in Year 1, aged five and six, at state primary schools across England. The resulting report, published this month, is the first instalment of a study that will follow the same children until they are seventeen.

It is a baseline, not a verdict. Nothing in it can yet tell us what causes what, because there is only one year of data. What it does give us is an unusually detailed photograph of a cohort at the very start of primary school, and that photograph is worth looking at closely.

What the study is

Five to Twelve is the second of three cohort studies commissioned by the Department for Education under the Education and Outcomes Pathways Studies programme, and it is delivered by the National Centre for Social Research, the National Foundation for Educational Research and the National Children’s Bureau. Between them, the three studies will track children from nine months old to Year 12.

The sample was drawn at random from the National Pupil Database, with children eligible for free school meals deliberately oversampled so that subgroup analysis holds up. Interviews were completed with 8,632 families, a response rate of 47%. Data comes from parents, teachers and the children themselves, including three cognitive assessments administered at home.

This cohort has a particular history. They were between one year seven months and two years five months old when the first Covid lockdowns began, and they are growing up through a sustained squeeze on household budgets and a period of rapid change in the technology available to them.

Six findings worthy of attention

1. The gap is already visible in Year 1

Teachers were asked to predict how each child would perform in their Key Stage 2 assessments, five years away. Across the cohort, 78% were predicted to reach at least the expected standard in maths and 77% in reading. Break that down and the pattern is immediate.

Predicted to reach at least the expected standard at KS2. Comparison group figures are for children not eligible for FSM and children with no SEN respectively.

The same gradient runs through household income and parental education. Children in households earning under £20,000 were roughly four times as likely to be predicted below the expected standard in maths as children in households earning £90,000 or more.

These are predictions, made by teachers who have known these children for a matter of months. That is precisely what makes them interesting.

2. School records and parents disagree about SEN

According to the National Pupil Database, 12% of the cohort had special educational needs. According to the parents interviewed, 16% did, with a further 3% unsure. Only 8% were identified by both sources.

Some of that gap is explained by timing, since the administrative data was recorded during the reception year. Not all of it is. The report is careful not to say which figure is right, and the more useful question is probably not which but why.

Where a child did have SEN, 94% of parents said they had discussed their child’s difficulties with the school. Of those, 65% said their child was receiving support as a result. That leaves roughly a third of families who raised a concern and did not see support follow.

3. Attainment grouping is more common than expected

Around two-thirds of children were in schools using attainment grouping for English (67%) and maths (66%). The report notes this is higher than other studies of this age group have found, which may point to the practice becoming more widespread in Key Stage 1.

The researchers also suspect that teachers had within-class grouping in mind when making their Key Stage 2 predictions. Children eligible for free school meals and children with SEN were less likely to be in higher attainment groups.

4. Behaviour is broadly positive, but the flags cluster

The overall picture of behaviour is reassuring. A quarter of children (26%) never displayed disruptive behaviour according to their teachers, 55% showed low levels and only 3% high levels. Seventy-one per cent were never picked on by other children and 72% never picked on others. Eighty-five per cent of child–teacher relationships were rated as having a high degree of closeness.

Within those numbers, though, the same groups appear repeatedly. Boys, children with SEN and children eligible for free school meals were more likely both to experience and to display bullying behaviour, and more likely to be recorded as disruptive. On the Strengths and Difficulties Questionnaire, 44% of children with SEN scored in the atypical range for total difficulties, against 10% of children with no SEN; for FSM-eligible children the figure was 30%, against 10% of their peers.

5. Home learning differs in kind, not only in quantity

Most households (83%) engaged frequently in home learning activities, and 59% of children were read with on five or more days a week. Households facing disadvantage were more likely to score low on the home learning environment index, and less likely to have more than a hundred children’s books at home.

The more interesting detail is what changes rather than what disappears. Households eligible for FSM, those on lower incomes and those with a child with SEN reported more drawing, painting, singing and educational activity on a device. Households not eligible for FSM were more likely to report reading with their child, helping their child to read, and talking about feelings.

Screen time averaged 1.74 hours on a school day. Twenty-seven per cent of FSM-eligible children used screens for three hours or more, compared with 15% of those not eligible; the figures for children with and without SEN were 28% and 16%.

6. Support is reaching the children who need it

Small group learning support was the most common form of help, received by 39% of children, followed by one-to-one support at 22%. Both were more common among the children the rest of the data would predict. Seventy-five per cent of children with SEN received small group support, against 35% with no SEN. Fifty-seven per cent of FSM-eligible children received it, against 35% of those not eligible. Speech and language support went to 65% of children with SEN and 7% of those without.

The report calls this promising. Whether it changes trajectories is a question only later waves can answer.

What this means for educators

The report is descriptive by design and does not issue recommendations. The points below are what the findings seem to ask of people working in primary schools.

1. Treat early predictions as provisional. Teachers were predicting outcomes five years out for children they had taught for a few months, and those predictions already tracked family income and free school meal status. The report itself raises the question of how early expectations and grouping practices might shape what follows. Worth asking how predictions are formed in your setting, how often they are revisited, and how permeable your attainment groups actually are in practice.
2. Check the SEN register against what families are telling you. School records put SEN at 12% and parents at 16%, but only 8% of children appeared in both, which suggests information is being lost somewhere between the two. So does the third of families who raised a concern and did not see support follow. Both are auditable at school level.
3. Read behaviour flags as need indicators. The children more likely to be recorded as disruptive, and more likely both to be bullied and to bully, are the same children scoring in the atypical range on socio-emotional measures. The report argues for identifying and addressing these concerns early to stop them escalating. A behaviour system that responds to the pattern rather than the individual incident is better placed to do that.
4. Do not read a low home learning score as a disengaged family. Families facing disadvantage were doing plenty with their children, and in several categories more than their more affluent peers. What differed was the emphasis, with less reading and less talk about feelings. That points towards specific, practical support such as book access, modelling shared reading, or bilingual materials, rather than general exhortations to be more involved.
5. Look at who is not at parents’ evening, and why. Attendance was high overall, but non-attendance was concentrated among FSM-eligible households, those on lower incomes, those with lower parental education and those with a child with SEN. Timing, transport, childcare, shift work and previous experience of school are all more tractable than they might appear.
6. Notice how parental interest is being judged. Teachers rated parents as more interested in their child’s education where the family was not eligible for free school meals, had a higher income, or had higher levels of education. This measures teacher perception rather than parental interest itself, and the distinction matters when those perceptions feed into expectations.
7. Watch the intersection of poverty and SEN. The report singles this out as a focus for later waves. Where the two coincide, the disparities in this dataset are consistently at their widest, and the mechanisms suggested include reduced access to advocacy and early intervention as well as marginalisation at school.

Caveats worth carrying

Three limits are worth holding in mind when quoting any of this:

• This is a single year of data. Associations are not causes, and nothing here establishes direction of travel.
• Behaviour, bullying and attainment predictions come from teachers, not children. They record perception as much as incidence, and the report says so.
• The home learning findings cannot distinguish between families who did not start an activity and families whose child would not join in. That is a meaningful ambiguity when the children in question include a disproportionate number with SEN.

Moving forward

Five to Twelve will follow this cohort for at least another five years, tracking some measures annually and treating others as one-off predictors. Later reports will be narrower and deeper. The data will also be made available to researchers, which matters for anyone with a question this baseline raises but does not answer.

Progress on closing the disadvantage gap has stalled since 2019, and the report frames its findings against the idea of cumulative disadvantage: early inequalities that reinforce themselves and widen. The corollary is the reason this study is worth watching. Each stage of education can either compound those effects or reduce them, and the primary years are where a good deal of that gets decided.

Fullreport available at https://www.gov.uk/government/publications/five-to-twelve-year-1

Technical report avaiable at

https://assets.publishing.service.gov.uk/media/6a7470b95a472b60f4ea6cd1/Five_to_Twelve_year_1_technical_report.pdf

Source: Allen, J., Wood, M., Light, R., Mitcheson, H., Classick, R., Harland, J., Poet, H., Speight, S., Moss, S., Hodgkin, J., Mallion, J. and Nugent, R. (2026) Five to Twelve: Year 1 Report. Department for Education.