We talk a great deal about what happens at either end of a young person's education. We measure Wales against other countries through PISA. Then, a few years later, we count the young people who are not in education, employment or training. But there is a question that tends to get lost between those two conversations: what happened to the child before they became the 17-year-old we now call NEET?
Wales scored 476 in science, 459 in reading and 456 in mathematics in PISA 2025. Those results deserve scrutiny. So does the Welsh Government estimate that 17% of 16 to 24-year-olds were NEET in the year ending December 2025. Yet neither figure can tell us how a particular child travelled from a classroom to the margins of education. A newly published Welsh study makes that missing journey harder to ignore.
In Alternative Educational Provision and Care Experienced Young People: An Administrative Data Study in Wales, Nell Warner, Phil Smith and Anna Rawlings examined linked education and social-care records for 93,012 young people in their final year of compulsory education between 2018/19 and 2020/21. Some had no recorded social-care involvement. Some received support from children's services. Others had entered care before eleven or during their secondary-school years. What happened to these groups at school was markedly different.
Among young people with no social-care involvement, 9.6% had experienced a fixed-term exclusion. For those who entered care before eleven, the figure was 33.1%. For those who entered care at eleven or older, it was 45.8%. Almost half. Permanent exclusion was also more common in that last group: 2.8%, compared with 0.3% among young people with no social-care involvement. By Year 11, 10.4% of young people who entered care at eleven or older were attending a pupil referral unit, or PRU, compared with 0.8% of those without social-care involvement.
There are children behind every one of those percentages. A school record might describe disruptive behaviour, an assault or an exclusion. It cannot tell us, by itself, whether that child slept the night before, had moved home again, was frightened, was struggling to understand the lesson or had lost someone they trusted. None of that excuses harm to other people. It does mean that understanding behaviour requires us to ask what was happening around the child, and what support they had, before the incident that made it into the record.
The study also found that additional and special educational needs were strongly associated with PRU attendance. Of those who entered care at eleven or older, 31.6% were receiving the then higher-level School Action Plus support and 16.5% had a statement; the corresponding figures among those with no social-care involvement were 7.1% and 2.5%. After statistical adjustment, special educational needs had the strongest association with PRU attendance in the researchers' model. Care, additional needs, exclusion and alternative education are often discussed in separate reports. A child experiences all of them at once.
This research does not prove that entering care causes exclusion, or that exclusion causes a PRU placement. The authors are careful about overlapping circumstances and the limits of observational data. Even after accounting for measured factors including special educational needs, free school meals and exclusion history, entering care at eleven or older was associated with 2.35 times the odds of PRU attendance in Year 11 compared with no social-care involvement. That is a reason to investigate pathways and support, not to assign blame to children or to the professionals who work with them.
The picture also challenges whom we imagine when we hear "pupil referral unit". More boys attend PRUs overall. Yet among girls attending PRUs in this study, 18.4% were care experienced, compared with 12.9% of boys. The researchers found a stronger relationship between care experience and PRU attendance for girls in their separate models and say it needs further investigation. I find myself thinking about which children we notice when they are angry and which ones we miss when they go quiet.
Alternative education can be the place where a young person finally feels seen. A smaller setting, a different pace and a trusted adult can make learning possible again. But provision is uneven. In its March 2026 review, Estyn found committed staff and strong examples of tuition for pupils educated otherwise than at school, alongside limited hours and curricula that were too narrow for some. If a young person is offered little beyond English and mathematics, we risk mistaking a limited offer for limited ability.
A child who cannot manage a classroom of thirty has not necessarily lost the ability to learn. Someone who has struggled through six lessons may still spend hours making music, editing videos, cooking, drawing or fixing a bike. We should be careful about confusing a young person's relationship with school with their relationship with learning. They are not always the same thing.
Then comes sixteen. For many teenagers it is a transition; for some it can feel like a cliff edge. They may leave a PRU and, at the same time, lose the adults who helped them re-engage. In research with care-experienced young people leaving Welsh PRUs, Smith and colleagues heard ambitious hopes for the future as well as the difficulty of moving on from trusted relationships. A place on a course means much less if the bridge into that course disappears on the first day.
This is where our national conversations stop joining up. We publish statistics on exclusions, children in care, EOTAS, PISA and NEET status. Researchers can link some administrative records, as the new study shows, and services hold other pieces of the story. Yet a parent, practitioner or policymaker looking for a clear public account of what happens to young people leaving Year 11 from a PRU or other EOTAS provision will struggle. What qualifications did they gain? Where did they go six months later? Did they remain there? How did the answer differ for young people with care experience or additional learning needs? The data available to the public do not readily answer those questions together.
That gap matters because an initial destination is not the same as a sustained one. A college enrolment in September is encouraging. Being able to stay, learn and move towards a future that feels possible is something else. If we cannot see where the route breaks down, we cannot tell whether our support arrived too late, ended too soon or was simply the wrong fit.
The NEET estimate has limits too: the Welsh Government cautions that its more timely Annual Population Survey series has become volatile and is less statistically robust than the annual statistical first release. The same release estimates, using three-year averages, that 18.6% of disabled 16 to 18-year-olds and 40% of disabled 19 to 24-year-olds were NEET. These are different age groups and measures from the cohort in the PRU study; we cannot treat them as the later outcomes of the same children. We can, however, ask why their journeys are so difficult to follow across the systems that serve them.
Eventually a young person may reappear in a dataset under a new label: NEET. Four letters tell us what they are not doing. They tell us nothing about what they can do, what they love or which adult once helped them believe they had a future. They do not tell us about the placement that broke down, the school day that became unbearable or the college course they wanted to finish but could not sustain.
This is one reason Fentro exists. Young people need a chance to explore interests, meet people beyond the classroom and discover a route that makes sense to them. Some will have experienced care; many will not. Some will have thrived in alternative provision; others will still be looking for the setting and support that lets them thrive. No single label tells us which conversation to have with the person in front of us.
So before asking how to get someone back into education, employment or training, perhaps we should ask what happened along the way. What are you good at? What matters to you? What would make you want to try again? And when a child begins to fall away from learning, who catches them? Somewhere behind the PISA score, the exclusion record and the NEET percentage is a young person. They were never a statistic to begin with.