England expanded children's mental health services – but disadvantaged young people are still being turned away

One in six children aged 5–16 in England has a probable mental health disorder. Since 2016, successive governments have invested heavily in expanding services – yet no one was systematically asking whether that investment was reaching the children who needed it most.

The OxWell Student Survey offered a way to find out. One of the largest school surveys in England, it captures the voices of tens of thousands of pupils each year, asking not just about their mental health but about whether they sought help, what happened when they did, and whether it made a difference. In 2023, nearly 33,000 secondary school pupils responded – providing a rare, nationally representative picture of who gets support, who gets turned away, and who never asks in the first place.

Our approach and partners

Researchers from the Nuffield Department of Primary Care Health Sciences and the Department of Psychiatry at the University of Oxford worked with ARC OxTV's social care theme and Oxfordshire County Council to analyse the 2023 OxWell data. They examined whether pupils' access to mental health support – and their experience of that support – varied by socio-economic background and ethnicity.

Rather than looking at services in isolation, the team used an established framework for understanding access that considers the full journey: from recognising a need, to seeking help, to reaching a service, to finding it useful. This made it possible to identify where the system falls short for different groups of young people.

What we found – and why it matters

1. Disadvantaged pupils face barriers at every stage. Young people from less affluent backgrounds were significantly more likely to be denied access to specialist children's mental health services (CAMHS). When they did access support – from GPs, community services, or other sources – they consistently rated it as less helpful than their more affluent peers.
2. Black pupils face specific barriers to non-clinical support. While ethnic disparities in access to clinical services were less pronounced overall, Black pupils were notably more likely to be turned away from social workers and charity-based services – sources of support that are often the first point of contact for young people in difficulty.
3. Expanding services is not the same as reducing inequity. The findings suggest that simply increasing the supply of mental health support does not guarantee that it reaches the right children, or that they experience it as helpful when it does. Service design, cultural responsiveness, and the quality of the interaction all matter.

The team presented these findings at the Health Economists' Study Group annual conference in June 2025 and has prepared policy briefings for Oxfordshire County Council. A full paper is in preparation.

What this means

For young people from disadvantaged or minority ethnic backgrounds, the current system is not working as it should. They are more likely to need support, less likely to receive it, and less likely to find it helpful when they do. Closing these gaps would not only improve outcomes for individual children and families – it would reduce pressure on crisis services downstream and help ensure that public investment in mental health delivers returns across the whole population, not just those best positioned to access it.

What needs to happen next

Two steps would significantly strengthen this evidence base. First, longitudinal analysis of OxWell data from 2024 and 2025 would show whether these inequities are narrowing, widening, or holding steady – a question that matters urgently as the NHS develops its 10 Year Health Plan. Second, linking OxWell's pupil-reported data with records from service providers – CAMHS, social care, and community services – would allow researchers to compare what young people report with what services actually deliver. Both steps require collaboration with commissioners, policymakers, and data custodians, along with resources for data governance and linkage.