The Department for Education's children's homes workforce census tells a story the sector knows intimately. The workforce has grown to more than 46,000, but turnover ran at 29% in 2024, only marginally down from 31% the year before. Nearly one in three of the adults caring for the most traumatised children in England leaves their post each year. In the same census, 62% of homes reported difficulty recruiting care workers, and 59% difficulty recruiting managers.
Set that against what we know about the children. Most young people in residential care have experienced repeated loss and broken attachment. That is often precisely why they are in residential care. The single most therapeutic thing a home can offer is a small number of consistent, reliable adults over a long period. A 29% turnover rate means the average child watches roughly a third of their care team walk out of their life every year. For a child whose working model of adults is "they leave", the workforce data is not an HR metric. It is a clinical outcome.
The census also quantifies part of the cause: pay in local authority homes runs materially above the independent sector (several pounds an hour for care staff, around £5 an hour more for registered managers). Independent providers competing for the same people on lower pay should not be surprised by the result, and pay is a lever providers control.
But the honest reading of the evidence is that pay is necessary and not sufficient. People leave residential childcare because of unsocial hours, emotional load, thin supervision, and the corrosive experience of working chronically short-staffed. The homes that retain people tend to share features that cost commitment more than cash: genuinely reflective supervision, visible and accessible managers, clinical support for staff processing vicarious trauma, and rotas built for sustainability rather than bare coverage.
Here is the practical implication for commissioners: staff stability data is arguably a better predictor of a child's lived experience than most of what appears in a tender response. Turnover, vacancy reliance on agency staff, and manager tenure are all measurable, and any provider should be able to produce them on request.
So two questions. To fellow providers: would you publish your turnover figure? To commissioners: why is nobody asking us to?
