The Agency Worker in the Children's Home: What the Reliance on Temporary Staff Costs Young People

Agency workers now cover a significant proportion of shifts in children's homes across the sector, particularly where vacancies are high and absence cover is a persistent operational challenge. This is not a failure specific to any single provider; it is a structural feature of a sector under sustained workforce pressure. But the fact that agency use is widespread and structurally driven does not make it neutral in its effects. A young person living in a residential home experiences an agency worker not as a staffing solution but as a stranger in their home — an adult they do not know, who does not know them, who arrives at the beginning of a shift and leaves at the end with no ongoing relationship with the young person they have just spent twelve hours alongside. The sector has been too slow to name this honestly, in part because naming it requires acknowledging that one of the things we most consistently promise to children who have experienced relational damage — stable, consistent, trustworthy adults — is precisely what high agency use most directly undermines.

For young people who have experienced developmental trauma, the sense of safety in a residential home is built substantially on the known-ness of the adults within it. The capacity to co-regulate, to begin to internalise a felt sense of safety, to tentatively allow an adult to matter — none of this is a generalised openness to any sufficiently warm adult who arrives. It is built slowly and often painfully with specific people over time. The keyworker relationship is the relational spine of a therapeutic residential placement, but it is not the only adult relationship that matters. The quality of a home's therapeutic milieu depends on the overall stability of the adult group: the consistency of the team, the familiarity of the faces, the accumulated knowledge that regular staff carry about each young person's world. When a significant proportion of any given shift is covered by people the young person has never met — or met once in an entirely different context — the young person's nervous system is doing something different from what it does when surrounded by familiar adults. Hypervigilance — the persistent orientation toward threat that is a defining feature of the traumatised stress response — is activated by unfamiliarity. A young person who is managing well in a week with a stable rota may present very differently over a weekend where agency cover has displaced most of the regular team. That shift in presentation is not fickleness or manipulation. It is the young person's nervous system responding accurately to a changed relational environment.

The therapeutic implications are concrete. Keywork sessions require a foundation of trust that cannot be established in the opening minutes of a first encounter. Life story work, the processing of trauma disclosures, conversations about contact with family members, support during mental health crises — none of these belong with someone the young person does not know and is unlikely to see again. They belong in relationships, and relationship is precisely what agency arrangements structurally cannot provide. This creates a burden for regular staff: they return from days off to find that a young person has had a difficult period with unfamiliar adults, that a situation has escalated in a way it might not have with a known worker, that relational repair is needed before anything else can happen. The therapeutic work is interrupted and must be restarted, often from a less secure position than where it was. Over time, this does not function as a series of discrete interruptions but as an accumulation of small breaks in the relational fabric of the placement — one that makes it harder for young people to invest in the adults around them. A young person who regularly encounters agency faces may conclude, with some justification, that the adults in this home are not constant enough to be worth the risk of depending on. This is precisely the relational conclusion that therapeutic residential care is designed to undo.

There is also a risk dimension that sits alongside the therapeutic one. Agency workers in a children's home make decisions about complex young people with limited contextual knowledge. Even the most thorough handover cannot transfer what a regular member of staff carries in their body: the read on what a particular young person's posture means at the start of a shift, the awareness that a particular tone of voice works badly with a particular child, the knowledge of what happened yesterday and how the young person is carrying it today. In moments of escalation, this matters enormously. De-escalation depends substantially on relational authority — a calm adult whom the young person trusts can achieve things in a moment of crisis that a stranger cannot. The well-intentioned, competent response of an unfamiliar worker may be exactly the wrong response in a specific situation because it is applied without the specific relational knowledge that the situation requires. Incident data across the sector is consistent with this: homes with higher rates of agency use tend to record more frequent escalations. While disentangling causation from confounding factors — vacancy rates, placement complexity, home size — is difficult, the relational mechanism is plausible enough to take seriously and to manage accordingly.

Some of the harm associated with agency use is reducible within the constraints of the current workforce market, and homes should pursue that reduction actively. Building a bank or associate model — a pool of familiar, vetted individuals called on regularly rather than sourced afresh each time — replaces the stranger dynamic with something closer to an extended team member. An agency worker who has worked in a home six or eight times is a materially different experience for young people than one arriving for the first time; building relationships with preferred agency workers and requesting them consistently is a concrete mitigation that many homes under-invest in. Induction for all agency staff — a genuine, thorough introduction to the young people in the home, their histories in appropriate summary, their key needs and co-regulatory strategies, the home's relational culture and practice language — is not optional; it is a minimum condition of safe and ethical practice, not a courtesy. Homes can also make explicit policy choices about what proportion of their rota can be covered by agency at any one time, treating this as a child welfare threshold rather than purely an operational variable. These choices create difficulty and cost. They also say something important about what is actually being prioritised.

None of this can be fully resolved at the level of individual homes or providers, however deliberate and thoughtful their mitigation strategies. The workforce conditions that produce agency dependency — pay that does not reflect the skill, relational complexity, and emotional demands of the role; insufficient career structures that drive good practitioners toward management or out of the sector; the absence of a recognised qualification pathway that signals professional standing; the cumulative toll of working with significant trauma without adequate support — are conditions that individual homes inherit rather than create. Until residential childcare is properly resourced as a profession, until pay and conditions are genuinely competitive, until the sector makes the political case for the investment the workforce requires, the vacancy rates that drive agency dependency will persist. Homes can work hard at this, and they should. But the young people in their care are entitled to a workforce that is stable enough to do genuine relationship-based work — and that entitlement is a policy question, a commissioning question, a pay question, that no single provider can answer alone. Being honest about this is not an evasion of local responsibility. It is a precondition of understanding the problem clearly enough to argue credibly for the changes that would actually solve it.