Local authorities placed nearly 800 children in unregistered settings last year, at a cost of £353 million. The same authorities, the same ministers, the same newspapers, spent the year attacking the registered private sector, the therapeutic providers working legally and above board to help children, for “profiteering”.

Both are true. The providers being attacked are not the sector causing the harm. This publication exists to say, with evidence, what is actually broken and who is actually responsible.

1. The registered private sector exists because the state asked it to. In the 1990s, local authorities told the private sector to open children's homes as a temporary measure, to fill the gap while fostering was scaled up. Thirty-three years later, the temporary market is ninety per cent of residential provision. The sector is now attacked for the consequences of decisions it did not make. Full history: Banning profit from care in Wales (and how we got here).

2. The dysfunction is broken information, not ownership. The sector has no common language for what children need or what providers offer. No outcome measurement. No way for a commissioner to tell a therapeutic provider with a decade of clinical evidence from a new entrant charging £20,000 a week for a child in a caravan. This is Akerlof's 1970 market for lemons, playing out in children's social care. When buyers cannot distinguish quality from noise, they assume average, pay average, and good providers exit. Fix the information and the market begins to work. Ban the ownership and it will not.

3. The operating model determines outcomes, not the capital structure. Private equity is in this sector because the state allowed it in. It is not evil by default. Neither is a charity good by default. There are bad local authority homes, bad charity homes, bad non-profit homes, and bad for-profit homes. There are also outstanding homes of every ownership type. The peer-reviewed evidence across healthcare is clear: long-term capital with a quality focus produces positive outcomes. Extractive practice across any ownership type produces harm. Regulate behaviour. Regulate extraction. Do not pretend that banning ownership fixes operations.

4. Councils place by process, not by child. Most local authority placement policies start with in-house fostering. If that fails, independent fostering. If that fails, residential care. The policy runs in sequence, regardless of what each individual child actually needs. An experienced commissioning consultant of twenty-five years, calls this sequential sourcing. In 2015, a judicial review on Section 22C of the Children Act required councils to choose the most appropriate type of placement for each child, not to run down a standard list. A decade later, most councils still run down the standard list. Children are bounced through three, four, five failed placements because of it.

5. Wales chose ideology over children's lives. The 1 April 2026 for-profit ban launched into a market where eighty-seven per cent of placements are for-profit, with no funded replacement in place. Welsh children are now being placed in unregistered accommodation run by the same councils who campaigned for the ban. Practitioners watching describe it as the worst sector intervention in a decade. The rhetorical control of the narrative is such that almost none of them will say so on the record. This was ideological policy imposed before a replacement was built, and children are paying for it.

6. We have a sufficiency crisis. Too many of the wrong homes in the wrong places. Most of what the market has built is small, generic emotional-and-behavioural-difficulty provision, heavily concentrated in the north of England where property is cheap. It sits half-empty. The specialist provision children actually need (therapeutic, trauma-informed, clinically supported) does not exist at scale. Some homes will close. Some will repurpose. That is the correction a broken signal has delayed for thirty years. Capping profit at zero does not put the right children in the right homes.

7. This has been solved before. The NHS created a common outcome language for elective surgery with Patient Reported Outcome Measures in 2009. UK adoption matching was transformed from 2013 by Linkmaker, a platform built on both sides, for adopters and children, not just for commissioners. Retail financial services was reshaped between 2001 and 2006 by Treating Customers Fairly, which replaced process compliance with outcome tests. Beyond the UK: Medicare Diagnosis Related Groups, Carfax, US hospital payment reform. In every case, one piece of information infrastructure, adopted first by the parties with the most to gain, became the de facto standard. The equivalent is identifiable here. It sits in a coordinating layer the sector has never had. This publication will work through what it looks like, piece by piece.

Analyses inspection, commissioning, provider economics and policy against the evidence, not the press framing. Shares the growing guide library (SCCIF preparation, Annex A, Regulation 44, statement of purpose, registration readiness) free to subscribers. Treats the sector as a professional community under pressure, not a problem to be solved.

Marcus Williamson. Builds technology sold into this market at elyndra.co.uk. The argument above defends the registered sector…it is also the argument commercially convenient for a vendor selling into it That is named here rather than buried!

Disagree with the evidence and the position will change.

Subscribe below.