Eileen Munro: Lessons learnt, boxes ticked, families ignored

The Government's 'never again' measures prompted by Climbié have made children less safe, not more

After the shocking abuse and murder of Victoria Climbié at the hands of her guardians eight years ago, an inquiry was set up, reforms were put in place and, supposedly, lessons learnt. The care community and the media vowed "Never again". Yet here we are again, horrified by the death last year of Baby P in Haringey.

Did we learn nothing from the Climbié case? We did learn something, but much of it was wrong. The quality of front-line work depends on the quality of the organisation. The unbalanced reforms instigated after the Laming report into the death of Victoria have harmed the quality of that front-line work.

If we want to learn from Baby P's death, we need to do more than list the mistakes made. We need to understand where the errors come from, and their cause. Research on human error shows how organisations can have latent weaknesses that make front-line error more probable. A chance combination of these weaknesses can lead to a major tragedy. The Government's reforms of children's services are creating such latent weaknesses; the damage they can do has been revealed in Haringey.

The Government has lost sight of the basic nature of child protection work. We can begin to understand the dynamic flow of a family's life and detect abuse only by spending time with parents and children. We can improve childcare only by forming relationships and working with parents. The organisation needs to be centred on supporting that human contact with the family. Instead, what we have now are organisations centred on feeding the Government's ever-growing appetite for hard data at the expense of the complex and subtle information social workers actually need to form a realistic assessment of child welfare.

The audit and inspection systems fail to make practice transparent. They place a mistaken emphasis on the easily measured aspects of practice such as forms filled in or meetings held. Whether these forms show accurate information and well-reasoned decisions, or whether intelligent discussion took place at the meeting, is harder to evaluate.

The current system is so skewed towards the simply measured that it sends out the wrong messages to workers about what is important. One social worker told me with pride of the care she had taken to reassure and comfort two little boys who were being taken into foster care. But she also told me that her seniors had not praised her work but criticised her for not giving higher priority to arranging a meeting that figured in the performance indicators. The "important" work is achieving a high score in the audit system; comforting frightened children comes second.

Many social workers go against the flow and keep a focus on children. The trouble is that they do so despite the systemic forces when it should be because of them.

We need to remember the importance of social workers having relationships with families, but we should also guard against social workers "going native" with the family. Psychological research has shown that people are very bad at policing their own biases. Social workers need regular critical supervision to ensure that their biases are not distorting their assessments.

The social worker in the Baby P case mistakenly saw the mother as compliant despite the existence of evidence to the contrary, such as the way Baby P continued to be injured but only while in his mother's care. Robust supervision should have challenged this flawed appraisal.

If hospital managers decided to save money by cutting back on operating theatre sterilisation, we could confidently predict a rise in infection, though we would not be able to predict which patient would get which infection. In the same way, if social work managers radically cut back on casework supervision, then we can confidently predict a rise in errors of reasoning.

The Government's software programme for case records – the Integrated Children's System (ICS) – is another latent weakness. It is widely loathed by practitioners who complain of spending too much time inputting data that are of little help in protecting children. Professor Sue White of Lancaster University has done extensive research that backs up their complaints. ICS not only poses a risk by taking social workers away from their time with families but also hampers good reasoning. Good practice, for example, requires assessment to be seen as an ongoing process with workers continually reviewing and revising their picture of the family. ICS treats revising a completed assessment form as a rare event and makes it technically difficult, thereby giving the implicit and dangerous message that workers should get it right first time.

The system needs to be rejected and a new one designed based on this question: "How can computers help social workers to make better judgements and decisions?"

The death of Baby P has fuelled a desire to improve child protection. The Government is right to order a review of children's services. The trouble is, their own reforms are part of the problem.

Dr Eileen Munro is reader in social policy at the London School of Economics