Evidence to the Children, Schools and Families Select Committee's inquiry into the training of Children and Families social workers.

1. Summary

(a) Risk assessment is a key task in social work with children and families. The most important decisions about removing children from their birth family involve an assessment of the relative risks and benefits of removal compared with leaving them in the home.

(b) Research shows that people's intuitive risk assessments tend to be biased. Training in the conceptual process involved can not only reduce these biases but also improve transparency, helping social workers articulate the reasoning behind the final judgment to families, their supervisors, and the judiciary.

(c) In responding to a child protection referral, assessments need to be made of the child's current safety or harm, the family's needs and strengths, and the risk of future harm. The current single form designed as an assessment of need contributes to workers failing to make a clear assessment of current harm and future risk.

(d) Defensive practice combined with weak risk assessment skills leads to risk assessments that focus more on protecting the worker or agency than children

(e) A number of actuarial risk assessment tools have been developed but either have poor accuracy or have not been empirically validated.

 

2. I have been asked to submit evidence on the issue of risk assessment in the training of children and families social workers.

 

3. My research focus includes the study of risk assessment and decision making in child protection. I have expertise in the use of formal risk assessment and decision making frameworks as well as in the psychology of risk and decision making. For several years, I taught a Masters' course at the LSE in 'Child Protection: Risk Assessment and Decision Making' to a multi-disciplinary group of experienced workers

 

4. The poor quality of social work assessments has frequently been noted in SSI and OFSTED inspections and in SCR reports.

 

5. Assessments need to be made of the child's current safety, the future risk of harm, and the family's functioning, noting both needs and strengths. There is some evidence that the use of a single assessment framework to cover all these dimensions contributes to an inadequate focus on current safety and risk of harm. Assessing current harm and risk requires a different mindset from assessing needs and strengths, with a heightened sense of suspicion and a willingness to be more challenging, even at the risk of upsetting the parent(s). It should be noted that in many of the most high-profile cases, such as Baby Peter and Victoria Climbie, the main failing was in the assessment of the maltreatment the child was already suffering, not in the assessment of future harm.

 

6. Risk assessment is crucial in deciding whether it is better to remove the child from the birth family or whether the family can be strengthened sufficiently to provide adequate care. Such decisions involve a very complex balancing of the risks and benefits of each option; only rarely is there is a clear choice between a highly dangerous family and a safe alternative.

 

7. The language used in discussing risks is ambiguous, contributing to misunderstandings between professionals in child protection. More attention needs to be given to ensuring there is shared usage of key terms. For example, 'high risk' can mean a risk of a highly adverse outcome or a highly probable risk of an outcome. Recently, ambiguity has increased due to the Every Child Matters' policy broadening the concept of an 'at risk' child to those at risk of failing to fulfil their potential. Consistency of use will be increased by people understanding the conceptual processes to which the words refer.

 

8. A number of actuarial risk assessment tools have been developed in the USA and Canada but either have poor accuracy or have not been empirically validated. They are being used in many jurisdictions but research has found that they are, in practice, used in inappropriate ways by many practitioners. The empirical evidence on their accuracy is limited but indicates low accuracy. This is perhaps not surprising because the conditions that lead to maltreatment are a complex and poorly understood interaction of a range of psychological and social factors relating to parents and children. In a blame culture, such tools have the attraction of transferring responsibility for a judgement that turns out to be inaccurate from individuals to the tool. The risk to the family of receiving an inaccurate risk assessment merits more attention. The main lesson to draw from the research on these tools is that our ability to predict future family behaviour is weak, whether by actuarial or professional calculation and therefore professionals need to hold their judgements with caution and be willing to review and revise them as new evidence emerges.

 

9. Conducting high quality risk assessments requires a high level of the basic social work skills of interviewing, observing, thinking critically, collecting information and analysing it.

 

10. Improving the training of social workers in risk assessment needs to be linked to ensuring that their subsequent work environment creates the conditions in which good risk assessments can be made. This involves recognising the time needed to reflect and formulate an assessment plus the crucial role of critical, reflective supervision.

 

11. Risk assessments require knowledge of numerous substantive issues e.g. the probability of a successful outcome from a mother's attendance at a drug addiction programme. The internet offers immense potential as a source of such knowledge and its use needs to be encouraged.