Perceptions of non-accidental child deaths as preventable events: the impact of probability heuristics and biases on child protection work

Abstract

Anxiety about the possibility of non-accidental child deaths has had a major influence on childcare policy and practice over the last 40 years in the UK and elsewhere. Formal inquiries into, and media coverage of, these rare events have served to maintain the perception that such incidents happen far too often and could have been prevented. This focus on individual events tends to distort a clear view of the actual probability of non-accidental child deaths. It reinforces the notions that potentially all childcare cases are risky, and that social workers face a high probability of being involved with a fatal incident. In consequence, work with children has become highly risk averse. However, in statistical terms, the probability of non-accidental child deaths is very low, and far less than that of a child being killed on the roads. This article examines the way in which perceptions of the ‘high’ probability of child deaths are sustained despite their very low statistical probability. The analysis draws on thinking from behavioural psychology, particularly the work of Kahneman and Tversky, to consider some of the biases in probabilistic reasoning which affect risk perceptions in a child protection context. The article explores the way in which inquiry reports into single past events reconfirm risk perceptions. It will be argued that recognition of the essentially unpredictable nature of future individual non-accidental child deaths would free up childcare professionals to work in a more positive and less risk-averse manner.