Abstract:
The early and accurate assessment of the likelihood that a child will be the victim of maltreatment in the future promises obvious benefits. Recent research suggests that predictive risk models – automated tools that gather and process information held in existing data sets in order to determine patterns and predict future outcomes – go at least some way toward making such assessments possible. However the application of predictive risk modelling to child maltreatment brings ethical risks and costs, including the possible stigmatisation of already vulnerable populations, predictable false positives, the use of data without consent, difficulties in designing and implementing effective interventions, and resource allocation issues. Not surprisingly predictive tools have been treated with suspicion in the child welfare area.1 This paper takes a predictive risk model developed in New Zealand as its focus and attempts to identify and respond to at least some of the ethical risks associated with the use of PRM in child protection, suggesting that the ethical costs associated with such modelling can be addressed or ameliorated or are outweighed by its potential benefits.