Bias, Prejudice and Shame in Predictive Policing: State-of-the-Art and Potential Interventions for Professionals

This chapter reflects on prejudice, bias and shame in the context of predictive policing. Predictive policing in crime prevention and crime science has gained interest during the past years and has, in some socio-cultural contexts, become a favoured technique to prevent crime and to detect offenders and victims before the crime happens. Predictive policing uses specific computer programs, algorithms and big data. Critical voices have pointed out that these programs are highly prejudiced and biased, particularly against members of minority groups. Such programs are said to predict crime scenes based on racist, gendered and stigmatising categories found in previous data sets and programming biases. Being categorised by computer programs as a potential offender or victim—either through personal ascription or according to crime location—can evoke shame in the person concerned. Previous research shows that shame often impacts negatively on the individual. However, it can be positive when it is understood as reintegrative shame or when it can be transformed towards personal growth and development. The aim of this chapter is to reflect on the interlinkages of the described concepts within the specific context, present a state-of-the-art intervention and reflect on a possible way forward based on provided conclusions.