Challenging the Actuarial Method in Penal Systems

In this second article about Profiling and Policing in an Actuarial Age, Bernard Harcourt is demonstrating with mathematical tools and examples how the implementation of actuarial methods and economic models, such as statistical discrimination in the penal field, is not effective and could be counter productive. The introduction of statistical methods in the prediction of behaviour does not work, on the contrary, it increases total criminality in society because the different groups do not react in the same way, they have not the same elasticity, facing the changes in policing. Moreover, the social cost of these methods is neglected besides it is characterized by the stigmatisation of a group. Then, why have these methods been adopted ? One of the explanations is the desire to make scientific the study of human behaviour.