Big data is being used to help parole boards keep us safe. Risk-assessment software, such as LSI-R and LS/CMI developed by the Canadian company Mutli-Health System, is used to process data in order to estimate the likelihood of a prisoner re-offending. The pertinent data includes a prisoner’s age at first arrest, education-level, nature of the crime, behavior in prison, criminal records of friends, and results of psychometrics tests. Profiles of prisoners are compared with others.
For the most part, parole boards have bias, whether they recognize it or not. Often the decisions of parole board members are based on personal experiences and boards tend to overestimate the likelihood that a prisoner will re-offend. The software is known to increase the proportion of applicants granted parole while reducing the proportion who re-offend.
In addition to aiding decision makers on which prisoners to release, similar software called COMPAS developed by Northpointe Inc, is used for rehabilitation. The software provides the factors that are likely fueling a person’s criminal behavior, which enables rehab staff to allocate resources efficacy by treating the individual’s deficiencies or problems.
Considering the tremendous fiscal and social costs of the United States prison system, hopefully more states will start using this technology.