Federal Sentencing Reporter
Use of predictions of recidivism is ubiquitous in American criminal justice systems from pretrial detention to parole release and proceeds largely oblivious to fundamental ethical problems that were widely recognized and examined in the 1970s. They include the false positive problem that most people predicted to commit acts of serious violence would not, and their confinement for that reason is unjustified, that common use of fixed characteristics such as age and gender punish people for matters over which they have no control is per se unjust, that commonly used socioeconomic factors such as marital status, employment, education, and living discrimination, because they are heavily correlated with race and ethnicity, effectively discriminate against members of minority groups, and that use of some of the same socioeconomic factors punishes people for making lifestyle choices that are neither the state’s nor the law’s business. prediction of recidivism, false positives, indirect discrimination
Michael Tonry, Legal and Ethical Issues in the Prediction of Recidivism, 26 Fed. Sentencing R. 167 (2014), available at http://scholarship.law.umn.edu/faculty_articles/525.