The Politics of Prediction

Abstract

Decisions have to be made about offenders. Predictive techniques can provide evidence that may reasonably be taken into account in such decisions. Uncertainty in prediction need not be disabling if it is transformed into probabilities. This is the interface between politics, science, and morality -- the interface where prediction techniques provide the medium of operational-level discussion and action. Because this is an interface, it is unreasonable for any sector of authority to arrogate to itself the task of making the determinations, and it is equally unreasonable to delegate that responsibility. The moral responsibility is for a shared responsibility. Those who firmly believe that they know with certainty what to do about crime or criminals have never studied the subject. Thus, the best thing that social scientists should do is to destroy myths. Perhaps the most important myth to be destroyed today is that decisions about individual offenders will have a significant impact on the incidence and prevalence or crime.