Methodological standards for the prediction of criminal reoffenses

摘要

Within the scope of particular legal decisions, German penal code provides assistance of expert testimony about the probability of criminal recidivism, to raise the judgments in their predictive facets to a scientifically founded basis. Basically there are two ways, leading to scientifically justifiable criminal predictions. The first is based on empirical knowledge about recidivism quotes and about predictive factors, which is used to estimate the individual probability to reoffend. The other way is based on a systematic analysis of the hitherto observable criminality in the specific case and its individual causation, which is then extra-polated to the future. This article introduces both approaches with their particular methodological principles and scientific probation and discusses their specific strengths and limitations against the background of the legal requirements for predictive expertise within German criminal law.