Using Re-Conviction Data to Measure Re-Offending: Incorporating Seriousness and Frequency into a Single Non-Binary Measure

Abstract

Re-offending is usually operationalized as a binary variable, signaling whether the criminal justice system encountered a given offender within a specific time-period. An increase or decrease in the proportion of participants committing a certain type of offense is usually considered the primary measure of effectiveness of any intervention. Yet such a binary operationalization of re-offending does not provide a complete picture of re-offending and might provide distorted or even false results by failing to take into account the seriousness and frequency of the re-offending. This paper begins by outlining why re-offending should be measured in a non-binary way, discussing the aims of evaluated treatments, the black-and-white nature of dichotomization, results from studies on desistance, chronic offenders and the effectiveness of sanctions including incapacitation. I then discuss possible ways of measuring the seriousness and frequency of re-offending, with a specific focus on their measurement in the continental European context and on the measure of re-conviction (whereas re-arrest is usually used as the equivalent measure in common law countries). I discuss the possible use of sentencing ranges set by legislators or of starting points set within various forms of sentencing guidelines to measure seriousness, in a similar way as the Cambridge Crime Harm Index does. I pay particular attention to the operationalization of the re-conviction measure in studying the effect of sanctions.