An experimental study of sentencing by magistrates

A pilot study was carried out with 23 magistrates to develop a sentencing severity scale. An experiment was then conducted with 168 magistrates deciding sentences for simulated cases in 56 groups of three. The results showed that sentences were more severe when offenses were more serious, when offenders had a more serious criminal record, when offenders were male, and when offenders were of higher social status. The age of the offender, the race of the offender and victim, the plea, the prevalence of the offense, and whether breack of trust was involved, did not have significant effects on sentence severity. A comparison between real and simulated sentencing decisions showed that they were similar, and a comparison between individual and group decisions showed that the group decisions were more likely to be relatively severe than relatively lenient.This research was completed while Mr. Kapardis was supported by a Social Science Research Council studentship.