NCJ Number
191912
Journal
British Journal of Criminology Volume: 41 Issue: 4 Dated: Autumn 2001 Pages: 726-737
Date Published
2001
Length
12 pages
Annotation
This article reports on a British demonstration project that displayed a way to investigate how the judiciary and the magistracy rank offense seriousness in assessing actual cases.
Abstract
The demonstration study used a particular set of offenders who were the source of a separate empirical study of sexual offending in England and Wales (Soothill et al. 2000). This study contains the criminal histories of 7,442 offenders convicted of an indictable sexual offense in 1973, with a follow-up period until the end of 1994. A straightforward approach for assessing crime seriousness from court data would be to model the custodial sentence awarded through a linear regression, and to try to adjust for a host of variables such as prior convictions, offender age, offender gender, guilty or not-guilty plea, victims characteristics, location of the court, the personality and remorse of the offender, and the demeanor of the judge. Some of these factors could never be routinely collected in court data sets. Once the above factors have been computed, then the effect due to the offense can be estimated and used to produce a ranking scale. This approach becomes troublesome as so many factors affect the sentence, and these all must be controlled. In addition, offenders receive one of a host of sentences, only a proportion of which are custodial. A new, alternative approach, which was used in the demonstration study, is the paired-comparisons method, which involves looking at pairs of offenses sentenced at the same court appearance for the same offender. If the person is convicted of both charges, the judge must then decide whether to award a more severe sentence for one offense than the other or to provide the same sentence for both offenses. This method provides a novel way of identifying how the judiciary and the magistracy identify "serious" offenses. 3 tables and 16 references