NCJ Number
182776
Date Published
December 1999
Length
141 pages
Annotation
This research uses sophisticated statistical techniques to provide a better understanding of sentencing trends that followed the implementation of various legislative changes in New Zealand.
Abstract
This was done by examining for the first time the combined effects of a wide range of statistically measurable factors on sentencing decisions. The factors included in this analysis ranged from the type and seriousness of the major offense committed to the number of charges proved, the plea, and the demographic and previous criminal history characteristics of the offender. A multivariate method (logistic regression modeling) was used to determine the independent effect of each statistical factor on the probability of receiving each sentence. Current sentencing practice was assessed by using models based on all proved cases that involved imprisonable offenses finalized in 1995. These models were then used to predict the probability of each sentence for offenders sentenced in each of three other years (1983, 1987, and 1991). As these predicted probabilities take account of the differences in statistical characteristics of offenders in different years, a comparison of actual and predicted probabilities indicates the magnitude of changes in sentencing practice, independent of any statistical trends. The study found that although imprisonment rates for many offenses and offenders have decreased, these decreases have been much smaller than the overall increases in the use of community-based sentences and suspended sentences. The widespread application of sentences intended as an alternative to imprisonment to offenders who would not otherwise have received a prison sentence is considered "net-widening." Extensive tables and figures, 11 references, and appended factors influencing the "no sentence" options and a discussion of the effect of sentence type on recidivism rates