Introduction
1.1 Statistical factors influencing sentencing
1.2 Changes in sentencing practice
1.3 Outline of the report
1.1 Statistical factors influencing sentencing
The aim of this research was to quantify the relative effect of various statistical factors on current and past sentencing practice in New Zealand. The research examines the individual and combined impact on sentencing of statistically-measurable variables, such as the seriousness of the offence committed and the previous criminal history of the offender. These factors are known to influence individual sentencing decisions (Hall 1998), but their overall contribution to sentencing across all criminal cases has never been quantified.
The two legal sources of guidance on sentencing are legislation and appellate decisions. Legislation limits sentencing discretion by setting a maximum penalty for each offence, in terms of the maximum fine or maximum term of imprisonment that can be imposed. In addition there are a limited number of mandatory and minimum penalties, such as life imprisonment for murder and driving disqualification for serious traffic offences.
However, for all but the most serious offences, the majority of convictions do not result in a prison sentence and, if they do, the sentence imposed is usually considerably less than the maximum penalty. For example, the maximum penalty for burglary is 10 years in prison, whereas the average prison term for burglary is less than one year and two-thirds of offenders receive a non-custodial sentence (Spier 1998).
Legislation offers less guidance on the use of the various community-based sentences (periodic detention, community service, community programme, and supervision) than it does on the use of imprisonment. Therefore, one objective of this research was to examine the use of community-based sentences, to determine what types of offence or offender are most likely to receive each community-based sentence and to compare the characteristics of offenders sentenced to community-based sentences with those sentenced to imprisonment.
The Criminal Justice Act 1985 also gives guidance on the use of imprisonment for violent offences and limitations on the use of imprisonment for young offenders and property offences. Less guidance is given on the relevance of factors other than offence type and seriousness, such as the characteristics of the offender (prior offending history, personal circumstances, ethnicity) and the particular circumstances of the case. The role of these factors in sentencing is discussed in detail in Hall's Sentencing (Hall 1998) and outlined briefly in section 2.2.1.
Therefore a further objective of this research was to quantify the effect of some of these variables. Two issues of particular interest are the influence of an offender's previous criminal history (such as the number of previous convictions) on the choice of sentence and the extent of sentencing differences between offenders of different age, sex and ethnicity.
A multivariate method (logistic regression) was used so as to take account of the interactions among the variables. For example, imprisonment rates are lower for women than men. However, a difference in imprisonment rates would be expected given the lower average seriousness of offences committed by women and their less extensive prior offending history on average. The use of multivariate modelling allows the effects of all variables - seriousness, criminal history and gender - to be assessed together, so that the independent contribution of each variable can be determined.
The main limitation of this research is that it incorporates only the effects of those variables currently collected in statistical databases. Therefore, the effect of many of the aggravating and mitigating factors relevant to specific cases (such as the unique characteristics of the crime and the personal circumstances of the offender) cannot be included in the analysis.
The sentences examined in this study are imprisonment, suspended prison sentences, periodic detention, supervision, community service, community programme, and monetary penalties (fines and reparation). The sentence of driving disqualification was not included in the analysis as it is used only for traffic-related offences. The factors influencing the probability of receiving no sentence (e.g. a discharge with or without conviction) are summarised in Appendix I.
1.2 Changes in sentencing practice
A further objective of this research was to determine whether sentencing practice has changed in recent years. A previous report (Triggs 1998) documented the very substantial changes in the use of different sentences over the 1986-1996 period. The key trends for imprisonable offences (Figure 1.1) are a substantial decline in the use of monetary penalties, compared to a large increase in the use of community-based sentences and in the percentage of cases that are not proved (especially those withdrawn following Police Diversion).
Such substantial changes in the use of different sentences indicates that either the type of offence or offender being dealt with by the courts has changed or sentencing practice has changed or both.
The previous study (Triggs 1998) found that these trends in sentencing have been influenced by a significant increase in the average seriousness of the cases being dealt with by the courts. However, at the time of writing of the previous report, detailed information was not available on the individual criminal histories of the offenders being sentenced. Therefore, the potential effect of changing offender characteristics could not be incorporated into the previous analysis.
Figure 1.1: Percentage of prosecuted cases involving imprisonable offences resulting in each outcome, 1986-97
Analysis of offender trends in the present report (section 3.2) indicates that the courts are not only dealing with more serious offences, but also with more persistent offenders. Therefore, at least some of the decrease in the use of monetary penalties and the increase in use of community-based sentences can be explained by these changes in the type of offence and offender being sentenced by the courts.
In addition, some changes in sentencing practice would be expected, given the legislative changes enacted over the last decade and a half. The Report of the Penal Policy Review Committee of 1981 (published in 1982) was influential in directing sentencing policy changes. Many of the recommendations of this committee were enacted by the Criminal Justice Act 1985. The terms of reference for the review included, amongst others, the following aims (page 10):
'(ii) To consider the means by which the incidence of imprisonment can be reduced to the greatest degree consistent with the maintaining of public safety.
(iv) To investigate means of increasing the availability of sanctions that keep the offender in the community.'
Thus the aims were to promote the use of community-based sentences as an alternative to imprisonment and to restrict the use of imprisonment as far as possible to the most serious, violent and persistent offenders. Two new community-based sentences were introduced in the early 1980s; community service in 1981 and community programme in 1985.
The promotion of community-based sentences would appear to have been successful, as indicated by the rapid growth of the use of these sentences in the latter half of the 1980s (Figure 1.1). However, the overall rate of imprisonment has not decreased by nearly as much as the use of community-based sentences increased, although imprisonment rates have decreased for many less serious offences (Triggs 1998).
In the present report, the causes of the recent significant changes in sentence use are re-examined, taking a greater range of factors into account and using multivariate analysis to examine the relative contribution of the various statistical factors. In particular, the focus was on estimating what proportion of the change cannot be explained by trends in statistical factors, which may indicate the extent to which sentencing practice has changed (i.e. the extent to which a person with the same statistical characteristics is now more or less likely to receive a given sentence than in previous years).
As the main objective of this analysis was to examine the shift in sentencing practice towards community-based sentences, only imprisonable offences were included in the analysis, as community-based sentences cannot be imposed for non-imprisonable offences. As sentences can only be imposed when the court case is proved, and criminal history data is only available for proved cases, this study is restricted to the analysis of proved cases.
The period covered by this analysis was also extended from the previous report, so that the legislative changes introduced by the Criminal Justice Act 1985 could be examined. Because it was not logistically feasible to examine criminal history information from every offender in every year, four evenly-spaced years were chosen for the analysis - 1983 (before the Criminal Justice Act 1985), 1987 (before the Children, Young Persons, and Their Families Act 1989), 1991 (before the Criminal Justice Amendment Act 1993), and 1995.
An additional objective of this research was to examine to what extent sentencing decisions are influenced by previous sentences or breaches of previous sentences. This issue is of interest as there is some concern that the rapid rise in the use of community-based sentences may be 'fast-tracking' offenders to more serious community-based sentences and prison sentences. This could be argued to have occurred if an offender who would previously have received a monetary penalty, but who now receives a community-based sentence, is more likely to receive a prison term if he or she breaches the community-based sentence or is reconvicted on another charge.
As the impact of fast-tracking can be affected by differences in reconviction rates between offenders who have served different sentence types, a brief analysis of reconviction rates by sentence type is included as an Appendix to this report.
1.3 Outline of the report
The statistical methods used in this analysis are presented in Chapter 2, including a summary of the explanatory variables examined and their relevance to sentencing. Chapter 3 presents an analysis of the interrelationships between the explanatory variables and documents recent trends in these variables. Chapters 4-10 present the results of the analysis for each sentence type, using both single-variable and multivariate analysis to assess the statistical factors influencing current and past sentencing practice. The factors influencing the probability of receiving no sentence are summarised in Appendix I.
The final chapter summarises the findings of the previous chapters, comparing the characteristics of offenders receiving each sentence and examining the changes in sentencing practice over the last decade and a half. The final chapter also looks in more depth at some of the interaction effects between different sentences, including the potential for net-widening effects for sentences introduced as alternatives to imprisonment and the potential effects of 'fast-tracking'.
A brief analysis of reconviction rates is included as Appendix II.
