- Diversion involves the redirection of offenders away from conventional criminal justice processes. Diversion interventions are broad-ranging and vary according to the type of offender targeted, the degree of offender supervision, the type of treatment offered and the stage in the criminal justice process at which offenders are recruited
- Police diversion interventions mostly revolve around education and treatment sessions for low level cannabis offenders
- Results from an outcome evaluation of police diversion interventions undertaken by the Australian Institute of Criminology (AIC) are positive.
Key findings include:
- in all jurisdictions the majority of diversion participants did not re-offend between 12 to 18 months after being cautioned
- most diversion participants with prior offending records were not re-apprehended for any offences post-diversion; of those that did re-offend, there was a decline in the rate of offending
- a high rate of compliance was observed, with the majority of participants completing the required attendance of education sessions or treatment, regardless of age, gender or Indigenous status
- two variables were identified as significant for predicting reoffending post-program; these were prior offending record (particularly multiple priors) and program-non-compliance
- although program participants varied considerably in terms of offending levels prior to and post diversion, changes in offending levels after diversion were relatively consistent across programs. This demonstrates the benefits of diversion, irrespective of variations in program design, demographics and the targeting of offenders
Diversion involves the redirection of offenders away from conventional criminal justice processes, with the aim of minimising levels of contact with the formal criminal justice system. Drug diversion programs have become increasingly popular in recent years for many reasons. These include the increased levels of incarceration of people for drug-related offences across much of the developed world, growing evidence that punitive responses alone have been unsuccessful in preventing the use of illicit drugs and the criminal activity associated with their use, and increasing awareness that (for many offenders) custodial sentences further compound harms associated with drug use (Bull 2003).
In its purest form the term diversion applies to processes that are at the very front-end of the criminal justice system (that is, at the pre-apprehension stage before any formal charges are laid) and are focused on diverting eligible individuals into education and/or treatment services, rather than to alternative forms of processing. In practice, diversion initiatives can be divided into several major clusters of interventions:
- police drug diversion programs
- court-based diversion programs (ranging from pre-court and pre- and post-sentence diversions, as well as programs at the higher end of the court system that include intensive pre- and post-sentencing drug court options such as long-term intensive treatment)
- drug treatment correctional centres which operate at the custodial level
Table 1 outlines police drug diversion programs currently available in Australian jurisdictions.1
Illicit Drug Diversion Initiatives (IDDI), funded under the Council of Australian Governments and introduced to provide a nationally consistent approach to drug diversion, often rely on police as the referral source. Each Australian state and territory has now implemented at least one police-based diversion program targeted at the use or possession of cannabis and/or cannabis implements. This bulletin attempts to highlight some of the key issues to do with police-based interventions.
Australian police drug diversion interventions
There has been significant investment in diversion by the Australian Government, however there is little knowledge about the implementation of these initiatives or a full understanding of their likelihood of success or comparative benefits against the costs and impacts of previous practice. In particular, literature on drug diversion as a whole is in consensus that there are significant gaps in knowledge about the true impacts of drug diversion and that information currently available is generally limited to findings from evaluations of interventions undertaken shortly after program implementation (Wundersitz 2007). In this context, while some interventions appear to demonstrate a positive impact, the extent to which they are instrumental in reducing drug use and related offending behaviour is inconclusive (Baker & Goh 2004).
The Drug Policy Modelling Program (DPMP) based at the National Drug and Alcohol Research Centre recently canvassed the opinions of senior drug policy bureaucrats in the health and police sectors around Australia to identify priority areas for illicit drug policy (Ritter 2007). Most respondents in this exercise suggested that the evidence-base for evaluating police diversion interventions was largely absent and that a greater evidence-base around the effectiveness (including cost-effectiveness) of diversion interventions was essential. Important questions that respondents felt were necessary to answer included inter alia:
- does mandated treatment have different treatment outcomes from voluntary treatment?
- is there an effectiveness difference between first versus repeat offenders?
- what is the return on investment for diversion interventions and is it different for different types of interventions?
- does the effectiveness of diversion differ depending upon the user?
One of the major barriers to police uptake of diversion and cautioning appears to be a perceived lack of feedback about the intervention. Another recent study (AIHW 2008) has found that police often do not receive information about the outcomes of diversion for the people they divert. In addition, they generally do not know if diversion interventions are considered effective and what the likely or anticipated outcomes are for offenders. The study suggests that the lack of feedback on diversion outcomes and effectiveness is likely to contribute to police perceptions of diversion being a soft option .
Diversion away from formal criminal justice system processes can be legislated by certain criteria or based on the discretion of police or magistrates. While discretion may be viewed as opening the door to inconsistency or discrimination (O Callaghan et al. 2004), an alternative view is that because of the complexities of cases presented to police, discretion permits necessary flexibility in the system. For instance, a review of the Cannabis Cautioning Scheme in New South Wales (NSW) (AIHW 2008) points to divergent views among law enforcement about the benefits of police discretion. Discretion is sometimes viewed as positive, enabling police to have flexibility and to not necessarily view a situation punitively. This increases the opportunity for an officer to identify and address other factors that may be impacting on drug use behaviour. However, some police indicated that discretion leads to inconsistent application between officers which could promote confusion in the community and could lead to internal problems (AIHW 2008).
Coercion into treatment
Policy makers and police have expressed interest in knowing whether coercion impacts on the effectiveness of diversion interventions (Ritter 2007) and literature on diversion is often concerned with the expectation that referral to treatment is less successful for coerced participants. However, studies have instead found that requirements surrounding the coercive aspect of police diversion are supportive of the treatment process and are associated with improved results. For example, a survey of treatment providers from the MERIT court-based intervention program reported treatment was more successful for diverted clients than for voluntary clients (Passey 2003). Similarly, a recent re-evaluation of the NSW Drug Court (Weatherburn et al 2008) also found positive outcomes for program participants. This greater success was attributed to longer treatment duration, regular court reviews and incentives to comply with program requirements. In another instance (Baker & Goh 2004), diversion interventions without a coercive aspect were found to result in poor treatment outcomes and demonstrate that voluntary treatment seeking was extremely low. On this occasion, only 0.7 percent of cautioned people voluntarily contacted a helpline in the NSW Cannabis Cautioning Program. A similar experience has recently been observed in Western Australia under the Cannabis Infringement Notice Scheme.
Diversion of indigenous offenders
Luke & Cunneen (1995) and Cunneen (2006) indicate that Indigenous young people are more likely than non-Indigenous young people to be arrested (rather than cautioned) and they tend to acquire a more extensive criminal record at a younger age than other young people. In addition, looking at variables that were significant predictors of diversion across Western Australia, South Australia and New South Wales, Snowball & Weatherburn (2007) found that past contact with the criminal justice system played just as important a role in shaping subsequent decisions about diversion as it did in shaping decisions about adult sentencing. Although diversionary alternatives appear to be effective in reducing Indigenous re-offending, Indigenous young offenders do appear much less likely to be diverted than their non-Indigenous counterparts (Luke & Lind 2002).
Costs and savings of diversion
There has only been one study (Baker & Goh 2004) that has examined the cost-effectiveness of diversion. This study, which examined the cost-effectiveness of the NSW Cannabis Cautioning Scheme, demonstrated positive results in savings to the criminal justice system. In the first three years of operation it was estimated that over 18,000 police hours were saved as a result of not having to:
- charge offenders at the time of detection
- prepare matters for court
- attend subsequent hearings
Baker & Goh (2004) calculated that the scheme resulted in total savings of well over $1 million during the first three years of operation. While these savings were balanced against the intervention s operational costs (estimated at approximately $1,096,000 over the first three years), the evaluators still concluded that the scheme paid for itself in its first three years. They also noted that because most of the costs identified were establishment costs expended in the first year of operation, these would reduce over time, thereby increasing potential savings.
Impact on court workloads
Baker & Goh (2004) found in their assessment of the NSW Cannabis Cautioning Scheme that there was an upward trend (of 16 percent) in the number of police charges laid every month for minor cannabis offences in the three years prior to the intervention; however, in the three years after the intervention s commencement, a downward trend (of 11 percent) was evident. Overall, 6,679 fewer charges were laid after the intervention s inception. While the number of court cases involving cannabis-only charges was relatively stable in the period preceding the intervention s commencement, there was a marked decline (26 percent) during the first three years of the intervention s operation. Overall, there were 5,241 fewer cannabis cases processed by the court in the three years after the cautioning intervention was established compared with the three years before.
Impact on crime and social costs
In evaluating the success of police drug diversion interventions it is also important to consider other outcomes, such as impact on crime and costs to the community. Tim Moore (2007) recently undertook a study that estimated the annual social costs of different drug types in Australia. He found that for every dependent cannabis user it cost the community $11,296 each year (compared to $105,342 for a dependent user of opiates). However, the social cost for a non-dependent cannabis user was estimated to be $192 per year. Total social costs due to dependent cannabis users were estimated to be a little over $2.7 billion every year, compared to $319 million for non-dependent users (dependent and non-dependent opiate users accounted for $4.3 billion and $212 million, respectively). As such, the costs generated by dependent users are estimated to be at least 8 times the average social costs generated by non-dependent users. If diversion interventions are indeed effective in managing dependent cannabis users of the drug then this would be compelling grounds for implementing targeted treatment and education diversion interventions (like IDDI) to reduce harms and costs to individuals and to the community.
AIC research on police diversion offending outcomes
The Australian Institute of Criminology (AIC) recently released a report that outlined findings from a systematic evaluation of police-based IDDI interventions across Australia (Payne et al 2008). As part of the evaluation, the research team collected outcome data from all jurisdictions and examined the impact of IDDI interventions on crime and re-offending outcomes of participants. The evaluation also acted as one of the first comprehensive studies describing police diversion and provided important insights into:
- client profiles of diversion participants
- compliance levels and predictive factors in compliance
- the offenders who respond best to diversion, particularly in terms of discouraging future re-offending
Key findings from this evaluation are highlighted below.
In terms of the demographic and offending profiles of people referred to police-based IDDI interventions, the evaluation found that the majority of IDDI participants were male from 70 percent in the Northern Territory to 86 percent in New South Wales (Table 2). This is consistent with the profile of all offenders dealt with by the criminal justice system, where males comprise the majority of people apprehended by police, prosecuted in court and sentenced to imprisonment. In contrast, there were marked inter-program variations in all other areas assessed (see next page).
The age of IDDI participants varied greatly, with interventions that focussed on juveniles having a lower mean age than interventions catering for both juveniles and adults or for adults only (Table 2). In the Northern Territory, where the majority of those referred were juveniles, the mean age of people referred to its drug diversion interventions was 15.2 years. In comparison, the mean age in South Australia’s Police Drug Diversion Program, which accepts referrals for both adults and juveniles, was 21.5 years. In Western Australia, where referrals are limited to adults, the mean age was 26.8 years.
The percentage of Indigenous IDDI participants varied across the interventions (from 1 percent in Victoria and the Australian Capital Territory to 31 percent in the Northern Territory). While the evaluation found that this was partly reflective of the size of the Indigenous population in jurisdictions, it did not appear to account for all variation. For example, although Tasmania has a relatively small Indigenous population compared to Queensland, 13 percent of those diverted in Tasmania were Indigenous, whereas only eight percent of those diverted in Queensland were Indigenous (Table 2).
Summary demographic profile of diverted people by jurisdiction
One explanation for these findings may lie in the type of intervention eligibility criteria applied, particularly those pertaining to prior criminal records (with Indigenous people potentially having longer criminal histories that may preclude them). A further explanation could be that individuals in most jurisdictions must admit to the offence that has been detected and consent to the intervention to be diverted; a course of action which has frequently been demonstrated to exclude Indigenous people (Payne et al 2008).
In the evaluation, prior offending records also varied widely among jurisdictions. Total figures indicated that people referred in NSW had the lowest levels of prior offending in the country, with only 13 percent charged with at least one criminal incident in the 18 months before diversion (including 8 percent with a prior property offence, 3 percent who had at least one prior violent offence and 2 percent with a prior drug offence). Moreover, of those NSW referrals with a criminal record, almost two-thirds had been charged with only one incident in the 18 months preceding diversion, while very few (8 percent) had been charged with four or more prior events. These results are in marked contrast to those of Tasmania, which had the highest levels of prior offending. In that state, 48 percent of its diversion clients had offended during the preceding 18 months, including 28 percent charged with a prior property offence, 11 percent who had committed at least one previous violent offence and nine percent who had a prior drug offence. A priori, these differences are marked. However, the report also indicates that differences in prior offending histories of clients are probably mostly due to program design and the marked variations in the eligibility criteria governing access to the various interventions.
Many of the interventions evaluated recorded compliance rates of 100 percent. This was because these interventions involved the simple issue of a caution by a police officer, with no further action required by the offender. For example, the NSW Cannabis Cautioning Scheme,Tasmania s First Level Diversion of cannabis offenders and Victoria’s Cannabis Cautioning Program each involved the distribution of education material only, with no further obligations placed on the offender.
Requirements for intervention compliance were found to vary greatly, as did compliance rate results. Other cannabis cautioning interventions assessed, together with those targeting any other illicit drug use, required attendance at one or more assessment and treatment sessions, with non-compliance potentially resulting in the individual being prosecuted in court for the original offence (or, as in the case of Western Australia and Queensland, facing a financial penalty). However, for interventions where some level of compliance was required, the majority of people referred actually fulfilled those requirements, which was a positive finding.
The evaluation also found that compliance levels varied according to the characteristics of offenders, although the direction of the relationship was not necessarily consistent across all interventions. For instance, the study did not find any pattern for gender and, no consistent patterns emerged in relation to juvenile versus adult status in terms of predicting non-compliance.
More consistent patterns did, however, emerge in relation to offending histories. Analysis of the interventions revealed that most diversion schemes had higher non-compliance levels among individuals who had at least one prior offence recorded against them in the 18 months before intervention referral compared with those who had no prior record. The importance of prior offending as a key explanatory factor was verified by regression analysis. Excluding cautioning schemes where no compliance was required, the evaluation found that in four of the five jurisdictions analysed, some form of prior offending (and in particular, property offending) remained a significant predictor of noncompliance once factors such as age, gender and Indigenous status had been controlled.
Shifts in pre and post-offending levels
Results of overall changes in the rates of re-offending for each individual were positive, particularly in relation to individuals who had a prior offending history. Among this group, the majority were apprehended for either no or fewer offences post-intervention than prior to entering the intervention. This finding was consistent across all jurisdictions. Figure 1 shows that of all those who had offended at least once during the 18 months before diversion, between 53 percent (Australian Capital Territory) and 66 percent (New South Wales and Victoria) recorded fewer offences in the 18 months after diversion. The percentage change across these interventions was comparatively similar, despite the marked variations among them in terms of absolute pre and post-offending levels. Of those individuals who had not offended in the 18 months before diversion, the majority (ranging from 70 percent in Tasmania to 86 percent in New South Wales) also remained non-offenders in an equal period after diversion.
Pre-post change in offending among those with prior offences (percentage)
Taken as a whole, the evidence strongly suggests that Australian drug-crime police-based diversionary schemes produce a range of positive outcomes. These are most evident in terms of impacts on offenders, the criminal justice system and on the general community. Findings from the AIC s systematic and comprehensive outcome evaluation of police-based IDDI interventions indicated particularly positive results. Overall, it appears that compliance levels were high and that the majority of people referred to IDDI interventions did not offend post-intervention, at least during the follow-up period studied. Moreover, even among those individuals with prior criminal records, the majority of individuals significantly decreased their offending behaviour after participating in a diversion intervention. As such, it appears that irrespective of the category of offender that is being targeted, interventions impacted positively on both entrenched offenders as well as predominantly non-offending drug users.
Given the significant organisational and operational differences among police-based IDDI interventions (for example, in relation to their set-up and eligibility criteria) and that the interventions varied considerably in terms of both pre- and post-diversion offending levels, one of the most telling findings from the evaluation was that the proportionate decrease in offending after diversion was relatively consistent across all jurisdictions. So even though one intervention may have started with higher levels of prior offending and recorded higher levels of offending after diversion, the degree of change among its clients was relatively similar to interventions with lower pre- and post-offending levels. This suggests that drug crime diversionary interventions that are tailored to particular communities and drug problems can be expected to have a generally positive impact as long as they are well designed and implemented.
- Australian Institute of Health & Welfare. (2008). The effectiveness of the Illicit Drug Diversion Initiative in rural and remote Australia. Drug Statistics Series. No. 19. Canberra: Australian Institute of Health and Welfare.
- Baker, J. & Goh, D. (2004). The Cannabis Cautioning Scheme three years on: An implementation and outcome evaluation. Sydney: NSW Bureau of Crime Statistics and Research.
- Cunneen, C. (2006). Racism, discrimination and the over-representation of Indigenous people in the criminal justice system: Some conceptual and explanatory issues. Current Issues in Criminal Justice 17, 329-346.
- Luke, G. & Cunneen, C. (1995). Aboriginal over-representation and discretionary decisions in the NSW juvenile justice system. Sydney: Juvenile Justice Council of NSW.
- Moore, T. (2007). Working estimates of the social costs per gram and per user for cannabis, cocaine, opiates and amphetamines. Drug Policy Modelling Program, Monograph 14. Sydney: National Drug and Alcohol Research Centre.
- Payne, J., Kwiatkowski, M. & Wundersitz, J. (2008). Police drug diversion: A study of criminal offending outcomes. Research and Public Policy Series. No. 97. Canberra: Australian Institute of Criminology.
- Ritter, A. (2007). Priority areas in illicit drug policy: Perspectives of policy makers. Drug Policy Modelling Program, Monograph 15. Sydney: National Drug and Alcohol Research Centre.
- Snowball, L. (2008). Diversion of Indigenous juvenile offenders. Trends & Issues in Crime and Criminal Justice. No. 355. Canberra: Australian Institute of Criminology.
- Weatherburn, D., Jones, C., Snowball, L., & Hua, J. (2008). The NSW Drug Court: A re-evaluation of its effectiveness. Crime and Justice Bulletin. No. 121. Sydney: NSW Bureau of Crime Statistics and Research.
- Wundersitz, J. (2007). Criminal justice responses to drug and drug-related offending: Are they working? Technical and Background Paper. No. 25. Canberra: Australian Institute of Criminology.