Reintegration of Indigenous prisoners: key findings

Foreword | The overrepresentation of Indigenous Australians in prison has long been one of the major challenges facing the criminal justice system. To date, there has been no national study of the extent to which Indigenous people reoffend and are returned to prison. Using data from all Australian jurisdictions, and covering 8,938 males incarcerated for violent offences and released from prison over a two-year period, this paper clearly shows that Indigenous offenders are readmitted to prison sooner and more frequently than non-Indigenous offenders. Analysis shows that Indigenous prisoners are nearly twice as likely to have been readmitted to prison within two years and more than twice as likely to return to prison for assault. Half of the Indigenous prisoners remained in prison until the expiry of their sentence, which makes post-release support particularly challenging. A key theme that emerged from stakeholder consultations was the need to improve support during transition back into the community through the involvement of family and community, and increased capacity to undertake throughcare, especially in remote settings. While there is a range of Indigenous-specific initiatives and programs in prisons and in the community, including those that target violent offending and substance abuse, an ongoing need is to increase participation in mainstream services and to include Indigenous cultural specificity without negatively impacting on program integrity, and to evaluate the outcomes of reform over time.

Judy PuttGeneral Manager, Research


There is abundant evidence, produced over an extended period, to show that Indigenous people are heavily overrepresented in the Australian criminal justice system. Within prison populations especially, the proportion of Indigenous people is far in excess of their representation in the general community.

Indigenous Australians are incarcerated at a far greater rate than non-Indigenous Australians. The age-standardised rate of imprisonment of Indigenous prisoners at 30 June 2007 (1,786.7 per 100,000 adults) was more than 13 times higher than that of non-Indigenous prisoners (133.5 per 100,000 adults) (ABS 2007). The crude Indigenous imprisonment rate has increased from 1,335.5 per 100,000 adults in 1996 when comparable figures first became available, to 2,255.5 per 100,000 adults in 2007 (ABS 2007).

Indigenous prisoners are most often incarcerated for violent (particularly assault-type) offences. While comprising 24 percent of the overall prisoner population, Indigenous prisoners comprised 42 percent of those whose most serious offence or charge was an act intended to cause injury, a category that includes assault (ABS 2007). Thirty-two percent of all Indigenous prisoners had this recorded as their most serious offence or charge. While this was also the most serious offence or charge for non-Indigenous prisoners, it only accounted for 19 percent of the non-Indigenous prisoner population.

Although statistics clearly show levels of Indigenous offending and overrepresentation throughout the justice system, there is little empirical evidence about Indigenous reoffending and reimprisonment. While it is clear that rates of Indigenous imprisonment exceed non-Indigenous imprisonment, it is not clear whether this carries through to higher rates of reoffending and reimprisonment.

Among the few studies that have investigated recidivism among Indigenous offenders, an examination of reoffending in the Northern Territory found that 45 percent of Indigenous prisoners released during 2001-02 were readmitted to prison for a new term of imprisonment within two years, compared with 15 percent of non-Indigenous prisoners (NT Office of Crime Prevention 2005). Offenders faced a higher risk of returning to prison if they had previously been imprisoned or incarcerated for assault.

A NSW study of reoffending among parolees estimated that Indigenous offenders were 1.4 times more likely than non-Indigenous offenders to reoffend at any time following release (Jones et al. 2006). A longitudinal study of ex-prisoners re-entering the community found Indigenous offenders were far more likely to return to prison during a nine-month period than non-Indigenous offenders (Baldry et al. 2003).

The purpose of national research undertaken for state and territory corrective services was to contribute to the national understanding of the reintegration of Indigenous prisoners. This was done using quantitative data on reimprisonment and qualitative data on the issues surrounding successful reintegration, with an emphasis on programs that might reduce violent offending. For the purposes of this research project, reintegration was defined in terms of not being readmitted to prison either on remand or for a prison sentence. This is an indirect measure of reintegrative success, but one that could be measured within the time and budget constraints of the research project.

This paper only summarises key findings from the full report (Willis & Moore 2008).

Readmission to prison: quantitative data

Administrative data on readmission to prison for a cohort of both Indigenous and non-Indigenous released prisoners was provided to the Australian Institute of Criminology (AIC) by all jurisdictions across Australia. This represents the first time that national data on released prisoners readmitted to prison have been analysed.

The sample was comprised of all distinct sentenced male prisoners (both Indigenous and non-Indigenous) who were:

  • convicted of and imprisoned due to a violent offence
  • released from prison between 1 January 2001 and 1 January 2003
  • released either on parole or due to sentence expiry.

De-identified unit record data were provided for each released prisoner on a range of variables, including readmissions to custody for a two-year period following release.

Data were provided for a total of 8,938 violent male prisoners meeting the above criteria. Some 35 percent of all prisoners were Indigenous.

Demographics and background


The majority of prisoners were aged between 18 and 25 years (35%) or 26 and 35 years (37%) at the time they were admitted to prison. More Indigenous prisoners were aged 35 or under (80%) than non-Indigenous prisoners (69%). Few Indigenous prisoners were aged over 45 years (4%) compared with non-Indigenous prisoners (13%).


Data on the level of completed education were available for five jurisdictions (excluding the Australian Capital Territory, New South Wales and Victoria). Where known, all prisoners in the cohort had relatively low levels of education. Indigenous prisoners had lower levels of education than non-Indigenous prisoners. Some 37 percent of Indigenous prisoners had less than a Year 9 level of education, compared with 21 percent of non-Indigenous prisoners. Less than seven percent of Indigenous prisoners had completed Year 12 compared with 16 percent of non-Indigenous prisoners.

Most serious offence

Consistent with the national prisoner profile, most Indigenous prisoners in the release cohort had been imprisoned for assault (77%). Less than half (46%) of non-Indigenous prisoners had been imprisoned for assault. Sexual assault was the most serious offence for eight percent of Indigenous prisoners, compared with 20 percent of non-Indigenous prisoners.

Also consistent with national census data, Indigenous prisoners tended to receive shorter sentences overall, with both the mean (614 days) and median (360 days) being lower across total offences than they were for non-Indigenous prisoners (908 days and 511 days respectively).

Prior adult imprisonment

Prior to the current imprisonment period, 74 percent of Indigenous prisoners had previously served an adult prison sentence, compared with under half (47%) of non-Indigenous prisoners. Again, this is highly consistent with the prison census data (ABS 2007).

Figure 1 : Released prisoners readmitted to prison within six, 12, 18 and 24 months (percentage)

Note: Includes NSW and ACT-sentenced prisoners transferred to NSW facilities

Source: AIC Reintegration of Indigenous Prisoners [computer file]; n=8,938

Type of release

Half (50%) of the Indigenous prisoners in the subject cohort had remained in prison until the expiry of their sentence, with a smaller percentage (42%) released before the end of their sentence on parole. A further eight percent of Indigenous prisoners were released through other mechanisms, such as bonds.

In contrast, most non-Indigenous prisoners (59%) had been released on parole, with only 39 percent held until sentence expiry. Two percent were released through other mechanisms.

Information provided for this project suggested that Indigenous prisoners, particularly in some jurisdictions, may decline offers of parole due to the difficulties of meeting reporting requirements and other conditions.

Readmission to prison

Across Australia, Indigenous prisoners in this release cohort were more likely (55%) than non-Indigenous prisoners (31%) to have been readmitted to prison within two years of release. As well as returning to prison at a higher rate, Indigenous prisoners tended to return sooner. Within six months of release (Figure 1), one-quarter of released Indigenous prisoners had been readmitted to prison, twice the percentage of non-Indigenous released prisoners (12%). Two-fifths of Indigenous prisoners (39%) had been readmitted to prison custody within 12 months, compared with one-fifth of non-Indigenous prisoners. There was some variation between jurisdictions as to the relative likelihood of Indigenous prisoners being readmitted compared with non-Indigenous prisoners, but in each jurisdiction Indigenous prisoners were more frequently readmitted to prison during the two-year period than non-Indigenous prisoners.

The finding that Indigenous prisoners tend to be readmitted to prison sooner than non-Indigenous prisoners is underscored by Figure 2. This shows the percentage of released prisoners remaining in the community on a daily basis within two years of release. At the time of release (day 0), all released prisoners (100%) remained in the community without having been readmitted to prison.

Figure 2: Number of days released prisoners remained in the community following release from prison (percentage)

Source: AIC Reintegration of Indigenous Prisoners [computer file]; n=8,938

At 365 days, 61 percent of Indigenous prisoners remained in the community compared with 79 percent of non-Indigenous prisoners.

As time progressed, an increasingly smaller percentage of released prisoners remained in the community as more were returned to prison. The percentage of Indigenous prisoners remaining in the community was consistently lower than the percentage of non-Indigenous prisoners.

Reoffence types

The reoffending patterns of offenders readmitted to prison within two years of first release were largely similar to the patterns of their original offending. It was markedly apparent that Indigenous violent offenders were far more likely to be readmitted to prison for assault (44%) than any other type of offence (Table 1). The percentage of readmitted Indigenous prisoners charged with assault was more than double the percentage of non-Indigenous offenders (20%), which is consistent with ratios seen for the 'original' offences for this release cohort. Non-Indigenous reoffenders were more likely to have been readmitted for robbery, break and enter, or theft offences than their Indigenous counterparts.

Table 1 : Most serious reoffence type for those readmitted to prison within two years
  Indigenous Non-Indigenous Total
  n % n % n %
a: Includes prisoners readmitted to custody for a new offence (n=1,670), and prisoners readmitted for a breach/technical violation (n=269, included in 'other offences') or other reason (n=4). Excludes cases where reoffence was shown as not available (n=89) or not applicable (n=220). Among these excluded cases may be some cases where readmission was for a breach of parole or other order that did not fit within the available categories.
Note: Excludes NSW and ACT-sentenced prisoners transferred to NSW facilities. Figures may not total 100 percent due to rounding.
Source: AIC Reintegration of Indigenous Prisoners [computer file]
Assault 472 44 171 20 643 33
Break and enter 78 7 117 13 195 10
Theft and related 60 6 105 12 166 9
Robbery 28 3 92 11 120 6
Traffic 103 10 74 8 177 9
Other violence 72 7 75 9 147 8
Other 255 24 240 27 495 25
Total 1,068 100 874 100 1,943a 100

Remand outcomes

Data on remand outcomes were available for a small group (n=350) within the cohort; 23 percent were Indigenous and 77 percent non-Indigenous. Indigenous prisoners were equally likely to have been convicted (46%) as released without conviction (44%) following remand. In contrast, most non-Indigenous prisoners had been convicted (65%) and fewer not convicted (26%).

Time spent in remand

Data were available for 306 prisoners showing how long they had spent in remand custody before receiving an outcome in court. For those who went to court and were convicted, there was no significant difference in the average amount of time Indigenous prisoners spent in custody before conviction, compared with non-Indigenous prisoners.

Conversely, Indigenous prisoners released without conviction tended to spend less time on remand (mean=53 days, median=33 days) than comparable non-Indigenous prisoners (mean=104 days, median=49 days).

It is interesting to note from remand-related findings from this study - alongside observations from both this study and earlier research - that Indigenous offenders typically receive shorter sentences than non-Indigenous offenders. It is possible that remand outcomes found in this study may be the result of the judiciary being unwilling to further punish some Indigenous offenders who have already spent time on remand. Any trend towards Indigenous offenders spending less time on remand would be a positive one. The limited remand data available for this study only allow speculative conclusions. Further study focusing on remand outcomes would be needed to examine differences in Indigenous and non-Indigenous remand outcomes.

Reintegration: qualitative data

To complement findings gained from the quantitative data, semi-structured interviews were conducted with 34 serving prisoners and seven ex-prisoners across the Northern Territory, Queensland, South Australia and Western Australia. Participants were asked about their views on programs and services they had accessed.

Interviews, surveys and consultations were conducted with a total of 44 prison and community corrections key respondents in these same jurisdictions. Participants were asked about their views on the delivery of programs and services for violent male Indigenous prisoners, how programs and services could be improved, and about the barriers to improvement and successful reintegration.

Participating in programs

A small number of serving prisoners provided information on the rehabilitative programs they had undertaken, how useful they felt those programs were and the elements that made them useful.

The interviewed prisoners had undertaken a range of violent offending programs and substance abuse programs, with most feeling that the programs were useful to them. The prisoners generally felt that violent offending programs were helpful in reducing their use of violence, and it was the development of skills to control anger and avoid violence that provided the greatest benefit. The prisoners also felt that substance abuse programs were useful in reducing their likelihood of using substances in the community. The programs were particularly educational on the impacts of drug and alcohol use, which tended to make them useful.

Most of the prisoners suggested that the programs they had undertaken had adaptations to make them Indigenous-specific. It appeared this was often the result of having mostly Indigenous participants, which tended to steer the program towards involving Indigenous perspectives, rather than adaptations in the program content.

Program improvements and barriers

Prisoners, staff and other stakeholders were asked for their views on how programs and services for violent male Indigenous prisoners could be improved to better facilitate reintegration and what issues might arise to create barriers to these improvements.

Most of those interviewed identified the need for programs and services to be adapted to meet Indigenous cultural-specific needs. Many interviewees saw a lack of Indigenous-specific programs and services as a major barrier to participation and successful reintegration. The absence of Indigenous-specific content in core programs was thought to reduce the effectiveness of interventions by lowering the responsiveness of Indigenous offenders to treatment.

Stakeholders highlighted the need for more knowledge on the role of culture and cultural content in prison-based and community-based programming, as well as more knowledge concerning the specific criminogenic needs of Indigenous offenders. A number of prisoners and stakeholders emphasised the need to address alcohol and substance use, particularly through Indigenous-specific approaches.

A range of suggestions was put forward to increase the relevance of programs for Indigenous offenders, including:

  • incorporating an understanding of Indigenous society and its collectivist approach, and the resistance of many Indigenous people to disclose information about themselves, rather than relying on programs developed from a Western perspective that emphasises self-disclosure, self-awareness and individual responsibility
  • recognising the place of violence in Indigenous communities and how it contributes to offending
  • applying holistic methods that address the mind, body and spirit
  • enhancing those elements that appear to make programs effective, such as skill development and education, by making them more directly relevant to the life experiences and circumstances of Indigenous people and their communities
  • increasing responsivity and participation by addressing issues such as anxiety, anger and resistance, which can interfere with Indigenous prisoners' willingness and capacity to participate in programs
  • making programs more enjoyable and engaging through visual content and physical activities
  • overcoming language and literacy barriers for those offenders with limited English language skills
  • involving elders and Indigenous facilitators in the development and delivery of programs.

Participants discussed some of the difficulties with increasing cultural specificity, including the diversity within Indigenous cultural perspectives and maintaining program integrity, as well as the importance of effective assessment and evaluation.

Interviewees highlighted the importance of community and family involvement in programs and services, pointing to the fundamental influences of community and family on the behaviour of Indigenous prisoners, as well as the impacts they experience in return. A theme that emerged in each consultation was the importance of throughcare in the delivery of programs and services. They felt that efforts to assist prisoners while in custody were often lost at the point of release. Stakeholders raised the need to achieve transition between custodial and community corrections, and to link with communities to provide ongoing support to offenders. They also recognised the difficulties of doing this in practice, given issues of remoteness and limited services.

The remoteness and the lack of services of many Indigenous communities were also identified more generally as major issues in the delivery of programs and services to Indigenous offenders. The disadvantages that Indigenous communities face can work against efforts to address disadvantage and achieve reintegration. This can impact on corrections in direct ways, such as by making it much harder for parolees to comply with the requirements of their orders, which results in breaches and return to imprisonment.

Other factors identified as particularly important to address in achieving Indigenous reintegration included:

  • addressing the grief and loss that consumes many Indigenous people
  • adequately responding to mental health problems
  • achieving reintegration for those serving short sentences and on remand who rarely receive correctional programs and services.

Many stakeholders saw resourcing constraints as a major impediment to make the kinds of improvements needed to achieve effective Indigenous reintegration. Many stakeholders identified limited funding as being a continuing barrier, a problem exacerbated by the difficulties in providing services to offenders in remote communities. Even where services are set up, it can be very difficult to find sufficient appropriate and qualified staff, especially those willing to work in remote areas.

An important point made by one group of stakeholders is that many of the limitations and barriers they experienced in developing and delivering quality interventions were not restricted to Indigenous offenders, but were problems across their entire agency or jurisdiction.


What emerges from the analysis of administrative data is a picture that clearly demonstrates the extent of Indigenous violent offending and, perhaps more importantly, reoffending. Indigenous men are being imprisoned for assault, released and reimprisoned for assault, and this tends to recur. The factors that contribute to their offending while they are in the community remain there when they returned to the community. Despite the best efforts of correctional administrators, practitioners and others to provide a range of support services, Indigenous men are not properly equipped to deal with the risk factors they confront, hence the aims of reintegration are not achieved.

While much has been written about Indigenous offending and been done to put that knowledge into practice, there remain areas to investigate. More needs to be known about what Indigenous offenders experience when they return to their communities, and how the benefits of correctional programming can be maintained and enhanced after release from prison. Further research is needed into how to increase program participation among Indigenous prisoners, and to identify the specific elements and delivery techniques needed to increase Indigenous cultural specificity without negatively affecting program integrity. Program development must also be supported and facilitated by evaluation, although determining success may be difficult unless Indigenous perspectives are considered.

The problems of Indigenous disadvantage extend well beyond the purview of corrections and all parts of the justice system. Correctional initiatives alone will not stop Indigenous men committing violent crimes. Efforts towards individual reintegration must ultimately be supported by major changes at the community level. Correctional approaches must involve throughcare principles and engage family, community members and respected persons like elders, within the context of much broader improvements to relieve social disadvantage, if lasting change is to be realised.


This project was funded by the corrective services agencies of Australia and New Zealand through the Corrective Services Administrators Council.


All URLs were correct at 20 June 2008

  • Australian Bureau of Statistics (ABS) 2007. Prisoners in Australia, 2007. ABS cat. no. 4517.0. Canberra: ABS
  • Baldry E, McDonnell D, Maplestone P & Peeters M 2003. Ex-prisoners and accommodation: what bearing do different forms of housing have on social integration? Melbourne: Australian Housing and Urban Research Institute
  • Jones C, Hua J, Donnelly N, McHutchison J & Heggie K 2006. Risk of re-offending among parolees. Contemporary issues in crime and justice no. 91. Sydney: NSW Bureau of Crime Statistics and Research
  • Northern Territory (NT) Office of Crime Prevention 2005. Recidivism in the Northern Territory: adult prisoners released in 2001-02. Darwin: Northern Territory Office of Crime Prevention
  • Willis M & Moore J-P 2008. Reintegration of Indigenous prisoners. Research and public policy series no. 90. Canberra: Australian Institute of Criminology.

About the Author

Matthew Willis is a research analyst with the Australian Institute of Criminology.