Corrections education and employment assistance ‘Down Under’: Current and emerging practices and paradigms 1

This article provides a view of contemporary Australia in terms of patterns of offending and incarceration, the characteristics of its correctional systems, vocational education and training (VET) within correctional settings, and post-release employment of prisoners and offenders serving community-based orders (CBOs). A two-year case study of employment assistance for 2,458 Australian prisoners and offenders serving CBOs was evaluated. The voluntary 12-month programme targeted participants at moderate to high risk of reoffending. Overall, employment outcomes were positive with more than one-third of registrations resulting in employment. Employment outcomes varied for gender and participant status (prisoner/offender). Recidivism outcomes were analysed for the whole programme and for a random sample of 600 prisoner participants. Results indicated a very low overall recidivism rate (7.46 per cent) for programme participants, and comparison of pre-programme and post-release recidivism showed reduced recidivism on three recidivism measures. The findings are contextualized in terms of current thinking and emerging practices in offender treatment, with a focus on reintegration as ecological system engagement and integrated systems of support as central to promoting positive lifestyle change.


Introduction
It is well understood that the high rate of growth in the prison population over the past decade has resulted in a large number of individuals re-entering the community following prison release. An important strategy in addressing this issue has been the development of programmes that promote successful community reintegration and reduced rates of reoffending. Employment is a key factor in this endeavour, with benefits relating to the individual and family (Hoare and Machin, 2010;Richie, 2001), wider system-level benefits (see Solomon et al., 2004;Uggen, 2000;Zhang et al., 2006), and potential social justice benefits as well. This article provides a contemporary view on post-release employment of prisoners and people serving community-based orders (CBOs) in Australia and on interventions that have been implemented to enhance employment opportunities. It then presents in depth the outcomes of one of these interventions in the state of Victoria. It concludes with a proposed model for an integrated system of support, which is considered a necessary element of provision to help offenders get work and stay in work.
In-prison programmes can provide valuable employment-relevant skill acquisition. However, nationally and internationally, low participation rates in vocational training and education programmes within prisons remain a problem. In Australia, the low proportion of prisoners achieving a certification or qualification may be explained to some extent by the impact of typically short sentences on programme participation. There are long waiting lists for courses, indicating that demand is much higher than availability (Callan and Gardner, 2007). Clearly, there is a chronic, yet urgent need to improve programme participation in prisons so that individuals leave prison with a skill base that can improve their chances of successful employment after release. Policy and programme reform is also urgently needed to reduce waiting lists for courses and to increase resources to provide vocational training and education.
Researchers have identified the key qualities of successful post-release employment programmes (Borzycki, 2005). These include networking with the labour market; modifying recruitment and placement procedures in line with labour market needs; and providing employer incentives, information about job opportunities, appropriate vocational training, work-release opportunities for suitable offenders, job-retention skills, assistance to offenders in appropriate disclosure of criminal history, and long-term follow-up support.
In addition to the need for employment assistance, there is the issue of stakeholder attitudes toward the employability of ex-prisoners. Employers, employment service providers and corrective services staff have been shown to rate the employment prospects of ex-prisoners as poor (Graffam et al., 2008), with variations related to the type of crime and whether training had been completed in prison. Attitude change can be promoted by exposing employers and employment service providers to ex-prisoners through job fairs and work trials, and by publicizing success stories. Employer incentives have also been shown to be effective in influencing employers to hire ex-prisoners (e.g. The Bridge Project, 2011).
Given research evidence that unemployment is related to re-offending and reconviction (Baldry et al., 2003;Webster et al., 2001), it is clearly important that ex-prisoners receive the support they need to become employed, productive, and integrated members of their communities. Ex-prisoners generally have poor work histories and are typically employed in low-wage jobs with few tangible rewards (Solomon et al., 2004). There are numerous identified barriers to employment that may affect ex-prisoners (Graffam et al., 2008;Graffam et al., 2005;Webster et al., 2001). These include a lack of personal and work-related skills; educational disadvantage/ low literacy levels; unfavourable employer attitudes; racism on the part of employers; lack of local work opportunities; lack of job contacts because of segregated social networks; financial difficulties affecting interview/job; problems making the transition from benefits to employment; behavioural problems; lack of basic skills and/or poor qualifications; low self-esteem, confidence and motivation; absent or poor work experience history; and difficulty adjusting to the routine of work (e.g. Fletcher, 2001;Heinrich, 2000;Mukamal, 2001; National Association for the Care and Resettlement of Offenders, 1997;Visher and Travis, 2003;Webster et al., 2001).

Australian correctional system(s): Offending, incarceration, and treatment
Australia does not have a unified, federal correctional system; each state and territory has its own system. There is some variability in terms of crime rates and incarceration rates which is primarily due to differences in composition of the state/territory populations. Sentence lengths are typically quite short throughout Australia with a third of prisoners serving less than two years, with the median length being three years, three months (Australian Bureau of Statistics, 2011). Nationally, approximately one-third of convicted offenders are incarcerated and twothirds are serving community-based orders.
The number of offenders in correctional institutions in Australia in June 2011 was 29,106, with a rate of imprisonment of 167 per 100,000 or 0.00167% of the adult population. Over half of prisoners in custody at that time had previously served a sentence in an adult prison (Australian Bureau of Statistics, 2011), with 38 per cent of prisoners released from 2006−07 returning to prison, and 45 per cent returning to corrective services but not incarceration (Australian Institute of Criminology, 2011). About a quarter of adult prisoners are reconvicted within three months of release, with between 35 per cent and 41 per cent being re-imprisoned within two years (Payne, 2007).

VET in the Australian correctional system(s)
The history of vocational and educational training (VET) within Australian correctional systems is a century long, and varied. Basic adult education has been available in Australian prisons since the early 1900s. Initially, this was in terms of literacy and numeracy education and vocational skills acquisition. Programme diversity expanded over time. In 1996, the federal government produced the Senate Report of the Inquiry into Education and Training in Correctional Facilities (Senate Employment, Education and Training References Committee, 1996), which recommended development of a national strategy to address the VET needs of offenders. In 2001, the National Strategy for Vocational Education and Training for Adult Prisoners and Offenders in Australia (Australian National Training Authority, 2001) was launched. There are four objectives in the national strategy: 1. improve access to VET 2. increase participation and attainment across a range and levels of VET 3. contribute to reintegration pathways 4. establish accountable system and equitable outcomes.
By 2006, all states and territories had a suite of VET courses in place within correctional facilities and available to those completing community orders or sentences. Within the general community, VET is delivered through a network of Technical and Further Education (TAFE) colleges. In some states/territories, TAFEs deliver VET within the corrections system as well. At present, VET is more valued and popular among prisoners than prison industry or service work, but less attractive to offenders serving CBOs (e.g. Callan and Gardner, 2007). Generally, there is a wide range of courses, from short courses to trade qualifications and tertiary diplomas. Within each system, there is good collaboration between educationists and clinical, industry, and custodial services. Due to differences in political priorities and economic conditions, there is variability across jurisdictions in the range of courses available. Everywhere, demand for courses exceeds availability.
In Australia, VET participation plays a part in sentence management, but participation is dependent on sentence length. As described, short sentences and waiting lists often preclude a person's participation (Callan and Gardner, 2007). Prisoners can engage in prison industry work, service work, and/or VET. Small gratuities are awarded for participation in each of these activities. VET activities range from short courses to apprenticeships and traineeships. Currently, more than three-quarters of prisoners are eligible for participation in VET. More than threequarters of prisoners work in prison industry or services and only a small percentage of prisoners and offenders undertake apprenticeships or traineeships (but this is rising) (Steering Committee for the Review of Government Service Provision, 2014). Generally, there is a high level of collaboration between VET services, clinical services, custodial services, prison industry, and employment services.
Education and training are clearly critical to successful reintegration and lifestyle change. Demand for VET is typically higher than interest in prison industry or service work. The need for a transition model that supports participants to completion is evident. Being able to commence a course while in prison or serving a CBO and seamlessly continue to completion post-release would be a significant improvement on the present situation.

Post-release employment assistance
Within Australia, employment assistance is provided through the Department of Employment (formerly Department of Education, Employment, and Workplace Relations (DEEWR)). Job Services Australia (JSA) commenced on 1 July 2009. Between July 2009 and March 2011 exoffender job seekers constituted approximately 11 per cent of total caseload. Over that period, 87,472 ex-offender job placements were achieved. With respect to outcomes where prisoners remained in employment for 13 weeks, 27,069 ex-offender outcomes were achieved and with respect to 26-week outcomes, 13,985 ex-offender outcomes were achieved (figures courtesy of DEEWR).
In addition to the mainstream employment-assistance services for which all Australians are eligible, in some states and the Northern Territory employment assistance is offered as part of transition support by correctional services. In Queensland, population of 4.7 million, Advance2Work is a transition employment-assistance programme that is jointly funded by the Queensland Department of Corrective Services and Department of Education and Training. It has been operating across Queensland since 2000. The programme provides support from five locations to prisoners who are released from all custodial centres in Queensland. Prisoners begin the programme within the six months preceding release. Programme providers ensure that the participant profile is representative of the state's prisoner population. The programme provides a range of employment-related supports including training-needs analysis, vocational training, job search skills, job placement, post-employment placement support, and referrals. Between July 2009 and July 2010, 3,094 persons were assisted. Of those, 717 (23.2 per cent) gained employment and 498 (16.1 per cent) retained employment for at least 13 weeks. Programme performance is comparable with mainstream JSA performance (Queensland Corrective Services, 2010).
The Prisoner Employment Programme is funded by the Northern Territory Correctional Services. The programme has three components: the Community Service Work Parties (CSWP), which entails completion of supervised work projects as part of restorative justice and transition; the Volunteer Employment Program (VEP) which entails individual, unpaid volunteer work within the community; and the Prisoner Employment Program (PEP) which entails engagement in paid employment. The model allows for progression through the components. In its initial offering, the programme showed considerable success in helping prisoners' transition from prison to the community, but lack of funding curtailed the programme. In the first 12 months since the PEP recommenced, sixteen prisoners participated in full-time paid employment, seven participated in paid training programmes (with two achieving full-time paid employment on completion of their sentence), and an average of four to six prisoners were on paid employment each month. Nearly two-thirds (65 per cent) of PEP participants were Australian Aboriginal and Torres Strait Islanders. Although the number is small, the prison population of the Northern Territory is also small (fewer than 2,000 in June 2012), and the programme is new (Graffam and Shinkfield, 2012a).
A different Prisoner Employment Program, which is offered by the Department of Corrective Services, Western Australia, is a five-stage pre-release-to-post-release programme that includes application, assessment, case management, placement, and post-placement support. Participants must have completed more than half of their sentence in order to be eligible for the programme. The programme operates from nine locations in Western Australia, including metropolitan and regional prisons. Paid employment is at award levels, and all standard clearances (police check, workers' compensation cover, etc.) are required. The programme is reported as successful in placing ex-prisoners in employment. Unlike the other state and territory programmes, no independent evaluation has been conducted and no figures are available.

Victoria: A case study of employment assistance
In Victoria, there have been various forms of a Corrections Victoria-funded employmentassistance programme for offenders and ex-prisoners operating since 2002. The programme has provided support from several locations to prisoners who are released from selected prisons and community corrections offices throughout the state. Prisoners begin the programme within six months preceding release. Offenders commence while serving CBOs.
The programme has provided a range of individualized employment-related supports including training-needs analysis, vocational training, job search skills, job placement, post-employment support, and referrals. Over time, there have been several thousand participants, and between 34 per cent and 40 per cent of participants have been placed into employment. Over time, between 16 per cent and 19 per cent have retained employment for at least 13 weeks. This performance is comparable with mainstream JSA performance. More than half of participants have been exprisoners, although prisoners comprise 35 per cent of the total corrections population.
An independent evaluation of the Victorian programme was conducted that covered the period 2002−04 and focused on employment outcomes and recidivism of participants (Graffam et al., 2005). Although historical now, the detailed results provide insights into elements of 'best practice' in transition and post-release support. The programme has a dual purpose of placing participants into sustained employment and reducing re-offending. During the period of investigation, the programme operated out of seventeen corrections locations, seven prisons and ten community-corrections locations, targeting participants at moderate to high risk of reoffending. Prisoners and offenders registered on a voluntary basis and were eligible to receive assistance for 12 months, commencing pre-release for prisoner participants. The programme adopted a long-term focus on development of employment-related skills, while also focusing on broader life skills. Employment assistance provided included activities related to work preparation and placement into employment, such as constructing resumés, being interviewed, and making job applications. It also included more general life skills and reintegration supports including referrals to housing, drug and alcohol treatment, and other relevant services. Work preparation involved skill building tailored to the individual to prepare them for entry into the workforce. Job-placement assistance included post-employment placement support and onthe-job training. The transition element of the programme ensured continuity of contact for prisoners (pre-release and post-release) and offenders (pre-and post-completion of community orders). In short, assistance provided was intensive in that it was long-term and multifaceted.

Participant files
Primary analysis of employment outcomes and recidivism among participants of the employmentassistance programme was based on the total programme-participant population. There were 2,458 registered volunteer participants in the investigation period (83 per cent male, 17 per cent female), of which 55 per cent were prisoner participants and 45 per cent were offender participants (those serving CBOs). Analyses of recidivism were derived from files of 600 employment-programme prisoner participants (M = 30.23 years; SD = 8.52).

Procedure
Ethics approval and permission to access data records were provided by the Department of Justice. Programme records were obtained from the Prisoner Information Management System (PIMS) through the programme providers and the Department of Justice. Data files were compiled from electronic reports. Recidivism data were provided electronically from PIMS records. The PIMS database only records offences resulting in a custodial sentence. Employment outcomes and recidivism outcomes were examined over a two-year post-release period for programme participants.

Data analysis
Beyond looking at the number of participants assisted and outcomes achieved, the effectiveness of the programme was examined in terms of its success in participants progressing through the 'steps' of the programme. The measure of this effectiveness is in terms of the proportion of participants moving from referral to registration, registration to placement, placement to a 13-week outcome, and registration to a 13-week outcome. Recidivism results are reported first for the whole programme, with specific reference to participants who obtained employment, participant gender, and participant status. A sample of 600 prisoner-programme participants was randomly selected from the state's PIMS and offences were analysed. These results are reported in terms of three measures of recidivism (offences per day, rated seriousness of offences, and extent of poly-recidivism). Table 1 presents results relating to the effectiveness of the whole of the programme in terms of programme progression. Results are presented in relation to gender, participant status (prisoner or offender participant), and total programme outcomes. The progression of participants in terms of conversion rates over the two-year initial pilot period is similar to intensive JSA provider performance. The programme converted almost twothirds of referrals to registrations, indicating relatively good performance in engaging participants and 'selling' the programme to them. Slightly more than one-third of registrations resulted in the participant obtaining employment (being placed). Although the percentage might seem low, it reflects the comprehensiveness of support needs, the low 'starting point' of many participants, and the labour intensiveness of the work. Of those who obtained employment, almost half achieved 13 weeks of continuous employment. Clearly, keeping participants employed proved almost as difficult as assisting them obtain employment. Taking all of this into account, the proportion of participants who registered and later achieved 13-week outcomes was approximately one-sixth. Results in the table suggest that patterns were very different for males and females and very different for prisoner and offender participants. The percentages shown are the rates of conversion or progression from one phase to the next in the programme. Virtually equal percentages of male and female participants progressed from referral to registration, but a relatively lower percentage of females obtained employment. However, of those who were employed, a relatively higher percentage of females achieved 13-week outcomes compared to males. Overall, females had a lower conversion of registrations to outcomes as well; approximately one-eighth for females and one-sixth for males.

Employment among programme participants
There were different patterns for prisoner participants and offender participants in terms of their progression in the programme measured by conversion rates. Prisoner participants had twice the registration rate of offender participants, with almost all prisoner referrals converting to registrations. Conversion of registrations to placements, however, was much lower for prisoners, with less than one-sixth of registered prisoner participants obtaining employment, compared to more than half of offender participants. Prisoner participants also had a lower conversion rate of placements to outcomes (41.1 per cent) than offender participants (50.9 per cent) and much lower conversion of registrations to outcomes (6.7 per cent) compared to offender participants (28.1 per cent).

Re-offending among programme participants
The most basic measurement of recidivism is the rate of re-offending among the programme participant population as a whole. Re-offending was investigated in relation to registered participants, distinguishing between participants placed into employment and those not placed. Table 2 presents those results for the whole of the employment-assistance programme. Results are presented in relation to gender, registration as a prisoner or offender (community corrections participant), and total programme outcomes. It is important to note that the timeframe for programme involvement (12 months) is shorter than the two-year timeframe used in many studies of re-offending. However, it is also clear from the research literature, and must be recognized, that a high proportion of re-offending occurs within three to six months of a prison release or completion of a CBO. Overall, the rate of re-offending by programme participants was low (7.46 per cent). This is well below re-offending rates reported in the literature. It is low for both participants who obtained employment (6.4 per cent) and those who did not (7.73 per cent), suggesting a positive programme effect in addition to any employment outcomes that were achieved. The re-offending rate was lower for participants who obtained employment, an expected result. It shows a relationship between employment and reduced recidivism. There were differences between male and female participants in terms of rates of re-offending. Females had an overall lower reoffending rate than males, regardless of whether they had obtained employment. However, for both males and females, employment had a positive effect on the re-offending rate. The effect of employment on recidivism was much greater for females than for males. Of course, the relationship between employment and re-offending is complex, and some caution around causal inference should be demonstrated. Moreover, selection bias may be a problem given that the programme was voluntary. Those who volunteered may simply have been more motivated to change and the re-offending results may reflect this inherent motivation. It is possible, therefore, that greater motivation to change on the part of the volunteers may have contributed to better re-offending outcomes regardless of whether they had participated in the programme or not. Although the programme was voluntary, it was not waitlisted, so selection bias was not introduced in this regard. Regardless of the potential limitations, programme employment and recidivism outcomes are both nonetheless impressive.
Prisoners had slightly more than half the re-offending rate of offenders overall, irrespective of whether the prisoners were placed in employment or not. Male prisoners in particular had a low re-offending rate compared to male offender participants. This difference may have been confounded somewhat by location differences in service provision. It is important to note that prisoners, upon release, may relocate to any one of the state's community corrections locations. In any case, for both prisoners and offenders, employment had a positive effect on re-offending.
The effect of employment on recidivism was much greater for offenders than for prisoners. Although offender participants had higher recidivism, the difference between employed and unemployed offenders was greater than that difference for prisoner participants. This result indicates that the programme has worked very well for prisoner participants, and that there is an overall 'programme effect' in addition to the positive effect of employment on re-offending. It has also obviously worked well for offenders, in that their rates of re-offending were also very low, well below non-programme statistics within the corrections system and reported in the literature.
Investigation of the three recidivism measures: offences per day; rated seriousness of offences; and extent of poly-recidivism are presented in Table 3. As indicated, there was a definite reduction in number of offences for the sample of prisoner participants following their release from prison while in the programme. The difference in number of offences per day was statistically significant, F(1,586) = 61.1, p<0.001, indicating that the difference could not be a 'chance' result. The decline in number of offences per day equals a decline of 82 per cent in offending (0.002/0.011 = 0.18). There was also a reduction in seriousness of re-offending. That difference too was statistically significant, F(1,586) = 594, p<0.001, indicating that the difference could not be a 'chance' result.
The decline in seriousness of offences equals a decline of 72 per cent in offending (1.21/4.35 = 0.28). For poly-recidivism as well, the number of different offences committed, there was a reduction in offending, and that difference was also statistically significant, F(1,586) = 1,156, p = 0.01. The decline in number of offences per day equals a decline of 80 per cent in the range of offences (0.77/3.87 = 0.2). In summary, the evaluation of Corrections Victoria's employment-assistance programme identified some very encouraging results. Employment outcomes were equivalent to those of Commonwealth-funded JSAs for the mainstream 'intensive assistance' jobseekers. Recidivism of programme participants was very low while engaged with the programme for both prisoners and offenders, whether they obtained employment or not, indicating a positive general programme effect. Comparison of pre-programme and post-release recidivism showed reduced recidivism after release with respect to all three recidivism measures.

In pursuit of a new paradigm
There is a growing recognition that existing and prevailing approaches to offender treatment are not very effective for a large proportion of offenders, at least, in terms of having a 'corrective' effect. With respect to recidivism, recent reports from the RAND Corporation (see Davis et al., 2013) indicate that correctional education can reduce recidivism. Likewise, other studies (e.g. Chappell, 2000;Steurer and Smith, 2003) show that education in prison has a moderate but significant effect on recidivism. Nevertheless, there is a recognized need for a new paradigm and treatment models to support that paradigm. Several lines of current thinking are contributing to a coalescing of ideas into a paradigm of reintegration. The reintegration construct relates to fitting the interdependent pieces of a fragmented life back together as well as finding a fit for that person within family and community.
Traditional models of offender rehabilitation have focused on correcting character defects and skill deficits of individuals. More recently, competency-based, social-capital-oriented models have been introduced. For example, the Good Lives Model (e.g. Ward, 2002;Ward and Brown, 2004;Ward and Fortune, 2013) provides a strength-based approach to rehabilitation. Characterand skill-building are important elements of a person's reintegration but not sufficient in their own right, and not at the cost of ignoring both assets and other needs. Extending this further, the concept of reintegration has been described as a consideration of the whole person within their 'context' or living environment (Graffam and Shinkfield, 2006).
An ecological system is a 'living system'. It is a system that sustains life and it also has a life of its own. Ecological systems are dynamic, changing systems comprised of a multitude of interrelated, inter-dependent, and interactive elements. Communities can be understood as large ecological systems, macrocosms; we manage, manipulate, act, react, and adapt to the conditions around us. We are interactive agents within those systems. An individual's living environment is a smaller ecological system, a microcosm comprised of: • intra-personal elements such as personality, knowledge, skill sets, experience, behaviours, health conditions, credits and qualifications, even aspirations • inter-personal elements such as social relations and networks, family relations, and formal relations such as those at work and in professional contexts, relations with service-providers, even mandated relations such as one might have within the criminal justice system • material elements such as food, housing, income, mode of transport, and formal support services of various kinds • wider community elements that impact on an individual's lifestyle such as local community culture, attitudes, socio-economic conditions, demographics and infrastructure, and other elements that contribute to conditions in the community • wider societal elements that influence conditions within the system as a whole such as laws and public policies, governmental structures and processes, and broad cultural, demographic, and economic conditions. This is illustrative of the complex set of interactive (and changing) conditions to which an individual must respond on release from prison or completion of a CBO and is committed to lifestyle change. The notion of reintegration, in terms of putting 'life pieces' together and healthy functioning within the community, is often rather a misnomer because many offenders have previously never really been integrated within their broader communities. Many experience multiple conditions of disadvantage across all five of the domains listed above. Many have lived marginalized or excluded from mainstream experiences. Given the complexity, multiplicity, and pervasiveness of disadvantage experienced by many offenders, comprehensive local networks of formal and informal support are needed. Support models of individualized case management that address the full extent of a person's support needs are critical. Also, progress must be understood and managed in terms of small steps and 'micro-gains' (Graffam et al., 2005). In short, our view is that reintegration is best understood as ecological system engagement, in terms of the compatibility of a person's resources and needs, and conditions within their living environment. A healthy 'fit' between a person's characteristics and resources and other factors such as their support needs being met and conditions in the physical environment being safe promote a positive lifestyle change. As conditions change, compatibility can improve or worsen over time. Support services must be responsive over time. Reintegration is an individual-specific process, not an event. It is important to monitor and support the whole process from offending through successful lifestyle change. It is possible to characterize the ecology of reintegration in many ways. The model depicted in Figure 1 is a very simplified one. It may be more useful to portray all five of the domains listed and described above. For purposes of considering 'everyday life', a simple model may be sufficient. This three-part model of reintegration allows one to concentrate on ensuring that conditions within all three domains are 'healthy' and needs are being met (Graffam and Shinkfield, 2012b).

An integrated system of support
It is essential that, with respect to reintegration within the community, support be individualized because of the diverse and dynamic nature of conditions, situations, and contexts. It is important to establish a focus on positive lifestyle change as the core objective of the process. Employment will/may come; stable housing will/may come; freedom from addiction may come; desistance from crime may come. But these should not be viewed as primary or core objectives; these result from positive lifestyle change. Transition-support programmes are really about helping to generate or facilitate transformative change. Achieving positive changes in specific life conditions is the means to achieving positive lifestyle change on a larger scale. In our view, this is the most productive way of understanding and describing reintegration and transition support programmes for prisoners and offenders.
An integrated local support system of networked providers, if organized and managed well, can provide the means for achieving sustainable, positive lifestyle change for prisoners. Such a system would be structured through a 'central' transition-support programme that serves to coordinate and case-manage relevant support services for individual clients. Figure 2 below depicts what such a system might look like. Such an integrated system of support has several advantages. It provides networked referrals, enhanced communication among relevant parties, and the potential for shared resources. Such a system can be individualized in the sense that elements can be added or deleted as appropriate to any individual. For example, the '?' that is included in Figure 2 is indicative of the 'open' nature of the system. Any relevant support services (one or more) can simply be added, with the assumption that the provider becomes part of the integrated system for the individual. Such a system can minimize duplication, reduce 'referral chasing', improve the performance of each provider, produce or approximate a 'seamless' network of services, and better protect programme participants from attrition and recidivism. The cost of establishing and maintaining such a system is no more than the cost of good business practices and can be absorbed by various resource savings gained by providers within the system.

Conclusion
The most important point we make is that transformative change is at the heart of the reintegration paradigm. It is really all about personal and lifestyle change. Acknowledging the depth of change required to transform a life, and that self-sufficiency is an ambiguous point well along the developmental path, is essential to building a support system that will serve ex-prisoners and offenders. Ex-prisoners generally require a slow rate of change to maintain more-or-less stable progress, and even then progress is likely to be difficult (Graffam et al., 2005). It is unreasonable to expect six-week rehabilitation programmes, or so-called 'intensive assistance', to be either intense or long enough to produce lasting results. Sustained change is rarely achieved via 'quick fix' approaches.
When we provide structures and pathways that assist people in this transformative change, we will begin to address the chronic problems of offending and recidivism. As long as we ignore the fact that there is a person at the heart of the complexity of offending and re-offending, we will struggle with our criminal justice systems. With respect to the specific focus of this article, education and employment are clearly two very important pieces of a reintegration jigsaw puzzle.

Note
1. This research has been funded, in part, by the Australian Institute of Health and Welfare and the Victorian Department of Justice.

Notes on contributors
Professor Joe Graffam is Pro Vice-Chancellor (Research Development and Training) at Deakin University, Victoria, Australia. He has been conducting research into community adaptation of people with disability and mental ill-health for more than 30 years and in prisoner reintegration for 15 years.
Dr Alison Shinkfield is currently a Visiting Research Fellow in Psychology at Deakin University. She has been conducting research into emotional state and community reintegration of ex-prisoners for 15 years. Prior to that she conducted motor behaviour research.
Dr Barbara Lavelle is currently a Research Fellow in the Faculty of Health at Deakin University. She has been conducting research into community adaptation of people with disability and mental ill-health as well as prisoner reintegration for more than 20 years.