This thesis addresses two questions regarding desistance through restorative justice (RJ) conferencing. First, it examines whether RJ conferencing supports change towards desistance. Although there has been a significant amount of research on what RJ conferencing can offer to victims and offenders, little is known about how and why RJ conferencing may support change towards desistance. This thesis addresses this gap by developing an explanatory model to understand change towards desistance through RJ conferencing. Second, this thesis asks how change towards desistance should be measured and evaluated in research on RJ conferencing. Scholars have generally agreed that desistance should be recognised as a process that involves lapses and relapses of offending until the complete cessation of offending. Despite this consensus, most research has adopted a prevalence measure of post-conference offending (i.e. involvement in reoffending). Such a measure may not be appropriate for capturing the complex processes of change towards desistance through RJ conferencing; however, there is no agreement as to what measure is best suited to examine such change. In my thesis, I employ six different measures of offending patterns towards desistance: (i) the prevalence measure of post-conference offending, (ii) change in the frequency of offending, (iii) change in the number of offence types, (iv) change in the seriousness of offending, (v) the time frame to reoffend and (vi) an integrated measure of offending patterns towards desistance. I investigate these measures and how they vary to better understand what aspects of RJ conferencing may support change towards desistance. Because my research seeks to show ‘how’ and ‘why’ RJ conferencing may support change towards desistance for young offenders, I employ a mixed-methods approach. Drawing on the South Australian Juvenile Justice (SAJJ) dataset, which includes more than 80 young offenders who have participated in RJ conferencing, my thesis contains two studies designed to answer the questions above. In Study A, I examine which aspects of RJ conferencing may support change towards desistance and whether these elements vary across the different measures of offending patterns towards desistance. In Study B, I explore why RJ conferencing may support change towards desistance. Using integrated measures of offending patterns towards desistance, I conduct a thematic analysis of young offenders’ responses to interview questions regarding their experiences with and attitudes towards RJ conferencing. By identifying common themes among offenders with similar desistance trajectories, and distinct themes among offenders with different desistance trajectories, I examine why desistance trajectories may differ between young offenders who have participated in RJ conferencing. The findings of Study A offer two important insights. They first show that regardless of the use of different measures of offending patterns towards desistance, residential instability is associated with an increase in post-conference offending at both follow-up periods. They also demonstrate that regardless of the use of different measures of offending patterns towards desistance, young offenders who reached a genuine consensus with victims on an agreement plan show the strongest movement towards desistance at both follow-up periods.
In Study B, I establish the concept of the ‘offender journey towards desistance’. This concept encapsulates how offenders experience RJ conferencing, and how their experiences are linked to desistance trajectories after RJ conferencing. Specifically, I identify three types of offender journeys: (1) the ‘optimal journey’, whereby young offenders completely desist from crime after RJ conferencing; (2) the ‘changing journey’, whereby young offenders experience lapses and relapses of offending after RJ conferencing; and (3) the ‘difficult journey’, whereby young offenders escalate their offending behaviour after RJ conferencing.
This thesis makes theoretical, practical and policy contributions to the understanding of the relationship between RJ conferencing and change towards desistance. My dissertation reveals the need to develop different theoretical accounts for different desistance trajectories through RJ conferencing, as no single theory or model adequately explains the complexity of the desistance process following RJ conferencing. In terms of RJ practice, my thesis highlights the importance of helping young offenders rebuild their life after RJ conferencing, because they struggle to ‘go straight’ after RJ conferencing. As for the policy implication, my findings suggest that to understand the complexity of desistance trajectories, an individual-based evaluation on what offending trajectories individual offenders experience (i.e. a qualitative approach) may be more appropriate than a group-based evaluation of whether offenders desist or not (i.e. a quantitative approach).