From Policy To Practice: A Study of the Queensland Youth Justice Service: Policy, Implementation and Outcomes for Young Offenders

Abstract

This thesis employs a broad evaluative framework to examine the impact of the Youth Justice Service (YJS) on the post-intervention offending behaviour of young people on community-based court orders. The YJS is a Queensland government policy initiative that aims to monitor compliance with community-based court orders, and identify and address causes of criminal behaviour. The evaluative framework views policy, implementation and impact as distinct but related dimensions of intervention. Reflecting this framework, three primary research questions are addressed: (1) Does the YJS concept represent a goal-directed, theoretically-informed, executable and assessable juvenile crime prevention policy?, (2) Is the YJS concept realised through service delivery?, and (3) What is the effect of the YJS on future offending behaviour? Three studies, employing qualitative and quantitative methods, examined these questions. Study one examined the YJS concept, drawing on some key themes from literature on policy development and implementation, developmental and life-course criminology and developmental crime prevention. This study synthesised key policy and procedure documents around six themes, including (1) rationale, (2) goals, (3) theory, (4) service delivery model, (5) method of operation, and (6) key performance indicators. Findings indicated that the YJS concept represents only marginal adjustments from the traditional Area Office (AO) model of service delivery, and integrates few new preventative mechanisms that would foreseeably lead to change at the operational level. Moreover, it suffers from goal ambiguity, fails to incorporate some key components of best-practice crime prevention that have proven successful when working with at-risk young people, lacks sufficient process-level specificity to ensure treatment fidelity, and places heightened importance on measuring impacts that have political value rather than benefits for the clients. In the second study, an in-depth case study of the Logan Area Youth Justice Service (LAYJS) was conducted to explore how the YJS operated in reality, and as compared with the policy directive. Information was drawn from a variety of sources including interviews with staff and clients, policy and procedure documents, direct observation, case management files and staff-researcher interaction. Evidence suggested that the LAYJS was focused primarily on ensuring compliance with court orders. Several organisational factors, such as staff workloads, the statutory basis for monitoring compliance, and the capacities of staff, have meant that comparatively little attention has been directed at addressing offending behaviour. For the most part, the LAYJS employs an individualised case management process, as distinct from the collaborative, team-based model that is prescribed in the YJS concept. Caseworkers have little faith in their ability to bring about positive behavioural change in their clients, and subsequently transferred the responsibility for intervention outcomes to the client. While acknowledging the importance of families in preventing offending, caseworkers emphasised that a number of organisational tensions have prevented them from engaging families in the case management process. The final study examined the impact of the YJS on post-intervention offending, controlling for developmental risk factors and key features of the intervention process. A random sample (N=190) of clients from three YJS offices and three AOs was drawn from the population of clients who had active community-based court orders between June 1999 and December 2002. Information from Department of Communities' case management files and rearrest data from the Queensland Police Service were entered into a purpose-designed database, and analysed using bivariate and multivariate methods including logistic regression and survival analysis. High proportions of missing data on non-statutory variables suggested poor record management practices, or alternatively that operational staff do not understand the role of developmental risk and/or protective factors and social contexts in preventing offending behaviour. Results indicated that the YJS was no better than the AO at preventing recidivism, as measured at 18-months post-intervention, even after controlling for risk factors that were significantly related to recidivism. The analyses found that some unmeasured variation in service delivery, even within service types, did impact upon recidivism, supporting the hypotheses of the first study and the contention that variation in intervention practice can influence offending behaviour. The likelihood of recidivism was increased if the client was using drugs or was influenced by delinquent peers, and decreased if he stayed in school until years 11 or 12, or where caseworkers addressed familial problems. This provides some sense of programs that may be appropriate for young offenders in the context of a community-based program. It also highlights the critical importance of incorporating families into case management, not only for the purpose of providing information, but also as viable targets of intervention. Survival analyses indicated that the YJS might have had some temporary deterrent effect, although this effect had dissipated by 18-months post-intervention. This result may reflect the increased focus on ensuring compliance with court orders as found in the LAYJS case study. However, given the hypothesis that the lack of process direction will result in variable practices across offices, it cannot be assumed that all YJSs place equal importance on compliance. Overall, findings suggest that the promise that the YJS would provide an innovative model of service delivery and generate improved outcomes for young offenders has not been realised. This research has added further weight to the perspective that examines both the individual and combined impact of theory, policy and implementation for measuring client outcomes. Deficits in any of these components ultimately have a ripple effect, making it difficult to achieve the predetermined goals of the policy at the operational level.