Counting family support? : the influence of performance measurement on child protection policy and practice

This research explores how child protection is constructed through performance measurement, and how the selection and reporting of performance indicators influences policy and practice. Performance measurement is now a mainstay of public sector management. At one level, it has a role in addressing ever-present questions about the quality and effectiveness of child protection services. However, there are concerns about assessing performance within a management framework, rather than a professional or client-centred framework. Therefore, the thesis aims to critically engage with the problems and potential of performance measurement in child protection with a view to maximising its capacity to promote better outcomes for children and families. It focuses particularly on the role of family support, because despite its proclaimed importance in legislation and policy, in practice family support work remains marginalised, situated 'outside' core child protection work. The research approaches performance measurement from a critical policy perspective, in which indicators have both technical and policy functions. While performance measurement may be viewed rationally as a means for government to manage resources, from a broader policy perspective the way it is conceptualised and implemented has major consequences for service delivery. The policy significance of performance measurement is explored through two studies. The first study analyses developments in policy and practice in three jurisdictions: England, Victoria (Australia) and Queensland (Australia). This analysis contrasts the prominent position of family support in child protection legislation and policy, with its marginal position in both practice and performance measurement. It outlines the socio-political context for performance measurement in child protection, drawing out the values and power relations embedded in performance indicators that give them a role in communicating policy intent. The study concludes that performance measurement in all three jurisdictions is directing policy attention (and therefore resources) towards investigation and out-of-home care, largely ignoring family support. The second study moves beyond the policy critique to more concrete concerns about how to reconstruct performance measurement in child protection to encompass family support indicators. Using a survey method, it seeks to describe and analyse the practice of family support in one jurisdiction (Queensland) according to a performance measurement model of inputs, outputs, processes and outcomes. It concludes that useful performance data can be collected to aid understanding of the contribution of family support to the overall child protection system. Such data will help to move family support beyond assumptions about its value, towards empirical scrutiny of its methods and results. It suggests that family support will command greater policy attention if included in the child protection performance measurement effort. The thesis concludes that viewed from a critical policy perspective, performance measurement can be harnessed in the interests of good practice. It can assist with the repositioning of family support, moving it from the realm of unproblematic moral goodness to more critical inquiry. It argues there is nothing inherently wrong with measuring effectiveness and efficiency, but it is important to examine the manner in which performance measurement is developed and implemented, its methods and consequences. Performance indicators are not neutral or merely technical. Establishing indicators can influence how social problems such as child abuse and neglect are defined, where resources are placed, what service types are funded, and how outcomes for children and families are defined. It questions the privilege performance indicators may be given within managerial approaches to government, as if they somehow provide an objective and unbiased account of how services are performing. However, this is not to suggest that performance indicators are worthless or meaninglessmrather that their meaning should be explicit, and their underpinning value stances debated within policy processes. Such debate will enhance the congruency between child protection research and indicators, and democratise the processes by which performance measures are developed, used and reported.