Living in a Parallel Universe: Using a pragmatic realist evaluation framework to evaluate the socio-spatial impacts of public housing renewal programmes

Public housing plays a central role in determining health and education outcomes as well as family stability and social wellbeing for low income households. However, since the 1980’s the stock of public housing in Australia has deteriorated and eligibility policies have shifted to prioritising vulnerable households. The effects of these policies are compounded in public housing estates, which increase the socio-spatial concentration of disadvantage.

In response, housing renewal programmes have traditionally focused inwards on estates to resolve the issues caused by spatial disadvantage. These solutions typically involve redevelopment processes that improve the physical attributes of estates to dilute the concentration by attracting new homeowners. The physical improvements are assumed to create positive social outcomes for tenants. but these assumptions have not been tested and it is not clear how or if the expected outcomes are achieved or who benefits from the programmes.

A pragmatic realist evaluation methodology is proposed and tested in this thesis using two case studies in Sydney. The methodology is a hybrid of the realist evaluation (Pawson and Tilley, 1997) and the theory of change (Weiss, 1998) evaluation frameworks. This new approach is considered beneficial because it combines a theoretical framework with a practical examination of how the programmes are being implemented in practice. It endorses multiple research methods to help unravel the assumptions of place-based interventions. Rather than focusing on outputs the methodology examines the causal relationships between the context, the way the interventions were implemented and outcomes that were achieved. The empirical findings provide insight into the limitations of social mix policy as it is currently promoted.

Internationally, the realist evaluation framework is growing in popularity, in academic and government institutions. Its application as an evaluation tool is still limited however particularly in the Australian context, which makes this thesis innovative and a contribution to both knowledge and practice.

Supervisors: Professors Bill Randolph & Simon Pinnegar.