Performance auditing in the public sector: crowded, contested and at a crossroads

This thesis examines the phenomenon of performance auditing (PA) as an instrument of accountability and regulation in the public sector. Using evidence from Australia, the thesis engages with the international literature on PA and makes findings that are relevant to how PA is practised around the world, and to the place of PA as one of many different instruments of public accountability and administration. After reviewing previous definitions of PA and finding them to have significant shortcomings, the thesis proposes a new definition based on the five elements of independence, authorisation, discovery, synthesis and publication. The definition avoids defining PA in relation to problematic elements such as ‘objectivity’ and ‘the audit method’, and enables new insights. Among the key insights explored in the thesis, and made possible by the new definition, are how PA resembles other public sector accountability instruments such as whistleblower laws, freedom of information laws and open-book policies, and how it also shares traits with non-government phenomena such as investigative journalism. The thesis uses the new definition to delineate the ‘PA space’ – the conceptual space in which PA is conducted – and then explores the phenomenon of ‘convergence’ whereby public sector oversight institutions from different heritages have come to cohabit the PA space. How different PA institutions interact in the PA space, and how PA can clash with other accountability technologies such as gateway reviews and probity audits, is also explored. Another important theme of the thesis is the impact of PA on public sector innovation and other aspects of public sector performance. After analysing the theoretical relationship between PA and innovation, the thesis argues that the conditions under which PA can improve public sector innovation may arise only rarely in practice. It is also argued that PA’s impact on public sector performance improvement more generally is contentious and not well supported by robust empirical evidence. Nor is the value and impact of PA well demonstrated in the performance reports of some audit offices. The thesis identifies a range of potential costs and distortions that PA may give rise to, and which should influence how PA institutions are monitored. The thesis concludes that the PA space is crowded and contested and that, for a variety of reasons concerning technology, institutions and governance, the law, public administration, society and the media, traditional PA is now at a crossroads. The thesis presents findings that have implications for the design and oversight of public audit institutions and for the design of public accountability frameworks more broadly.