An evaluation crucible: Evaluating policy advice in Australian central agencies

Policy advice is a core function of government that until quite recently remained outside the formal processes of performance evaluation. Evaluation, by its very nature, is designed to question both the effectiveness and relevance of government activities; applying it to policy advice opens up a traditionally confidential and politically sensitive arena, This paper reports on an evaluation experiment in Australian government-policy management reviews (PMRs) - that sought to evaluate the quality of central agency policy advice. It traces the development of the PMR model around interdepartmental committee processes, the bureaucratic politics that diluted the focus on policy outcomes, and examines how central agencies steered evaluation away from questions of public accountability towards arrangements for achieving more effective control of the processes underpinning production of advice. By targeting the process rather than outcomes of policy advising, PMRs sought unsuccessfully to adhere to the divide between management and policy and, in doing so, marked out the limits to performance evaluation.