AN ALTERNATIVE APPROACH TO IDENTIFYING COUNCILS ‘AT RISK’

Abstract

Murray and Dollery (2005) (MD) applied a statistical model utilising a series of key performance indicators (KPIs) published by the NSW Department of Local Government (DLG) to assess the DLG's procedures for identifying local councils that were ‘at risk’. While it appears that the authors were attacking a straw man (since the DLG's assessments were not based on these KPIs), and their underlying model has some serious flaws, it is contended that a statistical model may be of assistance to those concerned with monitoring the circumstances of local governments. An alternative approach to modelling is described, in which distress is interpreted as an inability to maintain service delivery at pre‐existing levels. It is argued that such a concept is more relevant to the public sector than conventional ‘financial distress’ prediction models. The estimates provided by NSW local councils of the cost of upgrading infrastructure to a satisfactory condition were used as a proxy for levels of distress. Independent variables were selected from a range of KPIs published by the DLG. While MD found that most of the DLG performance indicators were not statistically significant in their model (and therefore not useful discriminators of council distress), this study reaches different conclusions using a different selection of indicators and a distress variable construct that is more appropriately linked to service delivery.