ABOUT THIS RELEASE
In Australia, as in many countries, there has been a rapidly growing demand from policy makers for social and economic small area data to satisfy expanding decision-making requirements. The increasing user demand for small area estimates, together with practical difficulties in increasing survey sample sizes, has motivated the need to identify ways of finding reliable and defensible methods for producing quality small area estimates. A project has commenced at the ABS to produce a series of manuals on the theory, application and processes for producing small area statistics in the Australian context. The main purpose of this paper is to discuss a range of methodological approaches undertaken in the application of existing small area methods to the topic of disability. A set of diagnostic measures are used to assess the relative quality of the small area estimates produced by a selection of models. The results presented here are still preliminary and for this reason we also discuss the priorities for future work to improve on these models.