Local area population forecasts are used for a wide variety of planning, policy formulation, and service delivery activities. But not enough is known about the errors of such forecasts, particularly over the longer-term (20 years or more). Understanding past errors is valuable for both forecast producers and users. This paper evaluates the forecast accuracy of past sub-state population forecasts published by Australian State and Territory Governments over the last 30 years. It seeks to establish the extent and patterns of forecast error, whether there have been problems with data and methods which could be rectified in future, and whether error patterns are sufficiently reliable to inform prediction intervals for current projections. Population forecasts from the past 30 years were sourced from State and Territory Government websites and contacts. Estimated Resident Populations to which the projections were compared were created for the geographical regions of the past projections in ArcGIS. The results mostly confirm earlier findings with regards to the relationship between error and length of projection horizon and population size. Regression modelling of error on the basis of area characteristics was only modestly successful. Nonetheless, the results provide users with information on the approximate magnitude of forecast error that might be expected from different types of local area. Ways of applying this information to current population projections are suggested.
This study was supported by the Australian Research Council (Discovery Project DP150103343). The authors would like to extend their thanks to the State and Territory government demographers who generously supplied past projections and advice for this study.
Tom is Principal Research Fellow in the Northern Institute at Charles Darwin University. He is an applied demographer working mostly on population and household projections, migration analysis, educational demography, the indirect estimation of demographic data, Indigenous demography, very elderly demographic trends, migration analysis and subnational demographic change. He is co-editor of the book Demography for Planning and Policy: Australian Case Studies published recently by Springer.