Spatial Microsimulation Using Small Area Estimates of Income, Tax and Social Security Benefits

A spatial microsimulation technique using the ‘reweighting’ approach has been developed at NATSEM to produce synthetic estimates for small areas. This is a novel way to create small-area socio-economic data that are otherwise unavailable from other sources. This paper reports on the application of the technique to produce estimates of income, tax, and social security benefits for the Statistical Local Areas (SLAs) in three States and one Territory in Australia. The spatial technique utilises the 1998-99 Household Expenditure Survey (HES) data and the 2001 Census data, both from the Australian Bureau of Statistics. The process involves the ‘reweighting’ of the household unit record weights at the national level obtained from the HES to produce household unit record weights for each SLA. These SLA household weights are then applied to the selected output variables generated by NATSEM’s STINMOD Model to produce the required small-area synthetic estimates for SLAs. This paper explains how spatial microsimulation is undertaken to ‘regionalise’ STINMOD. It presents the results of the synthetic small-area estimates and measures the reliability of these estimates against Census counts and against data from independent sources. This paper also provides an example to illustrate how the spatial technique may be used to assess the local impact of government policy changes, by simulating the spatial effects of the tax cuts announced in the recent Australian Government Budget.