Regional Dimensions: Creating Synthetic Small-area Microdata and Spatial Microsimulation Models

This Technical Paper describes much of the work undertaken under the three-year ‘Regional Dimensions’ ARC Linkage project. The core aim of this project was to determine whether we could create synthetic small-area household microdata that were sufficiently reliable to be subsequently used by policy makers in such applications as needs-based planning, the analysis of disadvantage, and modelling the spatial impacts of policy change. The small-area household microdata were created by reweighting a national ABS sample survey to Census benchmarks for each small area. The validation undertaken suggested that the synthetic small-area data did provide reasonable estimates of the characteristics of households. The household microdata were successfully linked to NATSEM’s well-established STINMOD taxtransfer static microsimulation model, so we can now simulate the spatial impact of tax-transfer policy changes.