Beginning with the 2003-04 reference year, the Survey of Income and Housing (SIH) became a biennial rather than an annual survey. The ABS was asked to explore whether income micro data could still be made available for non-survey years. In response the ABS investigated methodological possibilities for producing synthetic micro data for non-survey years.
This paper presented to the Methodology Advisory Committee in November 2004 outlines the early stages of this investigation. It describes the static ageing technique and discusses how it could be used to produce forecasts of income micro data for non-survey years. Preliminary results from the exploratory study undertaken up to that point are also briefly discussed.
Results from the subsequent exploratory study indicated that the proposed methodology produced reasonably good estimates at aggregated levels of income. However, it was less suitable for producing acceptable estimates of income at the unit record level as would be required for research purposes. Synthetic income data produced through the proposed methodology is unlikely to capture the impact of policy changes on income distribution as accurately or adequately as would be the case if an actual survey was carried out. Consequently we have decided not to take forward the research into production.