Can reliable sub-state emigration estimates be obtained from an administrative population dataset? The case of the Person Level Integrated Data Asset (PLIDA) of Australia

To fully understand and forecast local and regional population change, it is necessary to obtain data for all the demographic components of change—births, deaths, internal migration, and overseas migration. In Australia, and many other countries, high-quality and geographically detailed data is available for most demographic components, but not emigration. The Australian Bureau of Statistics publishes good quality estimates of immigration and emigration at the national and State scales, but available sub-state emigration data is heavily modelled and suffers from quality issues. Good quality sub-state immigration data is available from the census. The aim of this study was to evaluate the quality of sub-state emigration data for Australia extracted from a large administrative population dataset, the Person Level Integrated Data Asset (PLIDA), which links several administrative datasets. We compared State and sub-state emigration flows from PLIDA to available emigration data, as well as sub-state immigration flows from PLIDA to census-derived immigration flows. We found PLIDA-based emigration flows to be considerably lower in volume than official State emigration estimates, likely due to undercoverage in PLIDA of temporary migrants and residents of remote regions. We discovered sizeable geographical, temporal, age, and visa/citizenship variations in emigration coverage relative to available emigration statistics. Our view is that PLIDA does not yet allow sub-state emigration estimates of sufficient quality for most uses. But we are hopeful that improvements to the population coverage of PLIDA will allow this in the future.