Regional internal migration estimates, provisional methodology

This product contains provisional statistics on internal migration between states and territories, and Greater Capital City Statistical Areas (GCCSAs) of Australia, by quarter. 

This is a temporary publication in response to COVID-19 and the heightened interest in internal migration data. The quarterly interstate migration component is released later and in greater detail in National, state and territory population, every three months. Internal migration for GCCSAs by financial year is released in Regional population annually, as regional internal migration estimates (RIME). The internal migration data in these referenced releases are the official estimates used to compile estimated resident population (ERP), the official measure of population in Australia. The quarterly internal migration data released in this product are provisional, and should only be used to provide more timely insights on internal migration patterns during COVID-19, and are not used in the compilation of ERP for GCCSAs. 

The commentary for this product is based on rounded data. In the commentary for this product, figures less than 1,000 are rounded to the nearest 10, figures over 1,000 are rounded to the nearest 100.

This release incorporates small revisions to the June 2020 estimates for GCCSAs.

Internal migration is the movement of people across a specified boundary within Australia involving a change in place of usual residence. It cannot be directly measured and is instead estimated using administrative data. The internal migration data in this product is based on aggregated Medicare change of address (postcode) data, provided by Services Australia. It is supplemented by aggregated Department of Defence change of address data as military personnel generally do not interact with the Medicare system. Internal migration data from the five-yearly Census is used to calibrate the Medicare and defence data. 

When producing estimated resident population, Medicare data received for one quarter is used to estimate migration for the previous quarter. This assumes that on average the time between a person moving and registering a change of address is three months, which has proven to be the best lag time for estimating the population. This quarterly lag assumption is reviewed after each Census. Consistent with this, the Medicare change of address information used to calculate these provisional internal migration estimates is also lagged by three months. For example, December 2020 quarter moves are used to reflect moves in the September 2020 quarter. As with estimated resident population, it is likely that some moves recorded in this publication occurred in the following quarter. While there may be some level of recent behavioural change due to the COVID-19 pandemic in terms of updating an address with Medicare, there is no statistical way to quantify this. 

The Medicare source data is assigned to a state or territory and GCCSA for a person's departure and arrival locations, based on the postcodes of their residential addresses as registered with Medicare. Postcodes are assigned wholly to a state/territory and GCCSA based on best fit. Where a postcode is split across areas, it is assigned to the area that contains the majority of that postcode's population.  

For further information on the method of compiling interstate migration estimates, including preliminary and revised methods, and the geographical coverage for Other Territories, see the Methodology section in National, state and territory population.

While the data sources and method used in this publication are the same as that used to derive official interstate migration in the compilation of ERP at state and territory level, it is different to the method used to derive official RIME in the compilation of sub-state ERP since 2016. As a result, the state and territory data in this provisional release is comparable to the official data that will be later released in National, state and territory population, however, there will be some differences at the GCCSA level that will be later published in Regional population.

The Medicare data used to compile the official RIME series is assigned directly to GCCSAs and other statistical areas, by geocoding to these areas based on reported residential address. The annual RIME series is compiled on an annual basis, determined by change of address one year ago, while the postcode-based data is compiled on a three-monthly cycle. While the geocoded Medicare data is only available for recent years, the postcode-based Medicare data is available starting from the mid-1980s, enabling a longer time series of postcode-based internal migration data to be compiled using this version. For these reasons, there are small differences between the GCCSA-based internal migration in this product and the RIME data published in other ABS products. For further information on the method of compiling RIME, including the method and status of pre-2016 RIME, see the Methodology section in Regional population. 

ABS publications draw extensively on information provided freely by individuals, businesses, governments and other organisations. Their continued cooperation is very much appreciated; without it, the wide range of statistics published by the ABS would not be available. Information received by the ABS is treated in strict confidence as required by the Census and Statistics Act 1905.

The ABS acknowledges The Treasury (The Centre for Population) for funding this new ABS product.