FEATURE ARTICLE: 30 YEARS OF ERP
On 31 March 2012 it will be 30 years since the ABS began releasing official population estimates on a 'usual residence' basis in its Estimated Resident Population (ERP) statistical series. The move to ERP was prompted by changes in international traveller behaviour, and also an aspiration of the ABS to have a consistent series of population estimates on a conceptual basis which aligned with United Nations and Organisation for Economic Co-operation and Development (OECD) standards. Although it is 30 years since ABS introduced ERP, the series was backcast on a consistent basis back to 1971, resulting in a time series longer than 40 years.
This article draws from previously published ABS material to give an overview of how the earlier measures of Australia's population evolved into ERP which is used today as Australia's official population measure.
Before 1966 - a consistent series - population by actual location
Prior to 1966, official population estimates for Australia and the states and territories were simply census counts plus estimates of natural increase and net migration. At the national level, net migration included both short term, long-term, and permanent movements. At the state and territory level, interstate migration was based on records of all movements by ship, plane, train and bus. Over this early period, up to 1966, the ABS was attempting to compile estimates on a consistent 'actual location' basis.
From 1966 to 1981 - a mix of actual location and usual residence
In 1966, the ABS recognised that the measure of interstate migration based on ship, plane, train and bus movements was inadequate as it did not cover all movements. The ABS changed its method of estimating interstate migration from one based on actual location to one based on usual residence. The key data source used to calculate interstate migration changed to interstate transfers of family allowance payments (Child endowment) and electoral roll registrations. This change introduced a conceptual inconsistency into the series, as interstate migration was based on usual residence but other components remained in an actual location basis.
In the early 1970's there was a large increase in the numbers of short-term overseas movements and also an increase in the volatility of that series. The volatility in overseas movements caused fluctuations in quarterly population estimates, for example, large gains in March quarters followed by large losses in June quarters.
From September 1976, it was decided that short term movements (duration less than one year) would be ignored in calculating net overseas migration. This solved the problem of quarterly volatility, but highlighted another problem which was that the Census base was still on an actual location basis and technically was inconsistent with the incoming net overseas migration data. This resulted in the 1976 to 1981 series of population estimates being a mix of actual location for the Census base and usual residence for the components of population change.
From 1981 onwards - a consistent series on a usual residence basis
The key changes which completed the move to a consistent ERP series were 1) adjusting the 1981 Census counts on to a place of usual residence basis; and 2) making an adjustment for Australian residents who were temporarily (less than one year) overseas as at Census night. This gave a consistent series forward of 1981, but did not fix the inconsistency caused by the mix of actual location and usual residence for the reference period 1971-1981.
To solve the issue that the series was still conceptually inconsistent for reference years 1971 to 1981, it was decided to backcast ERP to 1971. Because of inadequacies in the 1971 Census Post-Enumeration Survey, the 1971 ERP was derived by working backwards from 1976.
From 1986 onwards Medicare data became the key data source used to calculate interstate migration.
For more information, see the references below:
Timeline of changes to the ABS population estimates series
- Actual location basis.
- Short term overseas arrivals and departures were included in the population.
- Measured interstate movements by ship, plane, train and bus.
- A consistent concept.
- Change to interstate migration concept and measurement from actual location to usual residence based primarily on interstate transfers of family allowance payments and electoral roll registrations. This causes an inconsistency in ERP, which used a mix of actual location and usual residence components.
- Quarterly ERP is rolled forward on a usual residence basis (Census still based on actual location).
- Short term overseas arrivals and departures were not included in the population
- Census Post Enumeration Survey results used to adjust population estimates for under enumeration in Census.
- A hybrid method with some inconsistencies in concept.
- ERP on a consistent usual residence basis.
- ERP revised back to 1971 on a consistent basis.
Demography Working Paper 1979/1 - Population Estimates in Australia: A Discussion Paper, 1979
- Interstate Migration based on Health Insurance Commission (Medicare) data.
(cat. no. 3108.0) - released July 1979
Population Estimates: An Outline of the New Conceptual Basis of ABS Population Estimates
(cat. no. 3216.0) - released 29 March 1982
Estimated Resident Population, Australia, States and Territories, 30 June 1981, 1976 and 1971 (preliminary)
(cat. no. 3217.0) - released 31 March 1982
Technical Paper: Methods and Procedures in the Compilation of Estimated Resident Population 1981 and in the Construction of the 1971-81 Time Series
(cat. no. 3103.0) - released January 1983
Demography Working Paper 1996/1 - Evaluation of Administrative Data Sources for Use in Quarterly Estimation of Interstate Migration Between 1996 and 2001, 1996
(cat. no. 3109.0) - released 21 September 1999