• Summary
• Downloads
• Explanatory Notes
• Related Information
• Past Releases

QUALITY DECLARATION - SUMMARY

INSTITUTIONAL ENVIRONMENT

Estimated Resident Population uses data sourced from a variety of institutional environments. Much of the data is administrative by-product data collected by other organisations for purposes other than estimating the population. Births and deaths statistics are extracted from registers administered by the various State and Territory Registrars of Births, Deaths and Marriages. Medicare Australia client address data is used to estimate interstate migration. Passenger card data and related information provided by the Department of Immigration and Border Protection (DIBP) is used to calculate Net Overseas Migration (NOM).

ABS Census of Population and Housing and Post Enumeration Survey (PES) data are used to determine a base population from which Estimated Resident Population (ERP) is calculated and to finalise all components of population change. For information on the institutional environment of the Australian Bureau of Statistics (ABS), please see ABS Institutional Environment.


RELEVANCE

Estimates of the resident population (ERP) for the states and territories of Australia are published by sex and age groups, and estimates and projections of the Aboriginal and Torres Strait Islander population are also available. The ERP is the official measure of the population of states and territories of Australia according to a usual residence population concept. ERP is used for a range of key decisions such as resource and funding distribution and apportioning seats in the House of Representatives to each state and territory.



TIMELINESS

Preliminary ERP data is compiled and published quarterly and is generally made available five to six months after the end of each reference quarter. Every year, the 30 June ERP is further disaggregated by sex and single year of age, and is made available five to six months after end of the reference quarter.

Revised estimates are released once more accurate births, deaths and net overseas migration data becomes available. In the case of births and deaths, the revised data is compiled on a date of occurrence basis annually in the March reference period, released in September for the previous financial year.

In the case of net overseas migration, final data is based on actual traveller behaviour. Commencing with data for September quarter 2011, final estimates for NOM will be released quarterly and made available 18 months after the end of the reference period.

Final estimates are made available every 5 years after a census when revisions are made to the previous intercensal period. ERP data is not changed once it has been finalised. Releasing preliminary, revised and final ERP involves a balance between timeliness and accuracy.



ACCURACY

All ERP data sources are subject to non-sampling error. Non-sampling error can arise from inaccuracies in collecting, recording and processing the data. In the case of Census and PES data every effort is made to minimise reporting error by the careful design of questionnaires, intensive training and supervision of interviewers, and efficient data processing procedures. The ABS does not have control over any non-sampling error associated with births, deaths and migration data (see institutional environment).

Another dimension of non-sampling error in ERP is the fact that the measures of components of population growth become more accurate as more time elapses after the reference period. As discussed under Timeliness, the tradeoff between timeliness and accuracy means that a user can access more accurate data by using the revised or final ERP data. While the vast majority of births and deaths are registered promptly, a small proportion of registrations are delayed for months or even years. As a result, preliminary quarterly estimates can be an underestimate of the true number of births and deaths occurring in a reference period. Revised figures for a reference period incorporate births and deaths registrations that were received after the preliminary data collection phase as well as the estimated number of registrations that have still not been received for that reference period. For more information see the Demography Working Paper 1998/2 - Quarterly birth and death estimates, 1998 (cat. no. 3114.0) and Population Estimates: Concepts, Sources and Methods, 2009 (cat. no. 3228.0.55.001).

After each Census the ABS uses the Census population count to update the original series of published quarterly population estimates since the previous Census. For example, 2016 Census results were used to update quarterly population estimates between the 2011 and 2016 Census. The PES is conducted soon after the Census to estimate the number of Australians not included in the Census. Adding this net undercount of people back into the population is a crucial step in arriving at the most accurate ERP possible. For more information on rebasing see the feature article Rebasing of Australia's Population Estimates using the 2016 Census in the December quarter 2016 issue of Australian Demographic Statistics (cat. no. 3101.0). At the time of rebasing to the 2011 Census, estimates for the period September 1991 to June 2006 were revised in response to a methodological improvement in the estimation of Census undercount, for more information see Feature Article: Recasting 20 Years of ERP in Australian Demographic Statistics, December Quarter 2012 (cat. no. 3101.0).

COHERENCE

ERP was introduced in 1981 and backdated to 1971 as Australia's official measure of population based on place of usual residence. ERP is derived from usual residence census counts, to which is added the estimated net census undercount and Australian residents temporarily overseas at the time of the census (overseas visitors in Australia are excluded from this calculation). Before the introduction of ERP, the Australian population was based on unadjusted census counts on actual location basis. It is important to note this break in time series when comparing historical population estimates.

An improved definition for calculating NOM was applied from September quarter 2006 onwards. The key change is the introduction of a '12/16 month rule' for measuring a person's residency in Australia replacing the '12/12 month rule'. This change results in a break in time series and therefore it is not advised that NOM data calculated using the new method is compared to data previous to this. For further information see Information Paper: Improving Net Overseas Migration Estimation, Mar 2010 (cat. no. 3412.0.55.001).

The births and deaths data in this publication are not coherent with the data found in ABS births and deaths publications. This is because the revision cycle necessary to produce ERP results in a mix of preliminary births and deaths data, based on date of registration, and revised data which is a modelled estimate of births and deaths by date of occurrence. By contrast, the main tables of data in the births and deaths publications are based wholly on registration in the reference year, with some tables and analysis based wholly on date of occurrence data.



INTERPRETABILITY

ERP is generally easy to interpret as the official measure of Australia's population (by state and territory) on a place of usual residence basis. However, there are still some common misconceptions. For example, a population estimate uses the term 'estimate' in a different sense than is commonly used. Generally the word estimate is used to describe a guess, or approximation. Demographers mean that they apply the demographic balancing equation by adding births, subtracting deaths and adding the net of overseas and interstate migration to a base population. Each of the components of ERP is subject to error, but ERP itself is not in any way a guess. It is what the population would be if the components are measured well.

Population estimation is also very different to sample survey-based estimation. This is because population estimation is largely based on a full enumeration of components. In the case of the population base, only the PES used sampled data to adjust for census net undercount. In the case of the components of population growth used to carry population estimates forward, Australia has a theoretically complete measure of each component.

Another example of a common misconception relates to the fact that the population projections presented in this publication are not predictions or forecasts. They are an assessment of what would happen to Australia's population if the assumed levels of components of population change - births, deaths and migration - were to hold into the future.



ACCESSIBILITY

ERP data is available in a variety of formats on the ABS website under the 3101.0 product family. The formats available free on the web are: