While Australia has the advantage of taking a census enumeration every five years, businesses, public entities and non-profits often require updated and projected demographic information for their decision-making. Public sector demographers, especially those working for national statistical agencies and regional planning authorities, are often called upon to develop population estimates and projections for larger areas: nations, states/territories, and statistical regions. Applying the best available tools and techniques, along with the most appropriate assumptions about future trends, public sector demographers have consistently generated reasonably accurate estimates and projections that have proved useful to myriad public and private agencies. However, consumer-oriented businesses and some public sector agencies have characteristically preferred the current information for the smallest geographic entities as well as projections pushed out to ever more distant futures. The information produced by Precisely for Australia combines the best work of experts at the Australian Bureau of Statistics (ABS) with over 35 years of global data development experience on the part of Precisely’s demographers, geographers, and statisticians.
Precisely’s estimates and projections are based on the latest official ABS population estimates which are available at the time of development. The 2023 annual release was aligned with changes in the 2021 Census. Since 2023, Precisely revised and updated Estimates and Projections (Australia) based on the latest available inputs from ABS and other sources. The first step in this process is to analyze the latest national and state level current population estimates (Estimated Resident Population) and projection scenarios developed by the Australian Bureau of Statistics. Typically, current estimates are produced annually by the ABS whereas projections are produced less frequently.
The latest published ABS projection scenarios are built on the ABS Estimated Resident Population (ERP) 2022 base. Demographers at Precisely produced state-level projections through a cohort component model that utilizes the latest available ABS ERP and selected components of population change from both ABS and the Centre for Population (CFP). The cohort model reflects a reasonable range of input assumptions for fertility, mortality, and migration. All projections are for males and females by single years of age and aggregated to five-year age cohorts.
Next, a series of control totals for total population and total households are produced at the SA2 level of geography. The process of population estimation uses ABS estimates of resident population (ERP). These estimates typically incorporate new information about small area population trends that may have impacted specific areas since the most recent census enumeration.
While both the magnitude and trending of the ERP are important, the final SA2 control totals are forced to conform to the higher-level state projection series. SA2 control totals are also informed by Australia Post's Movers Statistics. Total households are estimated and projected for the SA2 controls through a combination of methods involving household headship rates, average household size, and an analysis of the “other population” whose living quarters are not considered housing units, such as military bases, prisons, and dormitories.
While standard demographic factors – fertility, mortality, and migration – are necessarily the key inputs to the estimation and projection process at higher levels of geography, at the smallest areas the key factors are those that affect the number of occupied housing units. Those factors include housing starts and new construction, which can be reflected in new dwelling approval data, but also by conversions, mobile home placements, demolitions, and trends in vacancy rates. Since the final result of the combination of all such factors is an estimate of occupied housing units or households, Precisely’s process focuses on occupied housing units at the smallest geographic level. Historical trends are combined with adjustments to higher-level control totals to arrive at an estimated current year number and a series of projected numbers. This trend analysis includes considerations of decline as well as growth, and relative stability along with high rates of change.
At the SA1 level of geography, Precisely’s analysts employ a variety of proprietary techniques to both establish the current year magnitudes for total population and total households, as well as establish a trajectory that informs the projection series. For example, new dwelling approval (NDA) data from ABS are incorporated in order to capture new housing developments. Time series aerial photography resources are used to confirm trends in special cases. Aggregated address counts from Precisely’s G-NAF Premium Product are used to corroborate areas of growth. The final step involves controlling the estimates and projections to higher-level control totals through the method of iterative proportional fitting.
Once the estimates and projections of total population and households are determined, the characteristics of households (such as household size and composition) and the characteristics of the population (age distributions, sex ratios, etc.) are generated according to a process that advances and updates those characteristics from the base year or census data. Precisely has incorporated the SA1 census population by ages and sex to create base year population characteristics. These population characteristics are updated annually utilizing post-censal SA1 estimates of resident population by ages and sex, once they are re-based on the latest/current census foundation. The key to this phase of the development process is to properly reflect the changing age structure of the population based on births, age-specific deaths, and the net impact of both domestic and international migration. As with all variables in this data, the results of the estimation and projection process at the lowest geographic levels are controlled to the data previously established at the higher geographic levels.
Another important household characteristic is household income. The core methodology in evolving household income in the intercensal periods is explained below.
For 2024, the estimate of current-year household income began with census distributions for weekly income. Distributions of households by income group, as captured in the census, were advanced by year in a two-stage process. The first state of household income estimates development moved census-based household income by year to 2023 using trends established form information based on the ABS release, Weekly Payroll Jobs and Wages. The survey utilizes Single Touch Payroll (STP) data from the Australian Taxation Office. The Tax Office estimates that 99% of all employers with at least 20 employees and 71% of businesses with 19 or fewer employees utilize this system. While there are gaps in the data, this dataset provides a very detailed view of the flow of income and jobs by broad industry and by Level 3 Statistical Areas (SA3).
The second stage evolved the 2023 income estimate to the 2024 estimate using an additional data source from ABS: the quarterly Monthly Employee Earnings Indicator (MEEI). This is also drawn from STP data and "presents experimental estimates of wages and salaries paid by active employing businesses and organisations to their employees in the Australian economy." The MEEI was used in modeling for 2024 because the Weekly Payroll Jobs and Wages report no longer published income statistics at the sub-state/territory granularity as of 2023. Overall, data from the MEEI release provides contemporary data, by industry, on the trajectory of wages. As this data is gross income and not per capita, data from the ABS quarterly release Labour Force, Australia, Detailed, was used to create overall per capita income growth.
Through the data and methodology outlined above, Precisely provides users with a seamless continuum of household income growth estimates that are informed by broader trends and granular data.
In summary, Precisely uses a variety of techniques, including both traditional demographic analysis as well as proprietary methods, to develop and annually update the estimates and projections database. Considerable care is taken to make the most reasonable assumptions as to births, deaths, migration, housing developments and other factors to arrive at the published final data. Nevertheless, users are advised to pay attention to the caveats discussed in the next section.