Population estimates
Estimates of the First Nations population used in the calculation of population rates
Australian Bureau of Statistics (ABS) estimates of the Aboriginal and Torres Strait Islander population are used in the calculation of population rates for many measures used in AIHW and other national reporting. The population rate is a measure such as the number of hospitalisations (numerator) expressed as a fraction of the total population (denominator).
Every five years, following each Census, the ABS recalculates the estimated resident population (ERP) of First Nations people, along with an updated time series population trend for previous periods ('backcast') and for forward periods ('projections') based on the best available demographic information (e.g. births, deaths and migration) and forward assumptions of demographic change.
In late July 2024, the ABS released 2021 Census-based estimates of the First Nations population.
Previously, following the 2016 Census population rebase, the AIHW used 'backcast' estimates of the First Nations population to calculate population rates.
However, between 2016 and 2021, the Census-based Aboriginal and Torres Strait Islander population estimates increased by around 25%, with less than half of this increase (around 44%) due to demographic factors (births, deaths and migration). The remaining increase (around 56%) was due to other (non-demographic) factors, including changes in the propensity of people to identify as an Aboriginal and Torres Strait Islander person.
Due to the large non-demographic change, there are concerns about using the 2021 Census-based backcast populations for historical reporting, as follows.
Potential for a numerator/denominator bias in historical rates
This may result when the backcast population estimates include large numbers of people who have not identified as an Aboriginal and Torres Strait Islander person in the numerator data. As a result, historical rates may be artificially deflated.
Plausibility of rates for younger age groups
A large proportion of non-demographic change recorded by the ABS was among young people and this seems to be affecting the 2021 Census-based backcast population results. For example, the 2021 Census-based backcast estimates suggest that the population of Aboriginal and Torres Strait Islander children aged under 1 declined between 2011 and 2021. This is in contrast to trends in registered births of Aboriginal and Torres Strait Islander babies which increased over this period.
AIHW advice on which populations to use to calculate rates
2021 Census-based estimates can generally be used for:
- demographic analysis of the Aboriginal and Torres Strait Islander population – e.g. number of people by and age and sex, and change over time in the size of the population
- rate calculations where the numerator is 2021 or later, except for analysis using the ‘MBS VII’ dataset (Medicare Benefits Schedule Voluntary Indigenous Identifier dataset). Data from the Medicare database are adjusted to reflect the size of the First Nations population, because identification of First Nations people in Medicare is not complete. Since 2002, those who choose to identify as being of Aboriginal and/or Torres Strait Islander descent have been able to have this information recorded on the Medicare database through the Voluntary Indigenous Identifier (VII). VII coverage varies by age group and state and territory.
For analysis of time series of rates that includes numerators prior to 2021, deliberations are underway across government agencies to agree on the most appropriate denominators to use. Until that collective decision is made, AIHW will use the 2016 Census-based population backcast estimates and forward projections as the source of denominator for analysis of population rates that includes numerators prior to 2021.
Implications for this report
The Ear and hearing health of Aboriginal and Torres Strait Islander people 2024 report contains a variety of analyses using population rates.
Based on the above advice, the following approach has been taken:
- tables with a reference period of, or starting at, July 2021 or later (most tables) – 2021 Census-based population estimates have been used
- tables with a reference period of, or starting prior to, July 2021 – 2016 Census-based population estimates have been used (pending a collective cross-agency decision).
This approach means that for some measures, the result for a rate presented in a point-in-time table (calculated using 2021 population denominators) will differ slightly from the same rate presented in the corresponding time series table (calculated using 2016 population denominators). In general, these differences are very small.
Each data table includes a footnote indicating the source and reference period of the population estimates used.
| Data tables | Analysis in the report | Population estimates used |
|---|---|---|
Newborn hearing screening (2.1.1a–b) Health checks (2.2a–c & e) HAPEE checks and diagnosis (2.4.1a–b, 2.4.2a–c) Hearing Australia clients (4.1a–d) EarTrain participants (5.3b–c) | Point-in-time, single year (calendar or financial) July 2021 or later | 2021 Census-based population estimates and projections |
Newborn hearing diagnosis 2021–23 (2.1.2) Emergency visits Jul 2021–Jun 2023 (3.1a–d & f–g) Hospitalisations Jul 2021–Jun 2023 (3.2a–e & g–j) Ear procedures Jul 2021–Jun 2023 (3.3.1a–d & f–g) Middle ear procedures Jul 2021–Jun 2023 (3.3.2a–c & e–g) Adenoidectomies Jul 2021–Jun 2023 (3.6a–c) | Point-in-time, multiple years combined All July 2021 or later | 2021 Census-based population estimates and projections |
| Deadly Ears analysis 2021–23 (1.2.4a–b) | Point-in-time, multiple years combined 6 months prior to July 2021 and year/s July 2021 or later | 2021 Census-based population estimates and projections. (as only 6 months are pre-July 2021) |
Health checks (2.2d) Emergency visits (3.1e) Hospitalisations (3.2f) Ear procedures Jul 2021–Jun 2023 (3.3.1e) Middle ear procedures Jul 2021–Jun 2023 (3.3.2d) Adenoidectomies Jul 2021–Jun 2023 (3.6d) | Time series, single year (calendar or financial) Starting before July 2021 | Use 2016 Census-based population estimates and projections, pending cross-agency advice. |
| Audiology services (2.3a–e) | Point-in-time or time series using the 2023 MBS VII weighted file | Use 2021 Census-based population estimates and projections. |
Reported ear and hearing problems, reported and measured hearing loss (1.1a–f, 1.2.1a–c, 1.2.2a–c, 1.2.3) Newborn hearing screening (2.1.1a–b) Waiting times (3.4.1a–e, 3.4.2a–e) Hearing Australia clients (4.1.1e–g, 4.1.2a–d) NDIS tables (4.2a–i) Audiologist etc and ENT workforce (5.1a–d, 5.2a–c) EarTrain participants (5.3a & d) HEBHBL (5.4a–c) | Analysis not requiring Census-based First Nations population estimates and projections | No action required |