2900.0 - Census of Population and Housing: Understanding the Census and Census Data, Australia , 2016  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 08/11/2017  First Issue
   Page tools: Print Print Page Print all pages in this productPrint All


This variable contains a person's age, and is collected for each person. Age data are also available in variables that group age in 5 year groups and age in 10 year groups.

How this variable is created

Respondents had the option to report either Date of Birth or Age. Where both sets of information were provided, Date of Birth was used to derive an age in years, except in circumstances where errors in the Date of Birth were identified. An image of this question is provided below.

Age imputation

Where no Census form has been returned for a dwelling that is identified by a Census Field Officer as having been occupied on Census night, people are imputed into that dwelling (both numbers and key demographic characteristics such as age, sex, marital status and usual address). The majority of imputed ages are attributed to imputed people. Refer to item non-response rates for more information.

A smaller proportion of imputed age was due to people who returned a partially completed form without reporting age. Where age was not stated or could not be derived, it was imputed using other information on the form and an age distribution of the population. There were also cases where respondents reported an age outside the acceptable range or where there were inconsistencies between age and relationship data.

Imputation Flag for Age variable indicates if a person's age was imputed in the Census. There are two categories:

    1 Age not imputed
    2 Age imputed

For 2016 Census data, the ABS introduced a new approach for age imputation. This methodology addressed data quality concerns with imputed age records in 2011 Census data and improved processing efficiency. A two step procedure was implemented together with age bounding rules to impute age. First, a linear regression model was used to model the likely age of recipients and rank their tendency be either younger or older. This ranking was then used in combination with other reported characteristics to impute an appropriate age value to recipients from similar individuals in the responding population.

The new methodology has improved age data quality for the 2016 Census. However, there are a small number of areas where there was a high level of non-response to the age question and age was imputed. The age distribution in areas where there are high levels of age imputation may not be representative for the underlying population and should be interpreted with caution. This is more noticeable in the small number of remote communities where there is a high level of age imputation (more than 20% of Census counts) combined with the reporting of Aboriginal and/or Torres Strait Islander status. The data for people where age was reported is not impacted.

Variable history

A question relating to Age has been asked at every Census since 1911. Since 2006, respondents have had the option to report either Date of Birth or Age in years. No changes were made for 2016.

Non-response rate

Using the Imputation Flag for Age, the non-response rate for this variable in 2016 was 5.5% (4.2% in 2011). For more information, refer to Understanding Census data quality.

Data usage notes

Respondents occasionally make errors in reporting their age or date of birth, such as reporting their last birthday or the date they filled out their Census form rather than their date of birth, parents filling in the form on their children's behalf and inadvertently reporting their own age or date of birth for their children, character transposition errors, or other typographical errors (e.g. sticky key repetition). The latter two are particularly relevant to the online version of the Census form.

Data captured from written responses carries a small risk of character recognition error, mainly caused by poor handwriting or respondents writing outside the question box. The vast majority of individual characters written on paper forms met pre-set recognition confidence levels and were accepted without further examination. Characters that failed given recognition confidence levels, or responses where there was a discrepancy of more than three years between age last birthday and date of birth, were sent to clerical officers for further determination based on visual inspection of an image of the response.

Other checks on age (such as unlikely combinations of age and other variables) were made at various stages during processing, to ensure an acceptable level of quality was maintained.

Further information

A definition of Age is available in the 2016 Census Dictionary.
A definition of Imputation Flag for Age is available in the 2016 Census Dictionary.
For the official ABS estimate of demographic data on the Australian population, users should refer to Australian Demographic Statistics.

Household Form question image

Question 4 as it appeared on the 2016 Census Household Paper Form:

Image: 2016 Household Paper Form - Question 4. What is the person's date of birth or age?
A text only version of the online Census Household form is available from the Downloads tab.