New approaches to undertaking a Census are being pursued by statistical organisations around the world. Australia, together with Canada, New Zealand, USA and the UK, is looking at approaches that use data obtained as Australians interact with government (administrative data) to improve the quality of Census data, inform our operations, and collect our data more efficiently.
The 2021 Census will use administrative data to improve how we gather and process data from Census forms. These innovations are a result of the following research agenda - Investigating administrative data in relation to the 2021 Census.
Better information increases participation
Comparing 2016 Census data results with administrative data is always done in a secure way that ensures confidentiality and provides information on the general characteristics of dwellings that were harder to reach in the 2016 Census. This research, along with more up to date administrative data, will be used to inform strategies that may improve the participation of hard to reach population groups.
Improving how we determine an occupied/unoccupied dwelling
To ensure the most accurate counts of people from the Census, we need to determine whether each dwelling was occupied or unoccupied on Census night.
In addition to the observations made by Census field staff, the 2021 Census will use a 'signs of life' indicator using a range of administrative sources to improve the decision on whether the dwelling was occupied or not.
Our research shows that using administrative data in this way will improve the accuracy of the Census.
Improving counts of Australians
The 2021 Census will use an improved methodology to provide counts where no Census form is returned, but where we believe the dwelling was occupied on Census night. This methodology will use administrative data to identify how best to impute (i.e. calculate) basic Census data (counts ages, sex and marital status) for such households.