AUSTRALIAN CENSUS LONGITUDINAL DATASET

The Australian Census Longitudinal Dataset (ACLD) uses data from the Census to build a longitudinal picture of Australian society.

The ACLD is a random 5% sample of the Australian population and three waves of data have so far contributed to the ACLD from the 2006 Census, 2011 Census and 2016 Census.

The 2011-2016 ACLD brings together a representative sample of over 1.2 million records from the 2011 Census with corresponding records from the 2016 Census. This will be expanded to include a third time point, with records from the 2006 Census, in mid-2018.

The 2006-2011 ACLD brings together a 5% random sample of approximately one million records from the 2006 Census with corresponding records from the 2011 Census.

The datasets cover all areas of Census data including:

• demographic information
• education and labour force data
• income
• caring responsibilities and disability
• voluntary work
• household characteristics
• family composition.

In taking a longitudinal view of Australians, the ACLD uncovers new insights into the dynamics and transitions that drive social and economic change over time, conveying how these vary for diverse population groups and geographies. The ACLD adds further value to Census data by providing insight into the pathways that lead to particular outcomes, and how these pathways vary for different population groups. It also enables the study of likely consequences of certain socio-economic circumstances for different population groups, as evidenced by the patterns in the longitudinal data.

For example, policy makers and researchers have used the ACLD to:

• better understand the socio-economic characteristics of people who have changed their self-identification as Aboriginal and/or Torres Strait Islander, and the resulting impact on statistics about Aboriginal and Torres Strait Islander populations;
• investigate employment outcomes for workers leaving the motor vehicle industry; and
• investigate changes in family relationships and fertility.

The ABS plans to continue to expand the ACLD by creating a new Panel for each Census, combining each Panel with successive Censuses and, where appropriate, linking the ACLD to other datasets.

Further information about the ACLD can be found in Microdata: Australian Census Longitudinal Dataset (cat. no. 2080.0).


AVAILABILITY OF THE 2016 ACLD

The 2011-2016 ACLD is available to registered users in TableBuilder, enabling them to create their own customised tables using weighted and unweighted data. In-built confidentiality processes prevent the identification of any individual or household.

In addition, the ACLD is available in the ABS DataLab as a microdata product to approved users undertaking approved projects. The DataLab provides high analytical utility using a range of current statistical software.

Initial analysis outlining broad employment and education transitions from 2011 to 2016 is available in Australians' journeys through life: Stories from the Australian Census Longitudinal Dataset (cat. no. 2081.0). This release includes aggregate data presented in Excel tables.

A subsequent release is planned for mid-2018 to expand the dataset to 3 time points (2006-2011-2016), and will include additional variables, such as Socio-Economic Index For Areas (SEIFA).

Information on how to access the 2011-2016 ACLD can be found in the publication Microdata: Australian Census Longitudinal Dataset (cat. no. 2080.0) or the How to Apply for Microdata page on the ABS website.

The 2006-2011 ACLD continues to be available to registered users in TableBuilder and to approved users via the DataLab.


HOW MUCH WILL IT COST?

For information about pricing, see Microdata prices.


HOW IS THE ACLD SAMPLE MAINTAINED?

Without sample maintenance, the ACLD would decline in its ability to accurately reflect the Australian population over time due to:

• people newly in scope of the ACLD (i.e. children born and immigrants arrived in Australia since the previous Census) not being represented in the sample;
• people selected in the ACLD sample no longer being in scope due to death or overseas migration; and
• missing and/or incorrect links (linkage bias).

Linkage bias occurs where certain populations are more difficult to link than others (e.g. Aboriginal and Torres Strait Islander people, young males), so links are more likely to not be identified for members of these groups and, if they are found, have a higher chance of being inaccurate. If left untreated, the representation of population groups suffering from linkage bias would worsen as each new Census is linked to the ACLD.

The ACLD sample is maintained through application of the Multi-Panel framework, developed by Chipperfield, Brown & Watson (2017). This framework provides an approach for selecting records in the ACLD to create panels which maintain the longitudinal and cross-sectional representativeness of the dataset over time, while minimising the impact of accumulated linkage bias on longitudinal analysis.

The Multi-Panel framework designs multiple overlapping panels, with each panel representative of a Census population (2006, 2011, 2016, etc.) that is linked to subsequent Censuses. The sample selection strategy for each panel is designed to:

• maintain a linked sample size of 5%;
• maximise sample overlap between the panels; and
• introduce new records to the dataset in each panel to account for new births, migrants and missed links in previous panels.

This allows flexibility for users, who can draw on the most appropriate panel for their research question. Each panel will be created as a separate dataset to minimise the complexity of the weighting strategies for users.


HOW IS CENSUS DATA LINKED IN THE ACLD?

Data from the 2011 ACLD Panel sample and the 2016 Census were brought together using data linkage techniques.

Data linkage is typically undertaken using a combination of deterministic and probabilistic methods:

• Deterministic linkage involves assigning record pairs across two datasets that match exactly or closely on common variables. This type of linkage is most applicable where the records from different sources consistently report sufficient information and can be an efficient process for conducting linkage.
• Probabilistic linkage is based on the level of overall agreement on a set of variables common to the two datasets. This approach allows links to be assigned in spite of missing or inconsistent information, providing there is enough agreement on other variables.

The 2011-2016 ACLD builds on the success of the ABS' data integration program from the past decade, benefitting from advances in linking methodology, technology and data availability to deliver a high quality integrated statistical resource.

To protect the privacy of Census respondents, we used an ABS encoded Census name for linking 2011 and 2016 Census records in the ACLD. Encoding was undertaken in 2011 for the purpose of protecting privacy by anonymising name and improving the future quality and efficiency of the linking process.

The codes are created by grouping people with a combination of letters from their first and last names using a secure one-way process, meaning that a code cannot be reversed to deduce the original name information. Each code represents approximately 2,000 people drawn from many different letter combinations, and therefore is not unique to an individual. Actual name information from the 2016 Census was not used to link to 2011 Census records.

The codes are only accessible to those ABS officers creating the linked dataset, and will never be released outside the ABS.