Re-imagining Household Income, Housing, Wealth and Expenditure Data
In July 2021, the ABS started work to look at how we could deliver more frequent and insightful data on the economic circumstances of households. Our work is focused on:
- re-designing the Household Expenditure Survey (HES) and the Survey of Income and Housing (SIH) to deliver better quality data and to make it easier for households to complete the surveys
- making more use of administrative data to deliver insights and to reduce the number of questions we ask in our surveys
- exploring the use of alternative big data sources, such as bank data.
Modernising ways to collect data from Australian households and making better use of existing data sources are organisational priorities for the ABS. The re-imagining project is helping deliver these priorities.
The focus of survey re-design includes:
- prioritising a digital-first approach
- improving question flow and sequencing
- sample design to maximise coverage across populations
- making it easier for respondents to complete the survey
- identifying where alternative data sources can replace questions.
While we have been looking to improve how we collect and release SIH and HES data, the fundamental goal remains the same; continuing to support primary uses of this data, including:
- supporting the compilation of key economic indicators including the National Accounts, Consumer Prince Index (CPI) and Pensioner and Beneficiary Living Cost Index (PBLCI)
- modelling the effects of taxation and income support systems on the cost of living and wellbeing for people and households
- understanding the distribution of economic resources among private households
- analysing the households most at risk of experiencing economic hardship.
Survey of Income and Housing
We have completed the re-design of the Survey of Income and Housing and the 2023-24 survey is underway. The re-designed survey includes:
- a digital first approach that better aligns to the expectations of survey participants
- an improved questionnaire with a reduced number of questions making it easier for people to accurately respond.
Some data that used to be collected in the survey will be replaced or supplemented with public sector data, including:
- government payments
- superannuation balances
- student debts.
We have started work to design the 2025-26 survey. We will continue to focus on making it quicker and easier for people to complete the survey by improving what we ask and how much we ask about. We're also improving our processes so that we can reduce the time it takes to publish our findings.
Household expenditure data
The work to improve the collection of household expenditure data is challenging and is taking longer than we first planned. Our new plan is to conduct a Household Expenditure Survey in 2027-28.
Over the course of the re-imagining project to date, we have completed several proofs of concept to explore how we might:
- reduce the volume of expenditure data we collect directly from households
- make use of private sector, big data sources
- underpin survey design with recall questions rather than a diary.
Research findings from this work include:
- Transactional bank or financial data can't seamlessly replace all detailed expenditure data from surveys without extra information like breakdowns of spending by merchant type.
- It may be possible to develop a hybrid transactional and survey dataset, producing metrics, by taking advantage of aspects of each source.
- It's difficult for respondents to remember all of their purchases and that relying solely on respondent recall poses data quality and accuracy concerns.
The pathway to measuring household expenditure remains complex. We will provide further updates on the nature and content of our future household expenditure survey as more details become available.
Administrative and big data developments
Outside of HES and SIH, the ABS has leveraged big data sources to gain new insights into household finances. Some examples are:
- Monthly Household Spending Indicator published monthly commencing December 2021 is an experimental indicator using aggregated, de-identified banks transaction data.
- Personal Income in Australia is compiled from the Linked Employer-Employee Dataset (LEED), which is built using Australian Tax Office (ATO) administrative data linked to ABS Business Longitudinal Analytical Data Environment (BLADE). LEED data is also used in Jobs in Australia.
- Using scanner data to estimate household consumption, September 2021 describes how supermarket "scanner data" will be used to produce current price and volume estimates of food consumption for the National Accounts. These estimates were implemented from the September quarter 2022 National Accounts, released in December 2022.
- Census insights on income in Australia using administrative data published in May 2023 using data from the ATO and Department of Social Services (DSS) to calculate the amount and types of income received.
An experimental Monthly Employee Earnings Indicator derived from Single Touch Payroll was first published in June 2023.
The following publications summarise insights from consultation around this re-imagining project:
- Household Expenditure Survey data - Report on user views, July – August 2021
- Insights into use of Household Expenditure Survey data, October 2021.
In 2022 we undertook a two phased co-design process with users including the Treasury, Reserve Bank of Australia, Australian Council of Social Services, and DSS. Phase 1 of the co-design process sought to capture the critical data requirements of both SIH and HES users. Phase 2 of this process enabled feedback on proposed content, elements of collection design, and planned uses of administrative data.
In March 2023 we held a webinar updating stakeholders of the status of the re-imagining project.
In September 2023 we wrote to all stakeholders involved in the co-design process, giving an update on the new timings of HES data collection and what to expect next for the re-imagining project.
Keeping up to date
If you are interested in sharing your perspective on this project, please send an email to: email@example.com.