2.3 Strategic objective three: new statistics to support Australia's emerging needs

2.3.1 Data integration is providing new insights

Data integration and increased sharing of administrative data has the potential to provide insight into “wicked policy questions” that cut across policy domains, different groups in the community, and different generations.

As a means for evaluating programs, data integration can provide insights into the effectiveness of interventions in areas such as health, social security, education, Aboriginal and Torres Strait Islander wellbeing, and the economy. This has the potential to underpin more cost effective Government expenditure and economic policy. There is also a potential to inform place-based analysis through a socio-economic and location based data resource.

With funding from the Government’s Data Integration Partnership for Australia (DIPA) initiative and user funding, there are two related aspects to the ABS data integration work that will continue in 2019-20:

• Building and maintaining an integrated data asset; and
• Using this data asset to undertake projects for policy agencies and researchers to inform “cross cutting public policy issues”.

Box 8 depicts the current ABS integrated data asset, which is built around three core domains of people, business and location. This asset has now reached a high degree of maturity with a large number of administrative and survey data included. In the last 12 months, for example, export and education data has been added. Maintaining the currency of this data asset and adding new data sources will continue to require resources in 2019-20.

BOX 8: Integrated data assets

The number of uses of this integrated data asset to inform official statistics and policy challenges is growing.

Data integration is now a critical element of official population estimates. Without integrating Medicare enrolment information with travel data from the Department of Home Affairs, the ABS would be unable to produce reliable estimates of state and territory population. This follows the cessation of the Outgoing Passenger Card for international travel.

A number of ground-breaking projects are now contributing directly to Government policy decisions and/or assessments of the economy. Examples include the allocation of school funding, assessment of wages growth by the RBA, and assessment of health outcomes from certain medicines (see Box 9).

BOX 9: Public policy focused data integration projects Improving the allocation of funding for non-government schools Integrated data delivers better measures for schools Background Based on capacity of families to contribute to the operating costs of the school Funding models so far have been based on aggregate data (SEIFA scores) More detailed data would mean more targeted funding Data Solution Direct measure of income is being produced using linked data of students to incomes of parents Allows for a more accurate analysis of parents’ capacity to contribute to school funding Outcomes Data integration pivotal for a new direct income measure More effective allocation of funding – flows to non-government schools that need it the most Potential improvement of education outcomes Better understanding of low wage growth in Australia Purpose of study To understand the factors potentially impacting on low wage growth by classifying firms in each industry into high, mid, and low productivity growth groups and then examining the relationship between wages, productivity, and other characteristics Data used Business Longitudinal Analysis Data Environment (BLADE) Findings Higher productivity businesses tended to pay higher real wages, and employees of these businesses experienced higher real wage growth. However, the increased wage growth was not in proportion to the productivity. Benefits Understanding the drivers behind wage growth is a valuable source of information for Government policy and RBA assessment of the economy. Preventing harm from medicines Purpose of the study To identify adverse health events associated with medicines Data used Pharmaceutical Benefits Scheme (PBS) data is currently being linked with hospital admissions Findings Initial findings include five medications newly identified as potentially associated with heart failure. A further 122 medications were confirmed as associated with heart failure (as an adverse event). Benefits Findings can help to better monitor the safety of medicines in Australia, improving health outcomes, targeting public expenditure, and potentially saving lives.

Linking Census data longitudinally (2006, 2011 and 2016) has provided a rich data set that is informing policies such as the COAG initiative Closing the Gap for Aboriginal and Torres Strait Islander People (see Box 10).

Current data integration projects with policy relevance that will be conducted by the ABS in 2019-20 include:

• A Treasury led project in collaboration with the Department of Social Services, Department of Education and the University of Sydney examining how parental receipt of income support and family circumstances can affect children’s social, health, education and employment outcomes;
• A Treasury led project in collaboration with the Reserve Bank of Australia and the Department of Industry, Innovation and Science to understand the drivers of multi-factor productivity;
• A Department of Industry, Innovation and Science project to identify the key drivers of innovation in Australian firms; and
• A Department of PM&C led project to examine the outcomes for job seekers who have left employment services programs in remote Australia since 2013.

BOX 10: Closing the gap targets Better measures of Closing the Gap targets Research using the Australian Census Longitudinal Dataset (ACLD) Research ACLD allows better measures of Indigenous population growth Impact of people newly identifying as Indigenous on outcome measures for all Indigenous Australians ACLD shows worsening employment outcomes between 2011-2016 While, looking at 2011 and 2016 Censuses in isolation shows a small improvement Outcomes Linking Census data over time allows us to measure the journey and progress of Australians better than the snapshot taken on Census night Government now has a better understanding of progress towards the Closing the Gap targets Policy agencies now have better evidence to set the right targets and adjust and target policies, programs and services

2.3.2 Alternative data sources

The ABS has long had an ambition of increasing its use of non-survey data, as well as emerging alternative data sources associated with innovation and technological change. Non-survey data includes for example, administrative data collected by public and private sector entities in the conduct of their business, retail transaction data, and web scraped data. Non-survey data has the potential to substitute or complement direct collection of information from business and household surveys.

Non-survey data already underpins a large proportion of official statistics including:

• Official population statistics, sourced entirely from administrative data;
• The CPI, which uses transactions data from the private sector and web scraped data;
• Trade statistics, which are primarily based on data from the Department of Home Affairs;
• The National Accounts and a number of our economic indicators which draw heavily on business and personal income tax data; and
• Financial sector statistics, which use APRA and RBA data.

Increased use of non-survey data can be more cost effective than surveys, and can eliminate duplication of effort and minimise ‘red tape’ burden on data providers.

In 2019-20, the ABS has an ambition to progress the use of innovative data sources as part of our strategic directions.

The new Regional and Agricultural Statistics Advisory Group has been established with the specific intent of leveraging alternative and emerging data sources to inform agricultural statistics, place-based socio-economic and environmental information at a regional level and further develop environmental economic accounts particularly for water and land use.

Pilot projects are being conducted with the Meat and Livestock Association and the Grains Research and Development Corporation to leverage industry data, and satellite imagery is being tested for use in gathering information about land use.

The ABS is conducting research to improve the Census count with the use of administrative data. The research is focused in two key areas: (1) improving the decision on whether each dwelling in Australia was occupied or not on Census night; and (2) developing improved methodologies to provide counts where no Census form is returned, but where the ABS believe the dwelling was occupied on Census night.

The ABS will also examine the Census experiences of peer international statistical agencies, and their use of administrative data, to inform future Census research.

2.3.3 Environmental and satellite accounts

Satellite accounts are a mechanism to re-use and combine existing information to provide a new perspective or increased transparency to information underpinning existing official statistics.

An example of this includes environmental-economic accounts which use a national account framework to allow for an expanded and enhanced view of economic activity through its relationship with the physical environment. Existing publications of water and energy accounts combine ABS economic information, some survey data and administrative data from agencies such as the Bureau of Meteorology, state governments and the DoEE.

Experimental estimates of a waste account, prepared in collaboration with the DoEE, have recently been released, and the development of a land account continues in collaboration with state government agencies and DoEE.

In 2019-20, the ABS will continue to produce a tourism satellite account in partnership with Austrade, and will look to update the experimental transport satellite account in partnership with Department of Infrastructure, Transport, Cities and Regional Development (both these accounts are funded by these partner agencies).