The Australian Statistician David W. Kalisch calls on economists to support publicly the importance of ABS statistics in informing important policy debates because too often they are taken for granted.
The ABS purpose is to inform Australia’s important decisions by partnering and innovating to deliver relevant, trusted, objective data, statistics and insights. The ABS informs Australia’s important decisions by Governments, by businesses, and by individual Australian citizens in their everyday lives.
ABS data informs fiscal and monetary policy settings, social support programs and infrastructure spending. ABS data informs many pertinent public policy debates, such as housing affordability, income and wealth inequality, cost of living, energy prices, the quality of life in our cities and regions, education and health outcomes, needs-based school funding, immigration policy and so much more.
In addition, ABS data is key to effective functioning of our democracy, with population data helping establish fair electoral boundaries and our official statistics informing choices by voters and political aspirants.
In 2016-17, the ABS released 727 statistical products relating to economic, social, population and environment domains. There were 16.1 million visits to the ABS website and 2.8 million downloads of ABS data. The National Accounts alone publish 12,500 statistical series each quarter. ABS statistical products form the bedrock for economic analysis and decision-making across both the private and public sectors in Australia. ABS has a significant social media presence among Government agencies, only eclipsed by the BOM weather site.
The five main dimensions
When we consider the public value delivered by the ABS we look at this from a number of angles, and consciously make choices across five main dimensions.
We seek to produce our core economic and social statistics to the highest standards. Everyone expects us to deliver “perfect” statistics, first time, every time. We deliver around 500 statistical series every year, and key data users point to one major error by the ABS on average every five years.“We deliver around 500 statistical series every year, and key data users point to one major error by the ABS on average every five years.”
We also seek to ensure our work program is relevant, and evolves to meet emerging policy needs. Our economy, society and environment keep changing, so this is an ongoing challenge.
We put effort into making sure our data are available for important uses, while also making sure we do not compromise the secrecy of individual personal and business information provided to us on trust. In past years, the ABS has had a default towards more limited use of data that did inevitably reduce the utility of ABS data to our nation and key decisions. Statistical techniques have since developed to enable a better balance between increased use and still safe use of sensitive data, alongside a contemporary understanding of our legislation.
We continuously seek to improve the efficiency of our operations, including data capture. We recognise the burden we place on households and businesses to respond to surveys and we look to capitalise as much as possible by using existing data collections.
At the same time as we are delivering our current statistical program, we are also building the future capability of the organisation so the ABS is well placed to continue operating as an effective national statistical agency into the future.“The ABS’ official data is more important than ever, in a world of contested ‘facts’, and we are constantly assessing how to maintain the quality and improve the relevance of our important statistical series.”
While we inevitably have to make trade-offs in terms of how we use our available (and generally declining) resources across each of these dimensions, my strong sense is that ABS has delivered more public value across all of these areas over recent years.
We have continued to deliver high quality statistics, reprioritised our statistical program, enhanced data access for users alongside strengthened data secrecy protections, made greater use of available big data options rather than imposing directly on households and businesses, and improved the efficiency and capability of the ABS as an organisation. We cannot do everything, that either data users or the ABS would like, but in the context that we have been operating, this is a remarkable achievement.
Contributing to important public debate
The ABS has reshaped its statistical program over its distinguished history of now more than 110 years. In earlier years, this was predominantly increasing the statistical outputs of the ABS, as the ABS received increased funding from Government.“ABS has been dealing with real resource reductions of over 20 per cent over the past decade from successive governments.”
ABS has been dealing with real resource reductions of over 20 per cent over the past decade from successive governments. We have pursued increasing prioritisation of our statistical work program. The ABS has also been very explicit around statistical priorities, as published in Essential Statistics for Australia (2013) and has more actively sought the views of key data users in recent work program consultations before we have made decisions about the scope of our work program.
We have made changes to our statistical work program in 2008, 2009, 2013, 2014 and 2017. More of our statistics are now user funded, and the majority of our social statistics are only possible through user funding, with our appropriation funding predominantly devoted to our economic and population statistics.
The ABS’ official data is more important than ever, in a world of contested “facts”, and we are constantly assessing how to maintain the quality and improve the relevance of our important statistical series.
While most public attention has been given to cuts in the ABS work program, the ABS has made a number of explicit improvements in our statistical activity:
The Australian Labour Account
Understanding the Australian labour market, including wages, is a key area of focus for the community and policy makers. To support this, the ABS has developed the first ever Australian Labour Account for the period 2010–11 and 2015–16, with work underway to both add to the time series, and to produce quarterly accounts.
The Australian Labour Account has been developed to provide a framework for combining different data sources (household and business surveys, and administrative data) to provide internally consistent estimates of key labour market variables related to jobs, persons, hours and payments for labour. The accounts provide a richer picture of the dynamics of the labour market and help make sense of the extensive range of labour related data.
They will also assist in assuring the quality of national accounts and will improve the reliability of both labour and multi-factor productivity statistics.
A key finding from the Australian Labour Account analysis showed the growth in the number of persons with multiple jobs. Over the six years to June 2016, filled jobs grew by 6.9%, from 12.4 million to 13.2 million. Over the same period, the number of employed persons increased by 6.8%, from 11.7 million to 12.5 million. The slightly slower growth in employed persons compared to filled jobs reflected growth in multiple job holding, with second jobs increasing by 64,100 (9.2%), compared with main jobs which increased by 791,700 (6.8%).
Further Labour Account information was released just a few days ago, and we will be releasing quarterly updates in the future, made possible through recent Budget funding.
ABS has also developed a cross-section Linked Employer Employee Dataset (LEED). An initial prototype was released in late 2015, with further cross-section releases anticipated through the second half of 2018. This can show the geographic dimensions of industry restructuring, features of secondary jobs and characteristics of people with multiple jobs.
The new Labour Account and LEED are adjuncts to our world-class ABS labour force survey, delivered every month, which will also soon provide monthly estimates of underemployment.
Our labour force estimates also clearly demonstrate the dynamism of the Australian labour market. The change in the number of persons employed which is published each month by the ABS of around 10-20,000 is a net figure reflecting sizable gross labour flows. Our monthly labour force data shows that around 350,000 persons commence employment while around 330,000 cease employment each month.
The Consumer Price Index (CPI)
The CPI is used by government to index a substantial component of its outlays, including payments to pensioners and other government welfare recipients, and is a key indicator used by the RBA board in their monetary policy decisions. We have recently made a number of innovative methods and data source enhancements that will enhance the accuracy of the CPI.
The CPI enhancements are significant. Specifically:
- A move from 6 yearly to annual re-weighting of CPI spending patterns, which will improve the accuracy of CPI estimates. From the December quarter 2018, ABS National Accounts data will be used to annually re-weight the CPI spending patterns;
- A move from direct price collection to the use of big data or scanner data and webscraped data.
This has significantly lowered data collection costs and improved the accuracy of the CPI. With these enhancements comes an increased possibility of the ABS being able to develop a viable and affordable monthly CPI, which would align the frequency of release in Australia with IMF international standards.
The ABS has made considerable progress extending and enhancing its measures of productivity. In 2016, the ABS extended the publication of productivity growth cycles to the industry level, and in early 2018, extended our productivity estimates to include states and territories, bridging a key data gap in the suite of productivity estimates.
Recent collaboration with the Department of Industry, Innovation and Science and the Treasury using the Business Longitudinal Analysis Data Environment (BLADE) facilitates a better understanding of Australian productivity dynamics at the firm level. Going forward, the Data Integration Partnership for Australia (DIPA) is expected to facilitate deeper productivity research using micro data from large, linked datasets. These statistical improvements will further assist informed policy debate on sustainable economic growth and material living standards.
Work is also underway to improve output estimates for the health subindustry, with future plans to review output indicators for the education industry. These are a prerequisite to extending economic performance indicators into the non-market and service sectors.
The ABS has a long history of leadership in the environmental accounts at both the national and international levels. ABS played a key role in the development of the System of Environmental Economic Accounts (SEEA) framework which has become an international standard. ABS continues to produce regular SEEA accounts in relation to energy, water, land and other environmental aspects. We also have an active experimental accounts program that is focussed on supporting key policy interests such as the Great Barrier Reef.
ABS is currently partnering with the Department of Environment and Energy (DOEE) to drive a national approach to environmental-economic accounting that will feature the SEEA framework. This approach has been agreed to by all Environmental Ministers as a means to improve decision making that involves environmental, economic and social dimensions.
This initiative is an important step forward in harnessing the potential of accounting to inform decision making that involves balancing economic and environmental benefits. Ross Gittins recently profiled this data in an article in the SMH.
Big data, integration and new data sources
I see data integration as the new frontier for statistical organisations. The ABS, alongside many other statistical agencies around the world, is focused on making better use of existing data, irrespective of whether the data has been collected by the statistics office or by other public and private organisations. Data is also being combined or integrated to provide new insights to inform the development of new policy; and to evaluate existing policies.
We’re already seeing new insights into the life journeys of Australians. Analysis to date of large integrated data sets has informed a broad range of important policy areas, including the contribution of small businesses to employment growth; the role of health services and education on life outcomes; and data insights to support needs based school funding.
The ABS is also using data integration to produce accurate population estimates without the need to fill in outgoing passenger cards – something which international travellers appreciate.
The ABS is also exploring big data sources for the production of official statistics. We are currently working with the Bureau of Infrastructure, Transport and Regional Economics and have demonstrated the feasibility of using telematics data from trucks to provide regular road freight statistics. Satellite imagery is another source the ABS has explored, demonstrating that it is possible to produce crop statistics through combining satellite data with small validation samples. And I’ve already mentioned use of supermarket scanner data in the CPI.
Safe and secure access to data is essential to maximise its public value. It’s a primary focus for me as Australian Statistician. The challenge for the ABS is to enable access to our more detailed data – what we call microdata – without putting at risk the confidentiality of the information we receive.
A balance is possible between the safe and effective use of national data to inform important decisions – without compromising sensitive information – but we are also progressing this cautiously given the strength of privacy concerns in Australia. The ABS has established a DataLab Environment which enables access by Accredited Researchers to ABS microdata products, and there are now 560 registered users of this facility.
ABS is transforming how we work
Statistical enhancements are also possible through improvements in how the ABS operates. We are not just Australia’s national statistical office, we are also a complex organisation with currently around 2500 office-based staff in nine locations and a further 400 field interviewers distributed across the nation.
Over the last three years, the ABS has been undertaking an extensive internal transformation program focused on enhancing our understanding of our external environment, and improving our organisational effectiveness through our strategy, governance, people culture and infrastructure. The infrastructure element of the program is introducing contemporary statistical systems over a five-year period.
This transformation program is challenging but will enable us to improve our efficiency and effectiveness, make it easier for people and organisations to provide information to us, and to use and access the statistics and data we produce, while we seek to ensure our products and services remain relevant for an evolving economy and society.
The ABS is now a more engaged organisation, more actively working with data users, inevitably prioritising its work program in response to resource reductions, strengthening its governance with particular attention to risk management, continuing to have a professional workforce but now with a greater range of expertise and diversity, and pursuing increased collaboration and accountability through the organisation.
SOCIETY AND POPULATION
Not everything goes to plan, all the time. As we saw in the 2016 Census collection process, the community has high expectations about service delivery from Governments, including the ABS. It is not just about the quality of the statistics. Our online Census system was offline for nearly two days within an eight-week collection period, which drew considerable media and social media reaction.
However, once the Census collection was complete, the ABS achieved an overall response rate of 95%, consistent with past Australian experience and best international Census achievements. The ABS worked hard to achieve this outcome, assisted by the Australian community’s strong support for the Census.
2016 Census data can be used with confidence. Thankfully for the many people who rely on Census data, including our governments and parliament, the Census did not fail and it was not the worst Census ever.
As an organisation, the ABS took many learnings from the 2016 Census process, and these were put in place securing the high-profile releases of 2016 Census data and then implementing the Australian Marriage Law Postal Survey (AMLPS) in late 2017.
The AMLPS demonstrated the ABS’ professionalism, expertise, and capability to inform important decisions. We achieved an outstanding voluntary participation rate of 80 per cent, bigger that Brexit and much larger than the Irish Same Sex Marriage Referendum. The integrity of the survey process was ensured through a robust survey design and impeccable implementation, all for only 2/3 of the total budget allocated by the government.
Meeting the information challenges of the future“Too often, our national statistics are taken for granted. I regularly hear the expectations they will always be provided, at the requisite quality, always available in time for whenever key decisions are being made, available for free.”
The ABS is working hard to meet the many measurement challenges of a modern economy. As the Nobel winning economist Joseph Stiglitz (2009) said: “What we measure affects what we do”. The ABS is acutely aware of current economic challenges and our role to contribute to their understanding and policy response.
Over the past fifteen years the Australian population has grown by about 26%, and the Australian economy has diversified and is much more complex. The service sector has averaged growth of 6% over this period, increasing its share of the economy to 63% of Gross Domestic Product (GDP). Measuring services is more difficult – conceptually and empirically – than goods. And the pace of change is ever increasing. As we all know, Australia isn’t an agricultural and manufacturing economy anymore. So, by any measure, the statistician’s job is getting bigger and more complex.
We are also faced with definitional, conceptual and measurement challenges arising from digitalisation and globalisation. The ABS is heavily involved in the international statistical collaboration currently underway that is considering the challenges of globalisation and digitalisation to economic measurement.
As an economist, I often think about the opportunity cost to the nation of not having the necessary statistics to inform important decisions. Official ABS statistics should form the basis for monitoring major developments across the economy, population, society and the environment. They should be key inputs to the development of major policy proposals, and can also contribute to rigorous evaluation of policy and program outcomes. Australia can make better use of data, recognising that effective use of data can improve our economic performance alongside social and environmental goals.
The ABS is more efficient and innovative now than in the past, which ensures Australians receive greater value from our data resources. While ABS statistics are generally free to the Australian community, they are not costless, given the expenses incurred to collect, process and disseminate key statistics.
I would suggest the economics profession has an important role to ensure important measurement challenges are met in the future.
From my experience across a number of policy domains, I have seen and heard many very effective advocates for public expenditure. They advocate directly to our politicians, to those who design policy options, and directly to the community. Economists are often heard in these public conversations, on both sides of the policy debate.
It would be great to see such interest and support for our critical national data resource. Too often, our national statistics are taken for granted. I regularly hear the expectations they will always be provided, at the requisite quality, always available in time for whenever key decisions are being made, available for free, that new statistics will be developed over time, and no past statistics ever stopped or reduced, etc, etc.
I believe there are greater opportunities to identify and publicly recognise the value of our national statistics; highlight the insights produced by our national statistics and associated research; support public funding of our essential national data infrastructure; and work with the ABS and others to improve the relevance and utility of our statistical system to measure a changing economy, society and environment.
This is a lightly-edited version of David W. Kalisch’s address to the Economics Society of Australia Conference, Canberra, on 12 July 2018.