Key note address to the 8th International Conference on Population Geographies (1 July 2015)
David W. Kalisch, Australian Statistician
I would like to start by acknowledging the Turrbal and Jagera people, the traditional owners of the land we are meeting on today. I'd also like to pay my respect to their elders, both past and present, and acknowledge all Aboriginal and Torres Strait Islander elders attending this conference.
I would also like to thank you all for the opportunity to speak to you today, as the keynote address of what looks to be a very interesting program.
Population is a topic that is near and dear to the ABS, and has been a core focus for us over the past 110 years. In fact, as Australian Statistician, I am conscious that there are only two statistical activities that are required by our legislation - the five-yearly Census of Population and Housing, and our quarterly estimates of the Estimated Resident Population of Australia and its states and territories. Such is the importance of population and its geographic distribution to our democracy, society, economy and environment.
Since the formation of the (then) Commonwealth Bureau of Census and Statistics in 1905, the ABS has provided a quality and trusted statistical service to governments, business, civil society and the community, producing an ever-increasing range of population, social, economic and environmental statistics.
Government, business and the community rely on ABS data to make very important decisions, and the ABS seeks to deliver the best quality and most timely information possible from the resources available to us from government.
In recent years the ABS has released more than 700 statistical products annually, with more than 13 million visits to the ABS website, and around 3 million downloads of ABS data. Quite obviously, ABS statistics are a critical part of today’s information environment.
While the current role of the ABS remains to provide official statistics for Australia, continuing the responsibilities and expectations set in 1905, we do operate in a very dynamic environment.
Much has changed in terms of our economy, society and environment, our data users have increased expectations around the quality and timeliness of information, and we live in an information age where information can be increasingly sourced from public and commercial sources. I expect that the pace of change we will face will only increase in the future.
To give a sense of some of the metrics, over the past fifteen years, the Australian population has grown by about 25%, the economy (as measured by GDP volumes) by 60%, and the labour force has increased by 30%.
In addition, the dynamic nature of our economy, the mobility of our population, and ever-changing household and family structures provide us with further measurement challenges.
More information, albeit of varying quality and utility, is being generated than ever before. A lot of information is collected on the transactions that individuals, households and businesses have with governments through the tax system, social and community services, health and education, as well as the range of regulatory agencies. In this digital age, more information is also being collected by businesses about their services and customers.
This is the context in which the ABS now operates and around which expectations for the ABS are being redrawn.
While the scale and complexity of the Bureau’s operations have increased over the past decades, the resourcing we have received from governments has been declining. We now have one quarter fewer employees now than we did in 1998. At the same time, the range and quantity of statistics produced by the ABS has increased enormously over the years.
Reconciling expectations with available resources, the classic economic dilemma, is confronted by the ABS every year. We need to continually reassess what we do and how we work, to deliver the best possible statistical program from the resources we have.
Technology and priority setting have been the two main mechanisms used to reconcile expectations with resources over past years, and I expect this will continue into the future.
Meanwhile, external scrutiny and reviews have highlighted a number of areas where we need to fundamentally change.
Since my appointment in December 2014, I have emphasised three main areas of focus for the ABS if we are to meet this challenge:
1. Continued delivery of robust, timely data
2. Increasing the level of innovation in order to meet new and emerging needs
3. Building strong partnerships to improve the availability and use of data.
The ABS faces the challenge of being an ambidextrous organisation –continuing to deliver reliable, timely, consistent time series information, particularly around our major population and economic information, while also innovating to take advantage of new information possibilities to meet new information requirements from key users. I will discuss this in more detail shortly.
But first, I want to recognise and comment upon the Government’s announcement in the 2015-16 Budget of a $250 million investment in the ABS across five years to refresh our statistical infrastructure, systems and processes.
Government investment in the ABS
This substantial Government investment in the ABS systems and processes is an important part of our transformation and complements the other transformation elements being pursued by the ABS.
This significant investment will ensure the ABS has the necessary infrastructure in coming years to continue delivering accurate, timely and quality demographic, economic, social and environmental statistics for Australia.
In making this investment, the Government has demonstrated its confidence that the ABS can effectively implement a major statistical business transformation program over coming years, to better manage our risks and put in place more efficient, enterprise-wide approaches. This is the first major infrastructure investment in the ABS for over two decades
As also agreed with Government, and required by our legislation, we will deliver a quality Census in 2016. I know this is close to the hearts of many of you here today, with the Census providing the basis for our national and regional population estimates and an array of critical information at the most detailed geographic level. This will also be Australia’s first predominantly digital Census, and while planning is well underway, this will be a major priority for the ABS throughout 2015, 16 and 17.
During this time, over the next five years, the ABS must go much further in transforming if we are to conquer the challenges we face and remain relevant. For our external stakeholders, I expect you will see continued delivery of quality statistics, innovation around what the ABS does and how it works, together with improved access to data and stronger partnerships.
Continued delivery of robust, timely data
I want to elaborate firstly on the continued delivery of rigorous, robust, timely data that meets a wide range of user needs and also has the trust of our data providers. Users of this data rightly see any interruption to the timeliness or quality of this data as unacceptable.
We must therefore continue to deliver and adapt our measurement approaches to respond to constant population, economic and social change. The delivery of quality information over our history has produced our Australian and international reputation. This needs to be protected, and if possible, enhanced over time. Our labour force estimates are one example of a key series where we are working to improve data quality within available resources.
This leads me to innovation. Over coming years, the ABS must innovate, both in terms of what we do and how we work so that we operate more efficiently, and deliver a more flexible and responsive statistical program.
Doing the same thing will not be sufficient into the future, even if we become just that bit quicker in delivery, or that much more efficient. I’ll say it again – we are in the information age. There are new opportunities. There are new needs. There are higher expectations of the ABS now than was the case previously.
We are re-assessing how we operate – our process for conducting surveys, the administrative and transactional data sources we use, and the way in which this data is processed, linked and disseminated.
We must embrace opportunities to collect information in different and better ways – more cheaply, more timely, from more diverse sources, and less intrusive and onerous.
We must make better use of currently available and emerging information sources, whether they are government, commercial or community sources, which can provide us with regular and timely data inputs, as well as reduce the burden we currently place on households and businesses.
Having said that, the extent to which innovation is already occurring within the ABS is not fully appreciated by the users of our data. There is little time or opportunity for users to ‘look under the bonnet’ of statistical production, and perhaps the ABS has not drawn attention to innovative developments as well as it might have.
For example, following the 2011 Census, the ABS provided a linked longitudinal Census dataset, or the ACLD (Australian Census Longitudinal Dataset). This dataset comprises 5% of the Australian population and is publically available for research. I have recently agreed to enhance access to this dataset by making it available through all ABS onsite data labs from July 2015. The publically accessible longitudinal dataset of nearly 1 million records is enabling rich research, especially into the changing propensity of Australians to identify as being of Aboriginal and Torres Strait Islander origin. There are now 67 organisations, or 1,400 registered users, with access to the file, and to date, 12,000 tables have been run in TableBuilder.
The Census longitudinal dataset, started in 2006, links Census responses over time to get more of a sense of the change across the Census time periods. In taking a 'longitudinal' view of the journeys of individuals, it provides new insights into the dynamics and transitions that drive population, social and economic change over time, as well as how these vary for diverse population groups and geographies.
There are further examples of important, innovative uses of data within the ABS and public sector more generally.
The ABS Centre for Data Integration has progressed projects including migrant settlement outcomes, early childhood education participation, a longitudinal labour force dataset, and a business longitudinal dataset.
We are working with the Departments of Social Services, Human Services, Health, ATO and to investigate linkages between Census, social security information, MBS and PBS data and income data and with the Queensland and Tasmanian Education Departments on experimental estimates of education outcomes.
ABS is also moving beyond the public data environment to draw insights from retail scanner data, and investigate use of satellite imagery to measure agriculture crop yields. The spatial opportunities of these big data approaches are considerable and have the potential to fundamentally change how we produce population information – especially the extent to which we can measure temporal dynamics which have generally been beyond the reach of traditional approaches.
While these are exciting, original developments, the rate of innovation needs to accelerate if we are to contribute insights to wicked government policy questions and dilemmas facing business and the wider community.
Building strong partnerships to improve the availability and use of data
I want to also highlight one of the hardest, but very important, area of transformation for any organisation, public or private – cultural change.
The ABS has an enviable reputation for the work it has delivered over past decades. However, as noted earlier, drawing upon the findings of a capability review undertaken in 2013, the ABS is also aware of some major improvements it can make to how it works.
We are a key, but not the only, player in an elaborate and inter-connected information environment.
We need to get better at partnering and collaborating with other government agencies and key stakeholders to deliver value to the ABS and our partners. Our partners and stakeholders have many insights we need to properly appreciate and understand. Together we can deliver better value for taxpayers and the community than if we work in isolation.
I am aware there is a widely held view that the ABS has not done enough to make data available, that we haven’t built relationships that help realise the full potential of our data.
I shared this view before I joined the ABS, and we have already taken some early steps to change this over recent months. While our existing legislation, largely established in the 1970s, does generally have a good balance between maximising the use of information and protecting the sensitive information of individuals, the ABS has traditionally taken a very conservative approach to the use of that information.
We do have some opportunity to take a more open approach to the use of our data and to develop workable solutions within our existing legislative framework.
We also now have technological solutions that enable us to manage sensitive data in ways that do not disclose personal information, such as our data integration facility and confidentiality tools.
A particular focus will be on access to micro-data. This data offers enormous analytical opportunities, but any access arrangements must be navigated with care in order to respect the confidentiality and privacy expectations of our data providers.
However, our partnerships need to extend beyond a transactional focus of we have data and you want it! It needs to encompass shared perspectives around research questions and the type of data you require in the future. More of a sense of co-designs beyond just the provision of data.
Transforming People Statistics
Some of you in the room will be aware of a further transformation program we have embarked on in our population and social statistics program titled ‘Transforming People Statistics’, which we exposed some early thinking on at the Australian Population Association Conference last December. This initiative seeks to fundamentally transform our population and social statistics programs into a truly 21st Century program.
Since the 1960s, ABS population and social statistics have been underpinned by a 5-yearly population Census, a program of varied, independently run population surveys, and various collections of administrative data. While people statistics have evolved incrementally over time, the broad paradigm has remained, with a few key exceptions (eg. data integration), largely unchanged.
The ‘Transforming People Statistics’ Program provides a focused strategy around the three areas of focus that I have outlined, and presents a future-focused vision for people statistics.
Firstly, it presents a roadmap for transforming people statistics to be more sustainable and of higher quality – to ensure their long term viability as robust and timely information.
Integration will be critical – both in terms of how we bring together multiple data sources (for example, Census and Survey data, Census and Administrative data, and so on), and in how we design our data to be more integrated. As part of this initiative, the ABS is proposing to bring all of its population surveys into a single integrated survey.
Secondly, the Transforming People Statistics Program outlines an approach for realising the potential of innovative methods to produce ‘statistical solutions’ that best meet the information needs of decision makers and users. Included within this is active exploring of how to address the demand for spatially enabled information to meet the needs for spatially focused infrastructure planning, service delivery and measurement of outcomes.
Lastly, it considers how to leverage partnerships with all data custodians, to unlock the potential opportunities for people statistics from existing statistical assets.
It also, importantly, provides the means by which population and social statistics can move from a traditional focus on collections, to being solution-centred. Official statistics of the 21st Century should be defined by the information needs of Australia and the best solutions for meeting those needs, more than the mechanics of how data are collected or acquired.
In this area, we aim to improve the frequency and timeliness of our small area information by more full taking advantage of administrative and survey data, rather than solely relying on 5 yearly Census snapshots. We will draw on new methods to integrated data to provide additional insights and improve our Aboriginal and Torres Strait Islander Statistics and labour force statistics, as well as introducing new data sources to better understand service populations, transport and other population dynamics. The ABS will extend on our work with the Australian Census Longitudinal Dataset to increase our understanding on individual pathways and how the population and society changes over time.
The ABS needs to continually improve how we work, if we wish to continue being seen as a world class statistical organization. Consistent with this, we are placing greater attention on the dimensions that are required for us to be a high performing organisation – such as high standards, focussed prioritisation, effective and efficient delivery of outcomes, and constructive internal and external collaborations.
I am committed to leading the ABS through the challenges and transformation program I’ve outlined today so that it leads a modern, relevant and responsive national statistical service.
Only if we provide an extensive range of robust and timely data to guide planning and decision making, only if we become more innovative and seize opportunities to meet contemporary and emerging information needs, only if we engage in partnerships to fully realise the potential of data, can we be confident that we are delivering the information required now and into the future to enable the effective functioning of one of the world’s great democracies.