2017 Menzies Oration: Democratising Indigenous Data
2017 Menzies Oration: Democratising Indigenous Data
Professor Ian Anderson AO, Deputy Secretary Indigenous
Ya Pulingina. Pangkerner Ian Anderson Palawa Trowunna: Trawlwoolway, Pairrebenne, Plairmairrerenner. Kartoometer mynee larrakia teeanner. Carnee meenee nenener nicer Lanena.
I would like to thank the Menzies School of Health Research and Director, Professor Alan Cass, for inviting me to speak tonight and for hosting this event.
It is a great honour to deliver the 2017 Menzies Oration.
I would like to take this opportunity to congratulate the Menzies School of Health Research for its ongoing commitment to examining and understanding Indigenous health. Research undertaken by the school has made a significant contribution in improving the health and wellbeing of Aboriginal and Torres Strait Islander peoples across Australia.
I would also like acknowledge the many distinguished guests attending today:
- the Honourable Vicki O’Halloran AM, Administrator of the Northern Territory;
- Professor Lawrence Cram, Deputy Vice-Chancellor and Vice-President Research and Research Training; and
- Professor Alan Cass, Director of the Menzies School of Health Research
Today, I would like to discuss the importance of data within the policy context.
Data can be an extraordinarily powerful tool.
Data shapes so much of our society. Governments, businesses, non-profit organisations and individuals all use data to inform their decision-making.
We live in a data age.
The Australian Productivity Commission in their inquiry on Data Availability and Use drew on IBM estimates :
‘the amount of digital data generated globally in 2002 (five terabytes) is now generated every two days, with 90% of the world’s information generated in just the past two years’
In addition to significant growth in the amount of data being collected, we also have the opportunity to better use our data assets. We have more sophisticated technologies to analyse and integrate data—in fact, we are also able to analyse more data, more quickly at lower cost.
By way of example, decoding the human genome involves analysing 3 billion base pairs—when this was first done in 2003, it had taken over 13 years at a reported cost of around $1 billion. Presently, it can be done in a day at a cost of approximately $3,000. 
Many people here tonight are in some way involved in the production of data. You might collect data as part of service delivery or part of a research project, analyse data to build information and evidence, or use this information and evidence to make policy decisions.
Data collected through research, survey and census or from the administration of government programs and activities, has played an important role in guiding policy and practice, particularly in Indigenous affairs.
Public data is a national asset. However, its value is only realised when data is transformed into information, and more importantly knowledge.
Data can help shape debate, identify areas of need, and be used to analyse the effectiveness of policy and other actions.
But, for this to occur, the data must be of high quality and we must collect the right data. Moreover, it must be accessible to all. I would also like to stress that data is not useful in isolation and must be combined with the skills and capabilities to make the most of it.
Democratising data – Data and Indigenous peoples, changing approaches
Now, in using the phrase—democratising data—I am signalling a change agenda to create the systems and platforms that ensure end-users are able to access data in a timely way. One that is relevant to a wider audience and allows data to be used.
The availability of data on Indigenous Australia has reflected broader politics.
It is reasonable to infer that in the late 19th century when representatives of the then six British colonies—New South Wales, Victoria, South Australia, Western Australia, Queensland and Tasmania—came together to draft up Australia’s constitution, Aboriginal and Torres Strait Islander people were not factored in.
It is probable that those who considered how the soon-to-be-federated country might operate did not want to include Indigenous people in financing arrangements for the states, or for the development of services or planning purposes.
It is also possible the founding fathers of the Australian federation did not think Aboriginal and Torres Strait Islander people were long for this world. That they were a dying race.
Under such scenarios, there would have been no need for information about Australia’s Indigenous population to be collated. The ‘race’ clauses included in our original constitution reflect this.
It was not until the successful 1967 referendum that paved the way for the Commonwealth to take on responsibility for the administration of Indigenous Affairs at a national level that Aboriginal and Torres Strait Islander people were formally included in our population counts. This marked the start of the first phase of Indigenous data development in Australia.
In the years that have followed, data sets concerning Indigenous Australians have slowly been built up. The Australian Bureau of Statistics, for example, has developed a clear picture of the lives and experiences of Indigenous Australians.
Over the past three to four decades, significant improvements have been made in the availability and quality of Indigenous data.
This includes Indigenous data obtained through our national survey program run by the Australian Bureau of Statistics. Equally important has been the collection of vital statistics—such as, data on births and deaths, hospital data and Medicare data.
However, the most significant innovation over the last two decades has been the development of performance measurement systems. These systems draw on data made available through our data collections systems. They have been developed with the specific purpose of enhancing policy decision making processes.
Perhaps one of the most sophisticated data sets relating to Australia’s Indigenous population can be found in the Indigenous health sector.
The move earlier this century to start strategically using data through vehicles such as the Aboriginal and Torres Strait Islander Health Performance Framework reports marked the beginning of the second phase of Indigenous data development in this country.
The first Aboriginal and Torres Strait Islander Health Performance Framework report was produced in 2006. Commissioned by the Australian Health Ministers’ Advisory Council (AHMAC), the data in these reports, updated every two years, and has been used by Australia’s political leaders to inform their decision-making in Indigenous Affairs.
Data is collated across three tiers: health status and outcomes; determinants of health; and health systems measures. Significantly, the measures had strategic value and ‘buy in’ from the states and territories.
The data contained in the Aboriginal and Torres Strait Islander Health Performance Framework played a pivotal role in the decision by the Council of Australian Governments to allocate resources in 2008 to address Indigenous disadvantage. However, just as importantly, the data informed where the resources would be allocated.
The collation, analysis and strategic use of Indigenous data by our political leaders is one thing—I will discuss this further when I describe our proposed approach to the refresh of Close the Gap—but making data accessible and relevant to a wider audience—and in this case, Aboriginal and Torres Strait Islander people—is another thing altogether.
Too often, data is not presented in a way that is accessible for Aboriginal and Torres Strait Islander people. Furthermore, it is interpreted in a way that is not connected to Indigenous reality.
Democratising data – Indigenous data governance
To improve policy effectiveness, policy must be developed with First Australians.
How we achieve this in the data space is pivotal. This goes to development of data governance.
An Indigenous led approach to data governance underlies the development of an Indigenous data system that addresses the priorities of Indigenous communities.
In a recent presentation at the Melbourne Institute’s Public Economics Forum in Canberra, University of Tasmania Pro Vice Chancellor for Aboriginal Research and Leadership, Professor Maggie Walter, spoke about the importance of Indigenous governance in relation to data collection.
Professor Walter draws on an Indigenous critique of statistics that points to the tendancy to decontextualize Indigenous context as a result of both an overemphasis in Indigenous statistics on aggregate reporting and its tendency to reductive analysis—in which representation of causality is often, by default, Aboriginality.
If data is irrelevant—or at least perceived to be—it is unlikely to be factored into decision-making processes.
In this regard, it is important that Indigenous Australians at a community level have the data they need to make decisions that impact on their local services and communities—I will resvisit this shortly.
Professor Walter further argues that Aboriginal people are pejoratively positioned according to five D’s: Deficit; Different; Disparate; Disadvantaged and Dysfunctional.
Professor Walter questioned what Indigenous data would look like in this country if Aboriginal and Torres Strait Islander people and interests led the process of collecting the data—if Indigenous values and socio-cultural realities were embedded in the process, and if Indigenous Australians decided what data was to be sought and were the intended audience.
The answer—and I totally agree with her—is data that is far more relevant and accessible for Aboriginal and Torres Strait Islander end-users. Our challenge is to create the institutional structures that enable Indigenous Australians to set priorities and create the analytical frameworks. As I have argued, this is an issue for local and regional governance. But it is also a challenge of inclusion of Indigenous peoples in our data systems governance.
As we reflect on the progress made over the 10 years the Closing the Gap framework has been in place, it is reasonable to question how different the framework and data collected on an annual basis would be had Aboriginal and Torres Strait Islander people led the process from the start.
It is impossible to know with any certainty how different it would be, but I can say more care would have been taken about how the Closing the Gap architecture was framed. It would have been more nuanced, less deficit-based and more focused on the strengths of Aboriginal and Torres Strait Islander people.
Nevertheless, this is a challenge. Our approach to developing a strength-based approach is to build an aspirational narrative that takes into account the importance and centrality of Indigenous culture; to choose measures that reflect Indigenous success and to frame our measures accordingly.
We cannot lose sight of the challenges that Indigenous disadvantage poses to us.
Although still being worked through by the Council of Australian Governments, in close consultation with Indigenous Australians, the refreshed Closing the Gap agenda will be based on the strengths of Aboriginal and Torres Strait Islander people.
Democratising data – local data for local decisions
With increased emphasis on local actions and place-based solutions, there is a greater need for high quality, granular, local level data to facilitate local decision-making.
Too often, the data we have remains at the national level—not at a regional or community level. A blanket national figure may have no relevance whatsoever to what is occurring in Yuendumu or on Groote Eylandt or in Wilcannia.
Disaggregated data can reveal the size of a task at the local or regional level. This is one of the learnings gained from the Empowered Community sites.
Local data can give us very specific objectives.
Providing greater access to local level data will allow communities to determine the areas and actions that are of most importance to them, and to hold governments and service providers to account.
Better access to data can help lead to greater community ownership of actions.
We are starting to see this with the Empowered Communities. However, access to local data, particularly administrative data remains limited.
It is important we explore platforms to make more local level data available to help communities and governments better understand local issues. We also need to look at ways to ensure more data, from all levels of government, is shared.
For instance, regional educators need to know local trends in school retention and schooling outcomes in order to improve education services. While improved access to local level health data can help better target services to need.
Local level data is a key to planning and development – improving the quality of local services.
Local decision-making driving changes in behaviour and values at the local level, leading to positive outcomes. Without the local data, this would not have been possible—or certainly less likely.
It is a key to accountability at a local level.
It is key to closing the gap.
Democratising data – making data more accessible
While there have been significant improvements in the type and quality of data collected, to realise the full value of our national data assets we need to look at ways of making them more accessible.
As I mentioned earlier, better access to high quality, granular data is critical for decision-making and accountability.
Improved access to higher quality, more relevant, data can help drive research and evaluation - building an evidence base to support policy.
Improving accessibility to administrative data sets has significant potential.
Governments hold extensive longitudinal administrative datasets that cover large parts of the population—with improvements in Indigenous identification over the past 10 years, this presents significant opportunities to better evaluate policies and programs and develop more effective and efficient ones.
As Martin Parkinson recently highlighted, the careful use of administrative data can help us better understand the pathways to prosperity—such data allows us to track what happens to individuals and cohorts over time.
We have a much to do to achieve this. Data is in fact difficult to access.
The reasons for this are complex. However, it was the view of the Productivity Commission that:
Lack of trust by both data custodians and users in existing data access processes and protections and numerous hurdles to sharing and releasing data are choking the use and value of Australia’s data. In fact, improving trust community-wide is a key objective. 
These hurdles include legal restrictions and concerns around privacy.
Concerns around privacy can be managed by building social licence or trust. This can be facilitated through increased transparency and accountability.
Trust can also be built by engaging Indigenous Australians in the development, design and management of data systems.
The Productivity Commission’s review into Data Availability and Use also suggested inertia or risk sensitivity continues to be a barrier to providing wider access to public sector data.
Changes here require a cultural shift—built around systems based on transparency and seeing data as an asset.
To put the challenge of data accessibility in perspective, there is already a considerable amount of high-quality data that relates to Aboriginal and Torres Strait Islander people.
In fact, Australia has—arguably—one of the most developed data collection systems in the world.
There are approximately 370 million Indigenous and tribal peoples living in 70 countries worldwide.
I led a research collaboration in which 65 collaborators surveyed Indigenous health and social outcomes across 23 countries. The results of this work were published in the Lancet last year.
Our approach to this project was designed to broaden the focus of international studies in Indigenous health from the four countries where the majority of previous international research studies had primarily focussed—Australia, New Zealand, USA and Canada.
We drew on governmental census and administrative data. This was supplemented by academic research and data from NGO’s. Local investigators were a key to finding this information, as they were able to navigate local data sources.
We collected data on seven indicators: population; life expectancy at birth; infant mortality rate; maternal mortality ratio; birth weight (high and low); nutritional status (child malnutrition; child obesity; adult obesity); economic status and educational Status.
Indigenous status was recorded in 68 per cent of measures, 88 per cent of which were by self-report. In the absence of recorded Indigenous status, we relied on proxy markers such as geo-proxies or language proxies.
The findings of this project were not surprising. As in Australia, Indigenous people around the world generally experience poorer health and social outcomes than their respective non-Indigenous populations. However, this was not always the case, and the size of the rate difference varied.
The number of indicators reported ranged from two to ten. Australia was one of two countries reporting against ten indicators. Educational attainment and infant mortality were the most reported indicators. High birth weight and maternal mortality were the least reported indicators.
Clearly there needs to be more effort internationally to improve the data systems relating to Indigenous people. Generally speaking, the poorer and more marginalised Indigenous and tribal people are around the world, the less data there is that relates to them. We only had data from one low income country.
What have we learned from this? That relative to other countries, we have decent data concerning our Indigenous population. It could be better—and what data we have could be better utilised—but we are very fortunate to have the data we have.
Democratising data – better harnessing our data assets
Making data available is only part of the solution; we must also make the most of our data assets.
From a public policy perspective, it is fundamentally important that we do not just collect data but we use it.
Data is one of the keys to informed decision making through the policy cycle.
However, currently we make very little use of our public data resources, and access for academics, researchers and communities is very limited.
Ideally, information on all government services and programmes by location should be brought together. To be useful, this would cover information programs and activities across governments.
In the future, there is no reason this data cannot be linked to other data.
For example, administrative data is useful for describing what happened. However, surveys can gather insights about the social and cultural elements of people’s lives and help us understand why people act in certain ways.
We have seen an expansion in data linkage projects, which, although they need to be handled sensitively, have the potential to provide crucial evidence of people’s interaction across multiple programs and services.
Data linkage provides significant opportunities to increase the sophistication of data usage—create more useful information and build knowledge and evidence.
Breaking down our silos and bringing together information from disparate sources, can help us gain a greater appreciation of how individual First Australians are interacting across a range of government programs and services—for instance in the social security, health and education sectors. It is good to see the Menzies school’s leadership in this area.
What are the implications of having better linked, more accessible, more relevant data relating to Aboriginal and Torres Strait Islander people? It means Indigenous end-users, be they Indigenous organisations, Indigenous communities or individual Indigenous people, are more likely to utilise the data to help them make informed decisions. Decisions informed by accurate information are more likely to be beneficial.
High-quality data also helps to increase the evidence base from which better, more targeted Indigenous policies can be developed. This goes to the heart of the refreshed Closing the Gap agenda.
Indigenous leadership in research and evaluation is also important.
Looking for opportunities to develop processes and mechanisms to work with Indigenous researchers, and build capability will be crucial—particularly those who specialise in quantitative research methods.
Developing greater Aboriginal and Torres Strait Islander research leadership will help build a more robust evidence base for policy. More Indigenous Australians leading evaluations of programs and activities can help improve accountability and improve policy design and service delivery.
In turn, building greater Indigenous research leadership will help to improve how we utilise data in policy making.
Democratising data – using data and evidence to inform policy
The Closing the Gap refresh is an important opportunity to better harness data and evidence to inform policy.
We are looking to use data as a fundamental tool to make the refreshed Closing the Gap agenda more meaningful.
The Closing the Gap framework, first agreed by the Council of Australian Governments in 2008, now includes seven education, employment and health targets.
It is nearing its 10-year anniversary. Some targets are due to expire in 2018.
There is political and community appetite for broader change.
In June 2016, the Council of Australian Governments agreed to consider a refreshed framework, targets and implementation principles for the Closing the Gap agenda.
The Government has highlighted the importance of using the best available data and evidence and ensuring this data and evidence is made available to decision makers across governments and communities.
Clearly, the targets we set, how they are measured, and the related performance indicators, will influence what governments, communities and individuals strive to pursue, and the activities, and programs they adopt.
Targets, to the greatest extent possible, should be grounded by science.
In this regard, when setting targets, we need to base them on a sound framework built upon a clearly defined goal and a related Theory of Change. It is important that targets are SMART—that is specific, measurable, achievable, relevant, and time-bound.
Where targets are not SMART, and not built on evidence, the causal links between actions to objectives are unclear—this can undermine policy action and contribute to a deficit mindset. Targets that are not SMART can also reduce credibility and overshadow success stories and areas where significant progress is made.
Building SMART targets, closely linked to objectives, will also help focus attention on desired outcomes for policy during the policy development phase, and support ongoing monitoring and evaluation. Targets that are achievable and time bound helps to improve accountability.
This will help ensure investments are targeted towards proving the greatest benefit to First Australians—allowing for rapid changes where actions are not working, or upscaling where actions have been seen to have a positive impact.
Democratised data is key to unlocking genuine regional empowerment in people taking responsibility for their own futures
We are committed to working with Aboriginal and Torres Strait Islander people.
Establishing data governance arrangements will be important.
We need to engage fully with Indigenous Australians in the process of developing, designing, collecting and utilising data.
We need to place a greater emphasis on the use of regional data. This will allow communities to take ownership of placed based actions and to hold governments and service providers to account.
We also need to look at ways to improve data access while at the same time always ensuring individual privacy is maintained.
Improved access to data can help drive research, and help build an evidence base to support and drive policy that improves outcomes for First Australians.
We know a one-size-fits-all approach is not the most effective way to improve outcomes for Aboriginal and Torres Strait Islander people.
It will put behind us the negative mindset, replacing it with a positive narrative focused on enabling Indigenous Australians to lead lives they value while at the same time supporting Indigenous advancement.
 Productivity Commission (2017) Overview of Inquiry to Data Availability and Use (page 4)
 Productivity Commission (2017) Overview of Inquiry to Data Availability and Use (page 2)