Unlocking the Value of Big Data: Implications for Social Policy and Social Justice

Collecting, storing, analysing and linking big data has the potential to provide significant public benefits. However the use of big data can also come with social costs such as seen with the recent ‘Robodebt’ initiative that resulted in a class action against the Federal Government. In today’s blog piece Katherine Curchin and Ben Edwards from ANU discuss the potentials and pitfalls in the use of big data and the role the Australian social research community has to play in examining the social costs and benefits, especially for Australia’s most disempowered people.

Unlocking the value of data is the Australian government's latest ambition. Government investment in big data infrastructure will increase the capacity to target and tailor government service delivery. It will enabled greater automation, potentially enabling staff to focus their energies on the most complex cases. It will also reduce reporting burden as customer information already held by the Government is used to help complete applications for new services. But as Virginia Eubanks argues data-driven decision-making also poses the risk of serious harm and injustice. 

Can the Australian government be trusted with the power its investments in data infrastructure will unleash? Some Australians will be unable to answer this question without thinking of the disastrous Online Compliance Initiative for debt recovery – commonly known as Robodebt. Launched in 2016, this initiative involved the automatic generation of debt notices using a faulty algorithm that averaged income across the year and relied on the cross-matching of Centrelink data with Australian Taxation Office records.

While there is history dating back to 1991 of these records being cross-matched to check for over-payment, the removal of the human from ‘the loop’ was key to the ramping up of the debt recovery program dramatically. From the discovery of 20,000 overpayment in 2015-16 to nearly 800,000 in 2016-17. In the face of mounting public criticism, decision makers allowed the political imperative to maximise revenues to outweigh the likelihood that it was demanding money from people unlawfully.

The program caused great hardship and anxiety to many low-income Australians who bore the burden of proof for establishing that they did not owe money and who were unable to understand how their debts had been calculated. The unlawful debt recovery scheme known as Robodebt is not only a "massive failure in public administration", as a Federal Court judge put it recently, but alarming evidence of the potential repercussions of linking datasets generated by different government departments.

The latest issue of the Australian Journal of Social Issues explores the use of linked data for research purposes and to carry out government functions such as child protection, allocation of resources, and debt recovery. We've put together articles that dive deep into the technical and governance challenges posed by data linkage, report important social policy insights derived from big data, and critically interrogate the  potential for big data to produce social harms.

In our editorial we look at the contradictory impulses towards openness and secrecy in the recent history of policy around big data. On 7 December 2015 the Australian Government released its Public Data Policy Statement, committing to “optimise the use and reuse of public data, to release non-sensitive data as open by default; and to collaborate with the private and research sectors to extend the value of public data for the benefit of the Australian public.”   

Significant reforms to the availability and use of public sector data have followed.  The Data Integration Partnership for Australia (DIPA), which is a three-year investment of $131 million, aims at “improving technical data infrastructure and data integration capabilities across the Australian Public Service.” A key element of the initiative is the Multi-Agency Data Integration Project (MADIP), which combines information on healthcare, education, income support payments, income tax and population demographics from the Census.

91 research projects had been initiated or undertaken using MADIP data.  Over 70% of these projects were undertaken by Australian Government departments, 8% by State Government Departments and the remainder by universities or the Grattan Institute. The findings from these projects are potentially of great interest to the Australian social policy community. However, they are very difficult to access. The Federal Department of Education, Skills and Employment is the only government department which publishes reports on all of their integrated data research.

There is a clear need for the public dissemination of the findings of the rest of the projects undertaken using MADIP. We believe that embracing transparency and maximising the public value gained from these research projects would help build the social licence for data linkage.

More developments are underway. The 2021 Federal Budget has committed $1.2 billion to a new Digital Economy Strategy which includes the development of a pilot program to develop data inventories for 20 per cent of Australian government agencies to facilitate better access for research purposes and further training in artificial intelligence.

The Australian social research community has an important and urgent role to play in uncovering the social costs and potential public benefits of collecting, storing, linking and analysing big data. Powerful actors can use their connections and economic resources to leverage a greater share of the benefits of big data, but it is important for social policy researchers to ask who is protecting the interests of Australia's least powerful people?

People experiencing hardship are disproportionately affected by the dilemmas created by innovative use of big data because social suffering leaves voluminous data in its wake. People living in poverty are more likely to be reliant on public provisioning and disproportionately visible in the records of criminal justice institutions. Their lives are monitored and scrutinised with greater intensity.

The Indigenous data sovereignty movement argues that Indigenous peoples are alienated from the governance of data about themselves and makes the case that exercising control over data is a crucial part of self-determination.  The rights and interests of Australia’s most disempowered people should be central to the discussion about the data which is generated by encounters between the state and citizens in moments of suffering and vulnerability. Their voices need to be amplified and heard.

Dr Katherine Curchin is a senior lecturer in social policy at ANU. Her research interests include social policy, Indigenous policy and public administration.

Associate Professor Ben Edwards is a Senior Fellow at the ANU Centre for Social Research and Methods. He is an expert in longitudinal studies of child and youth development, linkage of administrative data to surveys and longitudinal studies of disadvantaged groups such as refugees.

Moderator: Celia Green