Privacy is sometimes wrongly seen as antithetical to information sharing. It is not. Privacy and information sharing are different sides of the same coin. Our Privacy Act could just have easily and just as accurately been called the Information Sharing Act. Rather than arbitrarily limit what information can be shared, with whom, and for what purposes, the Privacy Act provides a framework for safe information sharing.
The new opportunities that are presented by better computer processing and storage capacity, and large administrative datasets that were previously merely the byproducts of service delivery have seen “information sharing” develop as a field of inquiry in its own right. The appetites of politicians and public servants for data solutions to public policy challenges, and complex social problems is increasingly insatiable, and is fed by data scientists, and software and hardware vendors.
The great promise of so-called “Big Data” technologies is the more efficient allocation of resources, the micro segmentation of the population to better target goods and services, and to monitor and evaluate the efficacy of those allocations. So how do we make the allocations of resources to the myriad of competing big data projects? What are the necessary preconditions to unlocking the value of the data? Who needs to be part of the conversation?
Joint data sharing seminar
Last month, a number of academic and government institutions came together in Melbourne to discuss just these questions.
The event, entitled Data and information sharing: Opportunities and challenges in the UK, Australia and New Zealand, was a joint project between Newcastle and Monash universities, the Australia and New Zealand School of Government, the Centre for Knowledge Innovation, Technology and Enterprise (KITE), the Economic and Social Research Council (ESRC), and Britain’s Centre of Excellence for Information Sharing.
The diversity of agencies and faculties represented suggested a challenge for a discipline of “information sharing”. Is “information sharing” the exclusive domain of data scientists, of IT systems architects and software engineers, of lawyers or public policy specialists?
Rather than tackle the whole world of information sharing, the session settled on a theme of “localism” to discuss the role of data and information sharing at the local level, where municipal government, NGOs and communities can look for opportunities to use shared data.
Britain takes this stuff seriously and has invested in establishing a Centre of Excellence for Information Sharing. We heard from its director, Stephen Curtis, who told us that in identifying opportunities to use data to help families under stress, or to reduce the cost and impact of domestic violence in our communities, we’ve got to put those we are trying to help at the center of our thinking - or as he put it, “it’s about service users, not data users”. You can view his short video summary of the event here.
Sue Bateman, who is Deputy Director of Better Use of Data at the UK Cabinet Office agreed, saying that an important part of putting the user at the center of any data solution or tool, is obtaining social license for the activity. Key to that is transparency with the affected communities.
Both UK speakers identified barriers to effective information sharing, with “risk aversion” high on the list. However, they pointed out these risk assessments did not seem to take into account the risks of not sharing information - which in the context of providing coordinated social and health services would seem high.
For example, with the increased focus on victims of crime, especially children, it was now also recognised that there needed to be a parallel focus on the perpetrators as they tended to repeat their behaviours on new victims, if the original victims were removed. Therefore, for the system to work better, there needed to be better tracking and sharing of information about the perpetrators and their subsequent behaviours.
Putting the needs service user at the center informs the New Zealand approach to information sharing. For example, a scheme to help offenders avoid imprisonment in favour of more productive programmes, or one that assists a released prisoner’s reintegration into society might both make use of information sharing to see that they are attending their scheduled programmes and appointments.
But considerable thought needs to be given as to how the resulting information is used. Does a missed appointment mean the person should be taken back to prison without further inquiry, or is it simply a prompt to understand what that person needs, whether it is transport, child care, or other support?
Increasingly, information sharing is becoming a topic of academic inquiry. Prof Ian McLoughlin of Monash University used the biological model of the “data ecosystem” to illustrate some of the challenges in getting to a data utopia. These include the related concepts of the “paradox of embedded agency” and the “installed base”. Agencies have so much invested in the status quo that it is difficult to shift them to more collaborative and productive modes of working. How to breakthrough? The answer is incrementally - by harnessing mediators who engage in “boundary-spanning practices” to forge novel relationships.
Behind the intimidating jargon is a plea for systems which can recognise and foster local, coal-face solutions which organically generate support – an approach that is different from the ‘big bang’ centrally-mandated, top-down big fixes imposed by governments.
South Australian model
We also saw how the theoretical models can be turned into practical tools which effect positive outcomes for vulnerable children. Peter Worthington-Eyre, Chief Data Officer at the Office for Data Analytics in South Australia’s Department of the Premier and Cabinet, explained the state government has managed to meet the operational needs of those directly involved in supporting families as well as the aggregate data needs of researchers, policy makers and politicians.
His role is to implement the state’s Public Sector (Data Sharing) Act. Its purpose is to reset the default to data sharing, overriding other legislation. The Act incorporates the Five Safes Framework - safe projects, safe people, safe data, safe settings and safe output. It is a framework, designed originally by the UK Office of National Statistics, for helping make decisions about making effective use of data which is confidential or sensitive. You can watch this video about the Five Safes here.
Like many in his field, Mr Worthington-Eyre has found that people who want data to solve problems often ask the wrong question. He says people who don’t understand data and its limitations often come with a pre-conceived solution. Stewards of datasets need to ensure data users bring a healthy dose of scepticism about the quality of data.
Even a sophisticated and well-developed information sharing system such as that employed in South Australia strikes common challenges such as a lack of standardised definitions for particular terms. For example, truancy might be measured in different ways and mean different things to each of the agencies which contribute and then seek to use a combined dataset.
New Zealand experience
New Zealand was represented by the Government Chief Privacy Officer, Russell Cooke, and Doug Gorman of the Social Investment Agency. They echoed the experience of speakers from other jurisdictions that privacy or data protection rules are often wrongly identified as obstacles to information sharing. They reiterated the key to success are social license, and clarity and transparency about the objectives and methodologies of the information sharing.
We learned there were no shortages of opportunities, and benefits to be gained from sharing data gathered for different purposes. But the challenges of information sharing require multi-disciplinary attention and the support of the communities which the programmes are intended to serve.
Image credit: Pixabay image