Better equity data through better (more equitable) data governance
Dr Bret Stephenson, La Trobe University; Australian Centre for Student Equity and Success (ACSES) Equity Fellow
Not everyone agrees with the ambitious growth and equity targets set by the Universities Accord, but there is one thing nearly all can agree on: we need more and better equity data.
The Accord final report urged the Government to improve equity data collection methods and develop more ‘granular’ indicators of disadvantage. It is a recognition that our student equity data ecosystem – a thirty-year-old ‘framework’ of equity group definitions, metrics and data collection methods – needs to be expanded and updated for the digital age.
Following the Government’s recent release of two post-Accord implementation consultation papers on equity policy, we are quickly (re)learning how exceedingly difficult this task is.
Post-Accord equity policy implementation
While the Accord final report advocates for the adoption of additional equity groups in the long term, it advised that four, now familiar, ‘target’ equity groups should remain the focus of policy in the short term – these include people from low-SES backgrounds, regional and remote backgrounds, First Nations people, and people with disabilities. Therefore, the two consultation papers focus on these cohorts alone.
The first paper proposes a system of ‘managed demand-driven funding for equity students’. It would guarantee a fully-funded Commonwealth supported place (CSP) to all equity students who gain course admission, but curiously, not necessarily at their chosen university – this caveat would not apply to First Nations students. The second paperdetails a ‘needs-based funding’ system that would provide additional funding to universities based on each equity student enrollment. Both proposals have attracted strong criticism for their impracticality and their perpetuation of student deficit-framings.
The proposals raise significant questions about whether our current equity data are still suitable for today's policy needs. Both papers emphasise the importance of improving data to accurately determine equity group ‘eligibility’ and ensure appropriately targeted resource allocation. However, they also demand much more from our equity group concepts and the data supporting them than ever before – perhaps too much.
Issues with current student equity data and its collection
Our current equity data framework was not designed to inform resource allocation or determine individual student needs and disadvantages. Developed in the early 90s when data was expensive and unwieldy, its purpose was to provide high-level, albeit low-fidelity, monitoring of underrepresented groups using minimal data. In this regard, it met its brief brilliantly but intentionally traded data quality and accuracy for reduced data management costs. The problems this creates are well-known in relation to the low-SES and regional and remote group definitions. These are imprecise proxy measures based only on a student’s home postcode and Australian Bureau of Statistics (ABS) census data.
What is less recognised is that students are identified with these groups automatically and without notification. In data governance terms, this process lacks transparency. As a result, few students would be aware they have been categorised or ‘targeted’ for support as a ‘low-SES student’. Important ethical questions must be navigated if we are to change SES status from a ‘stealth category’, used largely for monitoring, to one overtly used to distribute benefits such as CSP.
There are also important issues to consider with ‘self-disclosed’ cohorts, such as First Nations and people with disabilities. At most universities, to be identified with these ‘equity groups’ – though many object to the term – students must actively ‘tick the box’ during enrollment and sometimes re-tick each year.
The issues related to disability data are particularly challenging, and it appears that a third government consultation paper will be released focussing solely on these matters. For instance, there are alarming inconsistencies in how these data are captured and reported across the sector, undermining their utility for the proposed policies and raising questions of ethical and equitable data collection.
Policymakers should also consider that ‘self-disclosed’ groups are frequently underreported in the data, as students may decline disclosure for several reasons. Threats of stigma and discrimination are significant factors, as are concerns over privacy and control of personal information. Still others will decline self-disclosure in the belief that they don’t have sufficient ‘proof’ to establish group membership.
Students will also sometimes change their status, both ticking and unticking, for various justified reasons during their studies. For example, a student may retract their disclosure due to privacy violations, shifting self-identity or because of the temporary nature of some disabilities. These behaviours suggest that ‘equity group’ identification can be contextual, dynamic and sometimes deeply shaped by power asymmetries and lack of trust. This has important implications for the policies under consideration by government and for the way in which universities collect, govern and utilise student equity data.
The path to better equity data
While a fundamental rethink of our equity data framework may be on the cards, there are immediate actions that universities can undertake. To achieve better equity data, universities must work to become better, more transparent and trustworthy, stewards of these data. This means centring equity concerns and student voices within data governance policies and processes. It also means clearly communicating to students the purpose of equity data collection and being transparent in how the data will be used and disclosed.
The need for excellent and equitable data and digital governance has become increasingly urgent. As we await the Government’s pending reforms to the Privacy Act 1988 (Cth) the higher education sector should also prepare for what the Accord final report terms ‘enhanced [equity] data collection’.
‘In addition to setting [equity] targets, Government must improve data collection to understand and address more granular indicators of disadvantage better’. (p. 117)
Along with ‘more granular’ we could add more sensitive, as the report calls for new data collections to allow for the identification and monitoring of formerly unrecognised student equity groups, including: first-in-family students, mature-aged students, care leavers, refugees, carers, certain language groups, and prisoners (p. 117).
While these data are undoubtedly necessary, the collection, management, and use of sensitive student equity data involve a complex interplay of risks, responsibilities and interests that must be carefully balanced. Finding this balance is as difficult as it is important, but equity in the digital age demands it.
Dr Bret Stephenson, ACSES Equity Fellow; Principal Advisor, Data and Ethics, La Trobe University
Bret’s ACSES Equity Fellowship, Centring equity in data and AI governance: Informing policy to empower practice, will produce a digital governance framework to support stakeholders – equity practitioners, universities and government – in the responsible collection, management and use of student equity data in a time of rapid digital transformation.