Written by Yux Seow, Cultural Research and Development Lead
In the first essay we asked whose data it is and who sets the terms. That question is one half of a pair. The Maiam nayri Wingara Indigenous Data Sovereignty Collective describes two mechanisms that work together: the governance of data, which is about who decides the rules, and data for governance, which is about having the right data to make a decision in the first place. They are not separate functions. Settling who holds the data achieves very little if the data can't answer the question in front of you. This essay is about that second half - data for governance.
The data environment communities inherit
Most data about Aboriginal and Torres Strait Islander people was not built to support communities to decide anything. Palawa sociologist Professor Maggie Walter named this pattern BADDR data: blaming, aggregate, decontextualised, deficit and government-priority, and restricted in access. It's worth sitting with each one, because between them they explain why an Aboriginal organisation can be surrounded by data about itself and still walk into a meeting with nothing to bring.
Blaming data contrasts Indigenous against non-Indigenous, with the non-Indigenous figure standing in as the standard. Aggregate data reports at state or national level, which implies that communities across the continent are interchangeable or homogeneous. Decontextualised data isolates individuals and families from the social and structural realities they face. Restricted-access data is held behind the walls of the agencies that collected it.
Deficit and government-priority data is the one that matters most for decisions. The problem isn't only that the data is deficit-framed. It's that the data was built to serve institutional priorities. An institution decided what was worth measuring, and it decided based on their own questions. The result is a body of evidence that is genuinely useful, to somebody else's decisions.
Consider what this means in practice for managing Country. A dataset that reports vegetation condition averaged across a bioregion is real data, competently produced. It also cannot tell you whether the burn your rangers ran three winters ago on one stretch of a creek line did what you hoped. A state or territory water dataset can tell you a wetland holds water some percentage of years. But it can’t tell you whether the soak your old people relied on is still filling.
What a data-based decision needs
A data-based decision needs data that clears three specific hurdles, and the default data environment routinely fails all three.
It has to be about your Country, not a Country like yours. A regional average cannot answer a question about one soak, one fire scar, one stretch of river. Aggregation is not a neutral technical step; it’s authoritative in its destruction of the detail that local decisions turn on. Data that recognises cultural and geographic difference is what makes community-level planning possible.
It has to answer your organisation’s questions. If an Aboriginal organisation's priority is keeping young people connected to Country, and the available evidence measures school attendance and youth offending rates, then the organisation is being asked to plan for its own priority using instruments built for someone else's.
It has to arrive in time. Decisions have deadlines. A negotiation happens in a meeting. A funding round closes. A season for burning opens and shuts. Evidence that arrives after the decision has been made is not evidence, it's a record of what you could have claimed. This is the hurdle that data held elsewhere fails most reliably, because a request to an external agency runs on that agency's timeline, not your organisation’s.
What changes when data is held by the community
The shift is not that the community wins the negotiation. It’s that the community is in the negotiation on its own terms.
An organisation that holds its own monitoring can walk into a negotiation already knowing what's on Country, at what condition, and how that has changed. It can put a number on what it is being asked to give up. It can tell a funder what its programs achieved using measures it chose, rather than around indicators designed by someone else around problems it didn't nominate. It can say no to a proposal on evidentiary grounds instead of on principle alone.
None of that requires more data than these organisations already generate. It requires the data to be held, described, and reachable, which is the argument the first essay made about assets. Data treated as an asset enables practical capability.
There is a genuine cost to the alternative, and it is not only reputational. It's the survey commissioned over ground already surveyed, because nobody could find the earlier work. It's the negotiating position that weakens because a claim couldn't be contested inside the meeting. It's the submission that has to cite someone else's deficit framing of your own community, because that was the only evidence available.
What "need-to-have" actually means
Data governance is treated as a nice-to-have because of how it usually appears in an organisation: a project, a policy, a piece of software, something to get to once the core work is funded.
That framing has it backwards. The organisation's decisions are already being made on data. Somebody is already deciding what gets collected, how specific it is, and who can reach it. Data governance is already happening, it's just happening somewhere else, by default, in the settings of a platform and in the priorities of the institutions that funded the work.
Maiam nayri Wingara offers an important distinction in structural terms. In one model, Indigenous governance is built in: it operates at a high level in the organisational structure and it is the core business. In the other, it is added on: a limited structure, an advisory body, an afterthought. The paper is blunt that advisory groups, consultations, invitations to workshops, and opportunities to make submissions are not governance. They spend a community's time and intellect while deferring the decisions to someone else.
The need-to-have framing offers a rare opportunity for organisations to choose the data governance model and data for governance approach rather than inherit the default.
How local the data needs to be, the questions worth asking, and the people who hold the authority to decide them are different in every community. What is consistent is that an organisation which can answer its own questions negotiates differently from one that can't.
