The Arithmetic of Broken Promises: Closing the Gap

Closing the Gap at the Halfway Mark

Three out of nineteen. That is the sum total of what a decade of national agreement, signed by the Commonwealth and every state and territory in 2021 and carried since by successive governments, and by billions of dollars in program funding, has purchased for Aboriginal and Torres Strait Islander people by the halfway mark of the National Agreement on Closing the Gap. The Productivity Commission’s 2026 report, released this week, confirms what community organisations, health services and Aboriginal-controlled peak bodies have been saying in submission after submission for years: that the machinery of government, for all its talk of partnership and priority reform, remains largely unmoved by the targets it set for itself in 2021. Employment, land and sea rights, and legal rights and interests are the only three of fifteen assessed socio-economic targets tracking to be met by 2031. A fourth, preschool enrolment, has already been achieved. Everything else – the targets that speak most directly to whether a child lives, is safe, is educated, is free – is either limping forward off the pace required or going backwards altogether.

Productivity Commissioner Selwyn Button, a Gungarri man, put it plainly: governments will need to “redouble their efforts” for the promise of the Agreement to become reality. It is a diplomatic way of saying that the current effort has been insufficient, and that the insufficiency is not a mystery. The Commission’s own foreword to the report, co-signed with Chair Danielle Wood, states that the urgency to alter the trajectory of the targets not on track – and more importantly, those in decline – is “strikingly clear.” When your own auditor, using your own data, tells you in the plainest possible terms that you are past the halfway point of a ten-year agreement and travelling in the wrong direction on the targets that measure whether Aboriginal children are safe, alive and free, redoubling of effort is not a suggestion. It is a diagnosis of failure with a polite bedside manner.

At A Glance: The 19 Targets

The table covers thirteen of the fifteen targets the Commission assessed for trajectory this round; the remainder of the Agreement’s nineteen targets are set out in full in the report itself. The shape of the ledger is not in doubt: a small cluster of genuine wins, a larger cluster of targets moving too slowly to arrive on time, and a fourth, smaller cluster moving in reverse.

The Ledger Of Decline

Set aside for a moment the language of targets and indicators, because it has a way of anaesthetising what the numbers actually describe. Four measures are moving in the wrong direction: suicide rates, adult imprisonment, children in out-of-home care, and the proportion of children assessed as developmentally on track. Each of these is not an abstraction. Each is a family, a court date, a caseworker’s file, a funeral.

On imprisonment, the 2026 data shows the national rate for Aboriginal and Torres Strait Islander adults reached 2,500 per 100,000 in 2025 – 16.7 times the rate for non-Indigenous adults. In Western Australia the figure is 4,305 per 100,000. In the Northern Territory it is 4,166. The Commission does not attribute this rise to a single cause, and nor should this essay: remand populations have grown across the country for reasons that include court backlogs and changing charging practices as well as sentencing law. But the timing is hard to ignore. The rise has tracked closely alongside the stricter bail regimes that state and territory governments have introduced, often in the wake of high-profile crime incidents, in the years since the Agreement was signed – and Aboriginal and Torres Strait Islander people, who are already over-represented in contact with police and courts, are the population for whom a tightened bail test converts most readily into a prison bed. New South Wales, home to the largest Aboriginal and Torres Strait Islander population in the country, has seen both youth and adult incarceration worsen over the same period.

Out-of-home care tells a parallel story. Indigenous children are removed into out-of-home care at a rate of 52.2 per 1,000 – 11.5 times the rate for non-Indigenous children. Child protection authorities operate under enormous and genuine pressure of their own: statutory caseworkers carry caseloads well above recommended levels, kinship-care placements are chronically undersupplied relative to need, and every decision is made under the shadow of the opposite failure – a child left in danger. None of that is invented, and none of it should be waved away in an essay written from outside the system. But the Commission’s data still shows Aboriginal families being assessed as unsafe, and Aboriginal children being removed, at a rate that outstrips the rate for other families by more than an order of magnitude, a gap resourcing pressure alone cannot explain and that decades of inquiries have linked instead to how risk itself gets defined and by whom. Social Services Minister Tanya Plibersek was unusually direct about it this week, telling a First Nations justice conference in Sydney that the numbers are “terrible” and that the country is “going in the wrong direction.” She is right, and the candour is welcome, but candour without structural change is simply a more articulate form of the same failure.

Suicide has fallen among children and young people aged five to twenty-four since 2018 – a genuine and hard-won improvement, built substantially on the community-controlled and culturally grounded prevention programs that Aboriginal organisations have fought for years to have properly funded. But the rate remains highest among those aged twenty-five to thirty-four, and it has risen among all adults over twenty-five. The report does not draw a direct causal line between program design and this divergence, and neither can this essay from the data alone. What can fairly be said is that the age band showing improvement is also the band best served by community-designed, culturally specific programs, while the age bands still going backwards are those such services have historically reached less well – a pattern worth investigating rather than a proof already established.

“It is very concerning that we are still seeing outcomes worsening for early childhood development, children in out-of-home care, incarceration rates and suicide.” – Senator Malarndirri McCarthy, Minister for Indigenous Australians

A Preschool Paradox

The one unambiguous good-news story in the report deserves to be told honestly, including its limits. More than 95 per cent of Aboriginal and Torres Strait Islander children are now enrolled in preschool programs in the year before school, up from 61.3 per cent in 2016. Target 3 has been met years ahead of schedule. Coalition of Peaks lead convenor Donnella Mills was right to say that this shows what is possible “if governments took a more genuine and ambitious approach to working with our communities.” Community-led design was not the only factor at work – sustained funding commitments and existing capacity within the mainstream preschool system also mattered – but it is strong evidence, in a single data point, that the theory underpinning the entire Agreement – that outcomes shift when Aboriginal community control replaces top-down program design – works when governments allow it to.

But the Commission’s own report is careful, and honest, about what this achievement has not yet delivered. Preschool enrolment has “not translated into more equitable outcomes” for First Nations children more broadly. The proportion of Aboriginal and Torres Strait Islander children assessed as developmentally on track by the Australian Early Development Census remains lower than for other children, and that measure is one of the four still worsening. In other words: children are walking through the preschool door in record numbers, and the door alone is not enough. What happens before that door – in housing, in family income, in health, in the accumulated weight of intergenerational trauma and dispossession – still shapes whether a four-year-old arrives ready to thrive. The report puts it with a precision that governments would do well to absorb: “improving outcomes requires diverse, linked and complementary supports across early intervention, primary health care, community preschools and family support programs.” A single successful program, however well designed, cannot carry the weight of everything else that remains undone around it.

The Five In The Middle

Between the three targets on track and the four in outright decline sits a middle tier of five – life expectancy, Year 12 or equivalent attainment, tertiary qualifications, youth engagement in employment or education, and appropriately sized housing – improving, but not fast enough to meet the 2031 deadline. It is the quietly damning category, because it represents movement without arrival: a life expectancy target improving too slowly to close is one that, on current settings, will simply never close, drifting toward a horizon that recedes at the same pace it advances.

Housing sits uncomfortably here too, because it is upstream of much of the rest of the list. A family without secure, appropriately sized housing cannot easily sustain engagement with early childhood services or consistent school attendance, and is more likely to come into contact with child protection systems that too often read overcrowding as parental failure rather than as a housing shortage. A headline social program funded while regional housing supply goes untouched is building with one hand and undermining the foundation with the other.

What The Commission Actually Recommends

The Commission’s core finding is not really about the targets themselves – it is about the mechanism by which they move. Commissioner Button was explicit: “The best way to make progress is through governments acting on the four priority reforms – sharing decision-making, building the community-controlled sector, transforming government organisations and sharing access to local data,” and, critically, “where governments have started sharing power with communities, progress has followed.”

That is close to the entire thesis of the Agreement, and precisely the element most governments have been slowest to implement, because it costs them something real – not budget, but control. Every target on track or met – employment, land and sea rights, legal rights and interests, preschool enrolment – sits in domains where community control has had room to operate. Every target worsening – imprisonment, out-of-home care, suicide, early childhood development – sits substantially within systems still run by state and territory agencies on old assumptions, under tightened bail laws and unchanged child-protection thresholds.

The report also makes a structural point too rarely aired in public commentary: states and territories carry the practical weight of the Agreement without being formally bound by its targets. Justice and child protection – the two domains recording the sharpest declines – are overwhelmingly their responsibility. Every jurisdiction shows improvement across at least half its assessed targets, but the ACT, the Northern Territory and Tasmania recorded the highest number showing no improvement or outright decline. A national agreement whose most consequential levers sit with governments not directly accountable to its targets is a design flaw the Commission has now flagged clearly enough that it can no longer be treated as an oversight.

Two Responses, And The Space Between Them

The political reaction to the report has split along a fault line that will be familiar to anyone who has followed Closing the Gap reporting since 2020. Minister McCarthy called the preschool result “a significant milestone” and pointed to improvement in eight further targets, framing the report as evidence that “collective efforts can drive meaningful and lasting change” while also, to her credit, not shying away from describing the out-of-home care and incarceration data as terrible and worsening. It is a measured, incumbent’s reading of the data: real progress, real failure, sitting side by side, with a commitment to keep working within the existing framework.

Independent Senator Lidia Thorpe offered the opposite reading. “With no progress, Closing the Gap has now just become a sad reporting exercise on the racist state violence that continues to occur against First Peoples,” she said. “The problem is not a lack of information. The problem is a lack of political will for change.” National Commissioner for Aboriginal and Torres Strait Islander Children, Sue-Anne Hunter, went further still, suggesting on ABC that officials should be stood down for lack of progress: “If this was any other national agreement that was going backwards or just staying still, someone would actually lose their job.”

Both readings are defensible, and both are, in their own way, incomplete on their own. The incumbent’s reading risks using genuine wins – employment, preschool, land rights – as a shield against scrutiny of the domains where the state continues to actively cause harm through its own policy choices, particularly bail law and child protection thresholds that have hardened, not softened, since 2021. The abolitionist reading, for its part, risks discounting the material fact that some targets have moved, and moved because community control was allowed to operate – a fact that argues for extending the mechanism that worked, not for abandoning the Agreement altogether. Where the two readings meet, and where the Commission’s own data sits most comfortably, is this: the Agreement’s theory of change is sound: self-determination works where it is permitted; the Agreement’s implementation has been selective and half-hearted, permitting community control in the domains that were politically comfortable to relinquish while retaining full state control over the domains – justice, child protection – that are politically costly to release and that happen to be exactly where outcomes are worst.

The Languages Beneath The Ledger

Amid the grim arithmetic, the report carries one finding that speaks to something the target framework cannot easily quantify. For the first time since 2018, new data compiled by the Australian Institute of Aboriginal and Torres Strait Islander Studies, working with the Languages Policy Partnership and the community-controlled sector, has established that 177 Aboriginal and Torres Strait Islander languages are spoken or used across the country today. Commissioner Button was careful to note that the figure cannot yet be read as a trend, because it is the first survey of its kind co-designed with community to properly capture this dimension of culture and identity. But its significance runs deeper than the target framework was built to measure. Language survival, after more than two centuries of policies designed explicitly to extinguish it, is itself a form of resistance and continuity that no employment figure or imprisonment rate can capture. It sits outside the nineteen targets, and perhaps that is the point: the deepest measures of a people’s endurance are not always the ones a government agrees to be held accountable for.

What Redoubling Actually Requires

Commissioner Button’s phrase – “redouble their efforts” – will be repeated in press releases and parliamentary statements for months, stripped, in most of those repetitions, of the structural argument that sits underneath it. Redoubling effort does not mean more of the same programs at twice the funding. The report’s own findings do not support that reading. It means redoubling the specific mechanism the data shows actually works: shifting decision-making authority, funding and evaluation control to Aboriginal community-controlled organisations, in the domains – justice, child protection, health – where governments have so far been least willing to relinquish it, precisely because those are the domains recording the sharpest declines.

It means state and territory governments treating their child protection and bail settings as Closing the Gap questions, not merely as law-and-order questions to be resolved by the harshest available option and reported on separately, months later, as regrettable but unrelated data. It means a federal government willing to use the leverage available to it – National Cabinet, funding agreements, public reporting – to bring jurisdictions like the ACT, the Northern Territory and Tasmania, which the Commission itself has singled out for the highest number of stalled or worsening targets, into genuine alignment with the Agreement they signed. And it means resisting the temptation, visible in this week’s coverage already, to let one genuine success – preschool enrolment – stand in for a transformation that has, on the Commission’s own numbers, touched barely a fifth of the targets it was designed to move.

Halfway through a ten-year agreement, with five years remaining and four of nineteen targets moving backwards, the honest reckoning is not complicated to state, even if it remains politically difficult to act on. The Agreement works where it has been allowed to. It has failed, so far, chiefly in the places where the state has been unwilling to let it work – in the courtroom, in the caseworker’s file, in the housing waitlist, in the systems built long before 2021 and defended, still, as though they were neutral rather than causal. Selwyn Button’s report does not ask for sympathy or applause for the wins. It asks, in the driest and most careful language a statutory body can use, for governments to do the harder thing they agreed to do five years ago and have mostly avoided since: give up control, and watch what changes when they do.

Closing the Gap did not begin in 2021. Its earlier, cruder version ran from 2008, set and measured entirely by government, with Aboriginal and Torres Strait Islander people as the subjects of the data rather than the authors of the response – and it failed comprehensively; by 2020, most of those targets were off track too. The 2021 Agreement’s central innovation, won hard by the Coalition of Peaks, was to bind governments to the four priority reforms rather than to a scorecard alone: an admission that the problem was never the outcomes themselves, but who held the pen. This week’s report is the first serious test of whether that structural difference has been honoured. Sector by sector, the answer is: only where governments found it convenient to let go.

– – –

Bakchos is the founder of Blak and Black, an Australian media and advocacy platform established in 2010. Bakchos writes from the intersecting perspectives of Wiradjuri heritage, Jewish identity, and humanism.

© Bakchos, August 2026

BLAK AND BLACK  |  MEDIA AND ADVOCACY  |  EST. 2010