Empty the tribunal from the top, not the bottom
In 2025-26, 64 per cent of NDIS cases before the Administrative Review Tribunal resulted in a changed outcome. A review body that alters two thirds of the decisions it examines is not a backlog to be cleared. It is an instrument measuring the quality of the decision being reviewed.
Steps this would change
- Contest a Commonwealth decision at the Administrative Review Tribunal — Wait for allocation and case management
- Person with disability from access request to first funded support — Application to the Administrative Review Tribunal
- Employer sponsors a skilled worker — Skills in Demand visa — Application to the Administrative Review Tribunal
Rationale
This is the most important finding in this database, and unlike most of them it required no modelling at all. It is two published numbers put next to each other.
Number one. As at 31 May 2026 the Administrative Review Tribunal held 130,983 applications, received more than 77,000 during the reporting year, and is running an annual clearance deficit of roughly 30,000 cases. Median time to finalise has risen from 45 weeks five years ago; 43 per cent of applications are finalised within twelve months. NDIS applicants commonly wait about a year.
Number two. In 2025-26, 64 per cent of NDIS cases decided by the Tribunal resulted in a changed outcome — down from 73 per cent in 2024-25.
The review process models at 524 days, of which
86 per cent is queue and scheduling. It is the purest backlog in this
database.
The obvious response to a backlog is capacity: more members, more registry staff, more sitting days. That would help, and it is being argued for, and it treats a 64 per cent change rate as a fact of nature.
It is not a fact of nature. Every one of those cases is a decision the Commonwealth made, defended for a year, and then lost or conceded. The applicant carried the wait. The agency carried the cost of defending. And at the end the answer was different from the one first given.
What this proposal actually asks for. Not that the tribunal be weakened, restricted, or bypassed — the opposite. That the change rate be treated as the primary performance measure of the original decision-maker rather than as a statistic about the tribunal:
- Publish the change rate by decision type, by decision-making team, and by the reason the outcome changed. New evidence, misapplied criteria, and unreasonable findings are three different failures with three different fixes, and they are currently indistinguishable in the published data.
- Feed tribunal reasoning back into first-instance guidance systematically. A decision overturned for the same reason a hundred times is a training and guidance failure, not a hundred unlucky cases.
- Publish internal review change rates too. Nobody publishes how often internal review alters an outcome by portfolio, which is exactly the number needed to know whether that mandatory 90 day step is filtering or tolling.
Who acts: the Commonwealth alone — the NDIA, Home Affairs, and the tribunal, none of which need anyone else's agreement.
Why the days saved figure is what it is. 450 days is the addressable portion of the review process. It is the wait avoided by an applicant whose decision was right the first time, not a saving the tribunal can deliver.
What is being traded away
This proposal is the one on this site most capable of being turned into something harmful, and the risk section matters more than usual.
"Reduce the overturn rate" has an ugly solution available. The change rate can be lowered by making better first decisions, or by making review harder — narrowing review rights, raising evidentiary thresholds, restricting what the tribunal may consider. The second is cheaper, faster, and would be reported as success against exactly the measure this proposal asks government to adopt. NDIS advocates have already raised the fear that tightening the scheme will worsen tribunal outcomes for participants. Any version of this reform that touches review rights is not this reform, and this site would oppose it.
A high change rate is not proof that first decisions are bad. The sample is selected: only contested decisions reach the tribunal, and the people who contest are disproportionately those with a strong case, an advocate, or the resources to persist. A 64 per cent change rate among contested matters is compatible with a very high accuracy rate overall. That is a real statistical objection and it should not be brushed aside — though it cuts both ways, since it also implies that people without an advocate are wearing wrong decisions in silence.
Some change is the system working. Circumstances alter, new evidence emerges, and a decision right on the material available can properly be changed on better material. That is what merits review is for, and a change rate of zero would be its own kind of alarm.
Measuring individual decision-makers on overturn rates produces defensive decision-making. An officer judged on being upheld will refuse less and approve more, or refuse in ways that are hard to appeal. This is why the proposal asks for publication by decision type and reason rather than by officer, and this site's standing rule applies here as everywhere: name the body, never the officer.
Precedent
Published overturn and change rates are used as a primary regulatory quality measure in other jurisdictions and in Australian courts' own appellate statistics. The data here already exists and is already published; what is missing is its use as a management signal rather than a workload one.
Sources
- Administrative Review Tribunal, published statistics and processing times: 130,983 applications on hand at 31 May 2026; annual clearance deficit of about 30,000; 43 per cent finalised within 12 months; median finalisation time risen from 45 weeks five years ago
- Administrative Review Tribunal: 64 per cent of NDIS cases resulted in a changed outcome in 2025-26, 73 per cent in 2024-25
- ABC News, 'Fears NDIS tribunal delays will worsen as federal government tightens scheme', 2 August 2026