Submission 2079

Submission to the Senate Standing Committee on Community Affairs (Legislation Committee)

Inquiry into the National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Bill 2026

From: Lee Knights, Senior Process Engineer, Autistic adult, parent of two Autistic children, current NDIS participant, resident of South Guildford WA 6055 (federal electorate Hasluck; state electorate Midland).

Date: 1 June 2026 Synopsis

I am writing as a directly affected individual. I am an Autistic adult and current NDIS participant. I am the parent of two Autistic children. The Bill in its current form will land on my family in three distinct ways: directly on me as an existing adult participant; directly on my fourteen-year-old daughter as an existing 9-to-15 cohort participant facing reassessment under an unnamed instrument; and structurally on my sixteen-year-old son, who is Autistic but not currently a participant, by narrowing the eligibility pathway available to him in the 16-to-17 cohort for which no replacement supports program has been announced.

While my family’s experience is Autistic, the Bill’s eligibility, assessment and foundational supports provisions apply across all impairment categories and all ages. My submission addresses those broader effects on the whole disabled community alongside the autism- specific dimensions, because the Bill’s reach is whole-of-scheme and whole-of-cohort.

Beneath the Bill’s title and the Government’s public messaging, its central operative effect is straightforward, and the Government’s own modelling now confirms its scale. Departmental modelling tabled in the Senate on 27 May 2026 projects that 241,000 people who are participants before the new eligibility rules commence will be removed from the scheme by mid-2031, and that — once those who would otherwise have entered are counted — close to 350,000 fewer disabled Australians will be supported than would have been. That is the measure of this Bill: the exclusion of a very large number of disabled Australians from individualised support, in a package whose Budget measure books $37.8 billion in savings over four years, presented to the public as a response to fraud, rorts, and the scheme’s long-term security. On the Government’s own modelling, that exclusion is not an incidental consequence of the reform — it is the mechanism by which the savings are achieved. The integrity work the Government invokes is real, and it sits substantially in Schedule 2, the direction of which I support; the evidence does not connect it to the Schedule 1 access-tightening that produces the exclusion. Stripped of the framing, the Bill’s most significant effect is a budget saving delivered by removing disabled Australians

Submission 2079

from support. I ask the Committee to weigh the Bill on that effect, not on the framing around it.

I welcome the Committee’s inquiry. I submit this evidence in support of three things:

• Omit Schedule 1’s eligibility-tightening provisions from the Bill. The existing evidence base - the Disability Royal Commission, the National Autism Strategy 2025-31, the Independent NDIS Review (Bonyhady and Paul 2023), the Productivity Commission 2024 foundational supports paper, and the still-active Evidence Advisory Committee Third Round Consultation - does not support them.

• Defer Schedule 2 until after the Joint Standing Committee on the NDIS Integrity of the National Disability Insurance Scheme inquiry reports on 2 July 2026. The inquiry’s recommendations should then be integrated into the Schedule 2 design, and the amended Schedule 2 brought back for substantive community consultation under “nothing about us without us” principles.

• Omit Schedule 3 from the Bill, on grounds of Ministerial overreach in the new pricing decision-maker model and significant participant safety risks in the automation of administrative action under new s.59B to s.59D.

The three Schedules operate on different evidence bases and require separate procedural treatment. The clinical evidence base, the assessment instruments, and the replacement supports programs that Schedule 1 assumes do not exist, are not publicly named, are not funded, and are not demonstrably ready - and on the evidence in this submission Schedule 1 is incompatible with the statutory and treaty foundations on which the NDIS rests on four distinct grounds: the new s.3(1)(d) Objects clause amendment subordinating reasonable-and-necessary supports to financial sustainability; the new s.9B functional capacity definition that excludes environmental and personal circumstances, inverting the CRPD Article 1 social model that NDIS Act 2013 s.3(a) commits the scheme to give effect to; the wholesale removal of the s.31 Principles relating to plans at Schedule 1 Part 6, stripping the statutory framework for supported decision-making and community inclusion against CRPD Articles 12 and 19; and the new s.25A(2) codification of income-and-geographic regressivity against the Disability Discrimination Act 1992 and the National Agreement on Closing the Gap commitments.

I have prepared this submission with the assistance of accessibility tools, as the Committee invites participants to do where required. The substantive arguments, lived experience, factual claims, and primary-source verification are mine. A full disclosure on the use of those tools is at the close of this submission.

Section structure

This submission is structured in four parts. Sections 1 and 2 sit as introductory material - personal context and standing, what the Bill does, how it lands on my family, and the asks at a glance. Parts A, B and C then set out the case for treating each of the Bill’s three

Submission 2079

Schedules separately, mirroring the separability of their evidence bases and procedural treatment. Part D collects the cross-cutting evidence, the procedural concerns, and the recommendations to the Committee.

Personal context and standing - the family triple-take (this section)

What the Bill does, why it lands on my family, and the asks at a glance

Part A - Schedule 1: the case against the eligibility-tightening provisions

Schedule 1’s eligibility-tightening applies scheme-wide, across all ages. Section 5 addresses my fourteen-year-old daughter’s 9-to-15 cohort specifically. The permanence-and-treatment test in Section 3, and the structural under-subscription analysis in Section 6, apply to all ages - including my sixteen-year-old son’s 16-to-17 pathway and my own 18-plus pathway.

3. The clinical-evidence vacuum - the permanence-and-treatment test in new s.24(5),

s.25(1B) and s.25A

4. The diagnostic-language critique - “low to moderate support needs” against the

DSM-5, the Independent NDIS Review, the National Autism Strategy, and the peak clinical bodies

5. The 9-to-15 cohort and the plan-reassessment and renewal provisions - new s.48A

and s.50A

6. The structural under-subscription of the NDIS across the disabled community, and

the funding-below-cost provisions in new s.34A, s.32K(3C) and s.45C

7. Statutory inconsistency and discrimination - the new s.3(1)(d) Objects clause

amendment; new s.9B against the CRPD Article 1 social model; the wholesale removal of the s.31 Principles relating to plans at Schedule 1 Part 6 against CRPD Articles 12 and 19; and new s.25A(2) against the DDA, the CRPD, and the National Agreement on Foundational Supports 2026-31 commitment to align with Closing the Gap

8. Direct contradictions with Government-commissioned reviews on Schedule 1 -

the NDIS Review, the National Autism Strategy, the Disability Royal Commission, The Evidence Advisory Committee Third Round Consultation, the Productivity Commission, the House Committee Inquiry into the Thriving Kids initiative, and the substantiation of the Minister’s Royal Commission claim against the Explanatory Memorandum

Part B - Schedule 2: wait for the JSC NDIS Integrity inquiry

9. What Schedule 2 does well, and the integrity-rhetoric mismatch

10. Schedule 2 against the launch-era expert lineage - Productivity Commission 2011

and 2017, the academic critique on quasi-market failure, and the Quality and Safeguards Commission’s 2017-18 establishment and 2025 Australian National Audit Office audit

Submission 2079

The integration point — how Schedule 1’s new s.25A(2) actively makes Schedule 2’s thin-markets gap worse

Part C — Schedule 3: Ministerial overreach and participant safety

12. Schedule 3 Part 1 — Ministerial overreach in the new pricing decision-maker model and the funding-cap architecture under new s.45C and adjacent provisions

13. Schedule 3 Part 2 — automation of administrative action and evaluative determinations under new s.59B to s.59D, and the participant safety risk against the Robodebt Royal Commission findings

Part D — Cross-cutting evidence, procedural concerns, and asks

14. Economic evidence and impact assessment — the cost-benefit analysis lineage,

the cost-shift across every public-service system, the cost-impact on affected families with my family as a worked example, the missing Regulation Impact Statement, the BP1 3:1 spending-to-revenue ratio, and the gas-tax counterfactual

15. Procedural concerns and peak-body convergence — the procedural-sequencing critique (the Senate Legislation Committee reports sixteen days before the JSC NDIS Integrity inquiry reports on its own evidence base),

The Disability Representative Organisations Joint Statement of 14 May 2026, multi-party consensus on consultation inadequacy, the Butler office auto-acknowledgement, The Mosman Park case, and the cross-broadcaster editorial-architecture observation

16. Recommendations to the Committee

17. Conclusion and disclosure

18. Annexes — key primary sources and references

Submission 2079

Section 1 — Personal context and standing: the family triple-take

I am a Senior Process Engineer based in South Guildford, Western Australia, in the federal electorate of Hasluck and the state electorate of Midland. I am Autistic and late-diagnosed.

I am a current participant in the National Disability Insurance Scheme. I work in a chartered profession and have done so since well before my diagnosis. My professional life and my submission to the Committee sit on different sides of the same fact: that the experience of being Autistic in Australia is shaped by the supports — and the absence of supports — available across the lifespan.

I am the parent of two Autistic children. My daughter is fourteen years old. She is an NDIS participant. We recently moved her from mainstream schooling to online learning. The supports the scheme provides are what makes her continued education possible. My son is sixteen years old, also Autistic, and not an NDIS participant. He sits at the boundary of the Bill’s eligibility-tightening provisions, exposed to the absence of any announced replacement supports program for the 16-to-17 age band should his support needs escalate. My daughter, by contrast, is the family member most exposed to the Bill overall—her cohort sits in a triple gap (the new eligibility test applies, no assessment instrument has been publicly named for her age band, and no replacement program has been announced).

The Bill will land on my family in three distinct ways. I refer to this throughout the submission as the family triple-take.

First — directly on me, as an existing adult participant in the 18-plus cohort. The supports the scheme provides to me are not abstract entitlements. They are the supports that allow me to remain in paid work as a chartered process engineer. They are the supports that helped me remain in employment through a period of severe Autistic burnout that would otherwise, on my honest assessment, have ended my career. My support needs have not been static across the years I have been a participant. They have fluctuated with circumstances at work, at home, and with my health. The current scheme accommodates that fluctuation. The Bill narrows that accommodation in ways developed in Section 5.

The Bill’s new functional capacity assessment regime, scheduled to commence on 1 January 2028, will be applied to me on plan review. I will be tested against a new standard. I will be tested by a new assessment instrument. I will be tested against a new permanence definition. New s.24(5) provides—and new s.25(1B) imposes the identical test on the early intervention pathway—that “an impairment or impairments are not permanent, or likely to be permanent, unless: (a) the person has undertaken all appropriate treatment for the impairment or impairments (if any); and (b) any other treatment is unlikely to materially improve, reverse, or alleviate the impact of, the impairment or impairments; and (c) the impairment or impairments are likely to persist for the person’s lifetime.” This test is clinically incoherent for autism. There is no treatment that “improves, reverses, or alleviates” autism in the sense the test requires. Either reading of the test breaks coherence: a strict reading excludes all Autistic applicants because the imagined

Submission 2079

treatments do not exist; a vacuous reading admits all because there is nothing to exhaust. The gateway is being deployed as a tightening, so the intent is exclusionary. I develop this argument in Section 3.

Second — on my daughter, as an existing fourteen-year-old participant in the 9-to-15 cohort. The 9-to-15 cohort sits in a triple gap under the reform package. The new scheme- wide eligibility test applies. No replacement program has been announced (the Thriving Kids 0-to-8 cohort excludes her by age). No assessment instrument has been publicly named for her age band — the I-CAN instrument flagged in disability-sector commentary applies to ages sixteen and above; Thriving Kids has its own state-delivered methodology still being worked up; the 9-to-15 band has neither. She will be reassessed on plan review against an unnamed instrument and the same incoherent permanence test, with no announced safety net if she does not clear the gate. The supports she currently receives — the supports that are making her continued education possible — are the supports that would be withdrawn. I develop this argument in Section 5.

Third — structurally on my son, as a sixteen-year-old non-participant in the 16-to-17 cohort. He is Autistic but not currently an NDIS participant. The Bill does not change his support arrangements today — he has none from the NDIS — but it materially narrows the access pathway that would be available to him if his support needs were to escalate. The 16-to-17 cohort sits in a double gap: the new eligibility test applies, no replacement program has been announced for the band, and while the I-CAN instrument applies to ages sixteen and above (closing the methodology gap), the substantive program gap and the safety-net gap remain. If his needs were to escalate in the next two years to a point where applying for NDIS supports became necessary, he would be applying under the new regime, with no announced foundational supports program available as an alternative.

These three positions are not abstractions to me. They are the reality of daily family life. The arguments in the sections that follow are grounded in those three positions and in the broader evidence base I have drawn from the Government’s own published Budget papers, peer-reviewed clinical literature, peak-body submissions, and credentialed third-party economic analysis. Where the substantive arguments rely on figures, dates, or quotations, I have cited the primary sources and have independently verified each against the originals.

I make this submission in my personal capacity. The views expressed are mine and do not represent the views of my employer.

Submission 2079

Section 2 — What the Bill does and how it should be measured

This submission is structured in four parts. Part A makes the case against Schedule 1’s eligibility-tightening provisions, on the basis of the existing evidence base — the Disability Royal Commission, the National Autism Strategy 2025-31, the Independent NDIS Review, and the Evidence Advisory Committee Third Round Consultation. Part B argues that Schedule 2 should be held over until after the Joint Standing Committee on the NDIS Integrity of the National Disability Insurance Scheme inquiry reports on 2 July 2026, the inquiry’s recommendations are integrated into the Schedule 2 design, and the amended Schedule 2 is brought back for substantive community consultation.

Part C sets out the case against Schedule 3 on grounds of Ministerial overreach in the new pricing decision-maker model and the funding-cap architecture (Part 1) and significant participant safety concerns in the automation of administrative action and evaluative determinations (Part 2)—the latter against the findings of the 2023 Royal Commission into the Robodebt Scheme. Part D is the cross-cutting evidence, the procedural concerns, and the full list of recommendations to the Committee. The structure mirrors the three Schedule-by-Schedule procedural treatments I am asking the Committee to consider.

What a Bill drafted in response to the disability community would address

The Disability Royal Commission into Violence, Abuse, Neglect and Exploitation of People with Disability delivered its Final Report on 28 September 2023, after a four-year inquiry, with 222 recommendations across twelve volumes. The National Autism Strategy 2025-31, developed by the Government in consultation with the Autistic community, sets out commitments to close gaps in outcomes for Autistic Australians across health, education, employment, justice, and participation.

The Joint Standing Committee on the NDIS Integrity of the National Disability Insurance Scheme inquiry, also commissioned by this Government, is gathering the evidence base on the integrity questions Schedule 2 of this Bill purports to address, and reports on Thursday 2 July 2026. The Evidence Advisory Committee Third Round Consultation, the active institutional mechanism the Independent NDIS Review itself established under Action 23.2 for testing the evidence base on NDIS access decisions, closed for submissions on Sunday 29 March 2026 and has not announced a completion date for its findings.

There is a further evidence gap. The Government has commissioned three independent cost-benefit analyses of the NDIS over fifteen years — the 2011 and 2017 Productivity Commission inquiries and the 2023 Taylor Fry actuarial analysis prepared for the Independent NDIS Review. The Productivity Commission’s 2011 and 2017 reports both found, at the all-of-scheme level, that the NDIS delivers net economic benefit. Taylor Fry’s 2023 quantification measured net costs exceeding measured benefits at the all-of-scheme level under a deliberately constrained methodology that explicitly excluded major benefit categories (justice and child protection outcomes, post-school education, community benefits, fiscal multiplier effects); Taylor Fry caveated the measured benefit figure as “most likely an underestimate”, and at the cohort level found net benefits exceeding net costs for several cohorts including children-and-families. The three analyses do not,

Submission 2079

between them, support the proposition that the existing scheme delivers negative net economic benefit. No updated cost-benefit analysis has been published for the package of reforms this Bill enacts. The premise that NDIS spending is the right place to take $37.8 billion over four years from the federal Budget has no contemporary cost-benefit foundation.

A Bill drafted in response to the disability community and to the experts the Government has itself commissioned would be aimed at implementing the recommendations of the Disability Royal Commission, at delivering the National Autism Strategy’s outcomes commitments, at integrating the JSC NDIS Integrity inquiry’s findings and the EAC’s findings on the access evidence base, and at acting on the evidence an updated cost-benefit analysis would produce. That would be evidence of the Government listening. The Bill currently before the Committee is not that Bill. The Minister stated in his Second Reading Speech on Thursday 14 May 2026 that the Bill “takes up recommendations on measures to improve delivery of the NDIS that have been made to government by various independent reviews of the NDIS, including the Independent Review into the NDIS, in 2023, and the Royal Commission into Violence, Abuse, Neglect and Exploitation of People with Disability.” The Bill’s own Explanatory Memorandum names the Disability Royal Commission once and cites no specific recommendation of it by number or by content; the National Autism Strategy is not engaged with as a structural framework; the JSC NDIS Integrity inquiry has not yet reported; the EAC Third Round Consultation has not concluded; and no updated cost-benefit analysis has been commissioned. The Bill does not just fail to take up the Royal Commission’s recommendations — it operates against the direction of several of them, including Recommendation 6.34 on NDIS health navigators and the structural-reform recommendations on a Minister for Disability Inclusion and a dedicated department for disability equality and inclusion (developed in Section 8). The Bill is aimed at a different question than the question the Government’s own commissioned work has posed. I develop the Government-commissioned-review contradictions in Section 8, the economic and cost-benefit evidence in Section 14, and the procedural-sequencing critique in Section 15.

The Bill measured against the Minister’s four commitments

The Minister closed his Second Reading Speech on Thursday 14 May 2026 by naming four things the Government intends the Bill to do: “fighting fraud and stopping rorts; slowing rapid cost increases; making eligibility requirements clearer; and delivering quality services and supports to participants.” My substantive ask of the Committee is that the Parliament receive a Bill aimed at the work set out in the sub-section above — the Disability Royal Commission recommendations, the National Autism Strategy outcomes, the JSC Integrity inquiry findings when they report, the EAC Third Round findings when they conclude, and the evidence an updated cost-benefit analysis would produce. The Bill currently before the Committee is not aimed at that work; it targets the wrong things. Even measured against the four targets the Bill has set for itself, it fails. Two of the four are wrong as policy goals to begin with — slowing rapid cost increases conflates appropriate-cost growth for legitimately needed supports with rorts, and making eligibility requirements clearer

Submission 2079

assumes an unclarity the Bill does not demonstrate. On each of the four commitments in turn, the Bill falls short on its own terms — and on the second and third, the goal itself is wrong.

A note on terminology before I take each in turn. The Bill’s Schedule 3 Part 2 introduces, in new s.59B to s.59D, an architecture of automated administrative action — including the automation of evaluative determinations involving discretion, evaluative judgement, and state of mind — applied to NDIS participants whose engagement with bureaucratic processes is structurally affected by their disability. The 2023 Royal Commission into the Robodebt Scheme made specific findings on the participant-safety risks of exactly this architecture and identified external-accountability safeguards as necessary. The Bill’s safeguards in Schedule 3 Part 2 are internal-process safeguards (CEO oversight) rather than the external-accountability safeguards (statutory participant notification, automatic external review trigger, independent oversight) the Robodebt Royal Commission identified. I refer to the Bill’s Schedule 3 Part 2 architecture throughout this submission as RoboNDIS, in anticipation of the framing any post-failure investigation is likely to use on the Robodebt precedent. Section 13 develops the participant-safety case in full.

On fighting fraud and stopping rorts. The Government’s own Integrity of the National Disability Insurance Scheme inquiry, referred to the Joint Standing Committee on the NDIS by the Minister on 25 March 2026, is the evidence base on which a substantive integrity- and-fraud package should rest. That inquiry does not report until Thursday 2 July 2026 — sixteen days after this Committee is scheduled to report on the Bill on Tuesday 16 June 2026. Schedule 2 is the Bill’s integrity-and-fraud component, and it is being introduced before the evidence base for it has been completed, with the Senate’s review of it scheduled to conclude before the JSC’s findings are available. Schedule 2 also implements a partial subset of the existing Registration Taskforce 2024 Final Advice and is silent on the launch-era expert critique of thin markets, workforce capacity, and quasi- market failure that the Productivity Commission’s 2011 and 2017 reports — and a decade of subsequent academic research — have placed on the public record. A Bill that genuinely fought fraud and rorts would wait for its own integrity inquiry to report, integrate the findings into the design, and engage with the substantive expert critique that has been available for over a decade. Schedule 2’s current form does none of those things. I develop this in Part B.

On slowing rapid cost increases. The premise that “rapid cost increases” are themselves the problem requires examination. If a support is reasonable and necessary for a disabled Australian, the cost of that support is not a rort or a runaway expense — it is the scheme delivering what Parliament legislated for it to deliver. Cost is the right measure only where the scheme is paying more than it should for the supports it is providing. Where the scheme is paying the appropriate cost for legitimately needed supports, the slowing of that growth is, on its face, the slowing of the delivery of supports that are needed. The right target is value for money spent — efficiency, oversight, integrity, accurate pricing — not aggregate cost reduction in circumstances where the cost reflects accurately identified need. The Bill pursues the aggregate-cost target by reducing access (Schedule 1) and by

Submission 2079

enabling the Minister to set funding below cost (new s.34A, s.32K(3C) and s.45C). Neither mechanism distinguishes between rorts that a cost-reduction frame would legitimately remove and reasonable-and-necessary supports it would also remove. The fiscal asymmetry the package operates within — a $37.8 billion four-year NDIS reduction announced alongside revenue concessions that an industry-revenue measure could replace — is set out in Section 14. The Government has itself commissioned three cost-benefit analyses of the NDIS over fifteen years (PC 2011, PC 2017, Taylor Fry 2023); the three do not, between them, support the proposition that the existing scheme delivers negative net economic benefit. No updated cost-benefit analysis has been published for this Bill.

On making eligibility requirements clearer. The premise that the existing eligibility requirements are unclear requires evidence the Bill does not present. The current Act sets out the access requirements at s.21 to s.27 of the NDIS Act 2013, with the reasonable-and- necessary criteria at s.34. Those provisions have operated for over a decade and have been applied to the approximately 760,000 current participants and to the much larger volume of applications that preceded their current participation. The Explanatory Memorandum does not cite any participant survey, complaints data, or applicant-feedback evidence that the existing access requirements at s.21 to s.27 and the s.34 reasonable-and-necessary criteria are unclear in their operation; the reform rationales in the Explanatory Memorandum are about scheme cost, scheme growth, and alignment with the NDIS Review’s structural recommendations, not about clarification of provisions that are operating unclearly. No published review the Bill draws on — the Independent NDIS Review, the Productivity Commission 2024 paper, the Evidence Advisory Committee’s published work, the Disability Royal Commission — recommends the specific provisions Schedule 1 imposes on the basis that the existing provisions are insufficiently clear. The position that NDIS access is available to every Australian whose disability requires support is, in fact, clear; what has clouded it is the public messaging that has characterised participants as gaming the system, not any ambiguity in the access provisions. What the Bill does instead is impose a substantially tighter test through new s.24(5), s.25(1B) and s.25A — the permanence-and-treatment-exhaustion gateway — which has no foundation in the Disability Discrimination Act 1992 s.4 definition of disability, no foundation in the Convention on the Rights of Persons with Disabilities Article 1 definition, no clinical-evidence base in the published literature, no endorsement from any Australian medical college or peak clinical body, and is, for autism and for a range of other impairment categories, clinically incoherent on its face. RoboNDIS — the Schedule 3 Part 2 automation of evaluative determinations — does not make eligibility clearer either; it makes eligibility decisions automated, which is a different question and, on the Robodebt precedent, a more dangerous one. The Bill substitutes tightening for clarification and substitutes automation for transparency. I develop the clinical-evidence vacuum in Section 3, the eligibility-test critique in Sections 3 and 5, and the RoboNDIS architecture case in Section 13.

On delivering quality services and supports to participants. The Bill’s substantive provisions operate in the opposite direction on this commitment too. Schedule 1 reduces

Submission 2079

participation by approximately 160,000 existing participants by 2030 on the Scheme Actuary’s projection, and prevents the entry of approximately 140,000 more who would otherwise have joined - fewer people receiving supports is not the delivery of supports to participants. New s.34A, s.32K(3C) and s.45C authorise the Minister to set funding for groups of supports below the actual cost of those supports - providers paid below cost cannot deliver quality services indefinitely, and the participants relying on those service categories cannot expect supports to remain available. Schedule 1 Part 6 removes the s.31 Principles relating to plans wholesale, stripping the statutory framework that requires participant-centred, supported decision-making and community inclusion in the plan- making process - the architecture that gives the “supports to participants” commitment its statutory shape. Schedule 3 Part 1 concentrates pricing discretion in the Minister beyond the independent-monitoring architecture the Productivity Commission’s 2017 NDIS Costs report contemplated. Schedule 3 Part 2 - RoboNDIS - automates evaluative judgement in administrative action against a participant cohort whose engagement withureaucratic processes is structurally affected by their disability. Each of these provisions moves the scheme away from the fourth commitment, not toward it.

Measured against the Minister’s four commitments, the Bill operates in the opposite direction on each. Measured against the question a Bill drafted in response to the disability community would have addressed, the Bill does not engage with that question at all. The asks at the head of this submission are the asks that the evidence - the Government’s own commissioned work, and the Bill’s own structural choices - supports.

What the Bill Does

Schedule 1 — eligibility tightening across the whole scheme.

Schedule 1 introduces three structurally interconnected provisions that together narrow access to the NDIS from approximately 1 January 2028. New s.24(5) and new s.25(1B) impose an identical permanence-and-treatment test on both the disability requirements and the early intervention access pathways, providing that “an impairment or impairments are not permanent, or likely to be permanent, unless: (a) the person has undertaken all appropriate treatment for the impairment or impairments (if any); and (b) any other treatment is unlikely to materially improve, reverse, or alleviate the impact of, the impairment or impairments; and (c) the impairment or impairments are likely to persist for the person’s lifetime.” New s.25A defines “appropriate treatment” and includes at subsection 25A(2) the explicit statutory provision that treatment may remain “appropriate” for these purposes “regardless of whether the person’s individual circumstances restrict the person from accessing the treatment” - including, per a Note to the provision, financial circumstances and geographical location. New s.9B defines functional capacity as the person’s ability to undertake an activity “without assistance from other people, assistive technology or modifications” and “in a context that excludes, as far as possible, the impact of the person’s environmental and personal circumstances.” This is the medical model of disability. The framework excludes by definition the environmental and personal circumstances that CRPD Article 1 — which NDIS Act 2013 s.3(a) commits the scheme to

Submission 2079

give effect to — places at the centre of how disability is constituted. The full statutory and treaty inconsistency is developed in Section 7.

A scheme-wide gating to alternative supports is added in new s.25B, the Objects clause of the Act is amended at new s.3(1)(d) to subordinate reasonable-and-necessary supports to “financial sustainability”, and new ss.34A, 32K(3C) and 45C authorise the Minister to set funding for groups of supports below the actual cost of those supports. Schedule 1 Part 6 additionally removes the s.31 Principles relating to plans wholesale — the statutory framework governing supported decision-making and community inclusion in the plan-making process, which CRPD Articles 12 and 19 require — leaving only fragments of sub-sections (c), (ca) and (f) intact. The combined operative effect, on the NDIS Scheme Actuary’s projection cited in public reporting, is a reduction in scheme participation by approximately 160,000 existing participants by the end of the decade, and the prevention of the entry of approximately 140,000 more who would otherwise have joined.

Schedule 2 — provider-side and integrity reforms.

Schedule 2 introduces a substantial provider-side and integrity package: a narrowed NDIS provider definition in new s.10C, extension of the Regulatory Powers (Standard Provisions) Act 2014 monitoring and investigation regime to the National Disability Insurance Agency, new civil penalty provisions including parallel civil penalties to existing criminal offences, plan-management and support-coordination commissioning, and protection against participant overcharging. That work is broadly aligned with the 2024 NDIS Provider and Worker Registration Taskforce report. The three key concerns I develop in Part B are: that Schedule 2 is being introduced before the JSC NDIS Integrity of the National Disability Insurance Scheme inquiry reports on Thursday 2 July 2026 — sixteen days after this Committee is scheduled to report on the Bill — with the result that the Senate’s review of Schedule 2 will conclude before the integrity inquiry’s findings are available; that Schedule 2 implements a partial subset of the 2024 Registration Taskforce recommendations — omitting the universal-registration architecture from NDIS Review Recommendation 17, the worker registration scheme, the participant complaints process reform, and the four-tier framework the Taskforce’s Final Advice put forward — and is silent on the launch-era expert critique of thin markets, workforce capacity, and market stewardship that the Productivity Commission’s 2011 and 2017 reports placed on the public record; and that the NDIS Quality and Safeguards Commission — the body that exercises the substantive regulatory functions Schedule 2 expands — is publicly characterised as under-resourced.

Schedule 3 — pricing decision-maker and automation provisions.

Schedule 3 introduces two structurally distinct changes. Part 1 makes the Minister the pricing decision-maker for NDIS supports under new s.45C and adjacent provisions, with discretion to set maximum amounts payable, intensity ceilings, and worker-to-participant ratios that operate regardless of cost and regardless of the support needs identified through clinical and care-professional assessment. The Minister has no statutory requirement to hold or consult clinical or care qualifications in making those determinations. Part 2 — RoboNDIS — introduces statutory authority in new s.59B to s.59D for automation of administrative action, including evaluative determinations involving

Submission 2079

discretion, evaluative judgement, and state of mind. The pricing-decision-maker model concentrates discretion in the Minister beyond the independent-monitoring architecture the Productivity Commission’s 2017 NDIS Costs report contemplated. The automation provisions operate in proximity to the participant-safety risk territory the 2023 Royal Commission into the Robodebt Scheme made specific findings about. I develop the Ministerial overreach case at Section 12 and the participant safety case at Section 13. My substantive position on Schedule 3 is set out in Part C: the Schedule should be omitted from the Bill on either ground.

The definition of disability the Bill is operating against

A definition question runs through the Bill that the Explanatory Memorandum does not substantively address. The Statement of Compatibility with Human Rights at the end of the Explanatory Memorandum engages CRPD Article 4(3) on consultation but does not engage Article 1’s definition of disability against the Bill’s new s.9B functional capacity definition. The Explanatory Memorandum does not address the Disability Discrimination Act 1992 s.4 definition of disability against the Bill’s new s.9B either. Australia’s primary domestic disability statute is the Disability Discrimination Act 1992, which defines disability at s.4 by reference to total or partial loss of mental or physical functions, the total or partial loss of a part of the body, the presence in the body of organisms causing disease or illness, the malfunction, malformation or disfigurement of a part of the body, disorders or malfunctions resulting in learning difficulties, and disorders, illnesses or diseases that affect a person’s thought processes, perception of reality, emotions or judgement, or that result in disturbed behaviour. Australia is a State Party to the United Nations Convention on the Rights of Persons with Disabilities (CRPD), which defines disability at Article 1 as including persons “who have long-term physical, mental, intellectual or sensory impairments which in interaction with various barriers may hinder their full and effective participation in society on an equal basis with others.” NDIS Act 2013 s.3(a) commits the scheme to “give effect to Australia’s obligations under the [CRPD].”

Neither the DDA s.4 definition nor the CRPD Article 1 definition is contingent on whether the person has undertaken all available treatment, nor on whether further treatment might materially alleviate the impairment. The disability is not constituted by the unavailability of remedy. The Bill’s new s.24(5) and s.25(1B) introduce, into NDIS access alone, a test which has no foundation in either of the two statutory and treaty definitions of disability that govern the rest of the Australian disability framework. Schedule 1 is operating against a definition of disability that the Australian disability framework does not recognise.

The Bill is not silent on the definition of disability — it introduces a new one, for NDIS access alone, in the combined operation of the permanence-and-treatment test (new s.24(5), s.25(1B), s.25A) and the functional capacity definition (new s.9B). The new definition has no clinical evidence base — no Australian medical college, peak clinical body, or DSM editorial board has endorsed it (Section 3 develops the clinical-evidence vacuum). It is contrary to Australian statutory law (the Disability Discrimination Act 1992 s.4 definition) and contrary to international treaty law (the Convention on the Rights of

Submission 2079

Persons with Disabilities Article 1 definition, which NDIS Act 2013 s.3(a) commits the scheme to give effect to). Schedule 1 introduces a definition of disability that no clinical authority supports and no Australian law recognises.

The current Act already considers mainstream and foundational supports

A second observation runs through the structure of Schedule 1. The current NDIS Act 2013 at s.34 already requires, as part of the reasonable-and-necessary criteria for the funding of supports, that the CEO be satisfied that the support is “most appropriately funded or provided through the [NDIS] and not more appropriately funded or provided through other general systems of service delivery or support services”, taking into account family, carer, informal, mainstream and community supports already available. The reasonable-and- necessary test as it currently operates already accounts for whether a support could be more appropriately delivered outside the scheme. If the Government’s policy objective is for foundational supports outside the NDIS to carry more of the weight, the structural lever is the development of those foundational supports — not an amendment to the access gate, because the access gate already considers them.

The premise that Schedule 1’s eligibility-tightening is needed to direct people toward mainstream and foundational supports is structurally inconsistent with how the existing Act already operates. Until the Thriving Kids 0-to-8 program is operational from 1 October 2026, the unnamed adult and 9-to-17 foundational supports programs are designed and funded, and the $200 million Inclusive Communities Fund (Budget Paper No. 2 line 8937) is allocated and operational, there is nothing for the access gate to point participants toward. Once those programs are operational and demonstrably ready, the existing s.34 framework would direct participants toward them without a statutory tightening of the eligibility test at all. Schedule 1 is structurally redundant for the cost-shift purpose its Explanatory Memorandum advances.

What adequate support should look like — a four-step framework

I want to be concrete about what a functioning disability support system should provide, because the Bill’s structural choices are best understood against that baseline. My own experience of accessing the NDIS — and of the places where the system failed, and where the Bill would entrench those failures — maps directly onto a four-step framework. I name the steps explicitly because they provide the clearest account of what the Bill does and does not address.

Step 1 — Pre-diagnostic support, including access to diagnosis. Identification of a disability requires diagnosis. For most Autistic Australians, that diagnosis is not free: adult autism diagnosis costs approximately $2,500 out-of-pocket privately, against a twelve-month-plus public waitlist. The NDIS does not fund diagnostics. A person who cannot afford a private assessment and cannot wait twelve months is excluded from the evidence base required to apply for any support — NDIS or otherwise — before the access gate is even reached. The Bill does not address this foundational barrier. Diagnosis is also where many disabled Australians are most vulnerable. The cognitive and executive resources

Submission 2079

required to navigate the diagnostic process — finding a specialist, attending appointments, completing intake forms, sitting through assessments, and managing the emotional experience of receiving a diagnosis — are often the same resources that the emerging disability has already compromised. I came close to ending my career during a period of severe Autistic burnout, in the same period I was navigating the diagnostic and support- availability process; Step 2 below develops the burnout context in full. Disabled Australians need support during the diagnostic process, regardless of how that process is funded or delivered. The Bill provides no mechanism to address this either. The new permanence- an treatment test at new s.24(5) adds a further requirement at this step: the person must demonstrate not only that a disability exists, but that all appropriate treatments have been exhausted and further treatment is unlikely to alleviate the impairment. For autism, which is not “treated” in the sense the test requires, the practical effect is an additional layer of clinical incoherence laid on top of the existing diagnostic-access inequity.

Step 2 — Post-diagnostic support, including a functional capacity assessment and

identification of existing and missing supports. Following diagnosis, the system should provide a structured process to identify what the person can do, what assistance they need, what formal and informal supports are already in place, and what gaps remain. In the current NDIS, this is the role a functional capacity assessment (FCA) is designed to play. A participant may receive an initial plan with basic supports and funding for an FCA; The FCA then identifies the full range of supports the person actually needs, and a plan update is actioned to incorporate them. This two-pass process — initial plan, then FCA- informed update — is not an anomaly or an overreach. It is the designed mechanism for getting from a generic first assessment to an accurate individual-needs picture.

My own experience illustrates precisely where Step 2 fails. I received an initial plan as an Autistic adult. My plan included funding for a functional capacity assessment. I never completed that assessment — not because I did not need it, not because my support needs had been adequately identified — but because nobody explained to me what a functional capacity assessment was or why completing it mattered. I had been in severe Autistic burnout. Autistic burnout is not a temporary stress response; it is a period of significant functional reduction, often lasting months, during which the cognitive and executive resources required to navigate unfamiliar administrative processes are precisely the resources most depleted. I was barely managing daily life. I could afford, privately, some of the supports I needed during that period. Many people in my position cannot. I did not understand the FCA, I did not know what I was missing by not completing it, and no- one in the system had a statutory obligation to explain it to me. The result was that my plan never reflected my actual support needs.

There is a further dimension to that failure that the administrative record does not capture. For Autistic people, the anxiety of new things, new places, new people is not mere discomfort — it is a functional barrier in ways the neurotypical-designed assessment pathway does not account for. Knowing what a functional capacity assessment is helps. Knowing what it will involve procedurally — where it will take place, with whom, in what order, for how long — is what actually makes it accessible. Without that advance

Submission 2079

procedural clarity, the barrier is not motivational. It is disability-specific. The cognitive work of managing uncertainty about an unfamiliar process, with unfamiliar people in an unfamiliar setting, is not a background burden; for an Autistic person in burnout, with executive resources already depleted, it is the work that forecloses the option entirely. I received none of that procedural detail. The assessment pathway was designed around an assumption of navigability that the disability it was designed to assess had already compromised. This is a structural mismatch between the design of the FCA pathway and the disability cohort the pathway is meant to serve — not a general anxiety failure, and not a personal shortcoming.

There is a further barrier at Step 2 that does not appear in any administrative data. The language and rhetoric of government officials — Ministers, departmental spokespeople, the Government’s sustained public messaging characterising NDIS participants as gaming the system, the “uncontrollable growth,” the “rorts,” the scheme being used by people it was never intended for — and the reproduction and amplification of that language by media organisations without sufficient counter-framing, together generate a stigmatising public narrative that lands inside participants, not only in media coverage. Both layers are deplorable in their own right and require explicit naming. I question, regularly and genuinely, whether I deserve the supports I receive. Not from a rational assessment of my needs, but from the accumulated weight of hearing my participation in a statutory program described as a rort by Ministers, by commentators, by journalists, and by colleagues who express their disdain for NDIS participants without knowing that I am one. I am a chartered professional in well-paid employment. That employment is, in part, a consequence of having the supports the scheme provides. It is read by others — including well-educated and thoughtful people — as evidence that I do not need those supports. The contempt operates at the intersection of class and disability: visible professional success is taken as disqualification rather than as evidence that the scheme is working as intended. The guilt is not background noise. It is daily, and it suppresses my legitimate use of the scheme. I do not complete FCAs. I do not request plan reviews. I do not ask for what I actually need — not because those things are wrong, but because the rhetoric has taught me to read my own needs as excessive. I am not an outlier in this. When internalised stigma suppresses legitimate scheme use at scale, the cost models read low FCA completion and low plan-review requests as evidence that needs are adequately met. They are evidence that the rhetoric has succeeded in suppressing the signal. The Bill does not address this suppression mechanism. The permanence-and-treatment test and the “significant and ongoing change” threshold at new s.48A compound it.

The Disability Royal Commission’s Final Report Volume 6, Enabling Autonomy and Access, found that financial barriers are a systemic inequity for lower-income NDIS applicants, and Recommendation 6.34 specifically recommended health navigators to assist people with disability to access and navigate the NDIS — including the functional capacity assessment and plan-development process. [§90] That is precisely the gap my experience illustrates. No health-navigator program responding to Recommendation 6.34 is funded in the Bill or in Budget Paper No. 2. The $200 million Inclusive Communities Fund (line 8937) is group-based social participation, not access navigation.

Submission 2079

The period between initial application and plan approval is also the period of greatest vulnerability. A person has disclosed a disability and asked for help; they are receiving nothing while the system assesses them. I was in severe Autistic burnout during this waiting period. The clinical consequences of Autistic burnout — impaired executive function, communication difficulties, heightened sensory sensitivity, inability to engage with complex administrative processes — are at their most acute during the months a person is trying to navigate the most administratively complex process the disability support system provides. People who can afford private supports during this period receive them. People who cannot are exposed to that vulnerability without any safety net. The Bill provides no mechanism to address this waiting-period gap.

The Bill’s new s.48A makes the FCA-pathway failure the governing framework rather than remediating it. Under the Bill’s new plan-reassessment provisions, a request for an unscheduled plan change requires the participant to demonstrate that their support needs have significantly changed. A plan update following a functional capacity assessment is not a change in support needs — it is a better identification of existing needs that were never accurately measured in the first place. The two-pass process — initial plan, then FCA-informed update — IS the system working as designed. New s.48A treats the completion of the designed-in Step 2 assessment pathway as an unscheduled overreach rather than as the system working. The Minister’s public rhetoric characterising unscheduled plan updates as evidence of excessive demand on the scheme is, on the structure of the Act, misleading: it identifies the completion of Step 2 as the problem, in circumstances where the two-pass process is the system Parliament designed. The rhetoric reflects either a misrepresentation of how the system actually works or a lack of understanding of it, and the Bill’s s.48A entrenches that misframing as policy. I develop this argument in Section 5.

Step 3 — Access to identified supports. Once support needs are accurately identified, The person should be able to access them. Under the current scheme, the s.34 reasonable-and-necessary test already requires the CEO to be satisfied that a support is most appropriately funded through the NDIS rather than through other general systems, taking into account what is already available. The test as currently designed already considers foundational and mainstream alternatives. The Bill’s eligibility-tightening at Steps 1 and 2 does not improve the matching of people to supports at Step 3 — it removes people from the pipeline before Step 3 is reached, in circumstances where the foundational supports they are meant to be directed toward do not yet exist for most age bands and most support categories. This is the structural redundancy argument I develop in the subsection above.

A further structural problem at Step 3, often invisible in the policy framing the Bill rests on, is supply-side market failure. Even where participants have funding for identified supports, services are frequently unavailable. Provider thin markets — particularly for specialist autism supports and in regional and remote areas — produce long waitlists for participants holding plans they cannot fully use. The thin-market problem is worsened by NDIS payment limits often set below market value, which keeps service-worker pay low, drives

Submission 2079

Workforce attrition, and renders providers non-viable. A substantial portion of the “NDIS underspend” the Government cites as evidence of scheme excess is, in fact, evidence of supply-side market failure preventing participants from using their plans — not evidence that participants do not need the supports they have been funded for. Schedule 1’s new s.34A, s.32K(3C) and s.45C authorise the Minister to set funding for groups of supports below the actual cost of those supports, which structurally worsens the thin-market problem the Government’s own framing already misreads. Schedule 2’s silence on the launch-era expert critique of thin markets, workforce capacity, and market stewardship leaves the underlying problem unaddressed. I develop the under-subscription and below-cost-funding case at Section 6, and the launch-era critique at Sections 10 and 11.

Step 4 — Access to review for changes in circumstances. Support needs change. This is true for all disabled Australians, and it is true in distinctive ways for Autistic Australians. Autism is not a stable-state condition in the sense that the Bill’s s.48A “significant and ongoing change” framework assumes. Autistic support needs are often episodic and context-dependent: Autistic burnout, changes in the masking and compensatory strategies that make a person appear more functional than they are, changes in life circumstances, and changes in the external supports that sit alongside NDIS plans can all produce genuine changes in what a person needs. The “significant and ongoing” threshold in new s.48A, and its additional requirement at s.48A(2)(b) that any functional capacity alteration involve a “substantial reduction in daily activities”, are calibrated to a deterioration model — a medical model in which needs increase in a linear, observable, sustained trajectory. That model is structurally ill-fitted to autism and to psychosocial disability, where needs are variable, episodic, and not predicted by a prior clinical trajectory. I develop this in Section 5.

Step 4 must also accommodate changes in supports outside the NDIS. The design premise of the Bill — that foundational supports outside the scheme will carry more of the weight — implies that the scheme’s obligations to participants include an accounting for what happens when those external supports change. New s.48A(3) lists the categories of environmental change that can trigger a reassessment: living circumstances, education, work, informal support network. Changes to formal external support programs — including Thriving Kids, foundational supports programs that are withdrawn or underfunded, or school-based disability support services — are not in that list. The Bill assumes foundational supports will be adequate and durable; it provides no review mechanism if they are not.

Step 4 must also accommodate the reverse case: support needs that lessen. If the Government delivers genuine foundational supports outside the NDIS — better-funded Thriving Kids, properly-scoped 9-to-17 and adult foundational supports programs, restored disability funding to schools — in-scheme NDIS demand should fall, because participants would have a real alternative to direct some of their needs toward. That is the policy direction the Government’s own foundational-supports rhetoric claims to be moving in. The Bill’s s.48A architecture, calibrated only to deterioration, contains no mechanism for that lessening case. If foundational supports are delivered well enough to reduce a

Submission 2079

participant’s NDIS-met need, the Bill provides no plan-update pathway that recognises the reduction; if the foundational support is subsequently withdrawn, no pathway to restore the NDS plan to the prior level either. The Bill cannot achieve the policy direction the Government claims to want while the mechanisms it puts in place assume only escalation.

The equity dimension across all four steps. The failure points in this four-step framework do not distribute evenly across the disabled population. Peer-reviewed research published in the Medical Journal of Australia in 2025 — Disney et al., analysing 705,594 NDIS applications between 2016 and 2022 — found measurably lower NDIS eligibility rates for applicants in lower socioeconomic areas. [§89] This quantitative finding confirms what the structure of the system predicts: every step in the framework is harder to complete on lower income and in regional and remote settings. Step 1 diagnosis is more expensive. Step 2 FCA is less likely to be explained and supported. The waiting period is more harmful when private supports are unaffordable. Step 4 review is harder to initiate without administrative capacity and informal support networks. The Bill’s new s.25A(2), which provides that treatment is “appropriate” for the purposes of the permanence-and-treatment test regardless of whether the person can afford it or access it geographically, writes the Disney et al. and Disability Royal Commission finding into statute rather than remedying it. I develop the discrimination and Closing the Gap dimensions of this below.

Discrimination, equity, and Closing the Gap

A third observation is on the equity impact. The s.25A(2) provision that treatment remains “appropriate” regardless of the person’s financial or geographic circumstances codifies, in statute, an income-and-geographic regressivity. Adult autism diagnosis in Australia currently costs approximately $2,500 out-of-pocket privately against a twelve-month-plus public waitlist, with no NDIS funding for diagnostics. Treatment exhaustion can mean tens of thousands of dollars in private therapy and medication trials. The Disability Discrimination Act 1992 at s.6 prohibits indirect discrimination — a requirement that “is not reasonable having regard to the circumstances of the case” and that has, or is likely to have, the effect of disadvantaging persons of a particular impairment category. The CRPD at Article 5 obligates State Parties to prohibit all discrimination on the basis of disability. Aboriginal and Torres Strait Islander people are over-represented in lower-income and remote populations, and therefore disproportionately face the cost and geographic barriers s.25A(2) codifies; the National Agreement on Foundational Supports 2026-31, signed by all eight jurisdictions in early 2026, commits the Commonwealth and the States at clauses 24 and 25 to aligning Foundational Supports design and implementation with the four Priority Reforms of the National Agreement on Closing the Gap. The Bill is operating in opposite directions across these instruments.

Submission 2079

Section 3 — The clinical-evidence vacuum: the permanence-and-treatment test in new s.24(5), s.25(1B) and s.25A

The single most consequential operative provision in Schedule 1 is the new permanence- and-treatment test the Bill writes into the disability requirements (new s.24(5)) and the early intervention pathway (new s.25(1B)), supported by the new definition of “appropriate Treatment” in new s.25A. This section concerns the clinical-evidence basis for that test. My contention is that the clinical-evidence basis does not exist; that the test is structurally incoherent for the impairment cohorts to which it will be applied; and that the test cannot survive an evidentiary scrutiny the Bill has not been subjected to.

What the test requires

The Bill’s new s.24(5) provides — and new s.25(1B) imposes the identical requirement on the early intervention pathway — that “an impairment or impairments are not permanent, or likely to be permanent, unless:

 (a) the person has undertaken all appropriate treatment for the impairment or
impairments (if any); and
 (b) any other treatment is unlikely to materially improve, reverse, or alleviate the
impact of, the impairment or impairments; and
 (c) the impairment or impairments are likely to persist for the person's lifetime.”

The three limbs are conjunctive. All three must be satisfied for permanence to be made out.

New s.25A defines “appropriate Treatment” for the purposes of paragraphs 24(5)(a) and 25(1B)(a). Treatment is “appropriate” if it is “(a) evidence-based; and (b) can reliably be expected to materially improve, reverse, or alleviate the impact of, the impairment or impairments; and (c) is regularly undertaken or performed in Australia.” Sub-section 25A(2) provides that “treatment may be appropriate treatment for a person’s impairment or impairments regardless of whether the person’s individual circumstances restrict the person from accessing the treatment” — with a Note that “a person’s individual circumstances include the person’s financial circumstances and geographical location.” Sub-section 25A(3)(a) provides a narrow carve-out for “appropriate medical treatment the person cannot undertake for medical reasons”. Sub-section 25A(4) gives the Minister the power to determine, in delegated rules, further “circumstances” in which a person is taken to have undertaken all appropriate treatment; sub-section 25A(5) confirms that those rules “may make different provision in relation to (a) different classes of participants; and (b) different impairments or classes of impairments.”

The structure as drafted is therefore: a hard treatment-exhaustion gate in primary legislation, qualified only by a narrow medical-only carve-out, with the rest of the carve- out logic moved to delegated rule-making in which the Minister has the discretion to vary the gate by participant class and impairment class.

Submission 2079

The clinical-evidence problem

The test the Bill is constructing operates as an effective new statutory construct for what counts as a permanent impairment for NDIS access purposes: an impairment is treated as permanent only where the s.24(5) and s.25(1B) conjunctive limbs are satisfied, including the treatment-exhaustion limb. That construct is structurally inconsistent with the Australian disability framework’s foundational instruments — the Disability Discrimination Act 1992, the United Nations Convention on the Rights of Persons with Disabilities to which Australia is a State Party, and the National Agreement on Closing the Gap. Section 7 of this submission develops that statutory-inconsistency case in full. This section concerns the clinical-evidence dimension that operates alongside the statutory inconsistency.

The Bill applies this test to all NDIS access decisions across all impairment categories. To survive clinical scrutiny, the s.25A(1)(b) definition would need to be capable of identifying a body of treatment, for each impairment category to which the test will apply, that “can reliably be expected to materially improve, reverse, or alleviate the impact of” that impairment.

For a large share of the impairment categories the NDIS serves, such a body of treatment does not exist as a coherent clinical paradigm.

Autism.

The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) defines autism spectrum disorder as a neurodevelopmental condition characterised by persistent differences in social communication and social interaction across multiple contexts, and by restricted, repetitive patterns of behaviour, interests or activities. The diagnostic criteria explicitly require symptoms to be present in the early developmental period. The DSM-5 does not characterise autism as a condition for which treatment “remedies” or “alleviates” the underlying condition. Early intervention exists and supports Autistic children to develop adaptive strategies and skills; it does not make children non-autistic, and it is not described in the peer-reviewed clinical literature as remedying or alleviating an underlying impairment in the sense s.25A(1)(b) requires. There is no clinical “treatment” of autism that satisfies the test, because the clinical paradigm for autism is not a remedy-or-cure paradigm.

This is not a fringe view. No Australian medical college, peak autism body, clinical society, or peer-reviewed clinical-guidelines publication has publicly endorsed the test as proposed. The 2023 Independent NDIS Review, cited by the Minister in his Second Reading Speech and in the Explanatory Memorandum as the recommendation source, was led by Professor Bruce Bonyhady (an economist, founding chair of the National Disability Insurance Agency) and Lisa Paul AO PSM (a former senior public servant). The Review’s panel composition is in the public record. Neither lead was a clinician, and the Review did not commission a clinical-evidence study underwriting the specific test the Bill now imposes.

Either reading of the test breaks coherence. A strict reading excludes Autistic applicants by definition, because the treatments s.25A(1) imagines do not exist as a remedy-or-cure

Submission 2079

paradigm for autism. A vacuous reading admits Autistic applicants without any actual gating effect, because there is no “appropriate treatment” under s.25A(1) to exhaust. The Bill is being deployed as a tightening — the Government’s own cost-savings projection of 160,000 existing participants leaving the scheme and 140,000 prospective participants prevented from entering it depends on the gate operating exclusionarily — so the strict reading is the operative reading. Autistic applicants are excluded by the structure of the test, regardless of the clinical merits of their individual case.

Intellectual disability. Intellectual disability is defined in the DSM-5 by deficits in intellectual functions and in adaptive functioning, with onset during the developmental period. As with autism, there is no clinical paradigm in which “treatment” of intellectual disability “materially improves, reverses, or alleviates the impact of” the underlying impairment in the sense s.25A(1)(b) requires. Habilitative supports exist; they enable people with intellectual disability to participate in education, work and community life. They do not constitute treatment of an underlying remedial condition. The test is structurally incoherent for intellectual disability for the same reasons it is incoherent for autism.

Down syndrome. Down syndrome is a genetic condition. There is no clinical treatment that remedies, reverses, or alleviates the chromosomal condition. Associated health conditions (cardiac, gastrointestinal, thyroid, others) are routinely managed clinically, but the underlying condition is not the subject of a remedy paradigm. The test is incoherent for the cohort.

Cerebral palsy. Cerebral palsy is a group of permanent disorders of movement and posture caused by non-progressive disturbances to the developing brain. The clinical literature describes management, therapy, and surgical intervention for specific functional impacts; it does not describe a treatment paradigm under which cerebral palsy itself is reversed or alleviated. The same structural problem applies.

Progressive neurodegenerative conditions. Motor neurone disease (MND) is the public- comprehensible canonical example. MND is a progressive neurodegenerative condition for which the operative clinical paradigm is symptom management, multidisciplinary supportive care, and progression-slowing — not remedy, reversal, or alleviation of the underlying impairment. The same structural problem applies to other progressive neurodegenerative conditions and to a range of acquired conditions for which there is no body of treatment that “can reliably be expected to materially improve, reverse, or alleviate the impact of” the impairment in the sense s.25A(1)(b) demands. The Bill’s test does not engage with the distinction between conditions for which a remedy paradigm exists and conditions for which it does not.

Sensory impairments (vision, hearing). Severe and profound sensory impairments are, in many cases, conditions for which medical interventions are not curative. Assistive technologies and supports change the functional impact; they do not reverse the underlying impairment. The test is therefore either narrowly applicable (to a sub-set of treatable sensory conditions) or vacuous.

Submission 2079

The pattern is consistent across the largest cohorts on the NDIS. The s.25A(1) definition of “appropriate treatment” is operating in a paradigm that does not apply to the impairments it will most consequentially gate.

This is not abstract for me. I am a late-diagnosed Autistic adult and a current participant in the scheme; the Bill’s new functional capacity regime will be applied to me at plan review after 1 January 2028. There is no body of treatment for me to exhaust under s.25A(1)(b) — the clinical paradigm for autism is not a remedy paradigm — so the test will be applied to me as a structural exclusion gate, not as an individual-evidence assessment I can address by providing evidence on my own case. My fourteen-year-old daughter, whose access is the subject of Section 5, will be tested under the same provisions; my sixteen-year-old son, were his support needs to escalate to the point where applying became necessary, would be tested under the Bill’s amendment to the s.25 early intervention pathway that imports the same test. The cost of the test’s incoherence is not borne by the test. It is borne by the cohort the test is applied to.

The Royal Commission claim the Explanatory Memorandum does not substantiate

The Minister told the House on Thursday 14 May 2026 that the Bill “takes up recommendations on measures to improve delivery of the NDIS that have been made to government by various independent reviews of the NDIS, including the Independent Review into the NDIS, in 2023, and the Royal Commission into Violence, Abuse, Neglect and Exploitation of People with Disability.”

The Bill’s own Explanatory Memorandum, at its Consultation section, cites four specific NDIS Review actions (Action 3.1, Action 3.9, Action 2.15 and Action 11.3) and one NDIS Review recommendation (Recommendation 17), plus the 2024 NDIS Provider and Worker Registration Taskforce report. It also cites the Disability Royal Commission final report (2023), including Recommendations 9.4, 10.2, 10.3 and 10.4, as part of the consultation and research basis for the plan management and support coordination commissioning provisions of Schedule 2 (Explanatory Memorandum section 5.4). It does not cite any Royal Commission recommendation by number or by content as the implementation basis for any Schedule 1 provision, including the new permanence-and-treatment test.

The Disability Royal Commission delivered 222 recommendations across twelve volumes in September 2023. The Joint Australian, State and Territory Response was published on 31 July 2024. The Bill is being introduced into Parliament with a Ministerial Second Reading click that one of those review and inquiry bodies is part of its evidentiary foundation across the package. The Bill’s own Explanatory Memorandum substantiates that claim only in respect of four of the 222 Royal Commission recommendations, all of which are attached to Schedule 2 Part 6 plan management and support coordination commissioning provisions. The Explanatory Memorandum does not identify any Royal Commission recommendation as the source of any Schedule 1 provision. The Committee is entitled to require the Ministerial claim to be substantiated against the Schedule 1 provisions the

Submission 2079

Government’s public framing relies on the Royal Commission to legitimise, and to consider the implications of finding that it cannot be.

The s.25 early intervention pathway the Bill collapses

The current NDIS Act 2013 contains two distinct access pathways. Section 24 (disability requirements) requires permanence; section 25 (early intervention) does not. The s.25 pathway is the route by which children with developmental delay — for whom an impairment may not yet have been clinically established as permanent, and for whom the underlying clinical question is whether early support shifts the developmental trajectory — access supports without first having to satisfy a permanence test. Section 25(1)(b) of the current Act defines an early intervention support as one for which “providing the early intervention is likely to benefit the person by reducing the person’s future needs for supports.” The Parliament that enacted the original NDIS Act in 2013 built early intervention into the scheme as a separate pathway precisely so that supports could be provided to prevent the deterioration of impairment-related needs without first requiring the person to demonstrate permanence.

The Bill’s new s.25(1B) imposes the identical permanence-and-treatment test on the early intervention pathway. Functionally, this collapses the s.25 pathway into the s.24 pathway: an applicant for early intervention support must now also demonstrate that they have undertaken all appropriate treatment, that no other treatment is likely to materially improve the impairment, and that the impairment will persist for life. The architecture that allowed early intervention support to flow before permanence was established is removed.

This is significant beyond its immediate effect on the early intervention cohort. The Government’s spokesperson statement reported in The Guardian on 13 May 2026 framed the reform package as ensuring that the NDIS does not “replace health, rehabilitation and treatment services which play a critical role in preventing lifelong disability.” That framing presupposes that disability is a condition that treatment prevents. Applied to the cohorts the NDIS serves — autism, intellectual disability, Down syndrome, cerebral palsy, sensory impairments — the framing is clinically incoherent. Autism is not “prevented” by treatment. Down syndrome is not “prevented” by treatment. The s.25 early intervention pathway in the current Act was designed for a different proposition: that supports provided early reduce the future support needs of a person whose impairment is, and remains, permanent. The Bill collapses both the pathway and the conceptual distinction.

The Guardian operative-text grab and the absence of clinical evidence base

The operative text of the permanence-and-treatment gate was first publicly reported in Australian media in The Guardian on 13 May 2026 (Dan Jervis-Bardy and Sarah Basford Canales, “Australians will need to exhaust ‘all appropriate’ treatment options to access NDIS under proposed rules”). At the date of that publication, the Bill had not been introduced to Parliament — that occurred the following day. The verbatim text the Guardian article reported has since been confirmed by the Bill as introduced.

Submission 2079

In the same article, Senator Jordon Steele-John (Greens disability spokesperson) described the proposal as “extraordinarily alarming” and “cruel”; Senator Steele-John warned that the test would force people to spend years “running around, getting treatments that may be ineffective or inappropriate”. Shadow NDIS Minister Melissa McIntosh MP described the consultation process as inadequate, citing “an overwhelming amount of feedback that the government has once again neglected to consult with the Australians directly affected by this legislation. People with a disability are feeling ambushed by the Government.”

That two parliamentary parties of different political traditions — the Greens and the Coalition — have characterised the consultation process as inadequate is structurally significant for the Committee’s assessment of the Bill. A reform of the scale this Bill imposes — affecting access for hundreds of thousands of disabled Australians, with no published clinical-evidence underwriting the operative test — would, in a functioning deliberative process, be developed through substantive consultation with the clinical and affected communities, and through cross-party engagement before introduction. The Explanatory Memorandum does not name any specific clinical body — medical college, peak clinical society, or condition-specific clinical authority — that the Department engaged before drafting; the Consultation section names categories of engagement (the NDIS Review, the Royal Commission, the Registration Taskforce 2024) without identifying any specific clinical body that reviewed or endorsed the proposed test. The Bill was introduced thirteen days after the Joint Standing Committee on the NDIS Integrity inquiry’s first public hearing; the multi-party parliamentary critique on the inadequate-consultation point is on the public record. The procedural pattern is the data point — not the rhetorical characterisation either parliamentary party has used. Section 15 of this submission addresses the procedural-democracy concerns in full, including the Government’s own AGGPIA 30-day consultation minimum and the joint statement of twelve Disability Representative Organisations.

On the Government’s published material on the reform package — the Bill’s Explanatory Memorandum, the Minister’s National Press Club address of 22 April 2026, the Minister’s Second Reading Speech of 14 May 2026, Budget Paper No. 2 measure text of 12 May 2026, and the Department of Health, Disability and Ageing’s ministerial response MC26-008601 of 25 May 2026 — the Government has not identified any clinical body that has reviewed the proposed test and concluded that it is methodologically coherent for the impairment cohorts to which it will apply. The 2023 Independent Review of the National Disability Insurance Scheme (the Independent NDIS Review)’s lead members were not clinicians, and the Review itself did not propose the verbatim test the Bill imposes. The peak clinical and medical bodies — the Royal Australasian College of Physicians, the Royal Australian and New Zealand College of Psychiatrists, the Australian Psychological Society, and the Australasian Faculty of Rehabilitation Medicine — have not publicly endorsed the test. Nor have the peak disability and condition-specific representative bodies — Autism Spectrum Australia, the Australian Autism Alliance, AMAZE, Inclusion Australia, and the Cerebral Palsy Alliance / Cerebral Palsy Australia. The Committee is entitled, and I submit obligated,

Submission 2079

to seek clinical evidence on the test before it reports. The parallel absence of a cost- benefit analysis on the reform package is addressed in Section 14.

The parallel peak-body validation gap applies, with greater specificity, to the assessment instrument that will operationalise the new functional-capacity gate that sits alongside the permanence-and-treatment test. People with Disability Australia — the national cross- disability peak — has, in its 26 May 2026 member-blast submission template for this Committee, flagged the Indicator of Cognitive and Adaptive Needs (I-CAN) instrument being rolled out under the new framework planning regime as not validated for episodic or fluctuating disability, and as not validated for First Peoples cultural appropriateness. The First Peoples validation gap is material in the context of the Bill’s s.25A(2) regressivity codification (addressed in Section 7) and against the National Agreement on Foundational Supports 2026-31 commitment at clauses 24 and 25 to align with Closing the Gap. PWDA’s flagging is peak-body weight: the recognised national cross-disability peak has put on the public submission record that the instrument the Bill’s reform package will rely on for functional-capacity assessment is unvalidated for two of the cohorts on which it will most consequentially operate. The peak-body validation gap is therefore not confined to the permanence-and-treatment test; it extends to the assessment-instrument architecture the Bill builds around the test. Section 5 of this submission develops the assessment- instrument vacuum and the PWDA cultural-appropriateness critique in full.

The income-and-geographic regressivity layer of s.25A(2)

Sub-section 25A(2)’s express provision that treatment remains “appropriate” regardless of whether the person’s individual circumstances restrict the person from accessing it — with the Note that “individual circumstances include the person’s financial circumstances and geographical location” — codifies in statute an income-and-geographic gating that operates before and on top of the clinical-evidence gate. The clinical-evidence question therefore has a clinical-access dimension. Adult autism diagnosis alone, on which any subsequent treatment-exhaustion exercise would depend, currently costs approximately $2,500 out-of-pocket privately against a twelve-month-plus public waitlist. The NDIS does not fund diagnostics. Treatment exhaustion can mean tens of thousands of dollars in private therapy and medication trials, concentrated in major metropolitan service catchments. The s.25A(3)(a) carve-out for “appropriate medical treatment the person cannot undertake for medical reasons” is narrow and does not extend to treatment a person cannot access for financial or geographic reasons.

The cohort the s.25A(2) drafting most stringently excludes is therefore the cohort with the least capacity to absorb the cost and travel: Australians in lower-income households, and Australians in remote and very remote areas. Aboriginal and Torres Strait Islander people are over-represented in both populations. Treating treatment as “appropriate” regardless of the applicant’s financial means or geographic access is, on the face of the provision, an explicit decision to gate scheme access on income and location. The Australian disability framework’s foundational instruments — the Disability Discrimination Act 1992, the United Nations Convention on the Rights of Persons with Disabilities, and the National Agreement

Submission 2079

on Closing the Gap — do not contemplate disability access being gated on the affordability or geographic availability of treatment for the underlying impairment.

Section 7 of this submission develops the full discrimination and statutory-inconsistency dimensions of s.25A(2) under each of those three instruments.

The federalism dimension: the treatment-exhaustion limb against state mental-health-legislation rights to refuse treatment

A further dimension of the clinical-evidence vacuum surfaced on the parliamentary record at the Joint Standing Committee on the NDIS’s second public hearing in Sydney on 15 May 2026. Professor Jennifer Smith-Merry of the Centre for Disability Research and Policy at the University of Sydney - whose Australian Research Council Laureate Fellowship research is based on qualitative interviews with 86 NDIS participants, family members and stakeholders examining experiences of the scheme for people with psychosocial disability - testified that the Bill’s exhaust-treatment limb at new s.24(5)(a) and new s.25(1B)(a) is in conflict with state and territory mental-health legislation rights to refuse treatment.

The treatment-exhaustion requirement is a Commonwealth legislative requirement; the right to refuse treatment is a State and Territory legislative right. An applicant who exercises the State-conferred right to refuse a treatment categorically cannot satisfy the Commonwealth-conferred requirement that the same treatment be exhausted. Section 7 of this submission addresses the statutory-inconsistency dimension of that conflict in full, as the seventh pillar of the Schedule 1 statutory-inconsistency case, alongside the CRPD, DDA and Closing the Gap inconsistencies the other amendments produce.

Closing the case for Schedule 1

The clinical-evidence vacuum is the first of the three load-bearing arguments for the omission of Schedule 1. The diagnostic-language vacuum (Section 4) and the assessment-instrument vacuum (Section 5) build the case to its full structural form. The cumulative pattern is consistent: the access regime the Bill imposes is operating against a clinical evidence base the Government has not assembled, on cohorts for whom the framework’s underlying logic does not apply, with an assessment apparatus that has not been designed, and against an effective new construct of permanent disability the rest of the Australian disability framework does not recognise. The vacuum is not a procedural defect that can be filled by post-enactment work. It is a structural defect in the legislation itself.

Submission 2079

Section 4 — The diagnostic-language critique: “low to moderate support needs” against the clinical evidence base

The Government has placed on the Budget paper record, in the Thriving Kids measure of Budget Paper No. 2 (2026-27), a description of the cohort that will be transitioned out of NDIS individualised plans into the new 0-to-8 foundational supports program. The verbatim measure text reads:

 "children aged eight and under with developmental delay and/or autism with low

to moderate support needs, as well as their families, carers and kin.“

That phrase — “low to moderate support needs” — is the diagnostic-language frame on which the Thriving Kids cohort, and by extension the targeting logic of the broader reform package, depends. The phrase carries the rhetorical weight of clinical specificity. It is not, in fact, a term that any of the operative clinical documents in this area use.

The Department’s signed concession-and-contradiction

The framing has now been placed on the Departmental signature record in a way that makes the contradiction at the heart of the Bill’s targeting logic the Department’s own. The Department of Health, Disability and Ageing issued a signed Ministerial response to my direct correspondence to the Hon Mark Butler MP on 25 May 2026, Reference MC26- 008601, signed by L. Jackman, Assistant Secretary, Thriving Kids Taskforce. Paragraph 2 of that response reads:

"Terms such as ‘high functioning’, ‘low functioning’, ‘profound’, ‘severe’ and ‘mild’

autism are sometimes used to describe the challenges and barriers experienced by Autistic people and their families and carers. We know that every Autistic person has different strengths and faces different challenges and barriers, and these other terms can misrepresent the difficulties some Autistic people may experience and the capabilities of others.“

The immediately following paragraph reads:

 "Thriving Kids will support children aged eight and under with developmental

delay and/or autism with low to moderate support needs and their families.“>

The Department, in a single signed letter, acknowledges that “mild” and “severe” autism terminology “can misrepresent the difficulties some Autistic people may experience and the capabilities of others“ — and then, in the immediately following paragraph, operaionalises the equivalent “low to moderate support needs” terminology as the Thriving Kids cohort descriptor. The concession-and-contradiction sits inside two adjacent paragraphs of the same letter. The Department’s own correspondence supplies the strongest single piece of primary-source evidence that the diagnostic-language framing the Bill’s targeting logic depends on is one the Department itself has acknowledged can misrepresent the Autistic cohort it applies to.

Submission 2079

The same letter further notes that “Thriving kids is not diagnosis based” — yet the cohort descriptor used in the immediately prior paragraph (“low to moderate support needs”) IS a diagnostic-language descriptor. A second internal contradiction within the same letter compounds the first.

The descriptor is doing operational work, not only rhetorical work. “Low to moderate support needs” is the operational filter the Thriving Kids program is being designed around. The new functional-capacity assessment at new s.9B is the instrument that will operationalise the filter, with the threshold delegated to NDIS Rules under s.9B(2) and (3) and the design work now beginning in the Technical Advisory Group provisioned at Budget Paper No. 2 line 8919 with $3.3 million in 2026-27. Section 7 develops the statutory-inconsistency consequence in full: the Government’s stated rationale for the access-tightening package is permanence, but Autism is by definition permanent (DSM-5 lifelong neurodevelopmental condition with diagnostic criteria explicitly requiring symptoms in the early developmental period), so the permanence limb of the new s.24(5) and s.25(1B) gate cannot exclude Autistic applicants. The operational exclusion mechanism the Bill therefore relies on is the “low to moderate support needs” cohort descriptor operationalised via new s.9B — severity-by-another-name, with no statutory grounding in s.24 or s.25. The descriptor the Department’s MC26-008601 paragraph 3 places on the signature record is the operational gate through which Autistic children are diverted from NDIS individualised supports into the state-delivered Thriving Kids program in lieu of NDIS plans. The diagnostic-language acknowledgement in paragraph 2 of the same letter is the Department’s own admission that the operational gate cannot do that work without misrepresenting the cohort it applies to.

The remainder of this section sets out the corresponding absence of the framing in the operative clinical and statutory documents the Department’s diagnostic-language descriptor would need to be anchored in: the DSM-5, the Independent NDIS Review, the National Autism Strategy 2025-31, and the position statements of the Australian peak clinical bodies. The pattern across those documents is consistent. The framing has no foundation in any of them — a fact the Department itself, in paragraph 2 of MC26-008601, has now acknowledged in writing.

What the DSM-5 actually uses: three support-needs levels, not autism severity

The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), published by the American Psychiatric Association, is the diagnostic instrument used by Australian clinicians for autism spectrum disorder. The DSM-5 does not categorise autism by “mild”, “moderate” or “severe”. Autism is not a condition that exists in more or less severe forms; the diagnostic question is whether a person meets the diagnostic criteria for autism, which are themselves based on persistent differences across the lifespan in defined domains. What the DSM-5 does is identify three support-needs levels — explicitly tied to the supports the person requires in their actual context, not to the intensity of the underlying condition:

Submission 2079

• Level 1 — “requiring support”.
• Level 2 — “requiring substantial support”.
• Level 3 — “requiring very substantial support”.

These three support-needs levels are applied separately to two domains of presentation — social communication, and restricted/repetitive behaviours — and reflect the pattern and persistence of supports the person requires for daily functioning. The DSM-5’s framework is functional and support-anchored. It is not a “mild to severe” autism spectrum in the way colloquial usage often implies. The support-need level a clinician records under the DSM-5 framework is, by the framework’s own design, not a stable across-time-and-context attribute of the person — Autistic support needs change with life-stage transitions, with environmental demands, with co-occurring conditions, and with the supports already in place. The DSM-5 framework recognises this; the “low to moderate support needs” descriptor the Budget Paper No. 2 measure uses does not.

The Government’s “low to moderate support needs” phrase does not appear in the DSM-5. The DSM-5 does not use “low”. It does not use “moderate” as an autism descriptor. The closest analogue — Level 1 (“requiring support”) — is not the same construct. “Requiring support” is a pattern descriptor in the DSM-5: a person at Level 1 has noticeable difficulties without supports in place, with social-communication and behavioural impacts that are real and pattern-disrupting. A Level 1 designation is, by the DSM-5’s own definition, a finding that the person requires support — exactly the operational basis on which the existing s.34 reasonable-and-necessary NDIS test was designed to assess what supports are most appropriately funded through the scheme rather than through other systems. The existing test is the mechanism by which the cohort question is resolved at the individual level. “Low to moderate support needs” is not the same construct; it is a category of “low need” in the everyday sense the Budget Paper No. 2 phrase invites, applied at the cohort level rather than the individual level.

This matters because the public framing of Thriving Kids — and the eligibility-tightening provisions in Schedule 1 of the Bill that target the same cohort through the new schemewide functional capacity assessment regime from 1 January 2028 — depends on the implication that the cohort being moved off NDIS individualised plans, or kept off them in future, has clinically lower needs than the cohort retained. The DSM-5 does not support that implication. Within Level 1 alone, Autistic children present with a range of needs across the social-communication and restricted-and-repetitive-behaviour domains. Many of those children require — and the existing NDIS scheme has been delivering — individualised supports that cannot be reproduced through generalised foundational programs. The mechanism the Bill is constructing is the substitution of a cohort-level descriptor for the individual-level assessment the s.34 test was designed to perform.

Submission 2079

The Independent NDIS Review (Bonyhady & Paul, 2023) did not use the framing

The Minister and the Bill’s Explanatory Memorandum identify the 2023 Independent NDIS Review as the recommendation source for the reform package’s broader direction. The Review’s Working Together to deliver the NDIS report, delivered to Government in December 2023, did not use the “low to moderate support needs” framing as a basis for cohort design or NDIS access decisions. The Review’s recommendations on foundational supports were structured around a graduated tier model (general foundational supports; targeted foundational supports; NDIS-funded individual support packages); they were not anchored on a “low to moderate” diagnostic descriptor for any cohort. The phrase that has since appeared in Budget Paper No. 2 and in Government public messaging is not the Review’s framing.

This is consistent with the Review’s panel composition. The Review was led by Professor Bruce Bonyhady (an economist, founding chair of the NDIA) and Lisa Paul AO PSM (a former senior public servant). The Review’s panel did not include the clinical leadership of any peak Australian autism, intellectual-disability, cerebral-palsy, or general-disability body, and the Review did not commission a peak-clinical-body endorsement of the diagnostic-language framing that has since attached to the reform package.

The National Autism Strategy 2025-31 does not use the framing either

The National Autism Strategy 2025-31, published by the Commonwealth as the cross- portfolio policy framework for autism in Australia, does not use “low to moderate support needs” as a cohort descriptor. The Strategy’s framing of the Autistic population centres on lived experience, support need across the lifespan, and intersectional barriers. The Strategy explicitly acknowledges the heterogeneity of Autistic presentations and the inadequacy of single-dimensional severity descriptors as a basis for policy design.

The inconsistency between the National Autism Strategy 2025-31 (which the Commonwealth has published as the policy framework for the Autistic cohort) and the Thriving Kids measure (which the Commonwealth has published as the operational program for the largest single-impairment cohort of NDIS new entrants) is a cross-portfolio inconsistency the Committee is entitled to require the Department to explain. The two documents are operating in different conceptual registers. Only one of them — the National Autism Strategy 2025-31 — has a published clinical-evidence basis and the substantive Autistic-community and disability-peak-body engagement the “nothing about us without us” principle requires. The document being used to determine NDIS access — The Thriving Kids Budget Paper No. 2 measure text — is not the one with the clinical or the community foundation.

No Australian peak clinical body has publicly endorsed the framing

I have searched the public statements, position papers, and submissions of the relevant Australian peak clinical bodies for an endorsement of “low to moderate support needs” as

Submission 2079

There is no existing basis for autism cohort categorisation in NDIS access decisions. I have not found one. The bodies whose endorsement would carry the relevant evidentiary weight include:

• The Royal Australasian College of Physicians (RACP).
• The Royal Australian and New Zealand College of Psychiatrists (RANZCP).
• The Australian Psychological Society (APS).
• The Australasian Faculty of Rehabilitation Medicine (AFRM).
• Autism Spectrum Australia (Aspect).
• The Australian Autism Alliance (AAA).
• AMAZE (Victorian autism peak).
• Inclusion Australia (national peak for intellectual disability).
• The Cerebral Palsy Alliance / Cerebral Palsy Australia.

The Bill applies the test on which the cohort categorisations depend — both Thriving Kids’ 0-to-8 cohort framing and, from 1 January 2028, the scheme-wide functional capacity assessment regime in new s.9B — to autism, intellectual disability, Down syndrome, cerebral palsy, and the other impairment cohorts. None of the corresponding peak bodies has publicly endorsed “low to moderate” as the descriptor for who is in scope and who is not, for any age cohort. The Committee is entitled, and I submit obligated, to request from the Department of Health, Disability and Ageing the list of clinical bodies that have reviewed the cohort-categorisation framing — for Thriving Kids specifically, and for the broader functional-capacity assessment regime the Bill commences from 1 January 2028 — together with the dates of those reviews. If no such reviews have taken place, that is itself a finding the Committee should record.

Independent third-party convergence on the clinical-basis absence

The same conclusion — that the diagnostic-language framing the Bill operationalises has no foundation in the operative clinical documents — has been reached independently by credentialed third-party commentary on the predecessor “mild to moderate autism” framing used in Government public messaging since August 2025. Hannah Thomas, a lawyer and policy advisor writing in Cheek on 20 April 2026 (“Labor is manufacturing an NDIS crisis to justify cuts”), described the “mild to moderate autism” categorisation as “a misleading political term which has no clinical basis”, and traced its policy entry to August 2025, when then-newly-appointed Minister for the National Disability Insurance Scheme Mark Butler announced a plan to curb NDIS spending and expressed concern at the number of Autistic children on the scheme. The Budget Paper No. 2 “low to moderate support needs” measure text published in May 2026 is the operational successor to the same framing. Dr Clem Bastow, a cultural critic and critical Autism studies researcher (Naarm–Melbourne), writing in Crikey on 13 April 2026 under the headline “NDIS headlines are turning autistic people into the new dole bludgers”, made an adjacent point on the categorising’s effects — that the media apparatus accompanying the reform package’s roll-out has functioned to recategorise Autistic NDIS participants as a new welfare-suspect cohort, decoupled from the clinical literature on autism that the framing depends on for its

Submission 2079

apparent specificity. The convergence — peak Australian clinical bodies silent on the framing, and credentialed independent commentary across two registers describing the framing as a political construct without clinical foundation — confirms the absence of the clinical evidence the Committee should require the Department to identify.

The same critique has surfaced from a Government-side parliamentary register. Dr Mike Freelander MP, the Member for Macarthur (Labor) and a paediatrician of more than four decades’ practice, served as Chair of the House Standing Committee on Health, Aged Care and Disability for the Inquiry into the Thriving Kids initiative. The Committee’s December 2025 report, No child left behind, carries in its Foreword the Chair’s acknowledgement that “mild or moderate disabilities… are very subjective definitions.” A paediatrician Labor MP, serving as Chair of a Government-referred Inquiry into the program the diagnostic-language descriptor was designed for, is on the parliamentary record conceding that the framing the Bill’s targeting logic depends on is subjective. The Foreword’s concession sits alongside the Department’s signed concession in paragraph 2 of MC26-008601 — two adjacent Government-side registers (parliamentary committee Chair’s Foreword; Departmental Assistant Secretary’s signed Ministerial response) both acknowledge the limitation of the framing the Bill operationalises. The pattern is now four-register: peak Australian clinical bodies silent on the framing; credentialed independent commentary across two registers describing it as a political construct without clinical foundation; the Chair of the Government’s own Thriving Kids Inquiry conceding the descriptor is “very subjective” on the parliamentary record; and the Department’s signed Ministerial response acknowledging the underlying terminology can misrepresent the cohort. The absence of the clinical evidence the Department would need to point to is the only fact consistent with all four registers.

The connection to new s.9B (functional capacity)

Schedule 1’s new s.9B defines functional capacity as the person’s ability to undertake an activity “without assistance from other people, assistive technology or modifications” and “in a context that excludes, as far as possible, the impact of the person’s environmental and personal circumstances.”

This definition is in direct tension with the DSM-5’s Level-based framework. The DSM-5’s support-needs levels are explicitly defined by support need in context. A person at Level 1 “requires support” in their actual social and behavioural environment — that is what the Level descriptor refers to. The Bill’s new s.9B requires functional capacity to be assessed in the absence of the supports the DSM-5 framework treats as constitutive of the person’s functional profile.

The conflict between the two frameworks is the same conflict between the social model of disability and the medical model of disability that the international disability-rights instruments have resolved in favour of the social model. Autism disability is constituted by the interaction between Autistic neurology and a neurotypical environment — one whose social rules, sensory load, communication conventions, and routine administrative

Submission 2079

demands are not designed for Autistic functioning. Strip out the environment — the assistive technology, the supports, the accommodations, the neurotypical-norm-attuned cohabitants — and a substantial share of what is experienced as autism disability does not present at all. The same dynamic applies to a range of other impairment cohorts. The CRPD Article 1 definition of disability that Australia, as a State Party, is obligated to give effect to under NDIS Act 2013 s.3(1)(a) is anchored on exactly this relational understanding: -disability is constituted by the interaction between impairment and barriers. The Bill’s new s.9B is anchored on the inverse construct — the person’s ability stripped of the very interactions the international framework identifies as constitutive. The mismatch is not a clinical-language inconsistency at the margins. It is a structural inversion of the framework the scheme is committed to.

The inversion carries direct legal-exposure consequences. Section 7 of this submission sets out the statutory and treaty inconsistency between new s.9B and CRPD Article 1, alongside the parallel inconsistencies with the Disability Discrimination Act 1992 and the National Agreement on Closing the Gap. The diagnostic-language vacuum this section sets out and the s.9B definition the diagnostic-language vacuum will be operationalised through are not severable issues. They produce, together, a Bill whose targeting logic rests on a cohort descriptor without clinical foundation, applied through a functional-capacity test whose statutory frame inverts the international model the Commonwealth has committed to give effect to.

The assessment-instrument vacuum the framing depends on

A second consequence follows from the absence of a clinically grounded “low to moderate” category. A standardised functional-capacity assessment instrument is intended to be administered by a third-party assessor to produce a reliable, reproducible categorisation across many people. For the instrument to categorise applicants reliably, The categories themselves must be clinically reliable — that is, they must map to constructs that the diagnostic literature defines consistently and that trained assessors can apply consistently. If the categories themselves are not anchored in the operative clinical framework, the instrument designed to apply them cannot be reliable on its face.

The Bill imposes the new functional capacity assessment regime from 1 January 2028 (new s.9B; the commencement is confirmed in Budget Paper No. 2 line 8929). No assessment instrument has been publicly named for the 9-to-15 cohort. The I-CAN instrument flagged in disability-sector commentary is described as applying to ages sixteen and above; Thriving Kids has its own state-delivered methodology still being worked up. The Government’s Technical Advisory Group, established in Budget Paper No. 2 line 8919 with $3.3 million in 2026-27, is the body now beginning to design the instruments. The Group has not yet reported. The instrument it will design is, at the time of writing, not publicly named, not publicly described, not publicly evidence-tested, and not publicly endorsed by any clinical body.

Submission 2079

The Bill’s commencement date for the functional capacity regime - 1 January 2028 - is therefore set before:

• A clinically reliable category framework has been established for “low to moderate support needs” (because the diagnostic literature does not use this category);
• An assessment instrument has been named for the 9-to-15 cohort;
• An assessment instrument has been publicly tested for any of the new categories the regime will apply;
• A peak clinical body has reviewed and endorsed the assessment apparatus.

The Committee is entitled to take the position that an eligibility regime cannot commence on a date set in primary legislation before the assessment apparatus that operationalises it has been demonstrated to be clinically reliable. I develop the operational consequences of this for my fourteen-year-old daughter - and for the 9-to-15 cohort generally - in Section 5.

What the diagnostic-language vacuum means for the Committee

The diagnostic-language vacuum is significant for the same reason the clinical-evidence vacuum in Section 3 is significant. The Bill’s targeting logic depends on a clinical categorisation that the operative clinical documents do not use. The Government has put the term on the Budget paper record and on the Ministerial Second Reading record without identifying the clinical body that endorses it or the clinical document that anchors it. The Australian disability framework — the Disability Discrimination Act 1992 (s.4), the CRPD (Article 1), the NDIS Act 2013 (s.3), the DSM-5 (the diagnostic-instrument layer), and the National Autism Strategy 2025-31 (the cross-portfolio policy layer) — do not, between them, supply the framing the Bill’s targeting logic depends on. The framing has been generated, in policy register, by Government agencies and is being used to determine the operational scope of an Act of Parliament.

The Committee is entitled to require that the diagnostic-language framing be substantiated against the operative clinical and statutory documents, and that the Thriving Kids cohort description in Budget Paper No. 2 either be reconciled with the DSM-5 framework and the National Autism Strategy 2025-31, or be withdrawn as the targeting basis for the foundational supports program (Recommendation 22).

Submission 2079

Section 5 — The 9-to-15 cohort and the plan-reassessment and renewal provisions in new s.48A and s.50A

This section concerns two distinct ways the Bill operates against existing participants, both of which converge on my fourteen-year-old daughter. Part 1 addresses the cohort of children and adolescents on the National Disability Insurance Scheme between the ages of nine and fifteen — the cohort most acutely exposed to the gap between the date the Bill’s new tests commence and the date the assessment, replacement-program, and clinical-evidence apparatus those tests depend on will actually exist.

Part 2 addresses the participant-initiated reassessment and automatic renewal provisions in new s.48A and s.50A, which are not cohort-specific: they apply scheme-wide, to existing participants of every age, and they disproportionately disadvantage impairment categories characterised by fluctuating need. My daughter sits in both — she is a member of the 9-to-15 cohort and an existing participant whose plan will be reassessed and renewed under the new provisions.

My contention is that no existing participant should be reassessed against an apparatus that has not been designed; that the 9-to-15 cohort is the cohort most acutely exposed to that apparatus gap; and that the s.48A and s.50A provisions narrow participant agency over their own plans in ways that fall hardest on the cohorts whose support needs are not static.

Part 1 — The 9-to-15 cohort and the apparatus gap

The 9-to-15 cohort sits in a triple gap

The reform package has three operative strands that affect access to the NDIS. They are: an eligibility policy (the new scheme-wide permanence-and-treatment test and the new functional capacity definition in new s.9B); an assessment instrument (the standardised functional-capacity tool to be administered to applicants and to existing participants on plan review); and a replacement supports program (the foundational supports arrangements that will operate outside the NDIS for people the new eligibility test does not admit).

The three strands operate against different age bands in different ways. For the 0-to-8 cohort, Thriving Kids — commencing 1 October 2026 with $4 billion over five years per Budget Paper No. 2 — is the announced replacement program. For the 16-and-over cohort, the I-CAN instrument has been flagged in sector commentary as the assessment instrument that closes part of the apparatus gap. The 9-to-15 cohort sits in what I have referred to throughout this submission as the triple gap:

• The new eligibility test applies. The new permanence-and-treatment requirement in s.24(5) and s.25(1B), the new functional capacity definition in s.9B, the s.25A “appropriate treatment” construct, and the cost-shift gate in s.25B all apply to existing 9-to-15 participants on plan reassessment and to any 9-to-15 applicant entering the scheme from 1 January 2028 onward.

Submission 2079

• No assessment instrument has been publicly named for the cohort. The I-CAN instrument flagged for the 16-and-over cohort does not, on disability-sector commentary, apply to children under sixteen. The state-delivered methodology for Thriving Kids is being scoped for children up to age eight. The Technical Advisory Group provisioned in Budget Paper No. 2 with $3.3 million in 2026-27 has not yet reported on a 9-to-15 instrument.
• No replacement supports program has been announced for the cohort. Thriving Kids excludes the 9-to-15 cohort by age. The unnamed adult and 16-to-17 foundational supports programs the Government referenced at the National Press Club on 22 April 2026 have not been named, designed or funded as at the date of this submission. The 9-to-15 cohort is not within the scope of any announced foundational supports program.

The 9-to-15 cohort is therefore the cohort to which the Bill’s eligibility-tightening provisions apply with the fewest mitigating safeguards in place. The triple gap is not a matter of phasing. It is a matter of which cohorts have been resourced for the transition the Bill imposes and which have not.

Peak-body validation concerns about the I-CAN instrument

The 9-to-15 cohort sits in a triple gap because the I-CAN instrument flagged in sector commentary for the 16-and-over cohort has not been validated for younger ages. The parallel question — whether the I-CAN instrument is validated for the cohorts and circumstances to which it WILL be applied from age 16 — has been put on the public submission record by People with Disability Australia in its 26 May 2026 member-blast submission template for this Committee. PWDA — the recognised national cross-disability peak — flags two specific validation gaps in the I-CAN methodology: the instrument is not validated for episodic or fluctuating disability, and it is not validated for First Peoples cultural appropriateness.

The episodic-or-fluctuating-disability validation gap is material to the cohorts the Bill is argeting. Autistic support needs are episodic and context-dependent — they respond to Autistic burnout, to changes in masking and compensatory strategies, to life-stage transitions, and to changes in the external supports sitting alongside the plan. Psychosocial-disability support needs are highly variable across time and circumstance. Mental-health conditions and several physical-impairment cohorts share these characteristics. An assessment instrument that has not been validated for episodic or fluctuating disability will under-measure the support needs of the cohorts whose impairment is structurally episodic or fluctuating. The methodological gap maps directly onto the cohorts whose support needs are not static — the same cohorts the new s.48A reassessment thresholds, addressed in Part 2 of this section, also disadvantage.

The First Peoples cultural-appropriateness validation gap is sharper. Aboriginal and Torres Strait Islander people are over-represented in the populations the Bill’s new s.25A(2) regressivity codification will most directly disadvantage — lower-income and remote

Submission 2079

populations. They are also disproportionately represented among the cohorts the Bill’s eligibility-tightening will affect. The National Agreement on Foundational Supports 2026-31, signed by all eight jurisdictions in early 2026, commits the Commonwealth and the States at clauses 24 and 25 to aligning Foundational Supports design and implementation with the four Priority Reforms of the National Agreement on Closing the Gap. The Foundational Supports Agreement is the inter-governmental instrument the Bill’s reform package is rolled into. Applying an unvalidated-for-First-Peoples assessment instrument to First Peoples participants — at the access gate of the scheme that sits underneath the Foundational Supports Agreement — is structurally inconsistent with the Agreement’s own commitments to Closing the Gap.

PWDA’s flagging is peak-body weight. The recognised national cross-disability peak has put on the public submission record that the instrument the Bill’s reform package will rely on for functional-capacity assessment is unvalidated for two cohorts on which it will operate at scale. The Committee is entitled to require the Department to identify the validation evidence underpinning the I-CAN instrument’s application to episodic or fluctuating disability, and the validation evidence underpinning its cultural appropriateness for First Peoples participants — before the new framework planning regime commences on 1 April 2027 and before the new functional capacity assessment regime commences on 1 January 2028.

The rushed-rollout pattern is being repeated

The transition-gap risk for the 9-to-15 cohort is not a hypothetical concern arising from this Bill alone. It is the second iteration of a rushed-rollout pattern that senior figures across the disability sector have publicly identified as a significant contributor to the integrity, capacity and equity problems the Government now cites as the justification for the very cuts the Bill imposes. The evidence those figures give speaks to distinct dimensions of the problem, and this submission is careful not to merge them: the Agency’s own anti-fraud co-chair attributes the integrity gap to the design of the original payment system, built for speed of disbursement; the senior Labor architect of the scheme attributes the broader difficulties to a rollout whose early warnings were not heeded; and the academic voice closest to the deinstitutionalisation history attributes the capacity gap to the absence of a ready replacement service system. Speed of payments and speed of rollout are different things; both are in evidence, and neither is participant eligibility.

On 1 May 2026, the co-chair of the National Disability Insurance Scheme Fraud Fusion Taskforce — Mr John Dardo, Deputy Chief Executive Officer for Integrity, Transformation and Technology Services at the National Disability Insurance Agency — gave evidence to the Joint Standing Committee on the NDIS’s Integrity of the National Disability Insurance Scheme inquiry. Asked by the Chair to identify the contributing factors to the integrity problems the inquiry was investigating, Mr Dardo placed on the parliamentary record that “the program was established quite quickly. I think there were some assumptions about how the program, the market and the claiming systems would work.” He described the original payment system in detail: “the speed of payments getting out was the primary

Submission 2079

“motivator of the system design”, with manual extraction of integrity-flagged claims limited to “maybe 20 to 30 claims per day”, and “any claims between 5 pm and 6 pm were being paid without any human eyeballs on them.”

The Government’s own anti-fraud co-chair has placed on the parliamentary record that a primary cause of the integrity problems is the design of the original payment system — built for speed of disbursement, not for fraud prevention — and that the post-2022 remediation has been about putting payment-system checks and balances into place. The diagnosis the Government’s own evidence base supports is a payment-system-design diagnosis, not a participant-eligibility diagnosis. Schedule 1 of this Bill does not address payment-system design. Schedule 2 does, in part (Section 9). The Bill’s targeting logic — than who is on the scheme — is contradicted by the Government’s own anti-fraud co-chair on the parliamentary record.

The Honourable John Della Bosca, former New South Wales Labor health and disability minister and a pivotal advocate for the establishment of the NDIS, has publicly warned the Albanese Government against rushing the present changes (ABC News, 18 May 2026). On the original rollout, Mr Della Bosca observed “there were points along the way when I thought that there were things that were going wrong very early in the piece, and I don’t think the obvious warnings were heeded.” On the present reform: “in policy terms and human relations terms, you won’t be able to solve it immediately … it needs to be planned out carefully with people with a disability in mind.”

Professor Keith McVilly of the University of Melbourne, who began his career as a psychologist in one of the institutions that closed before the NDIS began, has made the same point (ABC News, 17 May 2026): “Unless [these new services] are well designed, we could end up in that crisis situation again.” Professor McVilly’s assessment is that “there [is] next to nothing from the old state systems that could be reactivated and much of their new systems would be starting from scratch” and that “we need this rebuild of the system now, but in reality, it’s going to take quite a number of years to achieve.”

The disability community’s own national peak bodies make the same point about readiness. Ms Megan Spindler-Smith, Acting Chief Executive Officer of People with Disability Australia, told the Australian Broadcasting Corporation on 18 May 2026 that “this is a huge reform map … and there’s such a short period of time to get this done. These systems cannot be set up overnight and we as a community can’t be used as a political football,” and that “we need to know what is going to be built … and that doesn’t mean going back to an old way of doing things that we know didn’t work and that were segregating many of us.” Ms Sophie Cusworth, Chief Executive Officer of Women with Disabilities Australia, in the Disability Representative Organisations Joint Statement of 14 May 2026, located the readiness concern in the assessment apparatus itself: “Women, girls and gender-diverse people with disability are already being failed by eligibility settings and assessment tools that do not properly recognise our lives, our needs, or the way disability presents for us. We know women and girls are less likely to be found eligible for the NDIS than men. Rushing through reforms without proper scrutiny risks locking that

Submission 2079

gender bias even deeper into the system.” These are the cohort’s own representative organisations, saying — before the Bill’s regime commences — that the systems it relies on are not yet built and that the assessment tools it will use do not yet work for the people they will be applied to.

The Bill, in contrast, commences the new functional capacity assessment regime on 1 January 2028 — set before the assessment instrument has been designed. The Thriving Kids program starts on 1 October 2026 — before bilateral implementation agreements with Victoria and Queensland are signed. The new framework planning rollout starts on 1 April 2027 — across the whole scheme, with the assessment instrument design and the s.25A(4) carve-out rules still to be made.

The Government’s framing of the reform package’s justification — that the current problems are a function of who is on the scheme and what supports they receive — is contradicted by the Government’s own anti-fraud official, by the senior Labor architect of the scheme, by the academic voices closest to the rollout history, and by the disability community’s own national peak bodies. Those voices identify a primary driver of the present problems as a function of how fast the original transitions were executed, with insufficient design, consultation, and replacement-program readiness. The Bill responds by repeating the same rushed-rollout pattern — and on the available evidence, with even more compressed Parliamentary scrutiny windows than the original 2013 NDIS legislation received. The 9-to-15 cohort is the cohort most directly exposed to the consequences. The reform package is not merely targeting the wrong source of the integrity problem the Government has diagnosed; it is also repeating, in real time, the implementation mistake that the Government’s own evidence base identifies as the source of that problem.

The psychosocial-disability precedent

The Australian disability system has recent experience of what happens when supports are withdrawn from a cohort without the replacement program being operational. The psychosocial-disability cohort experienced precisely this pattern in the transition out of the Personal Helpers and Mentors Scheme (PHaMs), which ceased on 30 June 2019. Of those previously receiving PHaMs support, approximately 20 to 30 per cent had been accepted into the NDIS as of 2018; 30 to 40 per cent had applied to the NDIS and been deemed ineligible. The remainder fell into the transition gap — neither retained on the existing program nor accepted into the replacement. The Productivity Commission’s 2024 report on review and accountability of mental health spending and Community Mental Health Australia’s Remember PHaMs paper (April 2024) both record the documented service-withdrawal pattern that followed and the lasting consequences for service availability.

The lived-experience consequence of the transition gap sits on the public record in the case of David Harris. Mr Harris had paranoid schizophrenia and was an NDIS participant. His NDIS plan expired in April 2019. He was found dead at his Western Sydney home in July 2019. A coronial inquest was announced in December 2020 following a Crikey investigation; NSW Deputy State Coroner Erin Kennedy ruled in October 2022 that the

Submission 2079

An inquest would be dispensed with, on the finding that the NDIA’s decision to extend NDIS plans constituted a satisfactory response. The Coroner’s ruling addresses the institutional response after Mr Harris’s plan expired; it does not address the structural question this submission raises - whether the original transition-gap pattern that left Mr Harris without active supports is one the 9-to-15 cohort under this Bill would be exposed to.

The Harris case has been reported across multiple Crikey articles between May 2020 and October 2022 and is referenced in the Remember PHaMs paper. It is the most prominent documented lived-experience consequence of a transition gap in the recent Australian disability-system record. It is the precedent the 9-to-15 cohort under this Bill will be operating against if the same pattern is permitted to repeat. The stake in the sequencing question is therefore not abstract. The transition-gap pattern this section documents is one in which people lose the supports that keep them safe and well, and the lived-experience record of such gaps includes deaths. The Committee is asked to weigh that the cost of commencing the new regime before the replacement apparatus exists is, on the documented evidence, borne in human terms and not only administrative ones.

The 9-to-15 cohort under the Bill is in a structurally similar position to the PHaMs cohort in 2019. The cohort’s existing NDIS access is to be reassessed against an apparatus that has not been designed, on a timeline that precedes the apparatus’s design, with no replacement program announced. The PHaMs precedent gives the order-of-magnitude picture of what proportion of the cohort can be expected to be deemed ineligible at reassessment - 30 to 40 per cent of those who actually apply, on the PHaMs experience - while the Harris case anchors the lived-experience cost when an individual falls into the gap and the existing safeguards do not catch them. The psychosocial-disability precedent is direct evidence that this kind of transition gap is not a hypothetical risk; it has been documented in both the systemic proportions and the lived-experience tail. The cost is borne by the cohort in service withdrawal during the apparatus build. The Committee is entitled to require the Department to identify why the 9-to-15 cohort transition will not reproduce the psychosocial-disability transition’s documented service-withdrawal pattern.

Part 2 — The reassessment and renewal provisions apply scheme-wide

The provisions addressed in this part - new s.48A on participant-initiated reassessment and new s.50A on automatic plan renewal - are not specific to the 9-to-15 cohort. They apply scheme-wide, to every existing participant on plan review, of every age. They fall hardest on the cohorts whose support needs fluctuate rather than follow a linear deterioration trajectory: autism at every age, psychosocial disability, intellectual disability with co-occurring conditions, and several physical-impaired categories whose presentation responds to context over time. They therefore operate against me, as an existing 18-plus participant whose support needs have fluctuated across the years I have been on the scheme (Section 1), and would operate against my sixteen-year-old son were he to enter the scheme, as much as they operate against my daughter. I develop the provisions here with my daughter as the worked example because she sits at the intersection of this part and the cohort gap in Part 1 - but the reassessment and renewal architecture is a whole-of-scheme concern, not a 9-to-15 one.

Submission 2079

What new s.48A does to participant-initiated reassessment

A participant’s plan is a living document. As a participant’s circumstances change, their support needs change. The current NDIS Act 2013 provides for the participant to request a reassessment of their plan under section 48(2)(b), with the CEO required to decide on the request within 21 days. The current section does not impose substantive gates beyond the request itself; participant requests are accepted into the reassessment workflow as a matter of course.

New s.48A, inserted by Schedule 1 Part 2 of the Bill, replaces that arrangement with a conditional one. A participant may request a reassessment only if all of the following are satisfied:

 "(a) there has been a significant change to those ongoing support needs of the participant that arise from an impairment in relation to which the participant meets the disability requirements or the early intervention requirements; and
  (b) the significant change to those needs results from either or both of the following:
    (i) an alteration in the participant’s functional capacity in relation to an activity covered by paragraph 24(1)(c) or subparagraph 25(1)(c)(i);
     (ii) an alteration in the participant’s personal or environmental circumstances;
  and
   (c) to the extent that the alteration is as mentioned in subparagraph (b)(i) — the condition in subsection (2) is met in relation to the alteration; and
   (d) to the extent that the alteration is as mentioned in subparagraph (b)(ii) — the condition in subsection (3) is met in relation to the alteration." 

Sub-section 48A(2) — the condition for functional capacity alterations — requires that the alteration in functional capacity is “significant and ongoing”, that it directly relates to a change in existing impairment or arises from a new or acquired impairment, AND that “the participant has experienced a substantial reduction in the participant’s ability to perform daily activities.” Sub-section 48A(3) — the condition for personal or environmental circumstances alterations — requires that there has been “an unanticipated, significant and ongoing alteration” in the participant’s living arrangements, education arrangements, work arrangements, or network for informal support.

The CEO’s decision window is also amended by Schedule 1 Part 2: from the current 21 days to 90 days (item 19, amending s.48(3)). The Explanatory Memorandum names a resourcing-side evidence base for the four-fold increase — that the National Disability Insurance Agency was meeting the existing 21-day legislated decision requirement in only 31 per cent of cases in the December 2025 quarter (Explanatory Memorandum footnote 99). The Explanatory Memorandum does not name a participant-impact evidence base. The change extends the Government’s allowable processing time fourfold in response to the Government’s own processing-capacity shortfall, with the participant whose request is

Submission 2079

pending waiting longer for the substantive reassessment to be triggered. A participant whose support needs have changed in ways the new s.48A gates do accept must still wait up to three months for the CEO to decide whether the reassessment is warranted, before the substantive reassessment can even begin.

A subtler structural concern operates inside the s.48A drafting itself. New s.48A(b)(ii) and new s.48A(3) permit an alteration in the participant’s personal or environmental circumstances to be a trigger for participant-initiated reassessment. New s.9B, however, requires functional capacity to be assessed in the absence of the impact of the participant’s environmental and personal circumstances. The reassessment-trigger threshold and the assessment methodology therefore operate on different conceptions of the same environmental and personal context: changes in that context can re-open the workflow, but the workflow itself proceeds on the premise that the context is stripped out. The asymmetry is not resolved on the face of the Bill, and it carries the same social-model- v ersus-medical-model conflict Section 4 develops in respect of new s.9B.

The combined effect is that participant-initiated reassessment is gated by a higher substantive threshold (“significant”, “ongoing”, and — for circumstances — “unanticipated”) and a substantially longer decision window. A participant whose support needs have changed gradually, or whose changing needs were anticipated as part of the impairment’s known trajectory, does not satisfy the gates. A participant whose circumstances have changed in ways the Act characterises as anticipated — for example, a foreseeable transition from primary to secondary school, or a planned transition from secondary school to post-school education or employment — does not satisfy the s.48A(3) gate, even if those transitions materially affect their support needs.

The drafting disadvantages impairment categories characterised by fluctuating need. Autism is one such category: the DSM-5 describes a condition whose presentation varies across contexts and developmental stages, and the Senate Select Committee on Autism’s 2022 final report, Services, support and life outcomes for autistic Australians, documented the significance of key life transitions for Autistic people and the whole-of-life nature of their support needs (Chapter 4, “Key life transitions”). Mental health conditions are another such category. So are degenerative impairments, intellectual disability cohorts with co-occurring conditions, and several physical-impairment cohorts whose presentation responds to environmental and contextual factors over time. The provision draws the threshold for participant agency over plan content at a point that excludes precisely the cohorts whose support needs are not static.

For my daughter, the operational consequence is concrete. If her support needs change in ways the Act’s drafting treats as anticipated — for example, if the transition from primary to secondary school is held to be an anticipated change, even where the actual support implications of that transition were not foreseeable in their specifics — she cannot request a reassessment, even though her plan no longer matches her needs. The provision moves agency from the participant to the CEO at exactly the point at which the participant is in the best position to identify a change.

Submission 2079

The FCA-to-plan update pathway and the structural mismatch in new s.48A

Section 2

This submission describes the two-pass process the NDIS assessment pathway was designed to complete: an initial plan provides basic supports and funding for a functional capacity assessment; the FCA then identifies the full range of supports the person actually needs; a plan update is actioned to incorporate those identified needs.

This is the designed mechanism for getting from a generic first assessment to an accurate individual-needs picture. It is not overreach - it is completion of the pathway the scheme’s own architecture built in.

New s.48A disrupts this mechanism at its core. To request a participant-initiated reassessment, a participant must demonstrate a “significant change” to ongoing support needs arising from either an alteration in functional capacity or an alteration in personal or environmental circumstances. A plan update following a completed functional capacity assessment is neither. It is not a change to support needs - it is a better identification of support needs that were always present and always stable but were never accurately measured in the first place. The first-pass plan was a starting point. The FCA was the designed mechanism for completing the measurement. The second-pass plan update is what completion looks like. New s.48A re-characterises that completion as a “change” event requiring substantive gating - “significant,” “ongoing,” and, for functional capacity alterations, a “substantial reduction in the participant’s ability to perform daily activities” — where the gate is calibrated to a deterioration model that does not describe what a post-FCA update is doing. It is not a threshold that is merely hard to meet. It is structurally wrong for what the FCA update is.

The deterioration model — support needs increasing in a linear, observable, sustained trajectory — is appropriate for some progressive degenerative conditions and structurally ill-fitted for autism and psychosocial disability. Autistic support needs are episodic and context-dependent: they respond to Autistic burnout, to changes in the masking and compensatory strategies that make a person appear more functional than they are, to life-stage transitions, and to changes in the external supports sitting alongside the plan. A participant whose support needs have increased gradually — not dramatically enough at any single point to satisfy the “significant and ongoing” threshold — does not satisfy s.48A(2)(b). A participant requesting a plan update that reflects the completed FCA rather than a changed needs profile does not satisfy s.48A(2)(a)’s framing as a “significant change.” Both of these participants are using the scheme as it was designed; both are excluded from the reassessment mechanism the Bill provides.

The autism-specific barriers to completing the FCA in the first place compound this. Section 2 sets out the personal account of those barriers - the functional obstacle that arises from uncertainty about what the assessment will involve, with whom, in what order, where; the need for advance procedural clarity rather than a generic explanation; and the internalised stigma that suppresses legitimate scheme use at scale, so that low FCA completion rates are read as evidence of adequately met needs rather than evidence that the rhetoric has succeeded in suppressing the signal. The structural consequence for this

Submission 2079

section’s purposes is this: the participants most affected by the post-FCA update gate are the same participants least likely to have completed the FCA in the first place, because the barrier to completing the FCA and the barrier to requesting a subsequent plan update are two instances of the same structural mismatch. The suppression is self-reinforcing. Disney et al.’s 2025 finding — measurably lower NDIS eligibility rates for applicants in lower socioeconomic areas across 705,594 applications between 2016 and 2022 — is the quantitative counterpart: the suppression of legitimate FCA use is not randomly distributed.

The formal external support change gap in s.48A(3) is a further structural problem. The environmental-circumstances gate covers changes to living arrangements, education arrangements, work arrangements, and informal support networks. It does not cover changes to formal external government programs on which the participant relies — Thriving Kids, school-based disability support services, the $472 million in disability loading to schools that the 2026 Budget simultaneously cut, or state-based foundational supports programs. The Bill’s design premise is that foundational supports outside the NDIS will carry more of the weight. If those supports are withdrawn, underfunded, or unavailable in the participant’s geographic area, no review pathway under s.48A(3) opens. The same gap applies where a foundational support technically exists but does not deliver the reasonable-and-necessary supports the participant actually requires: block-funded inadequate-support was the criticism of the pre-NDIS disability system that motivated the scheme’s individualised-plan architecture in the first place, and the Bill recreates that pattern by treating the existence of a foundational support as sufficient for s.48A purposes, regardless of whether it delivers what the participant requires. A participant whose plan was constructed on the assumption that school-based support, a Thriving Kids program, or a community service would carry part of the load — and who then loses that external support, or finds it is provided in a form that does not meet their needs — has no mechanism to trigger a reassessment on that basis. The Bill assumes foundational supports are adequate and durable; its own review architecture provides no means of testing that assumption.

The Royal Commission into Violence, Abuse, Neglect and Exploitation of People with Disability’s Volume 6 Recommendation 6.34 specifically identified health navigators to assist people with disability in accessing and navigating the NDIS, including the functional capacity assessment and plan-development process. That recommendation directly addresses the FCA completion gap — the absence of a support that explains what the FCA involves and why completing it matters, in terms accessible to the participant at the moment of maximum executive depletion. No health-navigator program is in the Bill or in Budget Paper No. 2. The Minister’s Second Reading Speech of 14 May 2026 invokes the Disability Royal Commission as a justification for the Bill. The FCA completion gap the Royal Commission identified at Recommendation 6.34 is entrenched by the Bill’s s.48A architecture rather than addressed.

There is a mechanism for reducing NDIS cost that this Bill bypasses entirely, and that the Bill’s own framing implicitly relies upon: building the foundational-supports ecosystem

Submission 2079

outside the scheme so that participants naturally reduce their NDIS footprint as adequate external supports become available. A participant who currently relies on the NDIS for supports that a functioning school-based disability service, an accessible community program, or a health navigator could provide will, if those services become available and adequate, draw on them rather than the NDIS. The existing s.34 reasonable-and-necessary test already enables this: it requires the CEO to be satisfied that a support is most appropriately funded through the NDIS rather than through other available systems.

Building the foundational-supports ecosystem first makes the existing s.34 test the cost-reduction mechanism, applied at the individual level, without an eligibility-gate amendment. Restricting access at the eligibility gate before the ecosystem is funded, named, and operational removes participants from the scheme without providing the alternatives the cost-reduction logic requires. Sequencing the access restriction ahead of the ecosystem is not cost reduction. It is cost transfer — to the health system, the education system, families, and the support networks that will bear the weight the scheme no longer carries. That this is the structural critique Schedule 1 cannot answer on its own terms is the heart of the case for Schedule 1’s omission.

What new s.50A does to plan renewal

The Bill’s new s.50A, inserted by Schedule 1 Part 5, introduces automatic plan renewal “by force of this section” on the end date of a participant’s old framework plan. The renewed plan is “the same as the old plan” except that:

• (a) the end date for the new plan is the 12 month anniversary of the end date for the old plan;
• (b) the new plan does not include any one-off funding for a support that was included under the old plan;
• (c) any other alteration determined in accordance with an instrument made under subsection (3).

Sub-section 50A(3) is a Ministerial legislative-instrument power: the Minister may, by instrument, determine further alterations the renewed plan will incorporate beyond the (a) date-shift and the (b) one-off-funding drop. The Explanatory Memorandum Note 1 indicates that the examples contemplated include the dropping of temporary variations (such as a variation that responded to an unexpected medical event affecting a participant’s carer) and the dropping of time-limited funding to the extent it has already been used. The scope of (c) is therefore delegated to Ministerial discretion to be exercised after the Bill commences.

Sub-section 50A(4) confirms that the new plan does not require a new statement of participant supports to be prepared with the participant and approved by the CEO, and that the making of the new plan does not involve the making of any reviewable decision.

The mechanism is, by its own terms, an automatic renewal without participant input, without a fresh statement of supports, and without administrative review rights.

Submission 2079

off funding line — that is, supports such as home modifications, capacity-building activities, or assistive technology purchased on a fixed-cost basis — is dropped on renewal.

The drafting again disadvantages cohorts characterised by fluctuating need, and several concrete examples illustrate the operational consequence. A child wheelchair-user whose plan included a one-off allowance for assistive technology has that allowance dropped on automatic renewal — even though the participant’s growth between plan years routinely requires wheelchair replacement, and the wheelchair replacement is neither a “significant change” in the participant’s condition under new s.48A(2) nor an “unanticipated” change in circumstances under new s.48A(3). The plan auto-renews without the replacement-funding line. A participant whose plan included a one-off allowance for capacity-building or environmental modifications has that allowance dropped on the same basis. Consider a participant whose plan was temporarily varied to add extra paid support because the family member who normally provides their informal care was hospitalised. On automatic renewal under new s.50A that temporary variation is stripped out by default — even if the carer has not recovered and the participant still needs the additional support — because the renewed plan is, by force of the section, “the same as the old plan” minus one-off funding and minus any variation a Ministerial instrument designates for removal. The participant must then re-establish the need through the very s.48A reassessment route the Bill has just narrowed. This is the Explanatory Memorandum’s own worked example (Note 1); the concern is that the removal is automatic and indifferent to whether the underlying need still exists. The renewed plan is operating on the assumption that the participant’s support requirements have not changed in any material way that the participant or the Agency would need to recognise — and that any support requirements which have changed materially can be addressed through the participant-initiated reassessment route under new s.48A, which the new s.48A gates have just narrowed.

The s.50A and s.48A provisions operate together. A participant whose support needs are changing gradually — the autism and intellectual-disability cohorts in particular — has their plan auto-renewed without a participant input mechanism and is gated out of the reassessment workflow that would update the plan. The architecture is silent renewal with restricted reassessment access. Plans drift further from need with each renewal cycle until the change has accumulated to the point of clearing the s.48A thresholds, by which point the gap between the plan and the need is no longer recoverable on the timescale the participant lives in.

My daughter’s exposure

My daughter is fourteen. She is an existing NDIS participant. The supports the scheme provides to her — and the existing s.34 reasonable-and-necessary assessment that produced those supports — are what makes her continued education possible. We made the decision to move her from mainstream schooling to online learning earlier this year. The decision was made on the basis of her support profile and the educational settings available; her NDIS plan was the resource we used to construct an arrangement that her support needs allowed her to engage with.

Submission 2079

The Bill’s provisions land on her in three overlapping ways. New s.24(5), s.25(1B) and s.25A apply to her on her next plan reassessment from 1 January 2028 — she will be reassessed against a permanence-and-treatment test that is clinically incoherent for autism, by an assessment instrument that has not yet been designed, against a “low to moderate support needs” categorisation that the DSM-5 and the National Autism Strategy 2025-31 do not use. New s.48A limits the conditions under which I, as her parent acting on her behalf, can request a reassessment if her support needs change gradually or in ways the Act treats as anticipated — including the transition from her current educational arrangement to a post-secondary one. New s.50A means that, on each plan end date, her plan is auto-renewed without a fresh statement of supports, with any one-off funding dropped, and without my input as her parent or her input as the participant.

She is fourteen now. By the time the new functional capacity regime commences on 1 January 2028 she will be sixteen, and the I-CAN instrument flagged in sector commentary will have come into scope. By the time her plan has been auto-renewed under s.50A once or twice, the plan that originally underwrote her education will have drifted from her needs without a participant-input mechanism to recover the alignment. The s.48A gates may or may not be met when we next attempt to act. The supports that are currently keeping her in education are the supports that, on the Bill’s drafting, would be progressively eroded by mechanism and rendered hard to restore by process.

What the Committee should consider

The 9-to-15 cohort question is, in compressed form, a sequencing question. The Bill commences the eligibility-tightening regime on 1 January 2028 (functional capacity assessment regime per Budget Paper No. 2 line 8929). The replacement supports program for the cohort is not announced. The assessment instrument for the cohort is not named. The clinical-evidence basis for the cohort’s place in the new categorisation is not on the public record.

This section grounds four of the recommendations at Section 16. The clinical-incoherence, apparatus-gap and statutory grounds developed across Sections 3 to 8 are why I ask the Committee to recommend that Schedule 1 be omitted from the Bill in its entirety (Recommendation 1). If the Committee does not adopt that course, the apparatus gap is why I ask that the commencement of the functional capacity assessment regime be made conditional on the assessment instruments being designed, independently validated against the cohorts they will be applied to, and operationally matched by a named, funded replacement program for the 9-to-15 cohort (Recommendation 2). The reassessment analysis in Part 2 is why I ask that new s.48A be omitted or substantially amended to preserve participant agency over plan content (Recommendation 7); and the renewal analysis is why I ask that new s.50A be omitted or substantially amended, so that no plan is auto-renewed without participant input, a fresh statement of supports, and a reviewable decision (Recommendation 8). No existing participant — in the 9-to-15 cohort or in any cohort whose impairment is characterised by fluctuating need — should be reassessed against tests that have not been designed, on the basis of categorisations the clinical

• instrument does not use, through a process that has narrowed their agency over their own plan and dropped components of that plan automatically with each renewal cycle.

Submission 2079

Section 6 — Structural under-subscription of the NDIS across the disabled community, and the funding-below-cost provisions in new s.34A and related sections

The Bill’s targeting logic presents the NDIS as a scheme whose participant count is rising beyond sustainable bounds. The Minister cited, at the National Press Club on 22 April 2026, current participation of “760,000 and rising to well over 900,000” as the basis for the reform package’s eligibility-tightening provisions. The framing presents the rising count as evidence of over-expansion. The framing is wrong on the evidence. The participant count is, on the evidence I set out in this section, an under-count of population need driven by structural access barriers — and the Bill’s package compounds those access barriers rather than relieving them. The same Bill also authorises, in primary legislation, the Minister to set funding for groups of supports below the actual cost of those supports, and to set funding caps that apply regardless of whether the cap meets the cost of the supports in the group. The combination — narrower access and below-cost funding — is structurally inconsistent with the Government’s own published statutory objects for the scheme.

Why the participant count is an under-count, not over-expansion

The current participant count reflects the population that has cleared four sequential barriers: diagnosis, awareness of eligibility, application, and successful planning meeting. Each of those barriers operates with material rejection or attrition that has nothing to do with the underlying clinical need.

Diagnostic access.

For adult autism, private diagnosis in Australia currently costs approximately $2,500 out of pocket, against public-sector waitlists of twelve months or more. The NDIS does not fund diagnostics; the cost falls on the individual or their family. Female and gender-diverse Autistic Australians face on average a later diagnostic age — frequently in the late thirties to forties — because the diagnostic literature on Autistic presentations in women and gender-diverse people has only recently begun to be brought into routine clinical practice. Aboriginal and Torres Strait Islander people face additional structural barriers to culturally appropriate diagnosis. People in remote and very remote areas face fewer clinicians qualified to make the diagnosis. The result is that a substantial share of the Autistic Australian population is undiagnosed at any given time; an additional share is diagnosed only in middle adulthood, decades after the impairment first affected functioning.

Application barriers.

The application process asks the applicant — frequently in the cohort whose impairment includes difficulty with administrative tasks, executive functioning, sustained attention, and written communication — to assemble extensive documentation, to describe their needs in a particular format, and to engage with multiple administrative interfaces. The application process is itself an access barrier for the very cohort the scheme is designed to serve. Eligible people drop out of the process. Eligible people never start the process.

Submission 2079

to successfully completes the application is a sub-set of the eligible cohort, not a representative sample.

Planning-meeting representation. The NDIS planning meeting is, in its current operational form, frequently a one-hour interview in which the applicant or participant must describe their needs to an Agency planner. Autism is, structurally, a presentation in which masking - the conscious or unconscious performance of allistic social behaviour at significant internal cost - is part of how the cohort moves through unfamiliar institutional encounters. A one-hour meeting with a stranger in an institutional setting is a high-masking context. Support needs are systematically under-described in that context, not because the participant is concealing need but because masking is part of the impairment’s interaction with the planning environment. The same dynamic applies, with different mechanisms, to a range of other impairment cohorts. The planning meeting under-measures what it is intended to measure.

Population scale. The Australian Bureau of Statistics 2022 Survey of Disability, Ageing and Carers identifies approximately 5.5 million Australians living with disability - approximately one in five of the population. Of that 5.5 million, the NDIS Scheme Actuary’s most recent published participant count is approximately 760,000 - approximately 14 per cent of the disabled-Australian population. The NDIS is the segment of the broader disability-support landscape that delivers individualised plans for people whose impairment meets the s.24 disability requirements or the s.25 early intervention pathway. Every Australian is eligible to apply if and when they acquire a disability with reasonable and necessary support needs not met through other systems; the eligible population is not a fixed sub-group but anyone whose circumstances bring them within s.24 or s.25. The gap between 5.5 million disabled Australians and 760,000 NDIS participants is not symptomatic of a scheme that has expanded beyond its intended footprint. Within the Autistic cohort the four-barrier picture is sharpest: Australian autism-prevalence estimates range from approximately 2.6 to 3.2 per cent of the population, equating to approximately 700,000 to 850,000 Autistic Australians. Autistic Australians are the single largest impairment cohort among new NDIS entrants (approximately 75 per cent of new entrants, per the Scheme Actuary’s projection cited in ABC reporting on the reform package). Even on a generous interpretation of overlap with other impairment categories, the share of the Autistic Australian population currently in the NDIS is a fraction of the share that would be on the scheme if every eligible person could clear the four-barrier sequence above. The participant count is sub-population scale, not population scale. This same cohort carries the worst documented life outcomes in the disability population: the Senate Select Committee on Autism’s 2022 final report, Services, support and life outcomes for autistic Australians, found that Autistic Australians have a life expectancy more than twenty years shorter than the general population and more than twice the general-population mortality rate - outcomes the Committee characterised as “unacceptably poor.” The cohort the Bill most directly targets for reduced access is, on the Parliament’s own prior findings, among the cohorts most in need of increased support. The targeting logic is inverted.

Submission 2079

The under-subscription reading is reinforced by the international comparator. Australia’s spending on disability and incapacity-related public services, as a share of GDP, sits below the OECD average. Nordic comparators in particular sit well above Australia, but the under-subscription point does not depend on the Nordic comparison alone — the OECD average itself, drawn across thirty-eight member countries with widely differing welfare-system models, is higher than Australia’s. The international-comparative evidence does not support the framing that the Australian scheme has expanded beyond comparator-country norms. The framing that the NDIS is over-expanded does not survive the international-comparative test, on either the broad-OECD or the Nordic-specific reading.

The Medicare comparator the Government itself has invoked

The Treasurer’s 2026 Budget speech on Tuesday 12 May 2026 invoked Medicare as the comparator the NDIS is being brought into line with: that the reform package would ensure NDIS growth is “in line with programs like Medicare.” The comparator does not survive analysis. Medicare growth is need-driven and uncapped; the Bill’s wider reform package, through Budget Paper No. 2’s bilateral funding mechanic, imposes an explicit 8 per cent ceiling on Commonwealth contributions to states from 1 July 2028. Medicare is fully utilised because the access barriers are structurally low — universal bulk-billable diagnosis, near-zero financial barrier to entry, light administrative process. The NDIS is structurally under-subscribed because the access barriers are structurally high — a privately funded adult autism diagnosis costing approximately $2,500 against twelve-month-plus public waits, gruelling application process, masking under-measurement in planning meetings.

The 760,000 current participant count is therefore an under-count of population need driven by the high access barriers — not over-expansion. The growth the Government cites as evidence of a scheme exceeding sustainable bounds is, on this reading, the scheme beginning to reach the eligible population it has structurally under-served since inception: a measure of the scheme working toward its statutory purpose, not failing against it. Capping NDIS growth at 8 per cent makes the scheme structurally less like Medicare, not more. The reform package widens the structural gap between the two schemes rather than narrowing it. The comparison the Treasurer is invoking is one the Bill is moving the scheme away from.

The funding-below-cost provisions: new s.34A, new s.32K(3C), new s.33(2EA) and s.33(2EB), and new s.45C

While the eligibility-tightening provisions in Schedule 1 narrow who is admitted to the scheme, a separate cluster of provisions narrows what people on the scheme actually receive. The Bill creates two parallel below-cost mechanisms — s.34A for old framework plans (Schedule 1) and s.32K(3C) for new framework plans (Schedule 4) — together with funding caps in s.33(2EA)/(2EB) and price ceilings in s.45C. Each operates by a different statutory route to the same effect: funding can sit below the cost of the supports the funding is intended to provide.

Submission 2079

New s.34A - Ministerial determination reducing funding for groups of supports.

Section 34A provides that “for the purposes of ensuring the financial sustainability of the National Disability Insurance Scheme, the Minister may, by legislative instrument, determine: (a) a percentage (lower than 100%) that is the percentage by which a funding component amount for a specified group of supports is reduced while the determination is in force…” The text of the affected plan is not altered; the funding is reduced unilaterally. Sub-section 34A(5) reads:

To avoid doubt, the determination has effect even if the result is either or both of the following: (a) the funding provided under a participant’s plan for a reasonable and necessary support is less than the total cost of the support; (b) the funding provided under a participant’s plan for all reasonable and necessary supports funded under the plan taken as a whole is less than the total costs of the supports.

Sub-section 34A(3) requires the Minister to “have regard to the safety of participants.” That is the only substantive safeguard in the section. The Minister may, by legislative instrument, reduce funding for any group of supports for the purpose of “financial sustainability”, and the reduction has effect even where the total funding under the plan is less than the total cost of the supports the plan funds. The Bill does not define “financial sustainability”: the term that triggers the below-cost power is left to the judgement of the Minister of the day, with no statutory criteria, threshold, or test against which that judgement can be reviewed. Because new s.45C bars registered providers from charging agency-managed or plan- manged participants more than the maximum amount determined, the participant generally cannot lawfully be charged the difference; the operative consequence is that providers withdraw from the below-cost segment and the participant forgoes the support or cannot find a provider willing to deliver it at the funded rate. The narrow exception is self-managed participants, who may meet the difference from their own funds - but self-management is itself inaccessible to many disabled people who cannot administer their own plans regardless of financial means, so the exception both reintroduces the income-based inequity the scheme is meant to avoid and is unavailable to a large part of the cohort. The drafting of new s.34A — with the Bill’s $37.8 billion savings target as its operative context — is itself the evidence that “have regard to the safety of participants” is a procedural rather than substantive constraint. If the safety obligation operated as a substantive constraint, a Bill structured around below-cost funding by Ministerial determination would not be the Bill before the Committee.

New s.32K(3C) - statutory permission for new framework plan funding to sit below cost.

Section 32K(3C), inserted by Schedule 4 Part 2 of the Bill (the new framework planning Part), provides verbatim:

A funding amount specified as mentioned in paragraph (3B)(d) may be more than, equal to or less than the actual cost of providing or acquiring the support or group or class of supports.

Submission 2079

The drafting authorises three states: funding above cost, funding at cost, and funding below cost. In context, the operative state is the third. A scheme designed on the assumption that funding meets the cost of supports has no need for primary legislation authorising funding to sit below cost; the “may be more than, equal to or less than” construction reads as a generic permissive drafting, but it is anchored in the same $37.8 billion savings target the Schedule 1 reforms operationalise. The above-cost and at-cost pathways are operationally inert in a fiscal context structured around savings. The below- cost pathway is the one the drafting was inserted to permit. This is not a Ministerial determination route — it is a built-in statutory permission.

Under the new framework planning regime (commencing 1 April 2027), the NDIS rules-making power for funding amounts by level of need (s.32K(3B)(d)) is exercised in the explicit knowledge — written into primary legislation — that the funding amounts produced may sit below the cost of provision. The Explanatory Memorandum confirms the design intent. The Ministerial below- cost authority under s.34A (which I describe above) applies to old framework plans; s.32K(3C) applies to new framework plans. Together the two provisions cover the full transition: old plans can be reduced below cost by Ministerial determination; new plans can be specified below cost by rule-making from the outset.

New s.33(2EA) and s.33(2EB) — funding caps that apply regardless of cost. Section 33(2EA) authorises the Minister to set, for a support or class of supports, “(a) a maximum amount of funding for the support or supports in the class of supports; or (b) a maximum intensity for provision of the support or supports in the class of supports; or (c) a maximum ratio of worker to participant for provision of the support or supports in the class of supports.” Sub-section 33(2EB) requires the Agency’s funding-component working out to be such that no specified maximum is exceeded. The Note to the sub-sections is the load-bearing text:

"These provisions apply regardless of whether a funding component amount meets the cost of the supports in the group to which it relates."

The Minister can therefore cap funding, intensity, and worker-to-participant ratio for any class of support, and the cap operates regardless of whether the resulting funding meets the cost of the supports. The funding becomes a notional ceiling, decoupled from the cost of provision.

New s.45C — maximum amounts payable for agency-managed or registered-plan- manager-managed funding. Section 45C empowers the Minister to determine, by legislative instrument, “the maximum amount, or the method for determining the maximum amount, for the acquisition or provision of an NDIS support or class of NDIS supports.” Where the determination applies, the Agency “must not pay an amount under the National Disability Insurance Scheme to any person in respect of the acquisition or provision of a support under a participant’s plan if the amount exceeds the maximum amount determined.” An NDIS provider “must not charge more than the maximum amount determined” for the support. Sub-section 45C(9) authorises different provision in relation to, among other matters, “(ii) whether the costs of providing the supports or classes of

Submission 2079

supports are greater due to the remoteness of the location in which they are provided; or (iii) whether the supports or classes of supports are provided remotely or in person.“ The price cap can be set differently for remote and non-remote provision, at the Minister’s discretion.

Read together, the four powers authorise the Minister to reduce funding below cost for old framework plans (s.34A), authorise rule-making to specify funding amounts below cost for new framework plans (s.32K(3C)), authorise caps that apply regardless of cost (s.33(2EA)/(2EB)), and authorise price ceilings that providers cannot exceed (s.45C). The Bill removes, in the same legislative move, the structural assumption that NDIS funding will meet the cost of the supports the plan funds — across both the old framework plan regime (which existing participants are on) and the new framework plan regime (which all participants transition into after 1 April 2027).

The pay-gap consequence already documented

The pay-gap consequence of NDIS funding sitting below cost is already documented in the sector. The Ability Roundtable’s 2025 workforce remuneration analysis found a 16.6 per cent pay gap for NDIS-funded psychologists against non-NDIS sector psychologist remuneration. The gap exists, in current arrangements, before any of the new Ministerial below-cost powers in s.34A, s.33(2EA)/(2EB), or s.45C is exercised. The Bill provides statutory authority to extend and entrench the existing gap by Ministerial determination, with the participant carrying the cost difference through reduced support quality, reduced support intensity, or the withdrawal of providers from the regulated market segment for the support in question.

The Productivity Commission’s 2011 and 2017 reports flagged thin markets, workforce capacity, and market stewardship as launch-era risks the scheme would need to manage to deliver on its statutory objects. The academic literature between 2013 and 2024 has confirmed quasi-market failure in specific support segments and geographies. The e61 Institute economist Pelin Akyol, in advance of her Economic Society of Australia (New South Wales) seminar on Wednesday 27 May 2026 titled “More Quasi than Market: Price Caps and Market Design in the NDIS”, has framed the question in these terms: “While much of the public debate has focused on tighter eligibility and reduced plan funding, a third lever has received far less attention: the design of the market itself … whether the NDIS functions as a competitive market, or whether price caps are doing most of the heavy lifting.” The credentialed-economics framing converges on the same architecture this section identifies. Schedule 1’s below-cost funding architecture imposes price ceilings that operate in those markets without addressing — and indeed exacerbating — the underlying thin-markets problem. The integration point — that Schedule 1’s funding- below-cost architecture actively worsens the conditions Schedule 2 needs to remedy — is developed in Section 11.

The combination: narrower access and below-cost funding

Section 3 covered the clinical-evidence vacuum in the permanence-and-treatment test. Section 4 covered the diagnostic-language vacuum that the eligibility-targeting depends on.

Submission 2079

Section 5 covered the assessment-instrument vacuum and the reassessment-and-renewal architecture that disproportionately disadvantages cohorts with fluctuating need. This section adds the under-subscription reframe and the below-cost funding architecture. Read together, the four sections describe a Bill that:

• Narrows access against impairment cohorts the existing scheme is already under-reaching;
• Does so using a test that has no peer clinical body endorsement and that is structurally incoherent for those impairment categories;
• Categorises the cohorts affected — under Thriving Kids’ 0-to-8 descriptor and, scheme-wide from 1 January 2028 under new s.9B, every age cohort the new functional capacity assessment regime applies to — using diagnostic language the operative clinical and statutory documents do not use;
• Operates the functional capacity assessment on an instrument that has not been designed;
• Restricts the agency of existing participants to update their plans as their needs change;
• Removes administrative-review rights at the plan-renewal point under new s.50A(4)(b);
• Subordinates the substantive reasonable-and-necessary obligation to financial sustainability at the Objects clause level (new s.3(1)(d)); and
• Authorises the Minister to fund the supports that remain at amounts below their cost (new s.34A(5), new s.32K(3C), new s.33(2EA)/(2EB), new s.45C).

These are not independent design choices. They operate together as a coordinated reorientation of the scheme. A participant whose access has been retained under the new permanence-and-treatment test, who is reassessed under an instrument the apparatus has not yet produced, whose plan is renewed under new s.50A without participant input and without a reviewable decision to challenge, whose one-off supports and time-limited supports are dropped on renewal, whose remaining funding has been reduced by Ministerial determination under new s.34A below the cost of the supports, and whose foundational-supports alternative has not yet been named for the relevant age cohort — still has a plan on paper. In substance, the plan no longer secures the supports it exists to provide: it funds less than the supports cost, the supports are delivered in markets the Bill has further thinned through below-cost pricing under new s.45C, the assessment apparatus is operating against a population the scheme has been structurally under-reaching for the lifetime of the scheme to date, and the avenues through which the participant could have challenged any of these outcomes have been progressively narrowed across Schedule 1, Schedule 3, and the Objects clause amendment together.

That is the structural argument for the omission of Schedule 1. It is not an argument that any single provision can be amended into coherence. The provisions operate together — and the asks at Section 16 reflect that the omission of Schedule 1 is the only remedy adequate to the architectural state of the package as introduced.

Submission 2079

Section 7 — Statutory inconsistency and discrimination

Sections 3 through 6 set out the substantive evidence base for the omission of Schedule 1 — the clinical-evidence vacuum, the diagnostic-language critique, the assessment- instrument and reassessment-and-renewal architecture, and the structural under- scription paired with the funding-below-cost provisions. This section turns to the statutory and treaty footing. The Bill amends seven load-bearing provisions of the NDIS Act 2013, set out below as seven pillars and developed in turn: Pillar 1 — the Objects clause itself at new s.3(1)(d); Pillar 2 — the new statutory definition of functional capacity at new s.9B; Pillar 3 — the new permanence-and-treatment gate at new s.24(5) and new s.25(1B); Pillar 4 — the Commonwealth-vs-State law conflict that gate produces against state mental-health-legislation rights to refuse treatment; Pillar 5 — the wholesale removal at Schedule 1 Part 6 of the s.31 Principles relating to plans; Pillar 6 — the removal of administrative reviewability at plan renewal at new s.50A(4)(b); and Pillar 7 — the regressivity codification at new s.25A(2). Each is structurally inconsistent with the Australian disability framework’s foundational instruments: the Disability Discrimination Act 1992, the United Nations Convention on the Rights of Persons with Disabilities to which Australia is a State Party, the National Agreement on Closing the Gap, and the federalism architecture of the state and territory mental-health framework.

The current NDIS Act 2013 at s.3(1)(a) commits the scheme to “give effect to Australia’s obligations under the [CRPD].” The seven pillars work against that obligation, and against CRPD Articles 1, 5, 12 and 19 in particular. They are not seven independent technical issues. They operate together.

Pillar 1 — The Objects clause amendment: new s.3(1)(d)

The NDIS Act 2013’s Objects clause at s.3(1) sets out the things the scheme is committed to doing. The current s.3(1)(d) reads: “provide reasonable and necessary supports, including early intervention supports, for participants in the National Disability Insurance Scheme.” The Bill amends this provision at item 60 to read:

 "> (d) provide NDIS supports for participants in the National Disability Insurance

Scheme that are reasonable and necessary, so far as is consistent with the financial sustainability of the scheme; and“

Two changes are operating inside the amendment. First, the words “including early intervention supports” are removed from the foundational statement of the scheme’s purpose. Second, the phrase “so far as is consistent with the financial sustainability of the scheme” is added as a subordination clause.

The first change is not a technical edit. The current Act treats the disability-requirements pathway under s.24 and the early intervention pathway under s.25 as structurally distinct routes into the scheme, each grounded in its own access test, each named separately in the Objects clause. Section 25 was not an oversight in the original drafting. It is the statutory expression of the framework’s recognition that early intervention is “likely to benefit the person by reducing the person’s future needs for supports” — to use the language of s.25(1)(b) of the current Act. The Bill removes early intervention from the

Submission 2079

foundational statement of the scheme’s purpose at the same time as it imposes, under new s.25(1B), the same permanence-and-treatment test that gates the disability-requirements pathway. The s.25 pathway is, in legal form, retained; in substantive form, it is collapsed into the s.24 pathway, gated by the same clinically incoherent test I described in Section 3, and removed from the Objects clause that names what the scheme is committed to providing.

The second change is the more structurally significant. The current Objects clause sets out coordinate principles — each grounded in s.3(1) without one being made conditional on another. The Bill makes financial sustainability a precondition to the principle of reasonable-and-necessary supports. The substantive support obligation now sits behind a fiscal-sustainability gate inside its own statutory home. New s.17B compounds the alteration by introducing “Principles relating to scheme sustainability” the CEO MUST have regard to when performing functions under the Act, including the sub-section 17B(3) principle that “Participants should be responsible for their day-to-day living costs, including day-to-day living costs incurred whether or not a person has a disability.” The combination — Objects clause subordination at s.3(1)(d), and operational direction at s.17B — restructures the scheme’s purpose around fiscal targets in a way the current Act does not.

The s.3(1)(d) amendment also creates an internal contradiction with s.3(1)(a), which is not amended. Section 3(1)(a) of the current Act commits the scheme to “give effect to Australia’s obligations under the [CRPD].” Australia’s obligations under the Convention are not contingent on the State Party’s domestic fiscal circumstances. The Bill leaves s.3(1)(a) in force while introducing, in the adjacent paragraph of the same clause, a fiscal-sustainability subordination of the substantive support principle. The two principles are now operating against each other inside the same Objects clause. The Bill’s drafting has not addressed the contradiction.

A separate concern attaches to “financial sustainability” itself. The Bill imports the phrase as a substantive criterion governing the scheme’s support obligation, and uses it again at new s.34A(3) (where the Minister determines below-cost funding reductions for the purpose of “financial sustainability”) and at new s.45C(17) (where the Minister has regard to “the need to ensure the financial sustainability” of the Scheme when setting maximum amounts). The Bill does not define what “financial sustainability” means, how it is measured, or against what evidence base it is established that the Scheme is not currently financially sustainable. Sustainability is a load-bearing operative term at three points across Schedule 1 and Schedule 3, and is undefined at all three. A statutory criterion that operates against participant access and provider pricing — but is not anchored in a measurable threshold or an evidentiary test — is not a substantive constraint on the Minister’s discretion; it is a rhetorical authorisation for it. The undefined trigger also carries no defined process: the Bill does not specify, when a sustainability determination is made, which supports, which cohorts, which age groups, or which participants (new entrants, existing participants, or the scheme as a whole) are affected, or on what basis those choices are made. A participant cannot know whether or when a determination might

Submission 2079

reduce the funding their plan depends on, and a provider cannot know whether the market segment they operate in will remain viable. The instability the undefined power introduces is itself a harm — it removes the stability disabled people need to plan their lives and the predictability providers need to stay in thin markets, and so compounds the thin-markets problem Section 6 and Section 11 develop.

Pillar 2 — The functional capacity definition: new s.9B against the CRPD Article 1 social model

The Bill inserts a new statutory definition of functional capacity at new s.9B. The provision reads verbatim:

 "(1) A person’s functional capacity, in relation to an activity, is the person’s ability

to undertake the activity: (a) without assistance from other people, assistive technology or modifications; and (b) in a context that excludes, as far as possible, the impact of the person’s environmental and personal circumstances.“

This is a hypothetical-ability test. The person’s functional capacity is measured by the activity they could undertake in a context that strips out the very supports they currently use and the environmental and personal circumstances they currently live in. The threshold against which the hypothetical ability is to be measured is not in the Bill. New s.9B(2) and (3) delegate the threshold and the methods for applying the definition to NDIS Rules made by the Minister. The Technical Advisory Group provisioned in Budget Paper No. 2 line 8919, with $3.3 million in 2026-27, is the body now beginning that work. The threshold the gate will operate against is, at the date of this submission, undesigned.

The methodology is structurally incompatible with the CRPD Article 1 definition of disability that Australia, as a State Party, is obligated to give effect to under NDIS Act 2013 s.3(1)(a). Article 1 of the Convention reads:

"Persons with disabilities include those who have long-term physical, mental,
 intellectual or sensory impairments which in interaction with various barriers may hinder their full and effective participation in society on an equal basis with others." 

The Article 1 definition is explicitly relational. Disability is constituted by the interaction between impairment and barriers — the social model. The framework recognises that the same impairment, in a different supports-and-environmental context, produces a different disability experience. The framework does not separate impairment from context; it identifies context as constitutive of disability. The 2025-2026 National Agreement on Foundational Supports uses the same social-model framing in its Schedule A definition of disability — the Commonwealth has put the social model on the public inter-governmental record three months before the Bill was introduced.

Submission 2079

New s.9B operates in the opposite methodological direction. Functional capacity is measured by stripping out the supports and the environmental context — exactly the factors the CRPD identifies as constitutive of disability. The Bill is, in effect, asking what the person’s ability would be in a counterfactual world in which supports and environmental adaptations did not exist. The CRPD does not recognise that counterfactual as the relevant frame for understanding disability. NDIS Act 2013 s.3(1)(a) commits the scheme to giving effect to the CRPD. New s.9B amends the scheme in a way the CRPD does not support, and in a way the Commonwealth’s own Foundational Supports Agreement does not support either.

The s.9B methodology has a further substantive problem beyond the framework conflict: it builds a catch-22 into the access test. Because functional capacity is measured with the person’s current supports stripped out, a participant whose needs are well-met by effective funded supports will measure as having higher functional capacity than they actually have unsupported — and that higher measured capacity is precisely what can cost them the supports on their next plan review. The supports that are working are the reason the test concludes the supports are not needed. A perverse incentive is built into the statutory definition: people who currently receive effective supports score higher on functional capacity and may lose those supports on the next plan review, because the supports themselves have raised the measured capacity. The fluctuating-need cohorts I described in Section 5 — Autistic Australians, people with intellectual disability and co-occurring conditions, people with mental health conditions, people with degenerative impairments — are disproportionately exposed to this dynamic, because the funded supports they currently receive are what is keeping their measured functional capacity at the higher level.

This critique has been advanced on the public record by the Villamanta Disability Rights Legal Service in its community legal forum on the Bill delivered on 20 May 2026 by Senior Lawyer Elise Almond, Lawyer Kelly Wilkins and Law Clerk Trinity Ford. The legal-service slide deck identifies new s.9B as the operative provision that strips out the supports-and-environment context the CRPD Article 1 definition of disability identifies as constitutive of the disability the scheme is designed to support. The substantive critique I set out in this section is on the legal-service public record.

New s.9B is also where the Government’s “original intent” rhetoric is most clearly contradicted by its own drafting. The Government frames the reform package as restoring the scheme to its original intent. But s.9B does not restore an original definition — it introduces a new one. The original NDIS Act 2013 contains no supports-stripped, environment-excluded hypothetical-capacity test; that methodology is created by this Bill. Changing how disability is defined and how functional capacity is assessed is a more fundamental departure from the scheme’s original design than any change in participant numbers, because it changes the thing the numbers are counting. Section 6 addresses the participant-number framing and Section 8 sets out the four primary-source venues in which the “original intent” rhetoric appears; the point here is narrower and structural — a

Submission 2079

Bill that genuinely returned the scheme to its founding definition of disability would not need to write a new definition of functional capacity into the Act.

Pillar 3 — The permanence-and-treatment gate at new s.24(5) and new s.25(1B): the rhetoric is permanence, the operational mechanism is severity-by-cohort-descriptor

The Bill introduces a new permanence-and-treatment gate at item 64 (new s.24(5)) for the disability-requirements pathway and at item 70 (new s.25(1B)) for the early-intervention pathway. The operative text imposes three conjunctive requirements:

 "(a) the person has undertaken all appropriate treatment for the impairment or impairments (if any); and (b) any other treatment is unlikely to materially improve, reverse, or alleviate the impact of, the impairment or impairments; and (c) the impairment or impairments are likely to persist for the person’s lifetime." 

I addressed the clinical incoherence of the treatment-exhaustion limb in Section 3 — autism, intellectual disability, Down syndrome, cerebral palsy, motor neurone disease, and the broad cohorts for whom remedy or alleviation is not the operative clinical paradigm cannot meaningfully exhaust treatments the relevant clinical evidence base does not name. The limb assumes a treatment model that does not apply to these conditions: there is no treatment that “reverses” autism or Down syndrome, and for a progressive condition such as motor neurone disease the operative clinical paradigm is symptom management, not remedy. For my own family this is not abstract — there is no course of treatment my children or I could “exhaust“ to satisfy s.24(5)(a), because the conditions the test is applied to are not the kind the test imagines. This sub-section addresses the second statutory inconsistency the gate carries: the gap between the Government’s permanence rhetoric, the criterion in the existing access tests in the Act, and the Bill’s operational exclusion mechanism for Autistic applicants.

The permanence rhetoric is a stable framing the Government has restated across multiple primary-source venues — the Minister’s address at the National Press Club on 22 April 2026, the Treasurer’s Budget Paper No. 2 measure text on 12 May 2026, and the Department of Health, Disability and Ageing’s signed ministerial response MC26-008601 on 25 May 2026 (Section 8 sets out the full venue map and the “original intent” drift). Across all three venues the Government’s stated rationale for the access-tightening package is the protection of supports for “people with permanent and significant disability.” The Chair’s Foreword to the House Standing Committee on Health, Aged Care and Disability’s December 2025 No child left behind report reproduces the same framing in the phrase “all people with severe disability.” Permanence is one half of the rhetoric. Severity is the other half. The two are run together in the Government’s public formulation; they do not appear together as criteria in the Act.

The naming of these instruments is worth pausing on, because in this reform the names and the effects point in opposite directions. A report titled No child left behind accompanies a reform that leaves the 9-to-15 cohort with no announced replacement

Submission 2079

program and narrows the early-intervention pathway thousands of children rely on. A Bill titled the National Disability Insurance Scheme Amendment (Securing the National Disability Insurance Scheme for Future Generations) Bill 2026 delivers a $37.8 billion reduction over four years and removes an estimated 160,000 people from the scheme. The titles convey protection and continuity; the operative provisions deliver contraction. The Committee is entitled to weigh the operative text against the framing the names invite, and to give the text the greater weight.

The statutory architecture the Bill is amending does not use “significant” or “severity” as a criterion. NDIS Act 2013 s.24, the disability-requirements pathway, refers to impairments that “are, or are likely to be, permanent” and to functional capacity that is “substantially reduced” in one or more of communication, social interaction, learning, mobility, self-care or self-management. Section 25, the early-intervention pathway, in its current form does not require permanence at all in the under-six developmental-delay limb — its access test turns on whether early intervention is “likely to benefit the person by reducing the person’s future needs for supports.” Neither pathway uses “severity” as a criterion at any age. “Substantially reduced functional capacity” attaches to the functional impact in the named areas, not to the impairment itself — a person with a mild impairment can meet the s.24 test if functional capacity is substantially reduced in the named areas, and a person with a severe impairment can fall outside the test if functional capacity is not substantially reduced in those areas. Severity language is a non-statutory rhetorical layer the Government has applied in different formulations across all of the venues named above. The Bill carries the rhetorical layer forward without grounding it in any new statutory criterion. Legislation that determines access to lifelong supports should rest on criteria established by appropriately qualified clinical authorities and developed with the affected communities genuinely consulted — not on a rhetorical framing imported into statute without a clinical or evidentiary foundation. New s.24(5) and new s.25(1B) introduce the permanence-and-treatment-exhaustion gate; they do not introduce a severity criterion. The operative text speaks to permanence, treatment exhaustion, and material improvement — not severity.

This is where the autism-is-permanent point bites. Autism is defined under the DSM-5 as a neurodevelopmental condition characterised by persistent differences across the lifespan, with diagnostic criteria explicitly requiring symptoms to be present from the early developmental period. The condition is, by definition, lifelong. An Autistic applicant categorically passes the permanence limb of the s.24 access test. An Autistic child seeking early-intervention access also categorically passes the permanence limb the Bill imports into s.25 via new s.25(1B). For autism the formal diagnosis is itself the confirmation, by a qualified clinician against the DSM-5 criteria, that the impairment is lifelong — so the permanence limb is already satisfied at the point of diagnosis, and the treatment-exhaustion overlay adds a second gate on top of a clinical determination the scheme already accepts. The campaign to “restore the NDIS to its original intent” of “permanent and significant disability” cannot exclude Autistic applicants on the permanence limb of the gate the Bill is constructing. Autism is permanent.

Submission 2079

Government’s rhetoric and the Bill’s permanence criterion do not, by their own terms, exclude Autistic applicants from access.

The Bill’s operational mechanism for narrowing Autistic access lies elsewhere. The Department’s MC26-008601 paragraph 3 names the Thriving Kids cohort it is rolling out from October 2026 as “children aged eight and under with developmental delay and/or autism with low to moderate support needs, as well as their families, carers and kin.” The cohort descriptor “low to moderate support needs” is the operational filter. The new functional-capacity assessment at s.9B — the methodology I addressed in the preceding sub-section, with the threshold delegated to NDIS Rules under s.9B(2) and (3) and the Technical Advisory Group provisioned in Budget Paper No. 2 line 8919 now beginning the design work — is the instrument that will operationalise the descriptor. The cohort- redirection mechanism is the composition of two architectural elements: a cohort descriptor that tags some Autistic children as having “low to moderate support needs”, and a functional-capacity-test threshold below which they will not access individualised supports through the NDIS and will be redirected into the state-delivered Thriving Kids program in lieu of NDIS plans. The same composition will operate scheme-wide from 1 January 2028 against existing participants on plan review, under new s.9B applied through new NDIS Rules.

The operational mechanism is severity-based. The cohort descriptor is expressed in the language of “support needs.” The assessment instrument will measure functional capacity against a threshold the Bill does not name. Either the descriptor or the threshold operates as a severity gate by another name. Severity has no statutory grounding in s.24 or s.25 — The criterion in the Act is permanence plus substantially reduced functional capacity in the named areas, not severity of impairment, and the Bill does not add severity to either access test as an explicit criterion.

The gap between the Government’s stated rationale and the Bill’s operational mechanism is evidence that the case for these provisions has not been honestly made. The stated rationale is permanence; the operational mechanism is severity-by-cohort-descriptor. The rationale cannot exclude Autistic applicants, because they categorically pass the permanence limb. The mechanism does the exclusion work the rationale cannot — through a criterion the Act does not on its face contain, applied via the s.9B functional- capacity test and the “low to moderate support needs” cohort descriptor. A reform genuinely about permanence would not need that machinery; its presence is the tell that the operative purpose is not the stated one.

I am not asking the Committee to resolve this by writing a clean severity criterion into the gate. That would entrench the exclusion, not cure it. The point is the opposite: the permanence-and-treatment gate at new s.24(5) and new s.25(1B), the s.9B definition that operationalises it, and the cohort-descriptor architecture that targets it are inconsistent with the CRPD and the Disability Discrimination Act 1992, and abandon the scheme’s statutory purpose under s.3 — and should be omitted on those grounds, whatever basis the rhetoric is ultimately said to rest on. The scheme should continue to do what the Act

Submission 2079

has always required: provide reasonable and necessary supports, including early intervention, to people whose impairment is permanent and whose functional capacity is substantially reduced, without a treatment-exhaustion overlay the clinical evidence base does not support.

Section 4 addresses how the “low to moderate support needs” cohort descriptor maps onto the diagnostic-language critique — the Department’s own letter at MC26-008601 paragraph 2 acknowledges that “mild” and “severe” autism terminology “can misrepresent” the support needs and capabilities of Autistic people, in the paragraph immediately preceding the paragraph that operationalises the equivalent “low to moderate support needs” descriptor as the Thriving Kids cohort filter. Section 8 sets out the full four-venue map of the Government’s permanence-and-original-intent rhetoric and the way the Bill operates against the existing statutory text.

Pillar 4 — The Commonwealth-vs-State law conflict at new s.24(5)(a) and new s.25(1B)(a) against state mental-health-legislation rights to refuse treatment

The permanence-and-treatment gate carries a further axis of statutory inconsistency that surfaced on the parliamentary record on 15 May 2026. Professor Jennifer Smith-Merry of the Centre for Disability Research and Policy at the University of Sydney testified to the Joint Standing Committee on the NDIS at the inquiry’s second public hearing in Sydney. Professor Smith-Merry’s research is funded by an Australian Research Council Laureate Fellowship and is based on qualitative interviews with 86 NDIS participants, family members and stakeholders across Australia, examining experiences of the scheme for people with psychosocial disability. Her evidence on the Hansard record:

“I am concerned about one particular change, and that is related to the need to exhaust all treatment or exhaust all other avenues first. What we’ve seen already in the operation of the NDIS — and we've done significant work on access and eligibility within the current rules. The processes involved in that are that people may be denied access, particularly under psychosocial disability, if they haven’t tried all medical treatments to address their disability. That can even include electroshock therapy or other types of therapies. Now, that conflicts with our state based legislation. Under our state based mental health legislation, you’re allowed to not take up a treatment.”

In response to Senator Steele-John’s direct question on whether this meant a conflict between the proposed change and the rights patients have to refuse care under State and Territory mental-health statutes, Professor Smith-Merry answered: “You’re allowed to refuse treatment. That’s a right that we have — that people have. That would conflict with NDIS eligibility criteria, where you had to be able to show that you had tried all alternative treatments.”

The State and Territory mental-health statutes each, in different statutory form, provide a right to refuse treatment in defined circumstances. The right is foundational to the regulatory framework State and Territory parliaments have built around mental-health

Submission 2079

treatment over decades. It operates at the level of the individual’s lawful capacity to decline a clinical intervention, and it does not turn on the medical considerations the narrow s.25A(3)(a) carve-out from the treatment-exhaustion limb contemplates.

The federalism axis of the inconsistency is structurally distinct from the CRPD, DDA and Closing the Gap inconsistencies the other pillars of this section address. The Bill amends Commonwealth legislation in a way that conditions access to a Commonwealth- admitted support scheme on the applicant having undertaken treatment State and Territory legislation provides the applicant with a statutory right to refuse. An applicant with psychosocial disability who exercises the State-law right to refuse treatment categorically cannot satisfy the Commonwealth-law requirement at new s.24(5)(a) and new s.25(1B)(a) that all appropriate treatment be exhausted. The Bill’s drafting does not address whether the Commonwealth provision operates to penalise the lawful exercise of a State-conferred right, or whether the Commonwealth provision is inoperative where the State right is exercised. The s.25A(3)(a) carve-out for treatment the person cannot undertake for medical reasons does not reach the conflict — the State right to refuse is a rights-of-the-person matter, not a medical-reason exception.

Professor Smith-Merry’s evidence was that this is not a hypothetical conflict. Her 86-participant qualitative research has documented psychosocial-disability applicants having been “denied access… if they haven’t tried all medical treatments to address their disability” under the existing NDIS access processes — under arrangements that are operational rules, not primary-legislation criteria. The Bill writes the operational arrangement into the statutory access test itself, at new s.24(5)(a) and new s.25(1B)(a), and applies it across both the disability-requirements and early-intervention pathways. The conflict the existing arrangement has already produced is what the Bill is codifying.

The Joint Standing Committee on the NDIS has Professor Smith-Merry’s evidence on its public record. The conflict between the two pieces of statutory architecture — the Commonwealth’s NDIS Act amendment requiring treatment exhaustion, and the State and Territory mental-health legislation providing the right to refuse treatment — has not been addressed in the Bill the Senate Community Affairs Legislation Committee is being asked to consider, in its Explanatory Memorandum, in the Minister’s Second Reading Speech, or in any of the Government public material the campaign has been able to identify.

Pillar 5 — The s.31 Principles relating to plans: the wholesale removal at Schedule 1 Part 6 against CRPD Articles 12 and 19

The current NDIS Act 2013 at s.31 sets out the Principles relating to plans. These eleven principles, (a) to (k), provide the operational architecture for how plans are prepared, varyed, reassessed and replaced, and how the funding for supports under a plan is managed. The principles state that plan preparation, variation, reassessment and replacement, and funding management, “should so far as reasonably practicable”: be individualised; be directed by the participant; consider and respect the role of family, carers and other persons significant in the participant’s life; recognise and respect the

Submission 2079

relationship between participants and their families and carers; strengthen and build capacity of families and carers to support participants who are children, and adults if the participant and carers agree; consider the availability to the participant of informal support and other support services generally available in the community; support communities to respond to the individual goals and needs of participants; be underpinned by the right of the participant to exercise control over their own life; advance the inclusion and participation in the community of the participant with the aim of achieving the participant’s individual aspirations; maximise the choice and independence of the participant; facilitate tailored and flexible responses to the individual goals and needs of the participant; and provide the context for the provision of disability services, coordinating delivery where there is more than one disability service provider.

Schedule 1 Part 6 of the Bill strikes out the eleven principles wholesale. The Villamanta Disability Rights Legal Service’s legal analysis of the Bill, delivered at the 20 May 2026 community legal forum, identifies that only parts of (c), (ca) and (f) — the family/carer-relationship principles and the community-responsiveness principle — find their way back into various sections of the Bill. The bulk of the architecture is removed. The principles removed are: (a) individualised; (b) directed by the participant; (d) and (da) capacity-building of families and carers to support participants who are children and, with agreement, adults; (e) consideration of informal support and other support services generally available in the community; (g) the right of the participant to exercise control over their own life; (h) advancement of inclusion and participation in the community; (i) maximisation of choice and independence; (j) tailored and flexible responses to the individual goals and needs of the participant; and (k) coordination of disability service delivery. The principles retained in fragmentary form are (c), (ca) and (f) — the family/carer-relationship principles and the community-responsiveness principle.

CRPD Article 12 obligates State Parties to recognise that persons with disabilities enjoy legal capacity on an equal basis with others in all aspects of life. Article 12(3) requires State Parties to take appropriate measures to provide access to persons with disabilities to the support they may require in exercising their legal capacity. Article 12(4) requires that all measures relating to the exercise of legal capacity respect the rights, will and preferences of the person, are free of conflict of interest and undue influence, are proportional and tailored to the person’s circumstances, apply for the shortest time possible, and are subject to regular review by a competent, independent and impartial authority. The s.31 Principles at (a) (individualised), (b) (directed by the participant), (g) (the right of the participant to exercise control over their own life), (i) (maximise choice and independence), and (j) (tailored and flexible responses to the participant’s individual goals) are the operational expression of the Article 12 obligation at the participant-plan level. Their removal from the Act is the removal of the operational architecture by which the scheme has given effect to Article 12.

CRPD Article 19 obligates State Parties to recognise the equal right of all persons with disabilities to live in the community, with choices equal to others, and to take effective and appropriate measures to facilitate full enjoyment of this right and the participant’s full

Submission 2079

inclusion and participation in the community. The s.31 Principles at (e) (consideration of informal support and other support services generally available in the community), (h) (advancement of inclusion and participation in the community), and (k) (coordination of disability service delivery in its community context) are the operational expression of the Article 19 obligation at the participant-plan level. Their removal from the Act is the removal of the operational architecture by which the scheme has given effect to Article 19. The same direction of travel is operating in the Bill’s wider funding architecture: Schedule 1’s funding-below-cost provisions disproportionately affect community-participation supports — The 50 per cent cut to Social, Civic and Community Participation funding the Office of Impact Analysis report surfaced through Guardian Australia on 20 May 2026 (which Section 14 addresses in full) operates directly against the community-inclusion obligation Article 19 imposes. The s.31 Principles removal at the planning-architecture level and the SCCP cut at the funding-architecture level are two instances of the same Article 19 inconsistency. The National Agreement on Foundational Supports 2026-31 Schedule A reinforces the social-model definition and the community-inclusion frame; the Bill removes the operational expression of that frame from the participant-level instrument while retaining the inter-governmental commitment at the Agreement level. The Commonwealth’s own documents are working against each other in the same period — and against the Commonwealth’s CRPD commitment that NDIS Act 2013 s.3(1)(a) requires the scheme to give effect to.

The pattern of which principles the Bill retains in fragmentary form is itself informative. The principles retained in fragments are the family/carer-relationship principles at (c) and (ca) and the community-responsiveness principle at (f). The principles removed are the choice-and-control principles at (a), (b), (g), (i) and (j) and the community-inclusion principles at (e), (h) and (k). The selectivity is significant. CRPD Articles 12 and 19 protect the participant’s own legal capacity, choice and community inclusion. Article 28 frames family and social-protection roles. The Bill removes the Article 12 and Article 19 principles while retaining fragmentary forms of the family/carer-relationship principles. The operational effect, at the participant-plan level, is the substitution of a participant-directed, choice-and-control-based plan-design framework with a framework in which family and carer relationships are recognised but the participant’s own choice, control and community-inclusion rights are not.

Pillar 6 — The removal of administrative review at plan renewal: new s.50A(4)(b) against CRPD Article 12(4) and the DDA review-rights architecture

Schedule 1 Part 5 of the Bill introduces a new s.50A providing for automatic plan renewal by force of law at the end of a participant’s reassessment period. New s.50A(4)(b) reads verbatim:

"the making of the new plan does not involve the making of any reviewable decision."

Submission 2079

The automatic renewal-by-force-of-law produces a new plan, and the new plan is not, on the operative text of the provision, the product of a reviewable decision. Administrative merits-review against a reviewable decision under s.99 of the Act, and the Administrative Review Tribunal pathway that follows from it, do not apply. A participant whose annual auto-renewed plan funds less than the previous plan funded — for example, because the Minister’s s.34A determination has reduced the funding component for a specified group of supports between the old plan and the new plan, or because the Agency has applied the new framework planning regime’s s.32K(3C) below-cost funding amounts to the new plan — has, on the operative text of s.50A(4)(b), no reviewable decision to challenge at the plan- renewal point.

This is incompatible with the CRPD Article 12(4) obligation. Article 12(4) requires that all measures relating to the exercise of legal capacity respect the rights, will and preferences of the person and are “subject to regular review by a competent, independent and impartial authority.” The automatic plan-renewal architecture is a measure relating to the exercise of the participant’s legal capacity in respect of the supports the plan funds. The Administrative Review Tribunal is the competent, independent and impartial authority through which the current Act gives effect to Article 12(4). New s.50A(4)(b) operates that obligation off the table at the annual plan-renewal point — the point at which the participant’s continuing access to scheme supports is being determined under either the s.34A reduced-funding regime or the s.32K(3C) below-cost regime that Section 6 of this submission addresses in full. Villamanta Disability Rights Legal Service’s independent legal analysis at the 20 May 2026 community legal forum identifies new s.50A(4)(b) as the operative-text removal of the review-rights pathway at plan renewal. The legal-service framing is on the public record.

The provision is also structurally inconsistent with the broader Disability Discrimination Act 1992 and Australian Human Rights Commission Act 1986 review-rights architecture the Commonwealth’s anti-discrimination framework operates through. Removing administrative-reviewability at the plan-renewal point — the point at which the participant’s continuing access to scheme supports is being determined — places the review-rights architecture of the Commonwealth’s anti-discrimination framework in tension with the access architecture of the disability supports scheme that anti- discrimination framework is supposed to operate alongside. Section 12 of this submission addresses the parallel review-rights stripping the Bill effects at Schedule 3 Part 1 through the Ministerial pricing-decision-maker model — the legislative-instrument carve-out from administrative merits review removes the review pathway at the pricing-determination point. The two provisions operate together: new s.50A(4)(b) removes review-rights at the plan-renewal point under Schedule 1; new s.45C removes them at the pricing- determination point under Schedule 3. The combined effect is the structural removal of the merits-review pathway at the two points the participant is most likely to need it.

Submission 2079

Pillar 7 — The regressivity codification: new s.25A(2) against the DDA, CRPD and Closing the Gap

I addressed the income-and-geographic regressivity layer of s.25A(2) under the clinical-access frame in Section 3. This section addresses the same provision under the discrimination and statutory-inconsistency frame.

Sub-section 25A(2) provides that, for the purposes of the s.24(5)(a) and s.25(1B)(a) treatment-exhaustion limb, “it is irrelevant whether the person can access the appropriate treatment, regardless of whether the person’s individual circumstances restrict the person from accessing the treatment”, with the Note that “A person’s individual circumstances include the person’s financial circumstances and geographical location.” Sub-section 25A(3) provides a narrow statutory exception only for “appropriate medical treatment the person cannot undertake for medical reasons” — financial inability and geographic inaccessibility are explicitly outside the exception.

This is incompatible with three of the foundational instruments of the Australian disability framework.

The Disability Discrimination Act 1992 at s.6 prohibits indirect discrimination — a requirement that is not reasonable having regard to the circumstances of the case, with which a higher proportion of people without the relevant attribute comply or are able to comply, and which has the effect of disadvantaging people with the relevant attribute. The treatments the test contemplates fall on the individual: adult autism diagnosis in Australia currently costs approximately $2,500 out-of-pocket against a twelve-month-plus public waitlist; full treatment exhaustion can mean tens of thousands of dollars in private therapy and medication trials. The s.25A(2) provision imposes a requirement — undertake all appropriate treatment, in financial and geographic circumstances the provision says do not matter — that disproportionately disadvantages low-income and remote applicants. The reasonableness test under s.6 of the DDA cannot be satisfied by a requirement whose unreasonableness in the relevant circumstances is its explicit statutory point. The DDA at s.4 defines disability without reference to whether treatment has been exhausted; the Bill’s permanence-and-treatment test introduces a treatment-exhaustion requirement into NDIS access alone, which has no foundation in the Commonwealth’s own anti-discrimination definition of disability.

The CRPD at Article 5 obligates State Parties to prohibit all discrimination on the basis of disability and to take all appropriate steps to ensure that reasonable accommodation is provided. Reasonable accommodation under the Convention is defined at Article 2 as “necessary and appropriate modification and adjustments not imposing a disproportionate or undue burden, where needed in a particular case, to ensure to persons with disabilities the enjoyment or exercise on an equal basis with others of all human rights and fundamental freedoms.” A statutory provision that codifies that financial and geographical circumstances are irrelevant to access is a refusal to provide that reasonable accommodation in the design of the access gate itself. CRPD Article 25 (right to health) is a separate right from disability-supports access; the Bill’s permanence-and-treatment test

Submission 2079

crosses two articles in one move by collapsing health-system treatment access into the eligibility gate for disability supports.

The Closing the Gap framework is the third instrument the s.25A(2) provision works against. Aboriginal and Torres Strait Islander people are over-represented in lower-income populations and in remote populations. The cost and geographic barriers s.25A(2) codifies disproportionately apply to Aboriginal and Torres Strait Islander people with disability. The National Agreement on Foundational Supports 2026-31, signed by all eight jurisdictions in early 2026, commits the Commonwealth and the States at clauses 24 and 25 to aligning Foundational Supports design and implementation with the four Priority Reforms of the National Agreement on Closing the Gap — including strengthening partnerships and shared decision-making with Aboriginal and Torres Strait Islander people, building the community-controlled sector, transforming government organisations, and shared access to data and information. The Foundational Supports Agreement is the inter-governmental instrument the Bill’s reform package is rolled into. The Bill amends the access gate of the scheme that sits underneath the Agreement in a direction the Agreement’s commitments work against. In three months the Commonwealth has signed an instrument committing to align with Closing the Gap and introduced legislation that codifies cost and geographic regressivity as features of the eligibility gate. The two Commonwealth documents are working in opposite directions.

The seven pillars operate together

The seven pillars are not seven separable technical defects. Taken together they reorient the scheme’s purpose, its methodology, its access gate, its federalism alignment, its plan-level participant rights, its review pathway, and the equity profile of the access gate. Each offends one or more of the foundational instruments:

• Pillar 1 (Objects clause, new s.3(1)(d)): subordinates the substantive support principle to fiscal sustainability and removes early intervention from the foundational statement of the scheme’s purpose — against the s.3(1)(a) CRPD commitment in the same clause.
• Pillar 2 (functional capacity, new s.9B): strips out the social-model factors the CRPD Article 1 definition identifies as constitutive of disability; the Disability Discrimination Act 1992 s.4 definition is likewise not contingent on a supports-stripped counterfactual.
• Pillar 3 (permanence-and-treatment gate, new s.24(5)/s.25(1B)): the stated rationale — permanence — categorically cannot exclude Autistic applicants because Autism is permanent by definition; the operational exclusion mechanism — severity-by-cohort-descriptor, applied through the s.9B test — has no statutory grounding in s.24 or s.25.
• Pillar 4 (Commonwealth-vs-State conflict, new s.24(5)(a)/s.25(1B)(a)): in direct conflict with State and Territory mental-health-legislation rights to refuse treatment, as Professor Jennifer Smith-Merry’s parliamentary-record testimony to the Joint

Submission 2079

Standing Committee on 15 May 2026 establishes — a federalism inconsistency the Bill’s drafting and Explanatory Memorandum do not address.

• Pillar 5 (s.31 Principles removal, Schedule 1 Part 6): strips out the choice-and- control and community-inclusion architecture by which the scheme has given effect to CRPD Articles 12 and 19 at the participant-plan level.
• Pillar 6 (review removal, new s.50A(4)(b)): operates the CRPD Article 12(4) regular- ter obligation off the table — and sits in tension with the Disability Discrimination Act 1992 and Australian Human Rights Commission Act 1986 review-rights architecture — at the very point the participant is most likely to need it: The renewal at which a Ministerial s.34A determination, or the new framework planning regime’s s.32K(3C) below-cost authority, may have reduced the funded supports.
• Pillar 7 (regressivity codification, new s.25A(2)): puts on the statutory record that access-cost and access-geography barriers are not grounds to consider treatment unavailable — against the Disability Discrimination Act 1992 s.6 prohibition on indirect discrimination, CRPD Article 5, and the Closing the Gap commitments the Foundational Supports Agreement is conditioned on.

The current NDIS Act 2013 at s.3(1)(a) commits the scheme to giving effect to Australia’s crpd obligations. The Bill amends the scheme in seven coordinated moves that work against those obligations — against crpd Articles 1, 5, 12 and 19 specifically — and against the Disability Discrimination Act 1992’s prohibition on indirect discrimination, against the Commonwealth’s own National Agreement on Foundational Supports 2026-31 commitment to align with Closing the Gap, and against the federalism architecture of the State and Territory mental-health statutes the seventh pillar identifies. The provisions in Schedule 1 cannot be amended into compatibility with these instruments by edits to the operative text. The structural reorientation is what is on the face of the Bill. It is the reorientation, not the operative text in isolation, that the Committee is being asked to assess — and the statutory and treaty inconsistency this section sets out is the load- bearing reason for the omission of Schedule 1 that the recommendations at Section 16 ask the Committee to record.

Submission 2079

Section 8 — Direct contradictions with Government-commissioned reviews on Schedule 1

Sections 3 through 7 set out the substantive evidence base — the clinical-evidence vacuum, the diagnostic-language critique, the assessment-instrument and reassessment-and-renewal architecture, the structural under-subscription and below-cost funding provisions, and the statutory-and-treaty inconsistency. This section turns to the review-framework footing on which the Government’s case for Schedule 1 publicly rests.

The Bill’s Explanatory Memorandum and the Minister’s Second Reading Speech of 14 May 2026 identify five Government-commissioned review streams as the evidentiary basis for the reform package. On the evidence in this section, the Bill is either selectively implementing recommendations from those reviews — taking the elements that support eligibility tightening and below-cost funding while contradicting or omitting elements that would not — or operating against the substantive direction of the reviews altogether. The Bill is not what the Government’s own primary documentation says it is.

The Royal Commission claim, against the Explanatory Memorandum

The Minister stated in his Second Reading Speech on Thursday 14 May 2026 that the Bill “takes up recommendations on measures to improve delivery of the NDIS that have been made to government by various independent reviews of the NDIS, including the Independent Review into the NDIS, in 2023, and the Royal Commission into Violence, Abuse, Neglect and Exploitation of People with Disability.” The reference to the Royal Commission is on the Hansard record.

The Bill’s own Explanatory Memorandum, in its Consultation section, identifies the specific recommendations the Bill is said to implement. They are:

• NDIS Review Action 3.1 (consistent assessment) — Schedule 1 Part 1, the functional capacity definition at new s.9B;
• NDIS Review Action 3.9 (the permanence test) — Schedule 1 Part 8, new s.24(5) and s.25(1B), anchored in the Federal Court decision National Disability Insurance Agency v Davis [2022];
• NDIS Review Action 2.15 (alternative supports / cost-shift gate) — Schedule 1 Part 9, new s.25B;
• NDIS Review Recommendation 17 (regulatory model) — Schedule 2;
• NDIS Review Action 11.3 (Minister-as-pricing-decision-maker) — Schedule 3 Part 1;
• the NDIS Provider and Worker Registration Taskforce 2024 — Schedule 2 Part 1, the new s.10C provider definition;
• the Disability Royal Commission final report (2023), including Recommendations 9.4, 10.2, 10.3 and 10.4 — Schedule 2 Part 6, plan management and support coordination commissioning provisions, per Explanatory Memorandum section 5.4.

Submission 2079

The Explanatory Memorandum identifies four Disability Royal Commission recommendation (9.4, 10.2, 10.3, 10.4), all of which inform Schedule 2 Part 6’s plan management and support coordination commissioning provisions. It does not cite any Royal Commission recommendation by number, by content, or otherwise as the implementation basis for any Schedule 1 provision, including the new permanence-and-treatment test in Schedule 1 Part 8, the new functional capacity definition in Schedule 1 Part 1, the alternative-supports gate in Schedule 1 Part 9, the funding-below-cost authorities, or the wholesale removal of the s.31 Principles in Schedule 1 Part 6. The Minister’s claim on the Hansard record that the Bill “takes up recommendations… of the Royal Commission” is substantiated in the Explanatory Memorandum only in respect of four recommendations attached to one Part of Schedule 2; the bulk of the substantive operative architecture of the Bill — the Schedule 1 access-tightening provisions — is not anchored to a single Royal Commission recommendation in the Government’s own primary documentation. I ask the Committee to require the Department of Health, Disability and Ageing to identify, for the public record, any further Royal Commission recommendations the Bill operationalises beyond the four cited in Explanatory Memorandum section 5.4, and the Schedules in which each is implemented. If no further mapping exists, the Royal Commission reference in the Second Reading Speech — to the extent it implies Royal Commission endorsement of the Schedule 1 access-tightening architecture — should be corrected for the parliamentary record.

The substantive force of the contradiction sharpens when the Bill’s narrow citation of Royal Commission recommendations is read against the Government’s own formal position on the 222 recommendations. The Joint Australian, State and Territory Government Response to the Disability Royal Commission, published 31 July 2024, records the Commonwealth’s formal acceptance position on the 222 recommendations the Royal Commission delivered in its Final Report on 28 September 2023. Of the 172 recommendations for which the Commonwealth has primary or shared responsibility, the Commonwealth accepts or accepts in principle 130, notes 6, and continues to consider 36. The Government has, on its own published record, a formal implementation framework for 130 specific Royal Commission recommendations. The Bill cites four of those 130 recommendations as the implementation source for one Part of Schedule 2; it does not cite the remaining 126 as the implementation source for any provision, including the Schedule 1 access-tightening architecture. The position is therefore simple to state: the Royal Commission is invoked at the rhetorical level for the package as a whole, but anchored at the implementation level to four recommendations attached to one narrow component of Schedule 2.

The substantive direction of the Disability Royal Commission’s Final Report is, on the elements relevant to the Bill, the opposite of the structural choices Schedule 1 makes. The Royal Commission heard evidence over four-and-a-half years and reported on 28 September 2023; the Government’s narrow citation of four recommendations cannot be read in isolation from the direction of the Report as a whole. The gap between acceptance and action is itself stark: independent analysis published by Queensland Advocacy for Inclusion in May 2026 found that, across the Royal Commission’s 222 recommendations, only 3 per cent have been implemented and 53 per cent have no reported progress.

Submission 2079

to record — near-total non-implementation of recommendations the Commonwealth has formally accepted — Schedule 1 advances the scheme in the opposite direction to four of them, each of which warrants the Committee’s attention in its own right.

Community inclusion, against the single largest saving in the package. The Royal Commission’s foundational finding was that the segregation of people with disability — in housing, education, employment and community life — is the structural condition in which violence, abuse, neglect and exploitation take hold, and that reducing segregation and enabling community participation is the central reform task. Departmental modelling tabled in the Senate on 27 May 2026 shows that the single largest saving in the present package — $13.2 billion over four years — is achieved by cutting the funding stream that allows participants to engage support workers to get out into the community. The Government’s own modelling makes the reduction of community-participation funding the financial centrepiece of the reform. A former Royal Commissioner has publicly warned that the package will leave participants more “isolated” and “segregated”, ignoring one of the central lessons of the Royal Commission. The largest single measure in a Bill the Government says takes up the Royal Commission’s recommendations produces the precise outcome — segregation and isolation — the Royal Commission was established to end.

Supported decision-making, against the removal of the s.31 Principles. Recommendations 6.4 to 6.10 of the Final Report called for a new supported decision- making framework placing the person’s will and preferences at the centre of decisions that affect them. Schedule 1 Part 6 of the Bill removes the s.31 Principles relating to plans — The operational architecture through which the scheme gave effect to participant choice, control and supported decision-making at the plan level (Section 7 develops the statutory consequence). The Bill dismantles the plan-level machinery of supported decision-making in the same package the Government says implements the Royal Commission.

Inclusive education, against the schools-disability cut. Volume 7 of the Final Report made fifteen recommendations directed at inclusive education and the reduction of segregated schooling. The same Budget that introduces this Bill cuts $472 million from the disability loadings that support students with disability in mainstream classrooms. The fiscal measure runs against the direction of the Royal Commission’s inclusive-education recommendations.

A dedicated disability portfolio, against portfolio consolidation. The Final Report recommended a dedicated Australian Government Minister for Disability Inclusion and a specific department for disability equality and inclusion. The current Government has instead consolidated the function into a combined Minister for Health, Aged Care, Disability and the NDIS. The machinery-of-government recommendation the Royal Commission made for the Commonwealth’s own disability governance has been answered in the opposite direction.

Underlying all four is a single methodological inversion. The Royal Commission’s framework was grounded in the social model of disability and the lived experience of

Submission 2079

people with disability; the Bill’s new s.9B functional capacity definition, addressed in Section 7, is grounded in the opposite frame.

A further, provision-level contradiction sits in Volume 6, Enabling Autonomy and Access, in one of the provisions the Bill’s Second Reading Speech invokes the Royal Commission to djustify. Volume 6 at Recommendation 6.34 specifically recommended health navigators to assist people with disability to access and navigate the NDIS — including the functional capacity assessment and plan-development process — with particular attention to the financial and geographic barriers that lower-income applicants face in navigating a complex administrative system. No health-navigator program responding to Recommendation 6.34 is funded in the Bill or in Budget Paper No. 2. The $200 million Inclusive Communities Fund provisioned at Budget Paper No. 2 line 8937 is group-based social participation, not access navigation.

More directly: the Bill’s new s.25A(2) provides that treatment is “appropriate” for the purposes of the permanence-and-treatment test “regardless of whether the person’s individual circumstances restrict the person from accessing the treatment” — including, per a Note to the provision, financial circumstances and geographical location. Volume 6 found that financial barriers are a systemic inequity for lower-income NDIS applicants. Peer-reviewed research published in the Medical Journal of Australia in 2025 — Disney et al., analysing 705,594 NDIS applications between 2016 and 2022 — provided quantitative confirmation of the Volume 6 finding: measurably lower NDIS eligibility rates for applicants in lower socioeconomic areas. The Bill’s s.25A(2) does not address the financial-barrier inequity the Royal Commission identified. It writes that inequity into statute — income and g geographic inaccessibility of treatment codified as irrelevant to whether that treatment is “appropriate” for eligibility purposes. The Government has invoked the Disability Royal Commission as the justification for a provision that is the statutory inverse of a specific Royal Commission finding about who the current NDIS is failing to reach.

The “original intent” rhetoric — second venue on the Department’s signed record

A separate rhetorical claim sits at the centre of the Government’s case for the Bill. On 22 April 2026 the Minister told the National Press Club that the Bill would “return” the NDIS to its original intent.“ The phrase has since become a signature framing in the Government’s public communications on the reform package. The Department of Health, Disability and Ageing has now placed the same framing on its own signed correspondence record. Ministerial response MC26-008601, signed by L. Jackman, Assistant Secretary, Thriving Kids Taskforce, dated 25 May 2026, reads at paragraph 6:

"Government is working with professionals and engaging with the community in
respect to these reforms and is restoring the NDIS to its original intent of
supporting people with permanent and significant disability and securing the

scheme for future generations.“

Submission 2079

The framing has therefore spread from Ministerial speech (22 April 2026) and Budget paper measure text (12 May 2026) to formal Departmental correspondence (25 May 2026), and the same rhetorical layer also appears in the Chair’s Foreword to the December 2025 House Committee report No child left behind — set out in the Thriving Kids Inquiry subsection that follows — in the phrase “all people with severe disability.” The framing is now operating across four primary-source venues of the Commonwealth’s NDIS-reform communications: three executive-side (Minister, Treasury, Department) and one parliamentary-committee.

The framing is rhetoric, not statutory grounding. NDIS Act 2013 at s.3(1) sets out the Objects of the scheme. Section 3(1)(d) commits the scheme to “provide reasonable and necessary supports, including early intervention supports, for participants in the National Disability Insurance Scheme.” Early intervention is named in the Objects clause as a coordinate Object of the scheme. Section 25 of the current Act establishes a distinct access pathway for early intervention supports — one that does not require permanence and that the scheme has used to admit participants since its launch. The framing that the scheme was “only ever” intended for “people with permanent and significant disability” is not what the Act says on its face — it is contradicted by the Act’s own Objects. This is not a point of emphasis. The “original intent” framing has been repeated through media coverage and absorbed into public understanding of the reform, and a meaningful part of the limited public resistance to the Bill traces to a premise — that the scheme has drifted from a narrow founding purpose — that the governing statute does not contain. The Department’s restatement of the framing in MC26-008601 is, measured against the statute the Department itself administers, incorrect.

A second statutory anchor compounds the point. NDIS Act 2013 s.3(1)(a) commits the scheme to “give effect to Australia’s obligations under the [CRPD].” CRPD Article 1 defines disability through the interaction between impairment and barriers — the social model — without conditioning access on permanence. The “original intent” framing the Department has now signed off on rests on a definition of disability the Act itself, in its first Object, has bound the scheme to give effect to a different definition of.

A third anchor sits inside the Bill’s own design. Schedule 1 Part 8 of the Bill amends new s.25(1B) of the Act to apply the new permanence-and-treatment test to the early intervention pathway. The Bill itself does not treat s.25 as if it had never existed; it amends s.25 to collapse its substantive distinction from s.24. Schedule 1 Part 6 separately removes the s.31 Principles relating to plans — the operational architecture by which the scheme had given effect to participant choice, control and community inclusion at the plan level. These provisions confirm that the Bill is not “restoring” a prior state of the scheme; it is restructuring the scheme by amending the provisions through which the original purposes of the Act have been operationalised. The “original intent” framing serves a rhetorical function the operative text of the Bill does not support.

A fourth anchor lies outside the NDIS Act. The Disability Discrimination Act 1992 defines disability to include past, present, future and imputed disability, and impairments without

Submission 2079

gregard to their permanence or severity; it contains no permanence threshold. The “original intent” framing — and the permanence-and-treatment gate the Bill constructs to operationalise it — narrows access by a criterion that neither the NDIS Act nor the DDA contains, at the same time as the Government asserts the reform restores a statutory purpose. Both propositions cannot hold.

Taken together, the four anchors settle a single question. The “original intent” claim is contradicted by (i) the Objects clause at s.3(1)(d), which names early intervention as a coordinate object of the scheme; (ii) the distinct early-intervention access pathway at s.25, which the Bill’s own Part 8 must amend in order to narrow — confirming the pathway exists to be narrowed; (iii) the first Object at s.3(1)(a), which binds the scheme to the CRPD’s barriers-based definition of disability; and (iv) the DDA definition, which carries no permanence criterion. A claim contradicted on four independent statutory footings is not a contestable reading of the Act’s purpose. It is a rhetorical frame the operative text of the Bill works against, now restated on the Department’s own signed record.

The Committee is entitled to require that any reliance on the “original intent” framing as a substantive rationale for the package be tested against (i) the statutory Objects of the Act at s.3(1), (ii) the existing s.25 access pathway and the Bill’s own Part 8 amendments to it, (iii) the s.3(1)(a) CRPD-grounding the Act commits the scheme to, and (iv) the Disability Discrimination Act 1992 definition of disability, which contains no permanence threshold. The framing’s appearance on the Department’s signature record at MC26-008601 makes the test available against a primary-source venue the Department itself has produced.

The provenance of the package’s framing — taxpayer-funded message-testing

A further matter goes to the provenance of the framing on which the package’s public case rests. In 2023, before the Independent NDIS Review reported, the National Disability Insurance Agency commissioned the Redbridge Group — a political strategy and research firm — to conduct focus-group message-testing on how to communicate reductions to the scheme. The work was funded from the public purse. Its output is on the public record through The Saturday Paper’s reporting and has been raised publicly by Senator Jordon Steele-John, the Greens disability spokesperson.

The October 2023 Redbridge report is explicit about its strategic purpose. It records that when fraud, rorts and unreasonable pricing were presented to participants as “posing an existential threat to the NDIS”, the researchers were “able to create an environment in which respondents were amenable to reforms designed to counter these things.” It records that the testing was able to “elicit a degree of qualified tolerance” for reforms that “would restrict… eligibility for the Scheme.” And it sets out the operative sequencing instruction directly: “demonstrating that we are first addressing providers and pricing is critical before discussing reforms that impact participant accessibility or funding.” The report also records the converse: without the fraud-and-rorts framing, participants resisted reform. That is consistent with the broader public sentiment Redbridge’s 2023

Submission 2079

to that the NDIS was so popular with the Australian public that it was regarded with the same protectiveness as Medicare. The significance is direct. The Government commissioned research that found the public would instinctively defend the scheme, and that presenting it as beset by fraud and rorts was what made restriction tolerable. On the Governments own commissioned research, the fraud-and-rorts framing is not incidental to the reform — it is the mechanism by which a scheme the public would otherwise protect is made cuttable.

The structure of the present Bill follows that sequencing. The provider-side and integrity measures sit in Schedule 2, and they are the measures the Governments public communications have foregrounded. The most consequential measures — the eligibility-tightening, the permanence-and-treatment gate, and the funding-below-cost authorities — sit in Schedule 1, which the public framing has discussed far less. The taxpayer-funded advice was to lead with fraud and pricing in order to build tolerance for access restriction; the Bill leads with fraud and pricing, and packages the access restriction alongside it.

Two limits should be stated plainly so the Committee can weigh the point precisely. The Redbridge testing was conducted in 2023 and surfaced through the inquiry into the predecessor Getting the NDIS Back on Track legislation of 2024; it goes to the provenance of the communications and evidence base the present Bill inherits, not to a finding about the drafting of this Bill. And the NDIS Reviews Co-Chair has stated the message-testing did not influence the Review’s recommendations. This submission does not assert that the testing produced the Bill. It asks the Committee to weigh a documented fact: public money was spent to identify the framing most likely to make access restriction tolerable to the very people it affects; the framing recommended was a fraud-and-pricing-first sequencing; and the Bills structure, together with the Governments public communications, follows that sequencing. Where a package is sold on a framing developed to manufacture tolerance, rather than on the evidence base set out across this section, the framing is not a substitute for the evidence the Committee is entitled to require.

The Governments own Thriving Kids Inquiry — sixteen recommendations the Bill does not implement, and Deputy Chair Ryans load-bearing dissent

On 2 September 2025 the Minister referred to the House Standing Committee on Health, Aged Care and Disability an Inquiry into the Thriving Kids initiative. The Committee was chaired by Dr Mike Freelander MP (Labor, Macarthur), a paediatrician of more than four decades’ practice, and deputy-chaired by Dr Monique Ryan MP (Independent, Kooyong). The Committee received 404 written submissions and held eight public hearings across Canberra (3 October, 10 October, 7 November, 17 November and 28 November 2025), Melbourne (20 October 2025) and Surry Hills NSW (21 October 2025). The Committee reported on 22 December 2025. The report, No child left behind, made sixteen recommendations.

The Department of Health, Disability and Ageing names this Inquiry on its own signed correspondence record as one of the consultation venues for the design of Thriving Kids.

Submission 2079

Ministerial response MC26-008601 of 25 May 2026 identifies the Inquiry — referenced as “the Parliamentary Inquiry on Thriving Kids and Foundational Supports” — among the venues that “helped informed the national model for Thriving Kids.” The Inquiry is Government-cited evidence; it sits inside the Government’s own claim about the consultation foundations of the program the Bill creates the legislative pathway into.

The relevant point is what the Inquiry’s recommendations actually said, and whether the Bill before this Committee implements them. On the operative text of the Bill and the operational follow-through documented in MC26-008601, the Bill is structurally non-implementing of five of the most material Committee recommendations.

Recommendation 1

The Committee recommended that the Thriving Kids Advisory Group, established to inform the design of the program, “include recognised organisations and peak bodies with an established history in providing services and support for children with… disabilities… as well as individuals with lived experience of disability or care of persons with disability, First Nations, and culturally and linguistically diverse backgrounds.” The Committee recognised, on the parliamentary record in December 2025, that the Advisory Group needed to include the cohort the program targets. The Department’s description of the Advisory Group’s composition, set out in full later in this section under the National Autism Strategy heading, names eight categories of professional expertise. It does not name any peak autism body. It does not name any Autistic-led self-advocacy organisation. It does not name the Autistic-adult community whose lived expertise is, on the National Autism Strategy’s own framing, a named tier of evidence. The Department’s operational follow-through, five months after the Committee’s recommendation, does not satisfy Recommendation 1 of the Government’s own Inquiry. “Nothing about us without us” — the long-standing principle of Autistic-led policy participation — sits documented on the parliamentary record as the standard the Government’s own Committee specified, and on the signed Departmental record as the standard the Government’s operational follow-through has not met.

Recommendation 3

The Committee recommended that the Government engage with the broader disability sector — including the cohorts on whom Thriving Kids will operate — in the design and implementation of the program. The Disability Representative Organisations Joint Statement of 14 May 2026 documents that the peak-body sector publicly considers the consultation pattern to date inadequate, with the Senate inquiry timeframe described as “extraordinarily short.” Recommendation 3 of the Government’s own Inquiry is not implemented by the consultation pattern the Department has applied to the Bill itself.

Recommendation 4

The Committee recommended the establishment of an Inspector-General of the NDIS — an independent statutory oversight role with public-reporting obligations and a charter to monitor the operation of the scheme. The Bill before this Committee does not legislate an Inspector-General. The oversight architecture of the NDIS remains the architecture the Committee, in December 2025, recommended be supplemented by a dedicated independent oversight officer. The recommendation is on

Submission 2079

the Government’s own December 2025 record. The legislative response five months later does not implement it.

Recommendation 9

The Committee recommended that a child’s participation in Thriving Kids should not preclude that child’s subsequent access to NDIS individualised supports if the child’s support needs increase or the Thriving Kids supports prove insufficient. This recommendation is the structural counter to the cohort-redirection mechanism the Bill operationalises through Schedule 1 Parts 8 and 9. The Bill’s new s.24(5) and s.25(1B) introduce the permanence-and-treatment gate; new s.25B authorises the Minister to declare entire support systems — including Thriving Kids itself — as “alternative supports” that exclude impairment categories from NDIS access. The Bill is not constructed to give effect to Recommendation 9’s non-precluding-pathway commitment. It is constructed to be a precluding pathway: Thriving Kids is the redirection destination, the new permanence- anD-treatment gate is the eligibility filter for that redirection, and the s.25B(4) Ministerial declaration is the closing mechanism. The Committee’s most directly applicable recommendation on the relationship between Thriving Kids and the NDIS is, on the operative text, structurally contradicted by the Bill.

Recommendation 10

The Committee recommended that the Government undertake a comprehensive review of Thriving Kids implementation 24 months after commencement. The recommendation reflects the Committee’s recognition that the design of the program needed to be tested in operation before being treated as a settled architecture. Thriving Kids commences on 1 October 2026; the Bill commences the new functional capacity assessment regime on 1 January 2028. Neither commencement date is contingent on, or paired with, a 24-month review mechanism in primary legislation. The Bill does not implement Recommendation 10.

The five recommendations the Bill structurally does not implement — 1, 3, 4, 9 and 10 — is the recommendations that, taken together, would have produced a different Bill. Recommendation 1 would have required the Department to consult the cohort the program targets in the design of the Advisory Group; the Department’s own correspondence shows it has not. Recommendation 3 would have required the Department to consult the broader disability sector on Thriving Kids; the peak-body record shows it has not. Recommendation 4 would have established an independent oversight officer for the scheme; the Bill does not legislate one. Recommendation 9 would have made Thriving Kids participation non-precluding of NDIS access; the Bill makes it part of the precluding architecture. Recommendation 10 would have built a 24-month review mechanism into the implementation; the Bill commences without one. The Bill’s relationship to the Inquiry is the same relationship it has to the NDIS Review, the National Autism Strategy, the EAC Third Round Consultation, and the Productivity Commission 2024 working paper: selective implementation, partial omission, or active contradiction.

The Chair’s substantive concessions on the parliamentary record

The Chair’s Foreword to No child left behind contains a number of concessions that bear directly on the Bill before this Committee. Dr Freelander, as Chair of a Government-

Submission 2079

referred Inquiry and a Labor MP, is making political statements rather than independent clinical-evidence findings, but the concessions are on the parliamentary record nonetheless.

First, the Chair acknowledges that Mr Butler’s Thriving Kids announcement “did create anxiety amongst a number of groups including parents and providers.” The Government’s own Inquiry Chair confirms on the parliamentary record that the program’s announcement produced community anxiety - the basis on which Sections 5 and 15 of this submission discuss the impact of the reform package on participants and their families.

Second, the Chair concedes that “mild or moderate disabilities… are very subjective definitions.” Section 4 of this submission addresses this concession in detail. A paediatrician Labour MP, serving as Chair of the Inquiry into the program the diagnostic-language descriptor was designed for, is on the parliamentary record agreeing with the substantive critique of the “low to moderate support needs” framing the Bill’s targeting logic depends on. Section 7 develops the statutory-inconsistency consequence: Autism is by definition permanent under the DSM-5, so the permanence limb of the new s.24(5) and s.25(1B) gate cannot exclude Autistic applicants, and the operational exclusion mechanism is the “low to moderate support needs” cohort descriptor operationalised through new s.9B - a descriptor the Inquiry’s Chair has now described as “very subjective.”

Third, the Chair acknowledges that “the scheme was rolled out with several structural flaws” - the structural-failure framing the present submission has applied to both the original 2013 rollout and the rollout architecture of the present Bill. The Chair’s concession aligns with Mr Dardo’s parliamentary evidence on 1 May 2026 (set out later in this section under the structural-evidence-base-mismatch sub-section) that the original payment system architecture was “established quite quickly” and built around speed of disbursement rather than integrity. The two concessions, taken together, are the Government’s own record that the integrity problems the Bill cites as its justification are the consequence of the rollout pattern the Bill is now repeating.

Fourth, the Chair commits that “we need to also make the NDIS and Thriving Kids transparent and allow scrutiny, with the ability to embed mechanisms to improve the systems over time.” The transparency-and-scrutiny commitment from the Chair of the Government’s own Inquiry is the procedural-democracy commitment Sections 13 and 15 of this submission identify as structurally absent from the Bill’s commencement architecture.

Deputy Chair Ryan’s Additional Comments — load-bearing dissent on the parliamentary record

dr Monique Ryan MP, Independent, served as Deputy Chair of the Inquiry. The report’s Deputy Chair Additional Comments - a substantive dissent from the Chair’s framing on five specific points - sit on the Government’s own December 2025 parliamentary record. The Additional Comments introduce evidence and framings that bear directly on the Bill before this Committee that the Chair’s Foreword does not, and warrant separate treatment.

Submission 2079

First, the Deputy Chair places the NDIS economic multiplier on the parliamentary record — that “every dollar spent on the NDIS generated $2.25 of economic benefit.” The $2.25 return-on-investment figure originates in Per Capita’s 2021 False Economy analysis rather than in the Productivity Commission’s 2011 report; Section 14 sets out its provenance. The load-bearing point for present purposes is the one the Deputy Chair makes: the economic- benefit side of the scheme — the multiplier on which its original case rests — has not been re-tested in the modelling that underpins the present Bill. Section 14 sets out the three Government-commissioned cost-benefit analyses to date — the Productivity Commission’s 2011 and 2017 reports and Taylor Fry’s 2023 quantification for the NDIS Review — and the absence of a fourth analysis testing whether the present reform package itself delivers net economic benefit. The three-CBAs-no-fourth point this submission develops in Section 14 now has a Government-record corroborating citation from the Deputy Chair of the Government’s own Thriving Kids Inquiry.

Second, the Deputy Chair adopts the defunding-of-alternatives framing as the structural causal account of NDIS cohort growth — the inverse of the Chair’s “overdiagnosis” framing on the same data. Dr Ryan’s Additional Comments place on the parliamentary record that the growth in Autistic NDIS participation is the consequence of the defunding of community-based alternatives over the preceding decade, not the consequence of clinical over-identification. This is the under-served-cohort thesis that Sections 4 and 6 of this submission develop in more detail. The substantive causal account of the data the Government cites as the rationale for the Bill is in active dispute between the Chair and the Deputy Chair of the Government’s own Inquiry, on the parliamentary record.

Third, the Deputy Chair’s Additional Comments call for the Thriving Kids rollout to proceed “only when… ready” — a deferred-rollout position that parallels the deferral framing this submission applies to Schedule 2 in Part B. The Deputy Chair of the Government’s own Inquiry is on the parliamentary record calling for a procedurally-deferred commencement of the Thriving Kids program, against the Bill’s 1 October 2026 commencement date.

Fourth, the Deputy Chair broadens the Chair’s “anxiety” characterisation. Where the Chair acknowledged “anxiety amongst a number of groups including parents and providers”, the Deputy Chair characterises the affected constituency as the “much anxiety within the disability community” — the disability-community-wide framing that Sections 5 and 15 of this submission have applied. The Government-record characterisation of the affected constituency is now on the parliamentary record in disability-community-wide terms, not the more limited “parents and providers” framing the Chair used.

Fifth, the Deputy Chair concedes that “there is a significant gap in our health care workforce strategy.” The workforce-strategy gap is material to the Schedule 2 deferral position this submission’s Part B addresses. The Deputy Chair of the Government’s own Inquiry has placed on the parliamentary record that no national workforce-planning mechanism exists for the support workforce on which both the NDIS and Thriving Kids de pend. Section 11 of this submission develops the integration point — the Bill’s new

Submission 2079

s.25A(2) regressivity codification compounds the workforce gap the Deputy Chair has identified.

Material requiring counter-context

The Inquiry’s record also contains material this submission does not cite uncritically. The Chair’s Foreword frames cohort growth as “overdiagnosis” and “upgrading of diagnosis” - uncited assertions of contested clinical framings, directly contradicted by the peer-reviewed evidence base this submission relies on (Disney et al. 2025; the under-served-cohort thesis developed in Section 6). The Foreword’s “all people with severe disability” scope phrasing applies the “original intent” / scope-narrowing rhetoric Section 7 has addressed - Autism is by definition permanent under the DSM-5, severity is not a statutory criterion in s.24 or s.25, and the rhetoric is rhetoric, not statutory grounding. The Deputy Chair’s “11 per cent of Australian five-to-seven-year-old boys, and 5 per cent of five-to-seven-year-old girls, are NDIS participants. This is unsustainable” construction reproduces the Minister’s growth-as-crisis framing; the campaign’s under-subscription thesis in Section 6 disputes the “unsustainable” conclusion that follows from those numbers. The Deputy Chair’s “dual system, with defined packages for those with more significant needs” proposal sits within the Butler-aligned policy direction; this submission does not endorse it. The Deputy Chair’s invocation of “social licence” is a frame this submission declines to accept; the NDIS’s legitimacy is statutory, anchored in NDIS Act 2013 s.3(1)(a) and Australia’s CRPD obligations, not in social licence.

These framings appear on the same parliamentary record as the material cited in this subsection. They do not displace the substantive concessions and load-bearing dissent the Inquiry’s December 2025 record contains. They sit alongside that material; on the points of disagreement, the campaign’s contradicting evidence base - Section 7’s canonical statutory statement, Section 6’s under-subscription thesis, and Section 4’s diagnostic-language critique - applies.

The three-layer consultation gap

The consultation gap this submission’s case for omission of Schedule 1 relies on is documented across three layers of the Government’s own primary-source record, no one of which is contested.

Layer one:

The Chair’s Foreword names the consultation network the Chair drew on for the Inquiry - “my many paediatric, health and education colleagues who have given me advice.” The Foreword names clinical and education professional advisors. It does not name Autistic-community voices.

Layer two: Recommendation 1 of the Committee’s own report - set out earlier in this subsection - specifies the cohort that the Thriving Kids Advisory Group should include: peak bodies with an established history in disability services, individuals with lived experience of disability, First Nations representatives, and culturally and linguistically diverse

Submission 2079

representatives. The Committee, in its own report, recognised the design gap and recommended its corrective.

Layer three: the Department’s operational follow-through five months later, the verbatim Thriving Kids Advisory Group composition set out later in this section under the National Autism Strategy heading, names eight categories of expertise. None include any peak autism body, Autistic-led self-advocacy organisation, or Autistic-adult community representative; nor any separately named “lived experience” category. The Department’s enumeration does not satisfy the cohort the Committee’s own Recommendation 1 specified.

The three layers, taken together, document the “nothing about us without us” gap on the Government’s own record. The framing is not the campaign’s framing of the Government’s consultation pattern; it is the Government’s consultation pattern documented across the Foreword’s named consultation network, the Committee’s own corrective Recommendation 1, and the Department’s non-implementation of that recommendation in the operational Advisory Group composition. The Bill before this Committee is the legislative product of a consultation pattern the Government’s own Inquiry recognised in December 2025 needed to be corrected, and that the Government’s operational follow-through has not corrected.

The NDIS Review (Bonyhady & Paul, December 2023)

The NDIS Review’s Working Together to deliver the NDIS report

The NDIS Review’s Working Together to deliver the NDIS report is the document the Government cites most often as the evidentiary basis for the reform package. The Bill’s relationship to the Review’s recommendations is more contested than the Government’s public messaging suggests.

Recommendation 3 — foundational supports alongside, not instead of, the NDIS.

The Review framed foundational supports as a tier complementing individualised NDIS supports - additional capacity, not a substitution. Schedule 1 Part 8 and Part 9 of the Bill create the legislative pathways out of the NDIS into foundational supports for participants whose access is revoked under the new permanence test or whose impairment is captured by a new s.25B(4) Ministerial declaration of an alternative supports system. The Review’s “alongside, not instead of” framing is structurally inverted in the Bill - foundational supports are now the destination for participants the eligibility-tightening provisions exclude. The structural inversion is the opposite of the Review’s recommendation.

Action 3.1 - consistent assessment, threshold deferred.

The Bill implements the recommendation to introduce a consistent functional capacity assessment definition at new s.9B. The threshold against which the definition operates is not in the Bill. Sub-sections 9B(2) and 9B(3) delegate the threshold and the methods for applying the definition to NDIS Rules made by the Minister, and the Technical Advisory Group provisioned at Budget Paper No. 2 line 8919 - with $3.3 million in 2026-27 - is the body now beginning that work. The Bill claims to implement the recommendation while leaving the operative threshold unbuilt.

Submission 2079

Action 6.3 — assessment tools as precondition to transition.

The NDIS Review identified the availability of validated assessment tools as a precondition to the commencement of any new access regime. The Bill commences the new functional capacity regime on 1 January 2028 (Budget Paper No. 2 line 8929). The application provisions across Bill items 11 and 94 confirm that the test applies to existing participants on plan review from that date, regardless of whether the tools have been validated by that date. The Bill commences the regime in the absence of the precondition the Review identified as necessary.

Action 23.2 — Evidence Advisory Committee to review the evidence base.

The Review identified the Evidence Advisory Committee as the body responsible for reviewing the evidence base on which NDIS access decisions are made. The People with Disability Australia submission to the EAC Third Round Consultation, dated 26 March 2026, documents that the consultation was active and ongoing at the time the Bill was introduced. The Bill pre-empts a consultation process that the NDIS Review identified as the institutional mechanism for reviewing the evidence base the Bill operationalises. The EAC has not endorsed the permanence-and-treatment test, the new functional capacity definition, or the design of the assessment instruments the Technical Advisory Group is now developing.

The National Autism Strategy 2025-31

The National Autism Strategy is the Commonwealth’s cross-portfolio policy framework for autism, published in 2025 with a six-year horizon. The Strategy’s framing of the Autistic population, the evidence framework it commits to, and the participation principles it adopts are inconsistent with the Bill in three distinct ways.

Co-design and lived experience as core pillar.

The Strategy commits the Commonwealth to working with Autistic Australians in the development of policy that affects them. The Bill was developed without co-design with the Autistic community, the broader disability community, or the peak bodies that speak for both. The Disability Representative Organisations Joint Statement of 14 May 2026 — signed by twelve peak bodies including People with Disability Australia, the Australian Federation of Disability Organisations, the Australian Autism Alliance, Children and Young People with Disability Australia, the Disability Advocacy Network of Australia, Inclusion Australia, Mental Health Australia, the First Peoples Disability Network, the National Ethnic Disability Alliance, Physical Disability Australia, Women with Disabilities Australia, and Down Syndrome Australia — publicly described the Senate inquiry timeframe as “extraordinarily short… approximately one month” that “does not allow time for written submissions, public hearings, and meaningful engagement with the disability community.” The peak-body consensus on the consultation inadequacy is the inverse of the co-design principle the National Autism Strategy commits the Commonwealth to.

The Department has further confirmed the co-design gap on its own signed record. Ministerial response MC26-008601 of 25 May 2026 describes the composition of the

Submission 2079

Thriving Kids Advisory Group — the operational body the Department names as having “helped informed the national model for Thriving Kids” — verbatim as follows:

• “The Thriving Kids Advisory Group included experts from a range of fields including paediatrics, child development, research, disability, child and family services, health care, early education and schools, and First Nations perspective.”

The eight categories of expertise named are the categories the Department considers material to the design of the 0-to-8 national model. The list does not name any peak autism body. It does not name any Autistic-led self-advocacy organisation. It does not name the Autistic-adult community whose lived expertise is, on the National Autism Strategy’s own framing, a named tier of evidence the Commonwealth has committed to. The Department’s description of the operational Advisory Group composition is, on its face, a description of an Advisory Group designed without the constituency the program targets.

“Nothing about us without us” — the long-standing principle of Autistic-led policy participation in Australia — is the principle that has not been observed.

Reducing discrimination and vilification. The Strategy commits the Commonwealth to reducing discrimination and vilification of Autistic Australians. The Bill’s public framing — through the Government’s ongoing characterisation of participants as a source of “leakage” and the Minister’s Second Reading Speech rhetoric on integrity — recategorises Autistic NDIS participants in a way that runs against the Strategy’s commitment. The pattern operates with particular force in respect of Autistic Australians because Autistic NDIS participants are the largest impairment cohort among new entrants (approximately 75 per cent of new entrants per the Scheme Actuary’s projection cited in ABC reporting). The Government’s “unsustainable growth” framing therefore lands on the Autistic cohort as the public face of the cost-side problem the framing constructs, rather than on the under-served-cohort interpretation Section 6 of this submission identifies: that the growth in Autistic NDIS participation is unmet need finally beginning to be measured, against a baseline of structural barriers (diagnostic cost, application complexity, masking under-measurement) that have suppressed legitimate scheme participation for years. Dr Clem Bastow, the cultural critic and critical Autism studies researcher I cited in Section 4, has described the cumulative effect of the framing as turning Autistic Australians into “the new dole bludgers.” That recategorisation is the framing the National Autism Strategy commits the Commonwealth to actively reducing.

Lived experience evidence framework. The Strategy commits the Commonwealth to an evidence framework in which research, clinical experience, and lived experience are all named tiers of evidence. The Bill’s new s.34(1E) and s.34(1F) introduce a statutory evidence hierarchy for NDIS supports funding decisions. The four tiers are: (a) published peer-reviewed generalisable research; (b) evidence as to effectiveness in the participant’s circumstances; (c) outcomes from the participant’s previous plan; (d) other matters. Lived experience is not named as a tier. The participant’s own clinicians and the participant’s own life evidence are structurally relegated to “(d) other matters” or excluded entirely. The Bill imposes an evidence hierarchy on NDIS access and funding decisions that is the

Submission 2079

inverse of the Strategy’s evidence framework — published research at the top, lived experience demoted to a residual category.

The Evidence Advisory Committee Third Round Consultation

The Evidence Advisory Committee is the body established under NDIS Review Action 23.2 with responsibility for reviewing the evidence base on which NDIS access decisions are made. The PWDA submission of 26 March 2026 to the EAC Third Round Consultation is on the public record. The submission documents the consultation as active and ongoing — PWDA’s recommendations include the maintenance and expansion of NDIS-funded Early Intensive Behavioural Intervention, which the Bill’s eligibility-tightening provisions (combined with the Thriving Kids reframe) operate against. Submissions to the Third Round Consultation closed on Sunday 29 March 2026; the EAC has not publicly announced a completion date for its findings on the Third Round.

The Bill commences the new access regime before the EAC has concluded the consultation Action 23.2 established. The Government’s case for Schedule 1 cites the NDIS Review as evidentiary basis; the EAC is the institutional mechanism the NDIS Review itself established for testing the evidence base; the Bill pre-empts that mechanism. That an evidence-review consultation specifically established by the Government’s own Independent NDIS Review for the purpose of testing the evidence base on which NDIS access decisions are made is active and ongoing while the Bill is before the Committee is itself a structural contradiction the Bill cannot answer on its own terms. The contradiction is a further load-bearing reason for the omission of Schedule 1.

The Productivity Commission

Three Productivity Commission reports sit in the documentary lineage of the NDIS: the 2011 Disability Care and Support report that established the scheme, the 2017 NDIS Costs report, and the 2024 Foundational Supports working paper that examined the cost-shift implications of moving cohorts off NDIS individualised plans into foundational supports.

The PC 2011 and PC 2017 reports both found, on the all-of-scheme cost-benefit analysis, that the NDIS delivers net economic benefit. Taylor Fry’s 2023 quantification commissioned for the Independent NDIS Review measured net costs ($16 billion in 2022-23) exceeding measured benefits ($7.5 billion in 2022-23) at the all-of-scheme level under a deliberately constrained methodology that explicitly excluded major benefit categories (justice and child protection outcomes, post-school education, community benefits, fiscal multiplier effects, value to the disability support sector, economic utility); Taylor Fry caveated the measured benefit figure as “most likely an underestimate”, and at the cohort level found net benefits exceeding net costs for the Autistic cohort in the lowest measured support-needs band and for children-and-families. The three Government-commissioned cost-benefit analyses do not, between them, support the proposition that the existing scheme is delivering negative net economic benefit; they identify cohort-level positive returns and methodological limitations on the negative all-of-scheme finding. The political characterisation of NDIS growth as “unsustainable” cannot be reconciled with the same

Submission 2079

Government’s commissioned-analysis findings of net economic benefit at the all-of- scheme level (PC 2011 and PC 2017), and of net cohort-level benefits within the lowest- measured-support-needs Autistic cohort and the children-and-families cohort the Thriving Kids measure now redirects (Taylor Fry 2023 under its constrained methodology). A scheme delivering net economic benefit on the Government’s own commissioned analyses is not, on the criterion the analyses apply, unsustainable. The cost-side framing the Government has placed on the public record is not the framing the Government’s own evidence base supports. Section 14 develops the economic-evidence dimension in detail. For Section 8’s purposes, the relevant point is that the Bill proceeds without a fourth Government-commissioned cost-benefit analysis testing whether the reform package itself delivers net economic benefit once cost-shift to adjacent public-service systems is counted.

The PC 2024 working paper on foundational supports identified a specific cost-shift consequence of moving cohorts off NDIS individualised plans into foundational supports: approximately $1.09 billion per year in additional costs to states and territories as the cost- shift lands on state-funded systems (health, mental health, schools, child protection, justice, family supports). New s.25B(4) of the Bill authorises the Minister to declare entire support systems as “alternative supports” and exclude categories of impairment from NDIS access by reference to those alternative supports. The Bill codifies the very cost-shift mechanism the Productivity Commission, as the Government’s own statutory advisory body on the scheme’s economics, identified as a structural risk. The contradiction is direct: a Government-commissioned analysis warns against a cost-shift mechanism, and the same Government introduces legislation that authorises the Minister to operationalise it.

The rushed-rollout pattern is being repeated, in the Government’s own evidence base

Section 5 of this submission set out the rushed-rollout-repeating frame from the standpoint of the assessment-instrument and reassessment-and-renewal architecture — those the same rollout pattern that produced the integrity problems the Government is now seeking to remediate is being replicated in the Bill’s commencement architecture, against the public testimony of multiple authoritative voices including Della Bosca, McVilly, Spindler-Smith, and Cusworth.

Section 8 returns to the same pattern from the standpoint of the Government’s own evidence base. On 1 May 2026, Mr John Dardo, Deputy Chief Executive Officer for Integrity, Transformation and Technology Services at the National Disability Insurance Agency and co-chair of the Fraud Fusion Taskforce, gave evidence to the Joint Standing Committee on the NDIS’s Integrity of the National Disability Insurance Scheme inquiry. Asked by the Chair to identify the contributing factors to the integrity problems the inquiry was investigating, Mr Dardo placed on the parliamentary record that “the program was established quite quickly. I think there were some assumptions about how the program, the market and the claiming systems would work.” He went on to describe the original payment system architecture: “the speed of payments getting out was the primary motivator of the system

Submission 2079

design”, with manual extraction of integrity-flagged claims limited to “maybe 20 to 30 claims per day”, and “any claims between 5 pm and 6 pm were being paid without any human eyeballs on them.” The Government’s own anti-fraud co-chair has placed on the parliamentary record that the cause of the integrity problems is the design of the original payment system — built for speed of disbursement, not for fraud prevention. The Bill responds to that diagnosis by tightening participant eligibility. The two are not the same problem.

The rushed-rollout pattern is then visible at the structural level of the Bill itself: the assessment instruments are not designed (the Technical Advisory Group is now beginning the work, against a 1 January 2028 commencement); the replacement programs are not named for the 9-to-17 cohort or for adults (only the 0-to-8 Thriving Kids program is named, and its bilateral implementation agreements are not yet in place for Victoria and Queensland — approximately 46 per cent of the Australian population — per Occupational Therapy Australia’s 12 May 2026 confirmation, unchanged by the 14 May Policy and Advocacy update); the consultation period is, on the peak-body consensus, “extraordinarily short”; and the assessment-instrument vacuum is embedded in the legislative text (Section 4 of this submission). The pattern the anti-fraud co-chair identified as the cause of the original integrity problems is the pattern the Bill is now repeating. This is structural, not strategic — the implementation pattern that produced the integrity problems the Bill is said to address is being replicated in the Bill’s own commencement architecture.

The structural evidence-base mismatch

The Joint Standing Committee on the NDIS Integrity of the National Disability Insurance Scheme inquiry was referred to the Committee by the Minister on 25 March 2026. The Chair confirmed at the opening of the 1 May 2026 public hearing: “The minister has actually referred this to us, asked us to take this up.” The inquiry is the Government’s own primary integrity-investigation vehicle. Its evidence base — captured across its public hearings to date, beginning with the 1 May 2026 Canberra hearing and continuing on 15 May 2026 in Sydney and 21 May 2026 in Melbourne — focuses on provider and payment- system integrity. The witnesses examined on the day were the NDIA’s Integrity, Transformation and Technology Services Deputy CEO; the NDIA’s Service Delivery Deputy CEO; Services Australia’s Payments and Integrity Deputy CEO; the General Manager of Fraud Control and Investigations at Services Australia; the Australian Criminal Intelligence Commission’s National Manager of Operational Analysis; the Executive Director of Mission Coordination and Analysis at the ACIC; the Commonwealth Ombudsman; two senior officers of the Ombudsman’s office; senior officers of the NDIS Quality and Safeguards Commission; senior officers of the Department of Health, Disability and Ageing; and the Australian Competition and Consumer Commission’s CEO and General Manager for Consumer and Compliance Strategies. The witness slate is provider-side, payment-side, regulatory, complaints-handling, and consumer-protection. It does not include witnesses examined on participant eligibility.

Submission 2079

The testimony on the day is correspondingly provider-and-payment-focused. Mr Dardo’s evidence on the original payment system architecture, and the post-2022 controls work (24-hour payment delay, bulk hold and release capabilities, mandatory provider digital identity, mandatory ABN/text-description/invoice evidence for participant claims above thresholds), is the centrepiece of the integrity narrative the Government has put on the parliamentary record. The Associate Commissioner of the NDIS Quality and Safeguards Commission, Ms Natalie Wade, framed the regulatory function as protecting participants from provider misconduct: “where we see noncompliance, our main concern is the impact on participants. Often it can be incredibly disruptive for participants. At best and at worst, it is significantly dangerous for participants when there is fraud and noncompliance with the code of conduct and other regulatory frameworks that are within the market and designed to keep participants safe.” The Commonwealth Ombudsman, Mr Iain Anderson, testified that complaints from NDIS participants overwhelmingly concern access to the scheme and changes to plans, not fraud or sharp practice: “Complainants raise a number of issues with us. Very few of those actually involve allegations of fraud or sharp practice. For the most part they’re about the experiences of participants who have concern about changes to their plan, accessing the scheme and things like that.”

The pattern across the evidence base is consistent. The Government’s own integrity- investigation inquiry has assembled a body of evidence pointing at provider conduct, payment-system design, and provider regulatory enforcement as the locus of the integrity problem — and pointing at NDIS participants as the parties harmed by that conduct. The Ombudsman’s complaints data confirms that participants’ complaints to the Commonwealth’s independent complaints body are overwhelmingly about access, plan changes, and delivery — not about fraud. The Quality and Safeguards Commission’s regulatory frame is participant-protective. Schedule 2 of the present Bill responds to that evidence base, in part (Section 9 develops this). Schedule 1 of the present Bill, by tightening participant eligibility, operates against the same cohort the inquiry’s evidencease identifies as harmed by the provider-side integrity problems the inquiry is investigating. The Bill punishes the cohort the inquiry’s evidence identifies as the victims of the very integrity failures the Government cites as the package’s justification. The Government has not, on the public parliamentary record, assembled an evidence base supporting the proposition that participant eligibility tightening responds to the integrity problem the same Government has commissioned the parliamentary inquiry to investigate.

Mr Dardo’s 1 May 2026 evidence on the post-2022 integrity-systems uplift quantifies what has already been done — at the NDIA, Services Australia, the Australian Criminal Intelligence Commission, the Australian Federal Police, the Commonwealth Director of Public Prosecutions, and across every State and Territory police force — to address the integrity problem the Bill cites as its justification. The NDIA now has approximately 600 staff working in the integrity space (110 in the Fraud Fusion Taskforce, 104 in payment integrity, ~100 in analytics through business-as-usual funding, plus support staff). The Taskforce conducts work across 24 Commonwealth agencies and every State and Territory police force. Warrant activity in 2025 was 77, compared with approximately 30 in total across 2018, 2019, 2020 and 2021 combined. Over 2,500 problematic providers have been

Submission 2079

removed from the scheme through administrative and Taskforce-coordinated work. The 8.3 per cent integrity leakage figure Mr Dardo placed on the parliamentary record at the 1 May hearing — “$45 billion of the scheme” annual payments × 8.2-8.3 per cent leakage = approximately $3.7 billion per year — sits within the Australian National Audit Office’s previously published 6-10 per cent integrity leakage range, and is of a similar order to — ineed somewhat lower than — the Australian Taxation Office’s published net tax gap of 9.1 per cent for 2022-23. The two figures are constructed differently and are not a precise like-for-like comparison; the relevant point is one of order of magnitude — the NDIS integrity-leakage rate the Government cites as the package’s justification is not an outlier among Commonwealth programs. Mr Dardo’s evidence on the parliamentary record is that integrity leakage “includes the full range, absolutely” — error, misunderstanding, inappropriate claiming, and criminal fraud collectively — not fraud alone.

The Committee is entitled to require the Department of Health, Disability and Ageing to identify the evidence base for participant-eligibility tightening as a response to integrity problems the Government’s own inquiry evidence places in a different domain. If the integrity problem the Bill cites as its justification is the problem the Government’s anti- fraud co-chair has placed on the parliamentary record, the response is in Schedule 2 and in the post-2022 integrity-systems uplift, not in Schedule 1.

The pattern across the review framework

The pattern is consistent across the six review streams. The Bill claims to implement a recommendation set that the Government’s own primary documentation either cites narrowly against one Part of Schedule 2 only (the Disability Royal Commission — four recommendations at Explanatory Memorandum section 5.4, none in respect of Schedule 1), implements only partially with the operative thresholds deferred (the NDIS Review Action 3.1), pre-empts before completion (the EAC Third Round Consultation under NDIS Review Action 23.2), commences before the precondition is met (the NDIS Review Action 6.3 on assessment tools), actively codifies the opposite direction of (the Productivity Commission 2024 cost-shift warning, the NDIS Review Recommendation 3 “alongside, not instead of” framing, the National Autism Strategy lived experience evidence framework), or structurally does not implement (the Thriving Kids Inquiry’s Recommendations 1, 3, 4, 9 and 10).

The Committee is entitled to require that the Government’s case for Schedule 1 be substantiated against the actual content of the reviews it cites, not the rhetorical citation of them. The substantive recommendations the Government is implementing — the Registration Taskforce 2024 provider-side measures, and partial elements of NDIS Review Recommendation 17 — are in Schedule 2, and Part B of this submission addresses them. Schedule 1’s eligibility-tightening provisions do not have the review-framework foundation the Government’s public framing claims. They have, by the Government’s own primary documentation, a different and substantially narrower foundation than the Second Reading Speech and Explanatory Memorandum cumulatively suggest.

Submission 2079

Section 9 — What Schedule 2 does well, and the integrity-rhetoric mismatch

Part A of this submission set out the case for the omission of Schedule 1 on the existing evidence base. Part B addresses Schedule 2 on its own merits. The substantive position I put to the Committee is that Schedule 2 be deferred until after the Joint Standing Committee on the NDIS Integrity of the National Disability Insurance Scheme inquiry reports on Thursday 2 July 2026, that the inquiry’s recommendations be integrated into the Schedule 2 design, and that the resulting amended Schedule 2 be brought back for substantive community consultation under “nothing about us without us” principles.

Schedule 2’s substantive direction is right — the provider-side and integrity reforms respond to documented evidence and align with what the Government’s own integrity inquiry is investigating. The procedural timing is wrong: Schedule 2’s evidence base IS the JSC’s evidence base, the JSC has not yet reported, and rushing Schedule 2 ahead of the inquiry’s findings risks repeating the original-rollout mistake Mr Dardo’s 1 May 2026 testimony identifies as a significant contributor to the integrity gaps Schedule 2 is meant to address. This section opens Part B by setting out what Schedule 2 does well, and the integrity-rhetoric mismatch that runs through the Government’s public framing of the package — the rhetorical move that uses Schedule 2’s substantive integrity evidence base to justify Schedule 1’s eligibility-tightening, which the same evidence base does not support.

What Schedule 2 contains

Schedule 2 introduces a substantial provider-side and integrity package across six Parts. Part 1 narrows the NDIS provider definition at new s.10C — a person is an “NDIS provider” if they receive NDIS amounts under a payment arrangement, or if prescribed by Rules, with Rules-based exclusion available. The Explanatory Memorandum confirms at lines 286-294 that the provider-definition change was informed by a public consultation between 19 December 2025 and 28 February 2026 receiving 269 submissions, and by the NDIS Provider and Worker Registration Taskforce 2024 Final Advice. Part 2 substantially expands the National Disability Insurance Agency’s regulatory toolkit by extending the Regulatory Powers (Standard Provisions) Act 2014 monitoring and investigation regime (Parts 2 and 3 of that Act) to the Agency, and by introducing new civil penalty provisions including parallel civil penalties to existing criminal offences. Part 3 expands the Agency’s information-gathering powers under sections 53 and 54 of the Act, including a new power for the Agency to require a person to appear before a specified Agency officer to answer questions on at least 14 days’ notice, with explicit notice that the person is not required to give information or produce documents if doing so might tend to incriminate them. Part 4 introduces a new s.45B record-retention regime: NDIS providers must retain records relating to claims for seven years (subject to Rules-based shorter periods), participants for three years, and other claim-makers for five years, with a 120-penalty-unit civil penalty ($39,600 at current Commonwealth penalty-unit value) for provider non-compliance and a new s.182(4) debt-creation mechanism where claim

Submission 2079

records are not retained. Part 5 reduces the time within which NDIS amounts may be claimed under s.45A(5)(a) from two years to 90 days. Part 6 restructures the plan-manager market: plan managers cannot simultaneously be registered to provide other supports under new s.73E(2B), addressing a documented conflict-of-interest concern.

The combined effect is a coherent provider-side and payment-side integrity package — the substantive part of the response to the integrity problems the Government has investigated through the Joint Standing Committee on the NDIS Integrity of the National Disability Insurance Scheme inquiry referred to that Committee by the Minister on 25 March 2026. It is important to be precise about what these provisions respond to. They respond to interim hearing evidence, to the 2024 Registration Taskforce Final Advice, and to the provider-definition consultation — not to the final report and recommendations of the Integrity inquiry, which the Committee has not yet delivered. That distinction is the whole of the deferral argument in this Part: the direction is sound on the evidence assembled so far, but the evidence base on which Schedule 2 should be settled is not yet complete.

What Schedule 2 does well

Four elements of Schedule 2 have substantive evidentiary support I want to recognise on the public record before turning to the amendments I ask the Committee to recommend.

• First — the provider-definition narrowing at new s.10C. This is the most consultation-evidenced single change in the Bill. The 269-submission public consultation between 19 December 2025 and 28 February 2026 supports it; the NDIS Provider and Worker Registration Taskforce 2024 Final Advice supports it; the Section 1 of this submission’s broader concerns about Schedule 1 do not extend to this provision. The provider-definition change addresses a documented regulatory perimeter gap.

• Second — the NDIA monitoring and investigation powers parity with other Commonwealth regulators. The Explanatory Memorandum notes that the Agency’s powers were “weak when compared to other comparable program administrators across the Commonwealth.” The Regulatory Powers Act extension brings the NDIA into rough parity with the Australian Communications and Media Authority, the Australian Competition and Consumer Commission, and the Australian Securities and Investments Commission. The civil penalty provisions, including the parallel civil penalties to existing criminal offences, address a midrange enforcement gap the Bill’s Explanatory Memorandum and the 1 May 2026 hearing testimony both identify as substantive.

• Third — the information-gathering powers expansion at Part 3. The Agency’s existing information-notice powers under sections 53 and 54 of the Act were not designed to support the modernised regulatory toolkit Part 2 introduces. Part 3 brings the notice power into rough alignment with the broader uplift: notices may now address the functions of the Agency under new s.53(2)(f); the notice form must explicitly inform recipients that failure to comply may attract a civil penalty under s.53(3), and that the recipient is not required to give information or produce documents if doing so might tend to incriminate them or expose them to a penalty; and the notice may require a person to appear before a specified

Submission 2079

Agency officer to answer questions on at least 14 days’ notice. The appear-and-answer- questions power is a substantive escalation from current arrangements and warrants ongoing oversight of how it is used in practice, but it is standard for Commonwealth regulators that the Regulatory Powers Act model contemplates. The substantive evidentiary support for the Part 3 uplift is the same as the support for Part 2’s broader regulatory toolkit extension.

Fourth — the plan-manager market restructure at Part 6. The conflict-of-interest restriction at new s.73E(2B) addresses a documented problem. The Minister’s own Second Reading Speech cited the 9-in-10 plan-managers-flagged-with-integrity-risk figure as evidence of the market segment’s documented dysfunction. The Government’s intention to commission a plan-manager panel under defined standards, governance and integrity requirements is substantively right on the provider side.

These four elements respond to the integrity evidence base the Government has assembled. They sit comfortably with the work the Fraud Fusion Taskforce has done since 2022, with the NDIS Amendment (Integrity and Safeguarding) Act 2026 (Royal Assent 8 April 2026), and with the 2020 NDIS supports legislation Mr Scott McNaughton, Deputy CEO Service Delivery at the NDIA, identified at the 1 May 2026 hearing as having “provided a lot more clarity around what people can and can’t spend their NDIS plans on. That created further safeguards that are really important.” The integrity work has been progressing through a multi-layered legislative architecture. Schedule 2 of this Bill extends that architecture. The substantive evidence base supports the direction of the extension. Recognising that direction is not the same as endorsing its enactment ahead of the Integrity inquiry’s final report. Legislating selected elements from a review that is not yet complete carries its own risk — the risk of settling in statute a design the inquiry’s own findings may yet revise. That is why this section commends the direction and, in the same breath, recommends deferral.

Two Schedule 2 provisions that warrant Committee scrutiny on participant- burden grounds — Parts 4 and 5

Two of Schedule 2’s six Parts impose new compliance burdens that operate primarily against participants and against the smaller end of the provider market, not against the conduct the Government’s own integrity evidence base identifies as the locus of integrity risk. They do not displace Schedule 2’s substantive direction, but they warrant the Committee’s scrutiny and, in my submission, refinement before the package is brought back after the Integrity of the National Disability Insurance Scheme inquiry reports.

Part 4 — record retention combined with the new debt-creation mechanism at s.182(4). New s.45B requires participants who make claims for NDIS amounts to keep and retain records of the claim, and of the supports the claim relates to, for three years. Item 86 of Schedule 2 substitutes a new s.182(4) under which an amount equal to the NDIS amount becomes a debt due to the Agency where the person required to retain records under s.45B does not comply with that requirement. The provider-side seven-year retention

Submission 2079

The obligation is proportionate to the regulatory uplift Part 2 introduces and to the administrative capacity of the regulated entities. The participant-side three-year retention obligation combined with the s.182(4) debt-creation mechanism is structurally different.

The 1 May 2026 Hansard, on the Ombudsman’s evidence and on the Quality and Safeguards Commission’s evidence, establishes that participants are not the source of the integrity risk Schedule 2 is designed to address — providers and individuals operating in the supply chain are. The participant record-retention obligation with debt creation does not respond to a documented participant-side integrity problem on the evidence base; it transfers the administrative burden of the integrity regime onto a cohort that the same evidence base identifies as the victims of integrity failures, not their cause. There is a more basic difficulty with imposing the obligation on participants at all. Self-managing and plan-managed participants already submit their claim evidence — receipts and invoices — to the Agency through its claiming systems at the time of claiming, and the Agency reviews and approves the claim before any payment is made. The records a participant is required to retain under s.45B are, in substantial part, records the Agency already holds, generated through a claim process that vets the evidence before money changes hands. It is difficult to identify the participant-side integrity gap that a three-year retention obligation backed by automatic debt-creation is designed to close, when a claim cannot be paid in the first place without the very evidence the Agency keeps.

The Committee should consider whether the s.182(4) automatic debt-creation mechanism is a proportionate response to participant-side record-retention failure at all — and whether it is required, given the Agency’s existing custody of the claim evidence — or whether, at a minimum, a graduated administrative response would better fit the participant cohort and the evidence base. As a baseline, the automatic debt-creation should not apply where the failure to retain records is attributable to disability-related administrative incapacity — which is, by the structure of the scheme itself, common in the participant population.

Part 5 — reducing the claim-time window from two years to 90 days. Item 89 of Schedule 2 amends s.45A(5)(a) of the Act by omitting “2 years” and substituting “90 days.” The Bill makes no other change to the claim-time architecture. The Explanatory Memorandum does not set out the evidentiary basis for the choice of 90 days as the new window, nor does it address how the new window will interact with cases in which support has been provided in good faith but the formal claim is late-discovered — for example where a provider has been disengaged or has had business interruptions, where a participant has been hospitalised or in crisis, or where intermediary arrangements have failed to surface a payable claim in time. The two-year window in the existing provision exists in part to absorb those operational realities. A 90-day window is workable for well-resourced corporate providers with mature billing infrastructure; it is materially harder for sole-trader allied health practitioners, small providers in regional and remote areas, and any participant who is self-managing supports — particularly participants whose disability affects executive function, organisation, or the capacity to action administrative tasks inside a tight window.

The Committee should consider whether the 90-day window is the proportionate setting, and whether a discretionary extension mechanism should be preserved so that a legitimate claim, delayed for a legitimate reason, is not rendered

Submission 2079

unrecoverable by the passage of time alone. A genuine claim does not become less genuine because a disengaged provider, a hospitalisation, or a disability-related administrative barrier pushed it past ninety days; rendering it unrecoverable is a different problem from fraud, and not one the Government’s integrity evidence base identifies as needing address.

These two provisions sit alongside the substantive elements of Schedule 2 the previous sub-section commends. They do not displace the broader Schedule 2 deferral the rest of this section sets out. They are flagged for the Committee as candidates for refinement during the post-JSC redesign, alongside the larger gaps Section 10 of this submission develops.

The integrity-rhetoric mismatch

The Government’s public framing of the present Bill has used integrity as the load-bearing justification for the package as a whole. The Minister’s Second Reading Speech on 14 May 2026 closed by naming four objectives: “fighting fraud and stopping rorts; slowing rapid cost increases; making eligibility requirements clearer; and delivering quality services and supports to participants.” The first of those four objectives — fighting fraud and stopping orts — is the public anchor of the package. The Government’s communications strategy has emphasised integrity. The Bill’s title is Securing the National Disability Insurance Scheme for Future Generations — a framing in which integrity is the through-line.

The mismatch is in the relationship between the public framing and the operative provisions. The integrity evidence base sits in Schedule 2 territory. The substantive integrity work the Government cites is in Schedule 2’s provider-definition narrowing, in the Regulatory Powers Act extension to the Agency, in the information-gathering powers uplift, and in the plan-manager market restructure. The Fraud Fusion Taskforce cross-agency information-sharing the Government invokes as the operational backbone of the integrity package sits within the broader integrity architecture — the 2020 NDIS supports legislation, The NDIS Amendment (Integrity and Safeguarding) Act 2026, and the Taskforce’s own established cross-agency arrangements — rather than in a discrete Part of the present Bill. Each of these elements responds to evidence already on the parliamentary record — though, as set out above, not yet to the inquiry’s final recommendations. Schedule 1 — the eligibility-tightening, the new functional capacity assessment regime, the permanence- anD-treatment test, the alternative-supports gate, the funding-below-cost authorities — operares against a different problem. Sections 3 through 7 of this submission set out the absence of a clinical evidence base, a diagnostic framework, an assessment-instrument design, a replacement-supports program, and a CRPD-and-DDA-and-Closing-the-Gap consistency for Schedule 1’s operative provisions. The integrity evidence base does not bridge the gap. Schedule 1 is sold publicly with Schedule 2’s evidence base attached to it.

The 1 May 2026 hearing evidence makes the mismatch concrete. Mr Dardo, the Government’s own anti-fraud co-chair, testified that the integrity problems trace to the original payment system design — built for speed of disbursement, not for fraud prevention

Submission 2079

— and that the post-2022 remediation has been a payment-system and provider-side undertaking: thousands of problematic providers removed, warrant activity multiplied several times over, and the NDIA’s integrity workforce scaled to roughly 600 staff working across two dozen Commonwealth agencies and every State and Territory police force.

Section 8 sets out the figures in full. The relevant point here is the locus of that work: the integrity effort is delivering, and it sits in Schedule 2 territory — provider-side and payment-side — not in the participant-eligibility provisions of Schedule 1.

What Schedule 1 does is different. Schedule 1 narrows access to the scheme. When the Minister announced the reform at the National Press Club on 22 April 2026, the figure given was a net reduction of about 160,000 — from 760,000 participants to around 600,000 by the end of the decade. Departmental modelling tabled in the Senate on 27 May 2026, and reported by Guardian Australia on 28 May 2026, puts the human figure substantially higher:

• 241,000 people who were participants before the new eligibility rules commence on 1 January 2028 are projected to be “exited” from the scheme by mid-2031 —
  • 33,000 by 30 June 2028,
  • rising to 125,000 by mid-2029, and
  • to 241,000 by mid-2031.

The scheme is projected to peak at 817,000 participants in 2027 and then fall to 598,000, against a no-change projection of 944,000; once those denied entry are counted, almost 350,000 fewer people will be on the scheme than would otherwise have been. The Minister’s “160,000” is the net landing point; the Government’s own modelling shows 241,000 existing participants removed and close to 350,000 fewer people supported overall. None of those people fall outside the scheme because of integrity behaviour. They fall outside because the eligibility gate has been narrowed. The integrity-rhetoric mismatch is therefore structural: the Government’s public framing connects integrity to the whole package, but the operative integrity work is in Schedule 2 and the rest of the integrity architecture, and Schedule 1 narrows access on a different basis altogether.

The funding-below-cost architecture in Schedule 1 is itself the sharpest single demonstration that the integrity rhetoric does not bridge across to the Schedule 1 operative provisions. New s.34A(5), inserted by Schedule 1 Part 4, reads verbatim:

“To avoid doubt, the determination has effect even if the result is either or both of the following: (a) the funding provided under a participant’s plan for a reasonable and necessary support is less than the total cost of the support; (b) the funding provided under a participant’s plan for all reasonable and necessary supports funded under the plan taken as a whole is less than the total costs of the supports.”

The text is from the Bill itself, not from campaign characterisation. The provision authorises the Minister, by legislative instrument, to reduce funding for a group of supports for the purpose of “financial sustainability”, and the “to avoid doubt” construction confirms the reduction has effect even where the funding falls below the cost of the supports the funding is intended to provide. This has no connection to fraud, to rorts, to provider integrity, to payment-system safeguards, or to any of the documented integrity work the Government cites as the rationale for the package. It is a cost-constraint

Submission 2079

provision authorising below-cost funding against participants. Its insertion in Schedule 1 — while Schedule 2 carries the genuine provider-side integrity work — is the operative-text proof of the mismatch the Government’s public framing creates, and it is the structure the taxpayer-funded message-testing set out in Section 8 expressly advised: lead with the provider-and-pricing integrity measures, and package the access-and-funding restriction alongside them. Section 6 sets out the funding-below-cost architecture in full; the citation here is to illustrate that Schedule 1’s operative provisions cannot be sold to the Committee using Schedule 2’s integrity rationale.

The Ombudsman’s framing, the Quality and Safeguards Commission’s framing, and the central premise

The Joint Standing Committee on the NDIS Integrity of the National Disability Insurance Scheme inquiry evidence on 1 May 2026 carries the operative framing for the integrity work in the words of the Government’s own regulatory and complaints witnesses. The Commonwealth Ombudsman, Mr Iain Anderson, testified verbatim: “Complainants raise a number of issues with us. Very few of those actually involve allegations of fraud or sharp practice. For the most part they’re about the experiences of participants who have concern about changes to their plan, accessing the scheme and things like that.” The Ombudsman’s complaints data confirms that NDIS participants do not, in any large numbers, generate fraud complaints; participants generate access and plan-change complaints, which are precisely the complaints Schedule 1 of the present Bill will generate at scale. The Associate Commissioner of the NDIS Quality and Safegards Commission, Ms Natalie Wade, testified verbatim that “where we see noncompliance, our main concern is the impact on participants. Often it can be incredibly disruptive for participants. At best and at worst, it is significantly dangerous for participants when there is fraud and noncompliance with the code of conduct and other regulatory frameworks that are within the market and designed to keep participants safe.” The Quality and Safeguards Commission’s framing centres participant safety, with provider misconduct as the locus of intervention. The Deputy CEO Service Delivery at the NDIA, Mr Scott McNaughton, described the 190-staff participant-safeguarding response branch the Agency operates: “They might be putting a banning order on a provider. Our team will then work with the affected participants to relocate them to alternative providers and make sure they’ve got the right supports around them. They proactively check in with those participants, very much from a welfare and safeguarding perspective.”

The central premise of the Government’s integrity work, as the 1 May 2026 hearing evidence establishes, is that participants are protected from provider misconduct, not that participants are the source of the integrity problem. Schedule 2 of the present Bill is broadly consistent with that premise. Schedule 1 of the present Bill — by tightening eligibility on participants, with the Schedule 1 application provisions confirming that the new tests apply to existing participants on plan review regardless of when they became a participant — is structurally inconsistent with the central premise the Government’s own integrity witnesses operate under. The integrity work protects participants from

Submission 2079

misconduct in the supply chain. The Bill’s Schedule 1 inverts the relationship, treating the participant as the locus of the integrity problem.

A contemporary documented allegation illustrates the central premise in concrete terms. On Wednesday 20 May 2026, a Perth man appeared in the Perth Magistrates Court charged with 264 counts of fraud, totalling more than $54,000, allegedly committed against the intellectually disabled man he was employed as a homesharer to live with and support. Police allege the support worker helped the participant set up a bank account for NDIS payments to be deposited into, kept the bank card, and used the funds across an alleged eleven-year period for personal purchases including holidays, alcohol, tools, and massages. The charges are alleged and the support worker has denied the allegations and indicated he will plead not guilty. The Australian Broadcasting Corporation report quotes the Western Australia Police statement on the broader pattern such cases sit within: “Individuals who rely on carers or support arrangements should be treated with dignity, respect and honesty at all times. Anyone who suspects financial abuse, fraud or exploitation of a vulnerable person is urged to report the matter to police.” The participant is the victim in this matter; the alleged perpetrator is the paid support worker. Schedule 1 inverts that relationship and then compounds it: it treats the participant — the person the evidence base identifies as the party harmed — as the locus of the integrity problem, and makes that person’s access to supports harder to obtain and easier to lose. The integrity work of Schedule 2 — the provider-definition narrowing, the monitoring and investigation powers, the information-gathering powers uplift, and the plan-manager market restructure — is the substantive response to the structural pattern this case sits within. Schedule 1 punishes the victim of that pattern.

The inversion has a human consequence beyond the legal-architectural one. The Government’s “leakage” rhetoric, the “unsustainable growth” framing, and the integrity narrative attached to a Schedule that does not address provider integrity at all, have generated a public-discourse climate in which disabled Australians who use the scheme are characterised — by direct implication, in print and broadcast coverage — as gaming the system or receiving supports they do not deserve. That characterisation is the cohort- blame frame the National Autism Strategy 2025-31 commits the Commonwealth to actively reducing (Section 8 addresses the contradiction in full). It is not a side-effect of the package’s framing; it is the framing. My family lives under that climate every time the hetoric resurfaces. The harm is not abstract.

Key gaps in Schedule 2 that Section 10 develops

Schedule 2 is also partial against the recommendation set the Government’s own evidence base supports. The full risk-proportionate registration framework (NDIS Review Recommendation 17) is deferred to future Rules and limited to high-risk-supports providers, leaving the 154,000 unregistered providers in the 2023 Review’s data outside the mandatory-registration scope; the four-tier model and worker-registration scheme recommended by the 2024 Registration Taskforce, and the Taskforce’s participant-complaints reforms, are not in the Bill; and the launch-era Productivity Commission

Submission 2079

lineage on thin markets, workforce development and market stewardship is not engaged. Section 10 develops these gaps, together with the 2025 Australian National Audit Office findings on the Quality and Safeguards Commission, as inputs to the post-JSC redesign of Schedule 2 — not as amendments to the Schedule 2 currently before the Committee.

The separability of the two Schedules — different evidence bases, different procedural treatment

The case for treating the two Schedules separately strengthens with this analysis. The Schedules operate on different evidence bases and need different procedural treatment.

Schedule 1’s access-and-funding tightening operates on an evidence base that has already been assembled and community-consulted. The Disability Royal Commission heard from the community over four years and reported in 2023; the National Autism Strategy 2025–31 is a published cross-portfolio framework with substantive community consultation; the Independent NDIS Review (Bonyhady and Paul 2023) consulted; the Productivity Commission 2024 foundational supports paper is on the record; and the Evidence Advisory Committee Third Round Consultation continues. That evidence base does not support Schedule 1. The Joint Standing Committee on the NDIS Integrity of the National Disability Insurance Scheme inquiry’s terms of reference do not cover Schedule 1’s subject matter. Schedule 1 should be omitted from the Bill on the existing evidence base, independent of any further inquiry. Sections 3 through 8 of this submission build that case on the existing evidence base.

A further point reinforces the case for Schedule 1’s omission on the JSC evidence base as well. Schedule 1’s funding-below-cost authorities (new s.34A, new s.32K(3C), new s.45C — Section 6) actively worsen the thin-markets and quality-provider problems Schedule 2 is designed to address. The Schedules are not severable in their substantive effects: Schedule 1 makes the structural problems Schedule 2 responds to materially worse. Omitting Schedule 1 is not only the response on the evidence base for Schedule 1’s own subject matter; it is also the response that allows Schedule 2 to operate against the integrity evidence base on which it was designed, without being undercut by parallel provisions in the same Bill that thin the very provider markets the integrity work depends on.

Schedule 2’s integrity reforms operate on the evidence base the JSC is currently assembling — the JSC NDIS Integrity of the National Disability Insurance Scheme inquiry, the 1 May 2026 Canberra Hansard, the testimony of Mr Dardo and the Ombudsman and the Quality and Safeguards Commission, and the further hearings on 15 May 2026 in Sydney and 21 May 2026 in Melbourne. That inquiry reports on Thursday 2 July 2026. Schedule 2 should wait until the JSC reports; the JSC’s recommendations should then be integrated into the Schedule 2 design; and the amended Schedule 2 should be brought back for substantive community consultation. The Schedule 2 deferral does not dissolve the existing integrity infrastructure during the wait — the 2020 NDIS supports legislation, the NDIS Amendment (Integrity and Safeguarding) Act 2026 (Royal Assent 8 April 2026), the approximately 600-strong NDIA integrity workforce, and the Fraud Fusion Taskforce’s 24-

Submission 2079

Commonwealth-agency architecture all remain operational. Schedule 2 is the next layer of integrity work, not its foundation.

The two arguments are independent on independent grounds. Schedule 1 doesn’t wait for the JSC because the JSC isn’t investigating Schedule 1’s subject matter. Schedule 2 waits for the JSC because the JSC’s evidence base IS Schedule 2’s evidence base. The Committee has the authority under its standing orders to recommend each Schedule on its own terms, on its own evidentiary footing, and with its own procedural treatment. This is the core ask of Part B of my submission, developed in the recommendations at Section 16.

Submission 2079

Section 10 — Schedule 2 against the launch-era expert lineage

Section 9 set out what Schedule 2 does well, the integrity-rhetoric mismatch in the Government’s public framing of the package, and the case for deferring Schedule 2 until after the Joint Standing Committee on the NDIS Integrity of the National Disability Insurance Scheme inquiry reports on Thursday 2 July 2026. This section sets out the broader implementation-risk lineage the National Disability Insurance Scheme has carried since its founding, and what Schedule 2 engages and does not engage from that lineage.

The Explanatory Memorandum anchors Schedule 2 to two recent documents - the 2023 Independent NDIS Review (Bonyhady and Paul) Recommendation 17, and the 2024 NDIS Provider and Worker Registration Taskforce Final Advice. The launch-era expert evidence base is broader and longer-standing than either. The Productivity Commission’s founding 2011 Disability Care and Support inquiry, the Commission’s 2017 post-launch NDIS Costs study, the academic literature on quasi-market failure between 2013 and 2024, the establishment of the NDIS Quality and Safeguards Commission in 2017-18, and the 2025 Australian National Audit Office performance audit of the Commission’s regulatory functions, together describe a set of structural risks Schedule 2 in its current form leaves largely open. The arguments developed in this section are not put as amendments to be made to the Schedule 2 currently before the Committee. They are put as substantive inputs that should be incorporated into the redesign of Schedule 2 after the Joint Standing Committee NDIS Integrity inquiry reports.

The founding inquiry — Productivity Commission 2011

The Productivity Commission’s Disability Care and Support report, delivered in August 2011 across two volumes, was the founding architectural document for the National Disability Insurance Scheme. Beyond its economic case (developed in this submission’s Section 14), the 2011 report flagged five implementation-risk areas the scheme would need to manage to deliver on the consumer-choice model the Commission was recommending.

First, on market design,

The 2011 report cautioned that the consumer-choice approach required active market stewardship to function. The Commission identified strong rationales for a consumer-choice approach, but conditioned that recommendation on the existence of supporting market infrastructure, and warned that the market “would not function well in the absence of intervention”, particularly in geographic and complex-cohort segments.

Second, on workforce development,

The 2011 report identified workforce growth as a critical implementation risk. The Commission’s 2017 report later quantified the projection: approximately one in five new jobs across the workforce as a whole over the years following its assessment would need to be in disability care if the scheme was to be staffed at the projected participant numbers.

Third, on thin markets,

The 2011 report explicitly acknowledged that geographic and complex-needs cohorts would face market failure absent government intervention.

Submission 2079

blueprint contemplated that, without specific intervention, “Indigenous people with complex needs may have to relocate from their communities to regional centres or major cities to receive appropriate care and support” — a structural admission, in the founding document, that the market model was not expected to deliver to remote and First Nations communities without targeted intervention.

Fourth, on quality assurance and provider registration, the 2011 report recommended that service providers should “pre-qualify” to claim higher price limits, with pre-qualification “typically requir[ing] registration”. This is the conceptual precursor to the 2024 NDIS Provider and Worker Registration Taskforce recommendations the Bill cites in its Explanatory Memorandum.

Fifth, on service standards and enforcement, the 2011 report noted that enforcement would be a structural challenge requiring dedicated regulatory infrastructure — a concern that ultimately produced the establishment of the NDIS Quality and Safeguards Commission in 2017-18.

Schedule 2 of the present Bill addresses the fourth and fifth of these areas partially. The narrowing of the “NDIS provider” definition at new s.10C and the extension of National Disability Insurance Agency monitoring and investigation powers under the Regulatory Powers (Standard Provisions) Act 2014 respond to the quality assurance and enforcement concerns the founding Commission flagged. Schedule 2 does not address the first three areas. There are no workforce development provisions in the Bill. There are no specific provisions for remote, very remote, or complex-cohort market intervention. The market stewardship architecture the 2011 report implicitly required is not established in the Bill. Worse, Schedule 1 works in the opposite direction: the funding-below-cost authorities at new s.34A, s.32K(3C) and s.45C (Section 6) thin the very provider markets the 2011 report warned would fail without intervention, and new s.25A(2) codifies the cost of that market failure onto the participant (Section 11). Each of these is a structural gap the post-JSC redesign of Schedule 2 should engage — and each is made harder, not easier, by Schedule 1 provisions operating in the same Bill.

The post-launch evaluation — Productivity Commission 2017

The Productivity Commission’s October 2017 NDIS Costs commissioned study report assessed implementation three to four years into the scheme’s rollout. Its findings on market and workforce remain unresolved by the present Bill in their substance and in the period since.

The Commission’s headline market concerns were on the parliamentary record verbatim. On intake quality — a concern about whether the scheme’s planning processes and supporting market could keep pace with rollout, not a suggestion that too many people were being admitted: “the speed of participant intake. This is impacting on planning processes, the quality of plans, supporting infrastructure and market development.” On workforce: “1 in 5 new jobs over the next few years will need to be in disability care, but workforce growth remains way too slow.” On thin markets: thin markets would persist for

Submission 2079

participants “with complex, specialised or high intensity needs, or very challenging behaviours” and “in the absence of effective government intervention, such market failure is likely to result in greater shortages, less competition and poorer participant outcomes.” On the policy framework as a whole: “Present policy settings are unlikely to see enough providers and workers as the scheme rolls out.”

The Commission’s recommended responses included independent price monitoring and regulation, better-tailored responses to thin markets, formal and informal carers being allowed to provide more paid care, and a targeted approach to skilled migration.

The Commission’s 2017 recommendation on pricing was specific: an independent price-monitoring and regulation function, set at arm’s length from the Minister of the day. Schedule 3 Part 1 of the Bill does close to the opposite — it makes the Minister the pricing decision-maker. The Explanatory Memorandum cites this as implementing Action 11.3 of the 2023 NDIS Review. The difficulty is the gap between form and substance: a Ministerial pricing power can be described as “setting prices,” but it does not deliver the independence the Commission’s recommendation turned on. A discretionary Ministerial determination — accountable only through the legislative-instrument and disallowance process — is structurally different from, and is the inverse of, the independent monitoring and regulation the Commission envisaged. On pricing, the Bill moves away from the Commission’s recommendation while citing a later review to claim it is moving toward it.

Schedule 2’s civil penalty extension addresses one part of the enforcement gap the Commission identified.

Schedule 2 does not address thin markets, despite the Commission’s explicit warning that government intervention is required to prevent market failure for complex-cohort participants. Schedule 2 does not address workforce development. The Bill does not contain provisions on formal or informal carer paid-care arrangements; that gap compounds the parental-responsibility provisions in Schedule 1, which restate at new s.34(1G) the principle that family members and other informal supports should provide what is reasonable in the participant’s circumstances. The skilled-migration framework remains a structural gap the Government has not engaged through this Bill. The intake-quality concern the Commission identified will be compounded, not relieved, by the new functional capacity assessment regime in Schedule 1 Parts 1 and 8 from 1 January 2028; the regime narrows the timeframe in which participant-facing decision-making occurs. On thin markets, workforce, and intake quality alike, the pattern repeats: Schedule 2 is silent where intervention is needed, and Schedule 1 actively worsens the conditions the Commission warned about.

The post-JSC redesign of Schedule 2 should engage each of these gaps that the Government’s own statutory body identified almost a decade ago and that remain open in the current text. But there is a prior point the Committee should not lose: that redesign cannot succeed while Schedule 1 and Schedule 3 remain in the Bill. Schedule 1’s below-cost funding and s.25A(2) regressivity, and Schedule 3’s Ministerial pricing control, work directly against the market-stewardship, thin-markets and workforce improvements

Submission 2079

Schedule 2 needs. Omitting Schedule 1 and Schedule 3 is therefore not only the right response on each Schedule’s own evidence base (Sections 3, 8 and 9, and Part C) — it is a precondition for Schedule 2 ever achieving its intent. Section 11 develops the interdependence in full.

The academic quasi-market failure literature — 2013-2024

A decade of academic research has documented structural failure of the pure-competition model in the NDIS quasi-market. The work has been conducted across the University of New South Wales, the University of Melbourne, the University of Western Australia, Curtin University, and the Centre for Social Impact. Daniel Reeders, Gemma Carey, Anne Kavanagh, Helen Dickinson, Gordon Duff, David Gilchrist, Satish Chand, Damon Alexander, Eleanor Malbon, Natasha Cortis, Celia Green and others have produced a substantial body of peer-reviewed research and policy commentary. The headline findings are consistent across this body of work.

On the model itself, the literature is direct: “The first decade of the NDIS was characterised by the assumption that pure competition alone could deliver choice in quality services… In practice, this has not happened for many participants and supports.” On the structural alternative, the same literature is direct: “The NDIS is best described as a ‘quasi market’ that requires a higher level of market intervention relative to private sector markets. Governments, as market stewards, retain responsibility for ensuring the NDIS delivers the intended outcomes for all participants, even when using market-driven service delivery approaches.”

On participant outcomes, the literature is direct: research with people with disability and their carers found that their experience of service delivery had “not improved despite reforms driven by a quasi-market model; instead of delivering increased consumer choice and improved efficiency, service users experienced inadequate service supply, service cutbacks, and an increased emphasis on cost subsidisation.”

On the remote and First Nations service-access gap, the literature carries a quantification the Government should not have allowed to remain unaddressed for a decade: “In remote and very remote Australia, over one in three participants who have been in the NDIS for over one year are not accessing daily activity supports, and over one in four are not accessing other needed supports.” These are not participants who have been refused supports. These are participants who have been admitted to the scheme, who are entitled to supports, and for whom the supports are not actually being delivered in their communities because of the market failure the 2011 Productivity Commission report contemplated, the 2017 Commission report quantified, and Schedule 2 of the present Bill does not engage. The Market Capacity Framework research produced by Reeders and colleagues at the University of Western Australia and via the Social Science Research Network proposed a structured approach to identifying thin markets that the Government has not adopted.

Submission 2079

The launch-era academic critique is not a historical artefact. On Wednesday 27 May 2026 — the same week the Committee was receiving submissions on this Bill — the Economic Society of Australia (New South Wales) held a seminar by the e61 Institute economist Pelin Akyol titled “More Quasi than Market: Price Caps and Market Design in the NDIS”. Akyol’s framing of the central question was verbatim: “While much of the public debate has focused on tighter eligibility and reduced plan funding, a third lever has received far less attention: the design of the market itself … whether the NDIS functions as a competitive market, or whether price caps are doing most of the heavy lifting.” The credentialed-economics community is, in May 2026, restating the same structural critique the academic literature has published since 2013 — and which the Government’s own Productivity Commission first raised in its founding 2011 report. The pattern is consistent across fifteen years of independent expert and academic work; across those same fifteen years the market failure it identifies has gone substantially unaddressed.

Schedule 2 of the present Bill addresses provider conduct regulation through civil penalties, expanded monitoring powers, and deeds of arrangement. Schedule 2 restructures the plan-management market segment to address a specific conflict-of-interest critique on the parliamentary record at the 1 May 2026 hearing of the Joint Standing Committee NDIS Integrity inquiry. Schedule 2 restructures the support-coordination market segment to address a specific market-segment critique on the same evidence base. Each of these is a substantive provider-side intervention that the post-JSC redesign should retain. Schedule 2 does not address the fundamental thesis the academic literature has set out — that the pure-competition model is the structural problem, that quasi-market delivery requires market stewardship rather than market regulation alone — and, in the segments where market delivery has demonstrably failed, raises the legitimate question whether a market model is the right delivery vehicle at all — and that the remote and First Nations service-access gap will not be closed by tightening the rules on the providers that are not present in those markets. Schedule 2 expands the National Disability Insurance Agency’s regulatory function and does not move toward the broader market-stewardship architecture the academic literature has called for and that the 2023 NDIS Review Alternative commissioning for remote and First Nations communities paper proposed in concrete form.

The Quality and Safeguards Commission 2017-18 establishment, and the 2025 ANAO performance audit

The NDIS Quality and Safeguards Commission was established by Commonwealth legislation in December 2017 and commenced operations in New South Wales and South Australia on 1 July 2018, expanding progressively across all jurisdictions. The Commission was the Commonwealth’s response to a launch-era critique that the fragmented state-and-territory oversight model previously governing disability services was inadequate to a national scheme. The 2017-18 architecture is the principal current institutional answer to the enforcement concerns the founding 2011 Productivity Commission report identified.

Submission 2079

Schedule 2 of the present Bill extends the Commission’s powers — through the extension of the Regulatory Powers (Standard Provisions) Act 2014 monitoring and investigation provisions, through Rules-based mandatory registration administration for high-risk supports, and through associated structural reforms.

The National Disability Insurance Scheme Amendment (Integrity and Safeguarding) Act 2026, which received Royal Assent on 8 April 2026, adds enhanced information-gathering powers, anti-promotion orders, whistleblower protections, and a ninety-day cooling-off period for participants exiting providers. The 2026 Budget contains a substantial Quality and Safeguards Commission funding measure in line with the expanded role.

The 2025 Australian National Audit Office performance audit of the Quality and Safeguards Commission’s regulatory functions is the most recent independent assessment of the Commission’s effectiveness. Schedule 2 expands the Commission’s enforcement scope substantially. Whether the Commission can effectively absorb the expanded role depends on its current operational state. The ANAO’s audit is a published public document; the point is that its findings should be formally before this Committee and weighed in its deliberations before Schedule 2’s Commission-expansion provisions are passed (Recommendation 23). Unless the audit is squarely on the Committee’s record and considered, the Schedule 2 expansion proceeds on an untested assumption — and the audit’s own conclusion, that the Commission is only partly effective in its existing regulatory functions, is directly material to whether it can absorb an expanded role.

The substantive inputs to the post-JSC redesign of Schedule 2

The launch-era expert lineage produces five substantive inputs that the redesign of Schedule 2 after the JSC NDIS Integrity inquiry reports should incorporate. The inputs are framed below as inputs to the redesign, not as amendments to the Schedule 2 currently before the Committee. The procedural reason is set out in Section 9 of this submission. The substantive reason is that the redesign should integrate the launch-era expert lineage alongside the JSC’s findings on the integrity evidence base, and the result of that combined work is a Schedule 2 the evidence and community-consultation foundations of which are stronger than the current text.

First,

the post-JSC redesign of Schedule 2 should incorporate workforce development provisions addressing the one-in-five-new-jobs-required projection from the Commission’s 2011 and 2017 reports. The current Schedule 2 has no workforce development provisions. Without workforce supply growth, the Schedule 2 expansion of National Disability Insurance Agency enforcement against providers will reduce the provider pool further rather than improve participant outcomes — particularly in the geographic and complex-needs market segments where workforce capacity is already the limiting factor.

Second,

The post-JSC redesign of Schedule 2 should incorporate thin-markets and alternative-commissioning provisions for remote, very remote, complex-cohort, and First Nations communities where pure-market delivery has demonstrably failed. The 2023 NDIS

Submission 2079

Review Alternative commissioning for remote and First Nations communities paper and the Market Capacity Framework academic research provide the evidence base. The one-in- three remote-participant service-access failure documented in the academic literature cannot be closed by provider-side regulation alone. This is one of the clearest places the Schedules collide. New s.25A(2) of Schedule 1 actively codifies onto the participant the cost of the decades-documented thin-markets failure the Government has known about since the founding 2011 report — but s.25A(2) is only the sharpest instance. The submission’s position is that the whole of Schedule 1 and Schedule 3 should be omitted (Section 16), because both operate directly against the thin-markets, workforce and market-stewardship improvements the post-JSC Schedule 2 must make. Schedule 1’s regressivity codification and Schedule 2’s silence on thin markets are two faces of the same Government decision: to legislate around the thin-markets problem rather than solve it. The full interdependence — why Schedule 2 cannot achieve its intent while Schedule 1 and Schedule 3 remain — is developed in Section 11.

Third, the post-JSC redesign of Schedule 2 should require the National Disability Insurance Agency or the Quality and Safeguards Commission, as the appropriate market- stewardship body, to produce a market stewardship plan parallel to the Commission’s regulatory function. The 2017 Productivity Commission report and the 2023 NDIS Review architecture both contemplated a market-stewardship function broader than provider regulation. Market stewardship is not the same activity as market regulation. It includes commissioning, market intervention in thin segments, workforce development, and demand-side investment. The current Schedule 2 expands regulation; it does not establish stewardship.

Fourth, the Committee should require the 2025 Australian National Audit Office performance audit findings on the Quality and Safeguards Commission’s regulatory effectiveness to be tabled and considered before any vote on the Schedule 2 Commission- extension provisions. The principle that the Commission can effectively absorb the expanded enforcement role is an assumption Schedule 2 makes. The ANAO has independently assessed that question. The Parliament should not vote on the expansion without the assessment on the record. This input applies whether the Committee adopts the deferral position set out in Section 9 or proceeds on a different procedural footing. The ANAO findings should be on the record regardless.

Fifth, the Committee should scrutinise whether Schedule 2’s regulation-expansion model resolves the launch-era quasi-market failure critique that a decade of academic research has identified as the structural problem. Schedule 2 assumes that the existing market- based delivery model can be made to work through stronger regulation. If the structural problem is the model itself rather than the regulation around it, as Reeders, Carey, Dickinson, Cortis, Gilchrist, Malbon and the broader academic literature have argued since 2013, and as the Akyol e61 “More Quasi than Market” seminar restates in May 2026, than Schedule 2’s regulatory expansion will not deliver the participant outcomes the Government claims. The Committee is entitled, under its standing orders, to require the Government to demonstrate the contrary by reference to evidence, not by assumption.

Submission 2079

Synthesis — what Schedule 2 still leaves open from the launch-era expert lineage

Schedule 2 addresses a defined sub-set of the integrity and provider-conduct problems the Government has diagnosed in its immediate-recent advice — the 2023 Independent NDIS Review Recommendation 17, the 2024 NDIS Provider and Worker Registration Taskforce Final Advice, and the integrity evidence base the Joint Standing Committee NDIS Integrity inquiry is presently assembling.

It does not address the broader implementation-risk areas the founding Productivity Commission inquiry flagged in 2011, that the post-launch Commission evaluation reaffirmed in 2017, and that a decade of academic research between 2013 and 2024 has documented in detail. The launch-era expert lineage — thin markets, workforce development, market stewardship, intake quality, the quasi-market failure thesis — is largely absent from the current Schedule 2.

The Section 2 framing of this submission is therefore confirmed. Schedule 2 is partial in two directions. It is partial against the 2023-2024 advice the Explanatory Memorandum cites: the central architectural reform of the 2023 NDIS Review Recommendation 17 is implemented only narrowly, with the full risk-proportionate provider visibility framework deferred to future Rules and limited to high-risk supports providers; the four-tier registration framework recommended by the 2024 Taskforce is not implemented in primary legislation; the worker registration scheme and the participant complaints process reforms the Taskforce recommended are not in the Bill.

It is also partial against the 2011–2017 founding evidence base: the thin-markets, workforce, market-stewardship, and intake-quality concerns the Commission identified in its founding report and reaffirmed in its post-launch evaluation are not in the Bill. The framing that Schedule 2 is the right corrective requires acknowledging both layers of partiality.

The corrected procedural position I put to the Committee is the position set out in Sections 1, 2, 3 and 9 of this submission — and reinforced by the interdependence this section has documented: Schedule 1 and Schedule 3 actively work against the intent of Schedule 2. Schedule 1 should be omitted from the Bill on the existing evidence base — the Disability Royal Commission, the National Autism Strategy 2025-31, the Independent NDIS Review of 2023, the Productivity Commission’s 2024 foundational supports paper, and the active Evidence Advisory Committee Third Round Consultation — independent of any further inquiry.

Schedule 2 should be deferred until after the Joint Standing Committee NDIS Integrity inquiry reports on Thursday 2 July 2026. The inquiry’s recommendations should then be integrated into the redesign of Schedule 2 alongside the launch-era expert lineage set out in this section. The amended Schedule 2 that results should, before any further legislative step is taken on the provider-side and integrity-side reforms, be subject to both expert review — to confirm the redesign actually meets the market-stewardship, thin-markets and workforce intent the launch-era lineage identifies — and substantive community consultation under “nothing about us without us” principles. Drafting the redesign is not the end of the process; testing it against expert and community scrutiny is.

Submission 2079

Section 11 — How Schedule 1 and Schedule 3 work against Schedule 2:

the interdependence the Bill ignores

The submission has so far treated the three Schedules largely on their own terms. This section draws them together, because they are not independent. Part B’s central proposition - omit Schedule 1, defer Schedule 2 - rests on a structural fact the Bill’s design obscures: Schedule 1 and Schedule 3 actively work against the improvements Schedule 2 is meant to deliver, and Schedule 2 cannot achieve its stated intent while they remain in the Bill.

Three interdependencies make the point; the third is developed in full as this section’s worked example. First, Schedule 1’s funding-below-cost authorities (new s.34A, s.32K(3C) and s.45C - Section 6) thin the very provider markets Schedule 2’s integrity and registration reforms depend on: a regulator cannot improve a market the same Bill is simultaneously starving of viable providers. Second, Schedule 3 Part 1’s Ministerial pricing control (Section 12) is the inverse of the independent price-setting the Productivity Commission identified as a precondition for a functioning market (Section 10), and pulls directly against the market-stewardship the Schedule 2 redesign requires. Third — the worked example below — Schedule 1’s new s.25A(2) and Schedule 2’s silence on thin markets are two faces of a single Government decision: to legislate around the thin-markets failure the Government has documented for fifteen years rather than to solve it, by codifying its cost onto the participant.

The thin-markets evidence base — what the Government has known for fifteen years

The Productivity Commission’s 2011 Disability Care and Support report identified, on the public record, that geographic and complex-needs cohorts would face market failure absent government intervention. The Commission’s blueprint stated explicitly that, without specific intervention, “Indigenous people with complex needs may have to relocate from their communities to regional centres or major cities to receive appropriate care and support.” That is a Commission acknowledgement, in the founding document, that the market model was not expected to deliver to remote and First Nations communities without targeted intervention.

The Commission’s 2017 NDIS Costs commissioned study report reaffirmed and quantified the warning. Thin markets would persist for participants “with complex, specialised or high intensity needs, or very challenging behaviours”, and “in the absence of effective government intervention, such market failure is likely to result in greater shortages, less competition and poorer participant outcomes.” The 2023 Independent NDIS Review’s Alternative commissioning for remote and First Nations communities paper proposed a place-based and community-driven framework specifically to address the thin-markets failure the Commission had identified. The academic literature between 2013 and 2024 documents the operational consequence: in remote and very remote Australia, more than one in three participants who have been on the scheme for over one year are not accessing

Submission 2079

daily activity supports, and more than one in four are not accessing other needed supports. The Pelin Akyol e61 Institute seminar held on Wednesday 27 May 2026 restated the same critique in the credentialed-economics register: “whether the NDIS functions as a competitive market, or whether price caps are doing most of the heavy lifting.”

This is a fifteen-year evidence base, on the public record, from the Government’s own statutory body, from the Government’s own commissioned independent review, and from independent academic research. The market failure is documented. The cohorts affected are identified. The geographic and complex-needs concentration is quantified. And still the Government has not acted — not across fifteen years of accumulating evidence, and not in this Bill, which fails to address the failure in Schedule 2 and, through Schedule 1 and Schedule 3, makes the documented problem actively worse.

What new s.25A(2) does

Schedule 1 of the present Bill inserts new s.25A(2), which provides verbatim:

"Treatment may be appropriate treatment for a person’s impairment or
impairments regardless of whether the person’s individual circumstances restrict
the person from accessing the treatment." 

The Note to the sub-section reads: “A person’s individual circumstances include the person’s financial circumstances and geographical location.”

The new s.25A(3) provides a narrow statutory exception for “appropriate medical treatment the person cannot undertake for medical reasons“. Financial inability and geographic inaccessibility are explicitly outside the exception.

Read together with new s.24(5)(a) and new s.25(1B)(a) — the operative limbs of the permanence-and-treatment test the Bill introduces into both the standard access pathway and the early intervention access pathway — s.25A(2) means that whether a person can actually access the treatment the test contemplates is irrelevant to the eligibility decision. A treatment which the person cannot access for financial reasons, or which is not provided in the geographic location in which the person lives, is still capable of being treated as “appropriate treatment” the person has failed to exhaust.

The integration — the cost of thin-markets failure assigned to the participant

ing statute

When the thin-markets evidence base and the new s.25A(2) regressivity codification are read together, the Bill’s operative effect on a person in remote Australia, or in a regional area where complex-cohort supports are absent, or in a financial circumstance that prevents access to private treatment, is direct and traceable: the person is treated by the eligibility gate as having failed to exhaust appropriate treatment that does not exist where they live, or that they cannot afford to access. The treatment the test contemplates is unavailable in their market not because the participant has chosen not to access it, but because the Government, as market-steward, has not solved the thin-markets failure the

Submission 2079

e founding 2011 Productivity Commission report identified and that the 2017, 2023 and 2024 evidence base has reaffirmed over fifteen years.

This is not a marginal or incidental effect of the provision. It is the statutory point of s.25A(2). The Government has chosen, in the operative text of the Bill, to write into law that the question of whether the treatment can be accessed is not the test. The test is whether the treatment exists somewhere as a conceptual category, irrespective of whether the participant can access it. The cost of the participant’s inability to access — the financial barrier, the geographic distance, the absence of providers in remote and First Nations communities — is assigned to the participant. The eligibility gate operates as if the market is functioning, in cohorts where the Government’s own evidence base has documented for fifteen years that the market is not functioning.

The Schedule 2 silence on thin markets is the same decision

Schedule 2 of the present Bill could have engaged the thin-markets failure. The 2023 NDIS Review’s Alternative commissioning for remote and First Nations communities paper provided a specific framework. The 2024 NDIS Provider and Worker Registration Taskforce Final Advice contemplated a broader market-stewardship architecture. The 2011 and 2017 Productivity Commission reports identified the policy responses. The Schedule 2 design choices the Government made — narrowing the “NDIS provider” definition at new s.10C, extending the Regulatory Powers (Standard Provisions) Act 2014 monitoring and investigation regime, restructuring the plan-manager and support-coordination market segments — address provider-side conduct in the segments where providers exist. They do not address the market segments where the providers are not present. Schedule 2 expands regulation; it does not establish stewardship.

Read together, Schedule 1’s s.25A(2) regressivity codification and Schedule 2’s silence on thin markets are the two faces of the same Government decision. Schedule 1 legislates that the participant’s inability to access treatment, in cohorts and locations where the market has failed for fifteen years, is not relevant to eligibility. Schedule 2 does not commit the Government to solving the market failure. The combined effect, on the operative text of the Bill, is the assignment of the cost of decades-documented market failure to the participant rather than to the Government as market-steward, and the codification of that assignment in statute.

The market-steward inversion

The current NDIS Act 2013 at s.3(1)(g) commits the scheme to “enable people with disability to be included in the community in which they live.” The Australian Government, as State Party to the Convention on the Rights of Persons with Disabilities, has obligations under Article 19 to ensure people with disability can live in the community with choices equal to others and that “a range of in-home, residential and other community support services, including personal assistance necessary to support living and inclusion in the community, and to prevent isolation or segregation from the community” is available. The launch-era expert lineage developed at Section 10 sets out the market-stewardship

Submission 2079

architecture the founding 2011 Productivity Commission report contemplated and the academic literature has called for since. Market stewardship is a Government responsibility. Where the market fails to deliver to remote, First Nations and complex-needs cohorts — as the evidence base has documented for fifteen years — the responsibility to address that failure sits with the Government as market-steward, not with the participant.

The Bill inverts that responsibility. New s.25A(2) writes into statute that the participant’s inability to access market-failed treatment is the participant’s problem. Schedule 2’s silence on thin markets writes into the Government’s chosen Bill design that the broader market-stewardship architecture is not being established now. The two design choices, taken together, are a Government statement that the market failure will not be solved, and that the consequence of the unsolved failure will be borne by the participants in the cohorts the Government has documented as most affected. That is the integration point.

What the post-JSC redesign of Schedule 2 should incorporate

I addressed at Section 10 the substantive inputs the redesign of Schedule 2 after the Joint Standing Committee on the NDIS Integrity of the National Disability Insurance Scheme inquiry should incorporate. The thin-markets and alternative-commissioning provisions sit at the centre of that redesign. The post-JSC Schedule 2 should establish a market-stewardship architecture that engages the remote, First Nations and complex-needs cohorts the Government’s evidence base has identified, drawing on the 2023 NDIS Review Alternative commissioning paper and the academic Market Capacity Framework research. The redesign should commit the Government, as market-steward, to addressing the thin-markets failure rather than legislating around it — or, as the present Bill does through Schedule 1 and Schedule 3, actively making it worse.

Schedule 1’s new s.25A(2) is the sharpest single operative provision codifying that regressivity, but it is not the whole of the problem, and removing it alone is not the submission’s position. The submission’s position is that the whole of Schedule 1 be omitted (Section 16): s.25A(2) is one instance of a Schedule that — across its below-cost funding authorities and its permanence-and-treatment gate — works against the market Schedule 2 must repair. Section 7 addresses the statutory inconsistency of s.25A(2) with the Disability Discrimination Act 1992, CRPD Article 5 and the National Agreement on Foundational Supports 2026-31 commitment to align with Closing the Gap; this section adds the structural-integration dimension. Omitting Schedule 1 is not only a discrimination-law correction — it is the step that stops the Bill from codifying onto the participant the cost of a market failure the post-JSC Schedule 2 redesign is the appropriate place to address.

This is the close of Part B. In summary, the procedural position the submission puts to the Committee is: Schedule 1 should be omitted in full, on its own already-assembled and already-consulted evidence base (Part A); Schedule 2’s substantive direction is supported, but it should be deferred until the Joint Standing Committee NDIS Integrity inquiry reports

• Submission 2079

on 2 July 2026, then redesigned with the inquiry’s findings and the launch-era expert lineage and brought back for both expert and community scrutiny ( Sections 9 and 10 ); and Schedule 3 - the ministerial-control provisions this Part has shown pull against Schedule 2 ’s intent - is addressed in Part C , where the submission asks that it too be omitted . The Schedules are not severable in their effects : Schedule 2 ’ s repair of the market cannot proceed while Schedule 1 and Schedule 3 continue , in the same Bill , to damage it .

Submission 2079

Section 12 — Schedule 3 Part 1: Ministerial overreach in the pricing decision-maker model and funding-cap architecture

What Schedule 3 Part 1 introduces

Schedule 3 Part 1 makes the Minister the pricing decision-maker for NDIS supports. New s.45C authorises the Minister to make a legislative instrument specifying maximum amounts payable for NDIS supports or groups of supports, intensity ceilings for specified supports, and worker-to-participant ratios — and those specifications operate, on the operative text of the provision, regardless of the actual cost of providing the supports concerned.

New ss.34A and 32K(3C), introduced through the Bill’s funding-adequacy provisions, authorise funding for supports to be set below the actual cost the market requires to provide them — s.34A by Ministerial legislative instrument for old framework plans; s.32K(3C) by Rules for new framework plans. Together, these provisions produce a single policy outcome: the Minister has statutory authority to set prices below cost, without a requirement to establish that the price is financially viable for the provider, without a requirement to model the market impact of the decision, and without any merits review pathway available to participants or providers who are affected.

The funding-below-cost authority is not a campaign characterisation; it is on the face of the Bill. New s.34A(5), inserted by Schedule 1 Part 4, reads verbatim:

"To avoid doubt, the determination has effect even if the result is either or both of
the following: (a) the funding provided under a participant's plan for a reasonable
and necessary support is less than the total cost of the support; (b) the funding
provided under a participant's plan for all reasonable and necessary supports
funded under the plan taken as a whole is less than the total costs of the
supports." 

The “to avoid doubt” construction confirms that the Minister’s determination operates regardless of whether the resulting funding meets the cost of the supports. Combined with the s.45C maximum-amounts authority that is the subject of this section, the architecture is coherent: Schedule 1 authorises below-cost funding amounts within plans; Schedule 3 authorises maximum amounts payable for the supports those plans fund. Section 6 sets out the Schedule 1 funding-below-cost architecture in full; this section addresses the pricing-decision-maker concentration in Schedule 3 that compounds it.

That is a significant concentration of power. It is not consistent with the pricing architecture the Productivity Commission contemplated when it designed the NDIS, and it introduces accountability deficits that I ask the Committee to examine carefully.

The pricing framework the NDIS has operated under

Since commencement, the NDIS has operated a published pricing framework — the Pricing Arrangements and Price Limits, previously the NDIS Support Catalogue, and before that the NDIS Price Guide — developed by the National Disability Insurance Agency

Submission 2079

through an annual process of cost analysis, market consultation, and sector engagement. That process has not been perfect, and the sector has documented its limitations extensively: below-cost pricing in thin markets, inadequate indexation, and workforce supply pressures that the price guide has been slow to reflect. None of that is in dispute here.

What matters for the argument in this section is that the current pricing framework has three structural accountability features that Schedule 3 Part 1 removes. The NDIA, as a statutory agency, is operationally accountable for its pricing decisions — it is required to publish its methodology, consult with the sector, and answer for its analysis through the ordinary accountability processes of Senate Estimates, Senate committee inquiries, and ANAO audit. The pricing decisions the NDIA makes are administrative decisions, made under statutory authority, and in principle subject to review. And the statutory functions of the NDIA at s.9(2)(f) of the current NDIS Act 2013 include market stewardship — a function that is operationally anchored in the pricing process, because the price the scheme pays for a support shapes whether the market for that support remains viable.

Schedule 3 Part 1 removes the pricing function from the NDIA and places it in the Minister’s office, to be exercised by legislative instrument. Each of those three structural features changes materially when this transfer occurs.

What the Productivity Commission’s design contemplated

The scheme the Productivity Commission designed in 2011 was built on a competitive market model in which, over time, provider competition was expected to discipline prices. The founding 2011 report was explicit that price guidance was a transitional architecture — necessary in the early years before the market matured, but intended to give way as the scheme scaled. The Commission’s 2017 follow-up report, NDIS Costs, found that the competitive pricing model had not materialised as anticipated: thin-market conditions were extensive across regional, remote, and outer-metropolitan areas; market concentration in certain support categories had increased rather than dispersed; and the scheme required active market stewardship, including independent pricing oversight, to manage the structural market failure the pure-competition model had not corrected.

Neither the 2011 nor the 2017 design anticipated what Schedule 3 Part 1 now proposes: than pricing authority would be held by the Minister of the day, exercisable by legislative instrument, with parliamentary disallowance as the sole accountability mechanism. The Productivity Commission’s architecture — imperfect as its implementation has been — maintained the operational separation between political direction and administrative pricing. Schedule 3 Part 1 collapses that separation. There is a deeper incoherence in the design. A genuinely competitive market — the model the Government invokes when it describes the NDIS as participant-led and choice-driven — would allow price to rise where supply is short, drawing providers and workers into under-served segments. A below-cost cap set by Ministerial instrument does the opposite: it suppresses the very price signal a

Submission 2079

market needs to correct itself, while retaining the language of a market. The Bill runs a market on one side of the ledger and overrides its central mechanism on the other.

Three structural failures in the accountability architecture

The “have regard to” discretion is not bounded by objective standards. New s.45C(17) specifies the matters the Minister must “have regard to” when making a pricing determination. They are three: any relevant advice provided by the NDIA; the need to ensure the financial sustainability of the NDIS; and the objects and principles of the Act. Note what is absent from that list. The NDIA’s own advisory obligation at s.45C(14) requires the Agency’s advice to have regard to the cost of safe, efficient, high-quality acquisition or provision of supports and value for money, alongside diversity and competition and financial sustainability. The Agency must consider cost; the Minister is not required to. The Minister’s “have regard to” obligations run to NDIA advice, scheme financial sustainability, and the Act’s objects — not to what it actually costs to provide the supports being priced.

“Have regard to” is not “must be satisfied that,” and it is not “must not determine unless.” It is a consultative obligation, not a substantive constraint. The legal effect is that the Minister must turn their mind to the listed considerations — the provision does not require the Minister to be governed by them, and it does not empower a court or tribunal to review the substantive merits of a pricing determination on the basis that the Minister weighted the considerations incorrectly. The financial-sustainability obligation sits in compound with Schedule 1’s new s.3(1)(d) Objects clause amendment, which subordinates the scheme’s reasonable-and-necessary supports objects to financial sustainability: the Minister must have regard to the objects of the Act; the objects now include financial sustainability as a primary purpose; both “have regard to” pathways run toward the same fiscal destination. For a power of this kind — one that sets the maximum price for every support in a scheme on which approximately 760,000 NDIS participants currently depend, once Australia’s obligations under CRPD Article 19 (living independently and being included in the community) and Article 28 (an adequate standard of living) are engaged — a three-item “have regard to” obligation that does not include the cost of provision is not an adequate accountability standard. A power capable of determining whether a support is funded at a level that allows it to exist should be bounded by an obligation to be satisfied of the cost of providing it, not merely to consider a list from which cost is absent.

Compare this to the NDIA’s current pricing process. The Agency publishes its methodology, grounds the price in cost modelling and market analysis, consults with the sector before deciding finalising the determination, and accepts challenge through established accountability processes where participants or providers believe the analysis is wrong. The shift from that process to a Ministerial legislative instrument with a three-item “have regard to” obligation, where cost of provision is not one of the three items, represents a degradation of accountability, not an enhancement of it.

Submission 2079

A further feature of s.45C reinforces the accountability concern. The Note following s.45C(1) exempts pricing determinations from the sunsetting provisions in Part 4 of Chapter 3 of the Legislation Act 2003. Legislative instruments normally sunset after ten years, requiring Parliament to actively renew them. The pricing determination is exempt. A determination made under s.45C can remain in force indefinitely without Parliament being required to revisit it. The accountability mechanism that sunsetting provides — forcing a parliamentary reconsideration of whether an instrument remains appropriate — is removed by design.

Disallowance is not routine accountability. A legislative instrument made under s.45C takes effect when it is registered (or on the date specified in the instrument) unless either House of Parliament passes a disallowance motion within fifteen sitting days of the instrument being tabled. The default, therefore, is that the instrument stands. Parliament must actively intervene to reverse it; the Minister is not required to demonstrate the instrument is appropriate before it takes effect.

For a government that holds, or can secure, a majority in either House when a disallowance motion is moved, the disallowance mechanism is not a meaningful accountability constraint. The pricing determination takes effect; the burden of reversal falls on an opposition or crossbench capable of passing the disallowance motion — which may not exist. Comparing this to the existing accountability architecture: the NDIA’s pricing decisions are made through a process the sector can engage with before finalisation, not after. The legislative instrument model reverses that sequence. The decision takes effect first; challenge depends on political will in both Houses of Parliament after the fact. This is not a marginal procedural point. It means that, in the ordinary case of a government with the numbers, the Minister can set a below-cost price that takes effect and stands, with no body — not a court on the merits, not a tribunal, and not the Parliament in practice — able to require that the price reflect the cost of the support. For a determination that governs the price of every support in the scheme, disallowance-only accountability is, in substance, close to no external accountability at all.

Reviewability-stripping — the most significant accountability failure. Administrative decisions of the NDIA are reviewable by the Administrative Review Tribunal. Legislative instruments are not. They are subject only to judicial review in the Federal Court on constitutional grounds or for jurisdictional error, and judicial review does not assess the substantive merits of a decision — it tests whether the decision-maker had the legal power to act and followed the required process. A provider whose pricing has been set below the cost of service, or a participant whose support has been intensity-capped at a level that doest not reflect their assessed functional need, has no administrative merits review pathway against a legislative instrument pricing determination. They can bring judicial review, but the threshold is materially higher than the threshold for merits review, the cost is materially greater, and the remedy — quashing the instrument — is a blunter tool than the merits review pathway of substituting a different price.

Submission 2079

This is the finding Villamanta Disability Rights Legal Service independently validated in its community legal forum on the Bill on 20 May 2026: the Minister’s pricing alterations under this architecture are not reviewable decisions in the administrative sense that currently applies to NDIA pricing actions. Villamanta reached that conclusion through their own analysis of the Bill’s operative text. It confirms that what the transfer from NDIA administrative function to Ministerial legislative instrument achieves, in practical accountability terms, is the removal of the merits review pathway that currently protects participants and providers against pricing decisions that are wrong.

The review-rights stripping is not unique to Schedule 3 Part 1; it is a pattern that runs across the Bill. New s.50A(4)(b), inserted by Schedule 1 Part 5, reads verbatim:

• “the making of the new plan does not involve the making of any reviewable decision.”

The annual automatic renewal of a participant’s plan by force of law under new s.50A is not, on the operative text of the provision, the product of a reviewable decision. Administrative merits-review against a reviewable decision under s.99 of the Act, and the Administrative Review Tribunal pathway that follows from it, do not apply at the renewal point. Section 7 of this submission addresses the CRPD Article 12(4) statutory-inconsistency dimension in full. What is relevant in the Schedule 3 Part 1 context is the architectural pattern: the Bill removes the merits-review pathway at the plan-renewal point under Schedule 1, and the Bill removes the merits-review pathway at the pricing-determination point under Schedule 3. The combined effect is the structural removal of the merits-review pathway at the two points the participant is most likely to need it — when their plan is being renewed below the funding level of the prior plan, and when the supports under the plan are being priced below cost. The Committee is asked to consider whether the Bill’s two-point review-rights stripping is consistent with the administrative-review architecture the Commonwealth ordinarily applies to decisions affecting access to substantive social-policy programs.

Taken together, the three structural failures produce an accountability architecture in which the Minister holds an unbounded discretion to set prices below cost, the decision takes effect by default, and the affected party’s only remedy — absent a parliamentary majority willing to disallow — is expensive Federal Court litigation on narrow jurisdictional grounds. That is not the accountability architecture the Productivity Commission designed for a scheme of this size and complexity.

The self-management exclusion creates a two-tier pricing system. Section 45C(2) provides that the pricing determination does not apply in relation to supports funded under a participant’s self-managed funding — it applies only where funding is managed by a registered plan manager or the NDIA. Self-managing participants can therefore engage providers at rates above the Ministerial cap; participants on NDIA-managed or plan-manager-managed funding cannot. This creates a two-tier price access system that tracks the self-management divide. Self-management is more accessible to participants with higher administrative capacity, stronger informal support networks, and greater familiarity with the scheme’s bureaucratic processes — broadly, the cohort already better positioned

Submission 2079

to navigate the system. The Ministerial price cap therefore operates most stringently on the cohort with the greatest structural barriers to self-management: participants in lower- income and remote circumstances; participants whose disability directly affects their capacity to manage their own funding. In thin-market areas where the only available providers charge above the cap, NDIA-managed participants lose access to those providers; self-managed participants retain it. This is the most consequential feature of the price cap, and it should be stated plainly: a participant whose funding is NDIA-managed or plan-managed cannot lawfully pay the difference between the capped price and the provider’s actual price — there is no gap-payment mechanism available to them. If no provider will deliver the support at the Ministerial cap, the support is not delivered at a lower price; it is simply not delivered. The cap does not discount the support — it forgoes it. The practical consequence is that the s.45C pricing power is also a power to remove supports from the scheme by under-pricing them: a Minister who sets the cap below the price at which any provider will deliver a support has, in operative effect, removed that support for every participant who cannot self-manage — without ever making a reviewable decision to do so. The equity gradient runs directly contrary to the CRPD Article 5 and Disability Discrimination Act 1992 s.6 obligations the scheme is required to honour.

The thin-markets compounding problem

Section 11 of this submission establishes the fifteen-year thin-markets evidence base: the structural under-supply of NDIS supports in remote, rural, and outer-metropolitan areas; the market concentration among a small number of large providers in certain support categories; and the ways in which the Bill’s new s.25A(2) — by deeming treatment “appropriate” regardless of whether the participant can afford it or reach it — makes the thin-markets gap worse. I do not repeat that analysis here.

What I add in the context of Schedule 3 Part 1 is the pricing dimension. Price caps set below actual cost do not reduce cost — they reduce supply. A provider operating in a thin- market area, whose cost structure reflects the genuine expense of delivering a support to a geographically isolated participant, cannot sustain operations below cost indefinitely. The effect of a below-cost price cap in a thin market is to suppress the incentive for existing providers to continue operating in that market and to suppress the incentive for new providers to enter it. Pelin Akyol’s 2025 research on quasi-market failure in the NDIS identifies the price cap as the primary lever through which the scheme’s market architecture is failing: the cap does most of the heavy lifting in the quasi-market failure (“More Quasi than Market,” e61 Institute / Economic Society of Australia (NSW) seminar, 27 May 2026).

A Ministerial discretion to set that cap, exercisable without objective-standards constraint, without independent review, and without a merits-review pathway for affected parties, compounds the thin-markets failure rather than addresses it. Schedule 3 Part 1 is being applied to a market the Productivity Commission identified as structurally deficient in 2017, and Akyol’s research confirms remains structurally deficient today. Concentrating the

Submission 2079

pricing power in the Minister without an independent oversight architecture is not the market stewardship intervention the evidence base supports.

This is not abstract for my family. My fourteen-year-old daughter, an NDS participant, eeds speech therapy and occupational therapy; even in metropolitan Perth we waited a great time to find providers with availability, and we now travel thirty minutes each way to reach them. Her plan funds one session a month against a clinical recommendation of weekly, and we manage fortnightly by stretching what we have. The supports are already under-funded and already hard to reach at current price levels - and that is before a Ministerial power to set those price levels lower still, in the same Bill that narrows her access on reassessment. A below-cost cap does not reach my family as a question of market design. It reaches us as a therapy session that is recommended, needed, and not funded.

The structural inversion — from independent market stewardship to Ministerial direction

The NDIA Act 2013 at s.9(2)(f) gives the NDIA market facilitation and stewardship functions. The scheme’s pricing architecture is the principal tool through which market stewardship is exercised — the price guide shapes what the market will bear, what providers can offer, and whether participants in thin-market areas can access supports at all. Concentrating the pricing function in the Minister removes it from an operationally accountable statutory agency and places it in a political office. This matters for the independence of the pricing process: the Minister’s pricing decisions will be made in the context of fiscal constraints (the $37.8 billion savings target over four years established by the Bill’s Budget measure), political priorities, and Budget cycles — not only in the context of what the market actually costs to operate. Independent pricing oversight is not a bureaucratic nicety; it is the structural protection against the pricing power being used as a savings instrument rather than a market management tool. Schedule 3 Part 1 does not merely remove that protection; in the context of a $37.8 billion savings target legislated in the same Budget, it hands the Government a direct lever to decide, as a political and fiscal matter, how much support the scheme will fund - by setting the price at which supports can be delivered. The concern is not only that independence is lost; it is that the power is configured to let fiscal targets, rather than the cost of supports, determine what participants actually receive.

The Productivity Commission’s 2017 report identified independent market stewardship as essential to the NDIS’s long-term sustainability. Concentrating market-stewardship pricing functions in the Minister — on the basis of a “have regard to” discretion with disallowance- once parliamentary accountability and no merits review pathway — inverts the design the Commission identified as necessary. I am not aware of any evidence the Government has produced that addresses why that inversion is appropriate, or how the accountability deficit it creates is meant to be remedied.

Submission 2079

What should happen instead

My position on Schedule 3 is that it should be omitted from the Bill. The underlying question Schedule 3 Part 1 gestures at is real - how maximum amounts, intensity ceilings, and worker-to-participant ratios should be set, by whom, and against what evidence base

• but the Bill answers it incoherently. There are two coherent models for a scheme of this kind: a genuinely market-led model with independent price oversight that lets supply and price adjust, or a publicly-commissioned, not-for-profit delivery model in which the Government provisions supports directly. Schedule 3 Part 1 is neither. It keeps the market’s risks - participants depend on private providers choosing to deliver - while removing the market’s price mechanism and the independence of the body that sets prices. That is the worst of both: political control of price, with neither market discipline nor a public-provision guarantee. But the answer to that question is not a Ministerial legislative instrument with a “have regard to” discretion, disallowance-only accountability, and no merits review pathway for participants or providers. The answer is an independent pricing oversight architecture - one that also engages the thin-markets problem directly, since price-setting and market stewardship cannot sensibly be separated - developed through substantive consultation with the disability sector under “nothing about us without us” principles, with published methodology, operational transparency obligations, and accessible review for affected parties. That architecture should be designed and consulted on, and tested against both expert and community scrutiny, before any legislative step on the pricing function is brought back to the Parliament.

The omission I am asking for is not a position that the pricing function should be frozen as it currently operates. The current Pricing Arrangements and Price Limits framework has documented limitations, and the sector has raised concerns about below-cost pricing under the existing NDIA-led framework for years. What I am asking is that the remedy to those limitations not be the concentration of pricing authority in a Ministerial legislative instrument with the accountability architecture I have described in this section. The evidence base - the Productivity Commission’s 2011 and 2017 design work, the academic thin-markets literature, and Villamanta’s independent legal analysis of the accountability architecture the Bill creates - supports the conclusion that Schedule 3 Part 1 makes the pricing accountability problem worse, not better. Omission of Schedule 3 from the Bill, followed by the co-design of independent pricing oversight through genuine community consultation, is the procedurally and substantively appropriate response.

Submission 2079

Section 13 — Schedule 3 Part 2: Automation of administrative action,

evaluative determinations, and the participant safety risk against the Robodebt Royal Commission findings

This is the provision the campaign and this submission refer to as RoboNDIS. Schedule 3 Part 2 authorises the automation of decisions about what individual disabled people receive from the scheme — including the reasonable-and-necessary determination at s.33 — under an oversight architecture whose central safeguards, on the face of the Bill, do not affect the validity of a decision even if they are breached. Australia has done this before. The 2023 Royal Commission into the Robodebt Scheme examined an automated welfare- decision system applied to a vulnerable cohort, found that it caused profound and systemic harm, and heard evidence connecting it to the deaths of affected people. The parallel here is not rhetorical; it is structural, and it runs through every safeguard this section examines. I set out below what Schedule 3 Part 2 provides, measured against what the Robodebt Royal Commission identified as necessary — and why, on that measure, it should be omitted from the Bill.

What Schedule 3 Part 2 introduces

Schedule 3 Part 2 creates a new Division 5 — Automation of administrative action — within Chapter 4 of the NDIS Act 2013, comprising new ss.59B to 59E. The provisions authorise the CEO to arrange, under the CEO’s oversight, for computer programs to take administrative action under designated provisions of the Act and relevant legislative instruments.

The definition of “administrative action” in new s.59B(3) is broad: it covers making, refusing or failing to make a decision or part of a decision; exercising, refusing or failing to exercise, a power; performing, refusing or failing to perform, a function or duty; and doing, refusing or failing to do, anything related to any of the above. New s.59B(4) then makes explicit what would otherwise have required careful inference: administrative action covered by this framework includes action that involves a discretion being exercised, an evaluative judgement being made, or a state of mind being formed. The Bill names these collectively as “evaluative determinations” and defines them in new s.9B by that label.

The provisions therefore do not authorise the automation of routine mechanical calculations only. They explicitly authorise the automation of judgement — the kinds of individual assessments the scheme is built on: whether a particular support is reasonable and necessary for this participant in this context; whether this person’s needs, in light of their circumstances, warrant this level of funding. The Bill is explicit about this. The safeguard question is what oversight architecture that explicitness demands, and whether the architecture the Bill provides is adequate.

Submission 2079

The designated provisions from commencement

New s.59C(1) specifies the provisions of the Act that are designated for automation from the moment Schedule 3 Part 2 commences. They are section 33; section 45; section 45A; and section 45C.

Section 33

Section 33 is the reasonable-and-necessary determination. It is the provision under which the CEO assesses, for each support included in a participant’s plan, whether the support satisfies the statutory criteria: whether it is related to the participant’s disability; whether it is likely to be effective and beneficial; whether it represents value for money; whether it is reasonable having regard to the participant’s informal and mainstream support networks, and to the capacity of the participant’s family, carers, networks and community to provide support; and whether it is most appropriately funded or provided through the scheme rather than through other general systems. This is not a mechanical calculation. Section 33 is the provision that determines what individual NDIS participants actually receive. It is an individualised assessment of need, circumstances, and competing support systems.

From the moment Schedule 3 Part 2 commences

Section 33 is on the initial designated list. From the moment Schedule 3 Part 2 commences, the CEO may arrange for computer programs to take administrative action — including evaluative determinations — under section 33, without Parliament being required to do anything further and without any parliamentary accountability for that decision (the arrangement under s.59B(12) is not itself a legislative instrument). The most consequential individual eligibility assessment in the scheme is authorised for automation before the framework has been tested, before any independent evaluation of its participant-safety implications has been conducted, and before any of the community consultation under “nothing about us without us” principles that decisions of this scale warrant. To be unambiguous about the sequence: the Bill authorises the automation of the single most consequential eligibility judgement in the scheme from the day Schedule 3 Part 2 commences — an untested method, applied to the determination that decides what each disabled person actually receives, with no statutory requirement that it first be evaluated, validated against participant-safety criteria, or consulted on with the cohort it will be applied to.

Expansion by Minister

The Minister may also, by legislative instrument under new s.59C(2), expand the list of designated provisions to any provision of the Act, any legislative instrument made under the Act, or any legislative instrument made under the NDIS Amendment (Getting the NDIS Back on Track No. 1) Act 2024. The only constraint where evaluative determinations are involved is that the Minister “must be satisfied that it is appropriate” to expand (s.59C(3)). That is a ministerial self-satisfaction standard. There is no objective test a court can review for compliance; there is no independent oversight body whose approval is required; there is no parliamentary approval before the instrument takes effect.

Submission 2079

The safeguard architecture and its structural gaps

New s.59E sets out the oversight and safeguard framework. I examine each element against what the 2023 Royal Commission into the Robodebt Scheme identified as necessary.

CEO oversight — s.59E(1) and (3)

The CEO must take all reasonable steps to ensure that automated action is action the CEO could validly take under a designated provision. This is the primary safeguard. It is, however, immediately qualified by s.59E(3): a failure to comply with the reasonable-steps obligation does not affect the validity of the automated action. The safeguard is real as a process obligation but it provides no protection to the participant whose decision was incorrectly automated. The automated action stands whether or not the CEO complied. This directly parallels the Robodebt experience: systemic failures in the automated calculation were not caught by internal oversight mechanisms, and the affected individuals bore the consequences of errors that the oversight system was not designed to detect contemporaneously. An oversight obligation whose failure does not affect validity is not a safeguard for the participant — it is an obligation whose breach attracts no operative consequence.

Notification — s.59E(4) and (5)

If another provision of the Act requires a notice of the administrative action to be given to the person, that notice must inform the person that the action was taken by a computer program. Participants will therefore know, at the point of receiving a decision notice, that a computer made the decision. This is a meaningful transparency measure as far as it goes. It goes nowhere, however, when paired with s.59E(5): a failure to comply with the notification obligation does not affect the validity of the automated action. A participant can be on the receiving end of an automated decision without being told it was automated. The validity of the decision is unaffected. The participant’s right to seek review is not triggered by the notification failure. There is no remedy at the point of breach.

Transparency of input data

The notification provision tells participants that automation was used. It does not tell them what data was fed into the program, what logic the program applied, or why the program reached the outcome it did. Villamanta Disability Rights Legal Service independently validated this gap in its community forum analysis of the Bill on 20 May 2026: participants cannot see the input data driving automated decisions. The standard operating procedure instrument (new s.59D) is a notifiable instrument — it will be published, and participants can in principle read the general rules the program applies. But “the SOP says the program applies rule X to criterion Y” is not the same as “here is the data the program used about you, and here is why it concluded that criterion Y was or was not met in your specific case.” Without input-data transparency, a participant’s ability to identify that a decision is wrong and to take steps to challenge it is materially impaired.

The arrangement is not a legislative instrument — s.59B(12)

The CEO’s decision to arrange for automation of a category of administrative action is not a legislative instrument. It is not tabled in Parliament, it is not disallowable, and Parliament has no advance

Submission 2079

visibility of it. The CEO is required to publish a statement on the Agency’s website that an arrangement has been made and which provisions it covers (s.59E(6)) — but publication after the fact is not the same as parliamentary approval before. The decision to automate the most consequential eligibility assessments in the scheme is one the CEO can make unilaterally, without parliamentary notice or disallowance, and with no external approval requirement.

The standard operating procedure is notifiable, not disallowable — s.59D(6). Where automation involves evaluative determinations, the CEO must make a standard operating procedure instrument specifying the circumstances in which the evaluative determination will always be made in a specified way (s.59D(2)). That instrument is a notifiable instrument: it is published and Parliament can read it. It is not a disallowable instrument: Parliament cannot vote to disallow it. The rules governing how automated evaluative determinations — automated judgements — are made in a scheme serving 760,000 disabled Australians are subject to publication but not to parliamentary accountability.

Retrospective accountability only — s.59E(7). The CEO must include, in the Agency’s annual report, the kinds of substituted actions taken (where the CEO intervened because an automated decision was not correct or preferable) and the kinds of automated action the CEO was satisfied were incorrect. Annual reporting is retrospective, aggregated, and occurs after the affected participants have already received the wrong decision. The Robodebt Royal Commission found that retrospective accountability — the “we can fix errors after the fact” defence — is not an adequate substitute for contemporaneous safeguards. The errors in Robodebt were not caught by retrospective reporting mechanisms; they were caught because individual affected people contested individual decisions, often at significant personal cost. The NDIS participant cohort, as I describe below, includes a substantial proportion of people for whom that individual-contestation pathway is structurally compromised by their disability.

There is a further dimension the Robodebt experience makes unavoidable. Robodebt’s errors persisted for years not only because retrospective reporting failed to catch them, but because warnings were dismissed and those who raised the alarm were resisted rather than heard; the Royal Commission documented a culture in which the scheme’s continuation was defended against the evidence. Schedule 3 Part 2 places the automation of these decisions, and the rules that govern them, under a degree of executive and Ministerial control — arrangements that are not legislative instruments, standard operating procedures that are notifiable but not disallowable, and expansion of the automated provisions on Ministerial self-satisfaction — that affords even less external visibility than the arrangements Robodebt operated under. A scheme this consequential for disabled Australians cannot safely rest on the assumption that the political and administrative actors running it will reliably surface their own systemic errors. The Robodebt record is that they did not.

Submission 2079

The plan suspension and revocation architecture

Sitting alongside the automation provisions, but operating through a different mechanism, is new s.40A — Participants who are not contactable. The provision authorises the CEO to suspend a participant’s plan if the CEO is satisfied that reasonable attempts to contact the elementary for the purposes of requesting information or reports have been made and the participant is not contactable. A suspension decision under s.40A is a reviewable decision (Note to s.40A(1)) — the participant can challenge it. But the challenge pathway requires the participant to engage with the review process, which requires the participant to know the decision has been made, to understand that review is available, and to have the communication and administrative capacity to initiate a review. For participants whose disability affects their communication and engagement patterns — the population s.40A most directly concerns, because the trigger is non-contactability — each of those steps presents a structural barrier.

New s.30(1A)(b) then provides that the CEO may revoke a person’s participant status if the participant’s plan has been suspended under s.40A for at least 90 days. The combination of the two provisions creates a 90-day pathway from suspension to revocation — not automatic in the sense that it requires no further CEO decision, but structural in the sense that the CEO’s revocation power opens once the 90-day period passes without the participant re-engaging. A participant in an episode of Autistic burnout, a mental health crisis, or a period of severe functional reduction — periods during which engagement with bureaucratic processes is hardest and support needs are likely highest — faces both the suspension of the plan that funds their supports and the opening of the revocation pathway if the episode extends past ninety days.

New s.50A adds a further dimension. Under that provision, a participant’s old framework plan is renewed by force of the section — automatically, through the operation of law — on its end date. New s.50A(4)(b) makes explicit that the making of the renewed plan does not involve the making of any reviewable decision. The Note to new s.59B(1) confirms that s.50A operates “through the operation of law and does not involve an administrative decision being made” — placing it outside even the automation framework entirely. The Minister may, by legislative instrument under s.50A(3), determine alterations to renewed plans — adjustments to funding components that take effect across all participants whose plans renew, without participant input, without individual assessment, and without any reviewable decision being made. Participants whose plans are renewed under s.50A have no review pathway for the renewal itself. The only path to altering a renewed plan is to request a reassessment — a process that places the initiative back on the participant. And if Schedule 1 passes, that escape route narrows to almost nothing: reassessment becomes subject to the new permanence-and-treatment gate and the tightened access tests, so the one participant-initiated path out of an operation-of-law plan outcome is itself gated by Schedule 1. Schedule 1 and Schedule 3 operate in tandem — Schedule 3 removes the review pathway at renewal, and Schedule 1 closes the reassessment pathway that is meant to be the alternative — and both run against the participant-protective intent the Government locates in Schedule 2 (Section 11).

Submission 2079

The Robodebt Royal Commission parallel

The 2023 Royal Commission into the Robodebt Scheme was not the first time an Australian government had deployed an automated welfare-decision system that caused systemic harm to vulnerable Australians. It was, however, the most thoroughly documented. The Commission documented that the scheme caused profound harm across the affected population, and heard evidence connecting it to the deaths of affected people. The stakes of getting automated decision-making wrong for a vulnerable cohort are, on that record, not merely administrative or financial — they extend to human life. The Royal Commission’s findings are directly relevant here.

Robodebt failed because it applied an automatic income-averaging calculation to individual welfare cases where the individual’s actual circumstances differed from the national data average the algorithm used. The result was mathematically plausible at a population level and wrong in individual cases — systematically, and over years. The failures the Royal Commission identified were not primarily failures of legal authority (the automated scheme lacked it, but that is not the NDIS parallel). They were failures of safeguard architecture: affected individuals were not told what data drove their debt calculation; the onus of challenging the debt was placed on the individual who had been wronged; contemporaneous external review mechanisms were absent; internal oversight did not catch systematic errors because the oversight was designed around the assumption that the algorithm was correct; and the cohort bearing the harm — welfare recipients, including many people with disability — was structurally least positioned to navigate the contestation pathway.

Compare that structure to what Schedule 3 Part 2 creates. The CEO’s oversight obligation doest not affect the validity of an automated action if the CEO fails to comply. The notification obligation does not affect validity if the Agency fails to comply. Participants cannot see the input data driving their automated decision. The standard operating procedure is published but not disallowable. Annual reporting is retrospective and aggregated. The s.40A suspension and s.30(1A) revocation pathway places the burden on the participant — the person least likely to be able to engage at the relevant time — tore- iniate contact within ninety days or face revocation. And s.50A plan renewals, which alter participants’ funding through the operation of law without a reviewable decision, operate entirely outside the automation framework’s transparency and oversight obligations.

The Government’s position, as reflected in the Bill, appears to be that the explicit statutory authorisation and the CEO oversight framework in s.59E distinguish Schedule 3 Part 2 from Robodebt. That is a procedural distinction. It does not address the substantive question the Robodebt Royal Commission identified: whether an automatic decision system applied to a vulnerable cohort has the external-accountability and individual-transparency architecture necessary to ensure that systematic errors are caught at the point where they occur, ot years later. On the evidence of what Schedule 3 Part 2 actually provides, the answer is that it does not.

Submission 2079

The Committee should weigh that conclusion against the gravity the Robodebt Royal Commission established. That inquiry connected an inadequately-safeguarded automated welfare-decision system to the deaths of affected people. Applying a comparable architecture to the NDIS participant cohort — disproportionately Autistic, intellectually disabled, and psychosocially disabled Australians — carries the same foreseeable category of risk. The foreseeable worst case of getting automated decision-making wrong for this cohort is not an administrative inconvenience to be corrected in a later annual report; it is the loss of disabled lives. On the evidence the Robodebt Royal Commission placed on the public record, that risk should be determinative of whether Schedule 3 Part 2 proceeds in its current form.

The pattern is already operational: Karen Shine’s parliamentary-record testimony

The structural risk Schedule 3 Part 2 would scale is not a hypothetical projection. It is the lived experience of NDIS participants under the existing administrative arrangements — arangements that operate without the explicit statutory automation framework the Bill introduces, but with the same pattern of computer-driven decisions, delayed agency response, and operative-text removal of review rights that the framework would systematise. Ms Karen Shine, a current NDIS participant, gave first-person testimony to the Joint Standing Committee on the NDIS at its second public hearing in Sydney on 15 May 2026. Her testimony, on the public Hansard record, tracks onto three of the Schedule 3 Part 2 risk features this section has set out. It also bears on two threads running through this submission: it is further evidence of Schedule 3 operating against the participant- protective intent the Government locates in Schedule 2, and it sharpens the procedural- sequencing concern Sections 8 and 15 develop — that the Bill is being advanced ahead of the Joint Standing Committee’s integrity findings, before the evidence the Committee is hearing (Ms Shine’s included) can inform the legislation.

Categorical application without individualised assessment — the s.33 risk. Ms Shine’s opening statement to the Committee:

"I’m an NDIS participant who was subjected to a prolonged, procedurally unfair

scheme integrity audit that resulted in a $21,000 debt with just 14 days to pay. Eighteen months later, that debt was withdrawn with no restoration of my rights. … I was audited against rules that didn’t even exist at the time I made the claim. I was told I should have claimed using categories on the self-managed portal that don’t even exist. The NDIA assumed fraud and raised debts whenever two or more suppliers invoiced for the same period. That’s not evidence of fraud. That’s evidence of my choice and control.“

This is the pattern the Robodebt Royal Commission identified as systemically harmful when applied algorithmically — categorical rules applied retrospectively against individual circumstances the rules did not contemplate at the time of the original administrative action. Ms Shine’s case demonstrates that under current arrangements, even without

Submission 2079

explicit statutory authorisation for automation of s.33 evaluative determinations, the agency is already applying rules that did not exist at the relevant time and raising debts on category-based assumptions rather than individualised assessments of the participant’s actual circumstances. The Bill’s authorisation of automated s.33 reasonable-and- necessary determinations from commencement would extend the pattern that has already produced this result to the most consequential individual eligibility assessment in the scheme.

Ms Shine’s experience also speaks to the integrity-leakage figure the Government relies on (Section 8). Her debts were raised on the assumption that overlapping invoices were fraud, when — on her account — they reflected legitimate choice and control. Where category- based assumptions of that kind are counted into a “leakage” total, the figure overstates fraud by folding in conduct that is not fraud at all. Automating the s.33 determination would scale precisely that over-attribution.

Operative-text removal of review rights — the s.50A(4)(b) / s.45C pattern. Ms Shine’s evidence on the administrative-review pathway:

  "it took the NDIA years, while I was only ever given 14 days to respond or comply,
 including paying large debts. Documents were inaccessible despite requests for
 alternatives. The debt letter stated the decisions were not reviewable, meaning no
AAT review, no internal review, no oversight and no way to clear my name." 

Schedule 3 Part 2’s framework would systematise this pattern at the statutory level. Section 7 of this submission has set out how new s.50A(4)(b) places annual plan renewals outside the reviewable-decision pathway and how new s.45C places Ministerial pricing decisions outside it as well. Ms Shine’s experience demonstrates the lived-experience consequences of the same pattern operating now under current arrangements: the participant left with no AAT review, no internal review, no oversight, no way to clear her name. The implicit defence that the existing review architecture will catch automated errors is falsified by Ms Shine’s case — where the agency’s debt letter expressly said the declarations were not reviewable, removing the pathway the existing legislative architecture was meant to provide. The Bill would put this removal on the statutory record across the renewal and pricing pathways the participant is most likely to need review for.

Retrospective correction without restoration — the s.59E(7) mechanism. Ms Shine’s account of the audit’s resolution:

 "in 2023 my lawyer received an email saying, ‘The agency will not pursue the debt,’
but I received no apology, no explanation, no correction of the record, no
education, and no reinstatement of self-management. The debt vanished, but the
consequences continue to have significant emotional toll and restrict my choice
and control." 

The Bill’s primary backward-looking accountability mechanism is the annual reporting requirement at s.59E(7). Ms Shine’s case demonstrates what after-the-fact correction

Submission 2079

means in practice for the affected participant: the debt was withdrawn after eighteen months, but the structural consequences — for choice and control, for engagement with the scheme, for emotional wellbeing — were not addressed. The participant did not receive an apology, an explanation, a correction of the record, an education on what the agency now considered correct, or a reinstatement of the self-management arrangement the audit had removed. The Robodebt Royal Commission’s finding that retrospective reporting is not an adequate substitute for contemporaneous safeguards is directly evidenced by what happened to Ms Shine under existing arrangements. Schedule 3 Part 2’s annual-reporting mechanism is the same kind of after-the-fact correction Ms Shine has on the public record describing as inadequate.

The Joint Standing Committee on the NDIS now has Ms Shine’s evidence on its public record. Her case is the falsification of the implicit “this won’t happen at scale” defence Schedule 3 Part 2 relies on. Under existing arrangements — without the explicit statutory automation framework the Bill introduces — the pattern is already operating. Schedule 3 Part 2 would scale it, codify the review-rights removal at new s.50A(4)(b) and new s.45C, and substitute retrospective reporting at s.59E(7) for the contemporaneous safeguards the Robodebt Royal Commission identified as necessary.

The Autistic-cohort-specific risk

The NDIS participant cohort is not a random cross-section of the Australian population. According to the NDIS Scheme Actuary figures cited in public reporting, approximately 75 per cent of new NDIS entrants are Autistic children. The scheme’s existing participant base is heavily weighted toward Autistic Australians and Australians with intellectual disability — cohorts whose engagement with bureaucratic and administrative processes is structurally shaped by their disability in ways the design of automated systems must account for.

Autistic Australians process bureaucratic communication differently. Phone calls are a significant source of sensory and social distress for many Autistic people. Complex written notices, particularly those using the language conventions of administrative law, may be difficult to parse under conditions of anxiety or executive-function load. Autistic burnout — a recognised phenomenon involving a temporary but severe reduction in functional capacity — can last days, weeks, or months and directly affects the capacity to engage with administrative processes at precisely the point where those processes are most consequential. The s.40A non-contactability suspension mechanism is applied to participants who cannot be contacted; a significant proportion of those participants will be Autistic, and their non-contactability will reflect their disability rather than an absence of need. In many cases it will also reflect a failure of the scheme itself — a failure to offer the participant accessible, individualised methods of contact suited to their disability — rather than any disengagement on the participant’s part. A suspension mechanism triggered by non-contact therefore penalises the participant for an access failure that is, in part, the Agency’s to fix.

Submission 2079

The NDIS automation framework as designed does not account for these structural features of the cohort it applies to. The SOP framework requires that circumstances covered by automated evaluative determinations be “sufficiently objective in nature” to be ascertained by computer program (s.59D(3)). The assessment of what is “reasonable and necessary” under s.33 is, for Autistic participants in particular, highly contextual and individual — it is the antithesis of the “sufficiently objective” circumstances the SOP framework requires. The risk is that the automation of s.33 evaluative determinations will produce systematic under-assessment of support needs for participants whose presentation is most variable and contextual, while appearing to comply with the SOP framework’s objectivity requirement at a categorical level.

I am one of the participants this sub-section concerns. I am Autistic, late-diagnosed, and a current NDIS participant. My disclosure at Section 17 names the executive function, communication, and information-processing characteristics for which I have used accessibility tools in preparing this submission. Those characteristics are the structural features the Schedule 3 Part 2 safeguard architecture does not contemplate. If a s.33 reasonable-and-necessary assessment about my supports is automated and produces an under-assessment of my actual needs, the pathway by which I would identify the error and challenge it is the pathway my disability structurally compromises. The notification provision tells me a computer made the decision, but not what data it used about my case or what reasoning produced the outcome; the review pathway requires me to initiate engagement, parse the decision, and respond administratively under conditions in which my capacity to do those things is exactly the variable the assessment turns on. The Bill’s safeguard architecture relies on the participant being able to do — at the operative moment, in respect of the most consequential individual eligibility assessment in the scheme — the very thing my disability makes hardest.

What should happen instead

My position on Schedule 3 is set out at the opening of Part C: it should be omitted from the Bill. The use of computers in administration is not in itself wrong — but the intent this Schedule reveals is. Schedule 3 Part 2 is built to let computers make the final decisions about what disabled people receive, with human oversight reduced to a retrospective check; on the face of the provisions, the direction of travel is toward a decision system that needs fewer people to run it, not one that supports participants better. That is the wrong objective. Automation should assist a human decision-maker — logging, surfacing information, flagging anomalies — not replace human judgement at the point a person’s supports are decided; and a human element must remain available to help participants understand a decision, know their rights, and challenge it where needed. The scheme processes large volumes of administrative actions, and there are genuine efficiency and consistency arguments for computer-assisted decision-making where it is appropriate. But “computer-assisted” — in which a computer presents information to a human decision-maker who exercises the judgement — is structurally different from “automated” — in which the computer takes the action and the human oversight is retrospective. Schedule 3 Part 2 authorises the latter, including for evaluative determinations, including

Submission 2079

for the s.33 reasonable-and-necessary assessment, with a safeguard architecture that does not provide contemporaneous participant-facing accountability, does not provide input-data transparency, and explicitly preserves the validity of decisions taken in breach of the CEO’s oversight and notification obligations.

The Robodebt Royal Commission’s central finding — that internal-process safeguards are insufficient where the system applies algorithmic rules to a vulnerable cohort whose individual circumstances are the primary variable — has direct application here. Schedule 3 Part 2 should be omitted from the Bill, and if the Government wishes to bring automation provisions back, they should do so with: external-accountability safeguards that affect the validity of decisions made in breach; input-data transparency as a statutory right; a contemporaneous review mechanism that is initiated at the point of automated decision, only on participant request; and substantive community consultation, developed under “nothing about us without us” principles, with the Autistic, intellectual-disability, and psychosocial-disability communities whose members are most at risk from automated decision-error in a scheme of this kind.

Submission 2079

Section 14 — Economic evidence, cost-shift, and the missing impact analysis

Sections 3 through 11 set out the access, cohort, statutory and integration arguments. This section addresses what the economic evidence the Australian Government has put on the public record actually shows, what the Bill omits from its own impact assessment, and where the cost-shift the Bill triggers will land.

The Committee is being asked to recommend a package of reforms with a $37.8 billion gross National Disability Insurance Scheme reduction over four years against the Scheme Actuary’s pre-reform projections, and a twelve-year cumulative reduction of $184.9 billion to 2036-37 per Budget Paper No. 1 Statement 3. The economic case the Government has assembled for that package is partial. The cost-shift the package will produce across other public-service systems is not analysed. The fiscal asymmetry against the revenue-side measures in the same Budget is documented in the Government’s own Budget papers.

The cost-benefit analysis lineage - and the fourth analysis the Bill does not contain

The National Disability Insurance Scheme has been the subject of three Government-commissioned cost-benefit analyses across its lifetime. The Productivity Commission’s 2011 Disability Care and Support inquiry produced the founding economic case for the scheme. It found the benefits would exceed the costs at the whole-of-economy level — estimating that the scheme, once mature, would add around 1 per cent to GDP — and identified approximately $11 billion per year in offsetting avoided spending across the public-service systems the scheme was designed to reduce dependence on.

The Commission’s 2017 NDIS Costs commissioned study report reaffirmed that net-positive finding at the all-of-scheme level on a post-launch evaluation basis. On both of those Government-commissioned analyses, the scheme returned more in economic benefit than it cost. The Taylor Fry 2023 Independent Evaluation of the NDIS — the most recent Government-commissioned analysis — measured net costs of $16 billion exceeding measured benefits of $7.5 billion in 2022-23.

The Taylor Fry 2023 measurement requires precise handling. The methodology was deliberately constrained. The evaluation report explicitly excluded major benefit categories from its measured-benefits column: justice-system outcomes, child-protection outcomes, post-school education outcomes, broader community benefits, fiscal multiplier effects, value to the disability support sector, and economic utility to participants. Taylor Fry characterised the $7.5 billion measured-benefits figure as “most likely an underestimate” on the face of the report.

At the cohort-band level, even within the constrained methodology, Taylor Fry found net benefits exceeded net costs for Autistic participants in the lowest measured cost bands and for children-and-families. The cohort the Thriving Kids measure targets - children aged eight and under with “low to moderate support needs” in the Budget Paper No. 2 measure text framing - is the cohort Taylor Fry’s analysis identifies as showing the strongest measured net benefits. The Government’s

Submission 2079

targeting logic for Thriving Kids fails on the Government’s own most recent commissioned analysis. (The “low to moderate support needs” clinical-language framing the Budget paper uses is itself addressed at Section 4 - neither term appears in the DSM-5, the Independent NDS Review, or the National Autism Strategy 2025-31, and no Australian medical college or peak clinical body has publicly endorsed the framing.) The headline net-cost finding is therefore a partial measurement against an explicit list of excluded categories, not a final assessment of the scheme’s economic value. The 2011 and 2017 Productivity Commission findings on the broader benefit categories remain on the public record and have not been displaced by subsequent analysis. There is no published Government-commissioned cost-benefit analysis of the present reform package, weighing the $37.8 billion reduction against the cost-shift the reduction will produce across other public-service systems. The Bill proceeds without a fourth analysis.

The independent (non-Government-commissioned) economic literature points the same way as the two Productivity Commission findings. Per Capita’s 2021 False Economy analysis estimated a return of approximately $2.25 in economic activity for every $1 of NDIS expenditure, and that every $1 billion of NDIS underfunding costs around 10,200 jobs and $2.25 billion in economic activity. The e61 Institute’s 2026 work on the scheme’s market design (Pelin Akyol, More Quasi than Market) locates the scheme’s inefficiency in price-cap and market-design failure, not in the existence of the supports. I have not been able to locate any credentialled, independent economic analysis that disputes the whole-of-scheme net-positive finding the Productivity Commission reached; the only contrary headline figure is the Government-commissioned Taylor Fry 2023 measurement, which is, on its own face, a constrained partial measurement that excludes major benefit categories. The Committee should require the Department to produce, for the public record and before the Bill proceeds, a full lifecycle cost-benefit analysis of the reform package that accounts for the cost-shift across other public-service systems (Section 16).

What the Bill says about its own economic impact

The Bill’s Explanatory Memorandum at lines 201 to 211 acknowledges this absence by characterisation. Two documents are attached as separate documents and described as “Impact analysis equivalent” - not as a Regulation Impact Statement. The phrase “Impact analysis equivalent” is the Government’s own description. A Regulation Impact Statement is the document the Government’s Australian Government Guide to Policy Impact Analysis requires for regulatory proposals with substantial economic, social or environmental impact. The choice to describe the attached documents as “equivalent” rather than as RIS documents acknowledges that the standard regulatory-impact discipline has not been applied to the Bill. The Office of Impact Analysis report that the Guardian Australia surfaced on 20 May 2026 — recording that the Department of Health, Disability and Ageing chose the 50 per cent reduction in Social, Civic and Community Participation funding over more cohort-neutral alternatives, with the Department’s own warning that significant cuts to participant supports could result in “a regression in daily living skills”, elevated risk of “injury, neglect and social isolation” and would “undermine the objectives of the NDIS, including the aim to provide reasonable and necessary supports to participants”

Submission 2079

calculates the cohort-asymmetric impact and proceeded with the cut anyway. The Department’s own report is the closest document to a substantive impact analysis. It is not what the Explanatory Memorandum attaches as the Bill’s economic justification. The Committee should require a full Regulation Impact Statement, prepared to the standard of the Australian Government Guide to Policy Impact Analysis, rather than accept an “equivalent” the Government’s own label concedes falls short. There is an irony the Committee is entitled to note: the Bill leans heavily on economic and actuarial framing to justify narrowing what is at its core a clinical and functional eligibility question, yet sets aside the part of the economic evidence economists are actually qualified to provide - the whole-of-scheme net-benefit finding, and the cost-shift the package will trigger.

Cost-shift across the public-service systems the scheme was designed to reduce dependence on

The 2011 Productivity Commission economic case for the National Disability Insurance Scheme included approximately $11 billion per year in offsetting avoided spending across other Government-funded systems. The avoided-spending categories - emergency department waits, public mental health system load, schools-disability support requirements, Centrelink and JobSeeker case-loads, justice-system contact, family carers exiting paid employment, aged-care and Medicare contact - are the systems the scheme was designed to reduce demand on by providing supports directly. When access to the scheme is narrowed, the demand returns to those systems. The Bill counts the National Disability Insurance Scheme expenditure reduction without counting the corresponding expenditure increases the narrowing will trigger elsewhere in the public service.

The ledger also has a positive side the Bill’s accounting ignores. NDIS expenditure is not money that leaves the economy; it largely pays the wages of the disability support workforce - approximately 325,000 workers - who pay tax, draw less on income support, and spend their incomes in their communities. Well-supported participants are themselves more likely to be in paid work, for longer hours, and in better physical and mental health, reducing demand on income support and the health system. Per Capita’s modelling captures part of this circular effect in its $2.25-per-$1 return and its estimate of 10,200 jobs per $1 billion. A $37.8 billion reduction over four years is roughly $9.5 billion a year; on that estimate, the implied job losses across the disability support workforce run to the tens of thousands - each carrying its own cost to revenue, in foregone income tax and increased income-support payments. A reduction in scheme expenditure is therefore not a clean saving of the headline figure: it withdraws wages from a large workforce, returns demand to the systems the scheme relieves, and forgoes the tax and participation gains the supports produce. None of this is counted in the Bill’s impact material.

The Productivity Commission’s 2024 Working Paper on Foundational Supports - the Government’s own statutory body’s most recent assessment - warned of a $1.09 billion per year cost-shift to the States as a structural risk of the Foundational Supports envelope and the broader access-tightening direction. Schedule 1 of the present Bill, in combination

Submission 2079

with the Foundational Supports Agreement signed by all eight jurisdictions in early 2026, codifies the direction the Commission’s 2024 paper warned against. The cost-shift to the States is a structural consequence the Government’s own analytical body identified prospectively. The Bill does not engage the warning.

The cost-shift is concrete in two Budget 2026 measures that operate together. The National Disability Insurance Scheme reform package reduces $37.8 billion in scheme spending over four years gross of new operational spending. The Education Portfolio measure in the same Budget cuts $472 million over four years from disability loadings to schools — the funding that pays for teachers’ aides and specialist classroom support for children with disabilities in mainstream education. The same Budget that narrows access to the National Disability Insurance Scheme cuts an adjacent foundational support system. The cohort the scheme is being narrowed for is the same cohort whose schools-system support is being reduced. The cost-shift to the schools system the scheme was designed to relieve of disability-support burden is being made simultaneously with the scheme’s own access narrowing. The Department of Health’s own warning on regression in daily living skills, elevated injury and neglect risk, and undermining the scheme’s objectives applies in substance to the schools-system cohort whose mainstream-classroom supports are also being cut.

The cohort-concentration mismatch

The Grattan Institute’s Better, safer, more sustainable: How to reform NDIS housing and support report records that “almost 40 per cent of the total costs of the NDIS are spent on only 7 per cent of its users who have intensive support packages. The costs of supporting these 43,500 people is at least $15 billion per year.” The fiscal concentration in the scheme sit in the intensive-support cohort that the Bill’s eligibility-tightening provisions barely engage. Schedule 1’s new permanence-and-treatment test, the new functional capacity assessment regime, the social-and-community-participation funding cut and the alternative-supports gate operate against the broader 93 per cent of the participant base — The participants whose plans are in the standard, non-intensive support categories. The Bill narrows access for the cohort that is not the source of the fiscal concentration the integrity-rhetoric attaches to. Sections 9 and 10 of this submission address the structural- policy implications of that mismatch. The economic implication addressed here is that the Bill’s reduction is being taken from the wrong fiscal concentration.

The fiscal asymmetry against the revenue-side measures in the same Budget

Budget Paper No. 1 Statement 3 Box 3.2 records the Treasurer’s own characterisation of the 2026 Budget repair task: “In this Budget, savings from spending measures are almost three times savings from revenue measures.” The 3:1 spending-to-revenue ratio is on the public record in the Government’s own Budget paper. The asymmetry has cohort consequences. The National Disability Insurance Scheme reform package delivers $37.8 billion of the spending-side savings over four years, and $184.9 billion over the twelve-year medium-term projection to 2036-37 (BP1 Statement 3 Box 3.2). The capital gains tax and

Submission 2079

gegative gearing reform package in the same Budget delivers approximately $3.6 billion of the revenue-side measures over four years against a Government-acknowledged total of approximately $13.5 billion per year currently foregone through the existing capital gains tax discount and negative gearing arrangements. The ratio between the NDIS spending- side reduction over four years ($37.8 billion) and the CGT and negative gearing revenue- side measures over four years ($3.6 billion) is approximately ten to one. The comparison with the revenue the Government continues to forgo is sharper still: at approximately $13.5 billion per year, the existing capital gains tax discount and negative gearing concessions forgo on the order of $54 billion across the same four years — more than the entire $37.8 billion four-year NDIS reduction. The Government is cutting disability support by less than it forges, every four years, through two tax concessions it has left substantially intact.

I support the capital gains tax and negative gearing reforms as good direction. The Division 296 superannuation tax reform of 2025 was the same direction. Good direction is not the same as enough in scope. The structural critique I put to the Committee is on the scale- asymmetry between the spending-side measure and the revenue-side measure, not on the substance of either reform.

The Petroleum Resource Rent Tax forecast adds a further asymmetry data point. The Australia Institute analysis of the 2026 Budget papers records that “PRRT revenue will fall from $1.9 billion next year to $1.3 billion by the end of the decade. The gas industry keeps telling us to wait, insisting PRRT revenue will go up. But the budget papers show it falling by a third by 2029-30.” A 25 per cent gas export tax — the reform Senator David Pocock’s Senate Select Committee on the Taxation of Gas Resources has been investigating since late 2025, and which the Australian Greens also pursue — would have raised, on the Australia Institute’s analysis, more than $17 billion per year. The four-year NDIS reduction at $37.8 billion is approximately two years and three months of foregone gas-tax revenue at the scale Senator Pocock’s Committee is investigating. The structural choice the 2026 Budget makes is to take spending-side savings from the disability-supports cohort at three times the rate it takes revenue-side savings from any source. The contrast in difficulty is itself telling: a gas-tax reform of this kind could be legislated and take effect quickly, with a positive fiscal and consumer outcome; the NDIS reduction is the harder path, carries a large hidden cost-shift that makes its headline savings almost certainly overstated, and falls hardest on disabled Australians — and disproportionately on Aboriginal and Torres Strait Islander disabled people in the thin-market communities Sections 10 and 11 address.

The “better and cheaper” logical incompatibility

The Government’s framing of the reform package is that the savings on the National Disability Insurance Scheme side will be partially redirected into Foundational Supports outside the scheme — Thriving Kids and the broader Foundational Supports envelope. The fiscal arithmetic does not support the framing. Thriving Kids: $4 billion total over five years ($2 billion Commonwealth plus $2 billion states-matched). Foundational Supports outside the National Disability Insurance Scheme: $3 billion Commonwealth over five years, matched-contingent on the States. The Inclusive Communities Fund: $200 million over

Submission 2079

tree years, held in Contingency Reserve pending further design and consultation. The total nominal envelope of $10 billion over five years (Commonwealth plus states-matched, state matching not locked) is against a $37.8 billion four-year reduction in the scheme. The Inclusive Communities Fund specifically - the group-based replacement-services measure - provides $200 million over three years, about $67 million a year, against the $37.8 billion reduction over four years, about $9.5 billion a year. Compared like-for-like, that is a ratio of roughly one to one hundred and forty - under a cent of group-based replacement funding for every dollar removed from individual plans. No per-person funding comparison has been published for any of the replacement streams.

This is where the Minister’s stated objectives collide. One of the four objectives named for the package is delivering better-quality services and outcomes for participants. A $37.8 billion four-year reduction - set against a replacement envelope of at most $10 billion over five years (state matching not locked), and a group-services fund running at under a cent for every dollar removed from individual plans - cannot deliver both better outcomes and savings at this scale. The Committee should not accept “better and cheaper” as a single proposition where the Government’s own arithmetic makes the two mutually exclusive.

The correct cost-reduction sequence - ecosystem first, access restriction second

The Government’s fiscal case for Schedule 1 rests on the proposition that narrowing scheme access produces net savings. The economic argument is more complicated than the savings-column calculus, because the mechanism the Government’s own policy logic relies upon - and the mechanism the Minister’s National Press Club address implicitly assumed - is the availability of alternative supports that participants can transition to. That mechanism requires the foundational supports ecosystem to exist before the access restriction is applied. The Bill reverses the sequencing.

The existing NDIS Act 2013 section 34 already provides the correct cost-reduction mechanism. Section 34 requires the CEO, in determining reasonable-and-necessary supports, to be satisfied that a support is most appropriately funded through the NDIS rather than through other available systems, having regard to what is available through family, carer, informal, mainstream and community supports. When the foundational supports ecosystem is built - accessible diagnosis, health navigators, school-based disability support, community-based allied health and mental health services - the existing s.34 test naturally directs participants toward those alternatives as they become available. Participants who currently rely on NDIS funding for supports that a functioning school system, accessible community service, or health navigator could provide will naturally draw on those services. The NDIS plan is reviewed accordingly. Cost reduction follows from ecosystem development without a legislative access gate to produce it.

The Disney et al. 2025 data maps this argument onto the existing participant population. Lower socioeconomic areas have measurably lower NDIS eligibility rates - not because

Submission 2079

disability prevalence is lower in those areas, but because the barriers to accessing diagnosis, completing the FCA, and navigating the administrative process are higher. If the Government’s $10 billion foundational supports envelope were deployed to build the diagnostic and navigation infrastructure in those areas first - accessible diagnosis at Step 1, health navigators and FCA support at Step 2 - eligible people currently excluded by structural barriers would enter the scheme. In the short term, scheme participation would increase. Over time, those participants - with adequate community and foundational supports available - would draw more from the ecosystem and less from the NDIS, and the existing s.34 mechanism would produce the natural cost-reduction the Government is seeking. The Disney et al. data does not describe scheme over-expansion. It describes the structural under-subscription of the scheme in under-resourced communities. The correct policy response to that data is ecosystem development and reduced access barriers - the opposite of the Bill’s access-tightening direction.

The fiscal asymmetry documented earlier in this section means the ecosystem-development investment the correct sequence requires is readily available from the revenue side: the $10 billion foundational supports envelope over five years - the envelope the Government relies on to provide the alternative supports the access-tightening presumes exist - is around six months of the revenue a 25 per cent gas export tax would raise on the Australia Institute’s projections. The fiscal choice not to fund the ecosystem from the revenue side, while restricting access before the ecosystem exists, has clear distributive consequences: they land on disabled Australians and their families, while the counterfactual benefits remain with the gas export sector.

The cost-impact on affected families — the dimension the Bill does not measure

The Bill’s “Impact analysis equivalent” documents do not include a cost-impact assessment on the families that will be affected when National Disability Insurance Scheme plans are narrowed or withdrawn. I write as the parent of two Autistic children and as a late-diagnosed Autistic adult myself. My daughter is fourteen years old, an NDIS participant, and recently transitioned from mainstream schooling to online learning. The NDIS-funded supports are what makes her continued education possible. When NDIS plans are reduced or removed for participants in our circumstances, the costs do not disappear; they shift. A parent leaves paid employment to provide the care the supports were funding. The household income falls. The participant’s access to education, to community participation, to allied health, and to specialist services contracts. Downstream Government-funded systems pick up the consequence - public mental health system contact, emergency department waits, schools-system load, justice-system involvement for adolescents whose support has been withdrawn. The downstream costs are real, and they are not measured by an Impact analysis equivalent document that records only the scheme-side expenditure reduction.

This is the missing dimension of the Bill’s economic justification. The Productivity Commission’s 2011 founding economic case included the offsetting avoided spending

Submission 2079

cross other systems as part of the scheme’s economic value. The Government, in narowing access fifteen years later, counts the scheme-side reduction without counting the corresponding cost increases the narrowing will produce. The fourth cost-benefit analysis the Bill does not contain is the analysis that would weight the two columns of the ledger that the Government’s own 2011 economic case for the scheme treated as inseparable.

Submission 2079

Section 15 — Procedural concerns and peak-body convergence

Sections 3 through 14 set out the substantive case on the merits: the clinical-evidence vacuum, the diagnostic-language critique, the cohort assessment-instrument gap, the structural under-subscription, the statutory and treaty inconsistency, the contradictions with Government-commissioned reviews, the economic-evidence limitations, and the cost-shift the Bill will produce across adjacent public-service systems.

This section addresses four procedural concerns that are independent of those substantive arguments but that the Committee should weigh when assessing the basis on which the reform package has been brought to Parliament and the adequacy of the process through which the Committee’s own scrutiny is being conducted.

The procedural-sequencing critique — the Legislation Committee reports sixteen days before the inquiry whose evidence base the Bill rests on

The Joint Standing Committee on the NDIS Integrity of the National Disability Insurance Scheme inquiry was referred to the Committee by the Minister on 25 March 2026. The JSC’s first public hearing was held on Friday 1 May 2026 in Canberra. The Bill was introduced to Parliament on Thursday 14 May 2026 — thirteen days after the JSC’s first hearing. The Senate Community Affairs Legislation Committee was referred the Bill on 14 May 2026, with this submission due by 29 May 2026. The Legislation Committee will report on Tuesday 16 June 2026. The JSC NDIS Integrity inquiry will report on Thursday 2 July 2026 — sixteen days after the Legislation Committee has reported on the Bill.

The Minister’s Second Reading Speech of 14 May 2026 characterises the Bill as taking up the recommendations of the independent reviews the Government has commissioned, including the integrity-investigation inquiry the JSC is conducting. The procedural-sequence consequence is this: the inquiry whose evidence base is said to justify the integrity provisions of the Bill will report on its findings sixteen days after the Senate has already been asked to form a view on those provisions. Either the JSC’s integrity-inquiry findings are not relevant to the Bill—which contradicts the integrity rationale the Minister placed on the Hansard record—or the legislative timetable has been structured so that the Senate completes its committee consideration before the Government’s own evidence base is reported. There is no third reading that survives scrutiny. If the Government’s integrity rationale were genuine, it would not have introduced the Bill ahead of the evidence base its own inquiry is still assembling; the sequencing is itself evidence that integrity is the framing of the Bill, not its driver.

This critique applies specifically to Schedule 2—the schedule whose subject matter (provider integrity, payment integrity, and provider registration) is the subject matter of the JSC inquiry. Part B of this submission sets out the case for deferring Schedule 2 until after the JSC reports. The procedural-sequencing argument is not merely a process objection; it is the structural evidence that the Bill’s integrity rationale would not survive the JSC’s own evidence base being before the Committee before the Legislation Committee reported.

Submission 2079

This procedural-sequencing concern is distinct from the consultation-adequacy concern addressed below: the first goes to the structure of the parliamentary review itself (the Legislation Committee reporting before the JSC), the second to the time given to disabled Australians to engage with the Bill. Both are live, and they operate independently.

The proportionate-scrutiny argument compounds both. Michael Brennan, the CEO of the e61 Institute — a former Chair of the Australian Productivity Commission and former Deputy Secretary of the Australian Treasury — published an independent analysis of the Bill’s Budget context on 14 May 2026. Brennan’s assessment, based on the Budget papers: “in fiscal terms, one item in this Budget looms larger than any other. It’s the NDIS save. That single policy measure fundamentally underpins the fiscal logic of the 2026-27 Budget in a way that no other issue does.” Brennan quantified the comparison: the NDIS save at $37.8 billion over four years is approximately eight times the trust-taxation measure and ten times the combined capital gains tax and negative gearing reform. His historical assessment: “Never before has a Budget been so dominated by a single program saving. And perhaps never before has the entire fiscal strategy rested so heavily on the successful implementation of one Budget measure.” A former Productivity Commission Chair on the public record that the Bill’s Budget measure has no historical precedent in scale is making the proportionate-scrutiny argument in terms a Committee considering whether its reporting timeline is adequate to the task is entitled to weigh. The combined effect of the DRO Joint Statement on community consultation inadequacy and Brennan’s credentialed-economist characterisation of the measure’s scale is that both the community-engagement record and the fiscal-magnitude record support a more extended and substantive scrutiny process than the current Legislation Committee timetable provides for. Read together with the provenance of the package’s framing set out at Section 8 — taxpayer-funded message-testing that advised leading with fraud and rorts to build tolerance for access restriction — Brennan’s finding points to the same conclusion: the measure that “looms larger than any other” in fiscal terms is a budget save, and the integrity framing is the vehicle for it rather than its driver.

The Government’s own Thriving Kids Inquiry — a five-month gap with no implementation

A separate procedural-democracy concern applies to Schedule 1. The Minister referred the Inquiry into the Thriving Kids initiative to the House Standing Committee on Health, Aged Care and Disability on 2 September 2025. The Committee received 404 written submissions, held eight public hearings across Canberra, Melbourne and Surry Hills NSW, and reported on 22 December 2025 with sixteen recommendations. The Bill before this Committee was introduced on 14 May 2026 — five months after the Inquiry reported, five months after the Government had on its own December 2025 record a substantive recommendation set produced from a deliberative process. Section 8 of this submission sets out in detail that the Bill structurally does not implement five of the most material recommendations the Inquiry made: Recommendation 1 (cohort composition of the Thriving Kids Advisory Group); Recommendation 3 (broader disability-sector engagement); Recommendation 4 (Inspector-General of the NDIS); Recommendation 9 (non-precluding

Submission 2079

either a pathway from Thriving Kids back into NDIS); and Recommendation 10 (24-month review of Thriving Kids implementation).

Five months is sufficient time to integrate substantive recommendations from the Government’s own Inquiry into the design of a Bill on the program the Inquiry investigated. The five-month gap was not a constraint on integration. The Department’s signed Ministerial response MC26-008601 of 25 May 2026 names the Inquiry as one of the consultation venues that “helped informed the national model for Thriving Kids.” The Inquiry’s place in the consultation record is the Department’s own claim. The non- implementation of the Inquiry’s five most material recommendations, five months after the report and against a venue the Department itself identifies as part of its consultation foundation, is the procedural-democracy concern. The Government cannot cite the Inquiry as part of the consultation record on which the Bill rests while simultaneously not implementing the recommendations that consultation produced. The Committee is entitled to require the Department to identify the basis on which the Government has selectively engaged with the Inquiry’s record - citing it as a consultation venue while not implementing the recommendations it made.

The Government’s own published consultation standard

A further procedural-democracy proof point applies to the consultation window for this submission itself. The Australian Government Guide to Policy Impact Analysis - the whole-of-Government framework maintained by the Office of Impact Analysis within the Department of the Prime Minister and Cabinet - sets the Government’s published standard: “consultation should occur for a minimum of 30 days where possible.” The Bill was introduced to Parliament on Thursday 14 May 2026. The Senate Community Affairs Legislation Committee’s original closing date for submissions on the Bill was Friday 29 May 2026 — a 14-day window measured from the day after introduction. The 14-day window did not meet the Government’s own published 30-day minimum. On 26 May 2026, in response to advocacy by People with Disability Australia and other Disability Representative Organisations, the Committee extended the closing date to close of business Monday 1 June 2026 — an 18-day window. The peak-body-secured extension still falls 12 days short of the 30-day minimum. The Disability Representative Organisations whose joint statement is set out in the next sub-section are applying the Commonwealth’s own published standard to the consultation-inadequacy critique, not a standard external to it. Section 14 separately addresses the parallel choice the Government has made to characterise the Bill’s economic impact analysis as “impact analysis equivalent” under the same Guide rather than as a Regulation Impact Statement; the two sub-sections together establish that the Commonwealth’s own framework for both the consultation period and the impact- analysis discipline has been applied selectively to this Bill.

The scale of the measure makes the truncated window harder to justify, not easier. On the assessment of a former Productivity Commission Chair set out above, this is the largest single program saving in the history of the Commonwealth Budget; a measure of that magnitude and reach warrants a consultation period beyond the 30-day minimum, not one

Submission 2079

that falls short of it. The Committee is entitled to ask why the Government has compressed scrutiny of its largest-ever budget measure below its own published standard.

The peak-body consensus on consultation inadequacy

On 14 May 2026 — the same day the Bill was introduced — twelve Disability Representative Organisations issued a joint statement titled “Call for Genuine Senate Scrutiny as Concerns Grow Over Rushed NDIS Reforms.” The twelve signatories were:

• People with Disability Australia (PWDA)
• the Australian Federation of Disability Organisations (AFDO)
• the Australian Autism Alliance
• Children and Young People with Disability Australia (CYDA)
• the Disability Advocacy Network of Australia (DANA)
• Inclusion Australia
• Mental Health Australia
• the First Peoples Disability Network Australia
• the National Ethnic Disability Alliance (NEDA)
• Physical Disability Australia (PDA)
• Women with Disabilities Australia (WWDA)
• Down Syndrome Australia.

The joint statement characterised the proposed Senate inquiry timeframe as “extraordinarily short… approximately one month” that “does not allow time for written submissions, public hearings, and meaningful engagement with the disability community.” The twelve organisations jointly called for “genuine Senate scrutiny” with a “robust” timeframe. The signatories are the Commonwealth’s own Disability Representative Organisations — the organisations funded through the operational-integrity envelope in Budget Paper No. 2 (line 8963, $15.9 million for Disability Representative Organisations) specifically to provide disability-community voices in the policy process. The organisations the Government funds to provide disability-community voices in the policy process have jointly and publicly described the consultation process for the Bill as inadequate.

The multi-party consensus on consultation inadequacy extends beyond the disability peak bodies.

Senator Jordon Steele-John (Greens, WA), the Greens’ disability policy spokesperson and a substitute member of the Senate Community Affairs Legislation Committee for NDIS-specific bills, publicly described the reform package as “extraordinarily alarming” and “cruel”. Ms Melissa McIntosh MP (Coalition, Lindsay), shadow NDIS minister, stated on the public record that the Government had “once again neglected to consult” with the Australians directly affected by the legislation, and that “people with a disability are feeling ambushed by the Government.” Two parties in the Parliament — the Greens and the Coalition — are on the public record describing the consultation process as inadequate. That is a cross-partisan consensus on the procedural failing, not a single-party critique.

The First Peoples Disability Network Australia’s inclusion among the twelve DRO signatories is the primary data point on the Aboriginal and Torres Strait Islander representation gap in the consultation record. Aboriginal and Torres Strait Islander people are over-represented in low-income and remote populations — the populations that the Bill’s new s.25A(2) most directly disadvantages by treating treatment access as “appropriate” regardless of financial or geographic circumstances. The DRO Joint Statement does not specifically address Aboriginal and Torres Strait Islander community

Submission 2079

consultation; the Committee is entitled to ask the Department whether the consultation process for the Bill included specific targeted engagement with First Nations disability communities, and whether the Department has assessed the impact of the s.25A(2) income-and-geographic regressivity provisions against the National Agreement on Closing the Gap Priority Reform 3 (building First Nations-led services and governance).

The “nothing about us without us” principle — embedded in the National Autism Strategy 2025-31, in the Disability Royal Commission’s Final Report, and in the social-model disability framework NDIS Act 2013 s.3(a) commits the scheme to give effect to — provides the standing basis for the consultation adequacy critique. Legislation affecting disabled Australians’ rights, access, and supports should be developed with substantive co-design engagement with disabled Australians and their representative organisations, not a one- month consultation window that the representative organisations themselves have jointly described as inadequate.

The responsible Minister’s engagement posture

The framing the Committee applies to the Government’s engagement with disabled Australians on the reform package is relevant to its assessment of the consultation-adquacy argument. The Office of the Hon Mark Butler MP — the Minister for Health and Ageing and Minister for Disability and the National Disability Insurance Scheme — initially operated an auto-acknowledgement template for NDIS correspondence from the community, redirecting participants and their families to the National Disability Insurance Agency and the NDIS Quality and Safeguards Commission, neither of which has any role in the design of the reform package. On 25 May 2026 the Department of Health, Disability and Ageing issued a signed substantive response to my 19 April 2026 correspondence reference MC26-008601, signed by an Assistant Secretary of the Thriving Kids Taskforce). The response is itself informative about the engagement posture. It was routed not to the Minister, and not to the NDIS-reform area, but to a topic-specific sub-Departmental taskforce. It acknowledged that “mild” and “severe” autism terminology “can misrepresent” the difficulties and capabilities of Autistic people, and then, in the following paragraph, operationalised the equivalent “low to moderate support needs” descriptor as the Thriving Kids cohort definition. And it engaged the cohort framing while not engaging the Bill’s Schedule 1, 2 or 3 operative provisions at all. The pattern is not silence; it is sub-ministerial routing that acknowledges the cohort while leaving the legislation’s substance unanswered.

I am a constituent of the electorate of Hasluck, in Western Australia. I submitted correspondence to the Minister’s office on 19 April 2026 through the aph.gov.au contact form. The response, five weeks later, was the Thriving Kids Taskforce letter described above — substantive in form, but routed away from the NDIS-reform area and silent on the Bill’s provisions. I engaged with the office of my local Member, the Hon Tania Lawrence MP (Labor, Hasluck), on 29 April 2026. Her Chief of Staff confirmed that the queries from my briefing were dispatched to the Minister’s office, with a response expected on an approximately six-month timeline. I note this not as a personal grievance but as a data

Submission 2079

point on what the engagement pathway between affected community members and the responsible Minister looks like in practice, during the period when the Bill was being finalised and introduced.

A documented systemic-failure incident

I include this element of Section 15 with care, and I am guided by the framing adopted by the Australian Human Rights Commission and disability peak bodies when they addressed it publicly. On 1 February 2026, police were called to a residential address at Mott Close, Mosman Park, Perth — in the federal electorate of Curtin, approximately eleven kilometres from my home in South Guildford. A person who provided care services to the family made the call. The family — parents Maiwenna Goasdowe (49) and Jarrod Clune (50) and their two teenage sons, Leon (16) and Otis (14) — was found deceased. The two boys were Autistic; their mother had publicly described them as being at the higher end of the autism spectrum, and they were the NDIS participants in the family. The incident is under coronial process as a suspected murder-suicide. I do not make predictive claims about the relationship between NDIS support decisions and the incident; that is a matter for the Coroner.

What is on the public record is the following. Multiple friends and support workers of the family stated publicly that the parents felt “failed” by the National Disability Insurance Scheme, that the boys’ NDIS funding had been cut, and that the couple were caring for their sons full-time with limited support and were at breaking point. These statements are sourced in multiple media outlets and in the Disability Advocacy Resource Unit’s summary.

The Australian Human Rights Commission’s response to the incident — from Commissioner Rosemary Kayess, the Disability Discrimination Commissioner — was: “We are seeing reforms that reduce supports or make them harder to access, leaving families anxious and frightened.” The Commissioner’s statement frames the incident not as an isolated event but as part of a pattern of reform-related support withdrawal. Multiple disability commissioners and peak bodies called for an inquest and for a broader review of family support arrangements for families caring for children with high support needs.

The Committee is entitled to consider that documented statement by the Disability Discrimination Commissioner of the Australian Human Rights Commission when assessing the human-consequences dimension of the Bill’s provisions. The Bill’s new s.34(1G) to s.34(1J) provisions increase the weight placed on informal supports — including family carers — in the reasonable-and-necessary assessment. Commissioner Kayess’ concern — that families with high-support-needs children are already anxious and frightened under the current scheme — is directly relevant to any legislative provision that increases the weight placed on those families’ caring capacity in the supports assessment. I reference the incident because the Australian Human Rights Commission has placed it in the public domain as evidence of a structural-harm pattern, and because the Bill’s provisions move the structural settings in the direction Commissioner Kayess identified as creating that harm. I put the consequence plainly, because the Commissioner’s framing

Submission 2079

invites no less: where supports that keep people safe are reduced or made harder to reach, the foreseeable consequences for some families are not measured in dollars. The Committee is being asked to weigh a Bill that moves the structural settings further in the direction a documented pattern of harm — including loss of life — already runs.

Submission 2079

Section 16 — Recommendations to the Committee

The substantive case for each of these recommendations is in the numbered sections cross-referenced. The recommendations are grouped by Schedule, consistent with the submission’s four-part structure, and then by the cross-cutting Part D asks. They are numbered sequentially for ease of reference.

Part A — Recommendations on Schedule 1

Recommendation 1.

The Committee recommend that Schedule 1 be omitted from the Bill in its entirety. The clinical-evidence base for the permanence-and-treatment test in new s.24(5) and s.25(1B) does not exist; the assessment instruments for the new functional capacity regime have not been designed (the Technical Advisory Group provisioned in Budget Paper No. 2 line 8919 is beginning that work in 2026-27); the replacement supports programs for the 9-to-15 cohort, the 16-to-17 cohort, and adults have not been announced; and the Schedule’s eligibility-tightening provisions are incompatible, on four distinct statutory and treaty grounds, with the National Disability Insurance Scheme Act 2013, the Disability Discrimination Act 1992, Australia’s obligations under the United Nations Convention on the Rights of Persons with Disabilities, and the National Agreement on Foundational Supports 2026-31 commitment to align with Closing the Gap. Schedule 1 also operates directly against the integrity improvements Schedule 2 is designed to deliver: its funding-below-cost authorities and its s.25A(2) regressivity thin the provider markets Schedule 2 must repair (Sections 10 and 11). Omitting Schedule 1 is therefore both right on its own evidence base and a precondition for Schedule 2 achieving its purpose.

(Sections 3, 4, 5, 6, 7, 8, 10, 11.)

Recommendation 2.

If Recommendation 1 is not adopted, the Committee recommend that the commencement of Schedule 1’s functional capacity assessment regime — currently set at 1 January 2028 per Budget Paper No. 2 line 8929 — be made conditional on the prior satisfaction of each of the following:

• The Technical Advisory Group has published its advice on the assessment instruments;
• The assessment instruments have been independently validated against the impairment cohorts they will be applied to;
• Bilateral implementation agreements between the Commonwealth and all eight States and Territories have been signed and published for the replacement supports programs that participants who do not clear the new access gate will be directed toward; and those replacement programs are operationally available in the geographic areas of the affected participants.

(Sections 5, 8.)

Recommendation 3.

The Committee recommend that new s.25A(2) — the provision that treatment is “appropriate” for the purposes of the permanence-and-treatment test regardless of the person’s financial circumstances or geographic location — be omitted from the Bill. The provision codifies in statute an income-and-geographic regressivity that the Disability Royal Commission (Volume 6, Recommendation 6.34) and peer-reviewed evidence (Disney et al., Medical Journal of Australia, 2025) identify as a systemic inequity in the current scheme. It is incompatible with the Disability Discrimination Act 1992 s.6

Submission 2079

prohibition on indirect discrimination, the CRPD Article 5 obligation to prohibit disability discrimination, and the National Agreement on Closing the Gap commitments on which the National Agreement on Foundational Supports 2026-31 is conditioned. (Sections 3, 7, 8.)

Recommendation 4. The Committee recommend that the wholesale removal of the s.31 Principles relating to plans at Schedule 1 Part 6 be reversed, and that the Principles be restored in full to the National Disability Insurance Scheme Act 2013. The s.31 Principles are the statutory framework for supported decision-making, participant agency, and community inclusion in the plan-making process. Their removal is inconsistent with CRPD Articles 12 and 19 and has been independently confirmed by Villamanta Disability Rights Legal Service’s community legal forum analysis of 20 May 2026 as a structural change to the plan architecture, not a technical amendment. (Section 7.)

Recommendation 5. The Committee recommend that new s.9B’s definition of functional capacity — which excludes environmental and personal circumstances from the assessment — be omitted from the Bill or substantively amended before any functional capacity assessment regime commences. The definition as drafted is incompatible with the CRPD Article 1 social model of disability, which NDIS Act 2013 s.3(a) commits the scheme to give effect to. This is, in substance, an attempt to redefine disability for scheme-access purposes against the barriers-based definition the CRPD establishes and the Disability Discrimination Act 1992 reflects: a test of functional capacity that excludes the person’s environment and circumstances is the medical model of disability the CRPD framework was adopted to displace. (Sections 3, 7.)

Recommendation 6. The Committee recommend that the amendment to the Objects clause at new s.3(1)(d) — which subordinates the provision of reasonable-and-necessary supports to the objective of financial sustainability — be omitted from the Bill. The amendment subordinates a participant-rights object to a fiscal management object in a way that has no parallel in any other Commonwealth rights-based statutory scheme and is inconsistent with NDIS Act 2013 s.3’s grounding in CRPD obligations. The “financial sustainability” the amendment elevates is not defined in the Bill, and the “unsustainable” growth it is said to address has not been demonstrated against the Government’s own commissioned analyses, which found net economic benefit at the whole-of-scheme level (Section 14). (Sections 7, 14.)

Recommendation 7. The Committee recommend that the new s.48A participant-initiated plan reassessment provisions be omitted from the Bill or substantively amended to preserve participant agency over plan content. Specifically: the requirement that a request for reassessment demonstrate a “significant change” in support needs should not apply to plan updates following a completed functional capacity assessment, which reflects a better identification of existing stable needs rather than a change in those needs; the “substantial reduction in the participant’s ability to perform daily activities” threshold at s.48A(2)(b) should not apply to impairment categories — including autism, psychosocial disability, and intellectual disability — characterised by episodic, variable or context-

Submission 2079

dependent need; and the extended 90-day decision window (from the current 21 days) should be reconsidered in light of the waiting-period vulnerability the submission documents at Section 2. (Section 5.)

Recommendation 8

The Committee recommend that the automatic plan renewal provisions at new s.50A be omitted from the Bill or substantially amended. Automatic plan renewal without participant input, without a fresh statement of supports, and without generating a reviewable decision is structurally incompatible with the scheme’s participant-centred design and with the s.31 Principles the wholesale removal at Schedule 1 Part 6 is simultaneously removing. (Section 5.)

Part B — Recommendations on Schedule 2

Recommendation 9

The Committee recommend that Schedule 2 be deferred - not opposed - until after the Joint Standing Committee on the NDIS Integrity of the National Disability Insurance Scheme inquiry reports on Thursday 2 July 2026. Schedule 2’s subject matter is the subject matter the JSC inquiry is investigating. The JSC’s findings and recommendations should be integrated into the Schedule 2 design before Schedule 2 returns to Parliament. Deferring Schedule 2 does not dissolve the existing integrity infrastructure: the NDIS Amendment (Integrity and Safeguarding) Act 2026, which received Royal Assent on 8 April 2026, remains in force; the approximately 600-strong NDIA integrity workforce remains operational; and the Fraud Fusion Taskforce’s 24-Commonwealth-agency architecture continues to function. (Sections 9, 10, 15.)

Recommendation 10

The Committee recommend that, when Schedule 2 is re-brought to Parliament following the JSC’s report, it be expanded to incorporate the elements of the 2024 Registration Taskforce Final Advice, the NDIS Review Recommendation 17 universal-registration architecture, and the launch-era expert lineage on thin markets, workforce capacity, and market stewardship identified in the Productivity Commission’s 2011 and 2017 reports - none of which is in the Schedule 2 currently before the Committee. (Section 10.)

Recommendation 11

The Committee recommend that any revised Schedule 2 be subject to substantive community consultation under “nothing about us without us” principles, with a consultation period that the Disability Representative Organisations Joint Statement of 14 May 2026 characterises as adequate, before it is returned to the Parliament for consideration. (Sections 9, 15.)

Part C — Recommendations on Schedule 3

Recommendation 12

The Committee recommend that Schedule 3 Part 1 - the Ministerial pricing decision-maker model in new s.45C and adjacent provisions - be omitted from the Bill. The concentration of pricing discretion in the Minister, exercisable through legislative instruments with disallowance as the sole accountability mechanism, is not consistent with the independent-monitoring architecture the Productivity Commission’s 2017 NDIS Costs report contemplated. Independent pricing oversight of the

Submission 2079

NDIS should be restored before any further legislative step on the pricing architecture. The Committee should note in particular that, because NDIA-managed and plan-managed participants cannot lawfully pay a gap above the cap, a price set below the rate at which any provider will deliver a support removes that support entirely for those participants — the pricing power is, in operative effect, a power to exclude supports from the scheme by under-pricing them indiscriminately.

(Section 12.)

Recommendation 13

The Committee recommend that Schedule 3 Part 2 — the automation of administrative action and evaluative determinations under new s.59B to s.59D — be omitted from the Bill. The provisions authorise automation of determinations involving discretion, evaluative judgement, and state of mind, with internal-process (CEO oversight) rather than external-accountability safeguards. The 2023 Royal Commission into the Robodebt Scheme made specific findings about the participant-safety risks of automated welfare decision-making and the inadequacy of internal-process oversight where vulnerable cohorts are affected. The participant cohort the Bill applies to — substantially comprising Autistic Australians, intellectually disabled Australians, and Australians with psychosocial disability — is a cohort whose engagement with automated decision processes is structurally affected by their disability. The submission refers to these automation provisions as RoboNDIS, reflecting their structural parallel with the automated welfare-decision architecture the Robodebt Royal Commission found caused systemic harm.

(Section 13.)

Submission 2079

Recommendation 16.

The Committee recommend that the Minister’s claim in his Second Reading Speech of 14 May 2026 - that the Bill “takes up recommendations… of the Royal Commission into Violence, Abuse, Neglect and Exploitation of People with Disability” - be substantiated against the Bill’s own Explanatory Memorandum, which cites four Disability Royal Commission recommendations (9.4, 10.2, 10.3 and 10.4) at section 5.4 as informing Schedule 2 Part 6 plan management and support coordination commissioning provisions only, and does not cite any Royal Commission recommendation in respect of Schedule 1’s eligibility-tightening provisions. The Committee should require the Department of Health, Disability and Ageing to identify, for the public record, any further Royal Commission recommendations the Bill operationalises and the Schedules in which each is implemented. If no further mapping exists, the Royal Commission reference in the Second Reading Speech - to the extent it implies Royal Commission endorsement of the Schedule 1 access-tightening architecture - should be corrected for the parliamentary record. The Committee should further require the Department to explain, for the public record, why the Royal Commission recommendations the Commonwealth has formally accepted (130 of the 172 for which it has primary or shared responsibility) are not implemented or engaged by this Bill, how the Government intends to satisfy those accepted recommendations, and how it has assessed the Bill against the targets it has set itself under the National Autism Strategy 2025-31 and the National Agreement on Closing the Gap.

(Section 8.)

Recommendation 17.

The Committee recommend that it seek, from the Department, the Technical Advisory Group’s terms of reference, membership composition, evidence inputs, and the projected timeline for the Group’s advice on the functional capacity assessment instruments — and that the Committee assess whether the Group’s timeline is consistent with the 1 January 2028 commencement date for the new functional capacity regime.

(Sections 3, 5, 8.)

Recommendation 18.

The Committee recommend that the consultation period for the reform package be assessed against the standard the Disability Representative Organisations Joint Statement of 14 May 2026 identifies, including whether the one-month Senate inquiry window is proportionate to the fiscal magnitude of the package - characterised by the CEO of the e61 Institute (former Chair, Productivity Commission) as the largest single-program Budget save in Australian fiscal history - and whether the Aboriginal and Torres Strait Islander disability community received specific targeted engagement in the development of the Bill.

(Section 15.)

Recommendation 19.

The Committee recommend that it seek from the Department the implementation modelling underlying the 2 per cent annual scheme-growth target: the assumptions behind the target, the operational mechanisms for achieving the save at the participant and plan level, and the analysis of why the 2024-25 Budget’s comparable savings measure (“Getting the NDIS Back on Track”, $14.3 billion over four years) was not achieved on projection — a history the e61 Institute quantified as a $13.9 billion upward revision to NDIS costs over five years prior to the current Bill’s savings measure.

Submission 2079

Budget’s fiscal strategy rests on the save; the Committee is entitled to understand the implementation basis for it. (Section 14.)

Recommendation 20

The Committee recommends that an Inspector-General of the NDIS be established as an independent statutory oversight role with public-reporting obligations and a charter to monitor the operation of the scheme. This recommendation implements Recommendation 4 of the House Standing Committee on Health, Aged Care and Disability’s December 2025 report No child left behind (the Inquiry into the Thriving Kids initiative). The Inquiry’s Recommendation 4 is on the Government’s own December 2025 record. The Bill before this Committee does not legislate an Inspector-General. The oversight architecture of the NDIS remains the architecture the Government’s own Inquiry recommended five months ago be supplemented by an independent oversight officer. (Sections 8, 15.)

Recommendation 21

The Committee recommend that, if Thriving Kids commences on 1 October 2026 as currently scheduled, a comprehensive review of Thriving Kids implementation be undertaken at 24 months after commencement — with the review’s terms of reference, methodology, evidence inputs, reporting obligations, and capacity to recommend operational and legislative changes established and published before commencement. This recommendation implements Recommendation 10 of the House Standing Committee’s December 2025 report No child left behind. The Inquiry’s Recommendation 10 reflects the Committee’s recognition that the design of Thriving Kids needed to be tested in operation before being treated as a settled architecture. The Bill commences Thriving Kids without the 24-month review mechanism the Government’s own Inquiry recommended. (Sections 8, 15.)

Recommendation 22

The Committee recommends that the Department of Health, Disability and Ageing be required to substantiate the “mild to moderate” / “low to moderate support needs” diagnostic-language framing on which the reform’s cohort targeting rests, against the operative clinical and statutory documents — the DSM-5, the Independent NDIS Review, and the National Autism Strategy 2025–31, none of which uses that framing — and that the Thriving Kids cohort description in Budget Paper No. 2 (“children aged eight and under with developmental delay and/or autism with low to moderate support needs”) either be reconciled with the DSM-5 framework and the National Autism Strategy 2025–31, or be withdrawn as the targeting basis for the foundational supports program. (Sections 4, 8.)

Recommendation 23

The Committee recommend that it seek and place on the parliamentary record the findings of the 2025 Australian National Audit Office performance audit of the NDIS Quality and Safeguards Commission’s regulatory functions (Effectiveness of the NDIS Quality and Safeguards Commission’s Regulatory Functions, Auditor-General Report No. 2 of 2025–26), which found the Commission only partly effective in its existing regulatory functions, and that it assess whether the Commission’s current operational state supports the expanded enforcement role Schedule 2 confers on it. The Commission’s demonstrated operational capacity is the proper evidentiary

Submission 2079

foundation for any expansion of its powers. These ANAO findings should be before the Committee, and integrated, on the same footing as the JSC NDIS Integrity inquiry’s findings (Recommendation 9): both bear directly on Schedule 2, and both should inform the redesigned Schedule 2 before it returns to the Parliament. (Sections 9, 10.)

Overarching recommendation

Recommendation 24. The Committee recommend that, as an overarching principle, the Government be required either to implement the recommendations of the evidence base it relies on for this reform — the Disability Royal Commission, the Independent NDIS Review, The National Autism Strategy 2025-31, the Thriving Kids Inquiry, and the Productivity Commission’s reports — or to place on the public record a substantive, reasoned analysis of each material recommendation it has chosen not to implement, and why. A Government that cites these reviews as the authority for the Bill should be held to the discipline of accounting publicly for the recommendations of those same reviews that it declines to adopt, and for how the Bill is assessed against the targets the Government has set itself under the National Autism Strategy 2025-31 and the National Agreement on Closing the Gap. (Sections 8, 10, 14, 15.)

Submission 2079

Section 17 — Conclusion and disclosure

Conclusion

The Bill before the Committee is not a single measure. It is three structurally distinct measures in a single document — an eligibility tightening, a provider-integrity package, and a pricing-and-automation package — each resting on a different evidence base, each warranting separate scrutiny, and each requiring a different response from the Committee. My submission has treated them separately because the Government’s own case for them is separate, and because the Committee’s recommendations will be clearer and more actionable if they follow the Bill’s own separable structure. But separable does not mean independent in effect: Schedule 1 and Schedule 3 actively work against the integrity improvements Schedule 2 is meant to deliver (Sections 10 and 11). And it is Schedule 2 — the provider-integrity package — that has carried the Government’s public case for the Bill, even though the measure that dominates it in fiscal terms is the Schedule 1 cost reduction. That ordering was not accidental: as Section 8 sets out, taxpayer-funded message-testing advised the Government to lead with fraud and integrity in order to build public tolerance for the access restriction that does the fiscal work.

Schedule 1 should be omitted from the Bill. The clinical-evidence base for the permanence-and-treatment test does not exist. The assessment instruments have not been designed. The replacement supports programs for the cohorts most directly affected — the 9-to-15 cohort, the 16-to-17 cohort, and adults — have not been named, funded, or scoped. The s.25A(2) provision writes into statute an income-and-geographic regressivity that the Disability Royal Commission identified as a systemic inequity. The s.31 Principles removal strips the statutory framework for supported decision-making from the plan architecture. Four distinct statutory and treaty inconsistencies are established on the face of the operative text. The existing s.34 reasonable-and-necessary framework already provides the mechanism the Government says Schedule 1 is needed to operate; the access gate is structurally redundant for the cost-shift purpose the Explanatory Memorandum advances; and the correct cost-reduction mechanism — building the foundational supports ecosystem so that the existing s.34 test can do its work — has been sequenced behind the access restriction rather than before it. The cost of getting that sequencing wrong is not only fiscal: the documented transition out of the Personal Helpers and Mentors Scheme left people without active supports, and Section 5 sets out the lived-experience record — including a participant found dead at home after his plan lapsed — that this Bill’s transition-gap structure risks reproducing for the cohorts it most affects. And the cost is not only fiscal and human: omitting Schedule 1 is also what the Commonwealth’s own legal obligations point to, given the four statutory and treaty inconsistencies — with the NDIS Act 2013, the Disability Discrimination Act 1992, and Australia’s obligations under the CRPD — established on the face of the operative text.

Schedule 2 should be deferred, not opposed. The provider-side and integrity reforms are broadly aligned with what the Government’s own integrity inquiry is investigating. The JSC reports on 2 July 2026. Schedule 2’s evidence base should wait for the JSC’s findings,

Submission 2079

which should be integrated into the design of a fuller Schedule 2 that is then brought back to Parliament with substantive community consultation. Deferral preserves the integrity infrastructure that already exists while getting the design right. The 2025 Australian National Audit Office findings on the Quality and Safeguards Commission’s regulatory effectiveness should be before the Committee, and integrated on the same footing as the JSC’s findings, before any expanded enforcement role is conferred.

Schedule 3 should be omitted from the Bill. The Ministerial pricing-decision-maker model is less independent than the architecture the Productivity Commission contemplated. The automation of evaluative determinations, in the participant cohort this Bill applies to, without the external-accountability safeguards the Robodebt Royal Commission identified as necessary, is a participant safety risk that the Bill’s internal-process safeguards do not adequately address.

The procedural record compounds the substantive case. The Government commissioned its own integrity inquiry; introduced legislation based on that inquiry’s ongoing evidencease before the inquiry reported; structured the Senate’s review to conclude sixteen days before the inquiry’s findings are due; and gave the community — through the Disability Representative Organisations, twelve of them, the organisations the Government funds specifically to provide disability-community voices in the policy process — approximately one month to respond to legislation that the e61 Institute, whose CEO is a former Productivity Commission Chair, characterises as the largest single-program Budget save in Australian fiscal history with no historical precedent.

I have written this submission as a directly affected person — an Autistic adult, a current NDIS participant, the parent of two Autistic children, a resident of the Hasluck electorate in Western Australia. The arguments in these pages rest on the Government’s own published Budget papers, on primary legislation and the Explanatory Memorandum, on credentialed expert analysis, on peer-reviewed clinical literature, on the Disability Royal Commission’s Final Report, and on the parliamentary record of the JSC hearings. Where I have asserted facts, I have cited them; where I have made arguments, I have shown the evidence they rest on. I have not been moderate where the Bill is immoderate. I have not softened the case where the case is not soft.

The scheme this Bill proposes to reform is a statutory scheme grounded in Australia’s CRPD obligations, enacted by Parliament, and designed to provide reasonable and necessary supports to people with disability across the lifespan. My daughter’s education depends on it. My working life depends on it. On the Government’s own modelling, 241,000 existing participants are projected to lose access by the end of the decade, and close to 350,000 fewer Australians will be supported once those denied entry are counted. They, and their families, are in positions comparable to ours. The Committee’s scrutiny of this Bill is their scrutiny. I am grateful for the opportunity to contribute to it.

Submission 2079

Disclosure — use of accessibility tools

This submission has been prepared with the assistance of artificial intelligence tools, used as cognitive accessibility infrastructure to support my participation in the democratic process. I have late-diagnosed autism, with associated executive function, communication, and information-processing characteristics that make drafting lengthy formal documents without structural and language support significantly more effortful and less accessible than it is for neurotypical submitters. Artificial intelligence tools have assisted me with structuring the submission, drafting sections from my research and analysis, and with language polish and source-tracking. The substantive arguments, the lived experience material, the primary-source research and verification, the factual claims, and the positions adopted throughout this submission are mine. I have reviewed every section and verified the primary sources on which substantive claims rest. The assistance is analogous to the support a neurotypical submitter with access to research staff and professional editors would receive, and reflects the Committee’s own invitation to participants to use accessibility tools where required. The underlying advocacy, the analytical judgements, and the views expressed are mine alone.

I make this submission in my personal capacity. The views expressed are mine and do not represent the views of my employer.

Submission 2079

Section 18 — Annexes

Annex A — Key primary sources and references

The following are the primary sources cited in this submission. Full source records, including URL verification dates and cross-document citation tracking, are held by the author.

Primary legislation and Bills

• National Disability Insurance Scheme Act 2013 (Cth) — ss.3, 4, 9B (proposed), 24, 24(5) (proposed), 25, 25(1B) (proposed), 25A (proposed), 25B (proposed), 31, 34, 34A (proposed), 48, 48A (proposed), 50A (proposed), 59B–59D (proposed).
• National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Bill 2026 (Cth) — as introduced 14 May 2026, including Schedules 1, 2, and 3 and the Explanatory Memorandum.
• Disability Discrimination Act 1992 (Cth) — ss.4, 5, 6.
• Regulatory Powers (Standard Provisions) Act 2014 (Cth).

International instruments

• United Nations Convention on the Rights of Persons with Disabilities — Articles 1, 5, 12, 19, 25.

Budget papers

• Commonwealth of Australia, Budget Paper No. 2 — Budget Measures: Budget 2026- 27 — “Securing the National Disability Insurance Scheme for Future Generations” measure (lines 8919–8963); “Thriving Kids” measure (line 9244); Education Portfolio — disability loadings to schools” measure; “Inclusive Communities Fund” measure (line 8937). Available: https://budget.gov.au/content/bp2/index.htm
• Commonwealth of Australia, Budget Paper No. 1 — Budget Strategy and Outlook: Budget 2026-7 — Statement 3, Box 3.2 (3:1 spending-to-revenue ratio; $184.9 billion medium-term NDIS reduction).

Government-commissioned reviews

• Bonyhady, B. and Paul, L., Working Together to Deliver the NDIS — Independent Review of the NDIS, December 2023 (Commonwealth of Australia). Available: https://www.ndisreview.gov.au
• Disability Royal Commission (Royal Commission into Violence, Abuse, Neglect and Exploitation of People with Disability), Final Report, September 2023 — in particular Volume 6, Enabling Autonomy and Access, Recommendation 6.34.
• Commonwealth of Australia, National Autism Strategy 2025-2031, 2025. Available: https://www.health.gov.au/our-work/national-autism-strategy
• Productivity Commission, Disability Care and Support, Report No. 54, 2011.

Submission 2079

• Productivity Commission, NDIS Costs, Study Report, 2017.
• Productivity Commission, Foundational Supports — Working Paper, 2024.
• Taylor Fry Consulting Actuaries, Independent Evaluation of the NDIS — Final Report, commissioned for the Independent Review of the NDIS, 2023.
• NDIS Provider and Worker Registration Taskforce, Final Advice, 2024.
• People with Disability Australia (PWDA), Submission to the NDIS Evidence Advisory Committee Third Round Consultation, 26 March 2026.

Parliamentary record

• Minister for Disability and the NDIS (Hon Mark Butler MP), Second Reading Speech — National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Bill 2026, Hansard, 14 May 2026.
• Minister for the NDIS (Hon Mark Butler MP), National Press Club Address, 22 April

• Joint Standing Committee on the NDIS, Integrity of the National Disability Insurance Scheme inquiry, Public Hearing Transcript, 1 May 2026, Canberra — testimony of Mr John Dardo (NDIA), Ms Natalie Wade (NDIS Quality and Safeguards Commission), Mr Iain Anderson (Commonwealth Ombudsman) and others.
• Senate Community Affairs Legislation Committee, Inquiry into the National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Bill 2026 — referral 14 May 2026; reporting date 16 June 2026. https://www.aph.gov.au/Parliamentary_Business/Committees/Senate/Community _Affairs/NDISFutureGenBill
• Joint Standing Committee on the NDIS, Integrity of the National Disability Insurance Scheme inquiry — referral 25 March 2026; reporting date 2 July 2026.

Agreements and national frameworks

• National Agreement on Foundational Supports 2026-31 — signed by Commonwealth and all eight States and Territories, January–February 2026. Available: https://federalfinancialrelations.gov.au/agreements/national- ageement-foundational-supports
• National Agreement on Closing the Gap, 2020 — Priority Reforms 3 and 4.

Peer-reviewed research and credentialed expert analysis

• Disney, G. et al., “Socioeconomic disadvantage and access to Australia’s National Disability Insurance Scheme: A retrospective cohort study of 705,594 NDIS applications, 2016–2022”, Medical Journal of Australia, 2025.
• Brennan, M. (e61 Institute), “Hiding in plain sight: the big item that underpins the Budget”, LinkedIn Pulse, 14 May 2026. Available: https://www.linkedin.com/pulse/hiding-plain-sight-big-item-underpins-budget-e61-institute-ltd-znjyc

Submission 2079

• Australia Institute, Budget analysis — PRRT and gas revenue, media release, 12 May 2026. Economist: Matt Grudnoff (Senior Economist).
• Per Capita, False Economy: The economic benefits of the NDIS and the consequences of government cost-cutting, 2021 — return of approximately $2.25 in economic activity per $1 of NDIS expenditure; approximately 10,200 jobs and $2.25 billion in economic activity lost per $1 billion of underfunding.
• Akyol, P. (e61 Institute), “More Quasi than Market: Price Caps and Market Design in the NDIS”, Economic Society of Australia (NSW) seminar, 27 May 2026.
• Grattan Institute, Better, safer, more sustainable: How to reform NDIS housing and support, 2025.

Legal service and disability community analysis

• Villamanta Disability Rights Legal Service, Community Forum Slides — NDIS Amendment Bill 2026 Analysis, 20 May 2026, presented by Elise Almond, Kelly Wilkins, and Trinity Ford.
• Australian Human Rights Commission (Commissioner Rosemary Kayess, Disability Discrimination Commissioner), media release responding to Mosman Park incident, February 2026. Available: https://humanrights.gov.au/about-us/media- centre/media-releases/disability-rights/commissioners-say-wa-tragedy- underscores-urgent-need-to-keep-all-children-safe-from-harm
• Queensland Advocacy for Inclusion, #TheDRCStillMatters implementation analysis, 2026 — finding that, across the Disability Royal Commission’s 222 recommendations, only 3 per cent have been implemented and 53 per cent have no reported progress.

Peak-body statements

• Disability Representative Organisations Joint Statement, “Call for Genuine Senate Scrutiny as Concerns Grow Over Rushed NDIS Reforms”, 14–15 May 2026. Signatories: PWDA, AFDO, Australian Autism Alliance, CYDA, DANA, Inclusion Australia, Mental Health Australia, First Peoples Disability Network Australia, NEDA, PDA, WWDA, Down Syndrome Australia.

Departmental correspondence and tabled modelling

• Department of Health, Disability and Ageing, Ministerial response MC26-008601, 25 May 2026 (signed by an Assistant Secretary, Thriving Kids Taskforce).
• Department of Health, Disability and Ageing, NDIS reform modelling tabled in the Senate on 27 May 2026 — projecting 241,000 existing participants exited by mid- 2031 and the $13.2 billion community-participation funding cut as the single largest saving measure (reported by Guardian Australia, 28 May 2026).