Submission 366

Submission to the Senate Community Affairs Legislation Committee

National Disability Insurance Scheme Amendment (Securing the NDI$ for Future Generations) Bill 2026

Prepared by:

• Niesha Illingworth (President) on behalf of the Association forBehaviour Analysis Australia

Approved by:

• Executive Board of Directors

Date:

• 29th May 2026

Association for Behaviour Analysis Australia

• PO Box 61
• Sandy Bay, TAS 7005
• Email: admin@auaba.com.au
• Website: auaba.com.au
• ABN 24165099392

Submission 366

Table of contents

Executive summary ……………………………………………………………………………………………………………………… 3

1. Introduction ……………………………………………………………………………………………………………………………… 4

2. Schedule 1, Part 9, item 97 (proposed section 25B): alternative support requirements ………………………. 6

• 2.1 The provision ……………………………………………………………………………………………………………………… 6

• 2.2 The substantive concern ………………………………………………………………………………………………………. 7

• 2.3 Why this matters for evidence-based early intervention …………………………………………………………… 8

• 2.4 Recommendations ……………………………………………………………………………………………………………….. 9

3. Schedule 1, Part 6, item 68 (proposed section 33(2EA)): maximum intensity determinations …………… 10

• 3.1 The provision ……………………………………………………………………………………………………………………. 11

• 3.2 The category error in moving from group-level to individual-level ………………………………………….. 11

• 3.3 The disproportionate effect on underrepresented populations ………………………………………………….. 13

• 3.4 The relationship between proposed subsection 33(2EA) and proposed subsection 34(1) ……………. 13

• 3.5 Recommendations ……………………………………………………………………………………………………………… 14

4. Schedule 1, Part 6, item 73 (proposed sections 34(1E) and (1F)): the evidence hierarchy ………………… 15

• 4.1 The provision ……………………………………………………………………………………………………………………. 15

• 4.2 The inversion of clinical practice ………………………………………………………………………………………… 16

• 4.3 The interaction with the dosage ceiling provision ………………………………………………………………….. 17

• 4.4 Recommendations ……………………………………………………………………………………………………………… 17

5. Schedule 1, Part 6, item 73 (proposed sections 34(1G) and (1H)): the parental responsibility presumption and “behavioural support” ………………………………………………………………………………………… 18

• 5.1 The provision ……………………………………………………………………………………………………………………. 19

• 5.2 The conflation of professional behaviour support with ordinary parental responsibility ……………… 19

• 5.3 Recommendation ………………………………………………………………………………………………………………. 20

6. The support needs of children with significant developmental disability ……………………………………….. 20

• 6.1 The reality of support needs in this cohort ……………………………………………………………………………. 21

• 6.2 The underestimation of these needs in policy framings ………………………………………………………….. 22

• 6.3 The cumulative effect of the proposed amendments on this cohort ………………………………………….. 22

7. The broader question of safeguards in the reformed Scheme ………………………………………………………… 23

Summary of recommendations …………………………………………………………………………………………………. 24

Attachments …………………………………………………………………………………………………………………………… 25

References ……………………………………………………………………………………………………………………………. 25

Submission 366

Executive summary

The National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Bill 2026 is presented as a measure to secure the sustainability of the Scheme and to deliver more individualised, participant-centred support. ABA Australia supports those objectives. Our concern is that four of the Bill’s mechanisms operate in the opposite direction. Each is a population-level power capable of overriding the supports an individual participant has been individually assessed to reasonably and necessarily require. There is, in our view, a tension between the policy intent of the reforms and the legislative mechanisms chosen to give effect to it. The reforms are described in terms of choice, control, and individualised support. The provisions discussed in this submission instead allow decisions about individuals to be made by reference to impairment class, population-level cost caps, and generalisable research, with limited safeguards for the individual whose circumstances fall outside the assumptions underpinning the rule.

ABA Australia’s four principal concerns:

• Exclusion of impairments (proposed s 25B). The Bill enables whole classes of impairment to be excluded from the Scheme by Rule, on a minimal threshold for ministerial satisfaction and with no individual exception for a person whose clinical presentation falls outside the class. The mechanism can revoke the status of existing participants, not only new applicants.

• Maximum intensity caps (proposed s 33(2EA)). The Bill permits a statutory ceiling on the intensity of an individual’s support to be set from group-level research. As the Explanatory Memorandum’s own worked example shows, this can reduce funding below the level a planner has assessed as reasonable and necessary, and it falls hardest on participants with the most significant needs.

• The evidence hierarchy (proposed ss 34(1E)–(1F)). The Bill ranks published, generalisable research above evidence of what has actually worked for the individual participant, inverting the integrative model of evidence-based practice and disadvantaging the participant groups least represented in generalisable research.

• Behavioural support as parenting (proposed ss 34(1G)–(1H)). The Bill lists “behavioural support” among the activities presumed to fall within ordinary parental responsibility, conflating everyday parenting with regulated, professional positive behaviour support delivered by qualified practitioners.

Submission 366

These provisions affect the same cohort — children and adults with significant developmental disability and high support needs — in combination, and their cumulative effect is greater than the sum of its parts.

ABA Australia’s recommendations, set out in full at section 8, are that the Bill be amended to:

• prescribe the matters of which the Minister must be satisfied before excluding an impairment or capping intensity, including the capacity, accessibility, and workforce of any alternative support system, and the state of the evidence relied upon;

• provide accessible, supported individual exception and review mechanisms wherever a population- level rule is applied to an individual;

• remove the hierarchical ordering of evidence in s 34(1E) and the power in s 34(1F) to disregard participant-specific evidence; and

• remove “behavioural support” from the parental responsibility presumption in s 34(1H), or confine it to ordinary parental management of behaviour.

ABA Australia’s position throughout rests on a single principle: where a population-level rule is applied to an individual, the legislation must require the decision-maker to weigh the evidence about that individual and must provide an accessible means by which a person whose circumstances fall outside the rule can be heard.

Introduction

ABA Australia welcomes the opportunity to make a submission to the Senate Community Affairs Legislation Committee on the National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Bill 2026.

ABA Australia is the national self-regulatory body for behaviour analysts in Australia and a qualifying member of the National Alliance of Self-Regulating Health Professions. A significant proportion of our membership delivers services to NDIS participants, including autistic children, children with developmental delay, and people with intellectual or developmental disability. Many of our members work directly in early intervention for young children and their families. The matters addressed in this Bill are therefore of substantial relevance to the participants we support and to the practitioners we represent.

This submission builds on ABA Australia’s prior engagement with the reform process, including our submission to the NDIS Evidence Advisory Committee on Early Intensive Behavioural Intervention (27 March 2026) and our follow-up correspondence to the Therapies Research Branch of the Department of

Submission 366

Health, Disability and Ageing (15 April 2026). The evidence base supporting that engagement remains

relevant to the matters considered here. We have not repeated it in detail in this submission and refer the

Committee to our submission to the NDIS Evidence Advisory Committee on Early Intensive Behavioural

Intervention (27 March 2026), which we provide as Attachment A to this submission and which sets out

the evidence base in full.

ABA Australia supports the overall objective of placing the National Disability Insurance Scheme on a sustainable footing for future generations. We also support, as a matter of principle, the development of foundational supports outside the Scheme to ensure that children with less significant needs are supported earlier and through community-based services that are universally available. The concerns set out in this submission are not directed at those broader objectives. They are directed at specific provisions of the Bill that, in our view, require amendment to ensure that the reformed Scheme continues to provide appropriate access to evidence-based supports for children and adults whose disability requires specialist, individualised intervention.

This submission addresses six matters arising from the Bill:

• Schedule 1, Part 9, item 97 (proposed section 25B), concerning alternative support requirements and the legislative mechanism enabling exclusion of impairments from the Scheme.

• Schedule 1, Part 6, item 68 (proposed section 33(2EA)), concerning maximum intensity determinations.

• Schedule 1, Part 6, item 73 (proposed sections 34(1E) and (1F)), concerning the evidence hierarchy for determining whether a support is effective and beneficial.

• Schedule 1, Part 6, item 73 (proposed sections 34(1G) and (1H)), concerning the parental responsibility presumption and the inclusion of “behavioural support” within ordinary parental responsibility.

• The support needs of children with significant developmental disability and the cumulative effect of the proposed amendments on this cohort.

• The broader question of whether the Bill, as currently drafted, provides sufficient safeguards to ensure that participants with significant disability who require evidence-based specialist intervention will continue to access it under the reformed Scheme.

A summary of recommendations is provided at section 8.

Submission 366

Schedule 1, Part 9, Item 97 (proposed Section 25B):

Alternative Support Requirements

Illustrative case example (proposed s 25B): loss of participant status by impairment class

Background. Mia is seven. She is autistic and has significant developmental disability. She uses few words, needs help with most daily living tasks, and is unsafe without constant supervision. Her family lives in a regional area and relies on a small visiting team, including a Certified Behaviour Analyst who designs and supervises her supports. Over the last plan period Mia began using a simple system to request help and a break, and her absconding and climbing reduced.

The problem under the proposed legislation. Under proposed section 25B, an impairment can be declared an excluded impairment by reference to a class of impairment, not Mia’s individual presentation, and the application provision permits revocation of status for existing participants. Mia could lose access to the Scheme on the basis of her diagnostic category, regardless of her severity, complexity, or recent gains.

The human consequence. In her region, foundational supports are unlikely to exist at the intensity and specialisation Mia requires. The foreseeable result is regression, renewed safety risk, and family burnout.

What the legislation should require instead. Before participant status is removed, the Bill should require an assessment of individual clinical presentation, and of whether the alternative system can actually deliver evidence-based supports at the required intensity and quality (Recommendations 1 and 2).

Illustrative, hypothetical example; not based on an identifiable individual.

2.1 The Provision

Proposed Section 25B, inserted by Schedule 1, Part 9, item 97, establishes the legislative mechanism by which an impairment can be made an “excluded impairment” for the purposes of the NDIS. Under proposed subsection 25B(4), an impairment is an excluded impairment if a support is declared by the National Disability Insurance Scheme rules to be an “alternative support” for the impairment. Where an impairment is an excluded impairment, the person does not meet the alternative support requirements and therefore does not meet the requirements for becoming or remaining a participant in the Scheme.

Submission 366

The Explanatory Memorandum makes clear that this mechanism is intended to support the rollout of the

Thriving Kids initiative from 1 October 2026, with the access changes for autistic children and children

with developmental delay aged 8 and under commencing from 1 January 2028. The Bill provides the

legislative power. The substantive declarations giving effect to Thriving Kids will be made by Rule.

ABA Australia recognises that the policy intent of Thriving Kids is to provide a coherent and accessible

foundational support system for children with less significant needs, delivered through community-based

services and universally available across Australia. We support that intent. Children with developmental

delay or autism whose presentation does not require specialist, intensive intervention are likely to benefit from earlier, more accessible support through foundational systems than from delayed access to the NDIS. This is consistent with the broader logic of a national tiered support system, and with longstanding clinical guidance on early identification and intervention.

2.2 The substantive concern

ABA Australia’s concern is not with the policy direction of Thriving Kids. It is with the breadth of the section 25B mechanism, the consequences of its use for children whose needs require specialist intervention, and the absence of safeguards in the Bill to ensure that the mechanism is not used in ways that exclude from the Scheme children whose clinical presentation indicates a continuing need for NDIS-funded supports.

Three features of the provision warrant the Committee’s attention.

• First, proposed subsection 25B(4) permits an impairment to be excluded by reference to a class of impairment, not by reference to individual clinical presentation. Where the Rule operates by impairment class, all participants with that impairment are affected, regardless of severity, complexity, or individualised support need. The section provides no mechanism by which an individual whose clinical presentation falls outside the categorisation underpinning the Rule may continue to access the Scheme.

• Second, proposed subsection 25B(6) provides that, before making such a Rule, the Minister must be satisfied that “it is not appropriate to fund or provide a support for the impairment through the National Disability Insurance Scheme”. The Bill does not prescribe what the Minister must take into account in reaching that satisfaction. It does not require the Minister to be satisfied that the alternative support system has the capacity to deliver evidence-based intervention at the intensity and quality required for the cohort to which the Rule applies. It does not require consultation with clinical experts or peak bodies. It does not require an evidence-based assessment of comparable outcomes between the alternative support system and

Submission 366

the supports that would otherwise have been available under the Scheme. The threshold for ministerial

satisfaction is therefore minimal and procedurally light, despite the consequential effect of the Rule.

Third, and most significantly, the application provision at item 101 makes clear that the revocation of

participant status under proposed paragraph 30(1)(c) applies to existing participants, regardless of when

they became participants. The section 25B mechanism therefore does not only affect new applicants. It

enables the Scheme to revoke participant status for children and adults already receiving NDIS supports.

For families currently receiving intensive early intervention for a young child with autism or developmental

delay, the consequences of a Rule made under section 25B(4) are not theoretical. They are the loss of access

to supports their child is currently receiving.

Underlying each of these features is a practical concern about burden. Where the only route to continued

access is an individual exception, the burden of invoking it falls on families — often families already

managing high daily care needs, living in regional areas, or with limited resources — who must identify

the pathway, assemble clinical evidence, and pursue a review against a decision that has already been made.

A safeguard that exists only in principle, and is difficult to access in practice, will not protect the children

most affected.

Why this matters for evidence-based early intervention

The cohort identified for transition under the Thriving Kids initiative is “children aged 8 and under with”

low to moderate autism and developmental delay“. ABA Australia’s substantive concern, informed by the

experience of our members, is that the criteria for categorising children as “low to moderate” will not

adequately identify the children for whom specialist intensive intervention is clinically indicated.

The evidence base for behaviour analytic early intervention, set out in detail in ABA Australia’s submission

to the EAC, supports the conclusion that, for an appropriate cohort of young children with developmental

needs, individualised intervention delivered at appropriate intensity by a qualified workforce produces

meaningful improvements in communication, adaptive functioning, daily living skills, participation, safety,

and family capability. The strongest effects are observed where intervention is comprehensive,

individualised, and delivered with sufficient intensity to address the child’s developmental profile. The

recent individual participant data meta-analysis by Eldevik and colleagues (2026), based on 621 children

across 15 controlled studies, reported clinically meaningful effect sizes across adaptive behaviour,

cognitive functioning, and autism severity.

These outcomes are not achievable through generalised community-based support. They are achieved by

qualified clinicians, working within a defined supervisory framework, delivering individualised

Submission 366

intervention with appropriate intensity over time. Foundational supports of the kind contemplated by

Thriving Kids are an important part of a tiered service system, but they are not a substitute for specialist

intervention where it is clinically indicated.

The risk created by the section 25B mechanism, in the absence of further safeguards, is that children whose

c临ical presentation supports specialist intervention may be categorised as “low to moderate” and

transitioned out of the Scheme. For some children, this will result in the loss of access to evidence-based

intervention during the developmental window in which that intervention is most consequential. The cost

will not only be borne by those children and their families. It will also be borne by the Scheme over time,

as support needs that could have been addressed early become more entrenched.

2.4 Recommendations

ABA Australia recommends amendment to proposed section 25B to introduce the following safeguards:

 a. Proposed subsection 25B(6) should be amended to prescribe the matters that the Minister must be

     satisfied of before making a Rule. These should include, at a minimum: that the alternative support

   system has the capacity to deliver evidence-based intervention at the intensity and quality required

    for the cohort to which the Rule applies; that the alternative support system is universally accessible

    to the cohort, including in regional, rural, and remote areas, and to culturally diverse communities

   and First Nations peoples; that the alternative support system is delivered by an appropriately

    qualified workforce; and that the comparable outcomes evidence supports the substitution.

b. Proposed section 25B should be amended to include an individual exception mechanism, by which

   a person whose clinical presentation falls outside the categorisation underpinning the Rule may

   apply for continued access to the Scheme. The provision should specify the criteria under which

   such an exception may be granted. The exception process should be accessible and supported, so

     that access to it does not depend on a family’s capacity to fund private reports or to navigate the

   system unaided.

 c. Proposed subsection 25B(4) should be amended to require consultation with peak professional bodies

   and clinical experts in the relevant intervention area before a Rule is made.

d. Rules made under proposed subsection 25B(4) should be subject to disallowance under the

    Legislation Act 2003, and should be supported by a published evidence statement explaining the

    basis on which the Minister is satisfied of the matters prescribed under amended subsection 25B(6).

 e. Proposed item 101 (application provision in relation to revocation of participant status) should be

  amended to provide transitional arrangements for existing participants, including: a minimum

Submission 366

• notice period before revocation takes effect; continuity of supports during the transition period;

• a supported, accessible right of review prior to revocation that does not rely on families to initiate and resource a challenge unaided.

3. Schedule 1, Part 6, item 68 (proposed section

33(2EA)): maximum intensity determinations

Illustrative case example (proposed s 33(2EA)): a cap overriding assessed need

Background. Eli is four. He has autism with co-occurring intellectual disability. He does not speak, needs support for every daily routine, and engages in self-injurious behaviour — head- banging and biting — when he cannot make himself understood. His Certified Behaviour Analyst has assessed him for a comprehensive early intensive behavioural intervention program: a cordinated, high-intensity model targeting communication, daily living, and safety during the developmental window in which these gains are most achievable. In its first months his self-injury has begun to reduce and he has started to use a small number of exchange-based requests.

The problem under the proposed legislation. Under proposed subsection 33(2EA), a determination may set a maximum intensity for a class of supports, and the Explanatory Memorandum contemplates a ceiling as low as 12 hours per year. For a child with Eli’s profile, evidence-based intervention is comprehensive and delivered at high intensity — this is the model the research base supports for this presentation, and the model his clinicians have, on individual assessment, identified as necessary for him. A statutory ceiling of that order would place the evidence-based standard of care beyond what may be funded, making the very intervention the evidence supports impossible to deliver. The cap would itself be derived from group-level averages drawn from a literature in which children with Eli’s profile are among the least represented — foreclosing the recognised, evidence-based model for the very cohort the underlying evidence describes least well.

The human consequence. The window for this intervention is now. A cap that forecloses comprehensive early intervention at four does not defer Eli’s progress; it forgoes gains in communication and safety that may not be recoverable later, and leaves his family managing escalating self-injury without the support that had begun to reduce it.

Submission 366

What the legislation should require instead. Where caps are introduced, the Bill should require

that the intensity available to an individual be set by clinical assessment of that individual, with

participant-specific data able to support an intensity above any population-level maximum

(Recommendations 6 and 7).

Illustrative, hypothetical example; not based on an identifiable individual.

3.1 The provision

Proposed subsection 33(2EA), inserted by Schedule 1, Part 6, item 68, permits a determination under

section 33(2E) to specify, for a support or class of supports, a maximum amount of funding, a maximum

intensity for provision of the support, or a maximum ratio of worker to participant.

The Explanatory Memorandum (page 50) gives two illustrative examples. The first is a maximum intensity

of “12 hours per year of a particular kind of therapy support” for a participant in a specified class. The

second is the worked example of Marco, an existing participant whose clinical reports recommend 30 hours

per discipline of allied health therapy for the next plan period. The example explains that, although the

planner determines 30 hours per discipline is reasonable and necessary, a support determination capping

therapy intensity at 25 hours per discipline operates to reduce the funding actually included in the plan to

25 hours.

The Explanatory Memorandum further states (page 50) that a determination of this kind would be made

where “research shows that any more than that intensity does not add any benefit for an individual”.

3.2 The category error in moving from group-level to individual-level

ABA Australia’s concern with proposed subsection 33(2EA)(b) is that it permits a statutory ceiling on the

intensity of an individual participant’s support to be set on the basis of group-level meta-analytic findings.

This involves a category error.

Group-level meta-analytic findings characterise average effects across populations. They describe whether, on average and across heterogeneous interventions and participants, additional intervention produces additional benefit. They do not characterise appropriate dosage for an individual participant. An individual participant’s clinical presentation, response to intervention, and goals are determined through functional clinical assessment, not by reference to meta-analytic effect sizes pooled across diverse populations. This

distinction is foundational to clinical practice across medicine and allied health.

Submission 366

The Explanatory Memorandum makes the category error explicit. At page 50, it states that “a determination could specify a maximum intensity for a particular therapy support if research shows that any more than that intensity does not add any benefit for an individual”. This sentence moves from a group-level claim about what research shows on average to an individual-level conclusion about what does not add benefit for an individual. The two are not equivalent. Group-level research findings about average effects across heterogeneous populations cannot, without further work, support conclusions about appropriate dosage for an individual participant whose clinical profile, response to intervention, and goals have been assessed by a qualified practitioner.

Where the group-level findings being relied upon are themselves contested in the peer-reviewed literature, the difficulty compounds. The meta-analysis by Sandbank and colleagues, published in JAMA Pediatrics in 2024, is the work most frequently cited in support of dosage ceilings for early behavioural intervention. That paper has attracted substantive published methodological critique, including a peer-reviewed rebuttal by Frazier and colleagues (2024) in the same journal demonstrating that the paper’s principal finding does not survive the inclusion of pre-intervention IQ as a covariate. In their published reply, Sandbank and Pustejovsky (2024) acknowledged that selective sampling due to IQ was an important concern they had not previously considered, and that their reproduced analyses confirmed significant positive dosage effects when this confound was addressed. They also recommended caution in the use of their findings to guide clinical practice. The body of evidence supporting a dose-response relationship for early intensive behavioural intervention, including the recent Eldevik et al. (2026) individual participant data meta-analysis and the 2025 Chetcuti et al. meta-analysis (which is co-authored by Australian researchers central to autism intervention policy), is substantial. ABA Australia’s submission to the EAC and our follow-up correspondence to the Therapies Research Branch set out these matters in greater detail.

A statutory ceiling that prevents a clinician from recommending, and a participant from receiving, a dosage that is clinically indicated by individual assessment is not justified by the current state of the evidence. The Explanatory Memorandum’s illustrative example of 12 hours per year would, for comprehensive early intensive behavioural intervention, render the intervention impossible to deliver. The evidence base does not support such a ceiling for the cohort of children for whom this intervention is clinically indicated.

Two further features of the provision warrant attention. First, a maximum intensity determination reverses the logic of early intervention. Sound practice begins from an individualised assessment of the intensity a child requires and adjusts it in response to data; a statutory ceiling instead begins from a cap and requires the case for anything above it to be argued against the determination. Second, once a maximum is set, there is a foreseeable risk that it will operate in practice as a default rather than a ceiling — that planning will gravitate towards the minimum intensity the determination permits, rather than the intensity the individual

Submission 366

Both features tend to detach funded intensity from assessed need, which is the opposite of the data-driven, individualised model that the evidence supports.

3.3 The disproportionate effect on underrepresented populations

Chetcuti et al.’s (2025) meta-analysis identified publication bias in studies of participants with high support needs, alongside publication bias in studies of naturalistic developmental behavioural interventions. The implication of the first finding is significant: the children most likely to require higher-intensity intervention are precisely those most underrepresented in the published evidence base on which generalisable findings rely. Applying caps derived from that evidence base to children whose profiles are underrepresented in it is not equivalent to applying them to the cohort the evidence describes. It is the application of a population- t level estimate to a population for whom the estimate is least reliable.

This concern applies more broadly than to autism. Women and girls, adults, those with high support needs, on-speaking individuals, culturally diverse communities, and First Nations peoples are systematically underrepresented across the disability intervention trial literature. Where statutory ceilings are derived from a literature that does not adequately represent these populations, the ceilings will most heavily affect those least well served by the underlying evidence.

3.4 The relationship between proposed subsection 33(2EA) and proposed

subsection 34(1)

The Explanatory Memorandum notes (page 50) that proposed subsections 33(2E), (2EA) and (2EB) do not affect the duty of the CEO (the Chief Executive Officer of the National Disability Insurance Agency, who is the statutory decision-maker under the Act) under section 34(1) to be satisfied that supports are reasonable and necessary. The Memorandum states that, once a support is considered reasonable and necessary, the determination operates only to constrain the amount of funding that may be included.

This framing understates the consequence of the provision. As the Marco example makes clear, where the planner has determined that 30 hours per discipline is reasonable and necessary, the support determination still operates to reduce the funded amount to 25 hours. The reasonable and necessary determination is not overridden in form, but it is overridden in effect. The participant receives less intervention than the planner has assessed they reasonably and necessarily require.

This consequence falls most heavily on participants with the most significant needs. The participants whose clinically assessed reasonable and necessary supports most exceed the cap are, by definition, those whose

Submission 366

clinical presentation requires the greatest intensity of intervention. They bear the largest absolute and

proportional reduction in their funded supports. The cap therefore operates regressively across the

participant cohort. As section 3.3 of this submission identifies, the participants most likely to require higher- intensity intervention are also those most underrepresented in the generalisable trial literature on which the

cap will be derived. The combined effect is that the cap is set on the basis of evidence that least describes this cohort, and applied in a way that most disadvantages them. For these participants, the gap between assessed need and funded support represents not an administrative inefficiency but a foreseeable harm.

Recommendations

ABA Australia recommends that:

 a. Proposed paragraph 33(2EA)(b) be amended to require, before any determination is made specifying
   a maximum intensity for a support, that the Minister be satisfied of prescribed matters including:
     that the evidence base relied upon distinguishes between interventions of materially different
    intervention class, intensity, and population; that the evidence base is not subject to substantive
    published methodological critique that has not been adequately addressed; and that the proposed
  maximum is consistent with the dosage range supported by the evidence for the cohort affected.

b. Proposed paragraph 33(2EA)(b) be amended to provide a mechanism by which an individual
    participant whose clinical assessment supports an intensity above the determined maximum may
   apply for an exception, with the determination of that application to be made on the basis of
    individualised clinical assessment by qualified practitioners.

 c. The Explanatory Memorandum’s worked example of Marco, which makes plain that the provision
    will operate to reduce the funding actually included in a plan below the level assessed as reasonable
   and necessary, be reconsidered in light of the policy intent stated at section 33(2EB), Note 2.

                                                                      14

Submission 366

Schedule 1, Part 6, Item 73: Evidence Hierarchy

Proposed Sections 34(1E) And (1F)

Illustrative Case Example (proposed ss 34(1E)-(1F)): Individual Evidence Outweighed By The Hierarchy

Background: Jordan is nine. He is autistic and does not speak. After many months of work with his speech pathologist and his behaviour analyst, he uses a speech-generating device to ask for help, for food, and for a break. Before that, when he was overwhelmed or in pain he would hit his head — hard enough, and often enough, to require medical attention. Twelve months of data across home and school show both supports working; his clinicians recommend continuing them to consolidate his gains.

The Problem Under The Proposed Legislation: Children like Jordan — non-speaking, high support needs, living regionally — are rarely represented in the generalisable research that proposed subsection 34(1E) ranks first. Under subsection 34(1F), the decision-maker could find his supports not effective and beneficial because that generalisable evidence is limited, even though the evidence specific to Jordan shows that they are working.

The Human Consequence: Jordan’s mother has not had to take him to a doctor for a self-inflicted injury in almost a year. If his supports are reduced, his clinicians believe he is at real risk of losing words — and with them, the means to show that something is wrong without injuring himself or others.

What The Legislation Should Require Instead: The Bill should require all available evidence to be weighed together, with the weight of each determined by its relevance to the individual participant (Recommendations 8 And 9).

Illustrative, Hypothetical Example: Not Based On An Identifiable Individual.

4.1 The Provision

Proposed Subsection 34(1E), inserted By Schedule 1, Part 6, Item 73, requires the CEO, when deciding whether A Support Will Be Or Is Likely To Be Effective And Beneficial For A Participant, to consider four classes Of Evidence “in the following order Of importance”:

• research and evidence In relation to the support That is published, peer reviewed and Generalisable;

Submission 366

b. evidence as to the effectiveness of the support, having regard to the participant’s circumstances

(including age and impairment);

c. evidence as to outcomes for the participant, arising from their use of the support in their previous

plan; and

d. other matters the CEO considers appropriate.

Proposed subsection 34(1F) provides that the CEO may decide that the CEO is not satisfied that a support is effective and beneficial if there is limited or no evidence of the kind mentioned in paragraph (1E)(a), even if there is evidence of the kind mentioned in paragraphs (1E)(b) or (c). The same applies where there is limited or no evidence of the kind mentioned in paragraph (1E)(b), even if there is evidence of the kind mentioned in paragraph (1E)(a) or (c).

4.2 The inversion of clinical practice

ABA Australia’s concern is that proposed subsections 34(1E) and (1F) embed in statute a model of evidence appraisal that inverts standard clinical care and is inconsistent with established principles of evidence-based practice.

Evidence-based practice, properly understood, integrates the best available research evidence, clinical expertise, and the values, preferences, and circumstances of the individual being supported. The three components are not hierarchically ordered. They are integrated. Generalisable research evidence informs clinical practice, but it does not substitute for it. The application of generalisable evidence to an individual participant requires clinical judgment about whether the participant’s circumstances fall within the population to which the evidence applies, whether the goals supported by the evidence are the goals appropriate for the participant, and whether the intervention as delivered for the participant matches the intervention for which the evidence was developed. These are clinical determinations.

The ordering in proposed subsection 34(1E), reinforced by the empowering provision in subsection 34(1F), substitutes a hierarchy for an integration. It permits a CEO to decide that a support is not effective and beneficial for a participant on the basis that there is limited generalisable research, even where individualised evidence demonstrates that the support is working for that participant. It also permits the CEO to disregard prior outcomes evidence for the participant where the generalisable research base is limited. Both outcomes are inconsistent with the way clinical decisions about individuals are made in practice.

Submission 366

This concern is not abstract. Many of the participants ABA Australia’s members support belong to

populations underrepresented in the generalisable trial literature, including adults with significant

developmental and intellectual disability, women and girls, non-speaking individuals, those with high

support needs, culturally diverse communities, and First Nations peoples. For these participants, the most

relevant evidence is often the participant-specific evidence captured in paragraphs (1E)(b) and (c). The

proposed hierarchy systematically disadvantages these participants by subordinating the evidence most

relevant to them.

The hierarchy also operates to exclude a category of evidence that is central to behaviour analytic practice. Single-case experimental designs — rigorous, replicable methods that establish experimental control at the level of the individual, and a recognised and long-established methodology in the field — sit uneasily within a hierarchy that ranks “published, peer reviewed and generalisable” research first. Although single-case designs are published and peer reviewed, their purpose is to demonstrate effects for individuals rather than to generalise across a population, and they are therefore liable to be treated as subordinate evidence under proposed paragraph 34(1E)(a). This matters because, for many of the participants ABA Australia’s members support, single-case data and the participant’s own prior outcomes are the most rigorous and most relevant evidence available.

4.3 The interaction with the dosage ceiling provision

Proposed subsections 34(1E) and (1F) interact with proposed subsection 33(2EA)(b) in ways that amplify the concern identified at section 3 of this submission. Where a maximum intensity determination is made on the basis of generalisable research, and where the CEO is empowered under subsection 34(1F) to discount participant-specific evidence in deciding whether a support is effective and beneficial, the practical effect is that the system is structurally tilted away from individualised clinical decision-making. A participant whose clinician has determined, on the basis of functional assessment and prior response to intervention, that an intensity above the statutory maximum is clinically indicated, faces a CEO empowered to disregard precisely the evidence base on which that determination was made.

4.4 Recommendations

ABA Australia recommends that:

 a. Proposed subsection 34(1E) be amended to remove the requirement that the matters listed be
    considered “in the following order of importance”. The four classes of evidence should be required

Submission 366

to be considered together, with their relative weight determined by the relevance of each to the

individual participant’s circumstances.

b. Proposed subsection 34(1F) be amended to remove the provision permitting the CEO to find a

support not effective and beneficial on the basis of limited generalisable research alone, where there

is evidence under paragraphs (1E)(b) or (1E)(c) supporting the effectiveness of the support for the

participant.

c. Proposed subsection 34(1E) should be amended to make clear that single-case experimental designs

are a recognised form of high-quality evidence and are not, by reason of their individual-level focus,

to be treated as subordinate to research of the kind described in paragraph (1E)(a).

d. In the alternative, if the structure of subsection 34(1E) is retained, the section should be amended to

require that the CEO take into account whether the participant belongs to a population

underrepresented in the generalisable trial literature, and whether participant-specific evidence is

the most relevant evidence available in the circumstances.

5. Schedule 1, Part 6, item 73 (proposed sections

34(1G) and (1H)): the parental responsibility

presumption and “behavioural support”

Illustrative case example (proposed ss 34(1G)–(1H)): professional behaviour support treated

as parenting

Background. Noah is ten. He is autistic, non-speaking, and engages in severe behaviours of concern

including aggression, property destruction, and absconding. A behaviour support practitioner has

completed a functional assessment and developed a positive behaviour support plan: structured

prevention strategies, environmental adjustments, explicit skill teaching, staff training, and ongoing

data monitoring across home and school. Early data shows fewer incidents and improved safety.

The problem under the proposed legislation. Proposed subsections 34(1G) and (1H) presume that

parents are responsible for “behavioural support” as part of ordinary care. Read literally, this could

treat Noah’s regulated, professional positive behaviour support as something a parent of any child

of similar age would be expected to provide, and so justify declining to fund it.

Submission 366

The human consequence.

Without skilled support, the foreseeable path for Noah’s family is escalation: increased harm, an unstable school placement, and risk of family breakdown.

What the legislation should require instead.

The Bill should clearly distinguish ordinary parental management of behaviour from professional behaviour support delivered by a qualified practitioner under a positive behaviour support plan (Recommendation 11).

Illustrative, hypothetical example; not based on an identifiable individual.

5.1 The provision Proposed subsection 34(1G), inserted by Schedule 1, Part 6, item 73, provides that, where a participant is a child, the CEO must take into account “the presumption that parents are responsible for providing substantial care and support for their children”. Proposed subsection 34(1H) provides that substantial care and support includes “supervision, personal care, transport, emotional support and behavioural support, and other assistance with the activities of daily living that, regardless of the child’s disability, would reasonably be expected of a parent of a child of a similar age”

5.2 The conflation of professional behaviour support with ordinary parental responsibility ABA Australia’s concern with proposed subsection 34(1H) is that it lists “behavioural support” alongside personal care, transport, supervision, and emotional support as an aspect of substantial care “that, regardless of the child’s disability, would reasonably be expected of a parent of a child of a similar age”. The term “behavioural support” has a specific and well-established meaning in the disability sector and in NDIS practice. It refers to professional services delivered by qualified practitioners, typically under a positive behaviour support plan, addressing behaviours of concern through functional assessment, evidence-based intervention, and structured monitoring. Behaviour support is a regulated practice area under the NDIS Quality and Safeguards Framework. The inclusion of “behavioural support” in a list of activities that are presumed to be within the scope of ordinary parental responsibility conflates two materially different concepts.

The everyday management of a child’s behaviour by a parent is one thing. Professional behaviour support, addressing behaviours that present risks to the child or to others and that require specialist intervention, is another. The two are not interchangeable. A parent of a child of similar age, without disability, would not

Submission 366

be expected to deliver professional positive behaviour support. The premise of the presumption, that the

activity would be expected of any parent regardless of disability, does not hold for this category of support.

The practical consequence of the provision, as drafted, is that the CEO may apply the parental responsibility

presumption to decline funding for professional behaviour support on the basis that “behavioural support”

is within the scope of ordinary parental responsibility under subsection 34(1H). This would represent a

significant narrowing of access to a category of support that is currently funded for children with significant

developmental and behavioural support needs, and that is delivered under regulated professional standards

by qualified practitioners.

Two consequences follow. First, the provision shifts responsibility for a regulated, professional service onto

parents. Positive behaviour support requires functional assessment, plan design, staff training, and ongoing

data-based adjustment by a qualified practitioner; it is not a task that can be transferred to a parent by

presumption, however capable and committed that parent is. Second, the burden of the presumption falls

unequally. Families with greater resources may be able to purchase support privately or absorb the work

themselves; families with fewer resources, or those managing several children, or their own disability or

ill-health, cannot. A presumption that treats professional behaviour support as ordinary parenting therefore

risks both displacing clinical responsibility onto families and widening inequity between them.

Recommendation

ABA Australia recommends that proposed subsection 34(1H) be amended to remove the reference to

“behavioural support”, or alternatively to specify that the reference is to ordinary parental management of

behaviour and does not extend to professional behaviour support services delivered by a qualified

practitioner under a positive behaviour support plan.

6. The support needs of children with significant

developmental disability

The provisions discussed in sections 2 to 5 of this submission are likely, in combination, to have their most

consequential effect on a specific cohort: children with significant developmental disability, including those

with intellectual disability, those with autism whose presentation falls within what is internationally

described as profound autism (Lord et al., 2022), and children under the age at which that classification is

typically applied who present with co-occurring intellectual disability and autism. ABA Australia considers

Submission 366

It important that the Committee have an accurate understanding of the support needs of these children before considering the Bill’s provisions in their final form.

6.1 The reality of support needs in this cohort

The needs of these children are very substantial. They may have limited or no functional communication and require structured, evidence-based intervention to develop alternative means of expressing needs and preferences. They may engage in behaviours that present serious safety risks to themselves or others, including self-injury, aggression, pica and absconding. They may require continuous supervision during waking hours to maintain safety, with disrupted sleep patterns that mean supervision continues through the night. They commonly require coordinated support across communication, adaptive functioning, learning readiness, self-care, participation, and behavioural domains, delivered with sufficient intensity over time to produce meaningful change.

The intervention required for these children is not generic. It is evidence-informed behaviour analytic early intervention, designed and overseen by qualified clinicians, delivered with appropriate intensity, individualised through comprehensive assessment, and adjusted on the basis of ongoing data and feedback.

ABA Australia’s submission to the EAC sets out the evidence base for this intervention in detail. The most rigorous recent syntheses, including the Eldevik et al. (2026) individual participant data meta-analysis, support the conclusion that this intervention produces clinically meaningful improvements in adaptive behaviour, cognitive functioning, communication, and reduction of behaviours that interfere with learning and safety. The strongest and most consistent effects are observed where the intervention is comprehensive, individualised, and delivered with sufficient intensity to address the child’s developmental profile.

A further feature of this cohort compounds the difficulty. Children with significant developmental disability frequently present with co-occurring conditions — autism with intellectual disability, with co-occurring physical or sensory impairment, or with complex medical needs. Multi-diagnostic, high-complexity profiles of this kind are rarely reflected in the generalisable trial literature, which tends to study more narrowly defined and less complex groups. Read against the evidence hierarchy in proposed subsection 34(1E), the consequence is structural: the children whose needs are greatest are also those least likely to be represented in the evidence ranked most important, and are therefore the most likely to fail the threshold the hierarchy creates.

Submission 366

The underestimation of these needs in policy framings

Public policy discussion of NDIS reform frequently underestimates the substance of these support needs.

This is understandable. The presentations are not widely visible in everyday community settings, in part because the support needs themselves limit the family’s capacity to participate in those settings. The result is that policy framings often default to a generalised picture of childhood disability that does not capture the reality faced by these families.

ABA Australia raises this concern not to dispute the policy intent of the reforms, but to ensure that the legislative framework is calibrated against the actual needs of the cohort it will most heavily affect.

Legislation drafted on the basis of a generalised picture of disability will operate differently when applied to individuals whose support needs sit at the upper end of complexity and intensity.

The cumulative effect of the proposed amendments on this cohort

The provisions discussed in sections 2 to 5 of this submission affect this cohort in combination. The section 25B mechanism creates the possibility of these children being categorised, by impairment class, as appropriate for foundational supports rather than the NDIS. The maximum intensity provision in section 33(2EA)(b) creates the possibility of statutory ceilings that would prevent the intensity of intervention clinically indicated for these children from being funded. The evidence hierarchy in sections 34(1E) and (1F) creates the possibility that participant-specific evidence, which is often the most relevant evidence for children in this cohort given their underrepresentation in generalisable trial literature, will be subordinated to evidence less applicable to their presentation. The parental responsibility presumption in subsection 34(1H) creates the possibility that professional behaviour support, required to address behaviours presenting significant safety risks, will be treated as within the scope of ordinary parental responsibility.

Each provision, considered alone, may operate in cases where the effect on individual participants is modest. Considered together, and applied to the cohort whose needs are most substantial, the cumulative effect is significant. The consequences include reduced access to evidence-informed behaviour analytic early intervention during the developmental window in which that intervention is most consequential; increased risk of harm to children whose behavioural support needs are unmet; isolation of families unable to access community, educational, or social settings due to the intensity of supervision required without funded support; foreseeable carer breakdown, with the long-term consequences for the child and the Scheme that follow; and the loss of opportunity to build the communication, adaptive, and self-regulation skills that reduce the intensity of support required later in life.

Submission 366

None of these consequences is consistent with the stated objects of the National Disability Insurance

Scheme Act. ABA Australia respectfully suggests that the Bill, as currently drafted, has not been calibrated

against the support needs of the cohort it will most heavily affect, and that the recommendations in this

submission should be considered with the needs of this cohort in mind.

7. The broader question of safeguards in the reformed

Scheme

The provisions discussed in sections 2 to 5 of this submission share a common feature. Each provides a

power, or establishes a structure, that operates at the population level. Section 25B excludes a class of

impairments. Subsection 33(2EA) caps intensity for a class of supports or participants. Subsections 34(1E)

and (1F) order classes of evidence. Subsection 34(1H) presumes a class of activities to fall within ordinary

parental responsibility.

Population-level provisions are appropriate in many contexts. The NDIS is a population-scale programme

and operates at scale by design. The concern is not with the existence of population-level provisions, but

with the absence of individualised safeguards within them. Where population-level rules are applied to

individuals whose clinical presentation falls outside the assumptions underpinning the rule, the individual

safeguard is the mechanism by which the system corrects for the limitations of the rule. Each of the

provisions discussed in this submission, as currently drafted, lacks adequate individualised safeguards.

There is a further dimension to this. The objects of the National Disability Insurance Scheme Act include

supporting participants to exercise choice and control in the pursuit of their goals, and the provision of

reasonable and necessary supports that are individualised to the participant. The provisions discussed in

this submission move, collectively, away from those commitments. An impairment excluded by class, an

intensity capped by population-level research, an evidence hierarchy that subordinates what is known about

the individual, and a presumption that reassigns professional support to families — each narrows the space

within which an individual participant’s assessed needs and choices determine the supports they receive.

The cumulative effect is a Scheme that decides more by category and less by the individual, which is the

opposite of the individualised, participant-centred model the reforms are said to advance.

The reformed Scheme will operate for many years and will determine the supports available to hundreds of

thousands of participants. The opportunity to insert appropriate safeguards at the point of legislation will

not return. ABA Australia respectfully suggests that the Committee give particular attention, in considering

the Bill, to whether the safeguards proposed in this submission, or others of equivalent effect, are necessary

Submission 366

to ensure that the reformed Scheme continues to deliver evidence-based, individualised intervention to the

participants who require it.

Summary of recommendations

In relation to Schedule 1, Part 9, item 97 (proposed section 25B):

• Amend proposed subsection 25B(6) to prescribe the matters the Minister must be satisfied of before making a Rule, including the capacity, accessibility, workforce qualifications, and outcomes evidence of the alternative support system.

• Amend proposed section 25B to include an individual exception mechanism for persons whose clinical presentation falls outside the categorisation underpinning a Rule.

• Amend proposed subsection 25B(4) to require consultation with peak professional bodies and clinical experts before a Rule is made.

• Require Rules made under proposed subsection 25B(4) to be subject to disallowance and supported by a published evidence statement.

• Amend item 101 to provide transitional arrangements, including notice periods, continuity of supports, and rights of review, for existing participants whose status would be revoked under proposed paragraph 30(1)(c).

In relation to Schedule 1, Part 6, item 68 (proposed section 33(2EA)):

• Amend proposed paragraph 33(2EA)(b) to prescribe the evidence-based matters the Minister must be satisfied of before any maximum intensity determination is made.

• Amend proposed paragraph 33(2EA)(b) to include an individual exception mechanism for participants whose clinical assessment supports an intensity above the determined maximum.

In relation to Schedule 1, Part 6, item 73 (proposed sections 34(1E) and (1F)):

• Amend proposed subsection 34(1E) to remove the hierarchical ordering of evidence classes, requiring instead that the classes be considered together with relative weight determined by relevance to the participant.

• Amend proposed subsection 34(1F) to remove the provision permitting the CEO to find a support not effective and beneficial on the basis of limited generalisable research alone where there is participant-specific evidence supporting effectiveness.

Submission 366

Amend proposed subsection 34(1E)

to recognise single-case experimental designs as a valid form of high-quality evidence not to be treated as subordinate by reason of their individual-level focus.

In relation to Schedule 1, Part 6, item 73 (proposed subsection 34(1H)):

Amend proposed subsection 34(1H)

to remove the reference to “behavioural support”, or to specify that the reference does not extend to professional behaviour support services delivered by a qualified practitioner.

Attachments

Attachment A: ABA Australia, Submission to the NDIS Evidence Advisory Committee on Early Intensive Behavioural Intervention, 27 March 2026.

References

Chetcuti, L., Uljarević, M., Schuck, R. K., Hardan, A. Y., Gengoux, G. W., Trembath, D., Vadgama, Y., Varcin, K. J., Vivanti, G., Whitehouse, A. J. O., Helton, M., & Frazier, T. W. (2025). Characterizing predictors of response to behavioral interventions for children with autism spectrum disorder: A meta-analytic approach. Clinical Psychology Review, 119, 102588. https://doi.org/10.1016/j.cpr.2025.102588

Eldevik, S., Strømgren, B., Eikeseth, S., Fields, A., Goetz, C. M., & Titlestad, K. B. (2026). Clinically significant outcomes of early intensive behavioral intervention for children with autism spectrum disorders: An individual participant data meta-analysis. Autism Research, 19, e70169. https://doi.org/10.1002/aur.70169

Frazier, T. W., Chetcuti, L., & Uljarević, M. (2024). Evidence that intervention dosage is associated with better outcomes in autism. JAMA Pediatrics, 179(1), 101–102. https://doi.org/10.1001/jamapediatrics.2024.4710

Lord, C., Charman, T., Havdahl, A., Carbone, P., Anagnostou, E., Boyd, B., Carr, T., de Vries, P. J., Dissanayake, C., Divan, G., Freitag, C. M., Gotelli, M. M., Kasari, C., Knapp, M., Mundy, P., Plank, A., Scahill, L., Servili, C., Shattuck, P., . . . McCauley, J. B. (2022). The Lancet Commission on the future of care and clinical research in autism. The Lancet, 399(10321), 271–334. https://doi.org/10.1016/S0140-6736(21)01541-5

Submission 366

Sandbank, M., & Pustejovsky, J. E. (2024). Intervention amount and outcomes for young autistic

children—Reply. JAMA Pediatrics, 179(1), 102–103.

Sandbank, M., Pustejovsky, J. E., Bottema-Beutel, K., Caldwell, N., Feldman, J. I., Crowley LaPoint, S.,

& Woynaroski, T. (2024). Determining associations between intervention amount and outcomes

for young autistic children: A meta-analysis. JAMA Pediatrics, 178(8), 763–773.

hhttps://doi.org/10.1001/jamapediatrics.2024.1832

ABA Australia thanks the Committee for the opportunity to make this submission and would welcome the opportunity to provide further information, including oral evidence, if that would assist the Committee’s consideration of the Bill.

Yours sincerely,

Niesha Illingworth President, Association for Behaviour Analysis Australia Psychologist and Behaviour Analyst (CBA/BCBA)