The fundamental challenge of the SIL market is a structural one, not a matter of regulatory oversight or provider goodwill. A single government buyer procuring a complex human service from a highly fragmented supply base cannot effectively govern either quality or price. The reform trajectory, starting with standardised needs assessment and culminating in formal commissioning, is designed to correct this by consolidating the market into a panel of providers the Agency can hold accountable.
The legislation currently before Parliament standardises how SIL packages are assessed and funded. On its own, that is only the first half of the reform, and in isolation a counterproductive one. The second, complementary half is the move to a commissioning model, a policy direction the Government will begin consulting on in July 2026. Commissioning will deliberately consolidate the market. The analysis that follows is a prediction of the logic and shape of that consolidation, grounded in the same market dynamics that saw Australian employment services contract from over 300 providers to 43.
The SIL market is defined by three critical figures. The NDIA purchases approximately $16 billion in supports each year for close to 37,000 participants from a supply base of many thousands of registered providers. The largest of these commands under 4% of national market share, and the top ten combined account for barely one-tenth of the total.
This level of fragmentation for a $16 billion procurement is highly unusual. It is the direct cause of two persistent and fundamental problems: the Agency's inability to differentiate the quality of support from one provider to another, and its inability to control the escalating price of that support. In economic terms, SIL exhibits the characteristics of both a market for lemons, where quality is unobservable and bad providers drive out good ones, and a common goods problem, where every party involved in setting a participant's funding level has an incentive to push it higher.
The proposition that a government buyer would deliberately consolidate a human-services market has a close and well-documented precedent. Federal employment services underwent an 86% reduction in provider numbers, from 300 organisations under the Job Network to just 43 under Workforce Australia.
The contraction was driven by a single, powerful instrument: the requirement for providers to service entire geographic regions, bundling high-cost and low-cost clients into a single obligation. Critically, while ownership concentrated, the physical delivery footprint remained remarkably stable. The lesson is that the corporate roof above a service can change without disrupting the service itself, a process already common in the SIL sector through mergers, acquisitions, and the orderly transfer of participants and staff when a provider exits.
While SIL is a place-bound service, this does not break the precedent. Consolidation does not require moving a participant; it requires moving the corporate entity that holds the lease and employs the staff. The key design constraint this imposes is that every provider must remain small enough to be safely removed and its participants rehomed by the Agency. This "removability" is both the Agency's core lever and the essential safety mechanism of the entire structure.
The quality problem is a direct consequence of information asymmetry. The things that matter most in a supported home (participant safety, dignity, and progress) are slow to materialise, hard to observe from outside the house, and difficult to attribute solely to a provider. Because the price is fixed, a provider that quietly under-services captures a higher margin than one that does not. In a fragmented market with a weak regulator, the honest provider is systematically disadvantaged.
The cost problem is driven by misaligned incentives in the planning process. The participant, their family, the provider, and the clinicians who shape the plan all have a reason to push the funding level up, and none of them bears the fiscal cost of that decision. This creates a structural bias toward escalation. The result is visible in the funding data: a fifth of participants consume close to half of all core daily-activities funding.
These two faults compound. A rational business model in the current market is to selectively recruit participants funded above their true cost of care and to escalate stated needs. The market structurally cannot reward a provider for moving a participant toward independence, because independence shrinks the package. It rewards the opposite.
The most damaging feature of the present structure is that it has made working against a participant's independence the commercially dominant strategy.
A clear understanding of how SIL funding is distributed is essential, as it reveals exactly which cohort the standardisation reforms will impact most heavily. Our analysis indicates that SIL funding is not a single population but two distinct groups combined into one distribution.
The majority of participants reside in shared, lower-ratio arrangements. A high-needs minority, however, are on one-to-one support and above, with annual packages starting around $650,000 and averaging closer to $850,000. This gap is too wide for a single smooth curve. As a bottom-up cross-check, a Roster of Care for a standard three-person shared house, priced at NDIS unit rates, produces a package of approximately $168,500 per participant per annum. The average core daily-activities package, however, is around $347,000. The difference between the two, measured from both the top-down distribution and the bottom-up roster, is the escalation margin that standardisation aims to close.
The National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Bill 2026, while framed as an integrity and sustainability measure, functions as the engine for this structural shift. Its practical effect is to fix each participant's plan within a reference range that cannot easily be escalated.
The assessment process is centralised within the Agency s32L(4A). While the validity literature for the likely assessment instrument, the I-CAN, is thin, the point is not clinical precision but standardisation. With every participant assessed against the same reference, packages can be normed.
On its own, in an open market, this first half of the reform is a trap. It will not succeed; it will create the crisis that necessitates the second half.
Introducing standardised, compressed funding into a market with no provider obligations will not cut spend. It will strand participants. Compressing a package pushes a large cohort of individuals below the threshold at which a provider finds them commercially viable. With no obligation to serve, the rational provider response is to decline these unprofitable referrals.
The only release valve for the Agency in an open market is to raise the package back up until a provider is willing to accept the participant. The standardised number, therefore, does not hold. It undoes its own saving, while participants who are accepted at the compressed number risk being quietly under-serviced. This is the trap that makes the second half of the reform, commissioning, a structural necessity.
Commissioning changes the unit of procurement from an individual participant to a bundled regional slot. A provider wins a slot through competitive tender, and that slot is an obligation to serve every participant in a given region at the standardised allocation. Profitable and unprofitable participants are bundled into a single, indivisible obligation.
This is a universal-service obligation, a standard regulatory tool for solving stranding problems. The tender is the lever that binds the two together. Only inside this bundle can the standardised assessment deliver its fiscal saving, because the provider is now obliged to serve the stranded cohort at that rate.
This structure transforms the provider's relationship with the buyer from a series of one-shot transactions into a repeated game across a national portfolio. The NDIA is a monopsony, the only buyer the provider will ever have. A defection in one region, under-servicing for instance, is visible to the buyer and can be punished across all other regions and in every future tender.
This echoes the economic logic of Bernheim and Whinston's multimarket contact theory, where the threat of aggregate punishment makes defection irrational. The fraud and quality problems in SIL have been misdiagnosed as enforcement failures. They are, in fact, problems of game structure. Converting a one-shot game into a repeated, portfolio-based one makes defection career-ending.
The real cost is a contraction of genuine participant choice, from a notional pool of thousands of providers to the few who hold slots in a given region. This is the trade-off: making the slot a meaningful obligation requires limiting the alternatives.
A repeated game disciplines a provider only if the buyer can see, compel, and punish. The bill provides the Agency with these powers. It establishes the Agency's own monitoring, investigation, and search-and-seizure powers (Part 3C of Ch 4), the power to compel answers (ss54, 53(3)), and a reduced claim window of 90 days (s45A). Crucially, it also enables automated, evaluative decision-making at scale (ss59B to 59E).
This is not an incremental regulatory tightening. It is a shift from policing three thousand ungovernable strangers to managing a panel of twenty to thirty known counterparties, where the close, case-by-case management that hard cases require becomes feasible for the first time.
If consolidation restores control, the logical extreme would be a market of just five or so mega-providers. This model, however, fails on the principle of removability. A provider holding several thousand participants becomes too large to be removed without stranding an unabsorbable number of vulnerable people. The Agency, which carries the ultimate safety obligation, would be unable to act.
The optimal provider is big enough to have everything to lose and small enough that the Agency can still take it.
The defensible level of consolidation sits in the space between a fragmented market that cannot be governed and a concentrated one that cannot be disciplined. This level can be derived from the Agency's own data.
Our model determines the fewest providers consistent with keeping each one removable, using the NDIA's own 80 service districts as the geographic unit. In the first step, each district is assigned a minimum number of provider slots based on two competing limits: a viability floor of roughly a three-house cluster, and a removability cap of about 10% of a district's participants. The second step allocates these slots to providers based on their capacity for geographic reach.
The total number of survivors is unstable, ranging from several hundred local operators to around forty national ones depending on the reach parameter. However, the number holding a genuinely substantial book, the "meaningful set", is remarkably stable, sitting at twenty to thirty across the entire plausible range.
The political impediment is smaller than it first appears. The Agency does not need to actively close providers. By controlling new participant allocation, and freezing it to targeted providers, it can let ordinary attrition bleed a provider's occupancy until it is no longer viable. This is consolidation by inaction, achievable even by a low-capability bureaucracy, and it is already occurring in the therapy market through the Thriving Kids reforms.
Furthermore, the constituency that usually defends incumbents is remarkably weak here. The providers facing consolidation are disproportionately small operators with little electoral weight. After the Royal Commission, the public is more likely to read a provider closure as the system cleaning itself up.
For the providers who will survive, this reform is not a punishment. It addresses the single largest destroyer of viability: vacancy. In a supported house, labour costs are fixed to the roster, while revenue is per-participant. A single vacancy can erase the entire margin on a thin 5% profit.
Commissioning solves the vacancy problem by directing referral flow. If the Agency allocates participants to the providers holding slots, the structurally guaranteed occupancy that results is worth more than any margin improvement a provider could extract from roster optimisation alone. It converts an unstable, uncontrollable cost into a predictable one.
If you operate above the viability threshold and below the removability cap, commissioning is designed to reward you. The providers who can demonstrate, with data, that they deliver compliant, efficient, and participant-centred services will be the ones that win slots.
The providers who will lose are those that cannot account for what they deliver, those that have relied on funding escalation rather than operational discipline, and those that are too small to meet the obligations of a regional slot.
This is not a prediction about policy timing. It is an analysis of structural inevitability. The two halves of this reform, standardisation and commissioning, are complementary and sequential. The first creates the problem that only the second can solve. The question is not whether this will happen, but whether you will be positioned when it does.
Each section reference below is drawn from the NDIS Amendment (Securing the NDIS for Future Generations) Bill 2026.
Occasional, considered analysis on SIL economics, NDIS reform, and provider viability. No noise, and you can unsubscribe anytime.