To paraphrase Charlie Munger: ‘show us the oversized government program and we will show you the scams and the rorts.’ The National Disability Insurance Scheme (NDIS) is on track to become the worst-designed Commonwealth government program in Australian public policy history. And that is saying something. The list of contenders includes the NBN, Snowy 2.0, RoboDebt, Pink Batts, School Halls, the Gonski reforms, Cash for Clunkers, and JobKeeper.

All badly designed. All economically destructive. All policy failures. And all sharing a common thread: they emerged in an era when Australian governments lost all restraint about the reach of their ambitions and all regard for whose money they were burning to fund them.

What unites them further is a near-complete absence of accountability for their architects. But why would there be accountability? It is other people’s money. The designers kept their generous pensions. Many were promoted into larger roles. Most have already received, and the rest will soon enough receive, their Order of Australia pins or Public Service Medals.

In Australian public office, failure is rarely a career setback. It is often a career strategy. Yet even among this crowded field, the NDIS stands apart. Not simply because of the scale of the program, but because its design makes reform extraordinarily difficult once the system begins expanding.

The scheme now covers roughly 760,000 Australians and rising. More than half of participants are under 18, and about 42 per cent are under 14. Consider the implications. These cohorts will age through every stage of life drawing on support funded by taxpayers. The long fiscal tail of the scheme is barely discussed in public debate, but it should be.

Defenders of the NDIS often respond that rising participation simply reflects long-overdue recognition of disability that earlier systems neglected. There is truth in that. Many participants clearly benefit from the support the scheme provides. The issue, however, is not whether the NDIS helps people. The issue is whether it directs finite public resources toward those with the most severe and permanent disabilities in a sustainable and well-governed way.

Increasingly, the answer appears to be no. Earlier this year the scheme’s fraud chief told Senate Estimates that up to 10 per cent of claims may involve inappropriate claiming, incorrect billing or outright criminal behaviour. That represents billions of dollars in potential waste each year. And this is not merely a theoretical concern.

In February the Australian Federal Police raided properties across Darwin following a months-long investigation into alleged fraud involving an employee of the National Disability Insurance Agency itself. The employee allegedly directed participants toward an NDIS provider business he co-owned. That business claimed more than $28 million in funding, with about $5 million flagged as suspicious.

This was not an isolated case. Investigations in the Northern Territory uncovered providers offering cash, alcohol and cigarettes to persuade vulnerable participants to sign service agreements, then claimed payments for services never delivered.

Such examples are often dismissed as enforcement failures. But when fraud appears repeatedly across thousands of providers in a rapidly expanding market, the explanation is unlikely to be enforcement alone. It suggests something deeper: incentives that make abuse unusually easy and unusually profitable.

Professor Steven Schwartz has identified a core mechanism. Over several decades the diagnostic criteria for autism have broadened dramatically. Children who once would have been considered severely disabled now share the same diagnostic category as those who will live largely independent lives.

The problem is not autism. The problem is that diagnosis has become the gateway to funding. Clinicians are paid to assess and certify. Families gain access to packages that average tens of thousands of dollars a year. Schools receive additional resources. Governments appear compassionate. The scheme is neither means tested nor subject to co-payments, meaning affluent families face exactly the same financial incentive to enter and remain in the scheme as those with no resources at all. With no price signal at the point of use and no personal cost to participation, every participant in the system faces incentives that point in the same direction: expansion.

Autism now accounts for about 43 per cent of all NDIS participants and more than $10 billion in annual spending. More than 62,000 people diagnosed with autism entered the scheme last year alone.

Some observers argue that this reflects better detection of previously unmet need. That is partly correct. But it does not explain why diagnostic expansion has occurred precisely in the conditions where clinical thresholds are most ambiguous and where the financial incentives to diagnose are strongest. Meanwhile the growth rate of participants with severe conditions such as psychosis or profound intellectual disability has remained stable or declined.

The deeper issue is that the NDIS rewards persistence rather than independence. A provider whose client becomes independent loses revenue. A family whose child improves leaves the scheme. Success reduces funding; continued eligibility sustains it.

The culture surrounding the program makes this dynamic visible. NDIS business conferences advertise strategies to ‘grow your provider business’ and ‘build powerful referral networks’. These events are filled with allied health providers, plan managers and consultants discussing how to expand their share of a $50-billion-a-year-and-growing market.

The NDIS was never supposed to be a growth industry. It was meant to be a support system for Australians with severe and permanent disabilities.

Advocates sometimes argue that the scheme resembles Medicare or other social insurance systems and therefore should not be expected to operate with strict financial limits. But that comparison misses the crucial design difference. Medicare pays for services used episodically by the entire population. The NDIS provides lifelong individualised budgets to a rapidly expanding subset of participants. Without tight functional eligibility and strong incentives for independence, demand becomes effectively open-ended.

This is the structural problem. The scheme combines individualised funding, porous eligibility boundaries and a rapidly expanding provider market while offering very few incentives for participants to transition off the program.

None of this means disability support should be reduced or abandoned. On the contrary, the principle behind the NDIS remains widely supported. The real challenge is designing a system that directs resources to those with the most severe functional impairments while supporting others through broader community and education programs rather than lifelong individualised packages.

Until that distinction is restored, the scheme will continue expanding faster than governments can finance it.

Unless the underlying design changes, the NDIS will continue growing until either the budget breaks or the public support that sustains the scheme collapses. Neither outcome would serve the people the system was designed to help.