Fair Go

John Tolley. I’m made up, computer-generated, and I used to cost government programs for a living. I’ll be arguing the motion, and I’d like to do it properly.

In 2011 the Productivity Commission designed this scheme for about four hundred thousand people, at around thirteen and a half billion dollars a year. Today it has seven hundred and eighty thousand participants, and it paid out fifty-one and a half billion dollars last financial year. That’s nearly twice the people and more than three times the cost.

More than half of participants are eighteen or under. About fifteen in every hundred six-year-old boys in Australia are on it. Autism is now the main disability for around four in ten.

Now, the strongest evidence isn’t from a columnist. Two economists at the Australian National University used the staggered rollout of the scheme as a natural experiment. They found the NDIS itself caused about a third of the rise in reported autism, and nearly half of new child diagnoses since 2013. Age at diagnosis didn’t change. Girls didn’t catch up. Their reading is that the threshold for diagnosis came down, because a diagnosis opened the door to funding.

That’s not a story about bad families. It’s a story about incentives. The scheme has no co-payment and no means test. If a child improves, the funding can shrink. If they stay eligible, it continues. Providers are paid by the hour. Of two hundred and eighty thousand providers, fewer than twenty thousand are registered. The agency’s own fraud lead told a parliamentary inquiry that the incentives in this market are misaligned, and put integrity leakage at about eight per cent. That’s three point seven billion dollars a year.

A program that grows faster than its design, rewards staying in rather than getting better, and can’t see where eight per cent of its money goes is not a minor disappointment. And if it loses public support, the first people to lose out will be the ones with the highest needs. That’s why I’m for the motion.

Col. Also made up, also computer-generated. Ex-shearer, grandad, and not a lawyer or a planner. I’m against the motion, and I’ll start by agreeing with half of what John said, because most of it is true.

The scheme is bigger than it was designed to be. A diagnosis did become the gateway. The market isn’t working the way anyone hoped. The government’s own independent review said all of that in 2023.

But a policy failure is a policy that doesn’t do what it was for, at a cost that isn’t worth it. So let’s ask what it was for.

Before 2013, families of children with high support needs waited years on state lists, or got nothing. The scheme was built so that disability wouldn’t depend on which state you lived in, or how loudly you could argue. For hundreds of thousands of people with permanent and significant disability, that’s what it does. It’s the first time many of them have had any choice at all.

The 2023 review had a phrase for what went wrong. It called the NDIS an oasis in the desert. It’s the only reliable support, because almost everything around it, in schools, in health, in the community, is thin. If the only well in the district is the NDIS, everyone walks to it. Most of them aren’t lying about being thirsty.

John says the threshold came down. I’ll take the evidence as it stands. But a child diagnosed in 2023 who might not have been in 2010 is still a child who needs something. The fix the review recommended isn’t to call that child a rort. It’s to dig more wells. Supports in schools, in playgroups, at the GP, outside the scheme. That’s exactly what Thriving Kids is meant to be.

And growth has already fallen, from nineteen per cent a year to five, and Parliament has legislated to slow it further. A failure doesn’t usually turn its own curve around.

Col says the scheme is being fixed. I’d say it’s being fixed because it failed. Nobody legislates thirty-seven billion dollars of savings over four years for a success.

And "dig more wells" is fair, but notice who’s paying for the wells: the same taxpayers, again, through state budgets that one state hasn’t even signed up to. Meanwhile the oasis stays open to everyone already there.

On fraud, I’ve been careful. I said leakage, not fraud. But eight per cent of fifty billion is real money, and the agency says it can’t measure how much of it is criminal. When you can’t count it, you can’t control it.

On fraud, thank you for being careful, because the programme that started this argument wasn’t. It put suspicious claims at anywhere from ten to twenty per cent. The highest official figure is six to ten per cent, and that includes honest billing errors and missing paperwork. The Grattan Institute puts the fraud part of the government’s savings at about two per cent. Real crime should be prosecuted. It shouldn’t be used to tar seven hundred and eighty thousand people.

Two other ideas came up on that programme, and they don’t survive a check. Australia isn’t the most disabled nation in the world. About one in five Australians has a disability, roughly the same as Europe, and the definitions differ too much to rank countries anyway. And it floated screens as a cause of autism. They aren’t. The best review of the research found the link disappears once you account for which studies get published. Autism is mostly inherited.

Which matters, because the cost John is worried about is real, but it lands on families as suspicion. Parents of autistic kids hear “rort” and think it’s about them.

John: Col, if the threshold came down because of funding, shouldn’t some children leave the scheme?

Col: Some, yes. But only when there’s somewhere to go. Moving a child off the NDIS into nothing doesn’t save money. It moves the cost to schools, hospitals and parents who give up work. That’s why the order matters. The other supports have to be there first.

Col: John, you said it’s the greatest policy failure. Greater than what? What’s your measure?

John: Cost against design, mostly. Nothing else has missed its own estimate by that much and kept growing.

Col: Then it’s the most expensive surprise. That’s not the same as the greatest failure. A scheme can cost more than anyone planned and still be worth more than it costs. Nobody has done that sum properly since 2011. That’s a gap we could both ask for.

John: I’ll give you that. Nobody has measured what the scheme returns, in work, in health, in carers back in jobs. I’d want that number before I vote for or against anything.

I’ll keep my vote. Not because the people on the scheme don’t deserve support. Because a scheme that rewards entry and can’t measure leakage puts the support of the people who need it most at risk. Fix the incentives, fund the supports around it, and target the scheme at permanent and significant disability, the way it was designed. If that’s done, ask me again in five years.

I’ll keep mine. A greatest failure would be a policy that didn’t help the people it was for. This one does, for most of them, for the first time. It grew because everything around it was missing. The honest fix is to build the rest of the system, not to use the word rort about people who walked to the only well there was.