If the public critics of the Coalition's welfare debt recovery scheme thought they were going for an easy ride, they would be sorely mistaken.
At the Senate hearings over the past two weeks, government senators have come armed with arguments to defend the program – most known as robodebt – and to attack those who want to see it dismantled or reformed.
Leading the charge, Liberal senator Hollie Hughes asked two legal academics if it was their usual practice to “use media slogans” in their submissions. She called that "disrespectful".
In the Western Australian town of Mandurah, community legal clinics appeared to be blinded when Hughes accused them of creating anxiety by using the "robodebt" moniker. It was a "misnomer", she said, because the system now included far more human oversight.
“Just wondering if you understand, or if you understand now, how the income compliance system works under it [the new system] and if you are aware of how those debts are generated, ”Hughes told the legal clinics.
Critics of the scheme – from social services groups, academics, Victoria Legal Aid, which is challenging the scheme to the courts, Labor and the Greens – say they know exactly how it works.
"Despite some improvements to the program's implementation, two fundamental design flaws remain inherent," Jacqui Phillips of the Australian Social Service Council told a hearing in Canberra.
“One is the use of ATO-reported annual income over the period someone has received income support, which leads to incorrect calculations of overpayments.
"The other is the reversal of the onus of proof, which requires people to disprove alleged debts on the basis of very limited information."
'It fails maths'
Terry Carney has a novel way of explaining the problem he sees with the scheme, which the government hopes will reap $ 2.1bn over the next three years.
He likens it to watching Don Bradman's last innings and using it as a basis to demand he prove his batting average. "Bradman got 99.94, not a duck," Carney says, noting Don's ill-fated final time at the crease.
Carney is a former member of the administrative appeals tribunal, which has the legal power to overturn Centrelink decisions. He has 35 years of experience dealing with social security issues.
The system, according to Carney, was among the critics who appeared before the Senate inquiry into robodebt over the past two weeks, "fails maths".
"What comes from the ATO is an average," he told the inquiry. "What is required … is an exact calculation."
Officials from the Department of Human Services told the inquiry that initial debt review letters were generated when a compliance staff member used what they collectively described as a "predictor" debt.
The process, as described by critics, uses a person's annual income from tax office data and compares it to what a recipient welfare reported to Centrelink fortnightly over the same 12-month period, dating back as many as seven years.
"You're going to get a letter if the discrepancy is significant," Jason McNamara of the Human Services Department told the inquiry this week. “More than $ 1,000… you think you have a better than 95% chance that you have a debt. Then we will start a review process. ”
For an exact calculation, Centrelink needs to know exactly how much a person was paid over the fortnightly periods they received welfare.
Before the scheme was ramped up by the Coalition in 2015, Centrelink would gather this information itself. It was a slow, manual process involving contacting former employers and banks to ask for a person's payslips and financial statements.
It would send about 20,000 letters to people each year. Now, about 10,000 are sent each week. Rather than getting the pay information on behalf of the recipient's welfare, Centrelink now demands they get it.
If they do, the department will use the average and the person will likely get a debt. If they get the information, the department says it will help people do so.
But at a hearing in Canberra, officials from the Department of Human Services confirmed that, since February 2017, Centrelink has contacted only 1,000 businesses to obtain payslips on behalf of welfare recipients facing a debt review.
Given that many people will have multiple employers in their work history, this means the number of people Centrelink has assisted will be even lower. Over the life of the scheme, more than 900,000 debts have been issued.
"There would be hundreds of thousands of people who would have difficulty contacting their employers," Labor Senator Deborah O'Neill told the inquiry.
A department official snapped back: “There is no evidence of that. We help the people who contact us and tell us they are in trouble. ”
Buoyed by evidence from the Australian Privacy Foundation, Hughes argued that it would "trample" on welfare recipients' individual rights if Centrelink obtained their pay or bank information without their knowledge.
"If they need consent, Senator, let's get consent," the legal academic and robodebt critic Darren O'Donovan told the inquiry.
Carney argued that, if they did not respond, Centrelink should gather evidence using special legal powers it has to compel companies and banks. That would always be better than calculating a debt without all the evidence.
Ken O'Shea is a licensed rigger who received a robodebt for $ 7,000. "The nature of the work means short bursts of intense activity of up to 100 hours a week, followed by nothing," he told the inquiry.
O'Shea asked Centrelink to help gather the information. "Historical bank statements are expensive – $ 4 per page over 12 months," he said. “This was the money I had. I asked Centrelink for a billing address so they could pay but they refused to do so. Almost two years ago, they said they would contact my employer for payslips. They have not done this.
"It seems very unfair that they expect me to incur costs and do their work for free because they have an idea in their head about my income."
Amid all that, there is the question of who was affected by this scheme. It was one thing to say that people should contact Centrelink – but those asked to be among the most vulnerable in the community.
The department's own data showed that of those contacted by registered mail or read receipt emails, 30% did not engage in the process, and another 30% did not complete it.
"These debts are shaped by the resources, the misunderstanding of ordinary people, rather than the best possible, objective evidence," says O'Donovan, who was lashed out by Hughes for using the term robodebt. He stands by it.
He told the committee: "Every Australian has a stake in this policy … Life-changing government decisions are being distorted by our status, our resources, our fears, our vulnerability."
With or without reform, Carney argued the current system needed to be "stopped and fixed".
Carney noted the department never appeals robodebt cases to the second division of the tribunal, which publishes its decisions. "I don't believe it has any confidence at all in its legal findings for its position," he says.