Yahoo News Government may not repay 200 Centrelink debts raised using ‘unlawful’ method
The federal government will not commit to repay more than 200 Centrelink debts raised using a method it says is unlawful and which it uncovered as part of a “historical review” of pre-robodebt cases.
Documents released to Guardian Australia under freedom of information laws detail the outcome of a “random sample” of pre-2015 welfare debts the government services minister, Stuart Robert, ordered his agency to conduct last month.
The “historical review” found of the 1,000 random cases analysed in 2009 and 2011, 206 were issued entirely or partially using a method known as “income-averaging”, which the government has conceded in public and in court is unlawful.
On Tuesday, a spokesman for the government services minister, Stuart Robert, refused to outline how the government could legally justify refusing to locate the recipients and repay debts the review had uncovered.
The document released under freedom of information showed in the majority of cases debts raised by Centrelink in 2009 and 2011 were “fully verified”.
The fact some were raised using income-averaging likely reflects a previous policy whereby Centrelink staff could use the method as a “last resort”, according to leaked government advice seen by the Guardian. In general, it is understood staff were strongly warned against the approach until 2015, when they were instructed to raise debts this way.
Robert told parliament last month he had asked Services Australia to conduct a review of 500 debts from 2009 and 2011, and he went on to use some of its findings in a political attack against Labor, arguing the “income-averaging” method central to the robodebt program was a long-standing practice of both Labor and Coalition governments.
Before 2015, Centrelink only issued about 20,000 debts a year through a mostly manual process in which staff would first seek payslips or bank statements after detecting a large discrepancy between a person’s fortnightly reported income and annual ATO pay data.
The Coalition ramped up debt recovery by instructing staff to issue debts without seeking this evidence and used automation to lift the number of debt “interventions” to 20,000 a week.
This effectively reversed the onus of proof by forcing welfare recipients to find old payslips dating back up to six years to prove they had not misreported their income to Centrelink.
Some experts believe the proportion of “income averaging” found by the “random sample” ordered by Robert is unrealistically high and the true figure could be less than 1% of debts raised each year.
Darren O’Donovan, a La Trobe University law academic, backed calls for a royal commission into the scheme, alongside Labor, the Greens, and community campaigners including the Not My Debt group.
“There are unanswered questions about the nature of these debts, which were produced under a very different business process to robodebt,” O’Donovan said.
“One certainty is that the minister should not be waving people’s debts around in a political debate without committing to refunding all unlawfully averaged debts.
Related: Centrelink staff forced to administer botched robodebt scheme deserve apology, union says
“We need a royal commission to get to the truth, and end the politics.”
Robert told the National Press Club last week the “income averaging” practice had been used for 20 or 30 years, and was done so “on some mass” since 2007, though again he would not give a commitment to find affected debts and repay the money.
When doorstopped by the Guardian after his press club speech, he dodged questions about repaying the debts identified by the “random sample” but claimed the agency’s investigations suggested 4,000 debts a year were calculated with averaged income data from 2007 onwards.
He said computing systems introduced from 2015 to automate income-averaging meant the government “knows who every single one of those individuals are” but “prior to that there was no computing system that existed”.
“That’s informing us as to what’s possible,” he said. “We’re still working out what [is] indeed possible, what datasets exist, which is why the sample was done and indeed how far it goes back.”
Asked if the government would find these debts and repay them, Robert said only: “The first thing was to understand what was the basis of numbers and whether indeed the practice was being done at scale.”
Last month, Labor lodged a request for correspondence between Robert’s office and the Services Australia related to the “random sample” in the Senate.
