X
Innovation

Services Australia says no Centrelink robo-debts raised since July 2019

6,888 debts have been reduced to zero and 1,421 reduced to 'non-zero'.
Written by Asha Barbaschow, Contributor

Services Australia has told the Senate Community Affairs Legislation Committee that since July 2019, no Centrelink debts have been raised under its Online Compliance Intervention (OCI) scheme.

It has in that time, however, raised 10,090 debts through its employment income confirmation system, under which the department says it checks an individual's employment income is correct "so you get the right payment amount".

During Senate estimates held in March, Services Australia was asked how many robo-debts had been reduced to zero, reduced to non-zero, partially waived, and fully waived since July last year.

Providing the answers in response to questions taken on notice, the department revealed a total of 6,888 OCI debts were reduced to zero; 1,421 were reduced to non-zero, which included debts valued below the original amount; and 167 were labelled as debts waived or written-off permanently.

Under a system called check and update past income, Services Australia has raised 82,099 debts since the start of the financial year.

In 2016, the department kicked off the data-matching program of work that saw the automatic issuing of debt notices to those in receipt of welfare payments through the Centrelink scheme. The OCI program automatically compared the income declared to the Australian Taxation Office (ATO) against income declared to Centrelink, which resulted in debt notices, along with a 10% recovery fee, being issued whenever a disparity in government data was detected.

One large error in the system was that it incorrectly calculated a recipient's income, basing fortnightly pay on their annual salary rather than taking a cumulative 26-week snapshot of what an individual was paid.

Centrelink's OCI program from 1 July 2016 through 31 August 2019 saw 1,159,662 assessments be initiated using the automated data-matching technique.

The federal government in November paused the automated data-matching element of robo-debt, but Services Australia has continued to use income information from the ATO to identify "significant discrepancies" with income information.

In another response to questions on notice from the same estimates round, the department said that since 19 November 2019, no debts have been raised from the Income Compliance Programme (ICP) through relying solely on income averaged data from the Australian Tax Office.

For the ICP, the number of debts raised was 822 in December, 968 in January, and 722 in February.

Also revealed in response to questions on notice was the number of debts raised under the robo-debt initiative that has been reviewed by the Administrative Appeals Tribunal.

A total of 1,214 debts received a first review by the AAT. 495 were affirmed, 30 were dismissed, 101 were set aside and remitted, 341 were set aside, 139 were withdrawn, and 87 are still awaiting decision.

180 debts made it to an AAT second review. 144 were listed as "varied or dismissed by consent" and 31 are still awaiting a decision.

Services Australia said the outcome of dismissed by consent often indicated a settlement, which only occurs at the AAT's second review level.

From 1 September 2019 to 31 December 2019, Services Australia said it received an estimated 6,247 requests for a quality check by a Subject Matter Expert (SME) for decisions connected to the OCI initiative. In the same period, 5,764 quality checks were finalised by an SME.

Services Australia has been undertaking an analysis of all income compliance reviews to identify where income averaging was used to determine a debt and was meant to have a solid number of how many people were affected by automation activities, as well as a plan of attack, by the end of January.

As these debts are identified, Services Australia said recovery would be frozen and it would remain in place until a decision was made regarding next steps.

"Should a customer contact the Compliance Assurance Division to enquire about the debt freeze, Compliance Officers are to advise the customer that the freeze has been applied until further decisions are made by government," the department wrote.

According to a report from The Guardian, over 400,000 debts were incorrectly issued and the federal government will reportedly be forced to refund around AU$550 million to recipients of welfare due to the automation faux pas.

Providing updated staffing figures, Services Australia said the number of people allocated to the income compliance identification process work as at 31 December 2019 was 222 Australian Public Service staff and 427 labour hire staff.

As at 31 January 2020, the number of staff assigned to the income compliance identification process was 632.

Also detailed was the cost of administering the OCI program from 2015-16 to 2018-19, inclusive. Services Australia said there has been approximately AU$606 million spent.

As of 31 December 2019, the cost of using registered post as part of the program was approximately AU$3.7 million.

After Minister for Government Services Stuart Robert's application for legal professional privilege was refused in providing responses to the Senate Community Affairs References Committee earlier this year, a bunch of responses to questions on notice from his department have followed suit.

"On 2 April 2020, the Minister for Government Services made a public interest immunity claim with respect to the disclosure of all communications involving Services Australia, Departments and Ministers about any litigation involving the Income Compliance Program," Services Australia wrote as an answer to questioning on how many debts were calculated solely using averaging and how many used averaging in part.

Asked if unlawful robo-debts recovered from deceased estates would be repaid, Services Australia said the question relates to a court case that is currently before the Federal Court of Australia and the department would "abide by any decision of the Court".

MORE ON ROBO-DEBT

Editorial standards