Robodebt in the Aftermath of Government Backdowns?

adminSocial security rights review

Terry Carney

The wash-up from Centrelink’s inability to defend the legality of robo-debt is becoming clearer, but many loose ends remain.  Government’s immediate (and appropriate but overdue) response was to abandon robo-debt,[1] and set about determining who was entitled to repayments.[2]

Six months after the humiliation of the November 2019 Federal Court ruling in Amato[3] the Government put some numbers on the projected cost of repaying false and unlawful working age debts based on ATO averaging in place of the required fortnight-by-fortnight information about earnings.  Repayments totalling $721,000 million are estimated to be owed to 373,000 individuals relating to 470,000 debts.[4]  But those figures are likely to understate matters.  The Government’s numbers do not appear to include around 70,000 members of the class action[5] who received a robo-debt but did nothing about it and it remains unpaid, or some 170,000 who did respond in some way and for whom repayments commenced.[6]  The overall costings appear to exclude paying interest on the monies improperly collected, which is being resisted by the Government[7] despite being a component in Amato.  Nor does it include damages for harms imposed (a key part of the class action), any exemplary damages component of such an award, or the very substantial in-house costs associated with redeployment to robo-debt repayment investigation, of nearly half of the public servants previously working on debt recovery and associated tasks.[8]

The ultimate fate of most of these questions hang on the progress and outcomes of the class action, currently scheduled for trial in late September this year.  In the meantime, there are a number of more pressing practical questions which remain unresolved.[9]

Some of the unresolved questions

What happens if bank details or contact information is not current?

Government appears to have recognised that electronic refunds are legally valid and administratively viable only where it holds current bank details of the person.  This is unarguably so.  While many people will anticipate the need to update bank records or respond quickly to any accurately addressed correspondence inviting them to do so, it can be anticipated that significant numbers will not do so.  What is the government’s obligation in this case?  Can it delay repayment until the update occurs?

The correct answer in my view is ‘no’.  Rectification of the unlawfully extracted debt must envision doing so by any reasonable and proportionate method.  Consequently if the residential address of the person is known with confidence, either from Centrelink or ATO or other records of the Commonwealth, then government is I believe ultimately obliged to forward a bank cheque for the requisite amount to that address.

What progress has been made on accessing ARO and AAT review?

This is a question of vanishing relevance for most robo-debt victims since refunds are supposed to arrive within months;  well overdue given it will for many be nearly a year after Amato and without the required interest on the amount in question, but a lot better all the same.  There are however three frustrations that a small number of people may encounter.

The first and most frequent of these frustrations will likely be that those who miss out on the routine roll-out of repayments, or rightly or wrongly believe they have missed out, may still encounter the tail-end of the previous ‘catch-22’ barrier to obtaining merits review of robo-debts.  This policy was where Centrelink refused requests for reconsideration by an authorised review officer (‘ARO’) on the basis that there was as yet no ‘decision’ to raise a debt, ‘merely’ insistence on provision of pay-slips or other information as a basis for a ‘reassessment’.  As previously explained,[11] this is an unlawful denial of review rights, but one that anecdotal evidence suggests continues to be a barrier for some.

The second of the potential frustrations – fortunately for very few people – concerns any past reviews of robo-debts.  Certainly most AAT first tier reviews tended to spot and set aside robo-debts as unlawful, but not all members will have done so.  For some AAT members this was because they did not adequately understand or agree with the law as subsequently determined in Amato, or because the case papers and hearing did not adequately bring the matter to mind.  Seeking to obtain belated AAT2 review of those AAT1 rulings will rest on obtaining leave to appeal outside the 30 day limit, and that leave is difficult to obtain.[12]

Third, any case previously ‘settled’ in the AAT even if under some pressure to see an end to the dispute will likewise be beyond revival as not involving a reviewable ‘decision’.[13]

As mentioned, these last two situations are likely to be rare occurrences and involve issues perhaps better left to ‘come out in wash’ by way of rectification from the roll-out of the repayments (which ought to catch any and all robo-debts irrespective of their later history).  But they are potential irritations none the less.


The sorry episode of the robo-debt debacle is slowly coming to a resolution, but messy loose ends unfortunately still look likely to cause some grief for victims and their hard pressed welfare rights advocates over at least the remainder of calendar year 2020.

[1]             Terry Carney, ‘Robodebt failed its day in court, what now?’ on The Conversation (28 November 2019)<>

[2]             For analysis of the immediate implications of the ruling, see Terry Carney, ‘Run-away Robo-debt Train Derailed: Pity about inadequate damage clean up?’ (2020) 6(1) Social Security Rights Review

[3]             Deanna Amato v The Commonwealth of Australia VID/2019/611-0  The initial test case was effectively spiked by waiver of all of the debt in question: Madeleine Masterton v Secretary, Department of Human Services of the Commonwealth VID73/2019.

[4]             Terry Carney, ‘Government to repay 470,000 unlawful robodebts in what might be Australia’s biggest-ever financial backdown’ on The Conversation (29 May 2020)<>

[5]             Katherine Prygodicz & Ors v Commonwealth of Australia VID1252/2019 at  Terry Carney, ‘Robo-debt class action could deliver justice for tens of thousands of Australians instead of mere hundreds’ on The Conversation (18 September)<>

[6]             Emily McPherson, ‘Hundreds of thousands still stuck in robodebt limbo’, Nine News 30 June 2020 <>

[7]             David Estcourt, ‘Senior ministers may be hauled into court over robo-debt class action’, Sydney Morning Herald Tuesday 16 June 2020 <>

[8]             Luke Henriques-Gomes, ‘Robodebt: total value of unlawful debts issued under Centrelink scheme to exceed $1bn’, The Guardian Wednesday 10 June 2020 <>

[9]             The suggestion that robo-debt methodology was applied before 2015 to raise working age ‘debts’ which also should be refunded as lacking a lawful foundation is too speculative to be addressed here. In sitting on the tribunal I never encountered such cases if they occurred, and it is possible that some or most were not true robo-debts because they applied averaging to pension payments, where averaging is more valid.

[10]            For a recent review, Shelley Bielefeld and Fleur Beaupert, ‘Income Management and Intersectionality: Analysing Compulsory Income Management through the Lenses of Critical Race Theory and Disability Studies (‘Discrit’)’ (2019) 41(3) Sydney Law Review 327.

[11]            Carney, ‘Run-away Robo-debt Train Derailed: Pity about inadequate damage clean up?’ (n 2 above).

[12]            Leave to belatedly appeal an AAT1 decision setting aside a robo-debt with directions that the debt be lawfully recalculated on the basis of obtaining the fortnightly employer payment details was recently refused in Re NPXL and Secretary, Department of Social Services (Social services second review) [2020] AATA 1891.

[13]            Re Elliott and Secretary, Department of Social Services (Social services second review) [2020] AATA 2176.