Uncovering the Grim Realities: Excavating Unlawful Income Apportionment
Recently we learnt that income apportionment goes back to the 1990s. But yesterday we learnt more about the scale of this unlawful practice.
Yesterday saw the release of some important freedom of information documents. They chart the ever-deepening fallout from the unlawful income apportionment disaster. These documents reveal that the Department of Social Services (‘the Department’) currently estimates that nearly 5.3 million debts may be affected by the practice of unlawful income apportionment. The unlawful practice, which involves calculating debts contrary to section 1073B of the Social Security Act 1991 (Cth), has been conceded by the Department to have occurred from at least the early 1990s until 2020 — that is, for about 27 years.
Whether and how these debts can be recalculated now is a question mired in litigation. The Department failed to get its preferred recalculation method over the line in a recent Administrative Appeal Tribunal appeal known as FTXB. However, the Department did not appeal the decision. Indeed, perhaps the Department was willing to accept the decision of the AAT. But we will never know, because now an appeal of the FTXB matter is on foot, scheduled for the Full Federal Court, and instituted by the social security recipient. (Their name, to be clear, is pseudonymised in the initialism ‘FTXB.’) Nevertheless, in responding to the appeal, the Department is trying to stave off both legal arguments and a practical reality: recalculating these debts will require detailed records of employment that in many cases no longer exist.
Scaling a Mountain of Unlawfulness
Importantly, the Department qualifies its estimate of 5.3 million debts. That figure is not an estimate of how many files are actually income apportioned, the FOI release says, but an estimate of those files that may be income apportioned. No doubt the Department would have been aware this release would prompt disquiet, and so has sought, reasonably enough, to put these figures into some context. That context, however, is one of uncertainty. The degree of evidentiary probity is still at a low level at this stage. Thus, the reason the qualification is necessary stems from the fact that the Department has not yet developed a method by which it can reliably satisfy itself that these files are affected by income apportionment. One imagines the Department has developed a model by which certain criteria or indices of income apportionment have been identified, and then has applied that model, en masse, to its historical archives.
But even if the figure of 5.3 million debts is an overestimate, the available evidence leads only in one direction: a significant proportion of historical Centrelink debts have been subject to unlawful income apportionment. After all, the review has been going on for many months now, and the figure of 5.3 million is much lower than it was a few months ago. In June 2023, for instance, the estimated total was more than 9 million debts. We are, in all probability, dealing with millions of unlawful debts. The Department’s own sampling exercises have found the following prevalence trends:
Based on FOI documents, in its first sample of debts, 72% of all screened employment income files were apportioned. These were ‘undetermined’ debts (ie, debts ‘in train’ — identified either before being alleged against or recently alleged against the recipient). This figure is worked out by combining two separate phases — but we don’t have a copy of the final report.
The Commonwealth Ombudsman then forced the Department to go further back in time. This second sample of historically deeper debts (going back to 2003) found that 54% of all screened employment income debts were unlawfully income apportioned.
In AAT matters, 78% of checked employment files have turned out to be unlawfully income apportioned.
The Department’s estimates have bounced around somewhat in the past number of months. They reached as high as 9 million in June (as noted above), then fell to 2.5 million in August. Two weeks later, they were back up to 5.3 million. This reflects a few key developments in 2024, which do tend to complicate the figures. Firstly, the Department finally came to terms with the reality that income apportionment goes back to the 1990s — at least. Secondly, there are ‘fresh streams’ of secondary debts emerging now, where compliance review officers had relied on faulty income apportioned income to calculate new debts.
It may be obvious to say, but getting a person’s earned income wrong has knock-on consequences for recipients: it can reduce supplements, family tax benefits and parenting payments. Bad data begets more bad data and the evidentiary fruit of a poisonous tree is itself tainted. We don’t yet know how many of the 5 million debts are in this category of ‘fresh stream’ debts, or how frequently they have occurred. But this new stream of unlawful debts is identified and analysed in a second freedom of information request disclosure released yesterday (discussed below).
Cascading Consequences
Earlier this year, the department webpage on income apportionment was quietly modified to read:
In early April 2024, we paused repayments of more debts we found that may have been indirectly impacted by income apportionment. These other debts may relate to payments that don’t need fortnightly income reporting, such as Family Tax Benefit.
Yesterday’s second FOI release confirms that this debt pause involved 41,000 debts belonging to over 30,000 Australians. The document also notes the existence of further potentially unlawful debts — those that fit into certain categories indicating that recovery has already been completed.
As discussed above, the knock-on effects of getting employment income wrong from the start are well-known. The wrongly calculated figures can disentitle a social security recipient to a supplement or payment that they would, on correct figures, be legally entitled to receive. One of the bitterest aspects of the Robodebt experience was the refusal of the Commonwealth to address these secondary or ‘knock-on’ financial impacts. What would have been if these debts were not miscalculated? In some areas of law, such knock-on effects are known as a ‘loss of chance’ or a ‘loss of opportunity.’
As we have discussed previously, some agency officials made efforts to highlight the full financial consequences of averaged debts and have them remediated as they arose. Legal representatives of Ministers of the Commonwealth government, however, argued against this. In settling the Robodebt class action, the government approved only ‘cut-price’ remediation. And as is underlined by the class action appeal now underway, the contours of so-called ‘robo-justice’ have been shaped by the immunities and preferences of the powerful. Will the income apportionment scandal prove any different?
Securing a People’s History of Social Security Law
In a recent senate estimates hearing, the Department finally recognised that income apportionment dates back to at least 1991. It described a process of consulting retired social security staff and relevant academics to arrive at this conclusion. But that conclusion is unsurprising. The historical practices of income apportionment and income averaging — differentiated only in terms of the source of the data (the one from the ATO, the other from the recipient) — are clearly visible even through basic desktop research.
It is there in a 1997 decision of the AAT called Osgoode — a matter in which the Department undertook income averaging decades before Robodebt. The Nolan (1997) case similarly shows the Department using daily rates across one or more Centrelink pay periods, contrary to what is now accepted as the requirements of the law. And income apportionment is visible in another case in the Federal Court called Danielson (1996), which was mistakenly used by Department to produce a formal guideline to raise debts in the absence of records. While some tribunal decision-makers attempted to resist the guideline, their efforts did little to stop the Department pursuing its own conveniences — its own policy preferences. Nor did they persuade the Department to question its own interpretation of the legal requirements.
One of the most striking aspects of the income apportionment debacle has been the considerable effort expended on distancing it from Robodebt. But there is at least one important intersection — real people. This 5.3 million cohort will include many thousands who received an initial robodebt letter or incurred a robodebt. In response, many will have gone to their employers and historical records and assembled evidence to prove the Department’s allegation wrong. These were the same people who told the Australian public of the stupidity of these allegations while politicians and bureaucrats dissembled. Some of those debts will have been proved wrong in toto (there was no debt to begin with). Others will have had the alleged robodebts significantly reduced.
But what happened when they handed over their corrective evidence? The Department then subjected them to an unlawful practice for a second time. All these years later, people in this category, who are entitled to nothing under the class action, may remain subject to unlawful calculations of their revised robodebts. And the Department does not yet know how to calculate these debts legally. The human cost of this situation — the stress, the confusion, the distrust — won’t be found in the cells of a spreadsheet.
The abiding lesson of this situation is that when you audit actual files — when you look properly at people’s income and their entitlements — the true realities tend to emerge. Prior to FTXB, the Department, in an apparent belief that its interpretation of the law would be upheld, ‘unpaused’ debt processing in respect of many debts earlier triaged as suspected unlawful debts. It was not until 349 cases had been trialled that that the Department attempted to recalculate them. In first two months of this recalculation effort, staff found and escalated 20 cases of unlawful income averaging. Yet the victims of these practices attract little attention, in the media or elsewhere, because their fate has already been shaped by the ordinariness of bureaucratic indifference. These are not the victims of headline-grabbing political decisions. They are the victims of institutionalised errors.
Waiting for Administrative Godot
So what happens next? It appears that we must now wait for the Full Federal Court decision in FTXB. A victory for the applicant — who will likely argue, as they did in the AAT, that detailed information about one’s earnings is required by the law before a debt can be asserted — would leave the Commonwealth without a recalculation method for these millions of debts. Such a result would be extraordinary, and its implications profound.
But, as we have discussed elsewhere, a win for the Department will just leave them with the same notional and functionally unwieldy recalculation method they are left with now. There is simply no efficient way of recalculating these files. No remediation will match with the mathematical or historical reality of these individuals’ welfare entitlements. But there are ways of risk-rating these debts and undoing the worst cases of harm. Writing them off and reimbursing them in full is one option. However, the ultimate answer as to the best course of action may not lie with the courts. Courts tend not to prescribe what is legal with such a degree of precision that the Department cannot then introduce its own method of meeting its legal requirements. In that sense, defining the best course of action may fall to the Department and what it does in the long-run. But the best course of action must surely involve confronting the reality — owning the administrative errors and ethical failures that led to this situation.