Feedback loops

Failures of bureaucracy and Robodebt

My friend told me a phrase in Chinese: 金玉其外 敗絮其中¹ (roughly translated to “gilded on the outside, rotting within”). It feels to me like some institutions in society have rotted within, while on the outside they still present shining exteriors. How does this happen?

Photo by Gwendal Cottin on Unsplash

I’ve become convinced that part of the problem stems from a failure of feedback loops, after I recently became unwillingly embroiled in a massive technology fail. The technology concerned was supposed to make everything easier, but instead, it was a disaster. I’ve spent the last three or four weeks mopping up the debris. It was a tremendous stress for everyone involved.

“Tell the relevant person responsible for it that the technology was a dud,” said my husband, after hearing me moan about it for the umpteenth time.

That’s when I realised: There was no relevant person to contact. No one seemed to be responsible for this particular “innovation”. It had been introduced by a bureaucracy, of course. There was a general email, as well as a general phone line, where you got a different person on the line every time. That’s it.

My fear is that the failures of this system won’t be dealt with, because there’s no one to receive feedback. Even if they do receive my feedback, there will no consequences: I’m just another moaning “stakeholder”. Nonetheless, I intend to try to ensure that someone in power gets my trenchant and detailed criticism of the system, and my seven recommendations for improvement. What’s the point of being a law professor if you don’t use your analytical skills and general argumentative nature to stand up to things like this?

As is often the case when one becomes entangled with a bureaucracy, my dilemma was an instance of the “principal-agent” problem: namely, that the person who makes the decision is not the one who suffers the adverse consequences of that decision. Indeed, in my situation, the adverse consequences landed on absolutely everyone other than the decision maker.

As a result, there’s no incentive for the bureaucracy to do anything about it. Nothing happens: the bureaucrats can indicate on their performance reviews that they successfully instituted a new system, tick the boxes, and continue on. Meanwhile, the system continues to fail. The process comes to dominate, and the fact that the process deals with real people is forgotten. The feedback loop is broken: no one knows (or perhaps no one has an incentive to care) that people aren’t happy with the system.

Photo by Cash Macanaya on Unsplash

While my experience with technology and bureaucracy over the last month has been unpleasant, it hasn’t been on the level of the recent Robodebt scandal, where vulnerable people on Federal government welfare benefits were subjected to automated debt recovery, on the basis that they’d been overpaid. To borrow the memorable phrase of my colleagues Toby Murray, Marc Cheong and Jeannie Paterson, the flaw was at the heart of the automated decision making process—the algorithm used to calculate the debts was wrong—and hence at least 470,000 of the alleged debts were incorrect or non-existent. Some alleged debtors committed suicide, but the process just kept rolling. Once you remove people from the decision-making process altogether, the principal-agent problem is intensified. Algorithms are simply a process, and they certainly don’t suffer adverse consequences if their process is flawed.

The scheme has just been subject to a Royal Commission. I’ve not read all the evidence, but what I’ve seen is frankly hair-raising and extremely distressing. How could the bureaucracy keep following a flawed process, despite the alarm bells which were ringing, and the concerns that people raised? Instead, it seems that they doubled down on the flawed process. It’s like a Kafkaesque nightmare: the bureaucratic machine, once set into motion, can’t be stopped, even if it’s crushing people along the way, and even if people are shouting at the machine to stop.

Catherine Holmes AC SC, the Royal Commissioner, began her preface to her Report in the following way:

At the outset of my inquiry, I had anticipated that the Commission would uncover how such a patently unreliable methodology as income averaging, without other evidence, to determine entitlement to benefit could become part of an Australian Government debt raising and recovery scheme. What has been startling in the Commission’s investigation of the Robodebt scheme has been the myriad of other ways in which it failed the public interest. It is remarkable how little interest there seems to have been in ensuring the Scheme’s legality, how rushed its implementation was, how little thought was given to how it would affect welfare recipients and the lengths to which public servants were prepared to go to oblige ministers on a quest for savings. Truly dismaying was the revelation of dishonesty and collusion to prevent the Scheme’s lack of legal foundation coming to light. Equally disheartening was the ineffectiveness of what one might consider institutional checks and balances – the Commonwealth Ombudsman’s Office, the Office of Legal Services Coordination, the Office of the Australian Information Commissioner and the Administrative Appeals Tribunal – in presenting any hindrance to the Scheme’s continuance. The report makes a number of recommendations. Some are directed at strengthening the public service more broadly, some to improving the processes of the Department of Social Services and Services Australia. Others are concerned with reinforcing the capability of oversight agencies. … [W]hether a public service can be developed with sufficient robustness to ensure that something of the like of the Robodebt scheme could not occur again will depend on the will of the government of the day, because culture is set from the top down. [emphasis added]

Of course, me being me, I immediately wondered if some of the government officials involved in this scheme had committed the tort of misfeasance in the public office. I’m evidently not alone: on page 659 of the Royal Commissioner’s report she wondered the same thing. Then I calmed down, and realised that the tort of misfeasance in the public office is difficult to prove: a person must show that the action by the public official was undertaken in bad faith, with an intention to harm the person, and that the actions were either malicious or known to be unauthorised, and that the person suffered actual damage as a result.² That being said, if the public official is found to have committed the tort, they may be personally liable, and the government may not be obliged to indemnify them. The losses covered may include losses to reputation, psychological injury, and damage to property.³

I can’t help hoping that some of the government officials involved in the Robodebt scheme are sued personally for damages, including exemplary damages (designed to punish the wrongdoer and make an example of them).⁴ It’s a drastic way of ensuring that the feedback loop is re-established—make the official pay personally for their wrong—but that’s part of the way in which private law can oil the wheels of society, and realign the incentives. If a government official behaves wrongfully in this way, then they are no longer acting in the public interest. It is a gross betrayal of the people whom they are supposed to serve.

I hope that the government takes the feedback and criticism of the Royal Commission’s Report on board. As the Commissioner points out, there must be the political will to deal with these issues, to ensure this misuse of power never happens again. Politicians, please stop gilding the exterior—forget the photo opportunities, the press statements and the grandstanding—and stop the rot within, when it occurs.

1. Pinyin: jīnyù qí wài bàixù qízhōng.

2. David Rolph, Jason N. E. Varuhas, Penelope Crossley, Michael Douglas, Balkin and Davis Law of Torts (Lexis Nexis, 2021) [25.28] - [25.33].

3. Ibid, [25.33].

4. Incontrovertibly available for misfeasance in the public office: Sanders v Snell (1997) 73 FCR 569 (rev’d on other grounds in (1998) 196 CLR 329). Although cf M Aronson, ‘Misfeasance in Public Office: A Very Peculiar Tort’ (2011) 35 MULR 1, 13: exemplary damages are rare. I would say this would be a perfect occasion for their award.