Robodebt: Can you recall a greater failure of public administration?

I am sure I am not the only retired public servant morbidly fascinated by the Robodebt Royal Commission. As well as following the excellent reporting of Rick Morton and Luke Henriques-Gomes, I have been reading the transcripts as they are published.

I have been particularly struck by Kathryn Campbell's exchange with Counsel Assisting Justin Greggery KC. Asked if she considered Robodebt to be a "massive failure of public administration", she replied: "I consider it a failure of public administration".

Mr Greggery persisted: "On a massive scale?"

Ms Campbell: "On a massive sca- ... on a significant scale".

Mr Greggery: "Can you think of a larger scale failure?"

Ms Campbell: "I think it is a significant matter, I have been involved in other significant failures, but I don't think it's helpful to talk about those."

Unfortunately Ms Campbell's appearance was interrupted at that point, so we will need to wait until December 5 to see if Counsel press her to identify other significant failures.

Of course any observer of Australian government and public administration can identify numerous failures of public administration, and over the four decades I worked in the public service I observed many and was involved in some.

It is clear that the Australian government is really bad at managing procurement. The worst example was probably the Coastwatch imbroglio, where a contract to conduct aerial coastal surveillance was awarded to a company without enough planes or pilots to do the work.

Defence acquisitions and information technology have been particularly problematic. The list of defence equipment procured on time, under budget, and with the specified capability is a lot shorter than the list of equipment delivered late, over budget, and without the specified capability. And in relation to information technology, the COVIDSafe app is only the latest in a long list of projects with poor outcomes.

The COVID-19 pandemic created many challenges for government, and not all were dealt with successfully. The vaccine procurement and initial distribution process was a shambles, described by former Prime Minister Malcolm Turnbull as "the biggest failure of public administration I can recall".

JobKeeper may have averted the worst economic consequences of the pandemic, but the failure to include a mechanism to recoup funding from recipients who did not suffer a revenue downturn cost taxpayers billions of dollars.

One can only imagine the then government's reaction to a proposal that JobSeeker entitlements should be based on estimated income, with no clawback from people who were unduly pessimistic.

Leaving aside procurement, information technology, and the exigent circumstances of the pandemic, the worst example of a failure in public administration I can recall is the maladministration of the detention provisions of the Migration Act in the 2000s highlighted by the cases of Cornelia Rau and Vivian Solon.

Ms Rau was a permanent resident who was held in detention for some months, while Ms Solon was an Australian citizen who was detained and then deported to the Philippines.

Inquiries into the cases conducted by retired police commissioners Mick Palmer and Neil Comrie were scathing about the performance of the Department of Immigration in the conduct of two cases. Mr Palmer observed that:

"There is a serious cultural problem within [the department's] immigration compliance and detention areas: urgent reform is necessary ... [there is] a culture that is overly self-protective and defensive, a culture largely unwilling to challenge organisational norms or to engage in genuine self-criticism or analysis.

"Officers are authorised to exercise exceptional, even extraordinary, powers. That they should be permitted and expected to do so without adequate training, without proper management and oversight, with poor information systems, and with no genuine quality assurance and constraints on the exercise of these powers is of concern. The fact that this situation has been allowed to continue unchecked and unreviewed for several years is difficult to understand."

However, despite the fact that the exceptional and extraordinary powers under the Migration Act had been abused to detain and deport people who were legally present in Australia and had a right to remain here, the policy framework under which the department operated did allow for detention and deportation of some people in some circumstances.

By contrast, Robodebt was conceived, approved, implemented, and maintained without any support in the policy framework.

The Royal Commission is constituted by a lawyer, and assisted by lawyers, instructed by other lawyers. It is not surprising that it has focused on why neither the initial legal advice ("a debt amount derived from annual smoothing or smoothing over a defined period of time may not be derived consistently with the legislative framework"), nor the subsequent draft Clayton Utz advice which reached the same conclusion, was heeded.

This is an important question, but with all due respect to the Commission, it is not the most important question.

The most important question is how a savings proposal so completely outside the policy framework was allowed to proceed by the DHS and DSS secretaries and senior executives. All of them ought to have understood that eligibility for social security payments is based on fortnightly income.

Even if they had received advice that it was technically legal to use an average fortnightly income derived from annual income in re-assessing eligibility, it was still clearly inconsistent with the policy framework set out in the Social Security Act.

READ MORE:

Middle-ranking staff in DSS were very clear in their views about this. On November 7, 2014, David Mason, an acting director in the Rates and Means Testing Policy Branch, sent an email to Cameron Brown, Director of Payment Integrity and Debt Management, setting out the various policy concerns raised by the income averaging proposal.

In his evidence to the Royal Commission on November 3 this year Mr Brown essentially accepted these concerns, and added that the proposal reminded him of "speculative invoicing".

He went on to say it was probably unethical: "Well, a lot of the individuals who would end up receiving these notices would be - the onus of proof on what they earned over the period of time, stretching back potentially years, would be a level of proof which they simply would not be able to substantiate. And unemployed people are, almost by definition, they have very vulnerable cohorts within them, and that there would be people who would enter into agreements to repay debts, which they had not incurred in the first place, and I felt that the practice as a result was unethical."

Yet despite these views Robodebt proceeded. So far the Royal Commission has not heard evidence about how this happened, partly because then DHS Deputy Secretary Malisa Golightly, who signed the ministerial submission initiating the formal policy development process, died last December.

The evidence of other senior DHS and DSS staff has not been very helpful. Ms Campbell has told the Royal Commission that she does not recall very much about the development of the proposal.

From the DSS perspective then Deputy Secretary Serena Wilson gave evidence that she thought the proposal had been killed off in early 2015, and did not realise for several years that income averaging was in use. Then DSS secretary Finn Pratt essentially said he relied heavily upon his staff for advice on policy issues.

Whatever the reasons, the fact the proposal proceeded resulted in about half a million people receiving debt notices premised on the fact that their average fortnightly income made them ineligible for benefits - even though average income was irrelevant in determining eligibility.

I cannot recall a more substantial failure of public administration. If Ms Campbell can identify "other significant failures" on the same scale when she resumes her evidence on December 5, the government may need to establish more Royal Commissions.