Australia’s Digital Transformation Stumbles Badly
The report stated, for instance, that, “… it has become clear to the committee that digital transformation is a policy area beset by soaring rhetoric and vague aspirations by government, largely unconnected to the actual policy activities actually undertaken.”
In addition, the Liberal-National coalition government’s execution of its digital services has been marked by a “litany of failures largely unprecedented in scale and degree,” and, further the report states, the “government to date has been unable to meet even the lower objective of being able to replace aging infrastructure without major mishap.”
Failures cited in the report include outages at the Australian Taxation Office, significant operational difficulties with the Department of Human Services’ Online Compliance Intervention program, and the cancellation of the Australian Apprenticeship Management System project as not being fit for purpose, among others. Together, the operational issues and development failures reflect “broader systemic problems” with the government’s capability to deliver information and communications technology (sometimes abbreviated as ICT).
The report from the Australian Senate’s Finance and Public Administration References Committee lists several familiar causes of these digital woes, such as a lack of leadership vision, direction, and accountability; government personnel who lack the necessary expertise; and an over-dependence on outsourcing and consultants; overly optimistic project objectives, costs, and schedules along with procurement strategies that do not reflect the myriad of risks involved. The report goes on to list several recommendations to address these digital transformation shortfalls.
A subsection of the report written by the government’s minority Liberal-National coalition committee members, however, vigorously rejected the report’s findings. They claimed that the main report, written by opposition Labor and Green party members, was essentially a political hit piece that was meant to embarrass the government and which reflected a “disregard for the facts.” There wasn’t a lack of leadership, vision, or accountability, for example, but just the opposite.
For instance, in 2015, the government created the Digital Transformation Agency to “help lead transformation across government and have central oversight of the government’s ICT agenda.” The agency, the government members of the committee argued, is already implementing the report’s recommendations, such as improving how government procures new systems, improving digital skills capability across government, and providing greater transparency of the government’s technical projects, costs, risks, and opportunities.
Furthermore, they claimed in their section of the report that there have been “hundreds, if not thousands of digital projects, both large and small, funded by the government that have been delivered successfully.” The failures cited in the report are “very isolated exceptions,” the minority members argued. Regardless, those failures are over-rated, since one has to expect failure “as part of the process of innovation,” they contend.
While there is no doubt a bit of political point scoring in the Senate report, it does speak to many of the fundamental problems that Australia, as well as many other governments, are experiencing while trying to implement “digital transformation.” This is especially true when the goal is to attain the Estonian level of government digitalization, which is considered the gold standard.
The operational difficulties mentioned with the Department of Human Service’s Online Compliance Intervention program are one example. In November 2016, the government rolled out an enhanced data-matching algorithm meant to compare social welfare payments received by individuals from Centrelink, the organization within the agency that delivers payments and services, against the information held by the Australian Taxation Office. The algorithm, which fully automates the mostly manual method previously used, determines whether the benefits received were correct, or if they were incorrect due to a benefits eligibility miscalculation or deliberate fraud.
While increasing the use of automation wasn’t unsound, the implementation was deeply flawed. The “robo-debt” algorithm, as it was soon dubbed, was developed in a way that assumed any discrepancy in data between Centrelink and the taxation office was essentially an indication of fraud on the part of the benefits recipient. In addition, where past policy held that the government needed to show that fraud existed, the policy was replaced with one where the recipient had to show they weren’t benefit cheats.
Further, the data used by the algorithm for discrepancy matching was unreliable and worse, the algorithm calculated income received by recipients incorrectly. This meant many recipients, at least one in six by some accounts, who appropriately received their benefits were automatically being sent letters telling them to prove they didn’t owe the government any money based on records the taxation office had said they did not have to keep. Previously, Centrelink workers would have manually checked for such discrepancies before the letters were sent to prevent such mistakes, but no longer.
A fundamental premise of digitized government is that it will increase citizens’ trust in government by showing that it is fair and transparent in its decisions. While the Australian government tried to frame the compliance algorithm in terms of fairness, saying that “Our aim is to ensure that people get what they are entitled to—no more and no less,” it has instead created the opposite belief.
Many Australians, especially the poor, now see their government using digital technology as an indiscriminate, uncaring, and illegal club to beat them with. The government’s planned use of facial recognition to determine if a welfare recipient should receive benefits will do nothing to change their minds.
Interestingly, the Digital Transformation Agency was “locked out” of working on the Online Compliance Intervention project. According to the former agency director Paul Shelter, the agency would have “failed it early on,” which is why it was probably never asked to review the project.
In the case of the Department of Education’s Australian Apprenticeship Management System project, the Digital Transformation Agency wasn’t locked out of it, it just didn’t provide much in the way of leadership. That effort was supposed to replace the outdated Training and Youth Internet Management System, but after spending six years and A$24.1 million, the effort was scrapped in May of this year with nothing to show for it.
However, the Education Department insisted the effort wasn’t a total disaster since the failure provided a teachable moment; the Department is now seeking ways “to ensure it doesn’t happen again.”
When the Digital Transformation Agency was asked about its role in that project, since part of its remit is to “ensure effective ICT and digital investments,” it basically shrugged its shoulders and admitted that really wasn’t its job. According to the Digital Transformation Agency, it’s a department’s leadership which is both responsible and accountable for whatever happens with an ICT system. This arms-length relationship seemingly undercuts the claims that the government is providing strong digital transformation leadership and accountability through the agency, which the Senate report identified as well.
Australian citizens’ satisfaction in the government’s digital services has dropped precipitously since 2011, and these and other services failures on the horizon will do nothing to increase it. A nation looking to implement digital transformation might wish to study the Senate committee report for lessons on what not to do, especially paying attention to former Digital Transformation Agency director Paul Shelter’s comments on the importance of successful ICT system delivery. It may then have a better chance of moving toward what Estonia has accomplished.