Policy morass, legacy haunt $1bn DHS project

The Department of Human Services (DHS) raised eyebrows with its recent request for expressions of interest (REOI) for a $1bn overhaul of its core systems, but government watchers believe the successful completion of the overhaul will take a lot more than just throwing money at the problem.

The details of the DHS upgrade were highlighted in Budget papers earlier this year, when the agency noted that the Welfare Payments Infrastructure Transformation (WPIT) Programme would be “one of the world’s largest social welfare ICT system transformations” and was designed to provide faster, more connected and automated digital services for customers as well as giving staff a “modern ICT platform that makes it easier for them to do their jobs”.

The system will, among other things, provide a platform through which millions of Centrelink recipients will engage with DHS to manage their eligibility for unemployment, family benefit and other social-services programs. WPIT will replace existing systems — in particular the current Centrelink online accounts environment, which has become so problematic for users, particularly with the release of new mobile apps branded “truly awful” by some users, that the agency has had to publish schedules of planned outages.

Yet even as the WPIT project (outlined in April by stakeholder ministers) kicks off, early scepticism is already rearing its head and there are concerns about scope creep in the estimated 7-year transformation project.

This year’s Budget papers allocated $169.3m over the next two years to WPIT, which will replace the legacy Income Security Integrated System (ISIS) whose 30 million lines of code handle delivery of $100bn in payments to 7.3 million people annually. ISIS has been progressively developed and upgraded since 1983, combining its original Model 204 database with more than 350 additional components added in the intervening years.

Paired with an enormous amount of policy change, WPIT is easily among the most complex projects ever attempted within Australia’s public service. In recognition of this, the government has argued that its implementation plan has been guided by “extensive consultation” with government, private-sector and ICT-delivering organisations including the UK government, Rio Tinto and Commonwealth Bank.

In May, the Department of Human Services appointed a permanent deputy secretary, Tamati Shepherd, to guide the program towards its scheduled 2022 completion. Yet Glenn Archer, a former Australian government CIO who is now serving as Gartner’s research vice president for public sector, believes the policy requirements around WPIT reflect the inherent challenges of the government’s digital transformation agenda — and will handicap efforts to quickly and efficiently rework the system.

“If you look at the range of different policies that we’ve created in the social welfare and taxation domains over the last 10 to 20 years, often online services or digital enablement of those services has been an afterthought,” he said, citing as an example the processes by which Centrelink recipients report their income to the agency.

“It’s one of the highest volume services that Centrelink run,” he said. “It’s all very straightforward, and the nature of the transaction that’s been built by Centrelink and other parties is sound. But the policy creates enormous complexity, requiring the citizen to report on a specific day within certain hours. You find yourself having to shoehorn an existing program into an online service delivery model for which it was never really considered.”

The sheer size of the program may pose other challenges for the project, which is being put to market under the lingering shadow of the Queensland Department of Health’s disastrous SAP payroll implementation, which has blown out budgets and recently landed the department in court against implementer IBM.

Amongst other issues, such projects will heighten the government’s aversion to risk — a key challenge that must be managed in any megaproject, where ambition and reality often collide to produce epic failures.

“In many instances, large projects get started on the basis of a well-intentioned proclamation from a group of senior leaders, who have a sense in mind about what they can afford to spend,” project management expert Jim Porter previously told GTR, citing research suggesting many big projects fail horribly due to factors including poor upfront planning and constraints on execution.

“People say that they need the project on a certain date, and that becomes the fundamental underpinning of the status of the project,” he explained.

“We need to learn from experience, but it seems we don’t do a particularly good job on that when we get into a very large project. People tend to do the same thing they did last time... there’s a real epidemic underway with respect to these large project failures.”

Just whether the project will doom yet another government agency to an impossibly complex IT implementation with unpredictable outcomes — and what vendor is willing to take it on, particularly in the wake of the Queensland Health disaster — will become clear after the REOI closes at the end of October.

Image courtesy Amanda Slater under CC