A super public service department could group several back-office functions. Photo: Michel O Sullivan
The push is on in Canberra for government departments to once again embrace a shared services model of procurement with a Commission of Audit report expected to add weight to calls for a super-department to aggregate back-office functions.
In theory, a public service super-department could eliminate duplication of work and garner greater purchasing power with potentially seductive savings, but the potential for disaster is huge, information technology industry analysts warn.
Gartner research vice president Jim Longwood says any plan to create a mega shared-services department in the capital is "doomed to failure".
Shared services have the greatest chance of success with a hub-and-spoke model, which organises agencies into small clusters around a lead agency, Longwood believes.
This arrangement keeps the number of stakeholders low, minimising opportunities for in-fighting and white-anting, Longwood says.
Shared services were traditionally justified by business cases promising operational efficiencies, standardisation and cost savings but sceptics have their pick of examples of shared services that have failed to deliver.
In 2011, the West Australian government disbanded its Office of Shared Services centre after an Economic Regulation Authority review found the project was over budget and unlikely to deliver promised savings of $57 million a year.
Instead, the ERA report found, the project had cost $401 million and achieved minimal savings for the 59 agencies it serviced.
The country's costliest and most notoriously bungled ICT project, the Queensland Health payroll upgrade, was developed under the auspices of a shared services group.
Originally with a budget of $98 million, and due for completion in July 2008, the project was the subject of a royal commission last year and is expected to cost taxpayers $1.2 billion by 2020.
The Newman government has since announced plans to sell its ICT service bureau CITEC and is evaluating the future of Queensland Shared Services which runs shared finance and HR/payroll software.
"IT shared services is great in theory but it is costly and hard to make the transition and implement," IBRS analyst Sue Johnston says.
For some agencies, entering shared arrangements can mean ratcheting service expectations down a notch, hardly a compelling reason for CIOs and senior bureaucrats to support the model.
"Because agencies have different operating environments and service levels, those with higher service expectations will probably have to accept a lesser service than now," Johnston says.
"In addition to not offering an optimised service, very little is done to develop a high-performing customer service culture, which is even more important when control has been removed from the agencies."
Allocating limited resources to projects equitably is also a juggling act, she warns.
"Disappointments are inevitable as it is impossible to please everybody. Shared-services management is constantly on the defensive to justify priorities, damaging morale and leaving little time to exercise leadership."
The collective decision-making process typical of multi-agency projects does not suit shared services, says Paul Wilkins, a former WA government CIO and now public sector consulting director at Ajilon.
"In good project management you have just a few people with decision-making power, within defined parameters," he says.
When no one and everyone is in charge, no one is powerful enough to pull the pin – even when it's clear things are going pear-shaped. WA's shared service centre limped along for four years before a decision was made to scrap it, Wilkins says.
"No one is brave enough to kill it – it gets a life of its own," he says. "A fundamental rule of project management is that, if it's not working, kill it early."
Sensible idea for savings or disappointing debacle? Share your experiences of shared services in the comments.