Can DTA succeed? A revue of ICT reviews

article-body

“Technology is power.”

Those were the three simple words put to me put to me by a now retired senior federal bureaucrat when asked why the federal government’s track record of strategic reform for $6 billion a year in systems, software and hardware has been so chequered and fraught with difficulty over the past 20 years.

It’s 2017 and, like many years previous, another review of federal technology strategy, spending, deployment and execution has been launched in an effort to get Canberra’s sprawling IT estate in order and delivering real service improvements and better value for money.

And like so many past IT reviews, there’s a burning recognition that things need to change – and for senior public servants and industry alike, a nagging question whether much will.

It’s fair to say no government has ever placed internal technology reform so far up the policy agenda as the Turnbull government, especially through the creation of the Digital Transformation Office cum Digital Transformation Agency within the Department of the Prime Minister and Cabinet to now prosecute a whole-of-government digital technology agenda previously vested to the Department of Finance.

Yet the frank and fearless departure of the PM’s personal choice of digital evangelist, Paul Shetler – who has since repeatedly broadsided the bureaucracy from the sidelines – has left many outside tech circles in government and industry wondering what really lies ahead and whether there is still sufficient courage and tenacity to effect real change.

After all, reviews of government technology spending are hardly a novelty.

A brief trip down memory lane, or a ‘revue of reviews’, certainly provides some lessons and pointers on where things may well head next. And where they shouldn’t.

Revue of reviews

Most incoming governments are keen to put their stamp of authority on the public sector’s tech estate, especially when they feel their predecessors used it as an ideological bludgeon to prosecute their agenda. Take the whole-of-government IT outsourcing push of the 1990s that neither side of politics wants to ever revisit, and with good reason.

When entire large departments divest control of their technology strategy to outsourcers, the results make #censusfail and some of Centrelink’s recent woes look almost bearable. Even before the creation of the Department of Human Services under John Howard, there was no stomach to outsource welfare technology because of the inherent risks and dire consequences that could hit society and government when pensions don’t get paid.

The fact that Centrelink’s core system from the 1980’s is still in place today is testament both sides of politics in Australia using a ‘hospital-pass’ to manage-up the difficulty and risk associated with such upgrades to successors. The very fact that it’s also now social services policy to recover ‘overpayments’, rather than initially underpaying people, is also telling.

As recognition grew over a decade that wholesale tech outsourcing within government had under performed — in fact had constrained and sometimes captured the public service — a bid to reassert internal control and influence over the bureaucracy’s systems came from the creation of the Australian Government Information Management Office (AGIMO).

Son of Gershon? Please spare us

Initially an exercise in gentle (read unenforceable) persuasion from the inside, the arrival of Labor after Howard’s defeat in 2007 was expected by many to put a broom through existing government ICT arrangements. It probably would have if not for the distraction of the Global Financial Crisis.

With telecommunications and broadband made proud front and centre of Labor’s technology agenda (applications? pfffft), the prosecution of post-outsourcing tech reform with government fell to the Department of Finance, which bowled up British cost-cutter Sir Peter Gershon when finance minister Lindsay Tanner requested a stocktake of the opportunities for improvements and savings.

The savings came. Around $500 million was found, earmarked for ‘reinvestment’ and swiftly clawed back by Treasury.

But the big opportunities for strategic and architectural reform were missed. Horribly missed.

Cloud, mobile computing and the massive shift to online service delivery may as well not have existed in the Gershon Review. And they largely didn’t until industry and some more courageous agencies kicked up a stink.

At a time when government needed to procure on-shore cloud quickly to keep pace and cut costs, a moratorium on new data centre spending was put in place, hobbling industry investment in new facilities in critical locations like the ACT.

If there was a single event that set the bureaucracy badly behind in harnessing technology to improve services, the Gershon Review’s dated, non-strategic focus on squeezing back-office ‘business as usual’ systems is arguably the tipping point.

If Paul Shetler found stiff resistance to a mandate for reform, part of it certainly stemmed from the immediate post-Gershon experience of negative consequences that followed backward-looking mandates from outside being brought to bear. Circle the wagons.

In the run-up to Labor losing office in 2013, even Communications Minister Stephen Conroy was openly questioning the need for responsibility for government technology to rest with Finance in his own government, an issue Turnbull has dealt with decisively even if the effects are yet to play out.

The reviews you don’t see

If the political class and senior bureaucrats had been in any doubt about the underwhelming impact of the Gershon Review or the Australian Government Information Management Office in propelling the modernisation of government services, a subsequent review of the Gershon Review’s impact didn’t take any prisoners.

Penned by former National eHealth Transition Authority chief Ian Reinicke, the document didn’t spare rod on the effects of Gershon’s review or the performance of the governance models it suggested.

“The casualty of this approach has been an inability to develop an ICT vision and the strategies needed to support it,” Reinecke said at the time.

Also lacking were “the vision and supporting strategies needed to continue to drive whole-of-government initiatives and ICT public sector reform.”

Even before the creation of the Digital Transformation Office, there were major technology stocktakes in play, the results of which wouldn’t make for pleasant reading in the morning’s press clippings.

Too big to fail

One bold evaluation actioned by Finance was a review of the value for money the government was getting from its ERP (enterprise resource planning) software investments – the enterprise-wide systems that can span across everything from asset management and operational reporting to finance, HR and payroll. They’re also are well known for creating technology vendor-lock-ins and undesirable proprietary dependence by organisations that can be costly to shift from.

Fortuitously, the announcement of the ERP review by Finance came literally days after the election of the Abbott government in September 2013.

“The devolved nature of the public service has resulted in Commonwealth entities, over time, implementing various, often customised ERP systems or systems for key functions. This has afforded Commonwealth entities the opportunity to pursue systems and processes that best meet their needs. However the fragmented approach may also have reduced the overall efficiency and effectiveness of government operations,” the notice to industry from Finance said.

“In a fiscally constrained environment, government investment must be smarter, and more innovative. This Project is an opportunity to undertake a whole-of-government analysis of ERP systems, to determine a strategic approach for the medium to long term.”

What the conclusions or the results from that review were have not been publicised, but the opportunities for consolidation, standardisation and centralisation are apparent enough.

The pre-review review … or just keep asking

There are some clues to the findings of Finance’s ERP probe in another even less publicised IT review released by the department.

Stemming from the Coalition’s promise in its 2013 election platform for a complete review of government IT systems, the document originally delivered in December 2014 was only made public under Freedom of Information in January 2016.

Referencing the ERP Review, Finance’s whole-of-government stocktake concluded “there are currently more than 200 unique ERP systems across the public sector.”

Despite the recommendation to centralise and standardise such systems (separate to the Shared Services play), it was estimated that “standardisation and automation of transactional business processes could avoid costs of between $233 million and $465 million over seven years (the average life of an ERP system).”

“An enabler to this is the consolidation of support staff and centralisation of procurement; on its own it is estimated this would reduce costs by $118 million over the forward estimates ($99 million from reducing staff and sharing application development activities and $19 million from centralising procurement and contract management and standardising technical configuration),” Finance’s audit concluded.

Those numbers aren’t exactly of a magnitude that would tempt even hardened cost-cutters to try their luck.

Especially given the high number of “unique” systems they touch. Moreover, many of those systems will have been bedded down and largely paid for, making them less appealing to ‘rip-and-replace’ unless there’s a pressing need.

But what they do tell you is that the level of disparate, disconnected and stand alone systems is simply far too high to be sustainable and their true legacy costs will leach out one upgrade at a time unless lasting reforms are made.

Where to now?

Finance is now quietly running the ruler over all the vendors seeking whole-of-government concessions. But the new reviews by the DTA will represent the first time a federal government has run a ruler over its tech estate with a central and strategic policy agency at the helm, as opposed to the head of government housekeeping. Moreover, technology and the services and efficiencies it can deliver are no longer a ‘back-office’ issue, even if there are still plenty of back-office systems.

As New South Wales has demonstrated, swift improvements to service delivery can be made if genuine reform is mandated from the top and followed through with logical and efficient changes to business processes and structures.

At a public level, the elephant in the room remains to what extent a personal digital identifier will be needed for individuals – irrespective of what the government calls it – for end-user improvements to occur and whether business will also get to participate within this identity ecosystem.

Perhaps most pressing is the need to reform the massive levels of complexity that underpin the business rules that any technology relies upon in areas like welfare and social security.

That will require not only political will, but an informed and judicious risk appetite from the public service as well.

The longer reform is put off, the harder it gets.