We seem to have been both naughty and nice this year. The Australian National Audit Office isn’t going to put its much-anticipated report on the future submarine program in our stocking before Christmas and we’ll have to keep our breath bated until January. That’s a shame, since the ANAO’s website says, ‘The objective of this audit is to examine the effectiveness of Defence’s administration of the Future Submarine program to date.’ That’s an $80 billion issue that many Australians would be interested in.

On the other hand, the ANAO has come through with Department of Defence Major projects report. As usual, it’s a very detailed piece of work that provides an unparalleled look at 26 of Defence’s major projects in 2018–19. Also as usual, it contains a lot of old news and some intriguing new news.

One piece of old news is that it doesn’t report on either the future submarine or future frigate program. Each of these now has an approved budget of around $6 billion, which would rank them at third and fourth in Defence’s portfolio, even before the government approves construction of a single vessel. Fortunately, we have unofficial advice from the ANAO that both will be in the next edition. Let’s keep our fingers crossed. The third plank of the domestic shipbuilding program, the offshore patrol vessel project, is included this year and is tracking well against its aggressive schedule.

Another piece of old news is that the MPR is hamstrung by terminology that seems designed to trigger the media’s worst ‘gotcha’ impulses. For example, the report lists $24.4 billion in ‘budget variations’ (again, old news) at table 3. But once we take out $14.1 billion in scope increases (i.e. the government agreeing to buy more stuff, such as 58 joint strike fighters), $1.3 billion in adjustments to scope and transfers between projects, $2.8 billion in indexation (i.e. inflation) and $4.4 billion in exchange rate variations (it really hurts when the Aussie dollar goes down against the greenback and euro), that leaves only $1.8 billion in real cost increases.

I should note that Defence presents a slightly different view of real cost increases in table 2C on page 84, but its total of $1.7 billion is broadly similar to the ANAO’s. Since the portfolio of 26 projects in the MPR is worth $64.1 billion, that’s a cost increase of about 2.6%. Moreover, the projects that have been completed and exited the MPR were collectively under budget by about $2.2 billion. So overall Defence is doing well on cost—but there’s a caveat I’ll discuss below.

Poor schedule performance is another piece of old news, and it’s not particularly good. The combined total of 691 months is bad (see table 4 on page 15). But it’s important to remember that while projects may be late, they are generally still delivering capability. The worst offender, at 112 months’ behind, is ‘Collins reliability and sustainment’, a large constellation of subprojects that’s been delivering a broad range of improvements since 2000. Many of them have been completed and the rest will be as the boats go through their docking cycles. The submarines have still been able to do their job, despite the delays.

Similarly, the second worst offender, the MRH-90 helicopter project, at 89 months, has delivered all 47 aircraft and Defence has been flying them. But not all of the individual capabilities under the project have been delivered, which is why the delay clock continues to tick. Some of these have real implications; for example, the lack of a fast roping capability means that Defence has had to keep the Blackhawk helicopter in service longer to fill the special forces role.

While we want the army’s helicopters to be able to support things like counterterrorism, this does raise the question of whether in every case it makes sense to keep a project going for years just to get the final few percent of capability. Sometimes it may be better to draw a line under it and apply the people and resources that are freed up to new priorities (autonomous systems, anyone?).

So, what’s the new news? The key worrying piece is that the F-35 project is under funding pressure. The MPR states that, ‘In September and November 2018, due to cost pressures, the Joint Strike Fighter project received government approval to transfer project scope of $1.5 billion to other phases of the Joint Strike Fighter program (none of which have been approved by government). There was no corresponding transfer of funds out of the project budget’ (page 26). Since the approved budget for the project is $16.5 billion, this means that 9% of the project scope has been moved off into some undefined point in the future, but Defence is hanging onto 100% of the budget to acquire the remaining 91%.

It would seem fair to classify this as effectively a $1.5 billion real cost increase since the capability Defence is getting is costing more. This brings the total of real cost increases to $3.3 billion. What’s more, the report states that, ‘Some level of known cost risk remains with a possibility that further scope transfers may be required’ (page 135). So more content may get pushed down the track.

The capability question is, what’s in the 9% that has been moved? And how important is it? That’s a little harder to determine. One element is the ‘beyond line of sight’ communication system (page 148). Since one of the key selling points of the JSF was its networked capability, this would seem to be a problem. Overall, there’s enough evidence in the MPR to indicate that, while the F-35 is still on track to achieve initial operational capability by the end of 2020, Defence doesn’t have a plan to get all of the elements required for final operational capability, scheduled for late 2023.

And then there’s the issue of the JSF sustainment system. I wrote earlier this year on the immaturity of the system. The MPR confirms that. Moreover, it states outright, ‘The F-35 future sustainment affordability has been affected by an increase in through-life sustainment cost estimates.’ With an F-35 flying hour already costing twice as much (page 63) as the classic Hornet it’s replacing, that’s some more bad news.

The government’s decision to pursue multi-decade domestic ship and submarine building programs sits oddly with decades of economic reform. Successive governments have striven to maintain competition in the Australian economy, and for good reason. Competitive markets drive efficiency, promote innovation and contain costs.

But what’s left of the local car industry attests to the unintended consequences of such policies. And of course that helps explain how these local mega projects—well over $50 billion and extending to beyond the mid-century—came to be. Even if they won’t replace jobs lost in the manufacturing sector, they’ll be large, symbolic and make for great photo ops.

But they are also going to provide the firms involved with effective monopolies. In practice, it will be so hard to change horses in midstream that the incumbents will have an assured position for years to come. Monopoly supply arrangements arise frequently in the defence sector, usually because there’s only one supplier of a unique piece of equipment. The cost, time and expertise required to develop alternatives is a high barrier to effective competition. The risk is that these new shipbuilding programs will become as inefficient as the government-owned shipyards of the 1970s and 80s.

In other areas where it’s infeasible to maintain competition—for example, in the provision of some utilities—the government regulates the industry to protect the consumer. For example, the US Government tightly regulates its defence industry to encourage innovation and efficiency and, just as importantly, to ensure that contractors don’t extract excessive rents, notwithstanding Mr Trump’s recent concerns about aviation projects. When we buy from the US we piggyback on America’s massive investment in research and development, and on its self-interested oversight of its defence firms.

We think that the government needs to be thinking much more about regulation of the monolithic enterprises it’s creating. The Department of Defence cannot effectively do the job itself; its agenda is not the same as the taxpayer’s.  We need a new arrangement.

A case study is illustrative. Until recently, submarine maintenance was largely a monopoly arrangement—albeit one involving a government-owned supplier ASC. Up until a couple of years ago, the results were increasing costs and diminishing vessel availability. It wasn’t until the government (acting in its triple role of owner, regulator and customer) commissioned the 2012 Coles Review that performance and efficiency started to turn around. The result has been dramatic, and Australia’s submarine availability is now approaching industry best practice (though costs remain an unanswered question). The key point is that it took intervention to shake Defence and ASC into action.

We don’t want to have to wait for things to fail again. There were many institutional and technical reasons behind the poor performance of the Collins fleet, but the core problem was that successive governments failed to take action. Given the sorry record of Collins sustainment, there’s a strong case for facing up to the challenge of regulating new monopoly suppliers right from the start.

That involves much more than ensuring that profit margins are kept in check, because it’s entirely possible for a firm to be inefficient and still make modest profits. And in the absence of countervailing pressures, a monopoly will tend to be less efficient than a firm subject to market competition. We’ve been making that argument for years, and we were heartened yesterday to see Senator Xenophon telling the AFR that he’s similarly seized with the idea.

Like the good Senator, we’re unpersuaded by the government’s response that the attention of the Australian National Audit Office is sufficient. We love the ANAO’s work in the defence space but, by definition, their audits are retrospective. Instead, we need to get these things right up front, and to have the ability to continually monitor progress.

Nor are we convinced that our own organisation is the right fit, as flattering as it was to be mentioned (along with the Productivity Commission) as part of the putative oversight mechanism. We’re happy to help to define the job of the eventual regulator, and we have some ideas. But the sums and timescales involved suggest to us a need for an enduring stand-alone office, perhaps on a statutory basis, that can push against the tide of political or bureaucratic enthusiasm when required.

This piece also appears in today’s Australian Financial Review and is reprinted here with their kind permission.

Having to resist the urge to take a cheap shot is an occupational hazard in my business, and one that I don’t always successfully avoid. Once such was the title I gave to a review of Project LAND 121 ‘Overlander’ (replacement of Army’s 7,000 field vehicles and trailers). Published in the 2010 Budget Brief, I called it ‘How many DCPs does it take to buy a truck?’

One of my first jobs in Defence back in 1994 was to conduct analysis in support of Overlander (and the then-Army 21 program), and I wondered even then why this was being made so complicated. To see it still making its way through the system 16 years later really had me scratching my head. Five years later, Overlander is now in its third decade; an acerbic Tweet last week observed that it’s now been running several times longer than WW2.

Two things happened last week that made me think about it again. First, I heard a talk from DMO’s Land Systems Division on the project at ASPI’s land forces conference. Then, I got a copy of the Australian National Audit Office’s new report on LAND 121 Phase 3B (replacement of medium and heavy vehicles). Those two sources allow us to piece together an interesting, if doleful, story.

The conference talk very much made the point that it’s not just about trucks, because the modern Army requires its vehicles to be integrated into its wider communications and command and control system. As the land forces move into the digital age, each vehicle becomes a node in Army’s network. But we know from far too many previous experiences that systems integration is a fraught part of projects. In my talk at the same conference, I noted the time it has taken to get the Tiger helicopter to communicate effectively with the rest of Army.

Like the helicopters, the schedule for the acquisition of vehicles has slipped as well. When the first second-pass approval was made back in 2007 the Final Operating Capability (FOC) was initially expected to be in 2016. (Note: ‘first second pass’ isn’t a typo: after a project reset, there was a ‘second second pass’ in 2013, with an ‘interim pass’ in between for good measure.) After the interim pass, FOC was expected in 2019. Now the estimate is 2023, seven years late (like Tiger). Of course, that could slip too as we get to the build, systems integration and delivery. ANAO notes (as usual) that cost and schedule risks tend to rise at the systems integration phase, adding that ‘Defence will need to maintain a focus on managing the remaining integration issues’. The ANAO must have their fingers crossed when they write these words, given the sorry history of this and many other projects.

The plethora of passes follows the cancellation of contract negotiations with the preferred supplier identified through a tender process in 2005–2007. The process suffered from a litany of shortcomings, but at the heart of the problem was another ‘old faithful’ error: ‘Defence originally considered that the medium and heavy vehicle acquisition was a relatively low risk military off-the-shelf procurement’. Read the report for the story, but it was rushed, with insufficient testing and poor communication of assessed risk. Once reality set in, another tender process was launched, and more time ticked away.

As a result, we’re also seeing other time honoured effects at work—the cost to capability and the operating budget of keeping older platforms in service. ANAO make it clear that there’s been a significant financial price to pay:

The delays in the medium and heavy fleet acquisition have placed considerable pressure on the existing Unimog and Mack vehicle fleet, which has now well exceeded its life-of-type and is increasingly difficult and costly to maintain. …the average sustainment cost per vehicle for the Mack fleet has increased by some 80% between 2009–10 and 2013–14, reflecting the advanced age of the fleet and difficulty in acquiring spare parts.

And it’s not just a dollar cost; capability is getting shaky:

… one third of the South Queensland Mack [fleet] is in the process of being written off due to excessive rust following annual technical inspections. …based on vehicle availability and the number of platforms written off each year, the Mack fleet will have difficulty supporting force generation activities from 2016 and in meeting strategic guidance from 2019.

It’s hard to know what to say about all this, and it’s all drearily familiar. The Army acquisition folk talking last week seemed to have a good grasp of what’s required, but the system as a whole is seemingly incapable of producing a sensible and timely outcome. The First Principles Review is throwing the whole lot up in the air, and it remains to be seen what we’ll get next time around, but the precedents aren’t good. I can’t help but think it’s time to take a much bolder approach—perhaps finding a private sector complex acquisition specialist, telling them ‘off the shelf by strong default’, and outsourcing the lot.