Megaprojects & mega holes: sublime lessons for Australia’s submarine enterprise

Megaprojects often go bad because of conscious decisions by political actors working in their own, often short term interests.

Written by

Marcus Hellyer

Earlier this year I started looking at what Bent Flyvbjerg’s analysis of megaprojects can teach us about the Department of Defence’s gigaproject: its efforts to acquire a nuclear-propelled submarine capability (SSN). I’ve wanted to get back to Flyvbjerg for some time, and after a few distractions here is the next instalment.

We will recall that Flyvbjerg has studied and assembled a database of tens of thousands of megaprojects of many different kinds from around the world. One of his key conclusions is that there is a ‘iron law’ of megaprojects: they are ‘over budget, over time, under benefits, over and over again.’ According to Flyvbjerg, 91.5% of megaprojects don’t deliver on time and budget. Once you add the third side of the ‘iron triangle’ of project management, namely the benefit the project is meant to deliver (or capability in defence terms), Flyvbjerg notes that increases to 99.5%. That won’t come as a surprise to the ‘beneficiaries’ of many megaprojects, such as Sydney’s recently opened $3.9 billion Rozelle interchange.

Moreover, and this is a key point, when megaprojects go bad, they don’t go bad in a linear fashion; they go really bad in an exponential fashion.

That doesn’t mean we shouldn’t do megaprojects. They are a fact of life, and indeed essential for delivering outcomes necessary for modern life. And they aren’t doomed to underdeliver—they can be done well; Flyvbjerg’s comparison of Frank Gehry’s Guggenheim museum in Bilboa with the Sydney Opera House illustrates this. But to do them well it’s vital to understand why megaprojects go wrong and to address those factors up front.

Considering that defence projects are amongst the worst offenders and nuclear power and waste disposal are the worst offenders (next to the Olympics), we should heed his advice in the case of the SSN enterprise which incorporates elements of all three.

Since both major parties are committed to this megaproject, it’s vital to understand the key factors that result in megaprojects going wrong and consider whether Australia’s enterprise to acquire nuclear-powered submarines (SSNs) exhibits some of those features. Only then can they be addressed.

Victims of ‘the 4 Sublimes’

A key reason for why megaprojects go bad is that their advocates and managers become captured by what Flyvbjerg terms the ‘four sublimes.’ The first is the technological sublime: ‘the rapture engineers and technologists get from building large and innovative projects with their rich opportunities for pushing the boundaries for what technology can do, like building the tallest building, the longest bridge, the fastest aircraft, the largest wind turbine, or the first of anything.’ Already we can see this at play in the SSN enterprise in the rapture its supporters express for the capability nuclear propulsion provides that apparently is so extraordinary it is worth the $368 billion opportunity cost involved in acquiring this rather than other capabilities.

The next is the ‘political sublime’, which he describes as:

‘the rapture politicians get from building monuments to themselves and their causes. Megaprojects are manifest, garner attention, and lend an air of proactiveness to their promoters. Moreover, they are media magnets, which appeals to politicians who seem to enjoy few things better than the visibility they get from starting megaprojects; except maybe cutting the ribbon of one in the company of royals or presidents, who are likely to be present lured by the unique monumentality and historical import of many such projects’.

Certainly in the Australian context, no politician can resist the opportunity for photo opportunities in a hi-viz vest and helmet in a shipyard. And already we are being bombarded with statements of the historical, ‘once-in-a-generation’ import of the enterprise.

The third is the ‘economic sublime.’ This is:

‘the delight financiers, business people, and trade unions get from making lots of money and jobs off megaprojects. Given the enormous budgets for megaprojects there are ample funds to go around for all, including contractors, engineers, architects, consultants, construction and transportation workers, bankers, investors, landowners, lawyers, and developers’.

We can already see this sublime at work, not just in the enterprises that will be building the submarines, but in all of the proliferating enterprises that hope to monetise AUKUS. Considering the timeframes involved in delivering SSNs, there will be sublime opportunities to make money from Pillar 1 for decades. On top of this there are the fantastical claims being made—on the basis of no evidence whatsoever—about the economic benefits that will be delivered to the nation by $368 billion of spending on one military system, these claims of course deeply intertwined with expressions of the political sublime.

Flyvbjerg’s final sublime is the ‘aesthetic sublime’. That’s ‘the pleasure designers and people who appreciate good design get from building, using, and looking at something very large that is also iconically beautiful, like San Francisco’s Golden Gate bridge or Sydney’s Opera House.’ While we hear few statements about the beauty of SSNs, in some ways, this echoes the raptures created by first sublime, namely the supposedly unique capability provided by SSNs.

We can see that the rapturous sense of the sublime that characterises megaprojects but leads to people turning off their critical faculties (or being directed to turn off their critical faculties so as not to be disloyal or defeatist) is clearly already well entrenched in the SSN enterprise.

That’s a problem, because the SSN enterprise is also marked by the factors that mark megaprojects that go badly off the rails. Flyvbjerg identifies 10 them. We’ll list them here and assess with traffic lights whether the SSN enterprise demonstrates those characteristics. Red indicates that characteristic is deeply embedded in the program.

Table 1: Megaproject characteristics and their presence in the SSN enterprise

You are free to challenge any of these assessments and the colour of the traffic lights, but in my view, any honest, informed assessment of the enterprise would reach a broadly similar conclusion, namely virtually all of the factors that can potentially lead megaprojects to underdeliver are present to a high degree in the SSN enterprise.

That’s not a good foundation, particularly when it’s combined with the four sublimes, which cause people to make bad decisions, or to stay silent in the presence of bad decisions. One of Flyvbjerg’s key arguments is that bad decisions are not merely the result of cognitive biases, of falling for the well documented fallacies that permeate human thought, such as the sunk cost fallacy. Flyvbjerg acknowledges that cognitive biases are part of the problem but only a part—we have to acknowledge that political bias is at work as well. Megaprojects go bad not merely because of people unconsciously falling victim to their own cognitive biases—they are the result of conscious decisions by political actors working in their own, often short term interests.

Getting the project started is where much of the misrepresentation occurs. For example, as Flyvbjerg notes, initial public statements of cost ‘aren’t intended to be accurate; they are intended to sell the project. In a word, they are lies—or spin, to use more polite language.’ We wouldn’t agree that is the case with the SSN enterprise; to its credit the Government has published an initial cost estimate that doesn’t appear to have been deliberately understated. That does however mean it is an enormous $268-368 billion figure. But there are other aspects of the ‘optimal pathway’ that are very optimistic around schedule, workforce, and industrial capacity—not to mention claims of the benefit to be delivered.

Flyvberg argues that the key to success is to plan slowly rather than rushing into action. Planning properly costs little. But this isn’t how megaprojects often proceed. Because megaprojects are prone to under delivery, stakeholders can reconsider, for example, when early cost and schedule estimates are shown to be hopelessly (or deliberately) optimistic. Since the project’s boosters don’t want commitment to waver, their aim is to dig ‘a big hole’ as soon as possible and let the sunk cost fallacy take effect. As former San Francisco mayor Willie Brown stated, ‘The idea is to get going. Start digging a hole and make it so big, there’s no alternative to coming up with the money to fill it.’

That’s why its important to be aware of both cognitive and political biases. Cognitive bias is not the cause of the drive to start digging the big hole as soon as possible, rather political bias is. It deliberately exploits the cognitive bias of the sunk cost fallacy that makes us reluctant to walk away from bad money and instead prefer to send more good money after it.

So far it may look like Australia’s SSN enterprise is still in the planning phase. We haven’t dug many actual holes; no construction work has started on shipyards or drydocks, for example. But the financial holes are being dug; major spending commitments have started with payments in excess of $4 billion each going to the United States and United Kingdom. Those sums alone would be Defence’s third biggest project ever behind the Hunter-class frigate and the F-35A, so the sunk cost fallacy will have lots of sunk costs to work on, particularly in the absence of any mechanism to have those funds reimbursed should no SSNs eventuate. That’s in additional to the sunk costs of individual and national reputations.

Ensuring megaprojects stay on track is not just a matter of planning carefully and challenging assumptions to ensure cognitive biases are exposed. Rather there needs to be public scrutiny to show how political biases are at work. Shedding light on the enterprise can only help it rather than hijack it. It’s not only Australian commentators who argue for this. Colleagues with deep experience of defence megaprojects in the US are stunned not only at complete absence of public scrutiny that went into the original AUKUS decision, but at the absence of real public disclosure since then.

The AUKUS enterprise has started to dig its big hole, but if it wants to fill it with actual submarines in a relevant timeframe, it needs to do a much better job at sharing information with key stakeholders, not least of all those people who are paying for it.

More information sharing means that the AUKUS enterprise needs stronger political leadership, not just the occasional press conference in high-vis gear. How much of the Prime Minister’s time is this megaproject worth? An hour a week, a day a month? The project is of such a scale that our political leadership should constantly have its attention on it and draw on the best advice, from inside and outside of Defence, testing every step of the enterprise. But three years into AUKUS, the endeavour seems to be treated as a standard, if large, Defence project. Officials do their best to avoid answering questions at Estimates and other public fora and in other respects shun public explanations and testing of their work. If the SSN is to be successfully delivered it needs to be a politically-led, national effort—and one whose cognitive and political biases and whose capture by the four sublimes are open to scrutiny. Burying it behind Defence’s walls and under its processes is a recipe for failure.

SHARE THIS ARTICLE