The federal government needs to adopt the 'measure twice/cut once' philosophy that often gets ignored in national projects, Oxford University professor Bent Flyvbjerg says.Orbon Alija/Getty Images
This past summer, a small squad of Ottawa’s top civil servants were busy assembling a bureaucratic dream team that will be known as the Major Projects Office, located in Calgary and headed by oil industry veteran Dawn Farrell. Prime Minister Mark Carney’s in-house disruptor, Michael Sabia, was involved, as were top-level folks from the provinces, territories and First Nations.
Farrell’s crew will be trying to figure out which “nation-building” projects should be green-lit as part of Carney’s pledge to streamline approvals, unleash internal markets and make Canada less dependent on U.S. exports. “There’s two pieces to this,” says Matti Siemiatycki, professor of geography and planning and director of the Infrastructure Institute at the University of Toronto. “One is, you have to make sure you’re building the right things. And two, you have to make sure they’re delivered well.”
No decisions have been made so far, but the government’s official pronouncements mention megaprojects such as highways, railways, ports, airports, oil pipelines, critical minerals, mines, nuclear energy facilities, and electricity generation and transmission systems—a.k.a. “PONIs,” as they’re known around the capitol: “Projects of National Importance.” It’s Canada’s Got Talent: Infrastructure Edition. What could go wrong?
Plenty, according to Oxford University megaprojects scholar Bent Flyvbjerg. He’s spent three decades studying the forces that bedevil such schemes, which are so ambitious that even the moving parts have moving parts. His research shows that an overwhelming number of megaprojects fail to deliver the desired economic benefits, much less achieve their budget/timeline targets.
“Only 8.5% of projects hit the mark on both cost and time,” Flyvbjerg wrote in How Big Things Get Done, a 2023 book (with co-author Dan Gardner) that distills years of research on more than 16,000 big projects in a wide range of sectors in 136 countries. “And 99.5%...go over budget, over schedule, under benefits or some combination of these.”
Opinion: What Carney’s underwhelming list of approval-ready national projects tells us
A Danish economic geographer and business school prof who founded a consulting shop called Oxford Global Projects, Flyvbjerg offers another caution. “It’s a warning sign that Canada wants to start so many projects at the same time,” he says. “Because one project is difficult, it’s not like 10 projects are 10 times as difficult. It’s 100 times as difficult.” The complexities, in short, multiply exponentially as governments scramble to marshal all the contractors, suppliers, equipment and workers to see through construction schemes as varied as LNG plants, ports and hydro lines.
Much of Flyvbjerg’s scholarship is grounded in the “measure twice/cut once” common-sense philosophy that often gets sidelined when vested interests and politicians control the narrative. His first commandment is, “think slow, act fast,” i.e., spend the time to plan properly, iterate those plans, consider alternatives and so on, all of which is far less expensive up front than when contractors are dug in and spending scads of money. Only after all the meticulous analysis is completed and verified should proponents put together experienced project teams and drive to the finish line.
One of his favourite examples is the way Pixar makes features. All the fussy creative planning takes place long before the expensive animation machinery kicks in. By contrast, bespoke infrastructure schemes, which are invariably touted as bold and innovative, demand customized and unproven technologies that greatly increase the financial risks. He cites the largest-ever boring machine that Seattle ordered up as part of a multi-billion-dollar effort to bury its elevated waterfront highway in a massive single tunnel; the untested machine expensively jammed part way through after it chewed into buried steel cables.
Flyvbjerg’s most rigorous empirical contribution, however, is something he’s dubbed “reference class forecasting” (RCF). Most big projects involve suspiciously inexpensive budget projections and upbeat timelines. That’s in part because politicians demand them, but also due to the vagaries of trying to quantify all the things that can go wrong and then assigning dollar values—an exercise, he says, that is a mug’s game for home renovators and government agencies alike.
Instead, Flyvbjerg and his researchers ferret out the actual final cost data on similar completed projects, make the necessary adjustments (i.e., for scale, inflation, etc.), and use those data as their benchmarks. In other words, instead of over-hyped best-case scenarios, RCF literally accounts for the myriad real-life problems that derail large infrastructure ventures. Researchers “have found that RCF actually results in more accurate forecasts,” he adds.
Flyvbjerg is a big fan of transparency—budget forecasts and the assumptions behind them should be widely available and subject to peer reviews. Lastly, he says there’s much to recommend highly scalable ventures that are modular in design: solar farms, transmission lines, server farms, etc. It’s the snap-together ethic of Lego, Denmark’s greatest invention.
Flyvbjerg is not naive when it comes to the raw power of inspiring stories about heroic megaprojects (e.g., the Hoover Dam), as well as the pressures imposed by politicians who espouse a damn-the-torpedoes outlook and demand to see shovels in the ground ASAP. (Prime Minister, are you listening?) The California high-speed rail fiasco, which U.S. President Donald Trump called out in July when he pulled US$4 billion in federal funding, is merely the most egregious example.
Reality check: The first ‘major projects’ for fast-tracking are already on the fast track
To tamp down the chaos that can envelope such projects, governments need to build their forecasts as rigorously as possible, ask hard questions about whether they or their partners can secure the necessary materials and labour, and question whether there are other means of achieving similar goals, says Flyvbjerg. Yes, pundits and opposition parties will shout that nothing’s happening (at least as far as they can see), but early course corrections cost a whole lot less in the long run.
Canadian infrastructure experts like Siemiatycki add a few other layers, such as the importance of incorporating Indigenous feedback early on to address constitutional issues that will get litigated and could pose existential threats to these big projects. Economic viability is another. “A lot of the projects don’t yet have financial backers and proponents,” he says. “The federal government already bought one pipeline and spent billions and billions and billions to build it.”
Drew Fagan, a former Ontario deputy minister for infrastructure, also cites the delicate matter of locating the optimal trade-off between political interests and the purported long-term benefits of a megaproject. “If you’re going to prioritize, what’s the balance between the policy analysis and the political imperative?” he says. “Nobody’s going to say that it shouldn’t be a decision of cabinet or a minister or the prime minister. Everyone understands this. You have to broker it.”
Flyvbjerg’s somewhat Spock-ish retort is that governments tend to make more rational decisions when they get better data and disinterested analysis from their policy teams. Siemiatycki, who has collaborated with Flyvbjerg, points to Australia’s national infrastructure agency as an example of how one state figured out how to shield megaproject delivery from excessive interference. “They publicize the criteria and how they’re going to apply it so it’s not this backroom negotiation,” he says. “It’s a much more transparent process.”
Oxford Global Projects, for the record, is advising on Ottawa’s high-speed rail scheme (known as Alto), a federal IT project for Global Affairs Canada and an expansion of the terminal shipping facility at the Port of Vancouver. There’s no indication that Farrell’s Major Projects Office has come a-calling—at least not yet.
Flyvbjerg keeps his distance from the consulting practice to make sure he stays in his own lane. “I’m a scholar before anything else,” he demurs, then adds, “I’m happy to share my knowledge.” To the federal mandarins evaluating all the PONIs that promise to steer the HMS Canada toward more bountiful harbours, his offer seems like a door well worth opening.