Challenger Cities EP100: How Big Things Get Done with Bent Flyvbjerg

This is the 100th episode of Challenger Cities. A bit of a milestone right? This thing started as an attempt to explore and make a difference specifically in Toronto, but early on took a bit of a turn to have conversations about cities and infrastructure that didn't seem to be happening anywhere else … … the ones that fall between planners, architects, engineers, economists and what 'regular' people want in cities. I didn't expect to still be at it, let alone to have talked my way through 100 guests to get here. I think Bent Flyvbjerg makes a fitting person to mark the occasion, having spent a career exploring perhaps one of the most consequential questions in the business of cities. How big things actually get done? And so often they aren't? Bent Flyvbjerg can point back to how his career began with a bridge that didn't go to plan. The Great Belt Link, joining two halves of Denmark across a stretch of water that had kept them apart for centuries, cost 17 (seventeen!) times more than anything the country had built before, and it ran into trouble more or less on contact. What caught his attention was not so much the overrun. Big projects do that. What caught his attention was that when he went looking for the data to explain why they overrun, there wasn't much of it. People don't really like keeping scores of their failings, so Bent started collecting them, and the database he opened that day now holds more than 24,000 projects. It makes him something close to the only person who can tell you, with a straight face and the spreadsheet, how likely your tunnel/power station/railway/hospital is to cause a big hole in your finances, or political credibility. One of the rules of thumb in his book with Dan Gardner (a Canadian, he was quick to point out) is'think slow, act fast'.It has the ring of a self-help or leadership airport paperback, however it is actually robust advice for building any major project. Acting fast is a privilege reserved for teams who have first taken the time to work through the idea, implications and desired outcomes slowly. That time affords the ability to develop understanding and competence, testing things out through simulation and experimentation. You build the thing in miniature, in the form of physical or digital models, and you have all your 'expensive' catastrophes there, where they are not exactly free, but at least cheap. 'It's much cheaper to make your mistakes in a simulation on a computer than to do it on the construction site.' Bent's favourite example is Frank Gehry, whose buildings look as though someone sat on the model, and who can nonetheless put them up quickly and close to budget, because by the time the first materials turn up he has already made every mistake twice, in toy models, and complex software. It leaves us the obvious question of why, if this works so well with one of the world's most famous architects, designing buildings that a notoriously complex to look at, so few people bother following such practices on more mundane infrastructure projects. I tested a bit of a theory with Bent, that an enemy could be reputation as much as it can be ignorance. In Alchemy, Rory Sutherland (who we talked about just before hitting record), has a bit about how it is always safer to be conventionally wrong than unconventionally right. So does that stay true in infrastructure projects? The minister or senior civil servant who follows the standard process, does the various feasibility studies, sets a budget, builds in some contingencies, hires the big consultants is doing a job. It's often one that won't be opened until they've left office or moved role. So the leaders that come next will seek to do the things you do when things are going well, deflect blame, and defer to a process. The projects that come along are then late, expensive and often rather underwhelming. But the one who does something more subversive, taking the time to explore before committing to'shovels in the ground',is taking a greater personal risk. The better route to a solution might be more oblique, as John Kay puts it, with Bent also channelling his own riff Daniel Kahneman, allowing for curiosity, exploration, new ideas and the discovery of unknowns before mistakes can be made. It does not help that the temptation runs the other way for the people actually holding the pen and the purse. Bent is wry about the profession's appetite for glory:'It's much more fun to build the longest bridge than just to build a bridge. The tallest skyscraper, the fastest plane, the deepest tunnel… much more fun than doing something standard.'A former guest, Jonathan English, divided the infrastructure world into engineering problems and phone-call problems … the slow, imaginative, experimental phase is the bit that looks to a nervous board like a phone-call problem being ducked. We spent huge amounts of money designing and building a more complex bridge or tunnel, when an alternative solution might not have required a bridge or tunnel in the first place. During our hour, I wanted the test some thinking around the demand side of the big project equation. Everything up to that point has been about building the thing right. There is a prior question of whether the thing is worth building at all. Bent has a principle of thinking from right to left, beginning at the far end, with the desired outcomes and benefits … the why, the person meant to end up better off … before reasoning backwards towards the budget and the program, rather than starting with the spreadsheet, or the Gantt chart, and then hoping the rest becomes clear later. "Money and time are means to deliver the benefits. The benefits are the real reason you're doing a project." "There's just one word you need to focus on, and that is why… if you're doing it for rail users, the answer to that 'why' had better have something to do with their experience." This is, more or less word for word, the argument I have been making about the demand side (or proposition design) of places and projects. It is the whole case for the infrapreneur, the value nobody currently owns, the job of thinking'will someone actually love this, or want this'in the room while everyone else is obsessing about the cost, financial rationale and engineering of the thing. Bent is more candid than most in his field about how hard demand forecasting really is. The Elizabeth line sailed past its projections; but Eurostar spent years missing forecasts that one may say were strategically misrepresented to unlock the funding. The Channel Tunnel has skirted bankruptcy twice, and it could physically run far more trains than it does, but doesn't. Ironically this is in part because nobody built anywhere to park them, so we got a multi-billion-pound pipe beneath the sea, hobbled by the want of a bigger storage unit. We see this sort of miscalculation around getting people to use a thing by not exploring the customer proposition sufficiently. Spanish HSR is another example of this where ridership pretty much doubled when competition was legislated. I suspect we will see the same with trains to Europe from the UK … Thanks for reading Challenger Cities! Subscribe for free to receive new posts and support my work. We should talk a bit about how you actually do build big things. Bent is verging on the evangelical about modularity. Construction is the one large industry where productivity has stagnated, while pretty much everything else relating to making things has improved significantly. There is a reason consumer goods became far better, and cheaper, while each new building, railway line, hospital or power station comes with more outlandish figures. "Why don't you build houses the way the auto industry builds cars? There's no reason. No reason at all. Car factories are assembly sites. They are not construction sites." And as a Dane, his remedy is actually stop treating each project as a unique work of art and start building with Lego. For by finding the standard, repeatable components, the way the likes of solar and wind installation eventually found theirs, you can let the learning curve do the work. A turbine that once took years to raise on site now turns up as modules and goes up in a day.'The construction site for wind turbines got turned into an assembly site instead,'he says,'which is what everybody in construction should be thinking about.' A deeper question the book questions is why construction, alone among the industries that make things, didn't go through a form of transformation or industrial revolution. We worked this out more than a century ago as the coach-builders who once made carriages by hand became the car industry and moved into factories, while construction stayed out on the building site. Bent has a blunt challenge to the profession, saying,"Why don't you build houses the way the auto industry builds cars? There's no reason.' The technology has been there for years, but the pressure to use it has been missing. Cars compete globally, so an inferior one is punished quickly and fatally, as many car manufacturers have found out. But houses are sold into a thousand small local markets that don't really compete, so the mediocrity survives."You can get away with building a bad house and still survive,"says Bent."If the car industry builds bad cars, they go out of business."Unless that reality changes, the reason to reinvent the process isn't there. Though some might say, that reality is now changing as populations become more mobile, and now de-industrialised economies see the standard of its infrastructure, and in many cases quality of life, declining, especially in comparison to newly industrialising nations. The demand side is relevant here too, because modular housing is not a new concept. Just in the post-war era, most places did it cheaply and badly, resulting in the term 'prefabricated' to be associated with 'damp' and dodgy. Bent knows exactly why the idea still struggles as'that's why nobody wants it … we have all these ugly examples from the past.'Modularity only works if you actually make good things in modular ways. We can now effectively print out solar farms, and China has effectively created moving factories that spit out thousands of kilometres of railway. In many ways, Lego is a supply-side miracle, but whether anyone wants what is built still requires some design thought. We need to apply the thinking of what are the Lego style components of a project, then get creative with how we assemble them, rather than constantly trying to design new shapes of Lego, before wondering why it takes so long to fit them together. Share This is the third Danish conversation we've done on the podcast after Mikael Colville-Andersen in EP66, and Ann-Britt Elvin Andersen & Jakob Norman-Hansen more recently. And it's interesting how Danes, as far as I can see, never stop telling one another the country is broken, while at the same time, being one of the beacons of desirable urbanism. In candid chats with each of them, I hear about how they open the paper to be freshly outraged about the state of the trains, or electricity prices, as a sort of daily national pastime. The place works, I suspect, partly because they decline to believe it's brilliant. What I envy is that impatience with themselves. Canada, where I've made my home, has arranged matters the other way around, with a settled contentment on top and, underneath, a much stronger sense of negativity that the country is simply not capable of figuring how to build housing, railways or modern institutions. Bent's own conclusion, after logging 24,000+ projects, is pretty blunt.'Go to the mirror,'he says,'point a finger at yourself, and say: your biggest risk is you.'We are, he reckons, hardwired to underestimate risk, forever mistaking the perceived version for the real one, convinced our own project is the exception, when almost none of them are. The way out is to hold two things at once, to be what he calls a realistic optimist:'realistic about the risk, and optimistic that you can actually do the project.' He boils it down to consistently learning. "If you give up your learning mindset, you can't expect to succeed. All innovation comes from learning." Which is more or less what this whole podcast has been trying to be, 100 episodes in. Thank you to everyone who has joined along on the journey so far, to Bent for being the special 100th guest, and here's to 100 more. To being Challengers. Thanks for reading Challenger Cities! This post is public so feel free to share it. Share