This is a guest post by George Weeks, reviewing a book called How Big Things Get Done by Bent Flyvbjerg and Dan Gardner
Want to know how to do something big? Ask someone who has done it before! Professor Bent Flyvbjerg has spent his career studying megaprojects. In 2023, he teamed up with author and journalist Dan Gardner to share the knowledge with us all.
How Big Things Get Done is a formidable title, but a readable book. I finished it in two evenings and enjoyed it so much I re-read it almost immediately and have had it on my mind more or less constantly. This review is by no means the first but I think this book is relevant to anyone who reads Greater Auckland, hence my putting pen to paper.
At the heart of the success of How Big Things Get Done is its synthesis of stories and statistics woven together to keep the reader interested. By taking us under the skin of familiar buildings, we can link broad principles to the places we know. We all know the Empire State Building exists, but how many of us know that it took just 14 months to build and came in 17 per cent under budget?
The book unfolds in nine chapters, each one anchored by examples and principles. Themes emerge throughout the book, for example fat-tailed statistical distributions which indicate a higher risk of unpredictable events which can cause major delays and/or cost increases. How do we shorten these risk tails? We investigate the risk factors and find solutions. Building a high-speed railway through England, with a high risk of discovering archaeological sites? Keep every qualified archaeologist on a retainer, to minimise the consequences of this type of inherently unpredictable event. HS2 Ltd actually did this. Expensive? Undoubtedly. But it’s cheaper than an unpredictable delay.
The importance of experience emerges consistently. We see a comparison between Frank Gehry’s completely successful Bilbao Guggenheim Museum and Jørn Utzon’s Sydney Opera House, a beautiful building but a legendary fiasco. Gehry won the Bilbao commission as a 62-year old architect, experienced at complicated public projects. Utzon meanwhile was an unknown 38-year old who’d won several architecture competitions in his native Denmark. Pitching his relative naivete against the immense task of delivering an unprecedented structure in the brouhaha of Australian state politics, Utzon faced an impossible task, departing in 1966. His successor, Peter Hall completed the seemingly-impossible building, but suffered for his work.
A more experienced architect, with many similar projects to their name, would have known how to handle the situation, potentially detangling the project’s knotty start. We see that it’s vital to understand specifically why you are undertaking a project. Flyvbjerg’s mantra is: Think slow, act fast. Slow down. Take the time to work out precisely what you need and how you’re going to deliver it. Then get on with it.
When reading How Big Things Get Done, one thinks of potential case studies. Earlier this year, Dr Sean Sweeney linked the high costs of City Rail Link with the loss of institutional experience. In an earlier generation, New Zealand’s ‘Think Big’ era is riddled with cautionary tales of large, one-off projects enthusiastically undertaken from the late 1970s to pursue energy independence. Billions of dollars went into a variety of complicated energy projects. As oil prices plummeted in the 1980s, the economics looked somewhat shaky. Yes, you could make petrol from methanol, but at a much higher cost than the conventional stuff.
Engineers in the UK seemed to go through a similar streak of how-not-to-do-its in the 1950s and 60s, for example:
- 1955 Rail Modernisation Plan – a vast, rushed programme to upgrade the pre-war railways, but with no real thought as to how a future rail system would work.
- Concorde – the supersonic perceived future of aviation, but zero commercial customers.
- Hawker-Siddley Trident: three-engined rival to the Boeing 727 but customised precisely to British European Airways’ specifications. Downsized, then lengthened again. Lousy sales.
- Advanced Gas-Cooled Reactors – Unique, technically-successful nuclear power stations that attracted no export customers.
If we must stick to experience, how do we build big things? Flyvbjerg’s advice is wonderfully Danish: Find your LEGO. Use mass-produced, well-tested components that work at any scale. This concept is known as “scale-free scalability” – as flexible as a flock of starlings. A wind farm can have one turbine, 20 or 2,000 – it really doesn’t matter; they’re just mass-produced components, copied-and-pasted. It’s no coincidence that wind farms rarely go over budget. Solar farms even less so, for the same reason. Flyvbjerg’s database contains details of over 16,000 projects; these assertions are solidly backed by data.
There are other great examples of teambuilding and culture (exemplified by Heathrow Airport’s Terminal 5) and we are advised to use experienced teams who clearly understand what they need to deliver. We also learn the importance of using precedents and reference-case forecasting (i.e. don’t think that your project is “unique” – chances are it isn’t). Precedents are always available and should always be used.
The book comes with two appendices containing mean cost overruns for different project types (nuclear waste storage is particularly high risk) plus further readings by Bent Flyvbjerg (many of these are available online). It is fully indexed and is clearly intended for professional use and thumbing-through. Despite being published as the UK edition, written by a UK-affiliated academic, the book referred to American “centers”, “travelers” and “favorites”. This to me is a mystery.
In all cases the material is readable, engaging and fascinating. How Big Things Get Done could in all likelihood carry on for a thousand pages of principles and case studies and still stay interesting. That said, its relative brevity of 190 pages is a virtue; people are more likely to read a short book than a long one. Many already have read it. I had to wait seven months for a copy to become available at Auckland Council Libraries. Needless to say, it was well worth the wait. Book reviewers seem to agree with me and it was shortlisted for the FT/Schroeders Business Book of the Year 2023.
How Big Things Get Done is superb. Easy to read, filled with valuable principles, thought-provoking and memorable. If you’ve read this far…do go and read it!
This post, like all our work, is made possible by generous donations from our readers and fans. If you’d like to support our work, you can join our circle of supporters here, or support us on Substack!
Processing...
Sounds great, I’ll hunt it out.
In one sense, Simeoon is right to carry on the RONs game because NZ does have institutional knowledge in road building. Just not enough though, because we can’t seem to be able to build them on time or under budget.
Or build them without needing rebuild and remediation shortly after opening. Looking at you, Waikato Expressway.
I favour steady improvement to existing assets over mega projects. Have a good permanent workforce and allocate a set amount to each project per year. That way engineers, workers learn from their mistakes and build institutional knowledge over time rather than having to learn it and lose every time a new project starts or finishes. This every important with roading and railways. A light rail system should start with a tram up Queen Street and expand. Mistakes will be made but they can be quickly rectified. The team will quickly learn what works and what doesn’t and patronage on the system will tell us how to expand. It will be subject to political interference but so are megaprojects.
Meh. Joke all you like Mr Plod, but I’ll take the bait because I’m not going to let that sit without putting it right: Simeon is not right to carry on the RoNS building. But you’re right to bring up that skill set. We do have a challenge in encouraging the transfer of those road building skills into projects that build the infrastructure we do need.
Using the LEGO concept makes sense (“mass-produced, well-tested components that work at any scale”). I’d extend that to simplified design processes, too. Eg for:
– active mode overbridges and underpasses for motorways and rail lines,
– lane dividers, bus infrastructure, raised crossings, modal filter furniture and kerb buildouts, etc, for retrofitting arterials
– low-cost active pathways joining regional towns
– cheap processes to remove vehicle crossings so we can retrofit our suburbs to having consolidated parking lots instead of driveways every few metres.
And so on… Thinking like this is already present in NZ (standardised bus shelters, park seats, etc) but we could have much more of it.
Yes, it was written a little tongue in cheek. I agree entirely with your list of lego solutions. I also agree with Royce around continuous improvement. Imagine if we’d set out to develop a way to commit to adding 50km of overhead power to the rail network each year. After a few years all parties would bring the price down on the guarantee of a long term pipeline of work. We just have to be careful ‘pipelines of work’ don’t become another form of economic rent for the providers.
Here’s a good article comparing kilometres of rail electrification in the UK and Germany.
https://www.samdumitriu.com/p/infrastructure-costs-electrification
Germany has electrified about 200km per year since the late 1960s in a rolling programme. Consistency reduces costs.
Over the same period, the UK has been feast-or-famine (mostly famine). There was a big spike in the late 1980s when the East Coast Mainline was electrified from London to Leeds, York and Edinburgh. This dropped off rapidly in the 1990s (privatisation was no help at all here) before growing again in the late 2010s.
Intervals of this length all but guarantee that most of the instutional knowledge will be lost and you have to start from scratch each time. Not cheap.
I’m really liking “Book of the Weeks”. Reviews that lead to books being read and at the same time making clear points even if you can’t get hold of the book for seven months.
LEGO concept doesn’t just apply to products. Plug-and-play design, reporting, risk management are even more important. And assembling teams out of LEGO helps – people who already know how to do their part of the project, however different the projects are.
Heidi’s first two items are under way (modular rail footbridges, prefab road components). Item 3 needs John Key back at work.
Item 4 – a pathway for redeveloping land fronting arterials – would be great. “My land, my rules” is the mindset that needs to be addressed before that happens.
“Book of the Weeks” eh? That is the best pun on my surname that I’ve heard in a long time. And thanks for the nice feedback. Glad to hear you’re enjoying the reviews. Maybe I should write some more?
Yes please, I’m enjoying them too.
+1
For point 4, perhaps that mindset can only be shifted if people can see the benefits and and start discussing it? There are some good (and some bad) road frontages in Hobsonville Point, and there are also some good (and plenty of bad) redesigned road frontages in areas that have been intensified. Perhaps AT and Council could highlight the good ones as part of a communications exercise?
NZ only has the population of a large city and way too many groups who think they know how to build stuff .Because of that we have 70 different plans because every council thinks they know best .We need to invest in getting the best 10 people we can from NZ and around the world and get them to work together on a lego type plan for road rail and water for a start and perhaps energy could come under them as well .That way every council would be using the same components for 3 waters ,for instance ,so the cost would fall dramatically as the ability to price gouge on one off setups would be gone .