The introduction to How Big Things Get Done, a new book about how megaprojects should be delivered, tells the cautionary tale of a high-speed rail programme that has veered massively off-track.
Costs ballooned to around three times the initial budget, leading to some drastic cuts being made. Instead of linking two of the nation’s most important cities, the line currently stops short of its end points on both sides, making the whole thing a white elephant.
The project in question isn’t HS2, though. The California High-Speed Rail line was initially meant to link Los Angeles to San Francisco at a cost of $33bn (£27bn), but the most recent estimate puts the cost of its completion at $106bn. A central section is being built in California between Merced and Bakersfield – “two towns most people outside California have never heard of”, author Bent Flyvbjerg points out – but the thorny question of where the High-Speed Rail authority will get the $80bn it still needs to complete the scheme remains unsolved.
Its failure serves as a reminder that it isn’t just the UK that struggles with the delivery of megaprojects – typically defined as projects with a value of at least $1bn.
The ‘Iron Law’
Flyvbjerg, a professor who teaches at the University of Oxford’s Saïd Business School and the IT University of Copenhagen, cites scores of other examples in his book, coining the 'Iron Law of Megaprojects', which states that they run “over budget, over time, under benefits, over and over again”.
After finding there was no data source on how frequently big projects went wrong, he created one. Of the 16,000-plus projects he has tracked in 136 countries, only 8.5 per cent were delivered on time and to budget. And only 0.5 per cent met expectations in terms of time, budget and benefits.
Although California’s high-speed rail project is used as the book’s opening example of how schemes can run perilously out of control, he tells the IC that he could just as easily have plumped for HS2. “It would have served the same purpose.”
When the UK’s flagship infrastructure project was first approved in 2012, the budget for building it – from London, through Birmingham and onto Manchester and Leeds – was £32.7bn. Although the Leeds leg was abandoned two years ago in a bid to save money, the cost of completing phase 2 from Birmingham to Manchester was estimated to have topped £100bn before prime minister Rishi Sunak scrapped this, too. According to data provider Tussell, £1.25bn has already been spent on phase two.
Flyvbjerg's book spells out many of the common mistakes made on big projects and what can be done to mitigate them. One of the most common is an “optimism bias” – an assumption that things will go much more smoothly than they inevitably do.
The main reason for this became the source of a debate between Flyvbjerg and one of the world’s best-known psychologists, Daniel Kahneman, he says in the book. Kahneman asserted that this is, in the main, a psychological trait but Flyvbjerg argues that the bigger a project is, the more likely it is to be swayed by political forces.
Budgets and timescales are often deliberately low-balled in order to gain approval, he says.
“I would actually say that the politics is the most damaging part of HS2. That is the thing that more than anything else has caused the budget and schedule to overrun,” he adds.
“The problem is that the government kept changing its mind. One of the basic banal truths that every project planner knows [is] if you keep changing, your project it is going to be incredibly expensive and incredibly slow.”
Taking lots of time during a project’s planning phase should be actively encouraged, as mistakes made on paper – or increasingly on screens via “digital twins” that replicate real-world environments – are cheaper and easier to fix. Once this is nailed down and the green light is given, though, the window for construction should be kept as tight as possible. The longer building drags on, the greater the potential for unforeseen disruptions, such as a global pandemic or the outbreak of a war that causes energy prices to soar.
Other Flyvbjerg recommendations, such as relying on experienced engineers and only using proven technology, were also ignored. Rob Holden, the former chairman of the UK’s first high-speed line from London’s St Pancras to the Channel Tunnel, told the BBC that HS2 was “ill-conceived” from the outset due to a requirement that the trains would need to run at 400kph. This made construction costs “exponentially higher” than those of HS1, which ran at 300kph. Holden also said his application for a role on HS2 was rejected.
Yet although many of HS2’s problems feature common mistakes Flyvbjerg identifies as universal, there were also uniquely British complicating factors.
Planning to fail
One of the trickiest is the planning system. Although there has been a framework in place for Nationally Significant Infrastructure Projects (NSIPs) since 2008, it has “deteriorated in recent years”, according to the National Infrastructure Commission (NIC).
The average length of time it takes to gain consent for an NSIP has increased from 2.6 to 4.2 years, the advisory body said last month.
Part of the problem has been the use of judicial reviews, which has increased from 10 per cent of major projects to 60 per cent, the NIC said.
Opponents to a scheme can use the judicial review process several times for the same project, says Hannah Vickers, chief of staff at project manager Mace Group.
“If you are just trying to block it, you can pick a number of aspects of that particular programme and just keep going after it,” she says. A true judicial review process should allow for legitimate objections, but should consider them all in the round, she argues.
Vickers, who recently co-authored a report on improving productivity for the Construction Leadership Council, also pointed out that planning reports require inputs from a variety of parties including local authority fire, transport and environmental departments.
However, they are not bound by delivery targets so this can lead to delays, which have costly knock-on effects. The NIC estimates that the cost of running a team assembled for a large infrastructure scheme can be as much as £1.5mn a month. The lack of any certainty about the length of a planning delay means it’s not possible to dismiss and subsequently rehire them, it added.
The NIC called for a reform of the NSIP framework to make sure infrastructure projects gain approval more quickly, “at a minimum” returning to the two-and-a-half-year consent level achieved in the early 2010s, it said.
The CLC has estimated that planning “presents a drag of £11.3bn annually” on the economy.
Flyvbjerg acknowledges that the planning system “is an issue”.
“But it’s not unique to the UK. Most highly evolved democracies have very complex planning laws, environmental laws and safety laws”, he says. “That’s not necessarily a bad thing.”
He adds that the UK is no worse in terms of cost overruns than similar nations, but that a focus on the most disastrous schemes that get the most media coverage means there’s a perception that it is.
“But it’s also the case in Germany. It’s also the case in the US. I very often hear this thing that people think they are uniquely bad in their country about this, and I can reassure that no, you’re just as bad as everybody else.”
The most recent update to the Government’s Major Projects Portfolio (which also includes big IT schemes) demonstrates just how poor delivery has been, though. Of the 244 live schemes, only 10 per cent are rated as green – meaning they are “highly likely” to be delivered on time, to budget and with no major outstanding issues. A slightly smaller number (9 per cent) of projects were rated as red, defined as schemes where “successful delivery appears to be unachievable”, but these represented a bigger proportion of the overall budget (12 per cent, red compared with 6 per cent green). The vast bulk of schemes (75 per cent) fell into an amber category – classed as those where delivery “appears achievable but significant issues already exist”.
Improving big project delivery is important, given that economic growth is already being dragged down by a lack of effective infrastructure investment. According to the NIC, the UK spent the least of any G7 nation on infrastructure spending – an average of around 19 per cent of GDP – in the 40 years to 2019. It estimates that infrastructure investment needs to grow from around £55bn a year over the past decade to £70bn-£80bn throughout the 2030s and 2040s. The bulk of this (£40bn-£50bn) will need to come from private sector sources.
Rail replacement service
Even with prime minister Rishi Sunak pledging to divert “every penny” of the £36bn that was budgeted for HS2’s second phase, many think it will be a blow to investment.
CBI chief executive Rain Newton-Smith said the decision to cancel the project’s northern leg “sends a damaging signal about the UK’s status as global destination for investment”.
Alasdair Reisner, chief executive of the Civil Engineering Contractors’ Association, says external investors will have “less confidence in the future profile of infrastructure investment in the UK” .
If all of the capital is reallocated, then the £650bn infrastructure pipeline highlighted by contractors such as Balfour Beatty (BBY) and Kier (KIE) in their pitches to investors should remain largely intact (see What the cancelling of HS2 means for listed companies) but Reisner said the government’s willingness to “rip up its own plans at short notice” meant that there would not be as much confidence in them seeing through the alternative schemes being proposed.
For the alternative asset managers currently sat on more than $4bn-worth of ‘dry powder’ (capital ready to deploy) in UK-focused infrastructure funds, the scrapping of phase two won’t have made much difference as “there was never really an inroad for them to invest”, says Alex Murray, a vice-president of research insights at data provider Preqin.
Indeed, since the private finance initiative (PFI) was abandoned in 2018 following criticism that it failed to deliver value for money – a National Audit Office Report that year found that the total bill for projects with a capital value of £60bn was likely to top £199bn – the routes for private sector investment into public infrastructure have narrowed.
Pricing methods such as contracts for difference (CfD), or the regulated asset base model, where governments agree to pay subsidies to developers of renewables (and in the future, nuclear) power projects are one method, with the cost eventually added to people’s utilities bills. The NIC’s latest five-year plan also calls for this method to be used to subsidise heat pumps to facilitate the transition to net zero. It argues that over the longer term, replacing a fossil fuel-based system with one that runs on renewables should bring down household spending on infrastructure (via utilities bills and tax payments) from £7,300 a year currently to between £5,500 and £6,500 by the mid-2030s.
Even here, there have been problems. Infrastructure investors in the water sector have been accused of milking bill payers and prioritising payouts to shareholders over investments in their networks. And the government’s latest CfD auction in September failed to attract a single bidder for offshore wind schemes, with the scale of subsidy not being seen as generous enough given recent hikes in construction, labour and materials costs. On top of this, higher interest rates make it “a lot harder for private capital to see the payback period that they would have in previous years”, Preqin’s Murray says.
Also, whether the criticism that investors into PFI or the water sector is deserved or not, “there is a read across as to how private capital gets treated”, says Edward Hunt, who runs the HICL Infrastructure (HICL) fund.
If there’s a rejection of the idea that private capital should be able to earn a decent return, investors will go elsewhere, he adds.
“The same sectors are being developed in France, Germany, the US, New Zealand. So investors have a choice… and they may find a better risk-return framework somewhere else.”
The government has paid attention to – and even acted on – some of the issues highlighted by Flyvbjerg. He was called in by ex-chancellor Gordon Brown back in 2003 and devised a process called “reference-class forecasting”, where budget forecasts take into account overruns on similar global projects and the “unknown unknowns” they faced in a bid to tackle the optimism bias issue.
Moreover, with the first major nuclear plant being built in decades (Hinkley Point C) already running 18 months behind schedule and 80 per cent overbudget, while the second (Sizewell C) has yet to attract the necessary funding, it is embracing Flyvbjerg’s promotion of modularity as the way to deliver “faster, cheaper and better”.
Little and often
Nuclear projects are the most expensive of all because they are bespoke, with everything typically built on site and standards that frequently change, Flyvbjerg said. However, both France and South Korea managed to bring costs down by building lots of them within a short space of time.
The government is currently assessing six different designs of small modular reactors (SMRs), including one designed by a consortium led by Rolls-Royce (RR.) that would only generate around 15 per cent of the power of a Hinkley Point C but would cost around £1.8bn per plant (although the first units to be built would be more expensive).
Yet the economies of scale and “positive learning” generated through could be lost if SMRs are forced to go through the same rigorous planning process – involving a five-person panel and a six-month examination period – as full-scale plants, according to Britain Remade.
“Without a proper planning strategy about where we're going to put them, and how we're going to override objections that might come up again and again, which should be really settled on the first planning application, we're not going to get the returns to scale and the decreased energy costs,” says policy researcher Ben Hopkinson.
Nuno Gil, a professor of new infrastructure development at Alliance Manchester Business School, argues that the UK’s higher project costs – including its planning costs – are reflective of its “liberal market economy” and the fact that governance structures differ to the likes of Germany and France.
In Europe, “people trust more on the ability to achieve consensus and put less emphasis on the cost-benefit analysis”, he argues. The proportional representation system also means local parliamentarians have less influence than the UK’s MPs.
He cites the example of Madrid’s metro expansion in the late 1990s as an example, which was overseen by central government and used modular construction to deliver 76 identical stations. By comparison, the “complex design features” in the 10 new stations built for London’s Crossrail project contributed to cost overruns on a project that busted its £14.8bn budget by £4bn.
Crossrail was approved using conventional cost-benefit ratios but when costs started racking up its bosses lobbied, “and built a large coalition” to convince governments to use a broader value for money metric that accounted for the scheme’s wider economic impact, Gil says.
“What we have right now in this country is a mess,” Gil argues, saying that strategic funding decisions are supposedly based on rational economic assessments but then other – often difficult-to-quantify – factors are then used when attempting to justify a project’s value for money.
Government and industry thus still “play the game”, with one side setting unrealistic targets and the other offering to meet them in the knowledge that expectations will be reset at a later stage.
In a roundabout fashion, this gets to the crux of the problem. Knowing how to get a job done is one thing but sticking to the plan is something else.
