From Prospect Magazine
Costs have spiralled, delays have mounted—and the heart of the project may never be built. The beleaguered high-speed two is a case study in how not to do an infrastructure mega-project
What to do about high-speed 2? It’s not helpful to say “I wouldn’t have started from here”, but we do need to understand how we got here to avoid repeating the mistakes. In short, the mistakes were: a lack of frankness about how much it would cost (about £2,300 per household in public spending, after allowing for revenues); confusion about objectives; a lack of regard for hard evidence on what we would be getting for our money; and a hopelessly expansive view of how much public money future governments would be able to find for HS2 in addition to a shopping list of other transport infrastructure aspirations. Further decisions on what to do about HS2 have to recognise the context of the current headache for the Treasury and for any future government, with the taxpayer currently having to find around £470 per household per year for the existing railway, quite apart from the costs of any new high-speed line.
In his magnificent—but quickly forgotten—independent transport study (2006) for the Treasury, Rod Eddington advised that in British conditions, high-speed rail was unlikely to produce sufficient benefit to justify the cost. Distances are relatively short, there is already a reasonably fast rail system, and construction costs are high. Public funds would be more productively put to use meeting transport objectives in other (less iconic) ways. But in 2009, Labour transport secretary and rail enthusiast Lord Andrew Adonis crafted HS2: a brand new “full Y”, running at very high speed from London Euston to Birmingham and then branching northeast to Leeds and northwest to Manchester, with trains beyond running on the existing railway. There would be up to 18 trains an hour between London and Birmingham, splitting off to serve an attractive list of northern destinations.
Successive governments stuck with the idea in spite of relentlessly increasing estimates of the costs and a stream of sceptical reports from parliamentarians (two particularly good ones from the House of Lords) and others. The arguments in favour changed over time from “speed” to “capacity” to regional development to “the north-south divide” to “levelling up”. Parliament granted powers for the compulsory purchase and destruction of personal property between London Euston and Birmingham on the grounds that there was an overriding national interest in building the whole scheme.
From autumn 2019, prime minister Boris Johnson faced the ultimate decision of whether to grant the final Notice to Proceed with Euston to Birmingham. The official business case, published in April 2020, advised that the “full Y” would take up to 20 years to open and was estimated to cost about £100bn (at 2015 prices, including 30 per cent to 60 per cent of contingency)—so an average of about £100m every week for the 20 years, or a total of £3,500 for each household in the UK. In due course that would be offset by about £45bn of revenues. By the official methods of measuring these things, the benefits relative to the then costs were mediocre for the scheme as a whole and poor for the Euston-to-Birmingham link alone (the regional links generate disproportionately high benefits). At the latest costs the returns must look dreadful, and that is before aiming off for the completion delays and inevitable impaired functionality.
Prime minister Johnson had the benefit of the Review that he had commissioned from Douglas Oakervee, published in February 2020. Few people read beyond Oakervee’s headline recommendation to go ahead. It was heavily qualified: it’s only worth doing if you do it all; the government should look again at whether the very high speeds and very high frequencies are necessary and feasible; renegotiate or relet the construction contracts because they have undesirable risk allocation and incentives; and put in place more effective governance and accountabilities in order to keep costs under control. The transport department should publish a revised business case for the project as a whole, said Oakervee.
Johnson pronounced his approval of the Notice to Proceed with Euston to Birmingham with a fulsome commitment to complete the whole scheme. HS2 was incorporated into his expansive National Infrastructure Strategy in November 2020, which included additional northern rail plans: “This document sets out our plan for a renaissance, backed by hundreds of billions of pounds of public and private investment… all but one of the major new capital projects in the next few years will be outside the southeast. Northern Powerhouse Rail will connect Leeds and Manchester, and we will help make bus and rail networks in those and other city regions as good as Greater London’s.” The National Infrastructure Commission suggested that the total cost of all major government commitments to rail in the midlands and north would rise above £180bn. In addition, “The Commission has separately estimated that more than £40 billion is required between now and 2040 to fund major transit schemes in the fastest growing, most congested cities, as well as increased multi-year settlements for transport in all cities.”
The problem with all this was that there was no discussion of how ministers were going to find the money. Of course, contrary to Johnson’s Strategy, you cannot have everything and choices have to be made. As the cold reality emerged that the annual funding for HS2 would have to be found, the government started shaving bits off. The branch from Birmingham to Leeds, with the possibility of allowing services to continue on the East Coast Main Line to the north, was truncated to terminate somewhere short of Nottingham. The link near Crewe that would have facilitated through-running towards Glasgow was withdrawn. Having already destroyed homes and businesses and created a moonscape in central London, development at the Euston site has been paused and expensive tunnelling machines purchased from Germany are to be mothballed underground, having done nothing useful. If Euston is not reinstated and the London terminus were to remain at Old Oak Common, it will be technically impossible to achieve the service frequency to remote destinations that was essential to the scheme.
Contrary to Johnson’s National Infrastructure Strategy, you cannot have everything and choices have to be made
Timescales until any trains start running have slipped by four or five years, and completion to Manchester looks 15 years away at best. Parliament has granted compulsory purchase powers from Birmingham to Crewe but not for the extra 28 miles to Manchester, where controversy rages about the destruction to be caused by the approaches, and whether to do the job properly with an underground, through-running station or to build a cheaper above-ground terminus.
One thing is clear: nobody would have proposed the emaciated and delayed scheme that is now being built, even if it had included the Birmingham-to-Manchester link.
So here we are: what next?
There is one overarching principle: whatever is decided, stick to the decision. If something is to survive, make sure there will be funding available and open it as quickly as possible. Delay is catastrophic for any long-gestation capital project, public or private. Every £1 spent is £1 lost to alternatives (or added to the national debt), and there is no offsetting benefit until the scheme opens. Delay also increases the risk of being further caught by unanticipated real cost inflation. It has been a factor in the poor economic performance of many projects—including, famously, the UK’s nuclear power station programme. Delay in order to square the circle of a current year’s public spending difficulty is always tempting, but always damaging.
The main options appear to be:
- Complete the core of the scheme, including the Manchester connection and the Euston terminus. Facilitate improved services to Scotland by alternative means. Maybe reinstate the link to Leeds in due course.
- Just finish what has been started: a high-speed shuttle between London Old Oak Common and Birmingham. Possibly complete to Euston in due course.
- Being in a hole, stop digging: forget the whole thing and repurpose the land and other assets. Spend the money on something else.
It is impossible for any of us outside the tent to form a really solid view without crucial information we do not have. Firstly, starting from now, what will be the costs of each option looking forward, ignoring previous irrecoverable spending but subtracting whatever cash could be recovered from assets released? Costs are bandied around but they are usually so ill-defined that it is quite impossible to make sense of them. When a cost is cited, here are some of the questions to ask.
- Exactly what is the scheme being referred to: what’s included?
- Is the cost of rolling stock (the trains themselves) included, as well as the cost of operations once opened?
- How soon will the line(s) open?
- Is the quoted cost net or gross of expected passenger revenues? Although the Treasury has to find the gross cost before opening, it’s the net cost that will be the burden on the taxpayer in the long run—and represents the amount that could have been used for other things.
- Which year’s retail prices are being used in calculating costs?
- What allowance has been made for future general price inflation, and importantly, what further allowance for inflation in specifics such as construction labour and materials?
- How much contingency is included? The costs cited when Johnson gave the Notice to Proceed included generous contingency—but in this case, costs have got so out of hand that the “generous” contingencies are not going to be enough.
- What is the true level of confidence in the engineering costs?
This last point is important: should we believe what we are being told? A 2023 report by the National Audit Office confirmed that there was—and there remains—very little basis for confidently costing the expensive works at Euston. It transpires that ground conditions between London and Birmingham are difficult—and will be difficult between Birmingham and Manchester—and not enough surveying had been done to enable robust costings. The government’s Infrastructure and Projects Authority concluded in July that successful delivery of HS2 “appears to be unachievable”.
In any case, throughout the gestation of the scheme, the government and the HS2 project have forfeited our trust in what we are being told. Cost estimates were repeatedly revealed to have been unreasonably optimistic. Between 2015 and 2019, the budget for the scheme was £55.7bn. Transport minister Nusrat Ghani told parliament in July 2019 that “there is only one budget for HS2, and it is £55.7bn.” It later transpired that the government and officials had known for some time that the scheme could not be delivered for that—at least not without unspecified “efficiencies”. The Commons Public Accounts Committee was scathing about this, and added: “HS2 Ltd’s annual report and accounts for the year ending 31 March 2019 similarly failed to give an accurate account of the programme’s problems… no adequate excuse was provided for not disclosing to this Committee and Parliament the risk and uncertainty the programme was facing… Lack of clarity and obfuscation about the budget issues with HS2 risks jeopardising the trust between Parliamentary committees and Government officials.” There appears to have been an implicit understanding among officials charged with delivering the project that they had a licence to protect it by being less than frank about the true situation. Who should we trust now?
Nobody would have proposed the emaciated HS2 scheme that is now being built
Secondly, one cannot make an intelligent decision without absolute clarity on the alternative uses those resources could be put to. Throughout the history of the project many people have been quick to claim benefits of many kinds—but loath to acknowledge the costs. They argue as if it were free money. But there are always alternative uses for the resources, including paying down the national debt or leaving money in the pockets of taxpayers. The question should never be “will HS2 produce benefits?”; it should always be “will HS2 produce enough benefit to justify taking the resources away from the alternative?” The regional authorities clamour for continuation of HS2 because the funding burden falls on central government. But what if they faced a real choice between HS2 and using the same funds for their own purposes?
This is related to the question of what we want to achieve. The £100m a week for 20 years could make a difference to objectives like “levelling up” (perhaps improving schools, education and training in deprived areas), or devolution (sums like this would offer many local authorities opportunities to improve their local transport systems as they saw fit), or improving local bus services all over the country, or making the conventional railway cheaper or better. From the beginning, there were studies that addressed how one could increase rail passenger and freight capacity on this route at a more modest cost, but they did not seem to be taken seriously.
Thirdly, we need to take stock of the evidence on what the benefits are likely to be. For a start, how many people will use whatever is now proposed and how much will they be willing to pay to use it? This was always even harder to predict than costs: the first passengers may not appear until a decade from now and any estimate of benefit needs to take a view on numbers of users for six decades or so after that. That is not a criticism of HS2: almost any major infrastructure investment has the same difficulty. It cannot be escaped, though smaller schemes like better bus services tend to be more flexible and less risky.
Nobody could have anticipated the impact of Covid on work and travel patterns, or the effect of improving communications technologies. It seems at least a possibility that the market for high-speed, long-distance rail travel has changed since Adonis presented the scheme well over a decade ago. How do the propositions about overcrowding on current routes look now? While it is certainly true that it is hard to fit more trains onto the existing tracks and that some trains are heavily loaded, the House of Lords procured the information to show that, on average, loads were not particularly high by the standards of the commute into London on other lines. In the official appraisals, crowding benefits were assessed and they were dwarfed by the value of time savings from faster journeys—even though the headline objective changed from speed to capacity. And as far as such benefits would exist, far from helping the north, much of the impact would be felt on medium-distance commuter services into London, using the capacity released by HS2.
If HS2 is to continue in some form, there are a number of unresolved issues to deal with. What will be the fare levels and structures, bearing in mind the competitive implications for the income of other rail services? How will the new railway be integrated into the existing rail system, as Oakervee recommended it should be? For instance, if high-speed trains run onto the existing tracks as they head towards Glasgow they will be slower than the existing trains, being less well-adapted to the old, twisting route. That will have implications for capacity, as well as the existing freight and remaining passenger services. What will be the regulatory regime for HS2, and how will it be dealt with as part of wider major reforms of the existing railway that are already under development?
There is also a need for an analytical eye to look at the hard evidence for other claims in favour of the scheme. HS2’s published work has always suggested to me that the emissions-saving argument is weak (seconded by Oakervee). Substance needs to be provided on the “levelling-up” claims. The “Keynesian” job-creation claims need to be queried when unemployment is low. There is a shortage of labour in construction and “digging holes and filling them in again” will just drive up prices rather than creating a multiplier effect to create jobs. In any case, alternative public schemes could equally claim to create employment.
And watch out for the sweeteners. At one time, HS2 was to provide a network of cycle paths alongside the route (an “emerald necklace”), but it is unclear whether this was ever funded. Meanwhile, four years after it opened, it was announced that HS2’s National College of High Speed Rail was to shut down. Over two sites it was reportedly meant to cater for 3,200 students, but typically had fewer than 100. It was found to be “financially inadequate” by the further education commissioner. Its successor institution has now also closed down.
Finally, there is governance. Oakervee recommended that before going ahead, the system of accountability and governance of HS2 should be reformed in order to keep costs under control. The government did respond: there are now six-monthly reports to parliament. Minutes of HS2 Board meetings are published, but they are so heavily redacted that they say little of substance. It is difficult or impossible for the outsider to follow the budgets and performance against budgets. Controlling costs in this gargantuan company was never going to be easy, yet HS2 Ltd was without a chairman for 18 months until February 2023. The chief executive—the highest-paid public employee in the country—leaves the company at the end of this week, and the newish chairman will become interim executive chairman, pending a recruitment process for a position that fell out of favour some years ago. For whatever reason, it seems that the current governance has proven unable to satisfy the Treasury that it is achieving adequate control of costs.
The time has come to stop treating HS2 as an icon, take a hard look at the evidence and decide whether or not it can earn its place in our plans to rejuvenate our inadequate public infrastructure.