Was HS2 always a fraud?

I've been reading an opinion piece in the journal Rail Engineer, titled HS2 offered so much*, lamenting the government decision on HS2.

The main point it makes is that people have been focussing on excessive costs but forgetting the benefits. That is not the case at all. The costs have been escalating and the various decisions to trim back the route have hugely reduced the benefits, a classic double whammy. Changes in business travelling patterns have probably also caused forecast passenger numbers to shrivel, deferring benefits even if growth is assumed.

Amongst the hand wringing it notes that the original 2013 cost estimate was £16.3 bn (in 2011 prices). They describe this as a "basic estimate prepared for business case" and it was for phase 1 to Birmingham. While one never has a detailed design on which to base a business case, I didn't realise that it was acceptable to punt that much public money on a "basic estimate" that wasn't considered at all reliable, but there you go. I suppose in fairness there were Parliamentary votes to approve the full scheme once designed with supposedly reliable cost estimates, but by then the project had political momentum and was uncancellable. And we were preoccupied with Brexit.

The current budget for phase one (which no-one believes) is £40.3bn plus £4.4bn contingency. In 2015 the cost for the full route was £27bn, based on route drawings but no ground investigation. By 2017 it was £37bn based on a finalised route and limited ground and site investigations (eh - even by 2017?). They note the estimate comprised 15,000 lines of data. I don't really care how much detail there was if a lot of it was clearly bunkum. By 2019 it was £44.4bn with the design 80% complete and based on 260,000 lines of rubbish, sorry, data but the DfT was estimating that the total cost would be between £65bn and £88bn (at 2015 prices).

We now know that, before the northern leg was cancelled, the total cost was realistically expected to top £100 bn and that it is being seriously suggested that HS2 Ltd should be investigated for fraud** for keeping quiet about estimates they knew were totally unrealistic. (Remember that the original scheme was for the Y route including the leg up to Leeds, not just one northern leg to Manchester).

The Rail Engineer also notes that there's been a good degree of inflation in the economy since. But that doesn't explain costs going up by a factor of three.

It's rather hard to feel sympathy for those crying into their beer about the cancellation: they've been let down from the start by the HS2 protagonists.

The Rail Engineer article also notes that, after various studies had shown that UK projects are typically 10-30% more expensive than those in Europe, the government commissioned a high-speed rail benchmarking study. It looked at 32 comparator European high-speed rail schemes and was overseen by an expert panel chaired by Sir John Armitt. (Armitt was chief executive of Railtrack and then Network Rail from 2001 to 2007 after which he was chairman of the Olympic Delivery Authority). 

So he ought to know what he's talking about. But I wonder... After finding that HS2 phase 2 was 49% more expensive than a European high-speed rail line with similar characteristics, the panel decided that the factors accounting for this additional cost were:

• strategic objectives requiring greater capacity and more intermediate stations (7%). This could be partly true but the "more intermediate stations" bit can't be. According to Wikipedia there are 20 stations in France that are served by TGV trains - and that's just those beginning with the letter 'A', I couldn't be arsed counting the rest!
• limited capacity of UK rail infrastructure requires dedicated high speed lines into city centres (15%). This is plain wrong. The mainline stations in Paris from which you can catch a TGV train are Paris Gare du Nord,  Gare de l'Est, Gare Montparnasse and Gare de Lyon. That's four of the seven large mainline Paris railway termini. I haven't checked other French cities but there's a TGV station at Part-Dieu in the main business area of Lyon, so go figure.
• fragmented UK construction industry and continuity of work (12%). I can believe this, it applies to nuclear power stations in spades
• Onerous design requirements (5%). That's the speed then is it? I simply refuse to believe it's only 5% and economist Bridget Rosewell, who was a commissioner at the National Infrastructure Commission, is with me on that***
• scope development compounded by limited experience of delivering high-speed rail in UK (10%). In other words, we don't know what we're doing while we're gold plating it.

This analysis seems to me to be missing the point. We know our projects tend to cost more than those of similar countries, though theirs are also subject to large overruns on time and cost (see Berlin airport for example). In April the High Speed Rail Alliance noted that American sources were lamenting the fact that the New York Second Avenue subway was costing 8 to 12 times more than the "composite baseline case"  compiled from data from several European countries. Factors identified included a lot of bespoke design. But on the other hand the Americans use a lot less tunnels than the Europeans (the Brits are using even more on HS2).

The real question is why, even though we know our projects cost more, do our projects cost a lot more than we think they are going to? Especially when we completed HS1 less than 10 years ago.

It seems it's because they did know (or at least strongly suspected) that was going to be the case but thought they'd never get the project approved if they fessed up.

Which is not new of course. It's exactly what the builder of some of New York's earliest transit infrastructure did. Get the project approved, start building it, then they won't dare to cancel it. Well Sunak just called your bluff chums.

And while I thought it was ridiculous to use the term 'fraud' when I first read it, now I'm beginning to wonder. The Sunday Times investigation** spoke to a number of whistleblower/malcontents who had worked at HS2, one of whom has been contacted by HS2's fraud and ethics team. The article went in some detail through the cost evolution, with numbers which are difficult to relate in detail to those given in the Rail Engineer article, though the general picture is even more stark. The main thrust of the article is that there were people working in the HS2 team who had felt for a long time - and sometimes popped their head up and said - that the costs were totally unrealistic.

Of course that's what happens in big projects that go off track. Management will always initially say that the cost estimate remains valid and press their team to find ways of finding savings. If you don't do that the steadily increasing numbers just become self-fulfilling prophesies and all discipline about containing the cost is lost. It's a game I played many times myself, with more success on some occasions than others, though my project totals were typically several millions rather than tens of billions.

What I found interesting in the article was that Chris Grayling, on appointment as SoS for Transport in 2016, had grave misgivings about the project. He called it a "crackpot scheme" according to a DfT source, while backing it publicly. Later that year a confidential report by Paul Mansell, a Treasury adviser, estimated that it was "highly likely" the scheme would cost more than £80 bn against the then budget of £55.7bn. But when the final vote to approve construction of phase one was taken in February 2017 there was no mention of that report and Grayling said it was "on track for delivery". The bill was passed with big majorities in the Commons and Lords. The DfT source said Grayling was unaware of the Treasury report at that time. I wonder if the chancellor, Philip Hammond, was aware of it?

By summer 2017 Michael Byng, a consultant who devised the method used by Network Rail to cost its projects, calculated that the overall cost of HS2 would be £104 bn. That may have been the first realistic estimate. It was nearly double the then official figure and 15 times the cost of the most recent high-speed TGV line in France (They have the benefits of much replication and an infrastructure of experienced personnel and suppliers as a result. But you wouldn't expect that to explain a factor of more than 2 or 3 at the most).

By late 2018 Grayling had appointed a new HS2 chairman, Alan Cook and had asked him to review the costs. The next month Grayling told Cook that, if the scheme couldn't be delivered for £55.7 bn "I think we will have a serious discussion about scrapping the project". One might have expected that would lead to HS2 closing ranks and standing behind the budget. But no, In June 2019 Cook sent a first draft of his report to Grayling. It suggested the project was billions over budget and years behind schedule.

This was not reflected in submissions to Parliament. In July 2019 the minister for transport, Nusrat Ghani, answered questions in a Westminster Hall debate on HS2 before the Commons vote on the bill to approve the Birmingham to Crewe phase two leg. She said "I stand here to state confidently that the budget is £55.7 bn and the timetable is 2026 and 2033". She repeated her assurances five days later during the third reading debate. But a freedom of information request later revealed Ghani had been told three months earlier that the project would breach its budget. In which case Ghani, now minister of state for industry, would appear to have misled Parliament.

By August 2018 Cook's final report revealed that the cost was now expected to be £88 bn, very close to Treasury expert Mansell's estimate. Even that number proved optimistic.

One wonders if the ministers of the day actually knew the estimates were unrealistic before the Commons vote and sent HS2 away to confirm the number, get the vote through and then come back with their request for more money. In that case they could hardly place all the blame on HS2 Ltd and were complicit in the "fraud". More likely they'd be able to hide behind the fact that costs on big projects are always under review and the "official" cost estimate hadn't been changed. Nevertheless this seems a matter ripe for an enquiry to me.

It's all so stupid. we could have built HS2 like HS1 and had a prefectly acceptable railway, bringing the overwhelming majority of the benefits. The Sunday Times article points the finger, as I have***, at Lord Adonis, who wanted the railway to be "super fast", meaning it had to be as level and straight as possible, constraining choices on the route and adding cost in land purchase, construction methods and materials. 

I accept that if we had gone for an HS1 spec the project would still have cost more than the original estimate, everything has gone that way. But it would be much more likely to have been completed in full, at an affordable cost and on a reasonable timescale. Basically it could have worked.

*published 18 October 2023, www.railengineer.co.uk

** We were instructed to lie about the true cost of HS2. Sunday Times Insight investigation, 22 October 2023. 

*** see Who to blame for HS2 20 October 2023

The cost of the New York subway project and the general lament about costs of projects in the USA can be found at https://www.hsrail.org/blog/why-transit-projects-cost-more-in-the-u-s-than-almost-anywhere-else-and-what-we-can-do-about-it/#:~:text=In%20the%20case%20of%20New,high%20even%20by%20U.S.%20standards.