More on West Coast Rail fiasco – and a strange choice of “independent” reviewers?

So what more do we know about the West Coast Rail Franchise affair since Friday?

We saw suggestions that different information was given to different bidders or they were treated differently. Clearly that wouldn’t be right – that breaks one of the most fundamental EU “treaty principles”, which underpin public sector procurement.  If that came about accidentally, then clearly it shouldn’t have, but it wouldn’t be the first or the last time it happens. If it was deliberate, then we’re into a whole different ball game. (Suggestions that “anyone but Virgin” emails were circulating, now denied, would be bad news for the Department and any individuals involved).

But other reports say, no, that’s not the case, it is all about “errors” in the calculations of risk, future passenger numbers and so on. But as we explained here, a difference in opinion over future events is not an “error”. I think inflation is going to be 4%, you say 2%. Neither of us is wrong – you might find that one of us has made our estimate in a more considered and professional manner than the other, but I’m not sure that’s an error as such. And I might still be proved right.

However, the Guardian had something that sounded much more like a “simple” cock-up.

But only on Tuesday did consultants discover that the spreadsheet on which all calculations were modelled was fundamentally flawed. The key mechanism, called the GDP resilience model, mixed up real and inflated financial figures and contained elements of double counting.

The Sunday Times had something similar.

As part of this, officials began to explore a complex economic model — known as the GDP resilience model — designed by the department to assess the riskiness of the bids and how much money should be put up as a deposit… a team of accountants from PWC were called in to investigate further. Their findings were devastating: the model was riddled with errors and double counting. Some figures had failed to take into account inflation. Formulas used to calculate variants such as future revenues or passenger numbers were simply wrong…. To make things worse, a version of the model had been sent to the train companies to devise their bids — meaning they were working on the basis of duff figures.

Then we had the statement from Kate Mingay, one of the suspended civil servants. As we said on Friday, it is most unusual for a civil servant to make a statement  like the one she made.

We also saw the Financial Times naming the other two suspendees as John Gilbert, Head of Rail Procurement and Supriya Bhol, a “procurement manager” . Gilbert is reasonably senior, as an SCS1 in grade terms, and his name had been reported earlier.  Bhol’s hadn’t - I haven’t come across him / her before, and LinkedIn doesn’t say much other than mentioning 7 years service at DfT.  From the job title though, I’m guessing a Grade 7 or even an SEO? It’s hard to see that blame for all this could rest with someone at that middle level, unless it was something very unusual.

Mingay obviously feels she is on pretty solid ground, and if it is a fault in a complex financial model, it seems unlikely that procurement people like Gilbert and Bhol are to blame. Indeed, it may be impossible to identify individuals if the flaws are in a long-standing model, perhaps adjusted over many years.

I would argue there’s been a failure of process in the sense that the model should have been verified, checked and double-checked. And why were there no financial advisers on the project? So the civil servants look at this stage increasingly like convenient  scapegoats – unless there is more to be revealed - and DfT leadership / Ministers might just be worrying now also about charges of defamation (or something similar) being brought by their own staff.

Just one other thought. Here was Francis Maude, speaking in December 2010, when a range of new non-executive Directors were announced for government departments.

“Today’s names include business heavyweights with huge experience of financial management and improving operational performance and they will play a key role in helping departments rise to the challenge and deliver further savings” .

Lord Browne, who was the ”lead” non-exec for Government said:

“Drawing on their experience, knowledge and expertise, the Non-Executive Board Members will play a principal role in bringing a more business-like ethos to the very heart of government”.

Well, the non-execs in DfT obviously failed to demonstrate their “huge experience of financial management”. Why weren’t they looking more closely at the process, the governance, the checks and balances? Indeed, as they have been around throughout the whole West Coast franchising process (unlike the Minister and the Permanent Secretary, who are both very new to DfT), isn’t there an argument that they should be picking up a bit of the blame as well?

Instead, two of the DfT non-executives have been given the task of reviewing what went wrong in this case. Let’s hope that includes looking in the mirror and asking themselves why the Board so singularly failed to pick up the problems.

Voices (8)

  1. Little Acorn:

    Maybe Virgin did know/found out about the spreadsheet issue and was holding back to see if they ended up in 1st place. Then, having found not, could not complain as they would have been out of time ascourt judgements clearly show that the unhappy bidder can’t ‘sit on its hands’ to hear the award decision. Maybe this helps to explain the media frenzy, to ‘help’ the DfT discover the problem and make a decision to cancel the award decision based on the ‘flaw’.

  2. Phoenix:

    We are falling into the trap of getting distracted by the data and formulae entry errors going into spreadsheets. This wasn’t what Branson cited when he launched his challenge. If he had, then it would have been a simple case of highlighting the errors to DfT straight away and the award would have been halted. Virgin’s issue was with the new risk assessment policy, whereby all future revenue risk – for the full 15 yrs – was transferred to the franchisee rather than shared with govt. It meant First Group’s optimistic demand forecasts got more credit than they would have under the old model. As I understand it, this change was made by ministers.

    We might speculate that the first DfT review will focus on the West Coast process and the data entry errors. Some people may be sacked. Then at 9hristmas the second review will find that he new model for revenue risk assessment doesn’t work after all. So it will be quietly changed and ministers will carry on.

    1. Final Furlong:

      Spot on

    2. life:

      Very good points and I agree, esp. last point – the issues around the model though, and attendant timing, do strike me as a little odd, and are of course related! Also of course we don’t know proportionally how “wrong” or otherwise either or both were, in terms of fiasco emphasis – I feel a spreadsheet coming on….

  3. Chris C:

    If the model is long-standing, how many previous rail contracts have been placed on the basis of incorrect calculations?

  4. Dave Orr:

    The plot thickens. The Blame Game becomes ever more murky.

    A large spreadsheet is a complex model akin to a computer program.

    It is good practice in devising a spreadsheet to have a set of test data that checks the calculation is correct and usually check sum totals that should match each other.

    The whole thing should be commented and have version control like any computer program and be tested after changes.

    When I worked for Shell International they had a published Good Spreadsheet Practice Guide and relevent training to ensure that someone didn’t wrongly calulate the value of oil reserves, profit margins etc.

    Who formulated this spreadsheet and who was responsible for supporting & maintaining it?

    Giving out the calculation used to devise winners and losers to the bidders would seem questionable practice wouldn’t it? The biddders can now simply tweak the numbers to get the final scores up.

    Does making the answer to the Rail Franchise Universe & Everything (is that 43?) from a single spreadsheet as the main indicator of the winning bidder mean that no human judgement was applied?

    Starting with First walking away from Great Western at a break point and letting the taxpayer down for their “profit share”…Score for Reliability and Trust in previous contracts?

    Why not simply do away with Rail Franchises and playing shops.

    Railway privatisation with multiple train sets and endless complex contract boundaries (franchises, tracks, signals, ticketing, train leases, new rolling stock etc)was questionable from the outset. Even Mrs Thatcher avoided that one.

    I predict lots more consultancy…….

  5. life:

    Very excellent points. The invisible man did it. Not the invisible woman, anyway.

    I encourage thorough QA of my earlier comments. However, assuming The Sunday Thunderer is right, and it does seem remarkably specific to think it’s not been well sourced, it does beg a few questions which I’m sure have very straightforward answers but at the moment seem puzzling. It may be that the model was too boring to QA from the DfT side, but if true that it was passed to Branson’s team and others, when did Virgin know it was wrong? He doesn’t seem, publicly at least, to have been complaining about the model specifically, rather the decision to favour the logic and timing of the opposing bid, and yet it seems odd that Virgin were throwing the kitchen sink at the “appeal” and yet hadn’t had someone spend their lunchbreak scenario testing the model with a few 0s and 1,000,000s etc. to flush out potential problems. And at the risk of insulting everyone, I would suggest he need only redeploy Mo Farah to do that rather than hire PwC. And of course however brief the tender process more odd still Virgin hadn’t done that before submission, I’m sure…..?

  6. Stephen Reeves:

    I was interested to read your article. I think you are right to question the role of non-execs on the DfT Board on the West Coast Mainline. One would also expect that a decision of this magnitude would have been discussed and formally agreed by the full DfT Board. Yet I have not seen any evidence that this process was followed, and the DfT website, strangely, does not have any Board minutes since late 2011.

Discuss this:

Your email address will not be published. Required fields are marked *