Wednesday, April 11, 2007

Dodgy Optimism Bias


Not just another load of Pollock's*

On BOM we hold these truths to be self-evident:
  • that virtually all public sector capital projects overrun their original budgets and timescales; and
  • that although the Private Finance Initiative (PFI) potentially offers taxpayers better value, in practice Gordo has bounced us into some shocking deals just so he can conceal the true extent of his public borrowing (eg see this blog)

We're also fully seized that it was the first of these truths that gave rise to the second: ie it was only because the public sector was such pants at managing capital projects that Gordo was able to get away with his PFI borrowing scam by arguing he was trying to save money.

As soon as he moved into No 11 Gordo got to work on boosting PFI spending. And to give expression- and teeth- to the policy, he got HM Treasury to cook up new financial appraisal rules for public officials to use when evaluating projects. Henceforth, all plans for traditionally funded public projects (ie projects funded by direct public sector borrowing) had to be costed and compared against the costs of the PFI alternative.

Fair enough, you say. Except that to make sure the dice were always firmly loaded against the "traditional" option, the new rules obliged officials to scale up all its costs with loadings for so-called "optimism bias" (the idea being that since public sector cost estimates are always understated, they need to be adjusted upwards).

Here's a summary of the loadings (taken from the Pollock paper discussed below):




(click image for bigger version)


As you can (hopefully) see, loadings for optimism bias are quite chunky. Range midpoints for cost loadings are typically 20-30%, and they run up to 200% for projects involving the development of new equipment (eg new military kit).

This makes it very difficult for traditional funding structures to compete with their PFI alternative. Little wonder then, that virtually all new schools and hospitals over the last decade have been PFI deals.

So where did these all-important loadings come from? Professor Allyson Pollock* of Edinburgh Uni's Centre for International Public Health Policy has taken a closer look at the underlying evidence. And her paper is a very interesting read (Evidence Base for Cost and Time Overrun Data in UK Value-for-Money Policy and Appraisal by Allyson M. Pollock, David Price and Stewart Player).

She quotes Gordo telling us PFI is all about value for money:

"PFI projects are being delivered on time and on budget. HM Treasury research into completed PFI projects showed 88% coming in on time or early, and with no cost overruns on construction borne by the public sector. Previous research has shown that 70% of non-PFI projects were delivered late and 73% ran over budget." (HMT 2003)

And hey, just like Bliar always promised, it's policy based on evidence, not dogma.

Except that- yes, you guessed it- the evidence actually turns out to be yet another dodgy dossier of apples and pears comparisons, inadequate data, and systematic bias.

To begin with, although the Treasury cites in evidence five separate research studies, only one actually makes a direct comparison of PFI projects with traditional non-PFI projects (2002 study by consultants Mott McDonald). And on that one Pollock comments:

"The study samples were not representative of projects procured either traditionally or under PFI. For all types of projects, the numbers in the PFI sample were so few as to provide no meaningful data. Measurement bias introduced additional sources of error. PFI performance, relative to public procurement, was not properly evaluated."

Nobel Prize research it ain't.

What should we conclude? Pollock sets out her conclusions here. In particular, she argues that taxpayers should be given much fuller access to the underlying data on PFI deals, which must be right. Gordo is a twister- as shown once again by the Great Pensions Grab revelations- and nothing he tells us can be taken on trust. We desperately need that disinfecting sunshine.

But what we should not conclude is that because Gordo has rigged his figures, that means we should junk the whole nasty business of private sector involvement and go back to traditional public sector capital investment.

Although Pollock shows that Gordo's preference for PFI is not supported by the evidence he produces, the evidence does underline the problem with traditional public sector investment- prior to project commitment, there is a systematic tendency to underestimate costs. Pollock's paper shows that for NHS hospital PFI projects, the cost escalation between the initial plan and the so-called "final business case" can easily exceed 100%- in some instances it's above 200% (see tables 2 and 3 here).

"Optimism bias" is a polite British way of putting it, but as blogged here, commenting on the results of an international study showing 90% of public projects blow their initial budgets, US think-tank the Cato Institute puts it more directly:

'The study concluded that lying, or intentional deception, by public officials was the source of the problem: “Project promoters routinely ignore, hide, or otherwise leave out important project costs and risks in order to make total costs appear low.”

Politicians use “salami tactics” whereby costs are only revealed to taxpayers one slice at a time in the hope that the project is too far along when true costs are revealed to turn back.'

However you slice it up, the reality is that public sector involvement in delivering anything is a surefire recipe for spending a lot more money than you were ever told about. Or would ever have dreamed possible.


*Footnote- BOM wishes to apologise unreservedly for such a crass and entirely predictable play on Allyson Pollock's name under her pic. She is obviously not a natural bedfellow for BOM- coming from the left, she's well known for opposing PFI, and anything else that "privatises" medicine (eg see here). But that doesn't mean she has to be wrong on everything. As we've noted before, Gordo's made such a hash of PFI that even free marketeers and Marxist planners find themselves on the same side. Some legacy.

1 comment:

  1. Bạn có muốn mua đồ bên Nhật mà lại không có người thân ở bên đó thì làm sao mà chuyển về? Hãy an tâm nếu bạn cần gửi hàng từ Nhật về Việt Nam.
    Hãy để giao nhận 247 lo giúp bạn. Khi bạn cần mua hàng bên Nhật chẳng hạn. Thì cứ việc liên hệ với chúng tôi, chúng tôi sẽ mua giúp bạn món hàng ấy và vận chuyển về Việt Nam.
    Ngoài ra chúng tôi còn nhận chuyển hàng từ pháp về việt nam, gửi hàng từ mỹ, mua hàng trên amazon,...
    Ở trong nước thì chúng tôi còn nhận chuyển hàng từ việt nam đi đức, chuyển hàng đi úc, dịch vụ gửi đồ từ việt nam sang nhật. Còn rất nhiều dịch vụ khác đang chờ bạn dấy, hãy liên hệ để biết chi tiết nha.

    ReplyDelete