Q91 Chair: That's not your job; that's their job. That's the whole point.
Graham Dalton: Well, it is until it goes wrong. We need to ensure that we will not have a contractor that cannot live with-
Chair: I don't think you understand our point.
David Finlay: I think that the evidence, as set out, is that the PFI costs for operation and maintenance over 30 years came in much lower than the agency's estimates, which were based on short-term conventional contracts. That raises two issues. First, why wasn't there a better estimate of what the PFI cost would be in advance of the bids coming in? Secondly, there are question marks over why you couldn't get better deals on conventional procurement for operation and maintenance over a long period.
Q92 Chair: Do you want to respond to that?
Graham Dalton: Sorry, I missed the last point.
David Finlay: The second point is that it raises question marks about whether you couldn't get better deals on operation and maintenance under conventional procurement over a long period.
Amyas Morse: You mean if you were offering the conventional maintenance over 30 years. Presumably, people would give you a pretty sharp pencil for estimating that for that length of time.
Graham Dalton: That's a fair question. If we have enough knowledge of the asset, the question is the viability of the contract. Some 15% of the motorway strategic road network is now under PFI, so we have a number of those contracts running. This is by far the most competitive that we have had.
Q93 Joseph Johnson: Can I dig into the comparison between the estimates of the present value of the cost of the contract versus the present value of the expected benefits of the contract? When you started the bidding process, what were the two figures?
Graham Dalton: Are we talking about the widening?
Q94 Joseph Johnson: Everything, the whole thing-construction and maintenance, cost versus benefit, at the start and then at the outcome.
Graham Dalton: Figure 5 gives us the table at the outset, which is the anticipated benefit-cost ratio initially.
Ginny Clarke: Figure 6 is the tender stage, and it looks at the costs and present value costs. Those were at the tender stage. Against it, we have the unadjusted standard tender.
Q95 Joseph Johnson: The reason why I am struggling is that in paragraph 12 of the report, we appear to be saying that the present value of the cost increased from £2.7 billion in May 2008 to £3.4 billion in May 2009, when the contract was awarded, versus what appears to be the expected benefit of £2.3 billion, which is mentioned in paragraph 7. Can you help me to understand that? On the face of it, it would appear that from the outset that the expected benefits were lower than the expected costs-if those two figures are comparable. If they are not comparable, can you tell me why they are not?
Graham Dalton: The net present cost of the contract is for the widening and for the maintenance and renewal of the asset over 30 years. When we work out the benefit-cost ratio, paragraph 3 might be useful. It states that the capital cost for the widening is £900 million and that it delivers £2.3 billion present value of benefits. That is the extra benefits bought by the widening.
Q96 Joseph Johnson: Why can't I have a figure that is directly comparable to the £2.7 billion and the £3.4 billion? Where is that directly comparable figure?
Graham Dalton: I don't think there was a benefit-cost ratio calculated at either stage.
Q97 Joseph Johnson: Why?
Graham Dalton: There was for the widening, because that is the benefits that you buy. For the maintenance and renewal, that is keeping a network in a steady state, so I don't believe that you do a benefits calculation for how much you are spending. You can compare the price and get the lowest net present cost.
Q98 Joseph Johnson: Would you never undertake a benefit analysis of a maintenance contract?
Graham Dalton: Because you are not buying new benefits.
Q99 Joseph Johnson: So you would never attempt to establish that. How would you measure value for money?
Graham Dalton: We look for the lowest net present cost. If it's for normal maintenance, then yes, the lowest cost-compliant bid. For this one, it's the lowest net present cost for undertaking maintenance and renewals.
Q100 Chair: That is a deeply disturbing statement to have made. We know that your ability to assess the costs of maintenance is, at best, flawed, because you have got them so wrong. We have no clue whether the figure you have ended up with actually reflects best value. All it reflects is lowest price, not best value. You do no work-and neither do your consultants, from my understanding of your answer to Mr Johnson-to assess whether those figures represent best value rather than lowest price.
Graham Dalton: We know what we are getting; there is a specification for maintenance and condition of the asset.