34. Calculations of costs imputed to a PSC are required to include an optimism bias (NAO, pp 127-129), which assumes a tendency to unwarranted optimism in cost estimates under traditional public procurement. The estimated cost by the traditional procurement path is raised by a percentage and the resultant higher figure is the PSC compared to the private finance cost of the project. The Treasury advises public authorities to generate their own estimates of optimism bias based on their own experience, and sector-specific studies.
35. The optimism bias figures employed in PSCs are significant. For example, the NAO pointed out that the PFP for GCHQ's new headquarters project relied solely on the highly uncertain assumption that procurement by conventional means would have over-run its budget by 24 per cent and expressed scepticism about the uses of optimism bias (p 129). Dr Chris Edwards, University of East Anglia, said: "The average runover in 2002 was running at about 13 per cent for fairly routine projects like general hospitals, and since then it has actually been reduced" (Q 297). Mr Ed Humpherson, NAO, told us that "Optimism bias has been a crucial contributor to the PFI net present value being lower than the public sector comparator net present value in a very large number of cases we have looked at. Indeed, what we have tended to see as a general rule is two numbers clustering quite close together so if you took the optimism bias out of one of them it would make the public sector comparator lower" (Q 256).
36. The NAO is sceptical about optimism bias uplifts in the context of Public Sector Comparators and about applying optimism bias solely to estimates of public sector costs. The projected costs of private finance projects may also be subject to optimism, although not necessarily at the same level as in conventional public sector procurement, and in practice any overruns would normally be met by the private sector.
37. The addition of optimism bias may in many cases have had the effect, even at reduced discount rates, of tilting the comparisons of net present value which public authorities have to make, in favour of PFP and against conventional procurement. We recommend that, in order to reach a fairer basis of comparison, where a percentage uplift for optimism bias is added to the estimated Net Present Value of Public Sector Comparators, an appropriate rate of uplift should also be added to estimates of the NPV of the cost to the client under PFP.