Select Committee on Treasury Fourth Report



24. In this Report, we use a definition of value for money which is used in the National Audit Act 1983, namely "the economy, efficiency and effectiveness with which [a body] has used its resources in discharging its functions".

25. The PFI contractor will usually borrow money from banks, etc., to finance the expenditure on building the asset, with repayment being made as payments for services begin. This borrowing will be at a higher interest rate than the Government can obtain, although the Chief Secretary said that the differential was narrowing.[43] PFI will be cost-effective, therefore, only if value for money savings more than outweigh the extra cost of borrowing. The main ways in which this can be achieved are as follows:

      (a)  private sector expertise: the private sector may be better at managing a project and save money (we return in paragraph 48 to the question of whether lessons learned for this reason can be transferred to conventionally-funded projects);

      (b)  appropriate transfer of risk: the private sector is better at handling certain types of risk than the public sector and can therefore reduce its cost (we return to this issue in paragraph 27);

      (c)  scope for innovation: if the authority specifies requirements in terms of the results to be achieved ("outputs"), this gives the contractor scope for innovation; in particular the contractor's responsibility for both building and maintenance encourages the construction of low-maintenance buildings, which cost more to build but result in a saving over the term of the project.[44] (This may also result in better maintained buildings, a point to which we return in paragraph 31.) The MCG said that there had been "a tendency to be prescriptive as to the materials used, rather than specifying quality and durability" in some of the early roads projects.[45] The BSA said that the area where the "biggest single savings" will be achieved is where the private sector has the "greatest scope for innovating [in] the way in which that service is delivered".[46]

26. The CBI characterised projects where PFI was likely to be successful as those "genuinely combining capital and service requirements; where the risks are primarily commercial; where scope for innovation exists; and with skilled and committed public sector management".[47] We agree with this analysis.

Transfer of risk

27. The cost of providing the PFI service can be affected by an unexpected event: the cost of the risks concerned depends not only on the potential amount of extra money involved, but also on the (estimated) likelihood of the event occurring.[48] The principle of transfer (or assignment) of risk is that each risk should be allocated to the party which is most able to reduce and manage the risk (by reducing its size, its likelihood or both), and therefore to reduce its cost. As the NAO puts it, "The Department should ... have sought to achieve not the maximum but rather the optimum transfer of risk, which allocated individual risks to those best placed to manage them".[49] Mr Adrian Montague told us that "if you transfer risks that the private sector cannot control either they will refuse to accept the risk or they will charge a price for that risk which may seem doubtful value for money."[50] Mr Keith Clarke of the MCG said that in the early stages there had been "almost ... a dogma that all risk shall be transferred to the private sector" and that it was the change to risk "being put where it is most efficient to be managed" which had allowed the increased flow of PFI contracts.[51] This was echoed by Professor Glaister.[52]

28. During the development of PFI various types of risk have been identified and their allocation considered. The NAO have recorded these as follows. The design and construction risks (e.g. that construction costs more, or takes longer, than expected) should be for the contractor (and indeed some of them would be borne by the contractor under conventional procurement). The risk that operating costs may rise faster than expected can be shared by linking it to an appropriate index.[53] Demand risk (the level of use of the service) is not usually controllable by the contractor, but can be transferred if it is not directly influenced by the public sector or if the cost is volume related. Residual value risk (the risk associated with taking over the project assets at the end of the contract) will vary depending on whether the assets are likely to be needed for their original purpose or can be used for other purposes. There is a technology/obsolescence risk associated with IT projects or others dependent on specialist equipment. Finally, there is regulation risk, where the NAO supports the general principle that the contractor should assume risk for regulatory changes applying to all industry, but not for "highly specific" legislation.[54]

29. UNISON pointed out that the public sector had to assume the risk that an essential service is not provided and quoted the problems with the Passport Agency in the summer of 1999 as an example where "regardless of who was delivering the service and on what terms, it was the government that had to take the final responsibility for what went wrong"; it said that these risks were not taken adequately into account.[55] The CBI pointed out (in the case of a PFI contract involving building) that if the contractor fails after the project has been built, "the Government really is in a better position ... they are going to get a completed hospital or a completed road. They can probably fairly easily replace the operators, often the costs relating to which are quite low in relation to the capital spend".[56] The NAO view is that any compensation to the contractor if the contract is cancelled because of contractor default (which may be required if a PFI bid is to obtain financing from banks etc.) should be limited so that the total cost (including the cost of appointing a replacement contractor, as well as the compensation) should not exceed the cost envisaged in the original contract.[57]

30. We recognise that the risks which are appropriate for transfer may vary from contract to contract. In the course of approving future contracts we expect that commissioning authorities will set out clearly the risks that are being transferred, how they have been valued and justification for their transfer. When risk is transferred to the contractor, we expect to see arrangements in place to make sure that the contractor may not escape liability for that risk by returning liability to the commissioning authority or to some other public body. For example, we are particularly concerned that the private sector should not be able to minimise operational risk through reduced service specifications once the contract has begun.

Effect on maintenance

31. One reason why a PFI contract may be seen as more expensive is that, in order to ensure the desired level of service, it will require the contractor to maintain the asset to a suitable standard. Although such maintenance will also be desirable in the case of a public procurement, it may be deferred because of a shortage of funds. The Major Contractors Group said that "public sector building stock has suffered from lack of maintenance and this has, in turn, adversely affected the public services which they house. ... In contrast, the buildings now being delivered under the PFI will be maintained to the specified standard for the full life of the concession".[58] The CBI also mentioned that PFI can protect expenditure on maintenance.[59] UNISON, however, pointed out that this might cause disproportionate reduction in non-PFI public services in any future cuts in funding.[60]

The public sector comparator

32. The purchaser of the services is usually expected to demonstrate that the overall cost of a PFI contract is less than that of a theoretical public sector procurement (the Public Sector Comparator or PSC), rather than simply relying on competition between different PFI contractors. Whether the PSC will demonstrate that PFI would be the best option depends, of course, on the assumptions made in calculating it.

33. The Taskforce issued new guidance in October 1999 about how to compare costs.[61] In the press release launching the guidance, the Government stated: "Public sector comparators alone are not the basis for assessing value for money". The NAO (quoted in the press notice) says that the PSC, though a "key part" of the evaluation, "should not be a pass or fail test, but needs to be seen in the context of a systematic evaluation of all the costs and benefits".[62]

34. The cost comparisons take into account that a payment made later effectively costs less, by discounting future payments using a discount rate specified by the Treasury, currently 6 per cent a year. In other words, a cost of £1 deferred by one year will be treated as 94p.[63] Because payments in a PFI project are generally made considerably later than in a conventional procurement, the discount rate chosen is crucial to the cost comparison: a small increase in the discount rate used would make the PFI option look very much more attractive. For example, the Treasury estimates the savings on the first four road deals was in the region of 12 per cent using a 6 per cent discount rate, but 23 per cent using an 8 per cent discount rate.[64] Professor Allyson Pollock, Head of the Health Services and Health Policy Research Unit at University College, London, and her colleagues calculated that a change of discount rate from 6 to 5½ per cent would change the economic advantage of the Carlisle Hospitals' PFI scheme from £1.7m to -0.9m (on a total cost of about £170m).[65] Mr Frank Dobson, then Secretary of State for Health, told the Health Committee on 22 July 1999 that the system of appraisal is based on a system established by the Treasury in 1983.[66] Ms Kate Barker of the CBI suggested that it would make sense, when assessing a project, to look at more than one discount rate.[67] The Chief Secretary pointed out that rates higher than 6 per cent were used in France and the United States and would be content if the Treasury's periodical reviews of the discount rate were externally audited.[68]

35. Several witnesses questioned the validity of PSCs: the Major Contractors Group said "despite extensive guidance, contractors continue to find that public sector comparators are often unrealistic, do not properly allow for the risks which the public sector would retain under a publicly-funded solution, and do not accurately cost the maintenance and operation elements of a project". In oral evidence their witnesses welcomed the new guidance but cast doubt on the extent to which it would be followed.[69] Mr David Clements of the Business Services Association said, "Frequently the output specification actually demands a higher standard from the private sector partner than the standard to which the services are [currently] being provided" and called for a corresponding adjustment to the PSC.[70] Professor Pollock and her colleagues argue that the PSC is increased to take account of risks to the overall cost, but that this element is to some extent included in the 6 per cent discount rate and in some cases NHS trusts have added amounts for risks which are also present in the PFI option as the contractor will not be taking them on. In addition, the adjustment for risk transfer is the result of a "very arbitrary judgment".[71] The TUC suspected that authorities which had no alternative to PFI calculated the PSC in "a half-hearted way" and did not revise it in the light of changes to the proposal.[72]

36. Another criticism was the lack of openness in PSC calculations. Professor Pollock said, "we have grave concerns about public and financial accountability, the lack of disclosure and the lack of transparency".[73] Professor Glaister pointed out the judgments which needed to be made in comparing the PSC with the proposed PFI project and suggested that these judgments should be "open to scrutiny and review" by a "visibly independent body", perhaps the National Audit Office or the Audit Commission, as well as some sort of public scrutiny.[74] The Chief Secretary agreed that the calculation should be subject to audit but did not believe that "a transfer of the responsibility for undertaking [the comparison] would be appropriate".[75]

37. The CBI suggested that "emphasis should move from a hypothetical Public Sector Comparator, which has all sorts of difficulties anyway, to one where you are looking for real comparisons, whether it is with other parts of the portfolio or the private sector or wherever else".[76] The Government expects benchmarking of this sort (as well as a PSC) "where there has been only one final bid or, exceptionally, where after a preferred bidder has been selected it is necessary to change the specification".[77]

38. The Public Sector Comparator should contain a clear statement of the risks which have been quantified and included in it; and there should be an independent element in its construction and in the evaluation process. The possibility of comparisons with other private sector projects should be pursued wherever such information is available; such comparisons will be particularly valuable when the Public Sector Comparator is unreliable because it reflects a course of action which is unlikely to take place.

Ensuring competition

39. Another contribution to value for money is achieved by ensuring competition between alternative bidders, and that the decision on a preferred bidder is not made too early.[78] The TUC said, "The common use of a shortlist of one provider during the latter stages of negotiations puts the shortlisted company in a monopoly position, which has resulted in a significant cost drift upwards on a number of occasions".[79] However, Mr Charles Cox of the CBI said that the issue is "quite finely balanced" and there was pressure on the contractor until the contract is signed. "If you keep a lot of people in for a long time it increases both the time and the cost of the process, which frankly is one of the major issues ...".[80] (The MCG and BSA estimated bid costs at about £1m per contract and, as about one contract in three or four was successful, this cost was being taken into account in the bids.[81]) Mr David Metter of the CBI explained that any changes in price after the appointment of a preferred bidder had to be justified "very carefully" and would arise because the authority changed the requirements of the project.[82] He added that if the Government attached importance to keeping more than one bidder until the contract was signed, this would increase costs and the Government should compensate for bid costs beyond a certain stage in the process. He said that the Dutch Government was doing this on a high speed rail link project.[83] Ms Barker said that a company which exploited its position would be in a poor position to negotiate any other contracts.[84] The TUC would be "concerned if unsuccessful bidders were to be reimbursed unless there was clear evidence that this would result in a net saving".[85]

40. Mr Montague said that the guidance was now that the authority "absolutely must not appoint a preferred bidder until the deal is done". He added that "it would only be in exceptional cases that some kind of tender cost support should be given".[86]

41. We accept that there is a trade-off between competition and the length and cost of negotiations. The best way to ensure competition is for the authority to clarify as early as possible in the process the outputs which are wanted, to avoid the need for too many changes as the negotiations proceed.

Monitoring of projects

42. We asked about the way in which PFI contracts would be monitored during their long life. The BSA said that the contracts contained mechanisms for making the changes which would be bound to be necessary.[87] Mr Chris Elliott of Barclays Private Equity Ltd told us that the PFI system encouraged an "adversarial relationship" where "the public sector is not sufficiently involved in the dynamic nature of the business" after the contract is signed. "It effectively acts only as a policeman and not as a partner". He pointed out that this attitude might make implementation of any required changes more difficult. He suggested that the "most conventional" way for the public sector to share in the performance of a PFI or PPP project would be to take an equity share.[88] The Taskforce has issued guidance entitled How to manage the delivery of long term PFI contracts,[89] covering such issues as "contract management" (checking that the contract terms are understood and fulfilled) and performance monitoring, the differences from a conventional procurement and the resolution of disputes.

Initial findings on value for money

43. National Audit Office reports on PFI projects examine the possibility of value for money gains. For the Skye Bridge a public sector comparator was not required but "most of the project's constituent costs ... were determined competitively or were clearly in line with market rates, and to this extent there is assurance that the Department selected the best available privately financed deal and secured value for money";[90] the NAO reported that the NIRS 2 National Insurance System "remains strikingly good value for money" subject to delivery of the final system to specification;[91] there were forecast savings of 17 per cent for Bridgend Prison but minimal savings at Fazakerley Prison;[92] the forecast savings on the first four roads contracts depended heavily on the discount rate and at the 6 per cent rate only two of the four roads showed savings.[93] The appraisal of the Home Office Immigration and Nationality Directorate's Casework Programme "forecast that the benefits of the PFI deal would greatly exceed its costs and that the individual costs in the winning bid were reasonable";[94] for the A74(M)/M74 Motorway the "apparent benefit of the privately financed road" may be somewhat less than the £17m estimated by the department, but the price can "still be expected to remain value for money";[95] the saving by transferring the estate of the Department of Social Security to the private sector is estimated as £560 million or 22 per cent of public sector ownership costs;[96] the PFI contract for the Dartford and Gravesham Hospital is expected to be £5.1 million or 3 per cent cheaper than the public sector comparator after errors in the calculation of the PSC have been removed;[97] the procurement of non-combat vehicles for the RAF is estimated to save £5.8 million (23 per cent);[98] and the project for the Contributions Agency Newcastle Estate had a higher direct cost than the alternative of remaining in existing accommodation; the Agency considered that the extra direct cost of the deal was justified by the operational benefits they expected it to bring, and a PSC had not been constructed.[99]

44. The Arthur Andersen Report mentioned in paragraph 15 examines estimated savings achieved by a sample of 29 PFI projects for which PSCs were available, and calculates the average at 17 per cent. For 12 of these projects, insufficient information was available on the valuations placed on the risks to be transferred to the contractor, a point which we raised with the Treasury witnesses, who said that the requirement for such information was now in the guidance.[100] For the remaining 17 projects the savings were heavily dependent on the risk valuations, which accounted for 60 per cent of the total cost savings. Six of these projects were shown to be cheaper than the PSC solely because of the risk valuations.[101]

45. As Mr Metter of the CBI pointed out, it will be necessary to wait until "the process matures further" before "one can see systematically whether there are savings been generated",[102] although Mr Cox pointed to the early NAO results.[103] Mr Clements of the BSA expressed disappointment that there was no comparative study "to take ... two projects that have been procured, one by traditional procurement and one by PFI and track them through their life. That is the only way that we will know what the real answers are".[104]

46. Some PFI projects have now reached the end of their construction phase, and it should be possible to draw firm conclusions about the savings obtained. The benefits for the operational phase are necessarily less tested at this stage.

47. The Government has announced that departments should evaluate all their PFI projects as well as conventionally procured projects and that the Government is also considering the "benefits and practicalities of establishing a central system to collect information on project performance and provide a facility to benchmark performance against comparable PFI and other projects".[105] We welcome the Government's new commitment that departments should evaluate PFI projects after they have been implemented. This evaluation should include both costs and service delivery, with comparisons against conventional procurement. Research in this area might also usefully be undertaken by the NAO as part of their value-for-money investigations. We also strongly support the suggestion of a central system to collect information on each PFI project to enable the comparisons to be made. In order that PFI can be used in the areas where it offers the greatest gains in value for money, the research should also establish which types of projects are most suitable for PFI, and also which are the particular features that make a project suitable for PFI, such as its size, the expected life of the asset, the type of risks involved and the relationship of PFI services to non-PFI services.

Learning lessons for conventional procurement

48. Mr Spackman supplied us with a note about the possibilities for drawing conclusions from the operation of PFI projects which could be used to improve conventionally-financed procurement. These could arise from methods of operation, from innovations in design and construction, or in financing.[106] UNISON also suggested that expertise gained in the procurement process could be used for future procurements outside PFI, thereby transferring part of the benefits of PFI to conventional projects.[107]

49. We welcome the reference in Public Private Partnerships: The Government's Approach to the possibility of the public sector benefiting from innovative approaches developed in PFI deals,[108] and expect that, as part of its work, the new Office of Government Commerce will monitor PFI projects systematically for features which can be transferred effectively to publicly-funded projects.

43  Q 307. Back

44  The MCG quote an example of better floor finishes in hospitals (evidence p 4, para 9). Back

45  Evidence p 6, para 20. Back

46  Q 96. Back

47  Evidence p 42, para 11. Back

48  See Taskforce Technical Note 5, How to Construct a Public Sector Comparator, p 34, Box 6. Back

49  NAO, Estimating the value for money of deals under the Private Finance Initiative, HC (1998-99) 739, August 1999, Appendix 2, para 1. Back

50  Q 326. Back

51  Q 58. Back

52  Q 114. See also Q 214 (Mr Charles Cox, CBI). Back

53  The CBI said that contracts could specify a "benchmarking" operation at 5-7 year intervals to reset such variables as utility prices to market levels (Q 211). Back

54  NAO, Estimating the value for money of deals under the Private Finance Initiative, HC (1998-99) 739, August 1999, Appendix 2. Back

55  Evidence p 58, para 1.6. Back

56  Q 213. Back

57  NAO, Estimating the value for money of deals under the Private Finance Initiative, HC (1998-99) 739, August 1999, Appendix 2, paras 24-6. Back

58  Evidence p 4, para 9. See also Q 3. Back

59  Evidence p 41, paras 5, 7. Back

60  Evidence p 58, para 1.5. Back

61  Treasury Task Force: Technical Note No. 5: How to construct a Public Sector Comparator. Back

62  HM Treasury News Release 177/99, 29 October 1999. Back

63  This does not include an allowance for inflation, as the costs are calculated at current prices. See Q 351 and Taskforce Technical Note No. 5, How to construct a Public Sector Comparator, para 3.8.9. Back

64  Evidence p 72, para 30 and footnote. Back

65  Declan Gaffney, Allyson M. Pollock, David Price, Jean Shaoul, "PFI in the NHS-is there an economic case?", British Medical Journal, 10 July 1999, p 118, Table 3 (quoted in Q 107). Back

66  HC (1998-99) 469-iii, Q 449. Back

67  Q 205. Back

68  Q 357, 360. Back

69  Evidence p 5, para 14; Q 16. Back

70  Q 17. See also Q 110 (Mr Chris Elliott). Back

71  Declan Gaffney, Allyson M. Pollock, David Price, Jean Shaoul, "PFI in the NHS-is there an economic case?", British Medical Journal, 10 July 1999, p 118; Q 107. The adjustment for risk transfer recognises that the costs assessed in the PSC are subject to risks not present in the PFI contract (because the contractor is undertaking them) and therefore need to be increased to take account of these risks (both their size and their probability). Back

72  Evidence p 67. Back

73  Q 102. Back

74  Q 104-5. Back

75  Q 357. Back

76  Q 202. Back

77  Evidence p 74, para 54. Back

78  See Committee of Public Accounts, Twenty-third Report, Session 1998-99, Getting better value for money from the Private Finance Initiative, HC 583, para 23. Back

79  Evidence p 4 section 2; p 66. Back

80  Q 193-4. Back

81  Evidence p 5, para 11; Q 34. Back

82  Q 194-5. See also Q 198. Back

83  Q 196. Back

84  Q 196-7. Back

85  Evidence p 66. Back

86  Q 365, 367. Back

87  Q 90-1. See also Q 234 (CBI). Back

88  Evidence p 24-5; see also Q 116-18. Back

89  Technical Note No. 6, 1999. Back

90  NAO, The Skye Bridge, HC (1997-98) 5, para 20. Back

91  NAO, The Contract to Develop and Operate the Replacement National Insurance Recording System, HC (1997-98) 12, para 12; but for subsequent delays see: Committee of Public Accounts, 22nd Report, Session 1998-99, HC 182, Delays to the new National Insurance Recording System. Back

92  NAO, The PFI contracts for Bridgend and Fazakerley Prisons, HC (1997-98) 253, p 46 fig 10. Back

93  NAO, The Private Finance Initiative: The First Four Design, Build, Finance and Operate Roads Contracts, HC (1997-98) 476, para 15. Back

94  NAO, The Immigration and Nationality Directorate's Casework Programme, HC (1998-99) 277, p 4 para 18. The net benefit of the deal would be increased if expected wider benefits were to have been included in the calculation. Back

95  NAO, The Private Finance Initiative: The Contract to Complete and Operate the A74(M)/M74 Motorway in Scotland, HC (1998-99) 356 p 6 para 20. Back

96  NAO, The PRIME project: The transfer of the Department of Social Security estate to the private sector, HC (1998-99) 370 p 3 para 9; these figures are present value savings over the 20-year duration of the project. Back

97  NAO, The PFI Contract for the new Dartford and Gravesham Hospital, HC (1998-99) 423, p 4 para 12. Back

98  NAO, The Procurement of Non-Combat Vehicles for the Royal Air Force, HC (1998-99) 738, para 1.36. The announcement by the RAF referred to savings of £17 m but this was without applying a discount rate to payments made later. Back

99  NAO, The Contributions Agency: The Newcastle Estate Development Project, HC (1999-2000) 16, p 5 paras 18-19 and p 8 recommendation 8. Back

100  Q 335-42. Back

101  Arthur Andersen Report, paras 5.3-5.9. Back

102  Q 209. Back

103  Q 214. Back

104  Q 67. Back

105  HM Treasury, Public Private Partnerships: The Government's Approach, p 32. Back

106  Appendix 6. Back

107  Evidence p 58, para 1.4. Back

108  HM Treasury, Public Private Partnerships: The Government's Approach, p 33 para 22-3. Back

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries

© Parliamentary copyright 2000
Prepared 23 March 2000