House of Commons - Private Finance Initiative

Private Finance Initiative - Treasury Contents

4 Future investment

Rules and principles for the use of private finance

72. As mentioned in the previous section of the Report all PFI projects have to undergo a VfM assessment. In a paper for the Lords Economic Affairs Committee the NAO made the following comments about this:

[...] like any financial model, they [the VfM assessment] cannot be relied upon as a sole source of assurance. They are susceptible to manipulation and we often find problems with their implementation.[143]

We consider some of the possible changes to the current VfM assessment later in the report. Any financial model, such as the current VfM assessment, can be subject to manipulation so it should never be used alone as a pass or fail test for the use of PFI.

73. In the 1980s the use of private or additional finance was governed by what were known as the 'Ryrie Rules'. These required that private finance could only be used if:

there were no favourable risk terms, such as a government guarantee;
projects yielded benefits in terms of improved efficiency and profit commensurate with the cost of raising risk capital; and
that use of private finance could not be additional to public finance. In other words, public expenditure would be reduced, pound for pound, in consequence of the use of private finance.[144]

David Heald and Alisdair McLeod went onto explain in their paper on the Ryrie rules that

The rationale for this provision was that there is little macroeconomic difference between the government borrowing on the market to finance public expenditure generally and the private sector borrowing for essentially public projects. The objective of the Ryrie Rules was to stop ministers from insulating private finance from risk so that it could be used to circumvent public expenditure constraints.[145]

74. The Private Finance Initiative was a departure from the Ryrie rules as it allowed private finance which would be additional to public finance. In 1992 the then Chancellor, Norman Lamont, said in his Autumn Statement:

[...] we will allow greater use of leasing where it offers good value for money. As long as it can be shown that the risk stays with the private sector, public organisations will be able to enter into operating lease agreements, with only the lease payments counting as expenditure and without their capital budgets being cut.[146]

The importance of risk transfer and value for money were evident as PFI started. In a speech to the CBI Conference on 8 November 1994[147] the then Chancellor Kenneth Clarke emphasised that the guiding principles of PFI were that:

the private sector must genuinely assume risk without the guarantee by the taxpayer against loss; and
value for money must be demonstrated for any expenditure by the public sector.

75. As the use of PFI has progressed there has been more detail on the approach in regard to risk transfer. In 1995 HM Treasury noted that "risk should be allocated to whoever is best able to manage it"[148]. This approach is detailed in HM Treasury's document Meeting the Investment Challenge:

The Government's approach to risk in PFI projects does not seek to transfer risks to the private sector as an end in itself. Where risks are transferred, it is to create the correct disciplines and incentives on the private sector to achieve a better outcome.

Successful PFI projects should therefore achieve an optimal apportionment of risk between the public and private sectors. This will not mean that all types of risks should be transferred to the private sector. Indeed, there are certain risks that are best managed by the Government; to seek to transfer these risks would not offer value for money for the public sector.[149]

76. Evidence we have seen suggests that the high cost of finance in PFI has not been offset by operational efficiencies. Much more robust criteria governing the use of PFI are needed. These should take precedence over the current VfM assessment. If and only if a project is deemed to pass these criteria should the option of private finance be considered. In our view PFI is only likely to be suitable where the risks associated with future demand and usage of the asset can be efficiently transferred to the private sector. We recognise that this may over time sharply reduce the aggregate value of remaining PFI projects but the higher cost of capital that remains will be easier to justify to the taxpayer.

Different sectors and circumstances

77. When gathering evidence for this inquiry we posed a number of questions. Two of these were: 'Are there particular projects which are suited for PFI?' and 'In what circumstances are PFI deals suitable for the delivery of services?'. We were interested to understand whether PFI was more appropriate within certain sectors and circumstances. Skanska told us that:

We believe that PFI is most suitable for the delivery of services that contribute to whole-life costing benefits and which are stable and predictable over the long term. PFI is less suitable for services that need to flex significantly over time to reflect changes in public service delivery, demographics or technology.[150]

Others also expressed the view that PFI was suited to operations where use was predictable over time and less suitable for services that were likely to need change. The CBI told us: "PFI works well when the risks of a project can be identified, quantified and transferred appropriately." It considered that "build and service contracts that have been used to provide schools, simple healthcare facilities and housing have been successful" as had "economic infrastructure projects, which have seen roads, railways and airports built and maintained". However it considered that schemes that "introduce complex technology risk, or in which future outcomes cannot be readily forecast may be less appropriate"[151] for PFI.

78. Transport for London told us that "PFI may be suitable" in circumstances where "the public sector can define its long-term needs and wants a single integrator of the delivery of that service". It also pointed out that projects that were the "least successful were all bespoke".[152] Steve Allen told us:

If you look at things like roads or new railways, where once you have designed where the transport scheme is going to go, you fundamentally are not going to change it, those have been more successful examples of PFIs than things that are intimately involved with the operations of transport; for example, the experience of the London Underground PPP, where it was much too closely intertwined with the day-to-day operations.[153]

Martin Blaiklock told us that in his opinion "Typically, 'accommodation'-type projects, where the underlying demand and service output over 30 years does not change, are suitable for PFI treatment."[154]

79. Owing to the current high cost of project finance and other problems related to PFI we have serious doubts about such widespread use of PFI. There are certain circumstances where PFI is likely to be particularly unsuitable, for example, where the future demand and usage of an asset is very uncertain and where it would be inefficient to transfer the related risks to the private sector.

Re-examining the VfM assessment

80. Prior to commencing the procurement of a PFI scheme, central government guidance requires the sponsoring authority to draw up an Outline Business Case (OBC), in which the rationale for the project is presented to ministers for approval. The OBC includes details of a Procurement Route Comparison, in which the projected Net Present Cost of the proposed PFI project is compared with that of an identical scheme carried out on the basis of conventional (i.e. non-privately financed) procurement.

81. Despite the significantly higher cost of private finance, the vast majority of projects submitted to the Procurement Route Comparison exercise continue to find in favour of the PFI option. To a significant extent, this reflects the incentives that public authorities (and their sponsoring departments) are subject to which favour the PFI methods (see Accounting and budgetary incentives of PFI), and the fact that the model is, as the National Audit Office has noted, "subject to manipulation".[155] We understand that the Treasury is making some changes to the guidance which determines the current quantitative element of this system, and will publish these in the autumn, but regards the system as fundamentally sound. We believe that a financial model that routinely finds in favour of the PFI route, after the significant increases in finance costs in the wake of the financial crisis, is unlikely to be fundamentally sound. The Treasury should seek to ensure that all assumptions in the VfM assessment that favour PFI are based on objective and high quality evidence.

82. In this report we are not able to examine every part of the VfM assessment in detail. We have therefore chosen to consider just two of the areas in more detail—the 'optimism bias' and the adjustment for tax. In the quantitative assessment component of the procurement route comparison, two models are constructed (one for a PFI and one for a conventional public procurement) where the specification of the facility is the same, as are many of the projected costs and risks. However, risks that in the public authority's view would be borne by them under conventional procurement, but which in the PFI solution would fall on the private sector, are valued and a percentage is added to the costs of the conventional procurement route. In the Treasury spreadsheet,[156] risks transferred to the private sector in this way are identified as 'optimism bias'—i.e. the likelihood that capital and operating costs will prove to be substantially higher than those estimated at the time of the OBC assessment.

83. This uplift is designed to reflect the fact that there is "a demonstrated systematic tendency for project appraisers to be optimistic".[157] This uplift to the estimated cost of the capital expenditure undertaken under a conventional procurement option, for example, is commonly in the range of 15-20%. However, the evidence base on which optimism bias adjustments are made is unclear. The key source appears to be a study published in 2002 by a prominent technical advisory firm within the PFI industry, Mott MacDonald.[158] This study has been called into question by scholars due to methodological concern—specifically, the non-comparability of projects; small sample size and numerous source of measurement bias. They noted that the "PFI sample contained only 11 projects, although 451 PFI construction schemes were completed" and there was an "over-representation of atypical schemes in the conventional procurement sample and under-representation of them in the PFI sample".[159]

84. The Treasury should ensure that guidance regarding Optimism Bias is based on objective, high quality and, as far as possible, contemporary evidence. The Treasury should not approve the PFI projects of departments or public authorities that fail to produce such evidence in support of their Outline Business Cases. We believe that the comparison of procurement routes should take place on the basis of the PFI model and a public procurement model, in which there is a serious attempt to fix prices and therefore transfer risk.

85. Hardwired into the Treasury Value For Money Guidance spreadsheet is an assumption that, owing to the use of private finance, PFI will lead to additional tax being received by the Treasury. The current guidance follows advice from the prominent PFI financial advisory firm KPMG[160] that corporation tax which would be received by the Treasury should be deducted from the costs of the PFI option (or added to the PSC, which has a similar effect). At the time of the KPMG report corporation tax was 30%. It is currently 26% and is due to fall to 23% by 2014.[161] The corporation tax adjustment can be between 2% and 22%, of the total costs of a project, depending on the assumptions in the spreadsheet. These projected corporation tax levels are significantly higher than would be expected by the companies involved in PFI projects[162] and assume a level of corporation tax exposure that is likely to be much greater than Special Purpose Vehicles actually pay on average. Corporation tax SPVs are normally a private 'shell' company with no assets, and owned by the shareholders in the project. The SPV typically borrows approximately 90% of the capital required to develop the project, and receives income when the project is operational.

86. As well as the fact that interest paid by the highly leveraged SPV is tax deductable the SPV is also likely to adopt sophisticated tax limitation strategies. Indeed, an SPV that can organize its tax planning efficiently, including the use of transfer pricing and off-shore registration and claim all possible tax allowances, might pay less tax than the conventional alternative. For example, the STEPs deal, providing property and premises for the two Departments which were to become HMRC, was negotiated with Mapeley STEPS Limited, a property holding company registered in Bermuda.[163] A study of the early PFI highways projects found higher rates of profit, and low payments of corporation tax, and questioned the assumption that high rates of tax would be paid.[164] Evidence we received shows that ultimate ownership of over 90 PFI projects has moved offshore.[165]

87. The current 'tax adjustment' is not based on the best available evidence and acts to bias the assessment towards choosing PFI. Private companies entering into contracts with the public sector will quite reasonably seek to minimise their tax liabilities. Governments may also vary tax rates. The assessment exercise which evaluates the value for money of different procurement routes must take this into account.

88. As part of the Lords Economic Affairs Committee investigation into PFI in 2009-10 the National Audit Office produced a paper which raised doubts about the use of the VfM assessment financial model. In particular the NAO noted that "we have yet to come across robust cost analysis between procurement routes, that tests the assumptions of cost efficiency set out in business cases."[166]

89. The National Audit Office should perform an independent analysis of the VfM assessment process and model for PFI. It should audit all of the assumptions within the model, and report on whether or not these are reasonable. This test of the VfM assessment model should, where possible, be based on representative and up to date samples of data.

Investment in public infrastructure

90. The importance of investment in infrastructure was expressed in numerous submissions. Skanska's submission said that there was "a proven need for infrastructure in the UK that is currently unfulfilled".[167] Balfour Beatty also made a similar point and considered that there had been underinvestment in the past:

Balfour Beatty believes that sustained investment in infrastructure is vital to the future of the economy. From 1999-2008, UK public investment as a percentage of GDP was lower than almost any other OECD country and almost half the average of G7 countries.[168]

This view agreed with a paper of December 2009 from the Institute of Civil Engineers. This noted that:

The UK suffers from a historic under investment in infrastructure, which the OECD has identified as a major factor holding back our economic performance. This has also been acknowledged by HM Treasury. As a result the World Economic Forum ranks the UK 33rd in the world for the quality of its infrastructure.[169]

Dieter Helm agreed with the need for investment in his additional submission to the Committee: "The UK requires a very significant increase in infrastructure spending, reflecting a combination of new policy priorities and the failure to maintain and enhance existing assets."[170]

91. Balfour Beatty also pointed out in its written evidence that "investment in infrastructure has a higher economic multiplier than other types of government expenditure".[171] The Office for Budget Responsibility own analysis agrees with this—its estimates show that capital expenditure has the greatest impact of the fiscal multipliers.[172]

Table 3: Estimates of fiscal multipliers

Impact multipliers

Change in VAT rate
Changes in the personal tax allowance and National Insurance Contributions (NICs)
AME welfare measures
Implied Resource Departmental Expenditure Limits (RDEL)
Implied Capital Departmental Expenditure Limits (CDEL)
0.35
0.3
0.6
0.6
1.0

Source: OBR, Budget Forecast, June 2010, p95, Table C8

The UK Contractors Group made the point that "more construction investment would help to stimulate the economy and employment".[173]

92. As investment in infrastructure is so important for the economy it is essential that the most efficient form is pursued and also that any changes should be phased to keep disruption in investment plans to a minimum. The most straightforward way for government to phase out PFI while continuing and even increasing investment is directly to fund capital spending. With the cost of government borrowing at historic lows and at a significant discount to other forms of finance there is a strong argument to be made that this would be the most efficient form of financing and therefore would release higher levels of investment at the same cost.

93. Any increase in direct capital investment would inevitably lead to higher borrowing figures in the short term as debt would be fully transparent—unlike with PFI where most of the liability is not part of government borrowing figures. However it is unclear why this should stop the government acting, particularly if in the long term PFI is less affordable. Dieter Helm considered that continuing with PFI would in effect be imposing a tax on us all:

if [...] the conventional PFI approach continued, the economy and society bears a considerable deadweight welfare loss through the high cost of capital. A rule based and flawed accounting methodology would be in effect imposing a tax on us all.[174]

An increase in government borrowing to replace PFI investment should not make it harder for the Government to meet the fiscal mandate which the Office for Budget Responsibility monitor. As the borrowing is for capital investment it will not increase the cyclically adjusted current balance which the OBR measure. Also if any increase in borrowing occurs before 2015-16 the supplementary target will also be unaffected. The coalition government increased capital budgets at the Spending Review without any direct impact on the fiscal mandate so this shows that such an approach is possible:

[...] the Spending Review has increased the capital envelope by £2.3 billion a year by 2014-15 relative to the Budget plan in order to ensure that capital projects of high long term economic value are funded. This change has no direct impact on the fiscal mandate, which targets the cyclically adjusted current balance, and will also not alter the year in which public sector net debt as a percentage of GDP begins to fall.[175]

94. Sustainable investment in public infrastructure is important for the long term health of the economy. We also recognise the paramount importance at the current time of stabilising the public finances. The Treasury will need to consider using more direct government borrowing to fund new investment. Replacing some PFI with direct public sector investment would not necessarily result in a higher financial liability for the Exchequer. It would mean that the debt was more transparent, as it would be held directly by government rather than through the intermediary of an SPV. An increase in government debt to replace PFI investment should also not necessarily make it any harder to meet the fiscal mandate. Continuing to use an inefficient funding system such as PFI is likely in many cases to increase the overall burden on taxpayers for the provision of public sector capital projects. If, rather than using PFI, the lower financing costs of government are utilised, we have seen evidence that investment can be increased significantly for the same long term funding costs.

95. PFI is a procurement model where the private sector manages the design, build, finance and operation (DBFO) of public infrastructure. If the public sector funds the investment this changes the financing element of the project but this can still accommodate a high level of private sector involvement. There may be merit in making more use of a design and build (DB) model using a fixed price contract to place risk with the private sector over the construction period. There will be other circumstances where a design, build and operate (DBO) model is most appropriate. Both the DB and DBO model allow government to benefit from its lower cost of funding while transferring significant risk to the private sector.

Current contracts and existing deals

96. £60 billion worth of capital investment (in 2010 terms) have already been committed to by PFI investors under successive governments.[176] One important question to consider is what is to be done with the PFI contracts that have already been signed and the assets which are already being delivered under PFI agreements. Andy Friend suggested that refinancing was one option: "I think maybe we are now in the territory where the public sector might contemplate having a right to refinance the senior debt and the capital structure of such propositions when you get into the operational stage."[177] Dieter Helm also agreed that there should be a right to demand refinancing once construction risk ends in a contract.[178]

97. Refinancing with government debt would be the most straightforward way to allow renegotiation of contracts, replicating what Steve Allen told us that TfL have been able to do with some of its contracts.[179] This refinancing would result in higher government borrowing figures, but this would only be because the debt was visible rather than hidden. As well as allowing for renegotiation, refinancing with low cost government bonds would significantly reduce the PFI unitary charge and make deals more affordable. The OBR estimate that the "the total capital liability" of PFI deals stands at around £40 billion.[180] If government refinanced this debt at a lower rate of interest then, for every percentage point the interest rate reduced, annual savings of £400 million would be realised for taxpayers.

98. The most straightforward way of dealing with current PFI contracts is for the government to buy up the debt (and possibly also the equity) once the construction stage is over. This would result in an increase in the headline level of government debt but it would not increase the structural deficit or prejudice the fiscal mandate as this debt would score as government borrowing for investment in the National Accounts. Interest rates on the financing of the deals would fall significantly, releasing savings. Although government debt levels would be higher the public finances would not be any less sustainable. This is because it would become more affordable to service the visible government debt rather than the hidden PFI debt. Every one percentage point reduction in the interest rate paid on the estimated £40 billion of PFI debt would realise annual savings of £400 million.

99. In most cases, PFI projects involve significant and long-term financial commitments for the public authorities involved. Given the fiscal challenges faced by government, and the degree to which public expenditure is currently being scrutinised for potential savings, there is pressure on public managers to secure a better deal from existing PFIs. A high-profile review of public sector efficiency recommended that the government audit all procurement contracts with a concession value of over £100 million, and explore ways of breaking contracts where these represent poor value for money. [181] At the time of writing, the Treasury was consulting with major DBFO investors on plans to introduce a code of conduct[182] on reducing the costs of existing deals, and individual public authorities are being encouraged to work with investors to identify where and how reductions in their charges might be achieved.

100. However, it should be recognised that there is a limit to the savings that can be achieved on existing contracts. Any attempt unilaterally to reduce PFI payments could have negative side effects. Mr Friend noted the potential for:

the hazard that it creates in terms of UK reputation [...] I think these things need to be borne in mind as we play the larger game, which is: how do we finance and fund the nation's infrastructure needs over the next decade?[183]

Mr Wardlaw also noted that:

I think there is a really important issue here about the perception of political and other risk around the UK and the UK's infrastructure, because those investors, those contractors, those utility companies outside the UK who are making these decisions have alternative places to put their capital.[184]

101. Mr Friend did however consider that it might be worthwhile looking at the contracts on a case-by-case basis:

What I think is more feasible is a vigorous, taskforce-based approach that would require the Infrastructure UKs, the Local Partnerships of this world and the local authority bodies, on a case-by-case basis to work through: what is the potential for varying scope? What is the potential for increasing productivity? What is the potential for taking back risks that were transferred at a price, like insurance risk, the energy risk I referred to earlier, and literally cutting a deal case-by-case? My perception is that there would be a willingness in the private sector on a case-by-case basis.[185]

Mr Rabin agreed telling us "I think that renegotiation is always feasible and, potentially in some cases, desirable." He added "I do very strongly believe that that should be at a local level between the buyer of services and the provider of services rather than at an omnibus global level."[186]

102. Where the return to an investor is significantly higher than the level projected at the time of contract signature, there may be an opportunity for such gains to be shared. If equities are sold by 'primary' to 'secondary' investors after the risky construction period has been completed, this can give rise to a capital gain, the post-tax value of which will under current arrangements accrue entirely to the primary investor. Similarly, it has become evident that the price paid for maintenance (which is agreed when the contract is signed) has often become unrelated to the actual cost for the provider[187], which can also generate significant additional investor cash-flow. Currently, the price paid for maintenance is locked in once the long term contract is signed and is therefore not subject to competitive pressures over its life.

103. We explored the issue surrounding sharing the capital gains on the sale of equity stakes. Mr Friend expressed a concern:

I think the problem with renegotiating at the equity level, as Richard has referred to, is that I think the current Treasury estimate is that 55%—perhaps slightly more—of the original equity has moved on, and it has been brought across a wide variety of institutions now, and that is the problem.[188]

Canmore Partnership did not believe that "calls to share investors' gains on disposing of their equity interests [...] is reasonable or practical" as:

these procurements are presumably already deemed to represent better VfM than alternative procurement models (ie Full Business Case approvals will have had to show this to be the case) and so profits on disposals are surely part of investors' reasonable "upside".[189]

104. Where the actual return on private capital is much higher than that projected return it is possible that the government could use its purchasing power to negotiate gain-sharing arrangements without eroding its credibility. There is, indeed, a precedent for this. In 2002, the Office for Government Commerce and several major investors in the PFI programme signed a code of practice[190] which committed the latter to share gains made via refinancing their debt (which significantly accelerates cash-flow and increases investor returns), even though contracts did not stipulate any such sharing.

105. We welcome the work that the Treasury is doing with the PFI industry on drawing up a code of conduct. We believe that it is in the interest of the PFI industry to cooperate as fully as possible with the government in this regard. In 2002 the government reached a voluntary agreement with industry to share refinancing gains with the taxpayer. Therefore in principle there is no reason why a non-obligatory gain-sharing arrangement could not also be considered in relation to the gains on the sale of equity stakes.

106. As well as the numerous PFI deals which have already been committed to there are also many deals already in procurement and others are in the pipeline. If the public sector wants to reduce costs it must ensure that the prices and profit margins charged by the private partner are at the market level at the start of the contract and also that efficiencies can benefit the taxpayer over the life of the contract. These issues are considered in more detail in Box 3.

Box 3: Returns on PFI investment

Analysis by Mark Hellowell - Specialist Adviser to the Committee
There are currently 61 projects in procurement, representing projected capital expenditure of £7 billion. In this context, it is essential that the government ensures that the ongoing revenue costs to the public sector of new PFI projects are minimised. This can be done in two ways: (a) ensuring that the prices and profit margins charged by the private partner are at the market level (and the market should be defined more broadly than PFI alone), and (b) that the likelihood of productivity gains throughout the contract period is recognised in contracts, so that there is a mechanism for sharing such gains. In respect of (a), it is important to consider both operational and financial costs. In terms of operational costs, it is clear that the costs of construction, maintenance and service provision must be benchmarked against best practice in the market more generally. As the National Audit Office has pointed out, this will require the compilation and active use of much better data than has been utilised by departments hitherto[191].
Treasury officials have made the observation that the cost of primary equity was too high as far back as November 2005.[192] However we have seen no evidence that equity rates of return have come down since that date. The cost of equity quoted by major investors in terms of constructing discount rates for valuing their portfolios of PFI projects is typically in the range of 7-9%.[193] The rates of return targeted by primary investors are around double those normal on the secondary market, and the risks borne by equity investors during the construction and early operational stage of contracts do not justify this. There is a case to be made for action to be taken to ensure that Equity Internal Rates of Return cluster around their efficient level, which should be close to the cost of equity quoted by PFI investors, and much closer to secondary market discount rates.
In respect of (b), a private company in charge of services over a 30 year contract is likely to find opportunities to reduce costs over this period. Of the services included within the PFI structure, only "soft" facilities management services, such as catering and cleaning, are benchmarked or market tested during the contract period. Currently, there is no mechanism under which the gains from such efficiencies in maintenance can be shared with the public sector-and thus the gains accrue to equity-holders in their entirety. Although maintenance services are subject to competitive tension in the tendering process, the standard PFI structure does not allow sharing from any efficiencies in building maintenance which contractors achieve over the contract's life.
This is because these services are not value tested and contractors do not share with public authorities information on their maintenance spend. This is both undesirable in its own terms, but is also likely to lead to opportunistic behaviour. Currently, because "soft" facilities management services are benchmarked/ market tested, there is an incentive for a bidder (working within the context of a strict public sector budget constraint) to under-price this element of the services at the point of financial close, and over-price the hard facilities management services (i.e. build in a profit margin above the market level). When, in subsequent years, the price of the soft facilities management services are benchmarked, this will lead to the price going up. The public sector will, in this event, pay a market price for the soft services and an above-market price for the hard facilities management, and it will pay this above-market price for the entirety of the contract period.
One option would be to examine the potential for broadening the benchmarking/market testing process to include all services. However this is complex. For example, there would be a need to consider how the price paid for hard facilities management payments interact with the costs of life-cycle replacement. In practice, this may not be possible. A simpler, and likely more effective, method would be to ensure that any free cash-flow (i.e. cash in excess of that required to pay operational or debt costs) in excess of that required to provide equity investors with the rate of return projected at financial close is shared with the public sector. This would ensure that the private sector retains an incentive to invest in productivity gains in maintenance (as under PFI now) but that the benefits from this are shared with the public sector (as is not currently the case).
It may also be useful to examine the experience of the "hub" private finance model in Scotland, in which similar principles are in operation.

107. We recommend that HM Treasury collates and compares data to ensure that it gets a good price on any deals already being negotiated. It should benchmark operational costs of PFI projects with market prices outside PFI. It should also compare the equity returns of investors with other investments with a similar risk profile. It should publish as much of this information as is commercially possible. Far more transparency is required. The Treasury should consider whether this should extend to publishing data and costings on existing contracts, where commercially possible, in addition to what is already published. The Treasury should also consider introducing a mechanism for deals in procurement to ensure that any productivity gains are shared with the taxpayer over the life of the contract. Based on the analysis presented in this Report, we ask the Government to give further consideration before proceeding with the procurement in its present form of the Royal Liverpool and Broadgreen Hospital in particular.

Improving procurement and project management skills

108. As part of our inquiry we received evidence about the importance of improving procurement and project management skills in the public sector. The CBI pointed out that:

For complex procurements to be successful it is essential that project teams have the appropriate skills and experience and are adequately supported by central bodies with strategic oversight.[194]

Mr Friend told us that of the many reports written about PFI projects "I reckon a good two-thirds of them refer to the need to invest seriously in commercial skills in the public sector, and I do not believe that we have done that consistently".[195] Dr Chris Lonsdale from the University of Birmingham noted in his submission that "Risk transfer under the PFI, and in public sector procurement generally, has been further affected by limited public sector commercial skills." He noted that there had seemingly been a lack of understanding in the higher levels of the civil service:

In terms of commercial skills and capabilities, the UK public sector has spent the 20-year life of the PFI attempting to create the necessary capacity. For many years, there seemed to be too little appreciation in the higher civil service ranks of the extent of the difficulties of complex procurements.[196]

109. TfL's submission pointed out that "TfL invests heavily to ensure that it has the right skills to manage its risks". It considered that there was a case for other organisations to be "supported by others that have the resources and experience, rather than having to buy the experience more expensively from the private sector".[197] Skanska agreed that there should be a "focus on development of public sector in-house skills"; this they believed would encourage "the public sector to take ownership of projects rather than relying on external advisers".[198] The NAO reported in 2007 that the "average cost of external advice for all projects was just over £3 million per project or approximately 2.6 per cent of the capital value of the projects".[199] We asked Mr Abadie how much PwC had received as financial advisors on PFI procurements. He explained that for a school they would "probably get £250,000 to £400,000"[200] and for a hospital it "may be £500,000 to £800,000"[201]. Mr Abadie also told us that PwC often had people on secondment, including himself, to the government.

110. The head of PFI policy at HM Treasury has often, like Mr Abadie (a partner at PwC), come from a banking or accountancy background rather than having design or construction expertise. Many have also been on secondment. Geoffrey Spence who preceded Mr Abadie as head of PFI at the Treasury was seconded from Deutsche Bank in 2001. Mr Abadie was succeeded by Charles Lloyd in 2009—another secondee from PwC. The current outgoing senior official at HM Treasury responsible for PFI is Andy Rose who also has a background in finance. PFI is a DBFO (design, build, finance and operate) procurement method. The expertise of those responsible for PFI policy at the highest levels in government has been primarily on the financing element of the project.

111. The need to improve procurement and project management skills in the public sector is something that all are agreed on. In some ways PFI may have exacerbated problems in this area. Rather than focussing on improving procurement methods and project management, public sector clients' attention has been diverted to financing arrangements and the other requirements unique to PFI. Owing to the complexity of PFI, the public sector has become too reliant on expensive external advisers. We are also concerned that PFI may have resulted in the balance of expertise within the centre of government being tilted too heavily towards financial skills with less input from those with experience in design and construction.

Other ideas

INFRASTRUCTURE ACCOUNTS

112. The need for a national set of accounts was raised by some of our witnesses. Dieter Helm told us "we have no national balance sheet to set against our assets and liabilities"[202] adding that this meant "we are not interested in the question of what level of investment we should carry out to set against so we can set assets against liabilities".[203] He explained in his written submission in further detail the benefits of this:

A national balance sheet would enable rational decisions to be made about borrowing and investing, and hence allow the low public cost of debt to be translated into lower costs of capital for infrastructure projects. The absence of proper balance sheet accounts therefore has a real deadweight welfare cost: the higher cost of capital on highly capital-intensive projects. The private returns on PFIs reflect this deadweight loss to society.[204]

He explained that without national accounts "I have no idea what our financial deficit in this country is, because I don't know whether we have just been eating up assets and depleting our infrastructure or not."[205]

113. Other evidence we received agreed that a set of national accounts would be an positive step. A joint submission from KPMG LLP, John Laing PLC and Lloyds Banking Group said:

We believe government should invest in systematically collecting and analysing evidence on the comparative performance of all procurement approaches.[...]

The single biggest step in this regard would be the institution of a set of national infrastructure accounts, which would show both asset investment and asset depreciation. Such accounts would provide a starting point for inquiries such as this to get under the skin of the infrastructure challenge and to compare like with like, and would force the public sector and its partners to think long-term.[206]

We asked Professor Helm how easily a set of national asset accounts might be created. He said that if you want a "perfect set of accounts, then it is really difficult". In his opinion it was therefore "best to just get on with it". He added "Let's have a look at what has happened to the oil. Let's see what electricity networks looks like. Let's have a look at the water networks. Do it pragmatically." This he explained would allow government to get a "handle on the big items pretty quickly and we can tell whether we are depreciating rapidly or not. So, the answer to that is it is not difficult."[207] In a recent Committee hearing on the Bank of England inflation report The Governor emphasised the importance of understanding both assets and liabilities. He explained that the sustainability of the public finances should "be judged in the context not just of future liabilities, whether it be pensions, PFI projects or any other kind of liability, but also assets."[208]

114. On 13 July 2011 the first summary of the Whole of Government Accounts (WGA) was published. This is an unaudited summary report for the year ended 31 March 2010. The audited version of the accounts will be published later in the year. The WGA is a consolidation of the financial accounts of about 1,500 public bodies and therefore does not include a value of all of the infrastructure of the country.

115. While there is an understandable focus on the current high levels of government debt, the government and the citizens of the country have no proper understanding of the assets which accompany these liabilities—there is no national balance sheet. The audited Whole of Government Accounts will be published for the first time later this year. This will provide further understanding of public sector organisations assets' for financial reporting purposes.

INFRASTRUCTURE FUND

116. One idea proposed to us to lower the cost of capital for projects was an infrastructure fund or bank. Professor Helm gave a hypothetical example of how such an fund could be beneficial:

Supposing the Government wants there to be 10 nuclear power stations in this country at £5 billion each. That is £50 billion over the next 10-15 years. Supposing it borrowed a fund called the Nuclear Bond Fund, and it borrowed £50 billion and it just asked the builders of those stations to bid for that money. So, it is the Government borrowing but the private sector is doing the CAPEX and the OPEX thereafter.

He explained that such a scenario would lower the cost of capital significantly:

It [the Government] would currently borrow probably [...] [at a] negative real interest rate. The private sector [borrowing cost] for a nuclear power station may be 10% or 15% real. It doesn't take first year undergraduate maths to work out there is a colossal difference between those numbers [...][209]

117. A 2009 paper to which both Dieter Helm and James Wardlaw contributed recommended that "The UK should establish an infrastructure bank (UKIB)". In the paper Mr Wardlaw gave some detail about the benefits that this could bring:

The prize is an institution which facilitates the introduction of private sector capital without crowding it out, finances itself with a government guarantee, aims to break even with any dividends reinvested, and whose liabilities do not score in the National Accounts but whose activities are defined by national priorities.[210]

Another witness, Richard Abadie of PwC, agreed telling us: "I am supportive of an infrastructure bank as well".[211] Healthcare Audit Consultants explained that "Just as the European Investment Bank helped finance the new Barts' and the London Hospital so a UK infrastructure bank could make funding easier for needed capital expenditure within the public sector".[212]

118. We explored the idea with Mr Wardlaw in more detail. He considered that the UK had "suffered" without such a bank:

Well, I think to some extent we have suffered without. Many other countries in Europe have benefited from having a national infrastructure bank or a state development bank: KfW, Eco in Spain, and CDC in France. I think that it has been an important part of the armoury of tools to enable infrastructure to be constructed in a public sector context.[213]

However he considered that part of the solution was already being planned. He told us: "I think we are sort of getting it. It is called the Green Investment Bank."[214] Dieter Helm however pointed out a difference between the Green Investment Bank (GIB) and an infrastructure bank:

The GIB is essentially a project finance vehicle, and hence it needs capital injections and equity finance. The infrastructure bank would be a debt vehicle, and would conform to the old-fashioned idea that an investment bank is all about matching savers with investors [...][215]

RAB AND LABV

119. One potential avenue for lowering the cost of capital is to use a Regulatory Asset Base (RAB). In a RAB an asset earns a regulated return for the investor. We asked Dieter Helm about this:

You do your capital project [...] it is finished, and that is the refinancing point. That in the utilities is the point where the asset goes into a regulatory asset base and then earns just a marginal cost of debt, which is very much lower than the marginal cost of debt that is being done in these businesses [...][216]

He provided more detail in a supplementary submission for the Committee about how the RAB model could apply to PFI.

There are a number of ways in which the PFI framework could be brought into a RAB-based model. The optimal approach would be to create an infrastructure "bank". The bank's role would be to match savings (in practice largely pension and life funds) with investments in infrastructure projects such as those currently included in the PFI contracts. The bank would "buy" completed projects, put a RAB-wrapper around them, and then sell them onto the pension and life funds.

He explained that this would "would capture the returns from the assumption of the financing requirement, and therefore limit abnormal profits to the construction phase". If however there was not an infrastructure bank he said that another approach would work.

The obvious starting point is to separate out the PFIs into three separate contractual parts: the construction phase, the operating phase and the financing phase after project completion [...]

For the financing phase (strictly the refinancing phase after project completion), there could be a separate contract, with an associated cost of capital and a repayment profile. This could be subject to a guarantee that the revenues will in fact be forthcoming. [...]

The advantage of the separation out of the contracts is that it provides a focus and opportunity to zoom in on refinancing on project completion. If the private sector demands a high return on the completed asset, then one of two possibilities arises: either the government can clarify the commitment to remunerate the capital; or the government itself could buy-in the completed asset at a lower cost of capital (or some part of it).

Once separated out, the capital cost can be accounted for in the same way as the utility RABs - for that is in effect what they have become.[217]

120. Some evidence also mentioned the possibility of using Local Asset Backed Vehicles (LABVs) to allow local authorities to use their assets to attract long-term investment from the private sector. Skanska explained that the LABVs "are dependent upon the public sector having appropriate land or other assets to transfer into the joint venture vehicle and upon the market value of the land/asset that is available".[218] KBR International Government & Defence believed that LABVs provided "attractive option for outsourcing of non core activities of government departments". They suggested this form of additional financing would:

Encourage a business management perspective where contracting authority and contractor work together to a common goal rather than creating an adversarial contracting environment. [...]
Avoid the lengthy procurement periods and start up costs of PFI models
Create a business relationship that can cope with a changing and dynamic environment.[219]

Equility Capital suggested an alternative form of financing which would be done "by employing lease-based funding programmes". They explained that the debt "will appear on the balance sheet but, as the funding is lease based, so will the asset." Also "at the end of the lease period the asset reverts in its entirety to the borrower". As the cost of capital would be much closer to the cost of government gilts they calculated the change would release significant savings.[220]

121. The Treasury should consult on the possibility of using other financing models, including the Regulatory Asset Base (RAB) and Local Asset Backed Vehicles (LABV), as a way of financing capital projects in competition or in preference to PFI.

143 National Audit Office, Private Finance Projects: A Paper for the Lords Economic Affairs Committee, October 2009, p48, para 4.10 Back

144 David Heald and Alasdair McLeod, Constitutional Law, The Laws of Scotland: Stair Memorial Encyclopaedia: Public Expenditure, 2002, para 502. Back

145 David Heald and Alasdair McLeod, Constitutional Law, The Laws of Scotland: Stair Memorial Encyclopaedia: Public Expenditure, 2002, para 502. Back

146 HC Deb, 12 November 1992, col 998 Back

147 HM Treasury, Private Finance: Overview of progress, News release 118/94, 8 November 1994 Back

148 HM Treasury, Private Opportunity, Public Benefit Progressing the Private Finance Initiative, November 1995 Back

149 HM Treasury, Meeting the Investment Challenge, July 2003, p35, 3.30 Back

150 Ev w42 Back

151 Ev w33 Back

152 Ev 37 Back

153 Q 108 Back

154 Ev w19 Back