Select Committee on Health Appendices to the Minutes of Evidence


Memorandum by Professor David Mayston (PS 32)



  Capital resource management in the UK National Health Service has a long history of tensions and problematic development. The introduction of the Private Finance Initiative (PFI) in recent years has offered managers and ministers the opportunity to escape from many of the capital constraints of the traditional governmental funding system. However, it has done so at a potentially high future opportunity cost. The extent and nature of the risks which are transferred from the NHS to the private contractors by the PFI schemes are unlikely to justify a higher cost of capital for private finance compared to that available to the government. Instead PFI schemes may introduce potential new risks of their own. The macroeconomic argument that public investment crowds out private investment is likely to apply even more strongly to PFI schemes, which compete more directly for scarce private finance funds. More efficient ways of raising finance for direct investment in the NHS could be devised that satisfy also a strong institutional need for low risk investments to protect future pension incomes. Greater transparency and value for money could be obtained through unbundling PFI schemes, and their associated contracts, into their constituent elements.


  The management of capital resources in the NHS has for many years raised interesting and important issues, and latent tensions [1-2]. This tradition has been continued in no small measure by the introduction into the UK public sector of the Private Finance Initiative (PFI) by a Conservative Chancellor of the Exchequer in 1992, and its continuation under the subsequent New Labour Government after the partial repackaging of PFI in the form of Public-Private Partnerships (PPP). Central to the initiative is the use of private finance to fund capital investment in hospitals and other capital assets, such as medical equipment, for which the NHS becomes in effect the lessee over a long contract period of 30 years or more, typically as part of a package that may include maintenance and other operating costs being borne by the private contractor, in return for an overall annual payment. The ownership of the assets remains with the private contractor during the contract period, though may revert to the NHS at the end of the contract if this reversion is included and paid for within the PFI contract.

  Partly because of the complexities which this new form of financing and contracting raised, the growth of PFI schemes in the NHS was initially at a low level. However, PFI projects in the NHS for which contracts have now been signed or which have reached preferred bidder status are expected to involve some £1.86 billion in cumulative capital spending by PFI contractors by the end of the financial year 2002-03 [3-6]. The capital spending element of many of these contracts will, however, represent a substantial under-estimate of the present value of the total costs which the NHS is likely to pay under as a result of entering into these contracts. Whilst figures are now available for the total annual spending commitments for existing individual PFI schemes [7], separate figures are not in general available for the different service elements, including future maintenance and operating costs, that these contracts involve, separated out from each other and from the cost of capital.

  The issues which the growth of PFI raises are multi-dimensional. The first dimension which we will examine is that of time, and the associated issues of inter-temporal efficiency and equity.


  One of the primary immediate benefits of PFI for the Government, NHS management, and patients is the upfront benefit of new capital facilities becoming available to the NHS under PFI sooner than they would have done under the traditional NHS system of funding capital investment directly out of public capital expenditure. With access to PFI funding, the Secretary of State for Health, Alan Milburn, has been able to announce another "29 new hospital schemes with a capital value of £3.1 billion" [8] shortly before the 2001 General Election, with obvious electoral appeal. This forms part of an overall NHS capital investment programme totalling £12.9 billion between April 2000 and March 2004 [9, p.31]. Even allowing for the fact that direct capital expenditure in the NHS was cut substantially over the period 1994-95 to 1997-98 as PFI funding was increased, paying for new hospitals and medical equipment through annual lease payments will typically enable several times more new assets to be available to the NHS in the early years, out of any given level of total NHS funding inclusive of upfront capital expenditure, than would direct capital expenditure by the NHS on these assets.

  However, it is important to appreciate the full force of the logic of compound interest in this context, particularly as the cost of capital that is implicit in many PFI deals is widely believed to be significantly in excess of the Government's own borrowing rate of interest. The annual "availability fees" for making PFI hospitals available to the NHS have been estimated by Gaffney et al [10] to range from 11.6 per cent to 18.5 per cent of the initial construction costs of the facilities, including an element of maintenance charges that makes the underlying cost of capital not directly specified. The cost of capital, r, that is implicit in a series of constant annual payments, Y, for the capital facilities, rather than for any service and maintenance elements, is given by:


where n is the length of the contract in years, Ko is the capital value that is make available now, and Arn is the relevant annuity factor [11].

  From (1), we can calculate the total capital value that can be made available now for a given stream of annual payments, Y, for different values of the annual costs of capital, r. One stream of annual values of Y that it is of interest is the case where Y is set equal to the construction cost of a new hospital, which we will initially assume to be the same for all new hospitals. This enables us to consider the trade-offs and opportunity costs that are involved in making total capital of Ko available now to the NHS through PFI schemes, rather than spending the associated annual payments on the direct construction costs of a whole series of one additional new hospital a year for the next n years.

  The second column of Table 1 considers the case of n = 30. When the cost of capital is three per cent, 19.6 new hospitals could be financed now by PFI out of the same stream of annual payments that would have paid for the direct construction costs of a whole series of one new hospital a year for the next 30 years. When the cost of capital rises to 15 per cent, only 6.57 new hospitals could be financed now out of the same stream of annual payments that would have paid for the direct construction costs of a whole series of one new hospital a year for the next 30 years. The third column of Table 1 considers the case of n=60. When the cost of capital is three per cent, 27.68 new hospitals could be financed now by PFI out of the same stream of annual payments that would have paid for the direct construction costs of a whole series of one new hospital a year for the next 60 years. However, when the cost of capital rises to 15 per cent, only 6.67 new hospitals could be financed now by PFI out of the same stream of annual payments that would have paid for the direct construction costs of a whole series of one new hospital a year for the next 60 years.

  The rate of time preference, and associated willingness to give up the prospect of 30 additional new hospitals for the NHS over the next 30 years in return for substantially less now, may then differ across the community. Current politicians and NHS managers, under immediate pressure for more capital facilities to meet current needs, may take a shorter term view than the longer term interests of many sections of the population at large whose health care needs may extend over future decades. The extent to which this is true is compounded by both demographic and technological change. The major increases in the size of the UK population over the age of 75 in the decades after 2010, from 3.4 million in 2001 to nearly 7.6 million in 2035 [12], are likely to place significantly increased demands on NHS funding. Even if the needs of this expanding age group are better met outside of hospitals, the foregone freedom to switch funds out of future potential hospital building programmes into other forms of health care implies similar long-term opportunity costs to current PFI schemes.


Cost of Capital
Number of Hospitals now 30 year
PFI contract60 year
PFI contract
3 per cent
6 per cent
10 per cent
15 per cent
20 per cent

  The existence of technological change would also lead one to expect that a more steady process of investment in new hospitals, that are able to incorporate recent technical developments as they become available, would result over time in a more efficient and effective capital stock in coming years. This will be particularly the case if there is a need for a long-term re-configuration in the pattern of acute hospital provision in the NHS [13], and there is a risk of PFI schemes being too hastily pursued by their sponsoring Trusts with an inadequate consideration of the long-term needs of the NHS [14]. The establishment of the Capital Prioritisation Advisory Group (CPAG) [9] in the NHS may help to ensure a greater degree of rationality in the choice of current PFI schemes, although will not change the basic inter-temporal trade-offs that a high PFI cost of capital implies.

  A fully rational system of investment planning over the current and the next n years would be one that seeks to maximise some valuation function V subject to an inter-temporal budget constraint, such as that given by

where I, is annual capital investment at time t, Ct is current expenditure (excluding PFI annual cost of capital payments) at time t, Ft is the total funding available to the NHS at time t from central government, net of any user charges, with F = (Fo, .., Fn) and a = (ao, .., an) is a vector of demographic and other relevant parameters, including the rate of technological change.

  Kt, is here the level of the effective capital stock available to the NHS at a time t, and is some function:

of the capital stock, Ko, at the start of period 0, and of subsequent investment levels and rates of technological change.

  In this context, PFI involves an initial level, Io, of capital investment in the NHS at time t = 0 plus current expenditure, Co, in excess of the direct funding for the NHS, Fo, from central government. Satisfying the inter-temporal budget constraint involved in (2) requires that total funding in at least some later period must then exceed the level of capital investment plus current expenditure in those later periods, to pay for the future PFI annual payments. If there is a tendency to defer the day of reckoning until later in the n year time horizon, this will imply that

ie increases in the cost of capital reduce the overall benefits that the NHS can achieve out of its future stream of funding.

  Assessing the relative health care needs of present and future NHS patients within the valuation function V in (2) raises issues of both inter-temporal efficiency and inter-generational equity, not only with respect to the levels of health care investment, It and current expenditures, Ct, in each future year, but also to how these should be funded over time through the time stream of tax-based governmental contributions Ft. Once PFI schemes and borrowing are permitted, Ft may diverge from the sum of It and Rt through the shifting of annual payments over time. The projected demographic changes over the next 35 years include not only an increased total population in older age groups, but also a declining tax base of younger age groups of working age [12], albeit with higher per capital real incomes, and consumption expectations, through economic growth. The time pattern of tax-based governmental contributions to the NHS is therefore unlikely to be a matter of indifference, but instead enter into the valuation function V in (2). It is then questionable whether an optimal policy of public finance does involve the accumulation of additional long-term obligations for annual PFI payments which must be met partly out of tax revenue in later years when these demographic trends become particularly acute. In contrast, there may be a strong case for the current generation of 40-60 year olds to pay more in tax in the next 10 years to fund direct NHS investment in a capital stock that will benefit them in future years, rather than shift forward these tax liabilities at a high rate of interest to a time when there will be greater pressures on the tax base, and when this generation may be less able to fund such additional liabilities.

  Nevertheless, even if present health care needs are given considerably more weight than future needs within the valuation function, Table 1 illustrates that significantly more new hospitals, or other capital facilities, could be made available now for the same annual cost of a 30 year PFI contract at a 15 per cent cost of capital, if the NHS had access to capital at a borrowing cost of six per cent or three per cent. Indeed such lower borrowing costs would enable, respectively, 2.1 and 3.0 times as many new hospitals of a given size to be made available now as the PFI contract does. Seeking to accommodate such an increased cost of capital within any given total annual level of funding may result in pressure to reduce the specification of the PFI schemes proposed. Reductions averaging between 26 per cent and 31 per cent have been found [15-16] in the beds to be made available by the first wave of PFI schemes compared to those currently available. These reductions are in addition to any resulting from the pressure that has already been exerted on NHS Trusts to reduce their bed stock through the internal NHS capital charging system (which includes a six per cent interest charge) since 1994, when the system became "non-neutral" in its incentives for such a reduction [2]. The planned reductions in bed numbers are also despite the findings of the National Beds Inquiry [17] that "the long term trend in reductions in general and acute beds is not compatible with the requirement to improve access to care and cannot be sustained" [17, p.42].


  A second dimension which we need to examine is that of risk and uncertainty. These will interact with our above discussion involving time through the adjustment which must be made to the relevant cost of capital, r, to include an allowance for any systematic risk which the investment projects involve. Such systematic risk arises from any correlation between the costs and/or demand for the project and general economy-wide uncertainties. It remains even after all project specific risks that are uncorrelated with such economy-wide uncertainties have been diversified away by being spread over the entire government sector [18], in the case of direct NHS investment, or over the capital market [11], in the case of PFI investment. Under the standard Capital Asset Pricing Model (CAPM) of finance theory, the existence of systematic risk implies a competitive cost of capital, and risk-adjusted discount factor, to be applied to the net cash flow of project j at time t of:

  If the project is run as a PFI scheme, a cost of capital above the risk-free rate is justified if the contract does transfer significant positive systematic risks to the PFI contractor. However, demand risk will remain with the NHS (see [19]) unless the PFI payments vary with the level of demand for the facility. Similarly if the PFI contract index-links payments to general construction and energy cost movements, the systematic risk will remain with the NHS.

  The net present value of the expected stream of payments, P;tt for t = 1...., n, for the jth PFI scheme, minus the expected construction costs and operating costs, using the risk-adjusted discount factors in (4), is given by:

  If there is competition in the supply of PFI contractors, we would expect the contract to involve a zero NPVj of excess profits and yield rates of return equal to the cost of capital. If the PFI contractors are bearing no significant systematic risks, this implies a competitive rate of return close to the risk-free rate of interest. An actual rate of return on the PFI contract, and implicit cost of capital charged to the NHS, in excess of this competitive rate would reflect a lack of strong competition in the bidding for PFI contracts. The relative effect of a higher cost of capital on the number of hospitals which could be made available now in return for a given future stream of constant annual payments is again illustrated by Table 1.

  The National Audit Office has stressed that "appropriate risk transfer is crucial to obtaining value for money in privately financed projects" [20]. However, the NAO does not distinguish between systematic and more specific risks which can reduced in total down towards zero through diversification. In stating that "the aim of a PFI procurement is to achieve an allocation of the individual risks to those best able to manage them" [21, p. 27], the NAO appears to move beyond the transfer of risk as the justification for the PFI route to one of managing a change also in the expected value of construction and other costs. If private sector contractors do have greater management skills than the public sector in controlling construction cost escalation, NHS options appraisals should include any such differences in expected values, net of management costs, in their evaluations of the Net Present Value (NPV) of the respective costs of publicly managed schemes compared to privately managed schemes, using the appropriate competitive cost of capital. A difference in the expected value of construction costs does not, however, justify the use of a higher long-term cost of capital for PFI schemes. If there is competition between private contractors, the gain from any generally higher sector management efficiency should be passed on to the NHS through lower PFI contract prices. Similarly, any superior ability of the private sector in general to reduce the variance of construction costs should be passed on to the NHS through competition in the PFI contract terms offered to the NHS, rather than through a higher cost of capital. If there remains some construction cost risk around a lower expected value, this would still only justify a higher cost of capital than the risk-free rate for the initial period in which such risk occurred, and only to the extent that such risk does involve systematic risk [11, p.408].

  Construction and design cost risks can themselves be addressed more directly through the alternative format of a fixed-price Design and Build contract. This can enable the risks of construction cost escalation, building completion delay, design faults and latent defects to be transferred from the NHS to private sector contractors in a potentially more efficient way than through a more complex long-term PFI contract, including scope for separate latent defects insurance [22] where appropriate. Such an approach would enable the different elements of the Design, Build, Finance and Operate (DBFO) requirements that make up a typical PFI contract (see eg [23]) to be unbundled and awarded to the most suitable contractors under the most suitable contract terms. This can include scope for the Facilities Management (FM) element of operating NHS capital assets to be contracted separately if this proves desirable because of genuinely lower operating costs, and for finance for the initial capital investment to be obtained from the most cost-effective source, independently of any bundling requirement.

  It should be noted the risk may not simply be transferred through the involvement of private contractors, but also increased. If contracts cannot fully determine the activities of the contractor, including the quality of the final product delivered, or fully anticipate changes in the environment, there is a risk of high agency costs, through increased monitoring and bonding costs, and high residual losses [24-27]. Problems in securing adequate hospital cleaning by private contractors, with potentially high costs [28], and problems in securing the adequate maintenance of railway track in the UK by private agencies suggest that tying the NHS into 30-year contracts with individual PFI contractors would not be a risk-free strategy. This may be particularly so if there is a financial incentive for the contractor to seek to increase the profitability of the contract by hiring inexperienced or unmotivated labour at the lowest possible cost. The inclusion within PFI contracts of explicit performance criteria may mitigate this problem to some degree. However, unanticipated changes, such as improved environmental and health and safety requirements, may lead to expensive disputes and costly litigation, as the experience of contractual disputes between EuroTunnel and its contractor Trans Manche Link, following stricter health and safety requirements for the Channel Tunnel after King's Cross Underground disaster, illustrates.

  In the case of health care, the need for future flexibility is underlined by both the multi-dimensional nature of demand across different forms of treatment and the technological uncertainty that currently exists over the nature and extent of future cost-effective forms of health care treatment that it might be desirable for the NHS to provide in future decades. As stressed in [19, 29], it is unclear that the NHS should want to assume additional risks from tying itself into 30-year contracts with specific PFI providers that may constrain the flexibility of the NHS to later respond efficiently to these future changes, and again risk expensive disputes and litigation if the PFI contract does not easily accommodate such future changes.

  The extent to which risk transfer is actually effective may also be reduced by the existence of a risk of insolvency of the PFI contractor, when it encounters higher construction costs, maintenance costs or operating costs, or lower revenue than it anticipated. In the case of Leeds Armouries, the risk of insolvency of the PFI contractor, when demand at the prices charged to the public under the PFI contract proved lower than anticipated, resulted in the sponsoring government department deciding to bail out the PFI contractor through additional funding for the costs, and losses, of much of the scheme [26]. The public sector may then be paying a risk premium for a risk which is not actually fully borne by the PFI contractor when the circumstances of the risk materialise. Despite the intended transfer of risk through a PFI contract, many of the costs which arose from delays in the implementation of a new computer system for the UK Passport Office fell on the public sector and general public, rather than the PFI contractor [30]. There is also some incentive for the Special Purpose Vehicles, that are often created to act as the private PFI contractors, to minimise their own exposure to risk by not leaving surplus fund in their financial structures. Brealey and Myers [31] emphasise the scope for risk-shifting games to be played in the private sector in order to shift unwanted risks on to unwary parties. This may result in a moral hazard risk that imposes additional agency costs on the NHS when it takes the PFI route, both in the ex ante monitoring of such risk and in bearing additional ex post insolvency costs if they are not adequately controlled.

  Given that some of the major PFI contractors have multi-million pound obligations not just to the NHS, but also to major other parts of the public sector and to privatised public utilities, such as Railtrack, fully assessing their insolvency risk may be a difficult task. The share price of at least one major PFI contractor has fluctuated dramatically in recent years. The need for public sector clients, such as the NHS, to ensure clarity within PFI deals over their liabilities and rights to take possession of key assets in the event of insolvency of the private contractor has recently been emphasised by the NAO [32].

  It should be noted that HM Treasury's [33] own prescribed discount rate of six per cent in real terms, for evaluating investment projects in the NHS and other parts of the public sector, appears to be based chiefly upon the approximate pre-tax rate of return to equity across the economy as a whole [33, p. 83]. This will include elements of risk to private sector equity, both from systematic risks and from the accentuating influence of debt leverage on equity risk, making six per cent above inflation is itself an excessive estimate of the relevant risk-free rate of interest. The sensitivity of the computed economic advantage of PFI schemes in the NHS to the choice of discount rate is demonstrated in [34].

  An increased cost of capital also brings with it an increased supply and affordability risk for the NHS, as a result of the reduced ability of the NHS to meet future PFI payments as well as many other future demands on its available funds. Such affordability risk may be compounded by a moral hazard risk if politicians face short-term pressures and local NHS managers seek to solve their immediate capital problems by agreeing to PFI schemes that impose onerous longer-term financial obligations on the NHS [29]. The willingness of PFI contractors to agree to such deals is increased by the introduction of the NHS (Residual Liabilities) Act 1996, under which the Secretary of State must take over the liabilities of any NHS Trust or Health Authority that becomes insolvent as a result of such obligations or is otherwise disbanded.


  The affordability and sustainability risk associated with the growth of PFI in the NHS is also complicated by the unclear nature of the future funding position of the NHS to cover the future annual payments that result from the capital costs of current PFI schemes beyond those which the NHS would otherwise incurred. There are a number of sources of such additional funding for PFI schemes. One is from reducing current expenditure on other health care activities in the NHS, which is a generally unattractive course of action. Another is from any operating cost savings that are generated by the PFI projects themselves, such as from the more efficient management of energy and maintenance costs. However, the magnitude of these savings is unclear, under a situation where both capital costs and operating costs are typically bundled together in an overall annual PFI payment. This makes difficult an external assessment of whether PFI schemes do offer value for money. It also makes difficult the prediction of the future public expenditure needs of the NHS compared to what they would have been under traditional ways of financing NHS capital expenditure.

  A third way is through viring monies which would have instead been used in future years to fund direct capital expenditure by the NHS into payments for current PFI schemes. It might be objected that this represents a form of "mortgaging the future" and of "eating the seed corn" of funds that would otherwise provide the new hospitals and capital equipment of the future. The traditional NHS separation between capital and revenue funds has been for the purpose of preventing such capital funds being raided to solve immediate current revenue expenditure pressures, a sentiment which would also seem in part to lie behind HM Treasury's [35] claimed desire for a "a clear distinction between current and capital spending". However, this separation appears to be breached by the new virement which HM Treasury itself appears to permit when it states [35, p. 26] that: "Departments will be able to channel funds earmarked for capital expenditure into current finance private-public partnerships". The willingness to divert capital funds to help finance PFI payments appears to be confirmed in several cases in the NHS [36].

  A fourth way in which current PFI schemes may be financed is through additional new money into the NHS over the life of the PFI assets. However, there is at present no indication or guarantee that this will be forthcoming. If it is not, there arises a potential risk of instability in NHS finances due to the circular nature of the present NHS capital charging system. This involves a positive feedback loop (see [37]) in which the total capital charges levied on the NHS's own capital stock are re-circulated to all NHS health authorities through the NHS's weighted capitation system [38] to form part of the "resource" income of individual health authorities. Under this system, each individual health authority, i, receives as its annual income, Mi, a proportion, ai<1, of the total income, M, that is circulated to health authorities, where ai reflects its age-weighted population and local need factors, and where:

M = R + c. KN with Mi = ai . M

  R is the extent of the total NHS non-capital Exchequer funding, KN is the value of the directly-owned NHS capital stock, and c is the total capital charging rate, of a six per cent real interest rate plus an annual rate of depreciation on the capital stock, where for the sake of simplicity we now omit time subscripts. The annual expenditure, Xi, of health authority i will equal:

Xi = REi + c.KNi + Pi
where REi is annual current expenditure, on items such as nurses' salaries, on health authority i's patients by the NHS Trusts with whom health authority i has contracts, KNi is the corresponding value of their own directly-owned capital stock, and Pi is their annual PFI payments. In order for the expenditure and available income to balance, we require:

Pi + c.KNi = ai . R—REi + ai . c.KN

  The ability of each individual NHS Trust and its health authority to finance any given level of annual PFI payments depends in part on the level of directly-owned capital stock, KN, which all NHS Trusts maintain. There is a risk here that individual NHS Trusts and health authorities may themselves over-estimate the affordability of their own PFI schemes because they may base their calculations (see eg [39]) upon the current levels of their income, Mi, and the associated current level of the directly-owned NHS capital stock KN. If PFI is used in place of direct NHS capital investment, the annual depreciation of NHS assets, together with any NHS land sales to partially finance PFI schemes, will cause the value of the directly-owned NHS capital stock, and hence (other things being equal) the incomes of individual health authorities, to decline.

  There is also a risk here of individual NHS Trusts free-riding on the capital charges income generated by other NHS Trusts. Holding constant the directly-owned capital stock used by the many other health authorities, the annual budget constraint (7) permits an almost a one-to-one switch from capital charges, c.KNi, on the directly owned assets used by health authority i into PFI payments, Pi, by the NHS Trusts with whom it has contracts. However, other things being equal, health authority i making this switch will cause the total NHS directly-owned capital stock, KN, and hence all other health authorities' incomes, to fall. PFI payments are here acting as a leakage from the circular flow of income in the NHS capital charging system. In a parallel way to the circular flow of income in macro-economics [40], there is a consequent potential multiplier effect of this leakage of PFI schemes and other expenditure items.

  Other health authorities will respond to their fall in income in general in part by reducing their own directly-owned capital stock and associated capital charges, through ward closures and other measures, and in part by cutting back on their own current expenditure and potential new PFI payments. The capital charges on the directly-owned NHS capital stock will then equal:

  where KoN is a baseline level of the directly owned capital stock that does not vary with the health authorities' incomes, and 0 is the marginal propensity of health authorities in total to vary their directly-owned capital stock as their total resource income varies. From the budget constraint (8), Q will equal unity if each health authority refuses to cut back on its own current expenditure and annual PFI charges in the face of such income variation, but will be less than one if it is willing to make such cutbacks. The value of Q will depend upon the annual net price c.(1— ai) each health authority i faces for additional capital stock in (8) under the now "non-neutral" system of NHS capital charges (see [2]).

  Inserting (9) into (8) yields the value of the multiplier effect of any reduction in the baseline value of the directly-owned NHS capital stock, due to a switch into PFI assets, to be:

  Recent years have indeed seen a decline in the value of the directly-owned NHS capital stock in real terms, after adjusting for annual depreciation and indexation, from £25.7 billion in March 1997 to £23.7 billion in March 1999 prices [41-43]. Fortunately, the next few years should see an increase, with direct capital investment, net of asset sales, under the new NHS Investment Strategy [9] projected to increase from £1.6 billion in 2000-01 to £2.6 billion in 2003-04.


  The sensitivity of the real cost of PFI in Table 1 to the cost of capital highlights the desirability of the NHS making use of cheaper sources of finance, whenever the additional risk premium associated with use of private finance is not justified by the risks which are actually transferred to the private sector. Even if some systematic risk remains with the public sector, NHS investment projects have a high capacity to utilise risk-free government borrowing as a source of finance. The recent reductions in direct public sector investment and government borrowing, that have accompanied the growth of PFI, have been associated with a substantial fall in long-term interest rates on government bonds/gilts to around 2.2 per cent for index-linked returns. These in turn have been associated with very low annuity rates now on offer to many thousands of pensioners currently retiring with "money purchase" pension funds to be converted into future annual income in their retirement. The result of such low annuity low rates will be substantially smaller future pensions for these pensioners.

  The changing demographic structure of the UK is itself associated with increasing numbers of individuals retiring with capital sums to invest in annuities. As the stock market looks a more uncertain prospect after the recent collapse of the long bull market and the Internet share bubble, current investors in pension funds may also increasingly seek safer havens for their money. The problems in recent years at both Maxwell Communications and the more highly respected Equitable Life Assurance Society in the UK illustrate the potential financial difficulties which can be encountered in the running of private pension schemes. Investors in the diversified "with profits" investment fund of Equitable Life have found themselves faced with unexpected undiversified risks when the future guaranteed annuity rates promised to investors by the Assurance Society before 1990 proved impossible to honour, illustrating the difficulty which even well-established private sector institutions have in offering risk-free rates of return.

  The mechanism which has been in place under the 1995 Pensions Act for ensuring the ability of private pension schemes to pay guaranteed levels of pension, once an annuity is purchased, has been the Minimum Funding Requirement (MFR), which encourages such schemes to invest a high proportion of their associated funds in government bonds. However, following the declining availability of government bonds, the Myners Report [44] has recommended the abolition of the MFR to permit other securities, such as corporate bonds, to replace pension fund holdings of government bonds, a recommendation which the UK Government has now accepted.

  The increased risk to private pension funds which may result from such a substitution of government bonds by corporate bonds is illustrated by the case of British Telecommunications (BT), whose debt level has increased dramatically in recent years to a peak of £30bn, following its multi-billion pound bids for third-generation mobile phone licences. The subsequent downgrading of the credit rating of BT's corporate bonds by both Moody's and Standard and Poor's, and the recent substantial decline in new mobile phone sales, underline the extent to which existing corporate bondholders are exposed to significant risks of loss of value as a result of both management and market behaviour. That the proposals to relax the MFR may have adverse effects on efficient risk-bearing is underlined in a recent report by a major firm of actuaries [45]. These effects include the discouragement of risk-taking by firms, an increased risk of insolvency of companies, the winding-up of company pension schemes, and the declining availability of defined benefit schemes for individuals investing in private pension funds. Financing direct NHS investment by issuing more government bonds would therefore have advantages not just for the NHS. Rather it would provide an efficient risk-free form of investment for pension funds, on whom increasing numbers of pensioners will be dependent, for their future ability to maintain a healthy standard of living into their old age.

  One major economic reason for limiting the size of direct public borrowing is that public borrowing crowds out private finance for investment in manufacturing industry and other parts of the private sector. However, the use of private finance to fund capital assets for the NHS via PFI itself directly crowds out private finance which might have otherwise been available for investment in manufacturing or elsewhere in the private sector. Moreover, such PFI finance competes directly for scarce risk-bearing capital funds that are likely to be more appropriate for financing private sector investment. In contrast, financing NHS direct investment through government-backed borrowing makes use of a different sector of the capital market, serving institutions requiring risk-free forms of investment, that is less suitable for financing many private sector investments. The substitution of direct borrowing by PFI finance therefore involves a double inefficiency in financial risk-bearing. It displaces private finance which might otherwise be used to finance genuinely private sector projects, whilst at the same time pushing pension funds that have a strong need for risk-free assets into riskier corporate bonds that are less suited to their financial needs.

  The use of public borrowing to finance investment in the NHS would be perfectly consistent with the Chancellor of the Exchequer's own Golden Rule [46], that "Over the economic cycle the Government will borrow only to invest and not to fund current spending". His second rule in the Code for Fiscal Stability [46], namely the Sustainable Investment Rule, states that "Borrowing to finance investment will be set so as to ensure that net public debt, as a proportion of Gross Domestic Product (GDP), will be held over the economic cycle at a stable and prudent level". UK general government net borrowing has now dropped from 2.0 per cent of GDP in 1997 to minus 2.1 per cent in 2000, whilst general government gross debt has dropped from 51.1 per cent of GDP in 1997 to 42.9 per cent in 2000 [47]. Even if Britain's entry into the Euro and Economic and Monetary Union within the EU is the dominant concern, these figures are well within the relevant Maastricht Treaty's Convergence Criterion for EMU entry, namely that total outstanding government debt must not exceed 60 percent of GDP, with general government debt not to exceed 3.0 per cent of GDP. The net borrowing figures are, moreover, expressed conservatively by not anticipating the receipt of £22.5 billion from the sale of third-generation mobile phone licences in the year of the sale [47]. Instead their receipt over the next 20 years will reduce the net borrowing figures well into the future. Overall, public sector net debt is expected to fall from 44 per cent of GDP in 1997 to under 30 per cent in 2002-03 [6]. The need to reduce public borrowing is therefore not a binding constraint in the context of the £500 million capital spending that is now involved in PFI schemes in the NHS.

  However, even if reducing formal government borrowing were a primary consideration, there are alternative mechanisms which could be devised to fund the expansion of direct NHS investment in place of PFI. One such mechanism might be the establishment of a supplementary pensions fund [48] in which prospective pensioners could invest and in which retired pensioners could deposit their accumulated capital sums in return for annuities. The available capital funds would be made available for direct NHS investment in new capital assets, in return for annual payments that represent a cost of capital lying between the current long-term index-linked yield of around 2 per cent on government bonds and the 6 per cent real cost of capital which HM Treasury currently impose as the cost of capital to the public sector. The need for adequate supplementary pensions provision is underlined by the falling adequacy of the basic state pension to meet the future retirement needs of many individuals, who otherwise risk imposing a large financial burden on future social security funding (see [49]). The value of the pensions coverage provided by the existing complex State Earnings Related Pensions Scheme (SERPS) has also declined in recent years (see [50]). A simpler supplementary pensions scheme might involve increasing current National Insurance contributions in return for increased future pension payments embodying a guaranteed rate of return on the increased payments, together with a right of individuals to opt out of these additional contributions and associated supplementary pension. Such a scheme would have similar characteristics to unfunded pension scheme contributions, even though the funds generated would be available to invest in NHS capital assets, and would not count as part of government net borrowing [51]. Investment in the NHS assets would be matched by a corresponding long-term future commitment, but a less onerous one than is involved in PFI schemes, whenever the latter involve a high cost of capital to the NHS.

  Providing an attractive risk-free form of long-term investment with reasonable rates of return through such a scheme may encourage increased saving for retirement, and add to the large potential market for the NHS to tap in order to finance its direct investment in capital facilities. The growth of PFI not only has the effect of causing the NHS to fail to tap this market in an efficient way. It also pushes the NHS towards raising finance from the private sector via an inherently more costly route, involving substantial additional costs in contracting, negotiation, consultancy and potential litigation costs if contract terms become difficult later to honour. Short-circuiting this route through financing direct NHS investment would eliminate many of these costs.


  HM Treasury's claimed motivation for the growth of PFI in the NHS and elsewhere is the pursuit of value for money [52]. However, an alternative explanation is the desire to satisfy the convergence/ Maastricht criteria [47] on public sector indebtedness, as a pre-condition for Britain's entry into Economic and Monetary Union (EMU) and the Euro zone. PFI offers the political attraction in this context that it can be used as a form of "off-balance-sheet finance" to reduce apparent public sector indebtedness. This is true under the Treasury's [52] revised guidelines on How to Account for PFI Transactions, provided that the different elements of the PFI scheme, such as capital construction and financing, cannot be separately identified within an overall service payment, and that "sufficient" risk is intended to be transferred to the private sector. As noted above, there must be doubt over whether these two features of PFI schemes help to promote value for money for the NHS. However, the role of accounting considerations in determining the details of PFI schemes is illustrated by the recommendations of HM Treasury's Taskforce on Private Finance [53, p 15] that:

    where the accounting analysis requires a PFI transaction to be treated in substance as borrowing . . . the public body should examine the scope for reworking the deal so that it is clearly for the provision of services.

  The use of off-balance-sheet finance, to avoid greater transparency in the long-term obligations that are being incurred, has in the past resulted in substantial financial difficulties for many private and public sector organisations [29]. The importance of achieving transparency in the presentation of public financial statements has itself been stressed by HM Treasury [35, 46] as "an indispensable hallmark of good policy" and to avoid "poor policies and bad decision-making" in its own pursuit of the goal of fiscal stability. The economic assessment of the suitability of the UK for entry into the Euro is of some national importance, with the experience of the UK's participation into the earlier Exchange Rate Mechanisms (ERM) illustrating the substantial risks which may arise if economic realities are ignored. The ability of PFI schemes to remain off-balance-sheet may itself prove to be short-lived if the UK Accounting Standards Board soon adopts suggested proposals [29, 54] for requiring the on-balance-sheet treatment of operating leases and similar arrangements that bundle together services, such as maintenance, with the availability of capital assets, in the same way that many PFI contacts do.

  A major consideration in whether or not a PFI scheme can be left off the balance sheet of the NHS Trust that makes use of the resultant hospital or other capital asset, both under HM Treasury's own guidelines [52] and under the ASB's regulations [55], is the extent of the risk transferred to the private PFI contractor. A strong desire to keep the assets and liabilities created by PFI schemes off-balance-sheet may well then lead to some risks being transferred to the private PFI contractor at an excessively high cost to the NHS. This is particularly the case as the risks which are relevant to the accounting regulations are not simply the systematic risks which finance theory [11, 31] suggests should be priced.

  The doubtful value for money that the desire to keep the arrangements off-balance-sheet may produce is compounded by a further factor. This is the significant weakening [56, p 13] that has occurred in recent years in the earlier Treasury-imposed Ryrie rules that privately financed schemes pass a value for money test against a public sector comparator. Without a readily available alternative source of public funding for direct NHS investment, such a test becomes unrealistic, with pressure on NHS managers in recent years to pursue PFI even at high future cost as the only option that was likely to be available to them to secure new capital facilities [10, 29].

  The traditional form of financing the NHS is reflected in the physical condition of many existing NHS sites, as an assortment of often poorly maintained buildings, with many smaller extensions and outbuildings when small amounts of capital became available [1]. Whilst PFI schemes provide some opportunity to rationalise current facilities and to inject more capital investment into NHS sites, its potential problems still reflect the underlying financial management framework of the NHS. This has fortunately now moved beyond the dominant annuality and cash accounting principles of the past that generated many inefficiencies in managing NHS capital resources [1]. The new Comprehensive Spending Review and Resource Budgeting [57] financial framework, within which the Department of Health now operates, involves a three-year planning and budgeting horizon. It also involves a separate Departmental Capital Budget for direct capital expenditure which may, or may not, be set an optimal level consistent with the overall optimal use of resources in (2). PFI has the attraction to current managers and politicians of having its main impact outside the three year budgets for which they have immediate concern, and avoiding the constraint on direct investment that is imposed by the Departmental Capital Budget.

  Failing to record adequately the assets and liabilities to which PFI schemes give rise on the public sector balance sheet can encourage the passing on of hidden liabilities to future generations of taxpayers. Under an inter-generational "Ponzi game" [58, p 97], the present generation gain an immediate benefit, here from securing new hospitals, but pass most of the cost on to future generations of taxpayers, at a cost which increases over time if they all play the same game. This danger will be mitigated to some extent under PFI by the fact that its costs are spread out over future generations who will also benefit in some degree from the PFI schemes, and who will be under some financial pressure not to allow additional PFI payments to escalate out of control within their own tax-paying lives. However, the precise extent to which future generations will benefit from current PFI schemes depends upon key details of the schemes, such as the implicit cost of capital, the future availability of beds and their flexibility to adapt to new medical needs and technology at reasonable cost.


  A higher cost of capital for the NHS under PFI schemes is unlikely to be justified by the extent and nature of the risk transfer that the schemes involve. Instead, securing capital via the PFI route is itself likely to impose significant new risks upon the NHS. More efficient ways of raising finance for investment in the NHS could be devised that would benefit also increasing numbers of pensioners in maintaining a healthy standard of living. Unbundling PFI schemes, and their associated contracts, into their constituent elements of design, build, finance, maintain and operate, could provide much greater transparency of the extent of the value for money obtained in each direction. Such transparency is necessary to overcome the suspicion that the PFI is driven mainly by short-term accounting considerations aimed at paving the way for British entry into the Euro, or by other political factors that are extraneous to the long-term needs of the NHS. Unbundling PFI schemes would also provide the opportunity to purchase the constituent elements from the most efficient sources, with a much closer association between risk and reward than PFI schemes at present provide. Thus if it is genuinely the case that private contractors can provide more efficient design, build and facilities management services for the NHS than the public sector can, these can be identified and priced separately. Such a separation would not imply a need to forego lower cost sources of public finance and would avoid the need for potentially inflexible long-term contracts that the NHS may prove to be very costly for the NHS.


  1.  Mayston DJ. Managing capital resources in the NHS. In Competition in Health Care: Reforming the NHS, Culyer AJ, Maynard A and Posnett J (eds). Macmillan: London, 1990.

  2.  Mayston DJ. Capital and labour markets for health. In Reforming Health Care Systems, Culyer A and Wagstaff A (eds). Edward Elgar: Cheltenham, 1996.

  3.  Treasury HM. New Ambitions for Britain: Financial Statement and Budget Report March 1998, HC620. TSO: London, 1998.

  4.  Treasury HM. Budget 1999: Financial Statement and Budget Report. TSO: London, 1999.

  5.  Treasury HM. Budget 2000: Financial Statement and Budget Report. TSO: London, 2000.

  6.  Treasury HM. Budget 2001: Financial Statement and Budget Report. TSO: London, 2001.

  7.  House of Commons Health Select Committee. Public Expenditure on Health and Personal Social Services 2000, HC882, Table 4.8.4. TSO: London, 2000.

  8.  Timmins T. New hospital schemes worth £3bn unveiled. Financial Times 2001, 16 Feb.

  9.  Department of Health. Departmental Investment Strategy. DoH: London, 2000.

  10.  Gaffney D, Pollock A, Price D, et al. NHS capital expenditure and the private finance initiative—expansion or contraction? BMJ 1999; 319: 48-51.

  11.  Copeland TE, Weston F. Financial Theory and Corporate Policy (3rd edn). Addison Wesley: Reading, Mass., 1998.

  12.  Office of National Statistics. National Population Projections. TSO: London, 1998.

  13.  Royal College of Surgeons. Provision of Acute General Hospital Services. RCS: London, 1998.

  14.  Meara R. Building Services and Servicing Buildings—The Private Finance Initiative in the NHS. Institute of Health Service Management, London, 1997.

  15.  Pollock A, Dunnigan M, Gaffney D, et al. What happens when the private sector plans hospital services for the NHS: three case studies under the private finance initiative. BMJ 1997; 314: 1266-1271.

  16.  Pollock, A, Dunningan M, Gaffney D, et al. Planning the "new" NHS: downsizing for the 21st century. BMJ 1999; 319: 179-184.

  17.  Department of Health. Shaping the future NHS: long term planning for hospitals and related services; consultation document on the findings of the National Beds Inquiry. DoH: London, 2000.

  18.  Arrow K, Lind R. Uncertainty and the elevation of public investment decisions. American Economic Review 1970; 60: 364-378.

  19.  Dawson D. The private finance initiative: a private finance illusion? Health Economics 2001;10:479-486.

  20.  National Audit Office. Examining the Value for Money of Deals under the Private Finance Initiative, HC 739. TSO: London, 1999.

  21.  National Audit Office, The Skye Bridge—Report by the Controller and Auditor General, HC 5, Session 1997-98. TSO, London, 1997.

  22.  Latham M. Constructing the Team: Final Report of the Government/Industry Review of Procurement and Contractual Arrangements in the UK Construction Industry. HMSO: London, 1994.

  23.  National Audit Office. The Private Finance Initiative: The First Four Design, Build, Finance and Operate Roads Contracts, HC476, Session 1997-98. TSO: London, 1998.

  24.  Jensen M, Meckling W. Theory of the firm: managerial behaviour, agency costs and capital structure. J Fin Econ 1976; 305-360.

  25.  Mayston DJ. Principals, agents and the economics of accountability in the new public sector. Accounting, Auditing and Accountability Journal 1993; 6: 68-96.

  26.  Chalkley M, Malcomson J. Contracts and competition in the NHS. In Reforming Health Care Systems, Culyer A, Wagstaff (eds). Edward Elgar: Cheltenham, 1996.

  27.  Hart O, Schleifer A, Vishny R. The proper scope of government: theory and application to prisons. Quarterly Journal of Economics 1997; 112: 1127-1161.

  28.  National Audit Office. The Management and Control of Hospital Acquired Infection in Acute NHS Trusts in England. HC230 Session 1999-2000. TSO, London, 2000.

  29.  Mayston DJ. The private finance initiative in the National Health Service: an unhealthy development in new public management? Financial Accountability and Management 1999; 15: 249-274.

  30.  National Audit Office. The Passport Delays of Summer 1999, HC 812, TSO: London, 1999.

  31.  Brealey R, Myers S. Principles of Corporate Finance (6th edn). McGraw-Hill: New York, 2000.

  32.  National Audit Office. The re-negotiation of the PFI-type deal for the Royal Armouries Museum in Leeds. HC 103 Session 2000-01. TSO: London, 2001.

  33.  Treasury HM. Appraisal and Evaluation in Central Government. TSO: London, 1997.

  34.  Gaffney D, Pollock A, Price D, et al. PFI in the NHS—is there an economic case? BMJ 1999; 319, 116-199.

  35.  Treasury HM. Stability and Investment for the Long Term: Economic and Fiscal Strategy Report 1998, Cm 3978. TSO, London, 1998.

  36.  Gaffney D, Pollock A. Can the NHS Afford the Private Finance Initiative? BMA: London, 1997.

  37.  Mayston DJ. Resource budgeting, capital charging and the private finance initiative. In Resource Accounting and Budgeting, Second Report of the House of Commons Procedure Committee, HC 438, Session 1997-98. TSO, London, 1998, 65-71.

  38.  NHS Executive. Resource Allocation: Weighted Capitation Formula. NHSE: Leeds, 1999.

  39.  Hodges R. Learning to live with the private finance initiative: case studies of three ambulance services, mimeo. University of Nottingham: Nottingham, 1998.

  40.  Dornbusch R, Fischer S. Macroeconomics, 3rd edn. McGraw-Hill: London, 1984.

  41.  Department of Health. The Government's Expenditure Plans 1998-99, Departmental Report, Cm 3912. TSO: London, 1998.

  42.  Department of Health. The Government's Expenditure Plans 1999-2000, Departmental Report, Cm 4203. TSO: London, 1999.

  43.  Department of Health. The Government's Expenditure Plans 2000-01, Departmental Report, Cm 4603. TSO: London, 2000.

  44.  Myners P. Institutional Investment in the United Kingdom: A Review. HM Treasury: London, 2001.

  45.  Mercer WM. Security for Occupational Pensions. William M Mercer: London, 2001.

  46.  Treasury HM. A Code for Fiscal Stability. HM Treasury: London, 1997.

  47.  Office of National Statistics. Government Deficit and Debt under the Maastricht Treaty. ONS: London, 2001.

  48.  Mayston DJ. NHS capital charging, PFI contracts and value for money in health care. Paper presented to Health Economics Study Group. University of York, York, 1997.

  49.  Hutton S. Current and future incomes for older people. Ageing and Society 1996; 16: 775-787.

  50.  Dilnot A, Disley R, Johnson P, et al. Pensions Policy in the UK—An Economic Analysis. Institute for Fiscal Studies: London, 1994.

  51.  Golland J, Savage D, Pike T, et al. Monthly Statistics on Public Sector Finances—A Methodological Guide. Government Statistical Service: London, 1999.

  52.  Treasury HM. Private Finance Technical Note No.1 (Revised): How to Account for PFI Transactions. HM Treasury: London, 1999.

  53.  Treasury Taskforce on Private Finance. Partnerships for Prosperity. HM Treasury: London, 1997.

  54.  Accounting Standards Board. Leases: Implementation of a New Approach. ASB: London, 1999.

  55.  Accounting Standards Board. Amendments to FRS5 "Reporting the Substance of Transactions": Private Finance Initiative and Similar Contracts—September 1998. ASB: London, 1998.

  56.  Heald D. Privately financed capital in public services. Manchester Sch 1997; 65: 568-598.

  57.  Treasury HM. Managing Resources—Full Implementation of Resource Accounting and Budgeting. HM Treasury: London, 2001.

  58.  Kotlikoff L. Generational Accounting. Basic Books: New York, 1992.

previous page contents next page

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

© Parliamentary copyright 2002
Prepared 15 May 2002