Value for money in UK healthcare public–private partnerships: A fragility perspective

Abstract

Healthcare private finance initiatives/public–private partnerships (PFI/PPPs) have favored the provision of health infrastructures and services, but they have been subjected to major criticisms since the value for money assessment of private finance initiatives/public–private partnerships does not consider uncertainties. Using a systematic literature review and content analysis, we identify the sources of uncertainty (i.e. fragilities) associated with UK healthcare private finance initiatives/public–private partnerships. Fragilities are rooted in their financial structure, inadequate ring-fencing of risks and contractual inflexibility. We then discuss such fragilities in light of the evolution of the private finance initiatives/public–private partnership policy in the UK, thus considering private finance 2 and non-profit distributing models. Although much has been done to overcome the shortfalls related to the financial structure of private finance initiatives/public–private partnerships, greater attention has to be paid to the adaptability of the infrastructure and to the risk governance of private finance initiatives/public–private partnerships. Based on these results we then identify new avenues of research considering the potential benefits stemming from diversification of the sources of funding, adaptability of both the contractual requirements and the physical infrastructure and the risk governance.

Keywords

Fragility healthcare public–private partnership value for money risk governance

Introduction

Healthcare service delivery is increasingly undertaken through private finance initiatives/public–private partnerships (PFI/PPPs) characterized by a long-term contract between private and public partners for the provision of health infrastructures and related services (see e.g. ACCA, 2012; Grimsey and Lewis, 2007). Healthcare PFI/PPPs constitute an alternative to conventional public procurement and they are signed only in cases they are assessed more value for money (VfM) than public provision (HM Treasury, 2006).

VfM options appraisal is undertaken at a pre-contract stage and it is not predictive of VfM in the long-term (Broadbent et al., 2003; Broadbent et al., 2008; Froud, 2003). For example, despite initial positive VfM assessments, some PFI/PPPs hospitals turned out to be vulnerable to the effects of the 2007–2009 credit crunch and subsequent financial, economic and political crises. This crisis has required the intervention of the public sector to bear the downside risks of PFI/PPPs, thus reducing their VfM (NAO, 2009, 2011, 2012a, 2012b). As such, the recognition and active management of uncertainties are required to safeguard VfM (Froud, 2003). Although the literature has evidenced that the sources of uncertainty of healthcare PFI/PPPs can be recognized at a pre-contract stage (Broadbent et al., 2008), a comprehensive systematization of the sources of uncertainty is still missing, and little is known on how to manage them.

The aim of this paper is to provide a comprehensive systematization of the sources of uncertainty (also referred to as fragilities) that have been so far associated to healthcare PFI/PPPs. The focus on the fragilities of healthcare PFI/PPPs is not meant to deny the potential merits of PFI/PPPs; rather, it is conducive to the provision of an overarching systematization of the shortfalls of healthcare PFI/PPPs that can be used to advance research and practice in the field. We then bridge academic literature with practice by discussing the so identified fragilities in light of the evolution of the PFI/PPP policy in the UK thus opening new avenues of research.

In the next section, we discuss the study context and the research questions. We then present the research method based on a systematic review and quantitative text analysis of the academic literature on UK healthcare PFI/PPPs. The research findings shows three main factors associated to fragility. We then discuss the research findings in light of the PFI/PPP policy evolution. We conclude advancing novel avenues of research to better safeguard VfM in the long-term.

VfM in healthcare PFI/PPPs

VfM assessment is a prerequisite for establishing healthcare PFI/PPPs and it is based on an option appraisal procedure at the pre-contract stage comparing PFI/PPP with conventional public procurement. The advantage is that VfM is assessed through a well-established methodology that can potentially increase the transparency of procurement procedures (Barretta and Ruggiero, 2008). Disadvantages have been associated with a questionable estimation of the parameters of analysis that tends to be biased in favor of PFI/PPPs (Grout, 2005; Shaoul, 2005), optimism biases leading to overestimation of benefits of PFI/PPPs (Flyvbjerg et al., 2002) and a questionable allocation of risks between the private and public partners (Grout, 2005; Sawyer, 2005). Overall, there is no consensus on VfM of PFI/PPPs (Andon, 2012), and the diverse research methodologies used in PFI/PPP evaluations make it difficult to compare VfM outcomes (Hodge and Greve, 2009). Furthermore, even if VfM options appraisal is required at a pre-contract stage, it does not consider the uncertainties that can affect VfM in the long-term (Broadbent et al., 2003; Froud, 2003). For safeguarding VfM it is thus necessary both to recognize and to manage uncertainty. Broadbent et al. (2008) evidenced that the sources of uncertainty (i.e. fragilities) can be recognized in healthcare PFI/PPPs but a comprehensive systematization is still missing.

To fill this gap, we use content analysis techniques to investigate the fragilities that have been so far associated to healthcare PFI/PPPs in the academic literature. Although most of the analysed studies do not aim at empirically testing causality between the fragilities and a reduction of VfM, they provide useful insights on the factors that have been associated with it. As such, sound content analysis techniques can help to provide a systematization of fragilities that can then be used to open new avenues of research.

Research method

To identify the fragilities associated with healthcare PFI/PPPs, we undertook a systematic review of academic literature and thereby clustered fragilities into three main factors using a fully automated text analysis (CATA) procedure. Our review is focused on healthcare PFI/PPPs developed in the UK, this being the first country to implement PFI/PPPs and the preferred ground for academic analysis (Marsilio et al., 2011; Roehrich et al., 2014; Torchia et al., 2015). Furthermore, healthcare PFI/PPPs account for 24% of the overall capital value of UK PFI/PPPs (Table 1) and they have been subjected to major criticisms, mainly related to their affordability and reduction of VfM in the long-term (see e.g. NAO, 2012b). Finally, the UK PFI/PPPs policy has evolved over time into new models, including the non-profit distributing (NPD) model in Scotland and private finance 2 (PF2) in England, thus making the UK's PFI/PPPs environment suitable for the identification of fragilities.

Table 1.

UK health PFI/PPP contracts signed at March 2013.

	Number of contracts	Capital value (£m)
Health PFI/PPPs
England	103	11,566
Scotland	27	1230
Northern Ireland	7	338
Wales	7	132
Total health PFI/PPPs	144	13,267
Total PFI/PPPs	717	54,211

PFI/PPPs: private finance initiatives/public–private partnerships. Total health PFI/PPPs: number and value of PFI/PPPs in the healthcare sector. Total PFI/PPPs: number and value of all PFI/PPPs.

Note: Own elaboration on Current Projects as at 31 March 2013 (Private Finance Initiative Projects: 2013; last accessed August, 2014) https://www.gov.uk/government/publications/private-finance-initiative-projects-2013-summary-data

Consistent with Torchia et al. (2015) and Roehrich et al. (2014), we performed a systematic review of academic contributions about UK healthcare PFI/PPPs based on a replicable and transparent process built to minimize search biases (Tranfield et al., 2003). Using ISI Web of Science, we expanded the search period from 1990 to 2014. The search resulted in 194 articles. We eliminated duplicates (59) and then read the abstracts. After having deleted abstracts not related to the topic, the final set of articles is composed of 21 papers (Tables 2 and 3).¹

Table 2.

Search settings

	ISI
Combinations of keywords	Public–private partnership (OR PPP) and health (OR hospital; OR National Healthcare System; OR NHS) and United Kingdom (OR UK); Private Finance Initiative (OR PFI) and health (OR hospital; OR National Health System; OR NHS) and United Kingdom (OR UK). Project Finance and health (OR hospital; OR National Health System; OR NHS) and United Kingdom (OR UK)
Search field	Topic
Search domain	Social sciences
Search period	1990 (October) 2014

Table 3.

Search results

	ISI
Total articles	194
Articles after deletion of duplications	59
Articles selected reading abstracts	21

The corpus for the content analysis comprises the text of the 21 papers, excluding references.² Content analysis is undertaken through a fully automated CATA. This approach is best suited for our analysis, since it is used for exploratory research and it is not subjected to coder biases because it does not require an a priori identification of codes (Bara et al., 2007).

Using a CATA software (Iramuteq), we analysed the co-occurrences of words. Co-occurrence analysis assumes that the meaning of sentences can be captured when words co-occur frequently in the corpus (Schonhardt-Bailey, 2013). We then undertook a class and factorial analysis. The method for class analysis is a descending hierarchical classification that identifies classes basing on the maximum separation of word co-occurrences (χ² is a measure of the relationship existing between words). The strength of the association between the word and its class is defined by the χ² of the word. Words with the highest χ² are the basic source for the interpretation of results. Finally, the factorial analysis merges those classes that, beyond the co-occurrence of words, share a similar content (Azevedo et al., 2013).

Findings

The study of terms co-occurrence evidences that the locus of academic debate has been the risks and costs related to PFI/PPP hospital projects (Figure 1). The relevance of these topics in the academic debate is further grounded by the class and factorial analysis performed through the quantitative text analysis.

Figure 1.

Co-occurrence of terms.

Specifically, CATA software identified four different classes in the academic debate. Class 1 comprises 20.7% of the analysed text, Class 2 comprises 23.2%, Class 3 comprises 25.1% and Class 4 comprises 30.9% (Figure 2). The factorial analysis resulted in the identification of three factors since Classes 2 and 3 were merged together because they share a similar content, while Classes 1 and 4 represent different factors (Figure 2).

Figure 2.

Class and factorial analysis.

In the following sections, we present each factor and some representative sentences comprising the key words with the highest χ² (identified in bold). Following Bara et al. (2007), by means of key words it is possible to place an interpretation on each class and factor.

Financial structure

The first factor, corresponding to the first class, is associated with the financial characteristics of PFI/PPP hospital arrangements. The sentences with the highest χ² contain the key terms equity, return and investors. As an example:

… policymakers can eliminate the problem of excess returns on the (increasingly prominent) equity element of a project's capital investment. (Vecchi et al., 2013: 269)

The sentences grouped in this factor relate to the financial structure of healthcare PFI/PPPs and the return for equity holders. UK PFI/PPPs are established through a legally independent company, namely Special Purpose Vehicle (SPV), through which the project is funded and managed. Highly leveraged SPVs have delivered high returns on equity even after adjusting for the higher financing costs (Engel et al., 2010), and it is claimed that PFI/PPPs' equity investors' profits have been much greater than can be justified by the risks they have actually borne (HM Treasury, 2012b; Vecchi and Hellowell, 2012; Vecchi et al., 2013), thus potentially increasing the costs for the public partner (Hellowell and Pollock, 2010). In this perspective, in 2008 the additional cost of private finance relative to public finance for 12 PFI hospital trusts in the UK was higher than expected at financial close (Shaoul et al., 2008). Furthermore, although long-term and binding PFI/PPP contracts are necessary to secure the repayment of debt over time and to provide an adequate return for the private partner, the private sector's returns are often guaranteed by the public sector and bankruptcy costs for the private partners are minimal when the public sector bails out the project (Bailey et al., 2009).

This is consistent with a transaction-cost approach where a high amount of debt in project finance operations subject to deep asset-specificity (as it is the case of PFI/PPPs) is associated to higher costs of governance and higher costs of funding when risks increase for both debt holders and shareholders (Williamson, 1988). The higher costs of debt governance (if compared to equity) in highly asset-specific transactions is rooted in the fact that ‘ debt is unforgiving if things go poorly’ (Williamson, 1988: 580) since it limits managerial discretion. PFI/PPP are related to high transaction costs associated with the simultaneous occurrence of asset specificity, the private partner's incentive to act opportunistically to extract more profits and bounded rationality on the part of the public partner (Lonsdale, 2005). This simultaneity allows the PFI/PPP's private partner to take advantage of the highly specific healthcare assets and commercial confidentiality can create bounded rationality on the part of the public partner so that the private partner can earn the profits (Toms et al., 2011). It is worth noting that the higher costs of debt governance for these transactions holds independently of whether debt is provided by the public or private sectors.

Finally, governmental and non-governmental reports highlight that UK PFI/PPPs have usually been overdependent on bank loans of shorter duration than the contract, meaning that debt has to be re-financed from time to time during the contract period. However, the 2007–2009 credit crunch caused a liquidity crisis in PFI/PPPs (KPMG, 2010). Subsequent to the credit crunch, Basel III capital requirements forced banks to lower their exposures to higher risk assets and so banks have become more inclined to refinance existing assets in regulated utilities rather than new capital expenditures, especially those outside government-regulated utility frameworks (Wagenvoort et al., 2010). Consequently, the refinancing of existing assets became very cheap, while funding of new infrastructure did not because of the bank's decreasing risk appetite for buying syndicated debt (Helm, 2010). Banks have not only become more risk averse, asking for higher risk premiums but also more short-term oriented (Burger et al., 2009; Hellowell, 2013; Regan et al., 2011), this being problematic for the UK's highly leveraged PFI/PPPs subject to the considerable power of banks and private sector funding (Aldred, 2008; Asenova and Beck, 2010; Toms et al., 2011).

Ring-fence of risks

The second factor identified by the content analysis comprises both Class 2 and Class 3, meaning that those classes refer to similar content. The sentences with the highest χ² refer to the risk arrangements of health PFI/PPP contracts. A representative quotation is:

However, the narrow framework used to consider risk transfer sidelines the consideration of both retained risks and the additional risks generated through the complex contracting arrangements…. (Aldred, 2008: 20)

The SPV isolates the risks and their impacts from the parent companies (normally a bank, construction company and facilities management company) through a ring-fence mechanism. Based on this mechanism, risks can be fully identified and managed within the PFI/PPP contract through their efficient allocation between the public and private sectors.

Nonetheless, not all risks can be explicitly considered and assessed within PFI/PPP contracts and even some of the anticipated risks are not fully managed within the contract (Asenova et al., 2007). It is impossible to anticipate all risks and the lack of an adequate expertise can hinder the ability to manage even the anticipated risks (Cuthbert and Cuthbert, 2010). Even if risks are best allocated between the private and public sectors, the default risk cannot be hedged by the private sector and, therefore, the public sector is ultimately responsible for bailing out the project in cases of default (Bailey et al., 2009).

The public partner takes the ultimate responsibility for the service and is perceived by the general public to be fully accountable for it, regardless of the selected procurement option. This need to minimize reputational risk means that the public sector partner is ultimately liable for a PFI/PPP hospital that fails to provide contractually agreed services to patients. In the event of failure, the public partner has to bear default risk, the cost of which is likely to put pressure on health authority budgets, thus increasing the fragility of PFI/PPP procurement option.

Another issue related to the risk ring-fence is that the SPV usually subcontracts service components to firms that are not included in the SPV arrangements (NAO, 2009; Roehrich and Caldwell, 2012). By subcontracting, the SPV passes risks through a network of firms outside the PFI/PPP contract, thus changing the balance of risk among partners (Akintoye et al., 2003; Shaoul, 2011).

Whether subcontracting increases or reduces fragility depends on the particular arrangements (Stulz, 2009). For example, fragility is reduced if subcontractors can actually bear those risks or, otherwise, if subcontracting reduces the high transaction costs of PFI/PPPs associated with their highly specific assets (i.e. hospital and medical infrastructures), reduces bounded rationality (unknown future service scenarios exacerbated by commercial confidentiality of contracts) and reduces scope for opportunism (e.g. reducing quality of health services to cut costs and thereby increase profits). Fragility is increased in the opposite case if subcontractors cannot bear the risks and transaction costs are not reduced or even increased (e.g. by increased scope for opportunism by multiple contactors because of insufficient coordination of ‘unbundled’ services). In effect, the passing of risks to multiple stakeholders outside the contract challenges the claimed benefits of bundling infrastructure and service provision into a single PFI/PPP contract since, de facto, the SPV subsequently unbundles risks outside the contract.

As such, the benefits of PFI/PPPs compared with conventional public procurement depend on the relative importance of subcontracting as a proportion of the PFI/PPP's contract value: the greater that proportion, the less is the net benefit of the PFI/PPP. This is because unbundled risks are no longer managed by the SPV itself. The risk ring-fence is too low and too weak to ensure that the private sector partner controls all risks for which it is contractually responsible. As such, the SPV is reliant upon other external parties to bear risks but subcontractors may be unable to manage and/or bear them. If so, failure of risk ring-fence is a cause of fragility for VfM.

Contractual requirements

The third factor, corresponding to the fourth class, is associated with the inflexibility of health PFI/PPP arrangements. The quotations with the highest χ² contain the key terms design, clinical and hospital. As an example:

… a key question is therefore whether the PFI model is structurally capable of stimulating innovative design solutions to reduce the risk of potential future obsolescence arising from fast changing care delivery and technology lifecycles. (Barlow and Koberle-Gaiser, 2008: 1394)

PPPs are established through long-term and transaction-specific contracts that bundle the infrastructure building and the subsequent service provision. Specific investments require parties to exchange with each other (i.e. lock-in). Asymmetric lock-in can create scope for opportunism, i.e. ‘hold-up’ (Williamson, 1979) that is facilitated where there are few alternative parties with whom to contract and there is a lack of transparency (i.e. knowledge is asymmetrically bounded). In PFI/PPPs, the public sector can be asymmetrically locked-in to the private sector provider and is in no position to enforce the risk transfer arrangements (Lonsdale, 2005). Furthermore, inflexible contracts can lock-in the public sector to complement the SPV's revenues and, therefore, decrease the public sector's ability to respond to financial and social changes (Edwards and Shaoul, 2003). Indeed, public resources cannot be diverted from PFI/PPPs to other uses over the whole duration of the contract since the public sector is legally committed to make unitary charge payments to the SPV irrespective of changing service requirements.

A substantial part of the healthcare PFI/PPP unitary charge relates to payment for service availability, the health authority having to ‘supply’ sufficient numbers of patients who can be treated in that facility over several decades if it is to achieve the expected VfM. This problem is not peculiar to healthcare, both demand risk and reconfiguration risk also being evident for other public services such as school PFI/PPPs (Accounts Commission, 2002). However, compared with schools, reconfiguration risk is greater for hospital PFI/PPPs because the treatment of patients is moving away from long-stay to short-stay hospital inpatient episodes and, for some operations, even outpatient treatment. However, contractual lock-in reduces the ability of PFI/PPPs’ agreements to cope with the new healthcare requirements thus creating strategic risks for the health authority and operational risks for the SPV. Such risks were not considered at the signing of the contract because the VfM option appraisal does not consider how the PFI/PPP fits within the service as a whole: it only considers the relative costs of provision by the PFI/PPP versus the public sector. Its VfM appraisal is therefore blinkered and this can cause operational fragility for the PFI/PPP and strategic fragility for the health authority that commissioned the PFI/PPP. In this regard, it is argued that PFI/PPPs do not provide a sufficiently flexible design to address potential future changes in demand for healthcare and in the healthcare models to meet that demand (Barlow and Koberle-Gaiser, 2008; McKee et al., 2006).

Additionally, there may be a lack of a strategic oversight whereby PFI/PPPs contracts are signed without considering their ramifications for the rest of the public health care sector, risking overcapacity through duplication of services. This was found to be the case for the UK's Peterborough PFI hospital that was declared unaffordable by the National Audit Office (NAO, 2012a). Built near a non-PFI/PPP hospital, the PFI/PPP hospital contract was approved following the project's positive VfM test without considering whether, in future, the new PFI/PPP capacity could create instability in the existing service system.

Finally, PFI/PPP contracts will effectively commandeer a significant proportion of the public budgets to the repayment of PFI/PPP liabilities over the next decades without considering the changing needs of the healthcare system (see e.g. Hellowell and Pollock, 2010). As an example, there is a long-term shift to the care-in-the community model for management of increasingly common chronic health conditions. These long-term ailments are associated with ageing populations' frailty and increased incidence of such conditions as type 2 diabetes amongst young adults. The contractual rigidity is particularly problematic because, unlike the SPV, the public sector cannot pass the default risk to other parties.

Overall, the public sector partner has increased the system fragility by bundling the infrastructure and related service over extended contract periods during which time its payment obligations are irrevocable irrespective of changing circumstances. This lack of flexibility decreases the ability of the public sector to deal with financial, medical, demographic and social changes. Indeed, 22 hospital trusts in England reported in 2012 that repayments were causing difficulties for the sustainability of their clinical and financial future, one subsequently going into administration because of the associated debts. This risk is magnified the greater the proportion of health budgets consumed by hospital PFI/PPPs, causing diversion of limited public funds not only from non-PFI/PPP hospitals but perhaps also from primary healthcare services and from rural areas without PFI/PPP hospitals to cities with them (Marriott, 2014).

Discussion

The content analysis allowed us to systematize the analysed academic literature in terms of fragility of the first model of healthcare PFI/PPPs. After two decades, the first wave of UK PFI/PPP contracts has evolved into alternative models meant to address the shortfalls of the initial form of PFI/PPPs (Table 4).

Table 4.

Alternative PFI/PPP models

Model	Differences with the traditional PFI/PPP model	Expected improvements
PF2	Introduction of public sector equity Introduction of funding competitions for a portion of the private sector equity before financial close Removal of soft services where contractors can include a risk premium in the pricing Deleveraging of the financial structure Increase access to capital market investors Standardization of contracts Disclosure and control of off-balance liabilities and equity returns	Decrease the overall cost of project Increase flexibility Increase transparency
NPD	Enhanced stakeholder involvement in the management of projects No dividend bearing equity Capped private sector returns Increase the leverage of the financial structure	Control of shareholders' returns and reinvestment of surpluses in the public sector

PFI/PPPs: private finance initiatives/public–private partnerships; NPD: non-profit distributing.

The PF2 model developed in England prescribes a lower use of debt relative to equity. On the one hand, increasing equity is meant to decrease the overdependence on bank funding. On the other hand, the participation of the public sector in equity investments is meant to bring a greater alignment of private and public incentives and to enhance transparency (HM Treasury, 2012a, 2012b). Furthermore, PF2 aims at decreasing the overdependence on bank funding through the attraction of new long-term investors.

In contrast, the Scottish NPD model prescribes an even higher debt to equity ratio than the standard PFI/PPP model to reduce the proportion of equity, to cap profits for shareholders to a predefined quota and so provide a better deal for the public sector by capping the private sector's profit (Scottish Future Trust, 2011). It was argued above that high financial leverage increases fragility, but this may be mitigated by NPD targeting institutional investors (e.g. pension and insurance funds) as a complementary source of funding – as does the PF2 model.

In this section, we discuss our findings in light of the current development of the UK PFI/PPP policy by highlighting how the previously identified fragilities have (or have not) been addressed by these newer PFI/PPP models and thus proposing new avenues of research.

Diversification of the sources of funding and financial structure

Both PF2 and NPD models differ from standard PFI/PPP mainly in terms of diversification of the sources of funding and financial structure. The major change between PF2 and NPD is in their financial structure that is, nonetheless, heavily dependent on debt funding (especially for NPD) mainly provided by banks (Table 5). In this regard, some authors argued that both PF2 and NPD might be only a cosmetic change to PFI/PPP schemes, questioning the real reduction of the overall cost of the project (Hellowell, 2013; Hellowell and Pollock, 2009; UNISON, 2011).

Table 5.

Financial structure of different models

	PFI/PPP	PF2	NPD
Financial structure	High leverage; limited diversification of the sources of debt funding (i.e. dominance of bank loans).	Lower leverage with public equity investments; Limited diversification of the sources of debt funding (i.e. dominance of bank loans).	Higher leverage; Limited diversification of the sources of debt funding (i.e. dominance of bank loans).

PFI/PPPs: private finance initiatives/public–private partnerships; NPD: non-profit distributing.

The necessity of diversifying the sources of funding is consistent across policies. As shown in the literature review, overdependence on an almost unique source of funding (specifically bank funding) has been considered a fragility of the standard PFI/PPP model. Diversification requires instead the involvement of a greater number and wider range of investors in both credit and capital markets as well as recourse to a wider range of financial instruments. In standard PFI/PPPs, the increase of the number of investors in the credit market has been limited to establishment of a consortium of banks for new PFI/PPPs with both pre-financial and post-financial close syndication (Bailey et al., 2009; Demirag et al., 2010). It is instructive that, typically, the originator bank involved in funding a new PFI/PPP now seeks to reduce its exposure to financial risk by creating a consortium of banks that jointly provide debt-funding to the SPV. Diversifying means increasing the number of investors including, for example, institutional investors such as pension, insurance and sovereign wealth funds that have an investment time horizon that matches the infrastructure timescale. Such investors have also considerable amounts of capital to be invested in ownership of a PFI/PPP and remunerated in the long-term (EPEC, 2010).

Although diversification is considered a way to reduce fragilities in both the novel models of PFI/PPP, its feasibility is still under investigation. Institutional investors require safe investments that may not fit the PFI/PPP market and, furthermore, the funding and solvency regulations can deter institutional investors from PFI/PPP investments (Inderst, 2009).

Diversification of funding does not necessarily change the financial structure of PFI/PPPs. In this regard, PF2 policy prescribes a decrease of the SPV's leverage (i.e. deleveraging) to limit the dependence on debt funding which, being a contractual liability, requires the binding repayment of debt liabilities irrespective of the flow of revenues. Deleveraging is therefore required to prevent both financial distress and default. First, in cases of financial distress (i.e. the SPV faces financial stress, but has not yet defaulted), overdependence on debt can decrease the SPV's flexibility by limiting managerial discretion (Williamson, 1988). Indeed, cash flows cannot be diverted from the repayment of debt to other uses since debt amortization and interest payment have to be paid even if the necessary financial resources are scarce. Second, a lower leverage decreases the impact of potential losses. Indeed, equity losses are capped to the investment, while debt losses are at least equal to the investment that represents only the minimum loss. Third, in cases of default on interest payments and amortization of debt, a bail-in (rather than a bail-out) strategy reduces the potential losses to the public sector. Debt-to-equity swaps are a form of bail-in that transform debt liabilities into equity liabilities and thereby cap the losses up to the investment. Nonetheless, preliminary studies on the effects of deleveraging on the cost of capital suggest that decreasing the amount of debt can lead to higher cost of capital, at least in the short term (Hellowell, 2013).

As already noted, NPD policy prescribes an even higher leverage and caps the potential profits for equity holders to a predefined quota. The rationale is that a limited dependence on equity funding deals better with the excess profits that have been often attributed to equity investors (Scottish Future Trust, 2011). Although profits can be constrained within politically acceptable limits, the effects of an extensive use of debt need to be considered. As discussed above, higher leverage means less flexibility of the sources of funding since debt has to be repaid irrespective of whether the PFI/PPP's revenues are sufficient or not, whereas equity does not. Furthermore, as noted before, in cases of default the public sector will ultimately bear the costs. On the other hand, capping equity returns could actually discourage new equity investors to participate in these kinds of projects thus undermining the rationale for diversification.

Adaptability of the infrastructure and risk governance

In terms of flexibility of the contract requirements, the novel models propose to reduce some of the services included in the contract (e.g. soft services) that were usually subcontracted to other firms by the private partners. If it is the case, the public partner can be less reliant on the prices negotiated between the private partners and their subcontractors (HM Treasury, 2012a). Nonetheless, it is also necessary to consider that those services will be still contracted-out by the public sector itself. Flexibility can be increased if the public sector establishes shorter term contracts for those services. Nonetheless, as noted above, fragility is reduced if those subcontractors can actually bear the risks associated to the service provision.

The adaptability of the infrastructure to the changing requirements of the healthcare sector is not explicitly addressed by the novel PFI/PPP models. For example, service adaptability can be enhanced by contract clauses that specify options for the change of use of the infrastructure (Cruz and Marques, 2013): the higher the volatility, the higher the benefits of embedding real options in long-term contracts. This would be a way of addressing the criticism of public sector partners being locked into long-term contracts but it would require additional costs in the definition of the contract. Another option is to reduce the infrastructural requirement for services. The UK National Health Service is likely to move away from institution-based provision of healthcare treatments to community-based provision (Future Hospital Commission, 2013). This also involves family doctors merging their practices and taking control of 60% of the former health authorities' (primary care trusts) budgets. Clinical commissioning groups are expected to choose to provide more mundane treatments in their own practices instead of sending their patients to hospitals. The public sector partner can reduce fragility via multisector PFI/PPP infrastructures designed (or having the potential) to be used for different purposes. For example, primary healthcare centers and hospital PFI/PPPs could be no longer required because of the changing medical and healthcare contexts noted above. If that is the case, the PFI/PPP hospital contract could specify that the physical infrastructure could be used (after suitable modifications) to provide residential care facilities for the forecasted growing numbers of very elderly people. In theory, this spatial adaptability would reduce asset specificity of hospital infrastructure and so reduce the high transaction costs associated with the simultaneous occurrence of asset specificity, bounded rationality and scope for opportunism. Although having the potential to increase the adaptability of the infrastructure, such solutions require additional research to evaluate their actual feasibility in practice.

Finally, neither PF2 nor NPD explicitly address the risk governance of PFI/PPPs that goes far beyond technocratic measurement of risks and VfM (Brown and Osborne, 2013; Foo et al., 2011). Forecasting in complex systems, such as PFI/PPPs, provides the illusion of control while ignoring the potential negative outcomes of unpredictable events (Makridakis et al., 2009). Therefore, although risk management is useful to address predictable risks, overconfidence in predictions ignores the dangers hidden outside the boundaries of predictability (Stulz, 2009). As such, it is necessary to ‘buffer economic systems against unexpected shocks’ (Orrell and McSharry, 2009: 742) or, in other words, to complement technocratic modeling with governance frameworks (customised to individual PFI/PPPs and involving focus groups) to address uncertainty through the detection and management of fragilities.

All PFI/PPP models require a robust system of governance engaging all stakeholders to address intangible risks and uncertainties that are not prescribed by contract. However, there is little evidence of PFI/PPPs adopting a broad socio-economic governance framework, instead being over-reliant on legal and regulatory approaches to preservation of VfM (Broadbent et al., 2008). The ‘one point in time’ VfM options appraisal is based on the contract requirements. As contracts are generally static, incomplete and unable to deal with the dynamism of the service context, the VfM assessment cannot fully appreciate the evolution of risks. Furthermore, PFI/PPP models are often perceived to be the only viable way to fund infrastructures and to overcome public capital budget constraints (Atun and McKee, 2005; Ruane, 2000) thus making the ex-ante VfM assessment purely artificial with little impact on the final procurement decision.

Risk governance does not avoid unpredictable events happening, but it aims at adapting the system to the evolution of the context so that it can cope with uncertainty. For Taleb (2012), the governance of complex systems has to be driven from the inside in order to make the system evolutionary. Consequently, managing fragility is built upon a governance framework that enables incremental but continuous changes in the system to ensure its evolution. Indeed, it is claimed that the impacts of a system's failure are minimized when decisions are small (i.e. do not radically change the system), continuous and kept local, rather being radical, discontinuous and centralized (Anttiroiko et al., 2011).

Conclusion

PFI/PPP and similar models have enabled provision of health infrastructures and services that were not financially (or politically) feasible through traditional public procurement. PFI/PPPs have been validated by ex-ante VfM tests that are inherently unable ex-post to identify and manage the fragilities. Fragilities identified above via our systematic literature review and content analysis are rooted in the financial structure of PFI/PPPs together with inadequate ring-fencing of risks and inflexible contract requirements.

In light of the current development of the PFI/PPP policy, our findings make it possible to advance novel avenues of research for managing fragilities. First, for proactive fragility management, it is necessary to consider if the potential benefits stemming from diversification of funding can be actually achieved. For PFI/PPPs already in place, the refinancing of debt provides an opportunity to increase the diversification of debt funding. For example, during the refinancing of debt the overdependence on bank loans can be progressively reduced by issuing project bonds and involving a wider range of investors. Nonetheless, such projects may fail to attract a wider range of investors given that institutional investors are more risk adverse and need to fulfill statutory EU solvency requirements.

Second, it is necessary to consider the pros and cons related to the move from higher to lower leverage. Although, compared with equity, borrowing reduces the flexibility of the SPV in using its cash flows, the pros and cons of moving from higher to lower dependence upon debt funding must be considered in light of the political context and public policy objectives underpinning use of PFI/PPPs, as made clear by the different approaches adopted by governments in England (PF2) and Scotland (NPD).

Third, further research is required to determine how adaptable are PFI/PPP contracts to the changing needs of the healthcare system and the difficulties of embedding (greater) flexibility into contracts.

Fourth, further research is required on the risk governance system of PFI/PPPs. Infrastructure projects are often designed and, sometimes, forced to pass the VfM test and obstacles to the subsequent implementation of a robust risk governance may arise from the vested interests of powerful contractors and financiers. Moreover, the pressures on EU member states to reduce their debts may create incentives for governments to outsource provision and management of overly bundled service infrastructure in place of conventional procurement, irrespective of VfM considerations and inhibiting the ongoing proactive governance of both risks and uncertainty which requires engaging all relevant stakeholders involved in the infrastructure and service provision in order to safeguard rather than predict VfM.

Finally, we acknowledge that our study is based only on the fragilities identified so far in our content analysis of the academic literature and that the evolution of PFI/PPP contracts can raise new sources of uncertainty that have to be identified and adequately managed after the signing of contracts in order to safeguard VfM over time.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work has been supported by FondazioneCariplo, grant no. 2011-0788.

Notes

References

ACCA (2012) Taking Stock of PPP and PFI Around the World, London: Certified Accountants Educational Trust.

Accounts Commission (2002) Taking the Initiative – Using PFI Contracts to Renew Council Schools, Edinburgh: Audit.

Akintoye

Beck

Hardcastle

(2003) Public-Private Partnerships: Managing Risks and Opportunities, Oxford, UK: Blackwell Science Ltd.

Aldred

(2008) Managing risk and regulation within new local ‘health economies’: The case of NHS LIFT (local improvement finance trust). Health Risk & Society 10(1): 23–36.

Andon

(2012) Accounting related research in PPPS/PFIs: Present contributions and future opportunities. Accounting, Auditing & Accountability Journal 25(5): 876–924.

Anttiroiko

Bailey

Valkama

(2011) Innovation in public governance in the Western world. In: Anttiroiko

Bailey

Valkama

(eds) Innovation in Public Governance, Amsterdam, The Netherlands: IOS Press.

Asenova

Beck

(2010) Crucial silences: When accountability met PFI and finance capital. Critical Perspective on accounting 21(1): 1–13.

Asenova

Stein

McCann

(2007) Private sector participation in health and social care services in Scotland: Assessing the risk. International Review of Administrative Sciences 73(2): 275–292.

Atun

McKee

(2005) Is the private finance initiative dead? It may have failed in the United Kingdom, but that won't stop it being exported. British Medical Journal 331(7520): 792–793.

10.

Azevedo

DM de

Costa

RKS

Miranda

FAN de

(2013) Use of the alceste in the analysis of qualitative data: Contributions to researches in nursing. Journal of Nursing 7(7): 5014–5022.

11.

Bailey

Asenova

Beck

(2009) UK public private partnerships and the credit crunch: A case of risk contagion? Journal of Risk and Governance 1(3): 1–11.

12.

Bara

Weale

Bicquelet

(2007) Analysing parliamentary debate with computer assistance. Swiss Journal of Political Science 13(4): 577–605.

13.

Barlow

Koberle-Gaiser

(2008) The private finance initiative, project form and design innovation. The UK's hospitals programme. Research Policy 37(8): 1392–1402.

14.

Barretta

Ruggiero

(2008) Ex-Ante evaluation of PFIs within the Italian health-care sector: What is the basis for this PPP? Health Policy 88: 15–24.

15.

Broadbent

Gill

Laughlin

(2003) Evaluating the private finance initiative in the National Health Service in the UK. Accounting, Auditing & Accountability Journal 16(3): 422–445.

16.

Broadbent

Gill

Laughlin

(2008) Identifying and controlling risk: The problem of uncertainty in the private finance Initiative in the UK's National Health Service. Critical Perspectives on Accounting 19(1): 40–78.

17.

Brown

Osborne

(2013) Risk and innovation: Towards a framework for risk governance in public services. Public Management Review 15(2): 186–208.

18.

Burger P, Tyson J, Karpowicz I, et al. (2009) The Effects of the Financial Crisis on Public-Private Partnerships. IMF Working Paper WP/09/144, Washington DC: International Monetary Fund.

19.

Cruz

Marques

(2013) Flexible contracts to cope with uncertainty in public–private partnerships. International Journal of Project Management 31(3): 473–483.

20.

Cuthbert

(2010) The Royal infirmary of Edinburgh: A case study on the workings of the private finance initiative. Public Money & Management 30(6): 371–378.

21.

Demirag

Khadaroo

Stapleton

(2010) Risks and the financing of PPP: Perspectives from the financiers. The British Accounting Review 43(4): 294–310.

22.

Edwards

Shaoul

(2003) Partnerships: For better, for worse? Accounting, Auditing & Accountability Journal 16(3): 397–421.

23.

Engel

Fischer

Galetovic

(2010) The economics of infrastructure finance: Public-private partnerships versus public provision. EIB PAPERS 15(1): 40–69.

24.

EPEC (2010) Capital Markets in PPP Financing, Luxembourg: European PPP expertise Centre.

25.

Flyvbjerg

Holm

Buhl

(2002) Underestimating costs in public works projects_error or lie? Journal of the American Planning Association 68(3): 279–295.

26.

Foo

Asenova

Bailey

(2011) Stakeholder engagement and compliance culture. Public Management Review 13(5): 707–729.

27.

Froud

(2003) The private finance initiative: Risk, uncertainty and the state. Accounting, Organizations and Society 28(6): 567–589.

28.

Future Hospital Commission (2013) Future Hospital: Caring for Medical patients. A Report From the Future Hospital Commission to the Royal College of Physicians, London: Royal College of Physicians.

29.

Grimsey

Lewis

(2007) Public Private Partnerships: The Worldwide Revolution in Infrastructure Provision and Project Finance, Edward Elgar Publishing Limited: Northampton, MA.

30.

Grout

(2005) Value for money measurement in public-private partnerships. EIB PAPERS 10(2): 32–57.

31.

Hellowell

(2013) PFI redux? Assessing a new model for financing hospitals. Health Policy 113(1–2): 77–85.

32.

Hellowell

Pollock

(2009) Non-profit distribution: The Scottish approach to private finance in public services. Social Policy and Society 8: 405–418.

33.

Hellowell

Pollock

(2010) Do PPPs in social infrastructure enhance the public interest? Evidence from England's National Health Service. Australian Journal of Public Administration 69: S23–S34.

34.

Helm

(2010) Infrastructure and infrastructure finance: The role of the government and the private sector in the current world. EIB PAPERS 15(2): 8–27.

35.

HM Treasury (2006) Value for Money Assessment Guidance, London: The Stationery Office.

36.

HM Treasury (2012a) A New Approach to Public Private Partnerships, London: The Stationery Office.

37.

HM Treasury (2012b) Equity Investment In Privately Financed Projects, London: The Stationery Office.

38.

Hodge

Greve

(2009) PPPs: The passage of time permits a sober reflection. Economic Affairs 29(1): 33–39.

39.

Inderst G (2009) Pension Fund Investment in Infrastructure. OECD Working Papers on Insurance and Private Pensions, No. 32. Paris: OECD.

40.

KPMG (2010) Project Finance and the Capital Markets – Bridging the divide. Amstelveen, Netherlands: KPMG International.

41.

Lonsdale

(2005) Post-contractual lock-in and the UK private finance initiative (PFI): The cases of national savings and investments and the Lord Chancellor's Department. Public Administration 83(1): 67–88.

42.

Makridakis

Hogarth

Gaba

(2009) Forecasting and uncertainty in the economic and business world. International Journal of Forecasting 25(4): 794–812.

43.

Marriott

(2014) A Dangerous Diversion: Will the IFC's flagship health PPP bankrupt Lesotho's Ministry of Health?, Oxford: Oxfam.

44.

Marsilio

Cappellaro

Cuccurullo

(2011) The intellectual structure of research into PPPS: A bibliometric analysis. Public Management Review 13(6): 763–782.

45.

McKee

Edwards

Atun

(2006) Public-private partnerships for hospitals. Bulletin of the World health Organization 84(11): 890–896.

46.

NAO (2009) Private Finance Projects, London: The Stationery Office.

47.

NAO (2011) Lessons from PFI and other projects, London: The Stationery Office.

48.

NAO (2012a) Peterborough and Stamford Hospital NHS Foundation Trust, London: The Stationery Office.

49.

NAO (2012b) Department of Health: Securing the Future Financial Sustainability of the NHS, London: The Stationery Office.

50.

Orrell

McSharry

(2009) System economics: Overcoming the pitfalls of forecasting models via a multidisciplinary approach. International Journal of Forecasting 25(4): 734–743.

51.

Regan

Smith

Love

(2011) Impact of the capital market collapse on public-private partnership infrastructure projects. Journal of Construction Engineering and Management 37(1): 6–16.

52.

Roehrich

Caldwell

(2012) Delivering integrated solutions in the public sector: The unbundling paradox. Industrial Marketing Management 41(6): 995–1007.

53.

Roehrich

Lewis

George

(2014) Are public private partnerships a healthy option? A systematic literature review. Social Science & Medicine 113: 110–119.

54.

Ruane

(2000) Acquiescence and opposition: The private finance initiative in the National Health Service. Policy and Politics 28(3): 411–424.

55.

Sawyer

(2005) The private finance initiative: The UK experience. Research in Transportation Economics 15: 231–245.

56.

Scottish Future Trust (2011) NPD Model Explanatory Note. Available at: http://www.scottishfuturestrust.org.uk/files/publications/Explanatory_Note_on_the_NPD_Model_pdf (Updated December 2011).

57.

Shaoul

(2005) A critical financial analysis of the private finance initiative: Selecting a financing method or allocating economic wealth? Critical Perspectives on Accounting 16: 441–471.

58.

Shaoul

(2011) ‘Sharing’ political authority with finance capital: The case of Britain's public private partnerships. Policy and Society 30(3): 209–220.

59.

Shaoul

Stafford

Stapleton

(2008) The cost of using private finance to build, finance and operate hospitals. Public Money & Management 28(2): 101–108.

60.

Schonhardt-Bailey

(2013) Deliberating American Monetary Policy: A Textual Analysis, Cambridge, MA: The MIT Press.

61.

Stulz

(2009) Six ways companies mismanage risks. Harvard Business Review. March, 1–8.

62.

Taleb

(2012) Antifragile: Things That Gain From Disorder, New York, NY: Random House.

63.

Toms

Beck

Asenova

(2011) Accounting, regulation and profitability: The case of PFI hospital refinancing. Critical Perspectives on Accounting 22(7): 668–681.

64.

Torchia

Calabrò

Morner

(2015) Public–private partnerships in the health care sector: A systematic review of the literature. Public Management Review 17(2): 236–261.

65.

Tranfield

Denyer

Smart

(2003) Towards a methodology for developing evidence-informed management knowledge by means of systematic review. British Journal of Management 14(3): 207–222.

66.

UNISON (2011) PPP/PFI Projects in Scotland. UNISON Scoltalnd, Policy Briefing.

67.

Vecchi

Hellowell

(2012) Securing a better deal from investors in public infrastructure projects: Insights from capital budgeting. Public Management Review 15(1): 109–129.

68.

Vecchi

Hellowell

Gatti

(2013) Does the private sector receive an excessive return from investments in health care infrastructure projects? Evidence from the UK. Health Policy 110: 243–270.

69.

Wagenvoort

de Nicola

Kappeler

(2010) Infrastructure finance in Europe: Composition, evolution and crisis impact. EIB PAPERS 15(1): 16–39.

70.

Williamson

(1979) Transaction-cost economics: The governance of contractual relations. Journal of Law and Economics 22(2): 233–261.

71.

Williamson

(1988) Corporate finance and corporate governance. Journal of Finance 43(3): 567–591.