Chapter 1. Public-Private Partnerships
- M. Cangiano, Barry Anderson, Max Alier, Murray Petrie, and Richard Hemming
- Published Date:
- April 2006
There is no clear agreement on what does and what does not constitute a PPP. A PPP has been defined as “the transfer to the private sector of investment projects that traditionally have been executed or financed by the public sector” (European Commission, 2003, p. 96). But in addition to private execution and financing of public investment, PPPs have two other important characteristics: first, there is an emphasis on service provision and investment by the private sector; and, second, significant risk is transferred from the government to the private sector. Some or all of these four features also characterize other means by which the role of government in the economy has been reduced over the last 20 years—including privatization, joint ventures, franchising, and contracting out.1 However, PPPs are distinct in that they represent cooperation between the government and the private sector to build new infrastructure assets and to provide related services. In fact, two methods that have been used specifically to reduce the role of government in the economy in favor of the private sector—concessions and operating leases—are in the first case a form of PPP and in the second case can be structured like a PPP.
A. Basic Features of PPPs
A typical PPP takes the form of a design-build-finance-operate (DBFO) scheme. Under such a scheme, the government specifies the services it wants the private sector to deliver, and then the private partner designs and builds an asset specifically for that purpose, finances its construction, and subsequently operates the asset (i.e., provides the services deriving from it). This contrasts with traditional public investment projects, under which the government contracts with the private sector to build an asset but provides the design and financing itself and, in most cases, then operates the asset once it is built. The difference between these two approaches reflects a belief that giving the private sector responsibility for designing, building, financing, and operating an asset leads to increased efficiency in service delivery. More specifically, such “bundling” is believed to provide an incentive for the private sector to design and build assets with features that enhance the quality or lower the costs of service provision over the long term.
The government is, in many cases, the main purchaser of services provided under a PPP. These services can be purchased either 1) for the government’s own use (a prison), 2) as an input to provide another service (a school), or 3) on behalf of final consumers (a free-access road). Private operators also sell services directly to the public, as with a toll road or railway. Such arrangements are often referred to as concessions, and the private operator of a concession (the concessionaire) pays the government a concession fee and/or a share of profits. Typically, the private operator owns the PPP asset while operating it under a DBFO scheme, and the asset is transferred to the government at the end of the operating contract, usually for less than its true residual value (and often at zero or a small, nominal cost). In this case, a PPP is often referred to as a build-operate-transfer (BOT) or build-own-operate-transfer (BOOT) scheme.
The term PPP is sometimes used to describe a wider range of arrangements. In particular, some PPPs exclude functions that characterize DBFO schemes. Most common in this respect are schemes that combine traditional public investment and private sector operation of a government-owned asset (note that the builder and the operator of the asset are not the same). This arrangement sometimes takes the form of an operating lease, although it can be considered a PPP if the private operator has responsibility for asset maintenance and improvement.2 Private sector involvement in asset building alone—which can take the form of a design-build-finance-transfer (DBFT) scheme or a financial lease—is not, strictly speaking, a PPP because it does not involve service provision by the private sector. This paper does not seek to explicitly exclude any type of arrangement from the definition of a PPP, and refers to cases in which the public sector partner is a public enterprise rather than the government.3 However, it pays most attention to PPPs that involve both investment and service delivery by the private sector, as well as private financing and ownership. Hence the focus is on DBFO schemes.4Box 1 describes some of the many variants of PPP schemes.
Box 1.PPP Schemes and Modalities
|Build-own-operate (BOO)||The private sector designs, builds, owns, develops, operates, and manages an asset with no obligation to transfer ownership to the government. These are variants of design-build-finance-operate (DBFO) schemes.|
|Buy-build-operate (BBO)||The private sector buys or leases an existing asset from the government; renovates, modernizes, and/or expands it; and then operates the asset, again with no obligation to transfer ownership back to the government.|
|Wrap-around addition (WAA)|
|Build-operate-transfer (BOT)||The private sector designs and builds an asset, operates it, and then transfers it to the government when the operating contract ends, or at some other prespecified time. The private sector partner may subsequently rent or lease the asset from the government.|
Uses for PPPs
PPPs appear to be particularly well-suited to providing economic infrastructure. This is primarily for three reasons. First, sound projects that address clear bottlenecks in roads, railways, ports, power, and other infrastructure are likely to have high economic rates of return and therefore to be attractive to the private sector. Second, in economic infrastructure projects, the private sector can be made responsible not only for constructing the infrastructure asset but also for providing the principal services related to it, allowing them to tailor asset design specifically to this purpose. Third, to the extent that these services are supplied directly to final users, charging is both feasible and, from an efficiency standpoint, desirable.
Social infrastructure is somewhat different in these regards. Although many social investment projects are clearly worthwhile, the private sector is not usually the principal supplier of social services. Thus, while PPPs may be formed to build and maintain public schools and hospitals, the government tends to continue to be the provider of the education and health care services deriving from them. Moreover, charging for government-supplied social services is not commonplace. Hence, social infrastructure PPPs offer smaller potential efficiency gains than either economic infrastructure PPPs or schemes that combine public investment and subsequent contracting out of the operation and maintenance of schools, hospitals, and other social infrastructure.5 That said, there are many examples of successful PPPs in social sectors.
The private sector can raise financing for PPP investment in a variety of ways. When services are sold to the public, the private sector can go to the market using the projected income stream from a concession (e.g., toll revenue) as collateral. Where the government is the main purchaser of services, collateral can comprise shadow tolls paid by the government (i.e., payments related to the demand for services) or service payments by the government under operating contracts (which are based on continuity of service supply, rather than on service demand). The government may also make a direct contribution to project costs. This can take the form of equity (where there is profit sharing), a loan, or a subsidy (where social returns justify a project). The government also can guarantee private sector borrowing.
PPP financing is often provided via special purpose vehicles (SPVs). An SPV can be a consortium of companies responsible for all aspects of a PPP, and as such it can be a means of exploiting the advantages from bundling. In practice though SPVs are often a group of banks and other financial institutions that combine and coordinate the use of their capital and financial expertise. Insofar as this is their purpose, an SPV can facilitate a well-functioning PPP.6 However, an SPV can also serve as a veil behind which the government controls a PPP either via the direct involvement of public financial institutions, an explicit government guarantee of borrowing by an SPV, or a presumption that the government stands behind it. Where this is the case, the risk is that the SPV will be used to shift debt off the government balance sheet. Private sector accounting standards require that an SPV be consolidated with an entity that controls it; by the same token, an SPV that is controlled by the government should be consolidated with the latter, and its operations should be reflected in the fiscal accounts.7, 8
Where the government has a claim on future project revenue, it can contribute to the financing of a PPP by securitizing that claim. With a typical securitization operation, the government sells a financial asset—its claim on future project revenue—to an SPV. The SPV then sells securities backed by this asset to private investors and uses the proceeds to pay the government, which in turn uses them to finance the PPP. Interest and amortization are paid by the SPV to investors from the government’s share of project revenue. Because the investors’ claim is against the SPV, government involvement in the PPP appears limited. However, the government is in effect financing the PPP, although this fact can be masked by the recording of sale proceeds received from the SPV as revenue.9
B. Country Experience
A number of advanced OECD member countries now have well-established PPP programs. Undoubtedly the best developed of these is the United Kingdom’s Private Finance Initiative (PFI), which began in 1992. The PFI is currently responsible for about 14 percent of public investment, with projects in most key infrastructure areas. Other countries with significant PPP programs include Australia (and in particular the state of Victoria) and Ireland, while the United States has considerable experience with leasing.10 Many Western European countries now have PPP projects, including Finland, Germany, Greece, Italy, the Netherlands, Portugal, and Spain, although their share in total public investment is quite small.11 Reflecting a need for infrastructure investment on a large scale and weak fiscal positions, a number of countries in Central and Eastern Europe have embarked on PPPs, including Croatia, the Czech Republic, Hungary, and Poland.12 There are also fledgling PPP programs in Canada and Japan. PPPs in most of these countries are dominated by road projects. In addition, greater use of PPP–type arrangements has been proposed to develop a trans-European road network (European Council, 2003).
In the rest of the world, PPPs have made fewer inroads. In Latin America, however, Chile, Colombia, and Mexico have used PPPs to promote private sector participation in public investment projects. Chile has a well-established PPP program that has been used mainly for the development of roads, airports, prisons, and irrigation. In Colombia, PPPs have been used since the early 1990s, but early projects were not well-designed. A relaunched PPP program emphasizes road projects. In Mexico, PPPs were first used, though unsuccessfully, in the 1980s to finance roads. Since the mid-1990s, Mexico has used PPPs with greater success for public investment projects in the energy sector through the PIDIREGAS scheme, and they are beginning to be extended to the provision of other services.13 Some other Latin American countries, most notably Brazil, are planning significant use of PPPs. There has also been some discussion of a regional approach to infrastructure development that would involve PPP—type arrangements.14
PPPs are beginning to take off in Korea, the Philippines, and Singapore (and, as noted above, also in Japan), but progress elsewhere in Asia is limited, despite strong interest in PPPs in some other countries, including India, Indonesia, and Thailand. In Africa, South Africa, a clear regional leader, has embarked upon or is developing PPPs in a number of sectors. Few other African countries have much experience with PPPs, although Mozambique has embarked on concessions to rehabilitate rail terminals and a port, while other countries have tried alternative forms of private sector involvement in infrastructure, especially in the water and power sectors (e.g., in Côte d’Ivoire and Senegal). Appendix 1 outlines the experience with PPPs in Chile, Ireland, South Africa, and the United Kingdom. Selected experiences of other countries are included elsewhere in the paper.
Although a number of countries have developed PPP programs, it is too early to draw meaningful lessons from their experiences. The U.K. government published a comprehensive assessment of the PFI (H.M. Treasury, 2003), which was informed in part by the results of independent studies and was favorable in terms of both the program’s procedures and its outcomes. Overall, however, while particular aspects of country experiences support some of the conclusions in this paper, few general lessons can be drawn yet, especially from the experiences of emerging market economies and developing countries.
C. Economic Principles
PPPs themselves have not been subject to extensive economic analysis. However, there is a good deal of analytical work that can be brought to bear on the issues raised by PPPs.15
Ownership and Contracting
The standard arguments for and against government ownership are relevant to PPPs. As a general rule, private ownership is to be preferred where competitive market prices can be established. Under such circumstances, the private sector is driven by competition in the product market to sell goods and services of a quality and price that is acceptable to consumers and by the discipline of the capital market to make profits. Various market failures (natural monopoly, externalities) can justify government ownership, although the result can be that government failure simply substitutes for market failure.16 Even then, there are those who argue that private ownership should be preferred because more often than not it offers potential efficiency benefits (Shleifer, 1998), and it leans against a possible bias in favor of government ownership. Against this background, PPPs can be seen as a means, on one hand, to combine the relative strengths of the government and the private sector in the ownership of assets and the provision of services that respond to market failure and, on the other hand, to minimize the risks of government failure.
Recent advances in the theory of ownership and contracting provide a more specific analytical justification for PPPs. The trade-off facing a government seeking to arrange for the provision of a particular service is between quality and efficiency. The government has the capacity to achieve a desired quality standard, but it may have difficulty doing so while also containing costs. The private sector can use its superior management skills and greater capacity for innovation to more actively pursue opportunities to reduce costs, but service quality may be compromised in the process. However, private sector provision of the service may be workable if the government can award a fully specified, enforceable (i.e., complete) operating contract to a private sector partner. Hence PPPs are well-suited to situations in which the government can clearly identify the quality of services it wants the private sector to provide and can translate these into measurable output indicators. The government can contract with the private sector in a way that links service payments to monitorable service delivery. As a result, PPPs tend to be better suited to cases where service requirements are not expected to vary substantially over time and where technical progress is unlikely to radically change the way in which the service is provided.
The case for PPPs is weaker when the government cannot write complete contracts. In general, services for which overall quality is not inherently suited to objective measurement (e.g., national defense, public law and order, diplomatic missions) are not candidates for PPPs. That said, elements of these services may be contracted (including the construction and maintenance of military bases, police stations, courts, and embassies), although the scope for efficiency gains may be limited for the reasons given above in connection with social infrastructure. A specific concern with PPPs is that even when service quality can be defined in a contract, asset quality is more difficult to define because poor construction may become apparent only after many years when the government is forced into costly repairs or difficult contract renegotiations (Grout, 1997). However, bundling can help address this problem by giving a private operator a clear interest in the quality of an asset (Hart, 2003).17
PPP projects involve a range of different risks. These can be usefully divided into five, somewhat overlapping categories:
- construction risk, which is related to design problems, building cost overruns, and project delays;
- financial risk, which is related to variability in interest rates, exchange rates, and other factors affecting financing costs;
- availability risk, which is related to the continuity and quality of service provision (which in turn depends on the “availability” of an asset);
- demand risk, which is related to the ongoing need for services; and
- residual value risk, which is related to the future market price of an asset.18
These risks are present in public, private, and PPP projects, but PPPs specifically seek to transfer some of them from the government to the private sector. While public projects can benefit from an inflow of private capital and a change in management responsibility alone, it is necessary to achieve significant risk transfer in order to derive the full benefits from such changes. The impact of risk transfer on financing costs and the need to price risk so as to ensure it is transferred efficiently then have to be addressed.
Risk Transfer and Financing Costs
Transferring project risk from the government to the private sector should not affect the cost of financing a project. This follows from the Modigliani-Miller theorem, which says that the cost of capital depends only on overall project risk. With complete markets in risk bearing, project risk is independent of whether a project is financed by the government or the private sector. However, with incomplete markets in risk bearing, project risk depends on how widely the risk can be spread, in which case the source of financing can influence overall project risk. Because the government can spread risk across taxpayers in general, the usual argument is that this gives the government an advantage over the private sector in terms of managing risk (Arrow and Lind, 1970). But the private sector can spread risk across financial markets, which means that it may not be at a significant disadvantage, and private sector risk managers may be more skilled than those in government. The outcome could often be that project risk is lower in the private sector.19
This result may appear to rest somewhat uneasily with the fact that private sector borrowing generally costs more than government borrowing. However, this mainly reflects differences in default risk. The government’s power to tax reduces the likelihood that it will default on its obligations, and investors are therefore prepared to lend to the government at close to the risk-free interest rate, even to finance risky projects. This being the case, when PPPs substitute private borrowing for government borrowing, financing costs will in most cases rise even if project risk is lower in the private sector. The key issue then becomes whether PPPs result in efficiency gains that more than offset the higher private sector borrowing costs.20 The impact of PPPs on efficiency is taken up below.
Pricing of Risk
When considering the PPP option, the government has to compare the cost of public investment and government provision of services with the cost of providing services through a PPP. Since risk transfer is key to realizing the increased efficiency available through PPPs, the government seeks to relieve itself of risks that it believes the private sector can manage better. To do this, the price that the government is prepared to pay to be relieved of these risks must be set at a high enough level that the private sector willingly assumes them. In this connection, it is important to distinguish between project-specific risk and market risk. Project-specific risk reflects variations in outcomes for individual projects or groups of related projects. Thus for a road, the project-specific risk could derive from interrupted supplies of building materials, labor problems, unfavorable weather, or obstruction by environmental groups. Project-specific risk is diversifiable across a large number of government or private sector projects and does not need to be priced by the government. Market risk reflects underlying economic developments that affect all projects, and it is not diversifiable and therefore has to be properly priced.
The government and the private sector typically adopt different approaches to pricing market risk. The government tends to use the social time preference rate (STPR) or some other risk-free rate to discount future cash flows when appraising projects, while private bidders for PPP projects typically include a risk premium in the discount rate they apply to future project earnings.21 Given this mismatch, the government may reject reasonable private sector bids for a PPP project. This may produce a bias against PPPs and in favor of public investment, which is counterproductive if the objective is to promote PPPs as a more efficient alternative to public investment and government provision of services.22 Moreover, even if the PPP route is chosen (maybe because of political preference), the allocation of risk between the government and the private sector may not be efficient, because the private sector may choose techniques of production or other project design features that are less efficient simply because they carry lower risk.23 Also, the private sector partner may respond to the underpricing of risk by compromising on quality to the extent possible without violating its contract with the government. On the other hand, it is also possible for the government to overprice risk and to overcompensate the private sector for taking it on, which raises the cost of PPPs relative to direct public investment. Finally, there may be incentives for the government to compensate for the underpricing of risk by extending guarantees, which may end up costing the government more over the longer term.
Competition, Regulation, and Efficiency
Much of the case for PPPs rests on the relative efficiency of the private sector. While there is an extensive literature on this subject, the theory is ambiguous and the empirical evidence is mixed. If there is a common theme, it relates to the importance of competition as a source of efficiency in both the private and public sectors. This explains the use of franchising as a means of having the private sector engage in repeated competition for markets that are inherently monopolistic yet still contestable (which is distinct from having continuous competition in a market). As explained, the scope for competition in the activities undertaken by PPPs is more limited because they tend to be less contestable—economic infrastructure involves large sunk costs and social infrastructure is undervalued—and because the provision of single-use assets inevitably creates a bilateral monopoly situation. One area where competition is clearly feasible is the award of construction and service contracts, and fostering competition in this area is crucial for realizing the benefits of PPPs in substituting private sector for public sector capital, improving management, and fostering innovation.
Incentive-based regulation is also important. Where a private operator can sell to the public, but there is little scope for competition, the government usually regulates prices. However, the challenge is to design well-functioning regulation that increases output (toward the social optimum), holds down prices, and limits monopoly profits while preserving the incentive for private firms to be more efficient and reduce costs. The two most common forms of regulation are rate of return regulation and price regulation. The former suffers from the problems involved in establishing appropriate cost benchmarks in a monopolistic situation and is therefore weak on incentive grounds. The latter caps price increases and therefore has potential for success on both counts. However, the fact that price caps are often adjusted to reflect rate of return considerations means that both types of regulation tend to be quite similar in their effects. Another type of regulation that has more promise is yardstick competition, in which rate of return regulation is based on costs in closely related domestic or international firms or in a hypothetical efficient firm, although this type of regulation is informationally demanding. Finally, profit sharing between the government and the private partner is an alternative form of regulation that preserves incentives, although it can still lead to excessive profits. This being the case, it tends to work better when the government is the main purchaser of services (Laffont and Tirole, 1999).
D. Institutional Framework
Successful PPPs deliver high-quality services to consumers and the government at costs that are significantly lower than those available through public investment and government provision of the same services. The preceding discussion suggests that PPPs are more likely to result in efficiency gains that offset higher private sector borrowing costs if they have the following three characteristics: 1) the quality of services can be readily defined and measured; 2) there is adequate risk transfer to the private sector; and 3) there is either competition or incentive-based regulation. These features should be reflected in the policy framework for PPPs, along the lines of that provided by the state of Victoria, Australia, which is summarized in Box 2. However, the success of PPPs is also dependent on the existence of an appropriate institutional framework. The challenges in this regard are greater in emerging market economies and developing countries, but they are also faced by advanced OECD countries. A PPP program should proceed with caution in the absence of an adequate institutional framework, which should be characterized by political commitment, good governance, government expertise, and effective project appraisal and selection.
Box 2.PPP Policy Framework in Victoria, Australia1
Victoria has developed a detailed and explicit policy on PPPs, Partnerships Victoria. An emphasis on value for money and the public interest is the key feature of the policy. There is, however, no presumption that the private sector (or, for that matter, the public sector) can deliver projects more efficiently or effectively. Instead, decisions are made on their merits and outcomes are judged on the basis of the public benefits obtained.
The policy stipulates that PPP projects should focus on the specification of the end result, rather than the means of delivery, and that performance measures should be established to ensure that the quality of services delivered meets the needs of the community. Private participation is to be the subject of competitive bidding, consistent with the government’s procurement policies, and there should be an emphasis on transparency and disclosure of processes and outcomes, while acknowledging the need to protect commercial confidentiality when appropriate. Moreover, standardized approaches are to be used whenever possible to minimize transaction costs, and, if needed, incentives should be provided to encourage high-level performance.
Partnerships Victoria projects are required to have a number of features. Outputs should be clearly specified (including measurable performance standards), and one or more private parties must be fully accountable to the government for the delivery of services. The clear specification of required outputs allows bidders to compete in devising creative means of delivering those outputs, with a view to reducing costs. Public agencies should limit detailed specification of inputs, such as the design of infrastructure or the means by which outputs are to be generated. There must also be a clear articulation of the government’s responsibilities, including the monitoring of outcomes. Finally, payments are due only upon delivery of the specified services, if they meet the required standards.1 Based on Victoria (2000, 2001) and material available at the Partnerships Victoria website (http://www.partnerships.vic.gov.au/).
Political commitment is a prerequisite for success. A PPP is a major commitment on the part of the private sector partner, who needs to know that politicians are also committed to the partnership. Uncertainty in this regard gives rise to political risk, which is not conducive to making long-term business decisions. Potential private sector partners also need to know that the government is fair in its dealings with the private sector and will meet its commitments under PPPs. In addition, it is also important to establish clear channels of responsibility and accountability for government involvement in PPPs.
Good governance is another prerequisite for success. Widespread corruption in government is a serious obstacle to successful PPPs, in the same way that it prevents successful privatization (Lora and Panizza, 2003). An appropriate legal framework provides reassurance to the private sector that contracts will be honored. This may require changes or additions to existing laws. This was the case in Italy and Spain, which recently revamped legal frameworks that for many years created obstacles to PPPs. In Italy, the 1994 Merloni Law has undergone a number of changes designed to facilitate private participation in infrastructure investment, while the 2001 Legge Obiettivo established a fast-track system for strategically important infrastructure projects.24 In Spain, the 2003 Concessions Law supplements a number of laws that already allow PPPs by extending private financing options.25 In both Italy and Spain, the laws have also been amended to better secure creditor rights.
The comparative success of Chile’s concessions program can be attributed in significant measure to the fact that it is backed by a comprehensive concessions law that addresses not only the basic requirements for effective concessions (the bidding process, rights and obligations of parties, property appropriation), but also the treatment of possible disputes and the cancellation and transfer of contracts. Brazil has recently enacted a PPP law, although some forms of PPP were already governed in part by legislation on concessions and procurement and by the transparency requirements of fiscal responsibility legislation. The provisions of the Brazil legislation are summarized in Box 3. The legal framework for PPPs should be supplemented by clear, credible, and efficient dispute-resolution mechanisms. Finally, it is important that PPPs should face nondiscriminatory taxation and regulation regimes. In India, while there is recognition of the need for a comprehensive legal framework, the current emphasis is more on reducing regulatory barriers and demonstrating sustained political commitment to private sector involvement.
Box 3.PPP Legislation in Brazil
The Brazilian law applies to PPPs at all levels of government, and it complements existing legislation in the fiscal area, including the Concessions Law and the Procurement Law.1 The law creates a new contractual modality through which a private partner is responsible for the construction and financing of a public asset that supports the provision of a contracted service. The law prohibits the use of PPPs to hire personnel, purchase equipment, or carry out public works and sets a minimum value for PPP contracts (R$20 million, about US$9 million). At the expiration of the contract, with a maximum duration of 30 years, the asset must be transferred to the public sector, with or without a final payment (depending on the contract).
The law includes safeguards for public finances by limiting the exposure to PPPs. Specifically, for all levels of government, the law limits total financial commitments undertaken in PPP contracts to a maximum of 1 percent of annual net revenue. If subnational governments exceed this limit, the federal government is authorized to withhold voluntary transfers. The law also limits financing for PPP projects that can be provided by the national development bank (BNDES) and public pension funds. Accounting rules for PPPs are being defined, including the valuation of guarantees and their treatment in relation to compliance with the 1 percent of net revenue limit.
The law contains provisions to minimize the exposure of private partners to institutional risk and reduce the implicit financial cost in PPP contracts. The key provision is the creation of a guarantee fund made up of government assets (e.g., equity shares of public enterprises, real estate, and budgetary contributions). The guarantee fund will be managed by a public commercial bank with an initial endowment of R$6 billion (about US$2.7 billion). The law also allows the earmarking of revenue to meet PPP contract payments.
The federal PPP program will be managed by a council (Conselho Gestor) formed by the ministers of finance and of planning, and the president’s chief of staff. The council will be in charge of establishing the criteria to select projects and designing the contracts. The authority to tender PPP projects resides in the council. A PPP unit has been established at the Ministry of Planning to provide support to the Conselho Gestor, and a working group at the National Treasury is working on macroeconomic issues related to PPPs, including the accounting.1 Law Instituting General Rules on Public–Private Partnerships Within the Realm of the Public Administration. Several states have approved their own PPP laws, which need to be compatible with the federal law. Under the Concessions Law, the private sector can build and operate public infrastructure, but the government cannot make payments to a concessionaire. Under the Procurement Law, the private sector is supplier to the public sector, but it cannot charge user fees, and contracts can have a maximum duration of five years.
PPPs require the development of expertise in the government across the full range of skills required to manage a PPP program. One common complaint about PPPs from the private sector is that bidding and contracting take much longer than in the private sector. Thus one of the functions of Partnerships UK, a joint private sector–government agency in the United Kingdom, is to promote PFI projects among government departments by providing financial, legal, and technical advice and assistance to support contract negotiations and procurement. The PPP Unit of the National Treasury of South Africa also provides detailed guidance and technical assistance related to assessing the feasibility and management of PPPs.26 In both of these agencies, however, the focus is on facilitating new PPP projects, even though managing a large stock of ongoing projects represents an equal or more demanding challenge. Particular attention also needs to be paid to developing PPP–related skills within subnational government agencies, because in many countries responsibility for spending in areas that are likely candidates for PPPs is devolved to them.
Effective Project Appraisal and Selection
Governments also must refine their project appraisal and selection processes. First and foremost, a decision to undertake a project, and the choice between traditional public investment and government supply of services or a PPP to implement it, should be based on technically sound cost-benefit comparisons. It is particularly important to avoid a possible bias in favor of PPPs simply because they involve private finance and, in some cases, generate a revenue stream for the government.27 The decision about whether a worthwhile project should be undertaken by the government directly or through a PPP should be informed by a public sector comparator indicating the cost of public provision. This should be used as a benchmark for determining whether the best private sector bid for a PPP contract—which will reflect the efficiency gains from private provision, higher private sector borrowing costs, and the costs to be borne by the government under the PPP—offers better value for money (VFM) for the government. The use of public sector comparators is the norm in advanced economies with considerable experience with PPPs, and Chile is making increased use of them to ensure that PPP projects offer good VFM.
E. Risk Transfer, Leasing, and Ownership
Risk transfer from the government to the private sector has a significant influence on whether a PPP is a more efficient and cost-effective alternative to public investment and government provision of services. This is clearly something the government should consider in deciding whether to embark upon a PPP and in negotiating the terms of a PPP contract. It should also be a focus of those seeking to assess whether a PPP will indeed yield the benefits claimed of it, and in particular whether it is being put forward mainly to move public investment off budget. Risk transfer is also relevant to determining the proper accounting and reporting treatment of PPPs, and indeed the discussion of risk transfer that follows draws in part on international accounting standards. However, risk transfer is an important topic in itself, which will be discussed before accounting and reporting issues are addressed.
Assessing Risk Transfer and Ownership
The private operator is typically the legal owner of a PPP asset for the period of the operating contract. However, if the government bears the risks (and derives the rewards) that are normally associated with ownership, it is in effect the economic owner of the asset. When this is the case, PPP investment is largely indistinguishable from traditional public investment, except that the payment profile for the government is different. Specifically, instead of the government making an upfront payment to cover the cost of building an asset, the private sector partner bears this cost and the government covers the opportunity cost of capital as part of its service payment to the private sector. This is how PPPs can be used to record lower government borrowing and debt than with traditional public investment.
In general, there are different risks entailed in owning an asset and in operating it. When the PPP contract distinguishes between the rights and obligations of the private partner in its capacity as the asset’s owner, as distinct from being its operator, risk transfer can be assessed by reference to these rights and obligations.
Private sector accounting standards provide guidance on how to assess risk transfer for leases. A standard lease contract is between the owner of an asset (the lessor) and the user of an asset (the lessee). An operating lease is similar to a rental arrangement in that a payment is made by the lessee to use an asset, and the lessor bears the risks related to ownership. A financial lease is a form of borrowing by the lessee to obtain the asset, and the lessee bears these risks. Whether a lease is an operating or a financial lease depends on the substance of the transaction rather than on the form of the contract. Factors that should influence decisions in this context are discussed in a number of private sector accounting standards for leases, such as those issued by the International Accounting Standards Board (IASB) and the Financial Accounting Standards Board (FASB) in the United States. The factors relevant to lease classification included in the relevant IASB standard are summarized in Box 4.
Box 4.Factors Determining the Substance of a Lease1
According to the IASB, the following factors would normally lead to a lease being classified as a financial lease:
- The lease transfers ownership of the asset to the lessee by the end of the lease term.
- The lessee has the option to purchase the asset at a price that is expected to be sufficiently lower than the fair value at the date the option becomes exercisable such that, at the inception of the lease, it is reasonably certain that the option will be exercised.
- The lease term is for the major part of the economic life of the asset even if the title is not transferred.
- At the inception of the lease, the present value of the minimum lease payments approximates the fair value of the leased asset.
- Leased assets are of a specialized nature such that only the lessee can use them without major modifications being made.
Individually or in combination, the following factors could also lead to a lease being classified as a financial lease:
- The lessee can cancel the lease and the lessor’s losses associated with the cancellation are borne by the lessee.
- Gains or losses from the fluctuation in the fair value of the residual fall to the lessee (for example, in the form of a rent rebate equaling most of the sales proceeds at the end of the lease).
- The lessee has the ability to continue the lease for a secondary period at a rent that is substantially lower than market rent.
PPPs can be specifically set up as operating leases, but it is unusual for them to take the form of financial leases. Financial leases tend to be used by governments to obtain major items of capital equipment such as airplanes, rather than to build infrastructure. Indeed, with a DBFO or similar scheme, the PPP asset is built and legally owned by the private operator, and this arrangement cannot on the face of it be a financial lease. However, an examination of the substance of a PPP transaction may indicate that the government, rather than the private owner, actually bears most of the risks associated with ownership. When this is the case, the view can be taken that the asset is in effect being acquired by the government through a financial lease and that the government is the economic, as distinct from legal, owner of the PPP asset.
Some criteria have been devised to assess the degree of risk transfer involved in PPPs. To a large extent, these derive from the private sector approach to classifying leases; indeed, the International Federation of Accountants (IFAC) has issued a standard for the public sector on leases that is closely related to the IASB standard for the private sector.28 However, IFAC acknowledges that the public sector may enter into a variety of arrangements for the provision of goods and services involving the use of dedicated assets for which it is unclear whether a financial lease is involved. Some national standards include quantitative criteria to establish the existence of a financial lease. For instance, the state of Victoria in Australia focuses on three criteria to determine whether a Partnerships Victoria PPP contract should be classified as a financial lease: 1) Does the government finance 90 percent or more of asset costs? 2) Does the service contract cover 75 percent or more of the useful life of the asset? and 3) Does the contract include a “bargain basement provision” whereby the government can purchase the asset at the end of the contract for substantially less than its residual value?29
Where PPP contracts do not provide a basis for applying the criteria for establishing the distribution of risks associated with ownership, the extent of risk transfer can be assessed by reference to the overall risk characteristics of the PPP. This is done in the United Kingdom, where the specific aim is to determine whether the government or the private operator “has an asset in a PFI property.” For nonseparable contracts (i.e., those where ownership and service elements of the contract cannot be distinguished), which are the norm, the U.K. approach is based, first and foremost, on the balance of demand risk and residual value risk borne by the government and the private operator. Demand risk, which is an operating risk and is the dominant consideration, is borne by the government if service payments to a private operator are independent of future need for the service. Residual value risk, which is an ownership risk, is borne by the government if a PFI asset is transferred to the government for more or less than its true residual value.30 Reference can also be made to various qualitative indicators, including government guarantees of private sector liabilities, and the extent of government influence over asset design and operation. The final conclusion is a professional judgment based on all relevant factors.
The Statistical Office of the European Communities (Eurostat) also provides guidance on the classification of PPP assets based on risk transfer. To this end, Eurostat recently issued a decision that a private partner will be assumed to bear the balance of PPP risk if it takes on most construction risk and either most availability risk—which is also an operating risk and relates to the continuity of service supply—or most demand risk. Further detail is provided in Box 5. While focusing on a few key risk categories for the purpose of assessing risk transfer is understandable, the Eurostat decision is problematic.31 Because the private sector typically bears most construction risk and availability risk, the decision is likely to result in the majority of PPP assets being classified as private sector assets, even if the government bears most demand risk. This being the case, the decision appears to be more liberal than Eurostat itself has been in practice. Thus, in the case of Ireland, Eurostat indicated that early PPP projects involved insufficient risk transfer and that investment in these projects would be classified as public investment. Subsequently, all PPP projects in Ireland were treated in this way. A concern is that the decision could open the door to PPPs that are intended mainly to circumvent the fiscal rules of the euro area’s Stability and Growth Pact (SGP).
Assessing risk transfer is likely to remain a difficult exercise. Certainly, it is essential that there be full disclosure of the relevant terms of original and renegotiated PPP contracts, and contract simplification and standardization would also help. However, the legal complexity of PPP contracts means that they will always be hard to interpret, and this will complicate assessments of risk transfer even when the focus is on a few key risks. Moreover, PPP contracts may not tell the whole story, because they only cover ex ante risk transfer. Political pressure for the government to bail out both large projects (those that are too big to fail) and providers of essential services may mean that the government in fact bears more risk than PPP contracts suggest.
Box 5.Eurostat Decision on the Treatment of PPPs1
The Eurostat decision covers long-term contracts in areas where the private partner builds an asset and delivers services mainly to the government.
Eurostat recommends that assets built by public-private partnerships be classified as nongovernmental assets and therefore recorded off the balance sheet for government, if both of the following conditions are met: 1) the private partner bears the construction risk, and 2) the private partner bears one of either availability or demand risk. An accompanying opinion of the Committee on Monetary, Financial and Balance of Payments Statistics indicates that these conditions refer to the private partner bearing “most of the risk” concerned.
Construction risk covers events such as late delivery, low standards, additional costs, technical deficiency, and external negative effects. If the government makes payments to the private partner irrespective of the state of the asset, this indicates that the government bears most of the construction risk.
Availability risk relates to the ability of the private partner to deliver the agreed volume and quality of service. Government payments to the private partner that are independent of service delivery indicate that the government bears most of the delivery risk.
Demand risk covers the impact of the business cycle, market trends, competition, and technological progress on the continued need for the service. Government payments to the private partner that are independent of demand indicate that the government bears most of the demand risk. Changes in demand due to changes in government policy are excluded.
It is the responsibility of national statistical offices to implement the Eurostat decision, based on information that is judged to be easily obtained from PPP contracts. However, when a clear classification is difficult to make, other contract provisions can be taken into account. In particular, if the government has an obligation to buy the asset at the end of the contract at a predetermined price, this would indicate that the government bears most PPP risk when other considerations are unclear.1 Based on Eurostat (2004).
F. Fiscal Accounting and Reporting for PPFs
There is not a comprehensive fiscal accounting and reporting standard specifically for PPPs. The accounting profession is taking steps to develop an internationally accepted standard, but the likely features are not yet clear.32 The absence of such a standard makes it difficult to close loopholes that enable PPPs to be used to bypass expenditure controls, to move public investment off budget and debt off the government balance sheet, or to hide the potentially high costs of using guarantees to secure private financing. An internationally accepted accounting and reporting standard could promote transparency about the fiscal consequences of PPPs, and in the process ensure that increased efficiency, rather than a desire to meet fiscal targets, is their main motivation. In any event, as PPPs become more commonplace, market analysts and rating agencies are developing the expertise to assess the fiscal risks involved and, in particular, the implications for debt sustainability of future commitments under PPPs and contingent liabilities. Thus any misuse of PPPs is unlikely to escape market scrutiny for long.
Existing standards provide a starting point to address the accounting and reporting treatment of PPPs. The 1993 System of National Accounts (1993 SNA) and the 1995 European System of National and Regional Accounts (ESA 95) cover some operations that characterize PPPs, including leases, while ESA 95, supplemented by the ESA 95 Manual on Government Deficit and Debt, covers public infrastructure built and operated by the private sector.33, 34 The fiscal reporting framework in the Government Finance Statistics Manual 2001 (GFSM 2001)35—which integrates flows and stocks and shifts the emphasis toward accrual reporting and balance sheets—is also well-suited to reporting on PPPs, although it does not currently provide comprehensive coverage of such operations. For a description of the GFSM 2001 analytical framework, see Appendix 2.
Current Treatment of PPPs
The recording of the following PPP operations is covered by existing accounting and reporting standards and is fairly straightforward: operating contracts, concessions and operating leases, financial leases, and the transfer of PPP assets to the government.36 This treatment is described below following the GFSM 2001 fiscal reporting framework.
- Operating contracts: Payments under operating contracts to private sector partners for services provided to the government are recorded in the government operating statement as an expense.37
- Concessions and operating leases: Concession fees and other payments by concessionaires to the government (e.g., profit shares) are recorded in the operating statement as revenue.38 When the government leases an asset it owns to a private operator, lease payments to the government by a private operator are also recorded as revenue.39
- Financial leases: The acquisition of a nonfinancial asset under a financial lease is recorded in the operating statement, together with incurrence of a lease liability to the private sector. The asset and liability are also recorded on the government balance sheet. Subsequent depreciation of the asset, and interest and amortization payments on the lease, are then recorded in the operating statement. As the lease liability is reduced, the PPP net asset value builds up on the balance sheet.40 When the lease concludes, the asset is recorded on the government balance sheet at its residual value.41
- Transfer of PPP assets to government: If there is provision for a PPP asset to be transferred at zero cost to the government, the asset transfer is recorded in the operating statement as the acquisition of a nonfinancial asset at its residual value, balanced by a capital transfer from the private owner. Any purchase price involved is an expense, and the capital transfer is reduced by the corresponding amount.42 The asset is also recorded on the balance sheet at its residual value when the transfer takes place, and subsequent depreciation of the asset is recorded in the operating statement.
Many countries are still working with the cash-based predecessor of GFSM 2001, A Manual on Government Finance Statistics 1986 (GFSM 1986).43 Under this framework, which is the basis of traditional fiscal accounts, only cash flows are recorded. However, with the exception of depreciation, other noncash transactions could be recorded in adjusted cash accounts (see Diamond, 2006). Since balance sheets are not part of GFSM 1986, PPP assets are not recorded as such, but the liability under a financial lease is recorded as government debt.
Accounting for Risk Transfer
When PPP projects involve limited risk transfer to the private sector, the practice of Eurostat and a number of countries is to classify PPP assets as government assets. This is done with a view to recognizing that the government plays a role in the economy and conducts fiscal policy through PPPs. For accounting purposes, Eurostat considers PPP investments that expose the government to significant risk to be public investment, while the state of Victoria in Australia and the United Kingdom consider them to involve acquisition by the government of the PPP asset through a financial lease.44 These two approaches—which are formally the same—raise some technical issues that are of concern to the accounting profession (discussed in Appendix 3).
More important, however, is the question of whether this binary approach, under which PPP assets are classified either as government assets or as private assets, is an appropriate way of accounting for risk transfer. The specific concern is that such an approach is insensitive to the fact that PPPs are intended to share risk according to which party can best manage it. The fact is that government exposure to PPP risk will vary widely across projects, and the accounting profession ideally should be seeking to develop a workable approach to identifying and quantifying the risk to which the government is exposed under PPPs and for assessing and disclosing the fiscal consequences of such risk. While this is a difficult task, Chapter 2 of this paper illustrates how this can be done for guarantees, which are the principal source of explicit risk for the government associated with PPPs.
Nevertheless, accounting bodies seem more likely to refine the current binary approach—probably by shifting the focus from ownership to control as the principal basis for establishing whether PPPs create government or private sector assets—and less likely to develop a new approach that is sensitive to the degree of risk transfer.45 If this is the case, there is a risk either that PPPs will be discouraged in cases where the private sector is prepared to bear significant but not the larger share of project risk or, more seriously, that governments will be tempted to tailor PPPs to meet the requirements for classification as private investment by trading off higher project costs for increased risk transfer to the private sector. This would defeat the objective of using PPPs for efficiency gains and disguise the medium- to long-term fiscal implications of many PPPs. To minimize these problems, it is important that governments disclose comprehensive information about PPPs, including their known and potential future costs.
Box 6.Detailed Disclosure Requirements for PPPs
For each PPP project or group of similar projects, government budget documents and year-end financial statements should provide information on the following:
- Future service payments and receipts (such as concession and operating lease fees) by government specified in PPP contracts for the following 20—30 years.
- Details of contract provisions that give rise to contingent payments or receipts (e.g., guarantees, shadow tolls, profit-sharing arrangements, events triggering contract renegotiation), with the latter valued to the extent feasible.
- Amount and terms of financing and other support for PPPs provided through government on-lending or via public financial institutions and other entities (such as SPVs) owned or controlled by government.
- How the project affects the reported fiscal balance and public debt, and whether PPP assets are recognized as assets on the government balance sheet. It should also be noted whether PPP assets are recognized as assets on the balance sheet of any SPV or the private sector partner.1
Disclosure Requirements for PPPs
Government budget documents and year-end financial statements should include an outline of the objectives of current or planned PPP programs and a summary description of projects that have been contracted or are at an advanced stage in the contracting process (their nature, the private partner or partners, and capital value). They should also disclose the type of detailed information specified in Box 6. In countries with sizable PPP programs, disclosure could be in the form of a separate Statement on PPPs. Within-year fiscal reports should indicate major new contracts that have a significant short-term fiscal impact. PPP contracts, or summaries of their key features (preferably in standardized format), should also be made publicly available. More detailed disclosure requirements for guarantees are suggested in Chapter 2.