You are on page 1of 24

Critical Perspectives on Accounting 19 (2008) 963986

An example of creative accounting in public sector: The private nancing of infrastructures in Spain
Bernardino Benito a, , Vicente Montesinos b,1 , Francisco Bastida c,2
Department of Accounting, Faculty of Economics and Business, University of Murcia-Campus of Espinardo, 30100 Espinardo (Murcia), Spain b Department of Accounting, Faculty of Economics, University of Valencia, Avga. Tarongers s/n, 46022 Valencia, Spain Department of Accounting, Faculty of Economics and Business, Technical University of Cartagena, Alfonso XIII, 50, 30203 Cartagena, Spain Received 11 January 2007; received in revised form 25 July 2007; accepted 30 August 2007
a

Abstract This paper analyses some proposals for private nancing of public works having emerged in Spain in recent years. We show that all the new nancing methods assessed are incorrectly named as private, for the payments are nally made by the Government by means of its budgetary resources. A deferral of accounting and budgetary recognition of these transactions, together with a false disclosure in nancial statements of the debt connected with the projects, are the main reporting consequences of the new funding methods. In short, it is a clear example of creative accounting with the aim of meeting the convergence criteria imposed by the European Union. 2007 Elsevier Ltd. All rights reserved.
Keywords: Public sector; Creative accounting; Infrastructures; PFI

The analysis and conclusions of this paper are part of a broader study, funded by the Spanish Fiscal Studies Institute. Corresponding author. Tel.: +34 968 363812; fax: +34 968 363818. E-mail addresses: benitobl@um.es (B. Benito), vicente.montesinos@uv.es (V. Montesinos), fco.bastida@upct.es (F. Bastida). 1 Tel.: +34 963 828294; fax: +34 963 828287. 2 Tel.: +34 968 325740; fax: +34 968 325782. 1045-2354/$ see front matter 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.cpa.2007.08.002

964

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

1. Introduction Most of infrastructures have been nanced in Spain through the public budget until the middle of the nineties. Concessions and private nancing have been scarcely used. However, private nancing of new infrastructures (private nance initiative, PFI, or more recently named also as publicprivate partnership, PPP) is growing in recent years. These are the main reasons for reducing the use of budgetary resources: European Monetary Union (EMU) requirements concerning government decit and debt limit the amounts that public entities can borrow to nance capital assets. According to the 1996 Stability and Growth Agreement, European Governments must reach a balanced budget in 2001, namely, a zero decit. This target must be achieved in spite of the social pressure for more and best public services and benets, making almost impossible for the public sector to reduce current expenses and thus generate savings. Important investment funds are searching in the nancial markets for safe and steady investments. Rigidity of traditional Public Administration brings about high management inefciencies. Financial earnings are normally improved by companies exibility and management skills. Investment priorities, as well as verication of public works protability and feasibility can be enhanced through the private business requirements if enough nancial rate of return is achieved. Furthermore, the infrastructures still remain in Spain quite below the most advanced countries equipments in the European Union (EU). It is generally known that endowment for infrastructures is one of the most important conditions for economic growth in any country. For example, the 20052020 Infrastructures and Transports Strategic Plan of the Spanish Ministry of Development highlights that, according to recent studies, the cumulative marginal productivity of the public capital is almost of 1.5, i.e., an increase of D 1 in public capital investment leads, in the long term, to a GDP increase of almost D 1.5. All the arguments set out here lead to the conclusion that lots of imagination are required when searching for new methods to fund public investments. On the one hand, a high quality in public services must be accomplished, and on the other hand, a rigid budgetary discipline has to be achieved. Accordingly, new funding methods have appeared in which private partners are involved in the development of public investment projects (PFI). Our main hypothesis is that many countries have used PFIs to defer payment and this way control their decits and debt without cutting investments in infrastructures and public services. Lack of a clear accounting standards on how to report PFIs has allowed governments to do it. The following pages analyse some proposals for private nancing of public works having emerged in Spain in recent years. After assessing all the new nancing methods, we nd that they are incorrectly named as private, for the payments are nally made by the Government by means of its budgetary resources. A deferral of accounting and budgetary recognition of these transactions and a wrong disclosure in nancial statements of the debt connected with the projects are the main reporting consequences of the new funding methods. In short, it is

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

965

a clear example of creative accounting with the aim of meeting the convergence criteria imposed by the EU. The remaining of the paper is organized as follows. Section 2 presents the theoretical background. Section 3 describes the main PPP contracts characteristics. Section 4 shows the opportunities for using new nancing methods of public works in Spain. In Section 5 we analyse different forms of private nancing of infrastructures and their implications for government decit and debt. Section 6 summarizes and presents conclusions.

2. Theoretical framework The choice between budget and PFI when nancing infrastructures has to do with the theory of property rights, which states that ownership matters when contracts are incomplete, thus preventing parties from stipulating all the potential eventualities in the contract (see Hart, 1996 for an insight on this theory). Therefore, a key aspect in PFI is the appropriate risk sharing between private and public sectors established in the contract. If adequate risk is not shifted to the private sector, then projects become quasi public, but with the funding removed from the governments balance sheet. A critical issue in the risk transfer arrangements is to achieve pricing that correctly reects the risks assumed by each party to the transaction (Brown, 2005). Hart (2003) developed a theoretical model of PPPs within the framework of the property rights theory. He concluded that conventional provision is good if the quality of the building can be well specied, whereas the quality of the service cannot be. In contrast, PPP is good if the quality of the service can be well specied in the initial contract (or, more generally, there are good performance measures which can be used to reward or penalise the service provider), whereas the quality of the building cannot be. The principal rationale for the adoption of PPP is that it addresses the decit of physical infrastructures of the countries. Since Public investment is constrained by the limits on public spending imposed by membership of the single EU currency, PPP becomes an attractive way of funding investments. PPPs involve contracting between government and the private sector under conditions of imperfect information. Theoretical developments such as principalagent theory provide an insight on PPPs. This theory focuses on the design of optimal contracts in the face of asymmetries in the information and objectives of contracting parties. Emphasis is placed on the optimal allocation of risk as a means of incentivising agents to achieve principals (government) objectives. PPP contracts specify which risks are borne by the government (principal) or contractor (agent). Moreover, as PPP contracts can connect different elements of infrastructure projects (for example, link the design and construction with one or all of the nance, operation and maintenance elements) there is better scope for transferring risk compared to traditional procurement methods. For example, payment may be withheld until assets are in operation thereby encouraging contractors to complete construction on time and within budget (Hurst and Reeves, 2004). PPPs can provide higher value for money compared with other approaches, if there is an effective implementation structure and if the objectives of all parties can be met within the contract. We must bear in mind that such contracts are complex to design, implement and

966

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

manage. Sometimes they are not the best option: indeed in some cases it has been reported that they lead to an increase in the costs of services to citizens. From a theoretical viewpoint, the main justication for PPP is the possibility to exploit the management expertise and the efciency of the private sector without giving up quality standards of outputs, thanks to appropriate control mechanisms from the public party. This result is achieved by setting up complex contractual arrangements with the private sector operator (agent) where the public sector acts as the principal. In principalagent relationships, the most complex issues are the precise denition of the tasks assigned to the agent, the measurement of the agents performance, and the extent to which the principal can control the agents performance for the whole duration of the contractual relationship. In PPPs, the cornerstone is the allocation of risk between the two parties: well-designed PPPs redistribute the risk to the party that is the superior insurer or the least cost avoider, i.e., the party best suited to control and/or bear the risk (European Parliament, 2006). The Public Choice theory argues that taxpayers perceive debt-funded projects less costly than tax-nanced ones, since they fail to perceive the actual cost of debt (Buchanan, 1967). Taxpayers do not properly evaluate inter-temporal government budget restrictions. When government offers decit-nanced outlay program, taxpayers overestimate benets of current expenditures and underestimate future tax burden. PFI contracts allow incumbents to invest in new infrastructures, which have a positive impact on voters opinion about them, while deferring the payments of the infrastructure.

3. Public works funding through publicprivate partnership In order to ll the growing gap between needed infrastructures and available resources, a key question arises: Which is the best way to make the investment in terms of value for money? In this context, since the nineties, PPPs have spread in Europe in general, and in Spain in particular. However, compared to PPPs, traditional contracts still are more widely used, and they may be more appropriate in many projects. Even in the UK, where PPP is frequently used, traditional contracts nance 85% of public investment (PriceWaterhouseCoopers-PWC, 2005). Thus, it is important to analyse the traditional model of public contracting as opposed to PPP. The main characteristics of the former would be The public sector buys assets, not services, to the private sector. Assets are ex ante perfectly specied: the public sector designs the asset before the contracting process starts. The private sector only takes responsibility for the asset construction. The private sector takes no responsibility for the assets long-run yield after the guarantee period. The public sector is directly involved in the management of the project contract. If there are several companies singing the contract, the public sector usually takes the responsibility of coordinating them. In spite of its common use, there is neither a generally accepted denition nor a unique model of PPP. It encompasses several relationship structures in which the private sector provides an asset or a service to the public sector. Utilities and transport projects based in

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

967

administrative concessions have been used since years ago in the EU, especially in France, Italy and Spain. In this type of contracts, the company receives the payments made by service users, as it is the case, for example, in highway tolls. PFIs, in the UK, broadened this concept to a wider scope of public infrastructures, and combined it with the introduction of services not paid by nal users, but by the public sector. PFIs include several types of contract, such as Short run management contracts with zero or small capital expenditure. Administrative concession contracts. These may involve the funding, design and construction of assets, with an important investment, together with the providing of related services. Joint ventures and partial privatizations, with a shared ownership between the public and the private sector (see in this respect Bult-Spiering and Dewulf, 2006; PWC, 2005). The main difference between a PPP and a traditional contract is that in a PPP, private holder earnings are connected to service outcomes and to asset yield during the contractual period. The private holder is responsible not only for delivering the asset, but also for the project management and for the appropriate provision of the service during the contractual period. The timing of payments received by the company is completely different from traditional contracts. The European Commission (2004) highlights the following PPP characteristics: The relatively long duration of the relationship, involving cooperation between the public and the private partner on the project. The method of funding the project, partly from the private sector, sometimes by means of complex arrangements between the various players. However, public funds in some cases rather substantial may complement the private funds. The key role of the economic operator, who is involved in different stages in the project (design, completion, implementation, funding). The public partner focuses primarily on dening the objectives to be attained in terms of public interest, quality of services and pricing, and it takes responsibility for ensuring compliance with these objectives. The distribution of risks between the public partner and the private partner, to whom the risks generally borne by the public sector are transferred. However, a PPP does not necessarily mean that the private partner assumes all the risks, or even the major share of the risks linked to the project. The precise distribution of risk is determined case by case. Besides, the European Investment Bank (2005) identies these aspects in PPP: It is initiated by the public sector. Involves a clearly dened project and the sharing of risks with the private sector. Is based on a contractual relationship which is limited in time. Has a clear separation between the public sector and the borrower: there must be a private partner providing nancial resources to fund the project (PWC, 2005).

PPPs introduce new elements in the relationships between public and private sector. The nancial reports must disclose the assets/rights and liabilities/commitments that arise in the provision of services and infrastructures under PPP. PPPs, as opposed to traditional

968

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

contracts or privatization, imply a risk sharing between public and private partners, in order to achieve a more efcient use of resources. Risk sharing and public resources involved determine the responsibility of the parties concerned in the project. An accounting discussion has started stemming from what we have just said. The issue at stake is the way PPPs should be reported in the public sector nancial statements. The concrete PPP where the problem arises is the one in which the private holder is responsible for designing, building or reforming and nancing an infrastructure. The contract enables the public partner to have the infrastructure at the citizens disposal, whereas the private partner commits to provide the infrastructure and is entitled to be paid by the public partner. When these new nancing methods emerged, there was no clear accounting standard providing guidance on how to report them. Thus, many countries used them to defer payment and this way control their decits and debt without cutting investments in infrastructures and public services. In this sense, Milesi-Ferretti (2004, p. 378) says that the imposition of numerical rules may encourage the use of dubious accounting practices, thereby reducing the degree of transparency in the government budget. These concerns have gained strength with the use of creative public nance by a number of European countries in order to facilitate meeting the budget decit ceiling established in the Maastricht Treaty. In the same way, Koen and van den Noord (2005, p. 7) state that creative accounting operations may have merits of their own. PPPs for instance have proliferated in several EU countries since the late 1990s, either at the national or sub-national level (e.g., in the form of PFI contracts in the UK and concessions in Spain). Instead of the government buying an asset and operating it, a private entity invests and owns the asset (at least partly and at least during the period of exploitation), selling the corresponding services to the government. PPPs may be justied on efciency grounds, but from the perspective adopted here their main feature is that they initially reduce the general government decit and debt for a given level of investment in publicly used infrastructure. In 1999, the Irish Government launched a programme of PPPs, in order to address the countrys acute decit of physical infrastructure. The rst PPP to reach the stage of operation was the contract for ve secondary schools. The evidence indicated that this PPP has not resulted in signicant innovations, and the public sector has failed to provide any evidence of value for money (Hurst and Reeves, 2004). Within the EU, Spain has been one of the countries taking advantage of the lack of clear accounting standard for new nancing methods for infrastructures. However, as we will see later, the last Eurostat guidance has banished all doubts, and accordingly, almost all these new methods must be included in order to compute the public decit and debt.

4. Opportunities for using new nancing methods of public works Soon after the Spanish Conservative Government took over in March 1996, the Minister of Public Works considered the possibility of nancing infrastructures based upon the use of funds coming from privatizations and other methods. The aim was to avoid an increase of government decit and debt rates. Besides, several circumstances did exist for resorting to private nancing: Plenty of available private savings.

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

969

Private business licensees had at that time remarkable earnings and capitalization rates. Therefore, they had nancial resources available for reinvesting. Considering the reduction of government investment gross xed capital formation building rms looked to be inclined to share direct infrastructure investments, in order to enhance certain bidding amounts in public works. A policy of monetary stability led to reductions of interest rates and thus private investment funds started to be interested in long-term investments with satisfactory prot margins. All these advantageous circumstances made the Government issue several rules to boost government investment. Thus, in December 1996 two acts were passed with the following decisions regarding private nancing of public works: Amendment of some provisions of Governments Contracts Act, setting up the so-called German or key on hand method (also called turnkey) for contracting investments, according to which, government investments are not paid for until they are completely nished. The regulation of this contracting method was later developed by a Royal Decree in May 1997. The next section will analyse in detail this type of contract. Amendment of some provisions of the General Budgetary Act, in order to allow longer terms for payments and to ease the use of revenues from public enterprises. Changes in the 1972 rules for building, maintenance and operation of granted highways: Extension of concession term limit from 50 to 75 years. Enlargement of the scope of the concession for licencees. Road maintenance and operation are no longer the only allowed activities for these enterprises, but also other could be carried out, such as service areas at the highways. Securities could be issued as toll rights supporting documents. Further provisions for constitution and operations of the funds corresponding to these securities have been developed by a 1998 Royal Decree. State-controlled rms entitled to borrow money can be established with budgetary resources. This variety has been known as Spanish method. We will be dealing with this in more detail later on. A public entity named Railway Infrastructures Management Entity (Gestor de Infraestructuras Ferroviarias, GIF)3 was created with the main objective of building and managing new railway infrastructures. This entity is a commercial corporation, not considered as Public Administration, since its revenues come from a fee the operators4 pay for using the infrastructures. These entities will not be taken into account when calculating the public sector decit and net lending/borrowing, according to European System of Accounts 1995 (ESA 95). This will be explained below.

3 Established in 1997, has disappeared and joined the Railway Infrastructures Managing Administration (Administrador de Infraestructura Ferroviaria, ADIF), which started working on 1 January 2005. 4 A 1998 Royal Decree regulated the railway infrastructures operation for international transport services, railways corporations licences and rates for the use of railways infrastructures. All theses rules were adopted according art. no. 104 of the 1997 Spanish Act and the European Directives 95/18/CE, regulating the railways corporations licences and 95/19/CE, regulating the concession of railways infrastructure capacities and the fees to be paid by its use.

970

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

The next section analyses these proposals, particularly: the full price payment contract (also known as German method or Turnkey); the establishment of enterprises or public entities building works on behalf of either the Administration or themselves (system known as Spanish method) and the shadow toll (nancing method used by some Spanish Regions). In spite of these methods being considered PPP or PFI, as we said in the introduction, all of them are incorrectly categorized as private, for the payments are nally made by the Government through its budgetary resources. A deferral of accounting and budgetary recognition of these transactions and a wrong disclosure in nancial statements of the debt connected with the projects are the main reporting consequences of these new funding methods. In short, it is a clear example of creative accounting with the aim of meeting the convergence criteria imposed by the EU. 5. Different forms of private nancing of infrastructures and their implications for government decit and debt First, we must make clear that no standards have been developed in Spain for PFI accounting treatment. Accordingly, we will focus on the EU regulatory framework, more specically, on the ESA 95 Manual on government decit and debt issued by Eurostat. We believe the analysis of these EU rules is of great interest, because they are legally binding in all member countries of the Eurozone, and almost all of them have considered the nancing formulas explained in this manuscript in a larger or smaller extent (see at this respect PWC, 2005), with the purpose of fullling the Maastricht Treaty requirements for public decit and debt limits. Fig. 1 and Table 1 show how PFI use and high government decit have been closely related in many EU countries. Thus, the higher the government decit, the larger the use of PFI. This relationship makes clear, as stated above, that in most cases the aim of using PPP or PFI has been the deferral of the payment and debt. In summary, we could label it as creative accounting.

Fig. 1. Average 20002005 PFI activity as a percentage of mean GDP. Source: PricewaterhouseCoopers (2005).

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986 Table 1 General government decit ()/surplus (+) as a percentage of GDP

971

Source: Eurostat. Shaded countries are those with the highest PFI activity (see Fig. 1).

The ESA 95 Manual is the result of European Commissions concern about the need to ensure the correct application of ESA 95 criteria to elaborate statistics and government nance accounts. PFI are among the operations with unclear treatment. This document supersedes other Eurostat regulations, and it is the result of some countries consultations to Eurostat regarding the accounting treatment of some PFI operations, which were being reported according to ESA 79. Usually, the answers to the consultations were published as News Releases, which we will refer to in the following pages (Table 2 summarizes them). 5.1. Full price payment when public works are accepted (German method) In this case, the Government signs a contract with the successful bidder for building and nancing the project. Building costs and interests are paid when the work is nished and accepted by the Government. Therefore, the Government does not pay any money until the project is nished (there are no partial payments). The contractor must nance the building, i.e., paying costs in advance while the building project is in progress, until the work is nally

972

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

Table 2 Main news releases of the Eurostat in relation with the treatment of PFI/PPP News Release no. 16/97 of 21 February 1997, Treatment of Public Finance Initiatives News Release no. 33/97 of 30 April 1997, Treatment of public nance initiatives News Release no. 15/2002 of 31 January 2002, Treatment of the transfer of Government real estate to a publicly owned corporation in Austria News Release no. 80/2002 of 3 July 2002, Securitisation operations undertaken by general government News Release no. 18/2004 of 11 February 2004, Treatment of publicprivate partnerships News Release no. 88/2007 of 25 June 2007, Securitisation operations undertaken by general government

accepted. This method has been widely used for road building in Germany (that is why it is known as German method) (Izquierdo and Vasssallo, p. 200). When the public work is nished, the Government will pay either in cash or through a maximum period of 10 years. It is possible for the contractor to convert these future payments into tolls to be paid by the infrastructure users. The following conditions have to be met to use this nancing method: Eligible projects are: road infrastructures, water and railway equipment and environmental and coastal installations. Total contracting price, nancial costs when deferring payments not included, must be over the following amounts: Roads, 24 million euros. Railway and water infrastructures, 18 million euros. Environmental and coastal installations, 6 million euros. This kind of contracts is not applicable to infrastructures improvements, repairs, conservation and demolition works. A specic contract will be signed for any individual work, within the abovementioned limits. Therefore, accumulation of works in a single contract is not permitted. Total annual amount committed using this contracting method cannot go beyond 30% of initial budgetary appropriations for investments in capital assets during the year. This German method has been used by the Spanish Central Administration. Although a specic regulation of this method for Municipal Governments was supposed to be issued more than 1 year ago, it has not been published so far. In the same way, some Spanish Regions planned to use this kind of contract, but no concrete regulation has yet been adopted.5 The main reason argued for using this nancing method is the possibility to defer accounting and budgetary recognition of investment expenses until the date the work is fully executed. In the same way, the recognition of the debt connected with these transactions is also deferred. Therefore, during the building period, neither budgetary expenses nor net borrowing increase, thus making easier for the Spanish Government to meet the
5 Public sector in Spain is organized in three levels: State or Central Administration, Regions (Autonomous Communities) and Local Entities. Out of the total government expenses in Spain, these three administrations spend approximately 50%, 35% and 15%, respectively.

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

973

convergence criteria (this is a patent example of creative accounting). That is to say, the Government argues that the infrastructure is not reported until it is nished and delivered. However, the bottom line is only the deferral of recognition of expenses and debt in the accounting books. This contract makes possible to do important public works, but at the expense of relevant commitments on future budgets and an increase of government indebtedness. According to the current EU accounting standard (ESA 95), in force since 1999, investment expenses must be reported along the years of construction (accrual accounting). The previous approach under ESA79 (cash accounting). reported the expenses according to the payments made (L der, 2000, pp. 118122). A consequence of the taking in consideration u of these circumstances has been the suspension of the use this nancing method during several years, according to additional dispositions of the laws passing the State General Budgets. 5.2. Constitution of public corporations and/or public entities As previously stated, the possibility to create public rms for road and water works building and operation was created by Law in 1996. This is the case of Railway Infrastructures Management Entity. This entity has its own legal status as a subsidiary of the Ministry of Public Works, it is able to hire and re its own staff and it holds its own assets and liabilities. The entitys main activities are building the high-speed railway infrastructures connecting MadridBarcelonaFrance, as well as the safety and control systems management. Revenues come from rates paid by users of operated lines and by the users of its own telecommunication networks. The entity can borrow funds by itself and the State is allowed to give it a nancial support. Similar corporations has been also established by the Ministry of Environment, for example, Ebro Basin Waters (Aguas de la Cuenca del Ebro, ACESA). Considering the Spanish Regions, the Catalan Government was the rst creating this kind of corporations, as in 1990 it established the rm Infrastructures Management Inc. (Gesti n de Infraestructuras S.A., GISA). An agreement denes the management and o nancial links between Catalan Government and GISA, which obtains revenues by the management responsibilities for building infrastructures. Under the guarantee of these revenues, the rm obtains funding in the nancial markets, without any specic governmental guarantee. When infrastructure construction is nished, it becomes a capital asset of the Catalan Government. Similar processes are going on in other Regions: Andalusia, CastillaLe n, Galicia, Castilla-La Mancha or Madrid. This kind of corporations is also usual in the o local public sector, but only in the largest municipalities. In addition to efcacy and efciency criteria, the main advantage of these entities has been that their borrowing is not consolidated when setting up public sector debt. However, in the new ESA 95 the borders for different sectors of economy are established according to a market criterion. Thus, if an entitys purpose is to produce goods or services for the market and it sells its production in the market, it must be included in the corporations sector, irrespective whether the owner is public or private (Jones, 2000). According to ESA 95 criteria, if sales are 50% or more out of the total operating expenses of an economic unit, the production of this unit is considered as market-oriented (paragraphs 3.33 and 3.37). In order to meet this criterion, amounts paid by the Government that contracted the

974

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

building of the infrastructure are not considered as sales. On the one hand, a consequence of ESA 95 criteria is the non-inclusion in the consolidated accounts of Public Administrations of market-oriented units, not necessarily with legal status. On the other hand, ESA 95 simultaneously leads to the inclusion of corporations, agencies or entities mainly operating on the base of ofcial nancial aids. This is the case, for instance, of the agent corporations we are analysing in this section. In the above-mentioned circumstances, new funding and management methods arise with the aim of establishing entities whose debt is not consolidated in public sector accounts. Special attention must be paid to the constitution of securities funds. A securitisation operation occurs when a government transfers ownership rights over nancial or non-nancial assets, or the right to receive specic future revenues, to another unit, named the securitisation unit, who in exchange pays the originator. In order to nance the purchase, the securitisation unit borrows on its own account by, typically, issuing bonds called asset backed securities (ABS). The securitisation unit uses income generated by the transferred asset or by the specic future ows, or by sales of the transferred assets, to service its debt. Usually the lenders will have a direct and legal claim on those assets or on those ows, in the event of the securitisation unit not paying the interest and principal due. When the proceeds obtained from the sale of the assets are higher than the initial price paid to government, and the securitisation contract includes, in addition to the initial payment by the securitisation unit, a clause on additional future payments to government, a deferred purchase price (DPP) is said to exist and all or part of the proceeds are allocated to government. Fig. 2 shows the basic features of the operating mechanism of this funding method. The key issue to be determined is whether a securitisation operation gives rise to revenue for the government, thereby reducing the public decit if there is one, or whether the proceeds should be considered as government borrowing. Eurostat has decided the following (News Release no. 88/2007 of 25 June 2007) (this decision complements and amends the decision taken on the same issue in News Release 80/2002 of 3 July 2002): 1. All securitisation of scal claims by government should be treated as government borrowing. 2. The existence of a DPP clause or of similar arrangements should lead to the classication of the securitisation operation as government borrowing. 3. A clause in the contract referring to the possibility of substitution of assets (except for marginal cases limited in scope and deriving purely from technical and material errors) should lead to the classication of the securitisation operation as government borrowing. 4. A clause of the securitisation contract stipulating ex ante government compensation to the unit (in case of government actions which are specically related to the unit and not to different economic units more generally) should lead to the classication of the securitisation operation as government borrowing. 5. When government compensates (for instance in the form of cash, of debt assumption, or of direct or indirect guarantee) the unit ex post for specic events, although compensation was not originally foreseen in the contract, a reclassication of the operation as government borrowing must occur, with an impact on the surplus/decit of government in the year in which the compensation is decided.

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

975

Fig. 2. Operating mechanism of securities funds. (1) Public Administration asks the public entity for infrastructure building. This request gives rise to rights in favour of this public entity. (2) Public entitys rights are handed over to a securities fund, which pays to the entity the present value of these rights. With this money, the public entity can afford to pay builders or nancial institutions for refunding previous borrowings. (3) Public Administration pays to the securities fund.

To sum up, when Spain has used this funding method, the aim was to hide public debt (as Koen and van den Noord, 2005 point out). Eurostat decision, however, concluded that it was creative accounting, and therefore the balance sheet should disclose the related debt. Something similar happened in other EU countries, such as Austria, Greece or Italy (see in this respect Milesi-Ferretti and Moriyama, 2006, p. 3288). 5.3. The shadow toll method Known also as DBFO (design, build, nance and operate), the infrastructure is built and operated by the concessionaire, who usually is a private rm. Government just pays the corresponding rates in order to get services provided using the assets constructed. This model is being applied for motorways, medical care in hospitals, elderlys homes, day nurseries, sport facilities, pools, waste treatment, water desalination, etc. The payment is made by means of periodic amounts of money that depend upon the use of the infrastructures by citizens. In order to avoid a low-use risk for the licensee rm, the Government guarantees the concessionaire enough revenues to achieve nancial balance. This kind of contracts is, in short, nearly pure concessions, with the only difference that it is not the user but the Government who pays the rates. The accounting treatment of this funding method for infrastructures resembles the operating lease. Sometimes, infrastruc-

976

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

tures ownership transfers to the Administration at the end of the operating period, and then the essence of the method is closer to a nance lease. Spanish Regions legislation also labels this funding method as demand levy, and the system is built upon a private sector involvement in the construction projects, on the base of foreseeable cash ows generation. Users do not pay any toll, but it is the Government who guarantees payments to private corporations according to previously xed rates. The main argument Spanish Regions make for the adoption of this funding method is to foster private involvement specially small and medium size rms in public works projects development. It is not hard to realize that the funding method based on the shadow toll is very similar to the nance lease. As in the cases of Germany and France,6 when the Government uses the nance lease, it cedes public property to a leasing corporation for some time. This corporation constructs and nances the work and when the building process is concluded, the construction is made over to the Government, which operates the installations as lessee, until the date of reversion. We think the Belgian experience on the shadow toll is worth commenting. In 1960 the Belgian Government decided to use this way to nance its road construction. The reason for using this method was to cope with the existing budgetary limitations and thus being able to avoid an increase in taxes or debt. As time went by, it was clear that the only real outcome of this policy was an increase of public nancial burden. Accordingly, in 1982, after more than 600 km of roads had been constructed (50% of total network), the Belgian State took up again the direct management of the highways program (Izquierdo, 1992, pp. 119120). In the United Kingdom, the 1991 Roads Act set up a new system for construction of toll roads by the private sector. This system operates through a contractual link with the State. On the one hand, each project must be self-nanced by means of the tolls paid by the users. On the other hand, no public guarantees or aids are allowed (this is purely the method known as project nance). The promoter or contractor takes on the building and operation risks; as it is not usual in UK to pay tolls by the users, it is the State which pays during the rst years the road runs (from this fact follows the famous name of shadow toll for this method). However, the possibility exists for a nancial support of the British State to these infrastructures, using operating grants, earmarked taxes, exchange guarantees, refunding advances, subsidized loans, scal benets, etc. Austria has also experienced this nancing method. In this way, different highways construction, operation and maintaining projects have been developed since 1982. Other experiences using this method for nancing roads or highways were implemented in Finland, Portugal and the Netherlands (Yates, 1992). Out of the PPP/PFI methods for infrastructure and public services provision, the shadow toll technique has been the most controversial, mainly in the British accounting practice, due to its implications in the Government nancial reports (see in this respect Allen, 2001; Broadbent and Laughlin, 1999; English and Guthrie, 2003; Froud and Shaoul, 2001; Heald,

6 In this latter country this kind of contracts is named public work management contracts (march dentreprise e de travaux publics); see Aguila (1995) and Besancon and Van Ruymbeke (1990).

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

977

2003; Heald and Geaughan, 1997; Hodges and Mellett, 2002; Mayston, 1999; Rutherford, 2003). This is a summary of the accounting problems at stake: (a) whether properties used in PFI contracts are assets and the amounts to be paid to operators are governments liabilities or if, by contrast, the Government only receives a service and pays for it; (b) whether the operator owns an asset to deliver a service or if, by contrast, the operator has a nancial asset representing the amount of payment obligations owed by the Government. The solution nally adopted in the UK is that the Government must report the asset if the Government directly assumes the assets risks and benets. Therefore, risk is the key element when dening the nature of a DBFO contract (Accounting Standards Board, 1994; Treasury Private Finance Taskforce, 1997). The rst time Eurostat issued guidance for estimating PFI incidence on Government decit and debt was in its 21 February 1997 News Release no. 16/97, through the analysis of two cases: Case no. 1. A Government asks a corporation to build and nance a capital asset. Corporations become the owners of these assets as the building is going on. In this case, the gross capital formation must be included in the Public Administration sector. Investment in this case increases government decit but it has no impact on government borrowing, as Council Regulation 3065/93 determines. According to this regulation, Public Administrations long and medium-term commercial commitments with corporations are not included in debt when calculating convergence criteria accomplishment. Case no. 2. A Government asks a corporation to build a capital asset which will be operated by this corporation right through its useful life. The corporation becomes too the owner of the asset. In this case, gross capital formation must be included in the corporations sector, and there is no incidence either on government decit or debt. Eurostat shows also concrete examples where both possibilities are simultaneously in use: case no. 1 is the method used in Germany for building and pre-nancing of roads by public sector agents. The bridge of Oresund, joining Denmark and Sweden, is an example of case no. 2. The construction began in 1996 by means of a Governmental Danish and Swedish Consortium of public corporations. The consortium funded the project by issuing government-secured debentures. The works ended in 2000 and the consortium has a license for operating the bridge and collecting tolls. Total debt is estimated to be refunded in 2026, while the consortium will operate the bridge indenitely. Gross capital formation is included in the corporations sector, with no incidence on government decit. Other PFI contracts in UK are examples of case no. 2. Instead of buying and operating a capital asset, the State hires the services of an operator in the private sector (Heald, 1997; Mumford, 1998). These operators manage to get the assets in order to provide the requested services. Gross capital formation is included in the corporations sector, with no incidence on government decit. A purchase of services provided by business corporations is recorded in Public Administrations accounts, thus increasing government decit every year. The Eurostat 30 April 1997 News Release no. 33/97 added a new case to the previously considered (case no. 3): a corporation builds an infrastructure and subsequently the corpora-

978

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

tion operates it during a given time period, collecting periodically certain amounts of money paid by Government. These payments concern the acquisition of infrastructure services. At the end of the operation time, the infrastructure comes back to the Public Administration with no additional payment. In this case, these rules must be taken into account: Initial investment (gross xed capital formation) must be recorded in the corporations accounting books, with no incidence on government decit. Government annual payments are services acquisitions that increase government decit every year. At the end of the operating time, the reversion of infrastructures to Public Administration does not generate a reduction in government decit. Considering the three above-mentioned PFI cases, we can conclude that Eurostat News Releases presents the German method (case no. 1) and shadow toll (cases no. 2 and no. 3) treatment. Focusing on cases no. 2 and no. 3, investment must be reported as a xed asset by the building corporation, with no increase of government expenses and debt. Only the amounts paid by Government when using xed assets are recorded as government expenses (current expenses to be more precise). However, as previously discussed, shadow toll method has been used in Spain, as a licence system where private sector is engaged to build and maintain the infrastructure, and Government pays a toll or rate for its use, until the settlement of nancial commitments. When the concession term is over, infrastructure becomes a public ownership, with no additional cost for Public Administration. As can be observed, the Spanish implementation of shadow toll differs a little bit of British PFI. According to British practices, any new investment the government tries to do using budgetary resources, must be supported by a specic study; this study must disclose that investment costs are lower than those incurred when using a new DBFO concession. As a result of this assessment, if PFI is preferred, a pure-toll concession or a service management or payment for services rendered contract will be implemented. Nevertheless, it is indeed a question of a deferred disclosure and by instalments payment of investment, along with maintenance costs, all through the concession time. In short, we can considered this PFI as a hire purchase, or at most a quite similar method to a nance lease, with the only difference that no option will be paid by the lessee (Government) to the lessor (private corporation) when receiving the denitive ownership of the asset. Another difference that can also be argued is that in the case of a nance lease, a xed amount must be paid all through the term of the contract, whereas the fees to be paid when using the shadow toll method depend upon the use citizens make of the infrastructure (e.g., for roads fees are paid according to annual trafc). No exact previous calculation can thus be done of the amounts to be paid every year by Public Administration with its budgetary resources. However, Government will guarantee enough revenues for nancial balance of the concessionaire. If no guarantee existed, probably nobody would bid for building the infrastructure. Besides, the Government will also x a ceiling amount to be paid (for example if trafc exceeds the ceiling volume of vehicles, no more fees will be paid). To sum up, we can say that Public Administration runs into debts that must be reported in the nancial statements by its maximum amount, as required by a conservative account-

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

979

ing philosophy. Spanish Regions PFI regulations require budgetary appropriations by the maximum amount the Government is bound to pay to the concessionaires (according to the foreseeable use of infrastructures), in order to guarantee both the effective operation of privately constructed works and the protability of the investment. This fact endorses our opinion about the accounting treatment of the commitments stemming from this PFI, as no risks are indeed assumed by licensees. We must bear in mind that, as the theory of property rights shows, a key aspect in PFI is the accurate risk sharing between private and public sectors. If adequate risk is not shifted to the private sector, then projects become quasi public, but with the funding removed from the governments balance sheet. The amount of debt will be included in a xed asset (a tangible asset or use duties). If the maximum amount taken for debt evaluation remains over the effective payments done when the concession term is over, asset and debt must be cut down by the difference. This reduction must be done every year by the difference between the maximum estimated payments and the real amount of commitments. In January 2000 there was a relevant shift in Eurostat initial opinion towards a position close to our arguments. Thus, the document ESA95 Manual on government decit and debt shows a clarication about accounting treatments of these PFIs by means of two examples (which are identical to the shadow toll): Case a. Government requires a corporation to build a prison according to the governments specications. Government agrees to pay the corporation a certain amount for 25 years for making the prison available for its use, provided it is adequately maintained. At the end of the 25-year period, legal ownership of the prison is transferred from de contractor to government. Case b. Government signs a contract with a corporation for the design and construction of a road. The corporation is also responsible for maintaining the road for 25 years according to an agreed standard. Government pays the contractor an annual fee linked to the number of vehicles using the road. According to Eurostat standards, the treatment for each case should be Case a. When the corporation is exposed to most of the risk/rewards of ownership during the period of exploitation, the infrastructure is recorded in the corporation balance sheet. The contract between government and the corporation has the characteristics of an operating lease. Only regular payments by government have an impact on government net lending/borrowing. If the infrastructure is given to government at the end of the operation period it will be presented in the government balance sheet through a gross xed capital formation, balanced by a capital transfer, with no impact on government net lending/borrowing. Case b. When the government is exposed to most of the risk/rewards of ownership during the period of exploitation, the infrastructure is reported in the government balance sheet. The contract between government and the corporation has the characteristics of a nancial lease. The infrastructure built by the corporation is allocated to government balance sheets through GFCF, balanced by an imputed loan of equal value. Accordingly, there is an impact on government net lending/borrowing for the value of GFCF, and government debt is increased by the amount of the loan imputed. During the operation

980

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

Table 3 Factors inuencing the distinction between operating and nance lease 1. Who is responsible for the maintenance and insurance of the asset? Assume government organizes and pays directly for the insurance and maintenance of the asset This suggests a nance lease since government is bearing the risk of variations in such costs 2. Who repays nance on early termination of a contract? Assume the government is responsible for repayment of the corporation debt in the event of early termination of the contract This suggests a nance lease since government is bearing that risk 3. Who determines the nature of the asset? Assume the corporation has signicant and ongoing discretion on how to full the contract; it makes the key decisions on design and construction of the asset; it decides how it is operated and maintained in order to provide the service required by the purchaser This suggests and operating lease 4. Who bears the demand risk? Demand for services provided by the asset might be greater or less than expected. Assume the corporation income is affected by the demand for the asset, such that government or other customers only pay for the amount of service consumed This suggests an operating lease 5. Are there any third part revenues? Assume the corporation uses the asset to provide services to customers other than just government, and the government is not exposed to the variability of third party demand, and these revenues are signicant part of the total income from the asset This suggests an operating lease 6. Does government pay less if the quality of service is not good enough? Assume government payments are reduced when the service provided by the corporation is not up the required standard, even if this is because of problems with asset rather than how it is operated This suggests an operating lease 7. Does government pay more if the corporation costs increase? Assume government does pay more if there is an increase in the corporation costs related to the asset. For example the corporation might have to undertake more maintenance than expected This suggests a nance lease 8. Who bears the residual value risk? Assume government has the option, at the end of the contract, to buy the asset at the current market price, and that it is not bound to buy the asset at a pre-agreed price if it does not need it nor if the asset is not in good condition This suggests an operating lease

period, annual payments should be subdivided into repayments of principal and interest payments related to the imputed loan. Interest payments have an impact on government net lending/borrowing. Risks/rewards of ownership should be assessed according to the factors presented in Table 3. In a contract, not all the characteristics have to point to the same assessment of the contract, but some may lead to classify it as nancial lease and some others may lead to consider it as an operating lease. In this situation, the relative importance of each characteristic of the

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

981

project must be balanced. In order to avoid confusion regarding risks evaluation, Eurostat issued 11 February 2004 News Release no. 18/2004 (which has been added to the ESA 95 Manual), making clear the treatment of PFI impact on government decit (net borrowing)/surplus (net lending) and debt. This news release states that Eurostat decision applies to contracts in which government is the main buyer of services provided by a private entity and the demand comes directly from the government or from other users (as for example, in medical and education services or some transport infrastructures). An important feature of these contracts is that they usually require the production of several assets. These assets, on the one hand, are specically designed for the provision of the service, and on the other hand, they need an initial capital outlay. The key issue is the beforehand classication of the assets involved in the PFI contract: either as government assets or recorded in the private partner balance sheet. Eurostat considers the assets involved in a PFI as non-government assets only if there is strong evidence that the partner is bearing most of the risk attached to the specic PFI. Therefore, the analysis of the risks borne by the contractual parties is the core element for the assessment of a PFI project, as far as the classication of the assets involved in the contract is concerned. This analysis, therefore, is essential to ensure the appropriate accounting treatment of PFI impact on the government decit. However, this assessment does not consider risks not closely related to the asset and that can be fully separated from the main contract. This the case when part of the contract might be periodically renegotiated, and subject to performance and penalty payments that do not signicantly depend on the condition of the main assets. Many risks may be observed in practice in such arrangements. Eurostat has selected three main categories of generic risks. Therefore, bearing a risk for one party means that this party bears the majority of the risk. A rst category is construction risk, covering notably events like late delivery, failure to keep specied standards, additional costs, technical deciency and external negative effects. Governments obligation to start making regular payments to a partner without taking into account the effective condition of the assets would be evidence that government bears the majority of the construction risks. A second category is availability risk, where the responsibility of the partner is quite clear. It may not be in a position to deliver the volume that was contractually agreed or to meet safety or public certication standards relating to the provision of services to nal users, as specied in the contract. It also applies where the partner does not meet the required quality standards relating to the delivery of the service, as stated in the contract, and resulting from an evident lack of performance of the partner. Government will be assumed not to bear such risk if it is entitled to reduce signicantly (as a kind of penalty) its periodic payments, like any normal customer would require in a commercial contract. Government payments must depend on the effective degree of availability supplied by the partner during a given period of time. Application of the penalties where the partner is defaulting on its service obligations should be automatic, should also have a signicant effect on the partners revenue/prot and must not be purely cosmetic or symbolic. A third category is demand risk, covering variability of demand (higher/lower than expected when the contract was signed) irrespective of the performance of the private partner. This risk should only cover a shift of demand not resulting from inadequate or low quality of the services provided by the partner or any action changing the quantity/quality

982

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

of services provided. To the contrary, it should result from other factors, such as business cycle, new market trends, direct competition or technological obsolescence. Government will be assumed to bear the risk where it is obliged to ensure a given level of payment to the partner regardless of the effective level of demand by the nal user. This makes irrelevant the impact of uctuations in demand level on the partners protability. However, this statement does not apply where the shift in demand results from an obvious government action, such as decisions of units of general government (and thus not just the unit(s) directly involved in the contract) that represent a signicant policy change, or the development of directly competing infrastructure built under government mandate. Eurostat recommends that assets involved in PFI should be classied as non-government assets, and therefore recorded off-balance sheet for government, if both of the following conditions are met: 1. The private partner bears the construction risk. 2. The private partner bears at least one of either availability or demand risk. If the construction risk is borne by government or if the private partner bears only the construction risk and no other risks, the assets are classied as government assets. This has important consequences for government nances, both for decit and debt. The initial capital expenditure related to the assets will be recorded as government xed capital formation, with a negative impact on government decit/surplus. As a counterpart of this government expenditure, government debt will increase in the form of an imputed loan from the partner, which is part of the Maastricht debt concept. The regular payments made by government to the partner will have an impact on government decit/surplus only for the part relating to purchases of services and imputed interest. Finally, we must bear in mind that the concept of private entity must be interpreted as opposed to the limits of the government. Therefore, it includes public, publicprivate or private companies providing services to the government on a market basis (Jones, 2000). All mentioned above supports the idea of real transfer of demand and availability risks for the not-consolidation of this kind of operations in the Public Administration sector, discarding models of xed and unconditional payments as those of some projects developed by the government. Furthermore, special attention is paid to the construction risk. This makes clear the strong need for two elements as a previous stage before the denition of the contract or before its bidding. On the one hand, the denition of investment costs before establishing the conditions of the provision of services. On the other hand, a thorough nancial planning of ows the operation will generate. Provided that in the Spanish case the Government has assumed the majority of the contractual risks, we can conclude that, once again, the aim was to defer the reporting of decit and debt. This deferral is no longer possible in the light of the last Eurostat guidance, and was a clear example of creative accounting. Renda and Schreer (2006, p. 1) state that in the case of Spain the legislation spells out that the State always maintains responsibility in contracts involving the public party. As a result, when a PPP turned out to be more expensive than what had been calculated ex ante, the government was expected to fund the project and ensure its viability.

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

983

Something similar has happened in Ireland, where Eurostat indicated that early PPP projects involved insufcient risk transfer to the private partner, and that investment in these projects would be classied as public investment. Until 2005, all PPP investment in Ireland had been treated in this way (International Monetary Fund, 2004).

6. Concluding remarks When PFIs described in this paper are used, it is the Government and therefore taxpayers but not users, who nally pay the public works. A common characteristic of PFI is that private corporations construct and they (where appropriate) also operate the public work. Business revenues come either from ofcial aids or from periodical payments for construction, maintenance and operation of infrastructures, the amount of which are linked to its actual use. Deferral of payments and instalment plans, as well as borrowing decentralization, have made possible for Spanish Public Administrations, as well as other European governments, to bid and contract new public works while tting EMU requirements without a dramatic reduction in public investment. However, PFI implementation has just been, at bottom, a nancial make-up. In other words, it has been a clear example of public sector creative accounting. As Hart (2003) points out, policy makers argue that PPPs are good because the private sector is a cheaper source of nancing than the public sector. This reasoning is strange since it is hard to imagine an agent that is more able to borrow than the Government through its taxation powers. It is important to say, according to PWC (2005), that all these PFIs are, in general, more expensive than traditional debt operations. As Hurst and Reeves (2004) show, PPPs implemented in Ireland have not resulted in signicant innovations, failing to provide value for money. PFIs are more complex operations and difcult to analyse, and sometimes they are extremely sophisticated from a legal perspective. However, we must bear in mind that governments must keep on investing on social centres, road infrastructures, water-treatment plants, etc., and therefore the rst point must be whether the investment is feasible from a traditional budgetary point of view. In case it is not viable, other alternatives will be considered (price subsidies, refundable advances, syndicated loans, subsidiary loans, exceptional nancial aids, among others). These alternatives should have the appropriate budgetary surveillance, so as to accurately control the volume of operations the government is able to nance. In the same way that limits upon future expenses prevent current politicians from imposing excessive nancial burdens to future politicians, PFI use should be justied from a budgetary and economic point of view. The theoretical underpinnings of our assessment of PFI is twofold. On the one hand, according to the property rights theory, a key aspect in PFI is the appropriate risk sharing between private and public sector in the contract. If adequate risk is not shifted to the private sector, then projects become quasi public, but with the funding removed from the governments balance sheet. If this is the case, projects become more expensive in the long run, while politicians have been able to present a good nancial situation in the short run. On the other hand, if we take into account the Public Choice theory, as we said

984

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

before, taxpayers overestimate benets of current expenditures and underestimate future tax burden. Thus, PFIs allow incumbents to create infrastructures, with an eye on winning the next elections. However, the sad part of the story is that these PFIs are more expensive in the long run than them being nanced through the traditional contracts. In the search for imaginative solutions, we would like to comment on a recent and interesting practice used by the Austrian Government, that could be used by other European public entities. The government establishes a new real estate company (Bundesimmobiliengesellschaft, BIG), which is 100% owned by the federal State. Most of the public buildings (schools, universities, ofces, etc.) have been transferred to this company for its management and maintenance. BIG has nanced this transfer by issuing securities and contracting loans. Most of the buildings transferred to BIG are subsequently rented back to the government units which previously occupied these buildings, by individual contracts based on market estimates. The BIG continues to employ all State civil servants who previously managed the maintenance of these buildings. Their civil servant status has been maintained. As far as government accounts are concerned, this transfer has raised three questions: (i) whether BIG is an institutional unit in its own right or an ancillary unit of general government, (ii) whether BIG had to be classied in the sector general government or in the sector non-nancial corporations, and (iii) whether the property transfer from the government to BIG should be considered as a sale of property (improving the decit) or as other volume changes in nancial assets/changes in classication and structure (neutral for the measurement of decit). Eurostat decision is as follows in this respect (see Eurostat 31 January 2002 News Release no. 15/2002): BIG is an institutional unit in its own right, should be classied in the non-nancial corporations sector, and BIG debt is not to be considered as part of government debt. Besides, the purpose of the transfer is to improve public management of real estate by rationalisation of buildings use, as well as management costs reduction. Moreover, because of the size of the transaction, it was not possible to organise a normal market sale of the property on the Austrian market. In addition, Austrian State wanted to maintain indirectly the ownership of the transferred property via the State owned company. For these reasons, the transfer of property was arranged bilaterally between the Austrian State and BIG. Considering these aspects, this transfer should not be treated as a market sale of real estate property in the sense of ESA 95. In consequence, the assets transfer to BIG has no impact on government decit/surplus. To sum up, PPPs are not a miracle solution and need long time to produce visible results. Governments should allocate the risk to the party that is the least cost avoider, i.e., the party best suited to control and/or bear the risk. Without this approach, the public sector runs the risk of using PPPs for the wrong reasons, for example to make up public accounts in the short-term while worsening the long-term nancial sustainability.

References
Accounting Standards Board. Reporting the substance of transactions, nancial reporting standard 5. London: Accounting Standards Board; 1994. Aguila Y. Le guide du march dentreprise de travaux publics. Pars: Ed. EFE; 1995. e

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

985

Allen G. Risk and the private nance initiative (PFI): library standard note. London: House of Commons Library, Economic Policy and Statistics; 2001. Besancon X, Van Ruymbeke O. Le march dentreprise de travaux publics: une forme particuli` re de concession. e e Pars: Ed. AJDA; 1990. Broadbent J, Laughlin R. The PFI: clarication of a future research agenda. Financial Accountability and Management 1999;15(2):95114. Brown C. Financing transport infrastructure: for whom the road tolls. The Australian Economic Review 2005;38(4):4318. Buchanan JM. Public nance in democratic process, scal institutions and individual choice. The University of North Carolina Press; 1967. Bult-Spiering M, Dewulf G. Strategic issues in public private partnerships. An international perspective. Oxford: Blackwell Publishing; 2006. English LM, Guthrie J. Driving privately nanced projects in Australia: what makes them tick? Auditing & Accountability Journal 2003;16(3):493511. European Commission. Green paper on publicprivate partnerships and community law on public contracts and concessions. COM(2004) 327 nal; 2004. Available online at http://europa.eu.int/eurlex/en/com/gpr/2004/com2004 0327en01.pdf. European Investment Bank. Evaluation of PPP projects nanced by the EIB. Synthesis Report; 2005. European Parliament. Publicprivate partnerships. Models and trends in the European Union. DG Internal Policies of the Union (IP/A/IMCO/SC/2005-161); 2006. Froud J, Shaoul J. Appraising and evaluating PFI for NHS hospitals. Financial Accountability and Management 2001;17(3):24770. Hart O. Firms, contracts and nancial structure. Oxford: Clarendon Press; 1996. Hart O. Incomplete contracts and public ownership: remarks, and an application to public private partnerships. The Economic Journal 2003;113:C6976. Heald D. Privately nanced capital in public services. Manchester School 1997;LXV(4):56899. Heald D. Value for money tests and accounting treatment in PFI schemes. Accounting, Auditing & Accountability Journal 2003;16(3):34271. Heald D, Geaughan N. Accounting for the private nance initiative. Public Money and Management 1997;17(3):116. Hodges R, Mellett H. Investigating standard setting. Accounting for the United Kingdoms private nance initiative. Accounting Forum 2002;26(2):12651. Hurst C, Reeves E. An economic analysis of Irelands rst public private partnership. The International Journal of Public Sector Management 2004;17(5):37988. International Monetary Fund. Publicprivate partnerships. Manuscript prepared by the Fiscal Affairs Department; 2004. Izquierdo R. La nanciaci n de las infraestructuras del transporte. Madrid: Ed. C mara de Comercio e Industria o a de Madrid; 1992. Izquierdo R, Vassallo JM. Nuevos sistemas de gesti n y nanciaci n de infraestructuras de transporte. Madrid: o o Colegio de Ingenieros de Caminos, Canales y Puertos; 2004. Jones R. Public versus private: the empty denitions of national accounting. Financial Accountability & Management 2000;16(2):16778. Koen V, van den Noord P. Fiscal Gimmickry in Europe. One-off measures and creative accounting. OECD Economics Department Working Papers, no. 417, OECD Publishing; 2005. L der K. National accounting, governmental accounting and cross-country comparisons of government nancial u condition. Financial Accountability & Management 2000;16(2):11728. Milesi-Ferretti GM. Good, bad or ugly? On the effects of scal rules with creative accounting. Journal of Public Economics 2004;88(12):37794. Milesi-Ferretti GM, Moriyama K. Fiscal adjustment in EU countries: a balance sheet approach. Journal of Banking and Finance 2006;30:328198. Mayston D. The private nance initiative in the NHS: an unhealthy development in new public management? Financial Accountability and Management 1999;15(3/4):24974. Mumford M. Public-worksprivate nance: the principles of the private nance initiative. London: Ed. Richard Grifn; 1998.

986

B. Benito et al. / Critical Perspectives on Accounting 19 (2008) 963986

PricewaterhouseCoopers. Delivering the PPP promise. A review of PPP issues and activity. London; 2005. Renda A, Schreer L. Publicprivate partnerships models and trends in the European Union. Brussels: European Parliament; 2006. Rutherford B. The social construction of nancial statement elements under private nance private nance initiative schemes. Accounting, Auditing & Accountability Journal 2003;16(3):37296. Treasury Private Finance Taskforce. PFI technical note no. 1, How to account for PFI transactions. London: HM Stationery Ofce; 1997. Yates ChM. Private sector nance for roads in Europe. In: Proceedings of a seminar held at the PTRC transport, highways and planning summer annual meeting. England: University of Manchester, Institute of Science and Technology; 1992. p. 1530.