Public-Private Partnerships: A Swiss Perspective

Abstract

While the last decades have seen a spectacular expansion in public-private partnerships (PPPs) in many high-income European countries, only two PPPs have been realized in Switzerland so far. In this chapter, we define what PPPs are and the optimal conditions under which they should be chosen. Our normative approach highlights that PPPs are not a panacea, and that the choice to use a PPP should be driven by the characteristics of the public service considered, according to transaction cost theory. We then consider the cultural and institutional differences that can explain the variations in the number of PPPs which are actually implemented accross countries with similar level of development. We point out that while there are probably too few PPPs in Switzerland, particularly due to the lack of a specific legal and institutional framework, there are clearly too many of them in some other countries.

1 Introduction

Since the first implementation of the Private Finance Initiative in the United Kingdom in 1992, the last decades have seen a spectacular development of public-private partnerships (PPPs) in many developed as well as developing countries. It has meant an increased participation of the private sector in providing a service itself as well as its accompanying infrastructure.

Interestingly, while most countries have adopted PPPs, the prevalence of such arrangements differs widely across countries, and the differences have persisted. In particular, while 722 PPP projects were launched in the United Kingdom between 1994 and 2016, Switzerland only had two during the same period. In Europe, Switzerland exhibits one of the lowest numbers of PPPs.

What could explain this low number, and is this a good or a bad thing? What is the right number of PPPs? In answer, we first define what PPPs are and what they are not (Sect. 11.1). We then develop a theoretical framework establishing the conditions under which PPP arrangements are optimal, or more optimal than other possible modes of provision (Sect. 11.2). This normative analysis highlights that the choice to use a PPP should be driven by the characteristics of the public service considered. As we expect public services to be quite similar across countries at a similar level of economic development, only cultural and institutional differences can explain the differences in the PPPs which are actually implemented. We then consider the bad reasons to use and not to use PPPs which might differ across countries. We point out that while there are probably too few PPPs in Switzerland, there are clearly too many of them in some other countries (Sect. 11.3). We conclude with some policy recommendations.

2 What Is a Public-Private Partnership?

2.1 Public-Private Partnerships Within the Myriad Ways of Providing Public Services

Once a public service has to be provided, public authorities can choose between a large number of modes of provision. To distinguish between these possibilities, it is useful to divide the life cycle of a project or an infrastructure into four main tasks: designing, building, financing and operating or maintaining. The allocation of these tasks between one and several agents (public and/or private) determines the mode of provision. The most frequent modes of provision are presented in Table 11.1.

Table 11.1 Overview of the possible modes of provision of public services

Under traditional procurement, the public authorities remain in charge of all four stages, except for building, which is often contracted to a private firm through a procurement contract. This means that the public authority keeps control over the infrastructure and the service provided, and that it also bears all the risks except for construction risks. The public authority may give some autonomy to the public provider of the service by creating specific public entities. These entities can be either autonomous public entities (e.g., the Geneva Airport) or state-owned limited companies under public law (e.g., the Swiss Federal Railways and the Swiss Post) or under private law (e.g., the air navigation service provider Skyguide). The relationship between the authority and the autonomous entity can either be based on a law or on a contract (as in FORS, the Swiss national centre of expertise in the social sciences; see Athias 2013 for details). All these arrangements correspond to a public provision of a public service.

Another way to provide a public service, quite specific to Switzerland,¹ is when a private operator builds and finances all or part of a public infrastructure in exchange for the opportunity to exploit the premises for commercial purposes unrelated to the public service (e.g., shopping malls or office buildings). This is made possible by granting a distinct and permanent leasehold right,² whose duration is typically between 30 and 100 years. This gives the private partner the right to build and own a distinct asset on ground whose ownership remains in the hands of the public authority. An example is the Tissot Arena, the new sport complex inaugurated in 2015 in Biel, which encompasses football, ice-hockey and curling fields, as well as a shopping mall. While these arrangements are often considered as PPPs due to the private financing of the public infrastructure, they are not PPPs as there is no involvement of the (same) private partner in the service provision.

When the provision of a public service does not require physical assets or when the public authority owns an infrastructure but wants to delegate only its operation to a private entity, the private party involvement is mostly based on service contracts, such as for child day-care centres or for services provided to migrants. Within such contracts, we can distinguish between lease contracts, where the private firm is paid by the users of the service, and management contracts, where the private operator receives a fixed-price payment from the public authority.

The provision of the public service can also be outsourced through PPPs. PPPs can be defined as long-term arrangements between a public authority and a private partner, chosen after a competitive tendering, in order to design, build, finance and operate an infrastructure that is used to provide the public service.³ This infrastructure can be either a new one or an already existing that needs to be renovated. The peculiarity of PPPs lies in the bundling of building and operation stages. As highlighted by HM Treasury , “Private sector expertise and experience has always been used in public sector procurement, but, where in traditional procurement , private companies built and then walked away, PPP seeks to ensure that the private sector takes responsibility for the quality of design and construction it undertakes, and for long term maintenance on an asset, so that value-for-money is achieved” (HM Treasury 2003).

PPPs can be either contractual or institutionalized. In the first case, the public authority concludes a contract with the project company (which can be a consortium) without being part of it. By contrast, in institutionalized PPPs, the public authority is a (minority or majority) shareholder of the project company. Boxes 11.1 and 11.2 below provide Swiss examples of both types of PPPs. Among PPPs, we can further distinguish availability and concession schemes. Whereas both are fixed-price, long-term arrangements to design, build, finance and operate a public infrastructure, the main difference relies in the sharing of risks between the public and private partners. In availability schemes, the public authority pays a fixed price to the project company according to performance criteria (demand risk is hence borne by the public sector). By contrast, in a concession scheme, the project company is remunerated according to the demand for the service (either directly by the users or indirectly by the public authority via shadow tolls) and hence bears the demand risk. In Switzerland, the term concession is used to describe the right to pursue economic activities that are regulated by the state, mainly because of the monopolistic nature of the market (e.g., local public transportation) or because this activity needs an access to a limited public resource (e.g., water, as in the case of dams or run-of-river facilities, or radio frequencies) or the use of the public domain (e.g., to build and operate a gas network) or due to the sensitive nature of the regulated activity (e.g., casinos) or to increased risks (e.g., airports or the storage of dangerous substances). Swiss “concessions” must then be distinguished from PPP concessions as they can be awarded to public or private service providers, and under all kinds of governance structures. For instance, the Federal Office of Civil Aviation (FOCA) can award concessions to operate airports to public authorities or to private firms. Although no such example curently exists in Switzerland, it would also be possible to award an airport concession to a PPP.

Finally, a public service can be provided through a regulated market. In this case, an authorization to provide the service is required, subject to compliance with some minimum requirements to ensure the quality of the service provided. Unlike PPPs, which are dedicated to specific projects, the requirements in regulated markets apply to all the firms in a sector; they are based either on federal, cantonal or municipal legislation, or on specific guidelines issued by a public authority. This is, for example, the case for nursing homes, which are more or less strictly regulated by the cantons (Athias and Wicht 2018a). This regulation can, in some cases, be the counterpart of public subsidies.

Box 11.1: The Administrative Centre Neumatt, the Only Contractual PPP in Switzerland

The first and only example of a contractual PPP in Switzerland is the administrative complex Neumatt, in the town of Burgdorf (BE). The availability contract was signed in 2009 between the Canton of Bern and the project company Zeughaus PPP AG, formed by the construction groups Marti AG and Royal BAM AG, as well as Hälg Facility Management AG. This contract covers a period of 25 years, starting in 2012, and includes designing, building, financing and operating a new complex which includes a regional prison for 110 inmates, four administrative buildings, a workshop for the canton’s Road and Civil Engineering Services, as well as an underground car park. The project company was chosen from among five companies, after three rounds of competitive tendering, and the total value of the project is 150 million CHF.

The operator of the complex is in charge of all tasks except those related to the custody and care of inmates. The services provided by the private partner encompass, among others, building facility management, the management of the car parks, refuse management, internal mail service, the staff restaurant, the signage, the management of the keys and the management of the office supplies. The private company is also in charge of the security and the surveillance of the complex, with the exception of the prison, which remains in the hands of the public authority.

Box 11.2: Cadiom, an Institutionalized PPP to Provide Distance Heating in the Canton of Geneva

Cadiom is an institutionalized PPP created in 1999 between the Services Industriels de Genève (SIG, an autonomous state company) and the consortium Vulcain, composed by CGC Energie, two engineering companies, and the construction group Zschokke (now Implenia). The public authority holds a majority (51%) in the project company Cadiom SA. The goal of this company is to design, build and operate a distance heating network in order to use the surplus heat coming from the incineration plant of Les Cheneviers in Aire-la-Ville. This network provides heating to more than 10,000 homes in five municipalities in the Canton of Geneva. The company Cadiom is paid directly by the clients (homeowners), hence corresponding to a concession scheme. The canton grants Cadiom the right to use the public domain and monitors the price and the quality of the service. The private partner was chosen after a competitive tendering process.

2.2 PPPs in the World and the Swiss Position

Between 1994 and 2016, 1458 PPPs were created in Europe (see Fig. 11.1), with a total value of 428.2 billion Euros. Among these projects, almost half were realized in the United Kingdom (49.1%), far ahead of France (11.9%) and Spain (10.9%).

Fig. 11.1

Number of PPPs in European Countries (1994–2016)

Switzerland has a very low number of PPPs. Indeed, though many projects are incorrectly described as PPPs by the authorities, only two genuine PPP projects have been realized in Switzerland so far: the administrative centre Neumatt in BE (a contractual PPP, see Box 11.1), and the distance heating network Cadiom in the Canton of Geneva (an institutionalized PPP, see Box 11.2).

3 When Should Governments Resort to PPPs?

This question is related to the more general question of optimal organizational choices, addressed in the work of the economists Ronald Coase (“Nobel Prize” 1991) and Oliver Williamson (“Nobel Prize” 2009).

3.1 Make or Buy for Public Services: The General Framework

A broad distinction can be drawn between the in-house provision of public services (“make”), such as through traditional procurement and private financing schemes and by outsourcing to the private sector (“buy”), either through service contracts or PPPs or regulated markets.

Ronald Coase posed the fundamental question of what the difference is between “make” and “buy”, in other words, why firms (or more generally public and private organizations) exist alongside the traditional market governance structure. He distinguishes between the hierarchy (the firm/organization) as a governance structure, where the coordination mechanism is the authority and hierarchy of the entrepreneur (through the labour/subordination contract),⁴ and the market, where it is the price mechanism that ensures the coordination of different players. As Coase (1937, p. 390) wrote: “The main reason why it is profitable to establish a firm would seem to be that there is a cost of using the price mechanism”. The concept of a cost of using the price mechanism—a transaction cost—was further developed in the early 1970s by Oliver Williamson,⁵ who developed the transaction cost theory (hereafter TCT) and formulated precise propositions on the nature of transactions costs, their measurement and the trade-off between “make” and “buy”.

According to the TCT , the market has a productive efficiency advantage due to stronger incentives related to the private ownership of profit and the competitive pressure that allows for disciplining and sanctioning the poor performance of agents. In some cases, the market governance structure might also achieve cost efficiency through economies of scale associated with high fixed costs services. That is the case, for instance, when the private partner has many clients and is present in many markets, which enables it to spread the average costs across a larger area of production, which is not possible for a single geographically limited public administration entity. Finally, the market governance structure might lead to productive efficiency gains when it allows an optimal allocation of risks. In particular, the market solution makes it possible for the public authority to transfer some risks to the private provider, who has a better ability to manage these risks (due to greater experience), such as risks associated with demand, availability and construction. By contrast, a public authority is better able to manage other risks, including the political and the environmental, and should bear such risks.

There is nevertheless a cost in using the market, because contracts are incomplete per se due to the assumption of the bounded rationality of agents. Agents are assumed to be rational, but they face cognitive limits in processing all the available information to design a complete contract. This contractual incompleteness leads to transaction costs that can manifest themselves ex ante (costs of redaction and negotiation, guarantees) and, above all, ex post (costs of contract maladaptation, renegotiation, contract enforcement, as well as the costs of breaching the contract). While the most important transaction costs are not observable, the very important contribution of the TCT is to highlight that their magnitude can be nevertheless assessed according to the characteristics of the transaction in terms of asset specificity, uncertainty and complexity.

Asset specificity is defined as the extent to which the investments made to support a particular transaction have a higher value to that transaction than they would have if they were redeployed for any other purpose (this difference of value constitutes a quasi rent). For example, if an individual learns Navajo, a language only spoken by a specific Amerindian community in the southwestern United States, he is making a very specific investment compared to those learning English, as the knowledge of Navajo is nearly without value outside this community. As a result , the presence of specific assets leads to the apparition of the so-called quasi rent (the difference of value of the investments for the transaction and outside the transaction). Asset specificity can be of different types such as physical specificity involving specific equipment, human capital specificity associated to the specific knowledge valuable to the transaction or site specificity involving specific geographical localization and other types. Asset specificity leads to transaction costs because it locks-in contracting parties into a situation of bilateral dependence, increasing the likelihood of occurrence of opportunistic behaviours to appropriate the quasi rent from both contracting parties, knowing that contracts are incomplete.⁶

In addition to asset specificity, uncertainty regarding the conditions that will prevail during the execution of the contract is another important determinant of transaction costs. As agents are supposedly rationally bounded, they might be unable to forecast all future contingencies during the life cycle of the contract. Thus, uncertainty often calls for welfare enhancing adaptation ex post by renegotiating the initial contractual terms, opening the door to potential opportunistic behaviour, and hence, overall, transaction costs.

Finally, transactions might be intrinsically complex in their object. The contractual difficulty generated by complexity can manifest itself either ex ante or ex post. Ex ante, it comes from the difficulty in specifying in the contract the expected service and the quality requirements, whereas ex post, it comes from the difficulty of observing and measuring the quality of the service provided. It might also be the case that even if the quality can be measured, it may be difficult to prove to third parties (e.g., a court) that an observed insufficient quality is attributable to the provider and not to exogenous causes. As a consequence, private providers can reduce costs to the expense of the quality of the public service. This is what Hart et al. (1997) observe in the particular case of US prisons. When operated by private operators, they observe that wardens are under-qualified, leading to an increase in violence and escapes. This would explain why, according to them, prisons delegated to private providers in the United States are prisons for those under 18 and not for dangerous prisoners.

Thus, the magnitude of potential transaction costs is determined at the transaction level, according to the above-mentioned transaction characteristics. This magnitude, in turn, drives the choice of the governance structure. Considering the respective advantages and drawbacks of the governance structures, the main theoretical proposition from the TCT is as follows: the higher the expected transaction costs, the more hierarchical the chosen governance structure should be. The optimality of the choice depends then on the adaptation of the governance structures to the characteristics of the transactions that they have to frame, defining the alignment principle.

3.2 Relative Optimality of PPPs

The trade-off developed above applies to the particular case of PPPs. As described in the first section of this chapter, PPPs correspond to a “buy” solution and hence exhibit advantages in terms of productive efficiency but drawbacks in terms of transaction costs, which will be more or less important according to the public service considered. Nevertheless, PPPs also have specific potential advantages and drawbacks.

First, a specific benefit of PPPs derives from bundling different phases of a project. In particular, bundling the design, build, operate and maintenance phases leads to life cycle cost savings. The private partner responsible for building a certain infrastructure has a stronger incentive to provide better quality when she also has to manage the maintenance of this infrastructure, in order to reduce her total costs. In other words, bundling induces the private partner to internalize, at the building stage, possible externalities during the operating phase, and thus to exploit the complementarities and synergies between the different phases of the project. This could lead to innovations at the building stage. The higher the externalities between different project phases, the higher the productive efficiency gains associated with PPPs. This bundling also allows for an improvement in global cost transparency.⁷ Bundling hence increases the alignment of incentives between public and private partners.

However, bundling involves a certain number of disadvantages in terms of transaction costs. In particular, bundling leads to a longer procurement process and to higher costs associated with bidding than traditional procurement processes (see Athias and Chever, 2018). Bundling different phases also increases the complexity and uncertainty of the project, and hence may increase transaction costs due to contract maladaptation and/or renegotiation. As a consequence, problems of adverse selection might arise; these are certainly the main source of transaction costs. More specifically, the winning private provider might not always be the most efficient one. Instead, this provider might be either the most opportunistic one (i.e., who best anticipates the future renegotiation of the contract) or the most optimistic one (regarding future demand or costs). It leads to the “winner’s curse” (see Athias and Nuñez 2008, 2015).

Thus, the specificities of PPPs increase both productive efficiency gains and transaction costs associated with the market solution. The alignment principle mentioned above would then call for a hierarchical PPP structure to minimize total costs. Athias and Saussier (2018) highlight that contractualized PPPs, more specifically concession contracts, might indeed be very hierarchical , with, for instance, contractual clauses that foresee not only that partners will have to renegotiate the contract every three or five years but also how renegotiations should take place. They also highlight that PPP contracts exhibit heterogeneity in terms of hierarchical structure, a function of the degree of uncertainty surrounding the transaction, as predicted by the theory. In addition, within PPPs, the features of institutionalized PPPs make them more hierarchical than contractual PPPs.

We expect then that institutionalized PPPs would more likely be chosen for public services which are potentially prone to significant contractual hazards. This is in line with the two case studies presented in Boxes 11.1 and 11.2. In the case of the Cadiom, the project involved the design, building and operation of a distance heating network with high uncertainty over the source of energy to be used. This uncertainty led to the choice of a more hierarchical form of PPP, which allowed for more coordination and mutual adaptation. By contrast, the Neumatt project involved simpler tasks, thus leading to a less hierarchical form of PPP, the availability contract.

As a result , the question of when governments should use PPPs, and which form of PPPs, depends on the characteristics of the service (s) to be delivered. It is hence possible to explain variations in the propensity to use PPPs across services, but not across countries at similar levels of economic development.

4 Why Is Switzerland Different?

Efficiency considerations should drive the use (or non-use) of PPPs, and that in turn depends on the type of service which is to be provided. To explain the low number of PPPs in Switzerland, one needs to look at institutional and cultural considerations that lead public authorities to use (or not use) PPPs.

4.1 Bad Reasons for Using PPPs

PPPs are frequently perceived by policy makers as a good way for a public authority to realize an infrastructure project when the financial means are constrained. While PPPs make it possible to avoid, or at least limit, an initial investment and hence the future interest and amortization of the debt, the counterpart is that the public authority will have to pay a contribution to the private provider (in the case of an availability contract) or to forego earnings from user fees (in the case of a concession), and this throughout the operation stage of the project. This boils down to the Ricardian Equivalence: the resources saved by the government by not paying the upfront investment under a PPP should be equal, in present value, to fees paid or user fee revenue foregone to the private provider. In addition, this leads to shifting the cost of a project to future generations. Thus, PPPs must not be considered as a means to get a “free lunch”.

In practice, choosing a PPP only because a public authority cannot bear the initial investment is often a means of circumventing a debt constraint imposed by law or upper tiers of government. As The Economist (2009) notes: “Cynics suspect that the government remains keen on PFI not because of the efficiencies it allegedly offers, but because it allows ministers to perform a useful accounting trick”. This also happened in the administrative centre Neumatt. The Canton of Berne decided the project did not fall under the debt brake, as the investment was borne by the private partner rather than the canton. However, the canton’s auditing office required the project be treated as a standard investment. The minister in charge of cantonal infrastructure then explained that the canton subsequently became less interested in engaging in new PPPs because, despite the success of the project, the goal of this approach, that is, alleviating the burden on the investment budget, has not been reached (Neue Zürcher Zeitung 2013, translation). This window-dressing of budget deficits as a way to get around the law is obviously not an acceptable, valid economic reason to justify using PPPs. Accounting rules for PPPs have been revised in order to avoid such behaviour from governments.⁸

In Switzerland nevertheless, another way to circumvent debt constraints is through the private financing project scheme, as highlighted in the first section of the chapter. Since the private provider finances the construction of the infrastructure associated with the particular public service being provided, no debt to the public authority is incurred. As with PPPs, it gives the authorities the illusion of a “free lunch”, and the opportunity cost associated with the alternative use of the ground is not assessed, even if it is potentially important. Thus, private financing is used in Switzerland as a substitute for PPPs to circumvent debt constraints. Privately financed projects can also be used by Swiss public authorities to avoid facing a referendum, which may become mandatory if proposed spending exceeds a specified amount. As private financing projects makes it possible to avoid the initial public investment, this could lead public authorities to use private financing schemes as a means to circumvent the semi-direct democratic instrument of the referendum.

PPPs are also often used in order to circumvent weaknesses in the traditional procurement process. In particular, a public authority might use a PPP to force itself to evaluate the overall costs of a project, by locking itself into a contractual or institutionalized relationship. In other words, PPPs could be used as a commitment device by public authorities. Although this might at first glance lead to better efficiency, it would be more appropriate to correct the organizational problems within the administration rather than turning to a PPP when it is not efficient. PPPs are not the quick-fix solution to the inefficiencies and bad practices of the public sector.

This would in turn explain why the use of PPPs is less frequent in countries where the public sector is considered to be more efficient, as in Switzerland. For instance, Afonso et al. (2005) constructed a Public Sector Efficiency indicator that measures the quality of the administration in terms of corruption, red tape, quality of judiciary and shadow economy. According to this indicator, Switzerland ranks well above all the other advanced OECD countries (see Fig. 11.2).

Fig. 11.2

Comparative public sector efficiency

4.2 Bad Reasons for NOT Using PPPs

However, if there are bad reasons to use PPPs, there are also bad reasons NOT to do so.

Public choice theory tells us there are private benefits for politicians to keep the provision of public services within the public sector. It allows policy makers to award jobs to their relatives, friends or political colleagues, though one should quickly add that Transparency International considers Switzerland one of the least corrupt countries in the world. Switzerland does have some forms of hidden corruption in its public service hiring and public procurement practices. This cronyism (referred to as the “B vitamins” in Switzerland, where B stands for the German word Beziehungen, i.e., Relationships) is widespread in Swiss administrations, especially at the local and cantonal levels. The high degree of decentralization gives substantial power to local politicians who, due to the militia system (meaning that most politicians are non-professional: less than 2% of local executive board members serve full time (Ladner 2011; Geser et al. 2012)), have closer connections with private interests (Meinhardt et al. 2014) and are hence more likely to make biased choices. In the same vein, at a higher level, it is a fact that the interest groups and lobbies have an effective influence on the militia members of the cantonal and federal parliaments. Their willingness and practices to advance their private interests which are often detrimental to the public interest are easier to hide in non-PPP projects for which it is easier to adapt the project in order to circumvent the obligation to use competitive tendering that applies to PPPs.⁹ This would also help explain why Swiss public authorities are biased against PPPs.

There are also “bad reasons” to avoid PPPs on the private sector side. Indeed, while PPPs allow the private sector to be involved in providing public services, the private sector may be sceptical about the rather hierarchical nature of most PPPs. In such a governance structure, the mission orientation of a public authority may conflict with the profit orientation of the private provider.

As a matter of fact, in Switzerland, the degree of mission orientation strongly varies between cultural groups, in particular between language areas, as German-speakers are more prone to consider that public firms must be managed as private firms (for more on this, see Athias and Wicht 2018b). There is a cultural reluctance, from a sizable share of the population, to accept the fact that for a market solution to be efficient in the provision of complex services (which are prone to higher transaction costs), it has to be associated with a coordination mechanism that tends towards hierarchy.

Switzerland also lacks a specific legal and institutional framework at the federal level for PPPs. Unlike in other countries where there are laws specifically designed for PPP arrangements (e.g., France, the United Kingdom), PPPs in Switzerland are based on the general rules of contract and procurement laws. Thus, the legal environment in Switzerland is weaker than in other countries, leading market solutions through PPPs likely to be less efficient in this country than elsewhere. In addition, whereas some countries have specific institutions to support the implementation of PPPs (e.g., the FIN INFRA in France), this is not the case in Switzerland. As a consequence, local decision makers might not have developed the skills and expertise to implement PPPs.

5 Conclusion

PPPs are not a panacea for providing public services, and they are optimal solutions only in very specific cases. There is thus no advocacy in Switzerland to systematically use PPPs, as has been the case, for example, in the United Kingdom. This might appear to be good news.

However, Swiss public authorities have shown a tendency to privately finance projects as a way to circumvent both debt and the constraints of a semi-direct democratic political system. Even when PPPs might be optimal solution, this tendency often biases the choice to not use PPPs. However, one could change the accounting rules for the private financing of projects by imposing competitive tendering for all projects, whether privately financed or as PPPs. Legislation should also be strengthened to provide a strong framework for implementing PPPs, and specific institutions should be developed to help the public authorities in implementing such arrangements and providing them with the required skills. Finally, the conception of the market might gain from evolving in Switzerland. Even in a competitive market, the existence of transaction costs implies that the provision of public services has to be regulated.