ایجاد و مدیریت امکانات برای خدمات عمومی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|10941||2006||18 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of Public Economics, Volume 90, Issues 10–11, November 2006, Pages 2143–2160
We model alternative institutional arrangements for building and managing facilities for provision of public services, including the use of the Private Finance Initiative (PFI), by exploring the effects on innovative investment activity by providers. The desirability of bundling the building and management operations is analyzed, and it is considered whether it is optimal to allocate ownership to the public or the private sector. We also examine how the case for PFI is affected by the (voluntary or automatic) transfer of ownership from the private to the public sector when the contract expires. Asset specificity and service-demand risk play critical roles.
The provision of public services is often organized through contracting out by governments to private profit-maximizing firms. Recently, however, governments in Western Europe and North America have developed new forms of public–private partnership (PPP) for public service provision (Rosenau, 2000). A form of PPP that has attracted particular attention is the Private Finance Initiative (PFI) developed in the UK (Grout, 1997 and HM Treasury, 2000). PFI contracts cover most forms of public service provision, including health, education, defence, prisons and roads. HM Treasury (2003) estimates that, over the period 1998–9 to 2003–4, private sector investment in public services through PFI was between 10 and 13.5% of total investment in public infrastructure, with 451 PFI projects completing construction, including 34 hospitals and 119 other health schemes, and 239 new and refurbished schools. There are two major differences between PFI and previous arrangements. First, PFI typically involves the bundling of the design, building, finance, and operation of the project, which are contracted out to a consortium of private firms for a long period of time, usually 25–30 years. The consortium includes a construction company and a facility management company and is responsible for all aspects of services. Second, a system of output specifications is used: the government specifies the service it wants, and some basic standards, but it leaves the consortium with control rights over how to deliver the service. The consortium has responsibility for the infrastructure facility during the contract period, during which it may implement innovative approaches to service delivery and it may use the facility for additional income-generating activities — provided the basic standards of service provision are not violated. However, there is no specific rule as to what happens to facilities at the end of the contract, although, in practice, in the few contracts that have been completed, in the cases of schools, hospitals and prisons, the facilities have been returned to the public sector, whilst for accommodation and general IT systems, they have been kept by the private sector (HM Treasury, 2002, 2004). PFI contrasts sharply with the way public services have traditionally been procured. Under traditional procurement (TP) the different stages of an infrastructure project are contracted out separately to different private firms and an input-specification approach is followed, with the government keeping ownership of the facility both throughout the contract period and after the contract ends (HM Treasury, 1998). The primary motivations for PFI schemes are to allow the consortium to exploit synergies between different stages of a project, and to incentivize the consortium to come up with innovative approaches to service delivery (Daniels and Trebilcock, 2000 and IPPR, 2001).1 Yet, evidence on the performance of PFI, relative to that under TP, is mixed. On the one hand, a greater proportion of projects is being delivered on time and within budget than under TP (National Audit Office, 2003b). In PFI prisons, for example, innovative solutions have been incorporated that had not featured previously, and there is evidence that greater benefits and lower costs are being achieved (National Audit Office, 1997 and National Audit Office, 2003b). On the other hand, the quality and cost of some early PFI schools have been worse than under TP (Audit Commission, 2003). In this paper we analyze the factors underlying these stylized facts.2 First, we study the desirability of the two defining characteristics of the PFI model: whether it is optimal to bundle the different stages of production and whether control rights should be given to the private firm(s). Second, we focus on an important practical concern for public infrastructure projects involving long-term private investments: the role of the residual value the facility, and of ownership of the facility once the contract expires.3 We consider a model in which there are three stages to a project. The first is the ‘building’ of a facility (we interpret this stage to include design), while the second is the ‘management’ of public service provision using the facility. The government delegates these two functions to the private sector, and, because the functions require specialized skills, two distinct firms carry out the tasks. We compare the case of unbundling, where the government contracts with the two firms separately, with the case of bundling, where the government writes a single contract with a consortium of the two firms. For bundling and unbundling, we consider both private and public ownership. Private ownership with bundling is interpreted as PFI; public ownership with unbundling is interpreted as TP. The third stage in our model relates to what happens to ownership at the end of the contract period. To reflect the emphasis given to innovation under PFI, we model investments by the firms as the undertaking of research into innovative approaches to carrying out their respective tasks. Such investments are noncontractible ex ante but verifiable ex post: whilst it is not possible to contract ex ante on the delivery of an innovation, once a potential innovation has been discovered, its implementation is verifiable. If implemented, an innovation may affect social benefit, management/maintenance cost, and residual value. We assume that, due to contractual incompleteness, ownership rights result in control rights: the owner of the facility during the contract period has the power to decide (and veto) whether any given innovative activity can be implemented. Thus, under private ownership, provided that basic standards are not violated, the owner (a firm or consortium) has the power to implement an innovative approach unilaterally. In contrast, under public ownership, renegotiation between the firm/consortium and the authority must take place before an innovation may be implemented. This modelling strategy seems to reflect existing evidence on the different nature of renegotiation under TP and PFI. House of Commons (2003) cites a census that compares the extent of renegotiation of the contract price under these two arrangements. Under TP, in 73% of construction projects the final price exceeded that agreed at contract. Under PFI, the corresponding figure was only 22%, and where there had been a price increase this had mainly been due to changes that had been led by the government department, not the contractor. Such price increases related to further work or facilities that had not been part of the original specification, or to changes to the function of a building.4 Our first result is to show that synergies between the stages of the project do indeed play a critical role, although they do not necessarily work in favor of PFI. A building innovation that increases the social benefit from a project may be associated with either reduced cost at the management stage (we call this a positive externality) or increased cost at the management stage (we call this a negative externality). For example, the design of a prison with better sight-lines for staff that improve security (i.e., social benefit) may yield the positive externality that the required number of security guards is reduced. In this case, bundling is always optimal, for it allows internalization of the positive externality. However, an innovative design of a hospital, using recently-developed materials, may lead to improved lighting and air quality, and therefore better clinical outcomes, but may have the negative externality of increased maintenance costs. In this case unbundling may become optimal, making a consortium undesirable. This is because, in a world where contracts are incomplete, the hold-up problem may lead to underinvestment even under the preferred ownership structure. To attenuate the underinvestment problem, it may become optimal to induce the firms not to internalize the negative externality (a second-best result) since internalization would depress incentives further. Our next set of results relates to the issue of optimal ownership and thus control rights. The intuition follows Hart et al. (1997). Under private ownership, the firm/consortium has ownership and control rights. Therefore, if there are private gains from implementing the innovation, it will implement the innovation without any further inducement. In this case a promise by the government to reward the firm/consortium for the increase in social benefit following the innovation would not be credible. Realizing this, the investor will disregard the effect on social benefit when it decides on its investment. In contrast, when control rights are left with the government, the firm/consortium has no power to implement the innovation unless renegotiation occurs. But, through renegotiation, it will internalize some of the benefits that its innovation brings to the government, and give up a share of the private benefits that the innovation brings to itself. In our context this implies that, with a positive externality, control rights should lie with the firm/consortium if the cost and residual value effects dominate the social benefit effect, and with the government in the reverse case. In contrast, with a negative externality, provided the cost effect is relatively small, a large residual value effect favors ownership by firm 1, and a large social benefit effect favors ownership by the government. Innovation during the contract period is also related to how the facility will be used after the contract expires. In practice, with PFI, the facility is in some cases retained by the consortium at the end of the contract period, but in others it is transferred back to the public sector, mostly automatically. In our analysis, assuming a positive externality, these arrangements do not affect the rationale for bundling. Furthermore, the prospect of a possible transfer of ownership to the public sector — if this can be achieved by voluntary negotiation, and the payment of agreed compensation — improves the consortium's investment incentives. But, given the unverifiability of investments and residual value, an automatic transfer clause in the PFI contract is necessarily inoptimal because it blunts these incentives. We explore these issues by assuming that when the initial contract is agreed, there is uncertainty concerning the relative sizes of the ‘public residual value’ (the residual value if the facility is used in the public sector when the contract ends) and the ‘private residual value’ (the facility being used in the private sector). Our analysis leads to specific results concerning the factors favoring the use of PFI. It is more likely to be optimal to use PFI if the externality is (more strongly) positive, if the effect of innovation on social benefit is relatively small, and if the effect on residual value is relatively large. Also, PFI is more likely to be optimal the more probable it is that the private residual value will be higher than the public residual value, and the lower the specificity of the facility for public, rather than private, use at the end of the contract. These results are generally consistent with empirical evidence on PFI, which we cite where relevant below. Several recent contributions apply the theory of incomplete contracts to the contracting out of public services. Hart et al. (1997) compare contracting out with in-house provision to a single private firm. They show that when there is contracting out the private firm will reduce costs excessively, at the expense of quality, whereas with in-house provision there are blunted incentives for both cost reduction and quality improvements.5Besley and Ghatak (2001) study the optimal provision of public goods and show that ownership of a public good should lie with the party that values more highly the benefit that the public good generates. Neither of these papers discusses bundling, but Hart (2003) finds that bundling is desirable if the quality of the service can be well specified in the initial contract, whereas the quality of the building cannot. In contrast to our analysis, he assumes that investments are never verifiable (so renegotiation never takes place) and in his model it does not matter who owns the facility at the end of the contract.6King and Pitchford (2001) also discuss bundling and consider the possibility of spillovers onto the value of other facilities. However, they model public and private ownership as distinct sets of rules that affect managerial discretion, and so no bargaining ever takes place. An alternative, complete-contract, approach to PFI is taken by Bentz et al. (2001). They show that the government will wish to buy services (as in PFI) rather than facilities (as in TP) if the building and service delivery costs are low. None of the papers cited here analyzes explicitly the role of residual value. Section 2 outlines the model, and Section 3 analyses the level of investments under alternative regimes. Section 4 compares different ownership structures first for a positive and then for a negative externality, while Section 5 explores issues relating to residual value. Section 6 concludes.
نتیجه گیری انگلیسی
In this paper we have studied the desirability of bundling the building and management of facilities used for the provision of public services; we have considered the appropriate ownership of these facilities; and we have focused on the role of the residual value of the facilities. We have paid particular attention to the case for PFI, and we have related our analysis to the evidence available. When there is a positive externality across the stages of production, bundling, with the firms organized as a consortium, is always optimal since it induces the internalization of the externality. This is consistent with the motivation commonly given for PFI contracts, which views the integration between the different phases of the provision of a public service as a device to promote investment. We have also shown, however, that, with a positive externality, ownership of facilities by the consortium (the PFI model) is not necessarily optimal: under some conditions it is preferable for the government to have ownership. Furthermore, if the externality is negative, the case for bundling is weakened, and, if the externality is weak, unbundled provision is preferred. PFI is more likely to be preferred (a) the more positive (or less negative) is the externality; (b) the stronger the effects that innovations in building and management have on the residual market value of facilities; and (c) the weaker the effect that innovations have on the benefit from provision of the public service. With a weak negative externality, if the period over which the service is to be provided is lengthened, public provision is favored relative to private provision. Provisions for ownership of facilities when the public service contract has expired also have an important role to play. In particular, assuming a positive externality, under PFI, automatic transfer of facilities back to the public sector reduces ex ante investment incentives, thereby weakening the case for using PFI. If, instead, it is specified initially that the PFI consortium will keep ownership at the end of the contract, the option to negotiate a mutually beneficial transfer of ownership to the public sector strengthens its investment incentives, and therefore also the case for PFI. Thus, the case for PFI is stronger for services for which there is no political or social prohibition on long-term private ownership of the facilities concerned. Our analysis has been based on the assumption that investments are verifiable. But suppose, instead, that they are unverifiable, as assumed by Hart (2003), in which case there can be no bargaining over implementation of innovations. Then a private provider can only be driven by concern for its own profit. In the revised model, the only innovation that will occur will be the building innovation a if firm 1 is the owner, the management innovation e if firm 2 is the owner, or both a and e if there is PFI; under public ownership there will be no investment. Hence, if the externality is positive, the case for PFI is strengthened: PFI dominates each other arrangement. A development of our analysis would be to take into account distinctions between different types of services. In particular, government reports have distinguished between core services, such as prison security or school classroom teaching, and ancillary services, such as cleaning and facility maintenance in prisons or schools (see, e.g., House of Commons, 2001). It is suggested that PFI works better where it provides both types of service (as in prisons), but less well when it is restricted to ancillary services (as in schools). This issue could be analyzed by incorporating into our model a public sector service agent that may provide the core service, and by allowing for possible synergies between core and ancillary service provision.19 The main limitations of our analysis are the assumptions that the contracting game is not repeated and that information is symmetric between the government and firms. Since PFI contracts are relatively complex, bidding costs are high, and few firms (or consortia) participate in the tendering process. On average there have been approximately four participants per PFI tender in the UK, and firms such as Jarvis and Carillion have won PFI contracts over a wide range of types of project. Thus, other things equal, we may expect reputational mechanisms to be particularly important under PFI. Although asymmetric information might be expected to weaken the case for any form of private provision, the implications for the choice between PFI and traditional procurement are not clear cut. Whereas a firm (or consortium) may have a good idea of the sign and magnitude of the cost externality, the government may not. Under PFI the firm implements an innovation without negotiation with the government, and so the government's lack of information has no direct effect. In contrast, with traditional procurement, since the size of the relevant bargaining surplus will in this case be unknown to the government, its bargaining position will be somewhat undermined. Whether this might lead to a stronger case for PFI is an interesting avenue for future research.20