21
Designing Public-Private Contracts for the Efficient Provision of Infrastructure Services
Draft
2006-03-31
Jan-Eric Nilsson
Dept. of Transport Economics
Swedish National Road and Transport Research Institute
Box 760
781 27 BORLÄNGE
Phone +46243736 79
Mobile +46 70495 1531
Abstract: The purpose of this paper is to discuss contractual tradeoffs in the extended use of private enterprises in the provision of infrastructure services, often referred to as Public Private Partnership (PPP): Which is the appropriate way to allocate responsibilities between the public sector principal and the private sector agent, taking the nature of the underlying optimisation problem into account? To this end, both matters related to allocative and productive efficiency are discussed. The former concerns the appropriateness of user charges to pay for new infrastructure and ways and means for ascertaining that a contractor takes due account of user effects in the design and maintenance of a new facility. The discussion of productive efficiency focuses the design of contracts with respect to sharing of risk between the parties. In addition, the incompleteness of the contracts and the concomitant risk for renegotiation is dealt with. The overall conclusion is that there is no straightforward “yes” or “no” answer to the question of whether PPP’s enhance welfare. Rather, the preconditions for projects differ, calling for a wide variety of solutions tailored to circumstances.
1. Introduction
Private-sector contractors have since long been used for building new infrastructure, contracts being awarded after a tendering process. When the new facility has been completed the builder is compensated according to the terms established in the contract where after the relationship between the parties is concluded. In some countries, private contractors are now also involved in the maintenance of existing roads and railways. (Reference?) Again, competitive bidding replaces the use of in-house resources.
There is however a further international trend with respect to the relationship between a public-sector infrastructure agency and a private contractor. What is often referred to as a Public-Private Partnership (PPP) changes the parties´ position in at least two ways relative to standard procurement:
· The contractor retains control over a construction project for a (long) pre-specified period of time and is made responsible for its maintenance before it is handed over to the procuring agency.
· The contractor may recover construction costs in non-standard ways, including tolls, shadow tolls or down payments over the contract period.
The purpose of this paper is to address contractual tradeoffs inherent in this change of roles: Which is the welfare maximising way to allocate responsibilities between the public sector principal and the private sector agent, taking the nature of the underlying optimisation problem into account? The discussion primarily addresses road and railway infrastructure projects, but much of the analysis generalises also to ports and airports. Most examples are, however, from the road sector.
It will be demonstrated that different contexts require different contractual constructs. There is therefore no single conclusion to be drawn with respect to the appropriateness of PPP’s. Rather, recommendations could range from using long contracting periods with user charging to recover all costs, to situations where there is weak evidence in favour of changing the role of the private sector as compared with today. A further conclusion is that the detailed design of contracts to deal with differences with respect to risk concerns and other information incompleteness’s, is of paramount importance. The devil is in the contractual fine print.
We start by reviewing in section 2 the development behind this change of roles in the sector. Section 3 focuses allocative efficiency and the need to account for the consequences for road users and for the rest of the society of a new piece of infrastructure and its maintenance; how can the agent be induced to take these effects into account. Section 4 addresses cost efficiency issues: The more of the responsibility for a project that is being transferred to a contractor, the greater is the chance that efficiency improves; but the higher responsibility also makes the contractor carry more risk, which is costly. Moreover, contracts are often incomplete, which may create incentives for costly renegotiation. The question is how these aspects are dealt with when contracts are initially designed. Section 5 concludes. In addition, appendix A formalises a simple welfare maximisation model for efficient construction and maintenance of new infrastructure.
2. Motives
There are several possible motives for the shift of roles between the public and the private sectors. A first reason is that PPP’s can be a means to enhance cost efficiency. This is so if the private sector is better at doing work that has previously been taken care of in the public sector. The principal-agent paradigm offers a motive. Within both sectors, any production process is plagued by incentive problems between “principal” and “agent”. In the public sector the parties may be the government and the head of an agency; in the private the shareholders and the manager of the firm. There are two reasons for these incentive problems: First, the respective parties have different objectives, and secondly, there are information asymmetries between them, making it possible for the more informed party, often the agent, to use this to his or her advantage.
There is reason to believe that it is easier for a private firm to handle these agency problems. One explanation is that the public sector principal tends to be more heterogeneous and dispersed; a private firm has one owner or a (relatively) homogenous board, while a public-sector agency formally is governed by a ministry but also has to account for its performance in parliament and in the public debate. The public-sector agency moreover has more ambiguous objectives with no single and clear measure against which to assess performance; a private firm is typically managed to maximise profits which can be relatively easier to monitor. The public sector agency is also more likely to face a soft budget constraint, meaning that it may be easier to make extra money available after budget overruns. An official who knows this may act differently, for instance be less prone to take painful decisions to cut costs, than if the budget constraint is absolute. Taken together, these reasons explain why it may be relatively easier for the private principal to discipline its agent into doing what it wants, in particular to be more efficient.
A second reason for using private companies to supply infrastructure services under contract with the public sector is that it may save on costs to make one and the same decision maker responsible for both construction and maintenance spending. This aspect is further developed in sections 3 and 4 below. So is also a third explanation, namely that a government may want to make users, not tax payers, foot the bill for a project and that it prefers having a private-sector agent take care of this task on its behalf.
A fourth motive can be identified when contrasting US and European transport sector policies towards financing road sector activities. Although there are differences across the continent, Europe seems to charge high taxes for all road use, with tolling used on an ad hoc basis. Taxes are levied both on fuel and vehicle ownership and increasingly also weight-distance taxes are being introduced (Parry & Small 2005). The revenue from taxation of road transport goes directly into the Finance ministry’s tax coffers rather than being earmarked for use in the sector.
America’s taxes are at a lower level but its fiscal paradigm is to earmark revenue from fuel taxes etc. for use in the transport sector, buy way of a road fund. There is, in addition, a political reluctance against jacking up these taxes, constituting a threat that the road fund will not suffice for maintaining and expanding the network in future.
The differences in taxation level could be expected to create a higher rate of traffic growth in the US than in Europe. The current US debate seems to be how to generate resources for future road construction, in the crunch created by constant tax levels and increasing fuel efficiency on the one hand and a growing traffic volume on the other. Recent legislative changes, and current actual examples, indicate a willingness to introduce road tolls as a means to solve the crunch. This could be interpreted as a price discrimination mechanism, i.e. to charge a premium rate for using uncongested sections of the network.[1]
A fifth possible motive for the development of PPP’s is well illustrated by the European Union’s Maastricht criteria. One of these fiscal rules caps the size of the public sector’s debt, and a PPP project may provide a means for reaching this objective. In particular, a private firm or a corporatised public body may take over debt and also raise capital for building the new infrastructure. It is thereafter reimbursed for these costs over a long period of time with tax revenue or with revenue from user charges. This provides a means for lifting off public debt and provides one reason for the current development of an Austrian commercial road agency and its interest also in PPP solutions (reference).
3. Allocative efficiency
Irrespective of the reason for interest in PPP contracts it is essential to design any deal between a public sector principal and a private profit maximising agent in order to provide infrastructure services so that social welfare is maximised. In particular, a contract that shall provide the agent with incentives for supplying efficient infrastructure services must account for the consequences of alternative investment and maintenance strategies for the future users of the new infrastructure and for society at large. If not, there is a risk that the agent minimises its financial costs without due regard for the consequences for the outside society. In this section we will establish how the contract must account for user and third-party consequences of the service provision.
There are two dimensions of relevance for others than the contracting parties. The first concerns whether the agent should be entitled to levy charges for using the services provided by the new piece of infrastructure. This is the demand side of the project, further described in section 3.1. Section 3.2 handles the cost side, i.e. the implications of alternative maintenance strategies for road-user and third-party costs, and how this must be dealt with in the contract. The discussion is based on the theoretical framework established in appendix A.
3.1 How should infrastructure facilities be paid for?
It is well known, and also shown in appendix A, that the welfare maximising policy is to charge the use of infrastructure facilities according to marginal costs. Users should pay for congestion, wear-and-tear as well as accident and environmental costs emanating from marginal variations in traffic volumes. Not least when a new piece of infrastructure is opened for traffic, capacity may be abundant, meaning that congestion is low. To the extent that the other cost components are recovered through fuel taxation, the toll charged for using the infrastructure should therefore be set to zero. Positive tolls would scare users away from the new facility in spite of that they add little to costs, hampering allocative efficiency.[2] This provides a motive for not giving an agent the right to levy tolls.
But the alternative to charging users in order to pay for an investment is to use tax revenue. This is also costly from a welfare perspective, since financing over the tax bill distorts consumption and production choices. The loss of consumer and producer surplus provides a measure of that loss which should therefore be compared to the loss of surplus as a result of tolling; if the efficiency loss from a toll is higher than the efficiency loss from marginal variations in the tax level, the project should be paid for by taxation, and the other way round.[3]
The size of the distortion from toll financing differs across circumstances. In particular, the presence of good alternatives to the priced facility is the key to understanding whether the choices of road users will be distorted, inducing them to use other “cheaper” routes. Several tolled bridges, such as the Öresund bridge between Sweden and Denmark, have fairly poor alternatives (reference) and so has many Norwegian toll roads where the topography makes bypasses long and costly (cf. Bråthen 2004). Tolls on these links may therefore have small distortionary consequences. Tolling of one link in a dense network, such as Hungary’s M1 motorway (reference) may, on the other hand, induce most traffic to use nearby alternatives, generating much higher social costs than if the motorway was not tolled. Putting focus on this substitution effect makes it feasible to discern on an early stage of an assessment whether a project may be eligible for user funding.
A different sort of situation could be envisaged when a motorway is built in a town etc. with high congestion on all or most streets. Under some circumstances it could be envisaged that tolling of the additional capacity if welfare-enhancing compared to the no-tolled-motorway benchmark. This is so if user benefits of the extra capacity, plus the benefits of reducing congestion on the non-tolled parts of the network, exceed costs for having it built. First best optimal would of course be to charge for short capacity in the whole network. Again, this illustrates that tolling may enhance efficiency under some circumstances while not in other.