The main dilemma of public-private partnerships is the extent to which the division of labor and responsibility between public and private spheres delivers the optimal infrastructure service or public good. Privatization has often been viewed as an excessive response to the inefficiencies of public sector, even when privatization comprises stringent regulatory environments (Martimot and Pouyet, 2007).
The literature stresses the trade-off between efficiency and rent extraction when the regulated firm has an informational advantage. Privatization reduces the information the government can access about the firm, thus makes privatization and asymmetry of information to raise firm's incentives to invest. Contract design and risk transfer in PPPs is on incomplete contracting. Market relations are problematic when they require relation-specific investments while taking place in a complex environment. This complexity makes contractual incompleteness unavoidable, leading to underinvestment by one party in the relationship due to fear of ex-post hold-up. Higher investment by one party can trigger tougher ex post bargaining by the other party, which is tempted to grab a share of the surplus generated by the investment. Asset ownership is assumed to confer residual rights to control over the asset and it motivates individuals to invest by giving them bargaining power ex post (since they retain control over the asset they own in case of disagreement) and thus higher returns of investment.
PPPs can mitigate or exacerbate inefficiencies arising from asymmetric information. Iossa and Legros (2004) present a general model of regulation under soft audit information, but it can easily be applied to PPPs. There are two periods, and one can think of the first as the building stage and the second one as the operating stage of a given infrastructure project. The public sponsor of the project can allocate the right to operate to the builder or another agent (e.g. an entrant). An entrant does not have information about building effort in the first period but may acquire it but doing costly auditing. In these circumstances, it is optimal to elaborate a contract that gives the potential entrant the option to buy the right to operate the project in the second period and to have the builder (e.g. first-period agent) receive monetary payment from the operator in this case. If the entrant does not exercise his option, the builder continues to operate the asset. In this way the willingness to exert the right produces the desired efficiencies.
Other authors (Bentz et al., 2002) stress the importance of whether operating costs are high or low. When costs are high the operator exerts effort during the building phase with a view to cutting operating costs. This leads to two types of inefficiencies in contracting. First, since the operator privately knows whether operating costs are high or low, even an optimal operating contract will leave an informational rent with the operator. Second, since there is moral hazard at the building stage with positive probability, the builder needs to be given incentives to invest in the reduction of operating costs. This arises from an increase probability of lowering operating cost which is relatively small compared to the required investment.
PPP then allows the ability to contract with only one agent, the consortium, who builds and operates the infrastructure. Without a PPP, the (risk-neutral) builder contracts without private information and, as a result, has to be paid only for the expected cost of the investment; but the informed operator gets an informational rent to reveal operating cost, enabling him to give incentives to the builder.PPP stimulates both the socially desirable and the undesirable investment, when is a PPP better than traditional public procurement? Hart concludes as follows: (i) if the cost-cutting, quality improving investment e can be verified, traditional procurement is better than a PPP because the investment is optimal and the initial financial requirements can be set at the efficient level by contracting with the builder; (ii) instead, if the cost cutting though potentially quality-shading investment i can be verified, PPP dominates traditional procurement since the contractor has better incentives to carry out the quality-improving investment.
Another source of contractual incompleteness that may influence the choice of procurement is a possible change in the objectives of the parties. To illustrate: a highway from point A to point B could be viewed as the right design when the sponsor invites contractors to bid for the project, but large migration or relocation of industries may make a highway from B to C the better design when project implementation starts. This could suggest that transferring ownership to the contractor reduces the risk associated with a change in objective of the sponsor. In that sense, transferring ownership to the contractor under a PPP could provide protection against political risk or 'soft objectives'. However, this view assumes that there is no way to protect ex ante the contractor under traditional procurement against such a change in objectives. Then, PPPs have potential to only help avoid cost overruns resulting from soft objectives.
Cost overruns are due to the desire of firms to get the contract and, therefore, to bid well below what they expect the cost of delivering the project to be (Flyvberg et al. 2003). Changes in the design of projects after awarding contracts often explain cost overruns. Besides this ex ante inefficiency, there will also be an ex post inefficiency if bargaining takes place under asymmetric information, a case of concern for PPPs. For instance, the sponsor may not know the quality of the investment made by the contractor before the bargaining starts, or the contractor may not be aware of future budgetary constraints that are known by the sponsor. In this case, the efficiency of bargaining will depend on the outside options of the parties, which are given by the initial contracting terms. There is then a trade-off between specifying many contingencies in the initial contract - which is usually positive in the absence of renegotiation - and modifying the outside options of parties - the effect of which is ambiguous when renegotiation cannot be prevented and when there is still asymmetric information. The issues of transparency, specification of strict performance targets, and completeness of contracts are key elements of the European Directive for public work contracts.
Project design, cost overruns, and competition Suppose there are two firms A and B that can bid for procurement of a public service. Like in Box 3, consider two potential designs for a highway: A and B. But now suppose that firm k (A or B) is specialized in design k: it has cost c of procuring design k and cost c + a (where a stands for 'additional cost') for procuring the other design. The sponsor has valuation v for having the right design and zero for the wrong one. Ex ante, there is an equal chance that the sponsor will prefer design A. The sponsor can invest ex ante to learn about his preferred design, for instance by doing macroeconomic simulations of labor demand, collecting information about the reliability of different designs, etc. We assume - to take the case most difficult for us - that this ex ante investment is not costly for the sponsor. Contractors know whether the sponsor knows his best design and then bid for the contract. If the design is not specified and firm A is selected, there is a 50-percent chance that the sponsor learns that design B is best: we assume that renegotiation enables the firm to extract a compensation r from the sponsor (obviously r ≤ v). If the sponsor knows his preferred design he specifies it prior to the bidding stage. We consider both situations in turn. Suppose the design is not specified: the two firms are in fact in a symmetric situation since they each face renegotiation with probability ½. Competing for the contract leads them to offer a price p = c + ( a - r ) / 2 and the sponsor obtains v - r / 2 - p = v - c - a / 2. We now turn to the other situation. If the sponsor specifies the design prior to bidding, the two firms are no longer symmetric: for instance if design A is specified, firm A can bid slightly less than p = c + a and gets the contract. Firm A obtains a rent of a while the sponsor has a payoff of v - c - a, which is strictly less than what he gets when the design is not specified. Hence, even if there are no costs of specifying the design, it is better for the sponsor to keep the firms in a symmetric situation: this increases competition for the contract between the firms, though the sponsor must accept an increase in the likelihood of renegotiation or cost overruns. Source: Adapted from Dewatripont and Legros (2005) |