Encouraging the private sector to generate innovative ideas can have merit. When a private party approaches the PWA with the idea for an infrastructure project and the idea is good, however, the question becomes how to structure a workable procedure. This requires the development of mechanisms for compensating the private parties for their ideas without affecting the transparency and efficiency of existing PPP awards. (See Hodges and Dellacha 2007, for details on unsolicited proposals.) Countries that have developed systems for receiving unsolicited proposals must deal with large numbers of proposals, running into the hundreds in the case of Chile, South Korea, and Taiwan, the countries with the most mature systems.
One possibility is for the PWA to contract with the proponent to develop the project as a PPP, but the lack of competition and transparency make this option unattractive.
The alternative is to design a clear-cut mechanism for remuneration. The first stage consists of the approval or rejection of the unsolicited proposal, according to clear guidelines (in particular, excluding obviousness). Once an unsolicited proposal is approved, there are various options that have been used to remunerate the proponent. In some countries, the proponent has an advantage in the competitive auction for the project (or the proponent can transfer its option). Its bid is chosen if it is no more than say, 5 or 10 percent of the best bid. In other countries, the proponent can match the best offer. The problem with these approaches is that the advantage possessed by the proponent may detract from participation in the auction, and therefore lead to projects awarded with little competition.
The alternative that we espouse is to separate the proposal stage from the award stage. Each year only a small number of proposals should be chosen by the PWA, rewarding the selected proponents with a fixed prize that is sufficiently attractive to attract good projects. The prize would be paid by the PWA, but it would be reimbursed by the winner of the project once it is awarded under standard competitive conditions. This proposal combines incentives for competition in unsolicited proposals but does not alter the competitiveness and transparency of the award process.
BOX 2
First Present-Value-of-Revenue Auction
The Route 68 concession, joining Santiago with Valparaiso and Vina del Mar, was auctioned in February 1998. It was the first road ever to be franchised with a PVR auction. The Route 68 concession contemplated major improvements and extensions of the 130-kilometer highway and the construction of three new tunnels. Five firms presented bids, one of which was disqualified on technical grounds. For the first time in the Chilean concessions program, minimum traffic guarantees were not included for free, but instead were optional and at a cost. Tat the pricing of guarantees by the government was not way of the mark can be inferred from the fact that two of the bidders chose to buy a guarantee, while the winner declined. Bidders could choose between two rates to discount their annual incomes: either a fixed (real) rate of 6.5 percent or a variable (real) rate given by the average rate of the Chilean financial system for operations between 90 and 365 days. A 4 percent risk premium was added to both discount rates. Tree firms, including the winner, chose the option with a fixed discount rate. Somewhat surprisingly, the PVR demanded by the winner turned out to be below construction and maintenance costs estimated by the ministry of public works (Ministerio de Obras Públicas, or MOP): the winning bid US$374 million while MOP estimated costs to be US$379 million. One possible explanation for this outcome is that the regulator set a risk premium (and hence a discount rate) that was too high, neglecting the fact that PVR auctions substantially reduce the risk faced by the franchise holder. A return on capital in the 10-20 percent range is obtained if a more reasonable risk premium (in the 1-2 percent range) is considered.
It is also interesting to mention that, apart from the pressure exerted by the ministry of finance, the main reason why MOP decided to use the PVR mechanism is that it facilitates defining a fair compensation should the ministry decide to terminate the franchise early. This feature of PVR is relevant in this case since MOP estimates that, at some moment before the franchise ends, demand will have increased sufficiently to justify a substantial expansion of an alternative highway (La Dormida) that competes with some sections of Route 68. Thus, the contract of the Route 68 concession allows MOP to buy back the franchise at any moment after the twelfth year of the franchise, compensating the franchise holder with the difference between the winning bid and the revenue already cashed, minus a simple estimate of savings in maintenance and operational costs due to early termination. As pointed out in the main text, no such simple compensation is available if the franchise term is fixed. (see Table 3). Portugal also recently adopted flexible-term contracts for all its highway PPPs.
A second advantage of a PVR is that it allows for nonopportunistic renegotiations of contracts. Indeed, an advantage of PVR contracts is that they provide a natural fair compensation, should the PWA decide to terminate the franchise early. It suffices to add a clause allowing the PWA to buy out the franchise by paying the difference between the winning bid and the discounted value of collected toll revenue at the point of repurchase (minus a simple estimate of savings in maintenance and operations expenditures due to early termination). No such simple compensation is available if the franchise term is fixed.
A third advantage is that the flexibility incorporated into PVR contracts is convenient for urban highways. Setting the appropriate ex ante toll for these projects is a complex task. Unless traffic forecasters are unusually accurate in their estimates, the resulting tolls are likely to be incorrect-either so low that they create congestion or so high that the highway is underutilized. In the case of the Orange County SR-51 HOT lanes, fees responded directly to congestion, but this made the franchise holder reluctant to consider expansions for the untolled adjacent road, leading to excessive congestion. In a PVR franchise, the regulator could set tolls efficiently to alleviate congestion without distorting the incentives of the concessionaire (although the regulator must take care to ensure that the tolls generate sufficient revenue to pay for initial capital expenditures).
A fourth advantage of a PVR approach is that it reduces the likelihood of bad faith renegotiations. Traditional fixed-term infrastructure contracts are often renegotiated by either extending the length of the concession, increasing user fees, providing a government transfer, or combining these approaches. Extending the concession term in a PVR contract is not possible because, by definition, the term is variable. Increasing user fees is ineffective because it shortens the concession term without increasing overall income. Government transfers are not logically impossible under a PVR but, because the partner cannot claim that it will receive less user fee revenue than it expected, a government transfer would be difficult to explain to the public. Furthermore, to the extent that firms are more likely to act opportunistically under financial duress, PVR contracts reduce the incentives firms have to lobby for renegotiations, since scenarios with losses for the firm are less likely under a PVR. Yet both fixed-term and PVR arrangements do not deter renegotiations that involve building additional infrastructure, which motivates the proposals we make below to improve PPP governance.
Although PVR schemes have a big advantage in terms of facilitating good faith renegotiations and deterring bad faith renegotiations, as well as reducing risk, they have a downside: the PPP franchise holder has fewer incentives to manage demand for the infrastructure project because any action that increases demand will shorten the contract term. Projects earn their income regardless of the concessionaire's efforts. By contrast, demand-increasing investments are more attractive under fixed-term franchises. This suggests that the PVR method is applicable in cases in which quality of service is contractible and demand for the infrastructure is inelastic to the actions of the concessionaire-that is, when demand is mainly exogenous. Another important assumption underlying our analysis is that major investments are not needed frequently. Thus, port infrastructure (not operations), water reservoirs, airport landing fields and highways are natural candidates for a PVR, while mobile telephony is not.