Adam Smith mentions that when infrastructure is privately provided and sustained with user fees, a market test filters out the waste: "When high roads are made and supported by the commerce that is carried on by means of them, they can be made only where that commerce requires them" (Smith 1776, V.1.III.1).
This filter works only when PPPs are financed mainly with user fees. Projects that are not expected to be profitable (and therefore are not socially valuable in many cases) will fail to attract a concessionaire. Financing capital expenses with user fees may lead to charges that are higher than socially optimal, an outcome that can be avoided under public provision. The large number of infrastructure projects that are evidently wasteful suggests that the benefits of having a market test that avoids overengineered (or outright unjustified) projects is likely to outweigh these costs. PPPs will not filter such projects out if they are financed with subsidies or if there is an implicit guarantee that the government will bail out a troubled concessionaire. This is the reason for using cost-benefit analysis for most infrastructure projects (with the exception of those fully financed by user fees). In the United States, many federal infrastructure projects do not go through a process of cost-benefit analysis, which explains the "pork barrel" projects that are valued by the federal legislature. Yet, as noted, such projects need generous government subsidies to be attractive to private firms, since user fees alone will not suffice.
Various public projects with private partners in the United States have gone bankrupt. The South Bay Expressway in San Diego, California opened in 2007, but fled for Chapter 11 bankruptcy in 2010, citing traffic at less than 40 percent of initial projections. Similarly, the Camino Colombia Toll Road in Texas was foreclosed by a district court in 2003-the only such case in the United States-due to vastly overestimated demand: effective revenues were only 6 percent of projections.
The Greenville Southern Connector in South Carolina fled for Chapter 9 bankruptcy in 2010. A demand forecast study predicted $14 million in revenue by 2007 while actual revenue only reached $5.4 million. That forecast failed to notice that the road made no sense as an access road to local commercial developments. Traffic barely justified a two-lane road, let alone the four-lane expressway that was actually built, suggesting this project was unnecessary.
This is one of three projects that have gone bankrupt, associated with so-called 63-20 nonprofits that benefited from tax exemptions, in which the promoters had nothing at stake. According to the TOLLROADSnews newsletter, these projects were enthusiastically promoted by a combination of consultants, engineering firms, financiers, and construction firms who made money at the expense of bondholders during the development, design, and construction phases, and who had nothing at stake thereafer.10
There are several ways in which lower-income users benefit from the existence of new or improved toll roads.
The evidence we could find from public sources suggests that demand turned out to be higher than forecasted for only one of the twenty projects in Table 2. It is therefore likely that firms have incentives to overestimate demand for PPP projects. One reason may be to profit at the expense of bondholders, as occurred with South Carolina's Southern Connector. Another reason could be an implicit agreement that the concessionaire will be bailed out by the government should demand be much lower than expected. As we discuss above in the section "Renegotiations and the U.S. Experience," concession terms have been extended and tolls raised to help concessions with revenues below projections.