Transportation Tuesday is the newest in a series of Georgia Public Policy Foundation policy briefs. Others are Medical Monday’s Checking Up On Health and Tax and Spend Tuesday.
The Transportation Tuesday post of October 27, 2020, shared an article from Cato Institute’s Regulation magazine that discussed an alternate – and, we thought, intriguing – approach to transportation public-private partnerships (P3s). You can read our post here. It elicited a response from our friend and Senior Fellow Bob Poole, Director of Transportation Policy at Reason Foundation. The response, published in full below, first appeared in the November edition of Public Works Financing.
A New Challenge to U.S. Highway P3 Concessions
By Bob Poole
The Cato Institute and the Georgia Public Policy Foundation are free-market think tanks that have supported both express toll lanes and long-term P3 concessions for transportation projects. Yet both have recently published articles that question P3 concessions for highways.
The six-page article in Cato’s Regulation magazine (which GPPF quoted from) includes statements such as that “PPP’s are routinely renegotiated” after the winner has been selected, suggesting a kind of crony capitalism. It also says that the only real benefit of these P3s is lower maintenance cost, and that the tax cost to the government is the same whether the project uses availability payments or tolls, because presumably the government would have pocketed the toll revenues (as if most governments had the political will, on their own, to expand the use of tolling). And even worse, the article claims that if the project is financed based on revenue risk borne by the company, it will not be financeable without government revenue guarantees (which it also says are common).
Many of those points do apply to Latin America, where renegotiations are common, and governments often do provide minimum revenue guarantees. The authors are respected Chilean economists who raise concerns about the P3 experiences of Latin American countries and propose their own solution: the variable-length concession. The bidders for a highway concession would each propose the minimum net present value of revenue they would accept to do the project, and the winner is the one bidding the least. The concession would go on for as many years as it took for the company to achieve the rate of return it had agreed to accept.
This kind of P3 concession has been used for some projects in Chile but has apparently not spread very widely. Proposing it for countries like Australia and the United States that are successfully transferring revenue risk to investors suggests solving a problem that does not exist in these two countries. Yes, there have been several bankruptcy filings of P3 highway megaprojects in both countries, but that has been part of the learning process for a new industry. And those bankruptcies have demonstrated that revenue risk really is being transferred, and taxpayers are being spared from bailing out companies that badly miscalculated.
In 2017 the editor of this newsletter and several colleagues published a research paper assessing various ways that revenue risk can be addressed in P3 concessions. They include the availability payment model in which government takes all the revenue risk (assuming tolls are used), the variable-length concession, minimum-revenue guarantees, and some form of revenue sharing. They note that no previous research has taken into account the lenders’ interests; their model studies the effect of each risk-minimizing method on the borrowing capacity of the project. The end result of their modeling is that the least-bad approach to reducing the revenue risk of a concession is a small minimum revenue guarantee and a high potential upside for the company — a kind of revenue sharing.
Their model explicitly finds that the variable-length concession “does little to increase project leverage and hence reduce the cost of financing.” But even more important, by providing essentially a guaranteed rate of return to the investors, this form of concession takes away the incentive to earn a higher return by doing a superb job of attracting toll-paying customers and delivering high value to them.
In a toll road project done as an AP concession, the company has no incentive to design and operate the road in a way that maximizes convenient opportunities for would-be customers to enter it or seek to win their ongoing loyalty. Their only two tasks are to build the project competently and on time and then to maintain it to the state DOT’s standards. That’s why this form of concession company can be described as a construction company plus a contract maintenance firm.
By contrast, a revenue-risk concession is a highway business, which has the kind of direct customer-provider relationship that you have with your cell-phone company and other service providers. And that is what is sorely lacking in the 20th-century U.S. highway model we still maintain (except for the few toll roads). Motorists and truckers pay both federal and state fuel taxes that are still called “user fees,” but politicians now spend them on as many other things as they can think of, to gain votes. Highway “users” have no idea what they pay in gas taxes, or whom to hold responsible when roads they use are in poor condition.
Highway “users” are not customers of state DOTs, because the fuel taxes they pay don’t go to the DOT; they go to the government’s treasury. The DOT’s actual customers are the legislators who decide how much money the DOT has to spend each year, and what it must be spent on, often short-changing proper maintenance in order to fund more ribbon-cutting opportunities in members’ districts. AP concessions keep most of that model in place, except for guarantees of proper maintenance.
To be sure, some worthy projects have high enough revenue risk that they can be difficult to finance solely on projected toll revenues. U.S. revenue-based P3 projects thus far have averaged about 14% state equity investment, along with an average of 29% private equity, a 27% TIFIA loan, and the rest either bank debt or tax-exempt PABs. In Texas and Virginia toll concessions, a revenue-sharing mechanism is built in, under which the state stands to gain a share of revenue if it rises above the baseline agreed to at the outset. TIFIA and modest state equity with revenue-sharing is a far better way of reducing the extent of revenue risk while preserving the powerful incentives it provides to attract and please toll-paying customers.
Transportation Tuesday is compiled by Benita Dodd, vice president of the Georgia Public Policy Foundation.