As editor Bill Reinhardt pointed out last month, there’s been a trend toward availability-payment (AP) concessions for U.S. highway projects the last several years. Does this mean demand-risk concessions have little or no future in U.S. highways?
I think such a conclusion is unwarranted, for a number of reasons. First, let’s get some perspective. Using the data in last month’s issue, since 1995 there have been 34 U.S. highways with long-term concession agreements (both greenfield and brownfield), with a total value of $36.75 billion. Of that total, 22 were revenue-risk, seven were AP concessions with toll revenues and five were pure AP concessions. The revenue-risk concessions account for 66.3% of the total investment.
Second, only four of those 22 revenue-risk deals have filed for bankruptcy, though one or two others are having trouble meeting their debt service payments due to below-forecast traffic and revenue. And the picture is no worse in other countries with sound legal frameworks that emphasize risk transfer as a key benefit of toll concessions. Australia has had four bankruptcies out of 16 toll concessions and has done only two AP concessions thus far. France, which pioneered toll concessions back in the 1960s, had no trouble selling its state-owned toll operators in 2004-05, and it continues using toll concessions even for high-risk mega-projects like the A86 West tunnel outside Paris and the world’s highest bridge, the spectacular Millau Viaduct.
Global infrastructure investment funds are eager to find more good transportation projects to invest in, and generally prefer a mix of brownfield and greenfield projects with a range of risk/reward trade-offs. There is no shortage of bidders for the bankrupt Indiana Toll Road concession, which can be refinanced less aggressively post-bankruptcy.
And while it is certainly true that more contractors will bid on AP concessions than revenue-risk ones, I count at least six global firms actively pursuing revenue-risk concessions in the United States in 2014. Pure greenfield toll projects may be too risky for most U.S. firms today, but if we can get Congress to allow states to toll-finance the reconstruction and modernization of their Interstates, most of those projects would be only moderate-risk as toll concessions, compared with greenfield projects. So I see a bright future for revenue-risk concessions in this country, since toll-financing is the only realistic way to pay for this much-needed trillion-dollar Interstates makeover.
The other side of the coin is that the potential market for AP concessions-especially of the pure AP variety-is inherently limited. To begin with, the stream of availability payments that a state government commits to over 35 to 50 years is a liability that must be recorded on its balance sheet. That is one reason why some state legislatures have not authorized AP concessions. It is why others, such as Florida and North Carolina, include AP liabilities in their annual calculations of how close they are to exceeding legislative or constitutional debt limits. Few state DOTs want their state to end up like New Jersey, where essentially all of their annual transportation capital budget must now be spent on debt service. Hence, responsible state governments will impose limits on the extent to which they can take on AP liabilities. As states make greater use of P3s for non-revenue-producing projects, highways will be competing against courthouses, schools, and many other sectors for a limited pot of AP debt.
There is also the question of appropriation risk, for which Portugal provides a dire object lesson. Despite previous success with toll concessions, in the 1990s its government switched to shadow tolls and availability payments, in order to do more projects. But by 2006 (prior to the global financial meltdown), its government decided that the liabilities were unaffordable; a federal judge estimated the liability at $65 billion, equal to 30% of Portugal’s GDP. Since 2010, under pressure from the International Monetary Fund, the EU, and the European Central Bank, Portugal has been phasing down availability payment amounts and requiring the concession companies to implement tolling.
I appreciate that a pure-AP highway concession offers efficiency gains compared with a plain-vanilla design-bid-build or design-build project-in particular, a design that will minimize life-cycle costs and ensure proper maintenance over the life of the concession. But what is lost in avoiding revenue-risk concessions is the use of toll feasibility as project screening tool. One of the downsides of our somewhat politicized highway system is the tendency of elected officials to pressure state DOTs to do projects whose costs exceed their benefits. Requiring large projects to undergo a toll feasibility analysis, and only proceeding with them if they come out looking pretty good on that score, provides a powerful way to filter out boondoggles (such as several recent projects originally proposed, but found not feasible, as toll concessions). As Skanska’s Karl Reichelt put it in the February issue of the Public Works Financing newsletter, “While rightly applicable for certain projects, AP projects don’t yield the full value proposition of true, market-risk PPPs.”
The Value for Money analysis that we urge state P3 units to adopt is based on quantifying risk transfer, and revenue risk is a major component in those calculations. Without the transfer of revenue risk to investors, the value of doing the project as a concession is considerably reduced, meaning fewer deals will emerge from Value for Money analysis to be done as concessions.
Finally, I’m increasingly worried about the political vulnerability of P3 concessions. As I’ve discussed in recent columns, concessions are under concerted attack from both the populist Right and the progressive Left. Most of the arguments they use are bogus or at best highly misleading. Certainly, the claim that concessions involve “government guarantees” is false for toll concessions. But when these people figure out how AP concessions work, I fear they will characterize that 35-year stream of payments to the concession company as “guaranteed revenue.”
My aim here is not to denounce AP concessions, but to point out that they should not become the sole or the main vehicle for private investment and management of highways. They have a role to play where tolling would be counter-productive to a project’s purpose (Port of Miami Tunnel) or for rebuilding hundreds of smaller bridges where tolling is not realistic. But for projects with serious toll potential (but not quite 100%), I think it makes better sense for the state DOT to invest a bit more of its own money into the project up-front and finance the balance as a toll concession, rather than putting taxpayers at risk for traffic and revenue, as this country moves increasingly towards per-mile tolling for major highways and bridges.
Robert Poole is director of transportation at Reason Foundation. This column first appeared in Public Works Financing.