I've received several queries about a Nov. 21st front-page story in the Wall Street Journal about financial difficulties at a number of U.S. toll-concession projects. Besides noting the bankruptcies of the South Bay Expressway in San Diego, Virginia's Pocahontas Parkway, and American Roads (a holding company for the Detroit-Windsor Tunnel and four Alabama toll roads), the article pointed to possible financial restructurings of the SH 130 toll road between Austin and San Antonio and of the Indiana Toll Road. In nearly every case the problem stems at least in part from overly optimistic traffic and revenue forecasts made prior to the start of the Great Recession in 2008. In several cases, the financial structure appears to be overly aggressive, with large payments coming due in the next year or two that exceed what is likely to be available from toll revenues and reserve funds.
Not all P3 toll roads are in such difficulties, which is hard to discern from the WSJ article. Not mentioned at all is the 91 Express Lanes in Orange County, CA—the country's first toll concession and one whose traffic and revenues remain robust (though it had been in operation nearly 13 years by the time the 2008 crunch began). Mentioned only briefly or not at all are the Chicago Skyway, the Dulles Greenway, the I-495 Express Lanes on the Beltway in northern Virginia, the Northwest Parkway in Colorado, and the 407ETR in Toronto. Several of those have less than projected traffic, but to the best of my knowledge, none is in serious distress—and the 407 in Toronto is thriving. In addition, though not a P3 concession, the Inter County Connector in Maryland recently completed its first year of tolled operations, with toll revenue almost exactly at forecast level, and 40,000 vehicles per weekday on average.
It's well-known that start-up toll roads are relatively risky endeavors, since highly accurate traffic and revenue forecasts are still more of an art than a science. And that's one reason I have long advised legislators and state DOTs against saddling taxpayers with traffic and revenue risk. One of the most important benefits of long-term toll concessions is shifting such risks (along with the risk of construction cost overruns, late completion, and operating & maintenance risks) to willing investors. I was interviewed about this issue by David Mildenberg for a Bloomberg article published on Nov. 27th, "Private Toll Road Investors Shift Revenue Risk to States." It documented the recent trend of companies that compete for mega-project concessions to push for availability-pay concessions rather than toll concessions. Both are long-term agreements in which construction, completion, and O&M risks are transferred, but in these newer concessions the company is paid by the state over the life of the agreement, with only minor variations depending on how available the lanes are 24/7 and what condition they are in. As Mildenberg's article points out, that leaves the largest risk—traffic and revenue--with the state, aka the taxpayers.
I've written about this point for several years, because one of the long-standing problems with US highway infrastructure is poorly justified projects that get approved and built more for political than for economic reasons. If a realistic projection of traffic and revenues doesn't come close to covering the capital and operating costs of a major highway or bridge, there's a serious question whether it's a wise investment. "Economic development" is usually trotted out in such cases, a kind of "field of dreams" premise. But let's face it: this country has a large and growing need for productive highway investment (such as rebuilding and modernizing the Interstates and building urban express toll networks) and a serious shortage of funding to meet those needs. Requiring such projects to pass a credible return-on-investment test is a way to separate the high-value investments from the cats and dogs.
To be sure, there are cases where, for policy reasons, not charging a toll on a new facility intended to divert traffic from other facilities can make sense (e.g., the Port of Miami Tunnel, being developed as a pure availability-pay concession). And there can be cases where a hybrid toll/availability structure is the best that can be done, due to specialized circumstances. But those should be the exceptions, not the rule. If this country shifts to a largely availability-pay model, we will lose a powerful means of ensuring the wisest allocation of inevitably limited highway investment funds.
Incidentally, Moody's early this month changed its outlook for toll roads from negative to stable. That change was based on the recovery of traffic growth, which Moody's projects as averaging 1.5% for 2014 for the toll roads it rates.