Toll concessions were supposed to be the answer to America’s massive highway funding shortfall. Yet after 25 years of active promotion by the P3 Division of the American Road and Transportation Builders Association (ARTBA), Public Works Financing, the Reason Foundation, and others, the United States has but a handful of toll concession projects.
Because of the high risk of greenfield toll projects, the private sector’s appetite for those is weak, with a significant shift toward availability-pay concessions. But unless those projects also include toll revenues, they do not solve the underlying highway funding shortfall. And while brownfield toll projects are very attractive to companies and investors, they are few and far between.
The elephant in the room for large-scale toll concessions is reconstructing and modernizing the most important segment of U.S. infrastructure: the Interstate highway system. With just 2.5% of total highway lane-miles, this network handles 25% of all vehicle miles of travel. Over the next two decades, nearly all of those lane-miles will exceed their original 50-year design life, and will therefore need full reconstruction. And since many of those corridors are undersized for current traffic levels, let alone those likely over the next 50 years, many corridors will need additional lanes. Moreover, at least a hundred interchanges on the system are major traffic bottlenecks that need redesign and reconstruction.
Toll-financed reconstruction and modernization would amount to a $1 trillion hybrid (brownfield/greenfield) opportunity for toll concessions. As such, it would be far less risky than greenfield projects and should be straightforward to finance via private activity bonds and Transportation Infrastructure Finance and Innovation Act (TIFIA) loans plus private equity investment.
That $1 trillion cost estimate comes from a new Reason Foundation study released earlier this month. The study’s purposes were (1) to estimate, using Federal Highway Administration (FHWA) data, the cost of reconstructing the entire Interstate system, (2) to identify specific corridors requiring lane additions between now and 2040 and estimate those costs, (3) to identify corridors where truck-only lanes are warranted by projected truck traffic levels and estimate their costs, and (4) to estimate the feasibility of paying for this endeavor via toll revenue generated by all-electronic tolling.
The quantitative effort was carried out on a state-by-state basis. That included adjusting FHWA unit cost data (for reconstructing existing lanes and for adding new lanes) to account for cost differences among states, as well as using separate vehicle miles traveled (VMT) growth rates for each state, and for cars and trucks separately (since truck traffic is expected to grow much faster than car traffic). The study used state-specific VMT growth rates derived from a model developed at the US Department of Transportation’s Volpe Center.
To provide a baseline for comparison of toll feasibility among states, the basic calculations used the same toll rates nationwide: 3.5¢/mi for cars and 14¢/mi for trucks, with both rates adjusted annually for inflation. Tolling was modeled as not starting until a corridor was reconstructed, following the principle of Value-Added Tolling (don’t ask highway users to pay a toll until you give them something much better than what they are already using). The basic feasibility measure was a comparison of the net present value (NPV) of toll revenue to the NPV of cost, both in 2010 dollars. For the Interstate system as a whole, that ratio was 99%, using the baseline toll rates-an unexpectedly positive result.
The results differ considerably when looked at on a state-by-state basis. Using the baseline toll rates, 30 states have NPV ratios greater than 100%–and nine of those, mostly in the South, could use somewhat lower toll rates than baseline. Nine other states have ratios in the 80-90% range, suggesting the need for somewhat higher (but still quite reasonable) per-mile toll rates. Another six are highly urbanized states with very high reconstruction and lane-addition costs (including California and New York). They would need per-mile rates in the 6-8¢/mi range for cars and 24-31¢/mi for trucks-rates that are in the ballpark of what’s being charged on new toll roads. Finally, there are five states with low population, low traffic, and mountainous terrain (such as Montana and Wyoming). They would need toll rates comparable to or a bit higher than those in the six urbanized states, which might or might not be politically feasible in those rural locations. The truly problematic state is Alaska, which does not need any lane additions but whose ratio came out at only 24%–clearly not toll-feasible.
Overall, these are very positive findings. When the Interstate program was being debated in the early 1950s, tolling was proposed (given the success of the Pennsylvania Turnpike and the toll-feasibility of the other northeastern turnpikes then under construction). But traffic levels in the South, Midwest, and West were seen as far too low to make toll-financing work for the nationwide system, which is how we got federal fuel taxes and the Highway Trust Fund. But the massive shifts in population and economic activity since 1956 have dramatically changed what is feasible for toll finance.
How should this massive make-over, which we have dubbed Interstate 2.0, be carried out? One approach would be to have the federal government take the lead, substituting a new federal toll for some or all of the federal highway user taxes, and building in some redistribution of revenues among the states, to address the problem of Alaska and the five other mountain-west states. The other approach is for states to take the lead, each creating a master plan for phased reconstruction and modernization and setting toll rates for each corridor based on investment-grade traffic and revenue studies. In this case, the federal role would be one of granting permission to toll for reconstruction and continuing to impose basic uniformity requirements to ensure that the Interstates continue to function as a system for interstate travel and commerce. That would include ensuring nationwide interoperability for all-electronic tolling.
Since Congress shows no sign of embracing a major new federal Interstate modernization program or of providing any sustainable source of increased highway funding, the Reason study recommends that in the next reauthorization Congress remove the current federal prohibition on tolling “existing” Interstate lanes. Essentially, this would amount to mainstreaming the three-state pilot program that allows for just three projects to use toll finance for Interstate reconstruction. Toll revenues would be dedicated exclusively to rebuilding and modernizing (via lane-additions including truck-only lanes, modern design and safety standards, etc.) existing Interstate corridors. During the Moving Ahead for Progress in the 21st Century (MAP-21) debates, a bipartisan amendment along these lines nearly made it into the Senate bill, and with the federal budget in even worse shape today, it should have a better chance in the 2014 reauthorization.
The case for going this route is very strong. The Interstates need this trillion-dollar investment in order to continue to play their vital role in travel and goods movement over the next 50 years. And there is no realistic alternative on the table to fund this massive program. Survey data summarized in a 2008 Transportation Research Board special report shows that voters overwhelmingly prefer tolling rather than any number of increased-tax alternatives to pay for major highway projects. And if implemented using the Value-Added Tolling principles, nobody would pay a new Interstate toll until the modernized (rebuilt and expanded, if warranted) facility was finished and opened to traffic.
Another key argument for this approach is that it would be the first major step toward replacing per-gallon taxes with mileage-based user fees. Remember, the Interstates handle 25% of all VMT, so converting that system to all-electronic, per-mile tolling would bring about this historic shift for a major fraction of all car and truck travel. And it would do so without requiring any Big Brother-type gear in motor vehicles. Current all-electronic tolling (AET) relies on well-accepted transponders and license-plate imaging. And AET should be interoperable nationwide within this decade.
From the public-private partnership (P3) perspective, in states with long-established toll agencies, those entities would likely get first dibs on the Interstate modernization projects in their state. But in most of the country, the logical route would be P3 toll concessions. Since these projects would all be mega-projects, the risk transfer benefits alone would make the concession approach desirable. And states without toll-financing experience should welcome the ability to make use of world-class companies with long toll road track records.
Debates over the future of the federal highway program will probably go on for the next decade or more. We cannot wait that long to get going on Interstate reconstruction and modernization. That’s why the top priority of the entire highway community in the 2014 reauthorization should be to get permission from Congress for all states to reconstruct all their Interstates with toll financing.
Sidebar: Answering the Pisarski Challenges
In an extended email discussion among transportation leaders, transportation expert Alan Pisarski said of the Reason proposal, “I am certain we will end up there, sooner or later.” But he identified four places where he “stumbled,” coming up with difficult questions. Here are his concerns and my initial responses.-Bob Poole
1 Toll Siphoning
While acknowledging the “virtuous triangle” among road owners, bondholders, and road users committed to the tolls being only used for the highways on which they are charged (i.e., as true user fees like electric or telephone bills), Pisarski says he sees this principle being frequently undermined.
This problem certainly does exist, with the most recent example being the ruinous Act 44 in Pennsylvania. Clearly, in the federal enabling legislation, Congress needs to spell out that the toll revenues may be used only for the capital, operating, and maintenance costs of the Interstates in question. And while all the non-highway interest groups would clamor for carve-outs for their purposes, the simple act of giving states permission to toll for reconstruction would not change anything in the rest of the federal highway and transit program. And at the state level, state DOTs would have their work cut out for them to persuade highway users that the tolled Interstate modernization plan was truly in their interest. That case would be a lot stronger if it ensured that the toll revenues would be used only to make the highways in question better-i.e., if the toll was a pure user fee, rather than a general transportation tax.
2 Diversion of Traffic
Toll traffic and revenue studies (including this one) assume that X% of the traffic projected to use a highway without tolls would choose not to use it if tolled. The impacts of diverted traffic on other roads, on increased travel time, and on safety, are negatives that must be addressed.
Diversion is a real issue, but I think its impact is often exaggerated. First, states control the principal alternatives to Interstates, and have every right to mandate and enforce sensible speed limits and weight limits on those alternatives. Second, some communities may welcome the return of some traffic that would patronize local businesses along the alternative routes. Third, over the 20-25 year period of Interstate reconstruction via toll finance, states will be in the process of replacing their state fuel taxes with some form of mileage-based user fee for non-Interstate highways and streets. This will likely be via low-tech systems charging 1.0-1.5¢/mi. for cars. While those are lower than the baseline Interstate toll rates assumed in our study, the perception that the alternative routes are no longer “free” will likely reduce the extent of diversion.
3 Trust Fund Diversion
Establishing a separate system for the Interstates would leave the federal Highway Trust Fund with a pile of cash without a purpose. And that would lead to a general-purpose program dominated by transit.
Given the dire shortfall in Highway Trust Fund user tax revenues, states would vigorously defend retaining something like the current split between highways and transit, given their long-established dependence on federal funding for major portions of their highway programs, not just for their portions of the Interstate. In addition, deleting the Interstates from the Trust Fund’s obligations (except maybe for Alaska and a few mountain states) would greatly reduce the pressure to increase federal fuel tax rates, so the various modes would still be fighting to divide up a limited pot of money.
4 Alternative Trust Fund Future
If we got everything else out of the Trust Fund and focused it exclusively on reconstruction and expansion of the Interstates, there would be enough money from current fuel taxes, and we wouldn’t need the tolls.
Adrian Moore and I proposed something along those lines in our 2010 Reason study, “Restoring Trust in the Highway Trust Fund.” It gained no traction, and I think its prospects are far less likely than Congress giving states permission to toll. To be sure, if $35 billion per year for 25 years could be dedicated entirely to Interstate reconstruction and modernization, it would raise $875 billion, in the ballpark of what is needed. But that isn’t going to happen. Moreover, there is a powerful case that per-mile tolling is a much better user fee than per-gallon taxes (see pp. 10-12 of the new Reason tolling study). That much better replacement is within our grasp.
Robert Poole is the Searle Freedom Trust Transportation Fellow and Director of Transportation Policy at Reason Foundation. This article first appeared in the Public Works Financing newsletter.