Layne’s Law

Map credit: VDOT

Proposed Interstate 66 improvements. Map credit: VDOT

by James A. Bacon

After spending a year and a half cleaning up the mess left by the previous administration — the Charlottesville Bypass, Norfolk’s Midtown-Downtown Tunnel, the U.S. 460 Connector in Tidewater — Transportation Secretary Aubrey Layne now has the opportunity to show whether or not he has the chops to handle a complex, politically charged transportation mega-project — the $2.1 billion in proposed improvements to Interstate 66 west of the Capital Beltway.

The Virginia Department of Transportation is focusing on two key dimensions to the mega-project, which is intended to address one of the most congested traffic corridors in Northern Virginia. The first is engineering. VDOT engineers have conceptualized a plan to build a corridor that will include three free lanes running both directions, two express lanes in each direction and high-frequency bus service  linking commuters with major activity centers. That plan will be subject to extensive environmental review and public input.

The second key question is how the state will finance and manage the project. Should VDOT use the controversial public-private partnership (P3) legal/financial structure to build and operate the project, much as it did with the Interstate 495 and Interstate 95 Express Lanes projects? Should VDOT undertake the project entirely on its own? Or should it create a hybrid approach? The question assumes tremendous urgency given the problems created by the P3 approach with the Norfolk-Portsmouth tunnels and the U.S.460 Connector.

Parenthetically, I would add that there is a critical third dimension to the project which appears to be getting less attention: What will be the impact of the project on Northern Virginia land use patterns? Will the project subsidize suburban sprawl (scattered, disconnected, low-density land use patterns) in the far western reaches of the Washington metropolitan area, which in turn will generate more demand — and more congestion — in the years ahead? In other words, will the state spend more than $2 billion to “fix” a problem that will re-emerge a decade or two after the project is completed?

In an interview yesterday, I sat down with Layne to discuss the financial dimension of the project. The transportation secretary has given enormous thought on how best to structure highway mega-projects, balancing the twin imperatives of controlling costs and limiting risk.

Layne sees transportation projects as forming a continuum between a model in which VDOT does everything, and a privatization model in which all functions are delegated to a private-sector concessionaire through an asset sale or a long-term lease. Any project is comprised of several parts:

Design
Construction
Financing
Operation
Maintenance

Any one of these pieces can be performed by the state or out-sourced to the private sector. In Layne’s view, because no two projects are identical, there is no standard configuration. He tends to think that the private sector does a better job of designing and managing construction than VDOT, but not by such a wide margin that jobs should automatically default to private players. In other areas, VDOT has a clear advantage. The fact that VDOT can tap tax-free bonds, which sell at lower interest rates, and requires no private equity financing often means that the state can finance a project at less expense than a private entity.

But there is more to managing a project than driving for the lowest cost or, as the previous administration tended to do, the lowest up-front public subsidy. Costs should be viewed over the lifetime of the project, which runs 50 years or longer, after adjusted for risk. One risk is the potential for cost overruns. In the past, VDOT used to eat construction cost overruns, which meant that the taxpayer paid. But that risk, says Layne, usually should be transferred to the private contractor. Another risk is that toll revenues might not materialize as forecast. Private entities pay especially close attention to that risk, and they try to negotiate all manner of protections into a P3 project to offset it.

Express lane projects get tricky because a private entity is motivated to maximize toll-paying traffic through its tolled lanes in order to maximize revenue, while the state’s goal is to maximize the throughput of riders, which means encouraging High Occupancy Vehicles and buses. In its negotiations over Interstate 95, the original plan called for Transurban to make a $250 million up-front investment to provide a robust mass-transit component. But as the negotiations proceeded, the mass-transit piece got whittled down drastically. If High Occupancy Vehicles exceed 24% of total traffic on I-495 (35% on I-95), the state is obligated to pay 70% of Transurban’s lost revenue. Meanwhile, the state pays for mass transit on I-95 out of its own funds — an expense that was not included in the announced project cost.

Layne’s Law is to start by first defining what the state wants in a project, and only ascertaining which mix of functions it should outsource to the private sector. “I’d love to have a private-sector partner,” Layne says, “but only if the risk transfer makes sense.” In the case of I-66, bargaining away a robust mass-transit capability to protect a private entity’s revenue stream probably would not make sense.

For the I-66 project, the state’s Office of Transportation Public Private Partnerships (OTP3) went through its usual procedure for determining a P3 structure for the project. Then the state hired public finance management consultants to determine, using the same inputs, how much it would cost the state to finance the project. Finally, Layne hired a third group to review the methodology and options proposed by the first two. This exercise showed that public financing with tax-free bonds and federal loan guarantees would not only be cheaper in the long run — saving $200 million to $500 million Net Present Value (NPV) over the lifetime of the project — but would require $400 million to $500 million less up-front public investment.

Preliminary analysis of I-66 financing options. Source: Secretary of Transportation.

Preliminary analysis of I-66 financing options. Source: Secretary of Transportation.

The state is scheduled to make a decision on how to structure the P3 deal in late summer. As it looks now, Layne says, “I-66 probably will be a P3 but not a concession [putting the private partner in total control of every step]. The private sector probably will build, design and operate it.”

Layne vows to keep “transparency and competition all the way through the deal.” If the state closes a P3 deal with a private party, the terms will be proprietary, says Layne, but he is obligated by law to personally certify that the terms are at least as good as if the state undertook the project itself. The new process laid out by P3 reforms law enacted by the General Assembly this year, he promises, will “stop another U.S. 460 from happening.”

There are currently no comments highlighted.

3 responses to “Layne’s Law

  1. FYI – you said: ” If High Occupancy Vehicles exceed 24% of total traffic, the state is obligated to pay 70% of Transurban’s lost revenue. Meanwhile, the state pays for mass transit on I-95 out of its own funds — an expense that was not included in the announced project cost.”

    The Contract says:

    “Section 5.07 Revenue Risk Related to Traffic Volume
    (a) Except for its specific obligations to the Concessionaire under the terms and conditions of this Agreement, the Department will not have any risk or liability related to actual traffic volume and revenue, including but not limited to the risk that actual traffic volume is less than the traffic volume projected in the Base Case Financial Model.

    (b) (i) From the period beginning on the second anniversary of the Service
    Commencement Date to December 31, 2030 (the “First Measurement Period”), the Department will pay to the Concessionaire amounts equal to 70% of the Average Toll for the number of High Occupancy Vehicles

    exceeding a threshold of 35% <<<<<<<<<

    of the total flow of all Permitted Vehicles in

    two consecutive Toll Sections that are then using such Toll Sections going in the same direction

    http://goo.gl/oqkuvQ page 13

  2. that’s in favor of the state, right? That’s a pretty hefty threshold before the state pays but it’s actually even more problematical given how it is measured and assessed. Basically there has to be substantial and continuous HOV above the 35% threshold.

    So you have a polynomial equation type problem involving real dollars.

    you have the “free” mainline
    then you have the “free” HOV
    then you have the variable toll SOV

    these three will vary in not entirely predictable ways but the concessionaire has to also provide a minimum performance – i.e. minimum speed – no matter the congestion -and the way it does this is by making the toll much more expensive when congestion starts to tick up.

    that also means if more people HOV -that there is less capacity for tolled SOV and the toll would go up also – not entirely in concert with mainline congestion but more so with HOV use.

    I would assert that this is not near as much a transportation problem as it is a heavy-duty math and calculus problem – and even then it’s still got an aspect of Art versus Science to it and it’s not something the very best transportation engineer is trained to do.

    So I’m not at all sure – that VDOT has the in-house talent to sit down and figure out how a variable toll road will work and especially so in figuring out how much money it would generate – relative to the cost and bond service.

    To give a couple of examples. Where did the 35% come from? why not 22% or 42%?

    on US 460 – how was it determined that there was no way to pay for it with tolls? What’s the analysis procedure that leads one to that conclusion – and is it something VDOT could do itself – either for it to build and operate a toll road OR for it to be able to analyze private sector PP3 proposals?

    You get two proposals. One of them says the road even tolled will have a deficit but is that number the same number as a competing proposal? not likely.

    and back at the front – how can it be said that Hampton and Virginia “need” a transportation corridor to move goods from the ports – but at the same time say that the shipping companies will not find it cost-effective to use a toll road verses available “free” alternatives?

    so “risk” is not some simple equation that numbers are cranked into and an answer pops-up – so I’m not at all convinced that VDOT has the expertise to do that kind of analysis.

    probably better – would be to have a 3rd party that has the expertise to analyze P3 proposals and that 3rd party would be an full-up investment firm.. capable of delving into all the nooks and crannies to come up with a prediction that is reliable enough to make decisions about.

    I suspect this is something that Mr Lane and company are pondering these days in general and specifically with I-66.

    Should VDOT develop in-house talent to do these kinds of analysis?

    isn’t this sort of where govt meets business when trying to determine what a value proposition is? If VDOT and Virginia are concerned with how much money they’d have to come up with – if predictions are wrong – isn’t that moving from a govt perspective to the business realm?

Leave a Reply