I appreciated last month’s defense of Availability Payment (AP) concessions in Public Works Financing, and agree with much of what the authors had to say.
Design/build/finance/operate/maintain (DBFOM) is a major advance over traditional design/bid/build which focuses only on the lowest bid to build a standardized design. It’s also a big improvement over design/build since it focuses on lowest life-cycle cost. I also agree that since either form of DBFOM is better than older procurement methods, there are cases where AP concessions, all things considered, can be a good choice for a transportation agency; I provided examples in my November 2017 Reason policy paper on the subject.
My difference with last month’s authors concerns only major highway projects financed with tolls. For that subset of all highway projects, I maintain that revenue-risk (RR) concessions have strong advantages that will end up with this being the mode of choice in the future. Let’s get down to specifics.
First, my critics maintain that the public will pay too much for transferring revenue risk to investors. Many professional experts disagree with that assessment. World Bank economist Michael Klein says having taxpayers bear revenue risk is asking them to provide “unremunerated credit insurance” for the project. A proper value for money (VfM) analysis seeks to quantify the value of revenue risk transfer, to make transparent how much “credit insurance” the taxpayers would be taking on in such cases.
A variant of this concern was expressed as follows: “All else being equal, [the] higher cost of capital [due to a larger amount of equity] increases the cost to the public via higher tolls to cover this difference.”
That sounds plausible, but traffic and revenue forecasters know their Economics 101. The price charged is based on market demand, not underlying costs. Other things equal, the demand to use a tolled corridor is the same, regardless of how it is financed. So the RR concession company cannot charge more than the market will bear, any more than the state department of transportation (DOT) can as the tolling partner in the AP concession.
The RR company, therefore, has an economic incentive to increase the volume of paying customers, to generate enough revenue from market-based tolls to cover its debt service and the target return on equity. This means the RR concession company seeks to offer innovative designs that increase traffic and revenue. A number of recent RR cases involved design innovation and added connectivity that produced large differences between the winning bid and the second-place bid—for example, a $1.3 million difference for the LBJ Express in Dallas and $300 million for the SH 288 project in Houston.
I referred obliquely to this in my original column, by noting the RR company’s incentive to attract more traffic than an AP company. The AP concession company’s compensation depends solely on keeping the facility in good condition. It has no incentive to attract additional vehicles since it gains nothing financially by doing so, and more vehicle miles traveled (VMT) than forecast will increase its maintenance costs.
My critics question the public benefit of a project attracting more VMT. The gains to the RR concession company (increased revenue) are obvious, but what about the costs and benefits to the public?
First, if the tolled project attracts a larger fraction of the VMT heading in that general direction, this reduces traffic diversion from a roadway that is tolled onto parallel non-tolled routes. That is one clear benefit. Second, if part or all of the toll facility is variably priced, the more VMT it attracts, the larger the fraction of vehicle traffic in that general direction that will be uncongested, thereby reducing the higher emissions generated from congested traffic. As for potential safety impacts of handling more VMT, the concession agreement presumably contains the same safety requirements regardless of whether the concession is structured as AP or RR.
Another concern is that the state gives up too much control of future transportation infrastructure by agreeing to RR concession agreements because the RR concession company usually negotiates some form of compensation clause in case the state builds competing facilities. (Since the AP concession company’s availability payments do not depend on the amount of toll revenue the state DOT collects, there is no need for compensation clauses in AP concession agreements.) Compensation clauses have evolved over time. The first one, for the 91 Express Lanes in Orange County, CA, flatly forbade Caltrans from building any other lanes in the corridor. That was because the lenders had never seen a toll road with free competition just a few feet away, and insisted on that kind of protection. Today, 25 years later, while most RR concessions do include compensation clauses (the Chicago Skyway does not), they (1) exempt all projects in the long-range transportation plan in effect when the concession is signed, and (2) put the burden of proof on the company to document the extent of diversion to a competing facility that was not in the long-range plan. This has been judged a reasonable trade-off for the benefits of RR concessions, though obviously, not everyone agrees with this judgment.
The critics also object to my emphasis on the use of RR DBFOM procurement to weed out boondoggle projects beloved by politicians. They note that some projects that would not have a commercial return on investment (ROI) may nonetheless have important public benefits, justifying their development. As a long-time researcher and occasional consultant, I’m well aware of this point and have written elsewhere that a transparent benefit/cost analysis should precede any decision to procure a large-scale transportation project. Yet the benefits of using an RR concession may still be realized if the project makes public policy sense but is unlikely to generate a commercial ROI. In such a case, the state DOT makes an equity investment in the project, in effect buying down the amount to be financed based on toll revenue. And in such cases, it is always prudent for the state to negotiate a revenue-sharing agreement, in case the revenues in the out-years turn out to be significantly higher than forecast in the base model.
Those who portray the AP concession model as the wave of the future, even for tolled mega-projects, have claimed that RR concessions are so difficult that few states are no offering them and few companies are bidding on them. Those judgments are premature, at the very least. Besides the RR concessions now under construction (I-66 Virginia, SH 288 Texas), the past year has seen three new RR megaprojects offered—each the first one in its state. And there is considerable private-sector interest in them:
- Maryland I-270/I-495 Managed Lanes: 27 firms responded to the Request for Information.
- Illinois I-55 Managed Lanes: 18 firms submitted Expressions of Interest.
- Alabama I-10 Mobile River Bridge: four consortia formed.
I will close with a broader point. In my column in the February/March issue, I argued that the 20th century U.S. highway model is broken and needs to be replaced. After three decades in transportation policy, my conclusion is that highways are in fact a vital public utility, but are not funded and operated like any of our other utilities. In electricity, natural gas, water supply, telecommunications, etc., customers pay the company directly, based on how much of specific services they use. The direct customer/provider relationship has many benefits, which are largely absent in the highway sector—except in state-run toll roads and in RR concessions. What we should want is a highway utility system in which the providers have strong incentives to attract and please their customers, and in which customers know what they are paying for highway services, as they do with other utilities.
Basically, our highways need to be de-politicized, to at least the same degree that applies to other vital public utilities. Even when the electricity or water provider is a government enterprise, customers pay it directly, rather than sending an electricity tax to the legislature, which then horse-trades and micromanages which facilities to add to those networks and how much to spend on maintenance. State and local toll agencies have long been the exception in highways, but now we also have RR concessions as a preview of a future highway utility industry.
Just as in electricity, highways need owners and a direct relationship between customers and providers. While AP concessions are a large improvement over traditional procurement, some can be characterized as a construction company married to a contract maintenance firm, whose only customer is the state DOT. And that state DOT depends entirely on the legislature for its funding. America has the world’s best electric, gas, water, and telecommunications utilities. We should be aiming to have the world’s best highway utilities, too. RR concessions are the prototype for that emerging industry.
This column first appeared in Public Works Financing.