As the elected establishment began touting how the financing model works so well for this project, we believe that is fuzzy math at best. If the state does not have $85 to $120 million to build a lane each way from Exit 23 to 30, but not only has $170mm to contribute to the $550mm HOT lane project – claims they are being made that the 4 access points will be 8 @ est $20m per adds $80mm to move the total to $630mm AND suddenly $200mm of apparently FREE money is falling out of the sky to pay for an assortment of local road projects. IF IT SOUNDS TOO GOOD TO BE TRUE ? Then re-read the reality of the situation based on referenced facts that we initially published on June 17, 2013
Some elected officials justify toll lanes as a means to generate revenues for roads. They note gas tax receipts are expected to decline while road construction needs will increase. These trend lines are headed in the wrong direction and to reverse that, these advocates point to toll lanes (and roads) as an important revenue-raising tool.
This implies that toll lanes must not only cover their operating costs, but also bring in enough revenue to pay for the lane itself. Ideally they would generate a surplus that can be used as a source of funding for other transportation needs. For I-77, the current plan is a $550M project consisting of toll lanes for 27.5 miles from downtown Charlotte to the middle of Iredell County. The project is proposed as a public-private partnership, where the NCDOT would lease the lanes to a private company under a 50 year contract. The $550M includes $170M in public funds, and the rest would be financed through a combination of private debt and equity.
Before we begin our analysis, a word about sources. Population data comes from the U.S. Census Bureau. The financial data was taken from the 2012 Managed Price Lane Guide published by the Federal Highway Administration. This document is basically a how-to manual for getting toll lanes built. It was written by Parsons-Brinkerhoff, an engineering consulting firmly deeply involved in the toll lane business. Among its authors is David Ungemah, the featured speaker at a local toll lane information session. He is nationally recognized as a “managed lanes expert.” Thus, any bias in the information presented is in favor of toll lanes.
With that in mind, we’ve undertaken a study of existing toll road financials.
The chart below plots annual toll revenues vs metropolitan served area (MSA) population for 11 toll lanes currently in operation. In addition, we plotted the required revenues ($20M) for the proposed I-77 toll lanes. Toll lane experts have placed this number as high as $30M; we erred to the low side.
A couple of observations. First, toll revenues are roughly correlated to the MSA. On the low end are the I-15 lanes through Salt Lake City with a metropolitan area of approx one million and revenues of $500K. On the high side- indeed an outlier- is SR 91 connecting North Orange County to Riverside. Combined, over 7 million people live along this route, and toll revenues were $41M. We’ll discuss SR 91 in more detail in a moment.
This correlation is not surprising as urban sprawl is necessary for the toll lane model to work. As we have stated elsewhere, toll lanes rely on congestion in the general purpose lanes to entice people to pay the toll. Robert Poole of the Reason Foundation affirms this in a recent article he wrote examining toll revenues in Miami. (Poole invented the term “HOT lanes” and is widely viewed as the father of the concept.) Poole states:
I could imagine that a managed lane project in Minneapolis or Salt Lake City, say, might not relieve enough congestion to cover its costs out of toll revenues. But congestion on I-95 in Miami was nearly as bad as on the Los Angeles freeways, and the prices charged during peak periods are considerably higher than on many other managed lanes around the country.
In order to cover its costs, Poole implies, toll lanes must be in a large metropolitan area.
The second observation is that the required revenues for I-77 are out of proportion with the region’s population. I-77 toll revenues will need to surpass those in Seattle, San Diego, Houston, Atlanta, and Miami, among others. In fact, the I-77 project would need to be the second-highest grossing toll lane in the country. It would require revenues greater than the eight lowest-grossing toll lanes combined. This should be especially concerning in light of the fact that the project’s northern terminus (Iredell) has the smallest population (160K) of any of the existing toll lanes.
Operating Income is defined as revenues minus operating costs and does not include debt repayment. Whatever is left over from operating expenses is what is available to pay for the capital cost of the project. Toll lanes incur several operating costs that are unique to toll lane operation. Among these are advertising, toll billing, toll collection, credit card fees, toll enforcement, occupancy enforcement, administration of the tolling authority, insurance claims and premiums, and ROI to private equity and debt investors. General purpose lanes incur none of these.
The chart below shows annual operating income vs population, with projects sorted in order of increasing population. Of the 11 projects, 7 have negative operating income. Three generate significant positive income and one generates a modest income.
With one exception, no toll lane serving a population of less than 5.5 million currently covers its operating cost. The exception (Denver) is an HOV-HOT lane conversion with a modest capital cost of $9M. It was built with public funds and is run publicly, thereby avoiding some of the expenses previously mentioned. The other HOT lanes with positive income are in Houston, Miami and Riverside/NOC. SR 91 (Riverside/NOC) generates a significant income of nearly $20M. This stretch of road is ten miles long and serves the relatively affluent North Orange County and more working-class Riverside. The total served population is over 7 million and traffic on this road is over 300K vehicles per day, about twice what the busiest freeway in North Carolina carries.
However, even this “success” story is not without a cautionary tale. The project originally opened in 1995 as a private, for-profit investment at a cost of $135M. When the state tried to make improvements to ancillary roads, the operator sued, citing a clause in the contract that would prevent the state from making improvements that would ease congestion and impact toll revenues. In 2003 the Orange County Transportation Authority bought the project back for $207M, 50% more than its original cost.
I-77 is not included here because the financials are not yet public. At 27.5 miles the I-77 toll project would be the second longest in the country, so operating costs can be expected to be on a par or greater than those currently in existence.
The chart below shows the total project capital costs. I-15 in San Diego is the highest, at $1.3B. The project was financed through general bonds issued by the state of California. This differs from the proposed financing in North Carolina, where bonds will be issued using anticipated toll revenues as surety (“toll revenue bonds”). Implicit in the San Diego project is the consideration that toll revenues will never be sufficient to retire capital cost. Conversely, as we reported here, North Carolina plans to issue hundreds of millions in toll revenue bonds. In fact, by the end of FY 2015, North Carolina will have issued over $2B of these types of bonds. These will be used for toll road construction (not toll lanes), but proponents of this method should take heed that even in Southern California, tolls are not viewed as a source of additional revenue.
The variation in capital costs are largely due to whether the project is a conversion of an existing lane or an entirely new project. Houston, Riverside/NOC and Miami are new construction while the remainder are conversions of existing HOV lanes.
Of course, new construction is significantly more expensive than conversions. I-77 is primarily new construction (10 lane-miles out of 90+ will be HOV conversion), and this is reflected in the project’s capital cost of $550M. The I-77 project would be among the most expensive currently proposed, and more expensive than any toll lane financed through toll revenue bonds except the Capital Beltway in DC. (The recently opened Capital Beltway cost over $2B. Operating financials are not yet available.)
As far as the implications for the rest of North Carolina, I-77 has the only HOV lanes in the state. Therefore, building toll lanes will require either new construction- and the associated expense- or conversion of existing general purpose lanes.
Compared to projects in the rest of the U.S, I-77 toll lanes would be:
- the second-highest grossing to make the project financially viable
- the second most expensive project financed through toll revenue bonds
- the second longest project in terms of miles
- the first to have four toll lanes beside four general purpose lanes
- located in the second smallest metropolitan area
In addition, I-77 toll lanes would be North Carolina’s first:
- privately operated toll lane
- toll lane or road to be operated on a for-profit basis
- road under a 50 year contract
Historically, no toll lane:
- serving a population the size of Mecklenburg-Iredell has ever had a positive operating income
- has ever repaid a debt obligation as large as the proposed I-77 project, regardless of metropolitan population
The proposed I-77 project therefore requires revenues and income that have been historically unattainable. This raises an obvious question: shouldn’t the bond underwriters- the people who do this for a living- walk away from this? When we were in Raleigh we asked this question. The answer was if they do not think the project is viable, they will simply not fund the project.
Not so fast. As we have pointed out previously, under the terms of the existing contract the taxpayer will be responsible for all but the equity portion of the total project cost. (Equity is typically about 20% of the total project.) If the project fails, advocates tout that taxpayers could buy the project for “pennies on the dollar.” To most people, eighty cents on the dollar is a lot of pennies.
Even if that were the case, the toll lane project requires several improvements that are otherwise unnecessary. Foremost among these are the replacement and/or construction of 9 bridges. According to the NCDOT none of the existing bridges require replacement. So taxpayer money that could be spent elsewhere would instead pay for replacing structurally sound bridges.
The above analysis shows that North Carolina lacks both the population centers and existing infrastructure of readily-convertible lanes to make toll lanes financially viable. If we pursue this route the scenario for North Carolina is one of high capital costs, low revenues, and negative operating income. Indeed, rather than generating additional income for roads, toll lanes pose a substantial risk of making our funding problem worse.
Toll lanes may work elsewhere, but they should be removed from consideration in North Carolina.