Surface Transportation Innovations Newsletter

Surface Transportation Innovations # 57

Topics include: cost-effective greenhouse gas reduction; private-sector toll roads' cost of capital; short-sea shipping; long-range transportation planning questioned; state DOT use of contractors; and other news.

In this issue:

McKinsey’s Calm Look at Greenhouse Gas Reduction

Everyone seems to have a pet solution for reducing U.S. greenhouse gas (GHG) emissions. Since many of the proposals for transportation call for drastic restrictions on driving (as measured by vehicle miles of travel-VMT), everyone in transportation has an interest in learning which GHG reduction measures are cost-effective and which are not. In that context, I want to call your attention to a major study by consulting firm McKinsey & Company, released in December 2007: Reducing U.S. Greenhouse Gas Emissions: How Much at What Cost? It involved a number of academic researchers plus inputs from various corporations and environmental groups.

Here is the central conclusion. While U.S. GHG emissions are projected to increase from 7.2 gigatons of CO2 equivalent in 2005 to 9.7 gigatons by 2030, there is realistic potential to decrease the 2030 total by 3.0 to 4.5 gigatons using abatement options whose marginal cost is no more than $50/ton. And almost 40% of the total could be achieved at “negative” marginal cost-i.e., they would pay for themselves over their lifetimes. The five “clusters” where most of the reductions would come from are (in order) electric power, buildings and appliances, energy-intensive industry, carbon sinks, and transportation.

Although transportation is the smallest cluster (with reductions ranging from 0.3 to 0.7 gigatons/year by 2030, depending on which of three scenarios were implemented), it’s clearly an important element. The two key themes are increasing average vehicle fuel efficiency and reducing the carbon intensity of transportation fuels. The three largest contributors to transportation GHG reduction would be cellulosic biofuels, inproved fuel economy for autos, and improved fuel economy for light trucks.

As in any such study, the results depend considerably on the assumptions made. One of the most important background assumptions was “no material change in consumer utility or lifestyle preferences.” Holding consumer utility constant “would imply no change in thermostat settings or appliance use, no downsizing of vehicles, homes, or commercial space, traveling the same mileage annually [as was assumed in the government reference case]. In other words, the reductions in GHGs projected in this study do not require the hair-shirt kinds of measures favored by many environmentalists and urban planners: major land-use changes to compel high-density living and transit use, etc. Rather, it posits that we can retain current middle-class expectations and aspirations while still reducing significantly the carbon-intensity of our lives.

And these results are even more impressive when you finally get to page 45 and learn that throughout the study, the assumed long-term price of oil was only $59 per barrel. (Remember, most of this work was done during 2007, well before the recent escalation in oil prices.) As the report notes, “If the long-term price of oil were higher, these options would become even more attractive.” I’m hoping McKinsey will put out a second edition, re-doing the calculations with something like $120/barrel.

That’s about all I have space for, but I urge you to download and read this important report-or at least the section dealing with transportation. You can find it at:

» return to top

Cost of Capital for Private-Sector Toll Roads

One argument raised by some critics of private toll roads financed under long-term concession agreements is that their cost of capital must be dramatically higher than that of public-sector toll roads. Public-sector toll roads are financed solely by debt, mostly or entirely tax-exempt toll revenue bonds. Private-sector toll companies use a mix of debt and equity-typically 20-25% equity and the rest debt. So the critics argue that since equity providers require a much higher rate of return than debt providers (they take on more risk, after all), and since the private sector can only issue taxable debt (which carries a higher interest rate than tax-exempt debt), the weighted-average cost of capital (WAAC) will be much higher for a private-sector toll road.

This argument was raised earlier this year by defenders of the status quo arguing against the possible long-term lease of the Pennsylvania Turnpike. And it was raised last year in Texas by those arguing against the development of SH 121 toll road in Dallas via a long-term concession rather than by the local public-sector toll agency. In the Pennsylvania case, a report by three academics that Peter Samuel and I criticized (see, they estimated the private-sector’s WACC at 7.75% compared with the Turnpike Authority’s 5.23%. In the Texas case, critic Dennis Enright initially compared a hypothetical private-sector WAAC of 6.98% with a public-sector figure of 4.44%. But after reviewing the final proposals actually made for the SH 121 project, he found a much smaller difference: 6.17% for the private sector v s. 5.0% for the public agency.

The big reduction in the private-sector figure came from the ability of the private bidder to make use of subordinated, tax-exempt debt under the federal TIFIA program. That is one of two measures enacted by Congress in recent years to encourage private-sector investment in highway infrastructure. The second was the private activity bond (PAB) provision in SAFETEA-LU that permits state agencies that work with the private sector to develop road projects under long-term concession agreements to issue tax-exempt revenue bonds on behalf of the project. And that can make an additional difference in the private-sector’s WAAC.

This difference was dramatically illustrated last month, when $589 million worth of PABs were issued as part of the Transurban/Fluor project to add HOT lanes to the Capitol Beltway (I-495) in northern Virginia. The interest rate on this senior debt was just 4.97%. When combined with the $587 million TIFIA loan at 4.45%, the weighted average cost of debt is just 4.71%. To be sure, 22.9% of the project cost is coming from Transurban and Fluor’s equity investment. Assuming a high single-digit (e.g. 9%) rate of return on that, the overall WAAC for this project is 5.69%.

Last year the Pennsylvania Turnpike issued $176 million in tax-exempt revenue bonds at 4.89%. But it also issued $68 million of federally taxable bonds at 6.11%. The WAAC for the Turnpike’s 2007 financing was thus 5.23%.

My point is that today’s financing opportunities for private-sector toll roads enable them to come very close to the weighted average cost of capital that is available to long-established public-sector toll agencies. Positing a major difference in cost of capital between the two options should no longer be taken seriously as an argument. What should be considered instead are the trade-offs involved.

Whether it’s a brand-new (“greenfield”) toll road or an existing (“brownfield”) one, there may be significant risk transfer to the private-sector partner, especially where major construction or reconstruction is involved. There are also differences between “system” financing and project financing. A large existing toll road system may be able to cross-subsidize a new toll road for a number of years, whereas the private sector typically finances each project solely based on its own revenues. System financing has its pluses, but if the new toll road has larger than expected losses, the toll agency will have to raise tolls on the rest of its system to meet its debt-service obligations. Under a long-term concession, toll rate increases on that road are generally capped; hence, if traffic and revenues are below forecast on the new road, the company has to absorb the losses. ( That’s what risk-transfer means!)

Those are the kinds of issues we should be debating when there is a choice between using the private sector or an existing toll agency. Yes, the private sector will likely have a slightly higher cost of capital. But the benefits may well be worth it.

» return to top

Is Short-Sea Shipping a Serious Alternative?

Under the twin pressures of increased fuel prices and concern over greenhouse gas emissions, there is growing interest in looking at whether it might make sense to shift some freight from trucks to other modes. Intermodal rail has been a successful alternative for some types of freight in some corridors. And then there is “short-sea” shipping-moving containers on barges along the U.S. coast or up rivers to inland ports. It’s been around for a long time, but has never seemed to gain traction. Advocates can point to greater fuel efficiency (ton-miles per BTU), but the very slow speeds that produce high fuel efficiency don’t match the needs of many shippers of goods in containers.

Be that as it may, in the Energy Independence and Security Act of 2007, Congress provided for $2 billion in loan guarantees for short sea shipping (SSS). The act also ordered the U.S. DOT to set up an SSS program and designate specific SSS projects and routes, and ordered DOT to use SSS to carry federally owned or generated cargo “when practical.” So this spring, the Maritime Administration launched its Marine Highways Project to identify corridors in which cargo could be shifted from highway to barge. Its first such corridor will be 100 miles of the James River, from Norfolk to Richmond, VA.

As these developments were taking place, a colleague sent me a paper by Robert Mulligan of Western Carolina University and Gary Lombardo of the US Merchant Marine Academy: “Short Sea Shipping: Alleviating the Environmental Impact of Economic Growth.” According to its abstract, its aim is to quantify the potential environmental benefits of SSS. I did find it useful in showing how fuel consumption increases exponentially with operating speed, leading the authors to conclude that a ship carrying 80 containers or trailers would have less fuel use and environmental impact than 80 trucks only at speeds below 27 knots.

Well and good, but what about overall cost? Here the quantification breaks down, thanks to the paper’s gross misrepresentation of highway costs. To begin with, the paper assumes that new Interstate-type lanes would be necessary to accommodate growth in truck traffic. Then it asserts that such new lanes cost $32 million per lane-mile plus $100 million per interchange. The citation for these extremely high numbers was listed as an FHWA document, “Typical Interstate System Cost per Mile,” Document Route Symbol HNG-13, March 21, 1997. After my email to Prof. Mulligan went unanswered, I contacted FHWA’s Highway Needs & Investment Analysis Team for help. A highway engineer there told me that this was “a one-page calculation sheet that was apparently updated annually and used internally for reference. It has been discontinued,” and he was unable to provide me with a copy. V ery helpfully, though, he sent me a recent sheet of costs used in their HERS software, giving costs per lane-mile in 2006 dollars. Note that this is nine years of highway cost inflation later than the 1997 date on the previous citation.

For rural Interstates, the cost to add a lane in “rolling” terrain is from $2.5 million to $4.0 million per lane-mile. For urban Interstates, in large urbanized areas, the comparable cost range is from $6.7 million to $22.5 million. Let’s take that 100-mile corridor from Hampton Roads to Richmond, traversed by I-64. That corridor is mostly rural, with some small and large (but no “major”) urbanized areas. So if that corridor is 80% rural and 20% large urban, the weighted average cost of adding a lane-mile to I-64 would be about $5.5 million per lane-mile, in 2006 dollars. That’s a far cry from $32 million in 1997 dollars!

If short-sea shipping makes economic sense, even as a small niche market, it’s going to have to be proven with real numbers, not wishful thinking. For now, I will remain a skeptic.

» return to top

Should We Keep Doing Long-Range Transportation Planning?

I purposely used that provocative question to introduce you to a very provocative report, “Roadmap to Gridlock: The Failure of Long-Range Transportation Planning,” by Randal O’Toole. ( Even if the premise makes you bristle, I urge you to download and read this challenge to the way most metro areas do long-range transportation planning today.

O’Toole reviewed 75 long-range plans, representing the 67 largest metro areas plus some smaller ones. He makes two kinds of critiques-procedural and substantive-and I think there is merit in both areas. Procedurally, he contrasts the current state of the practice with the “rational planning model” taught in planning schools. According to that model, planners first define goals and criteria, with the latter being mostly quantifiable outputs. Next, they forecast future travel needs and expectations, and test their modeling assumptions via sensitivity analysis. Then they develop a comprehensive set of possible transportation projects, with costs and benefits estimated for each (and with the benefits based on the quantifiable criteria-such as reduced congestion). Projects are ranked in terms of cost per unit of benefit, and sets of projects are created to be weighed as alter native plans. Overall costs and benefits of each alternative are then estimated, to enable selection of a preferred alternative.

Of the 75 plans O’Toole reviewed, he found none that came close to this model. None did sensitivity analysis of critical assumptions, none systematically estimated benefits and cost-effectiveness of proposed projects, most failed to include any realistic alternatives, and many failed to project the effects of the proposed plan on transportation. Most of the evaluation criteria were qualitative, not quantitative. And none of the plans listed the effects of individual projects on quantifiable criteria such as congestion. Let me interject here that I have read nowhere near 75 LRTPs, but I’ve been frustrated by the lack of such information in most of the ones I have read–but I naively thought they must be exceptions; apparently they are the rule.

Interestingly, O’Toole cites the one exception he is aware of-a splendid report that I remember reading with admiration about eight years ago. It was the “Bay Area Transportation Blueprint for the 21st Century: Evaluation Report,” produced by the staff of the Metropolitan Transportation Commission in 2000. For each of what must have been several hundred possible projects it provided quantitative estimates of costs and impacts on congestion. Because it revealed “inconvenient truths,” the report was not widely distributed, and no such report has been produced since then, by MTC or any other metropolitan planning organization that I know of.

O’Toole also raises some provocative substantive points. He groups the 75 LRTPs into two sets: those that emphasize technical tools (signal coordination, bottleneck removal, etc.) and those relying more on behavioral tools (land-use planning, getting people out of their cars, etc.). His critique of the latter approach is one that I share, and have written about from time to time in this newsletter. But he marshals some new arguments and evidence to make this case anew in this report. And he goes on to suggest that, in principle, long-range (25 years or more) transportation planning is unlikely to be effective. First, it requires information about the future that is basically unknowable. Second, when done in the behavioral mode, it becomes too complex to analyze in a meaningful way. Third, these inherent problems lead to many of the decisions being made politically, rathe r than by following a rational process. Fourth, the process offers planners and policymakers little or no incentive to make sure that their decisions are the best ones-or to change course if new information comes along. This very brief summary does not do justice to his arguments, which I hope you will read and consider.

So should transportation planning be abandoned? Not even O’Toole goes that far. But he does suggest that Congress should eliminate the federal mandate for 25-year plans, encouraging MPOs to focus on short-term planning. He also urges a greater reliance on user-fee funding, to provide stronger incentives for matching outcomes to needs (such as congestion reduction). His case may not persuade you, but I guarantee it will challenge your thinking on this subject.

» return to top

Outsourcing Transportation Engineering

One of the things I’ve become far more familiar with since moving to the east coast from California five years ago is the extent to which most state DOTs contract with private engineering firms. What was unthinkable at Caltrans (thanks to the political power and litigation expertise of that agency’s engineers’ union) is increasingly standard practice in DOTs all over the country. Two recent reports provide insights into this large and expanding trend.

The first is from the Government Accountability Office (GAO-08-198) and was released in January. “Federal-Aid Highways: Increased Reliance on Contractors Can Pose Oversight Challenges for Federal and State Officials,” despite its ominous-sounding title, provides comprehensive documentation of the extent and scope of this trend. GAO surveyed all 50 state DOTs-and got responses from all 50. Its researchers found that states have increased their use of outsourcing to the extent that “consultants and contractors have substantial responsibility for ensuring quality and delivery of highway projects.” More states are using design-build, for example, even though only 25 of the 50 states have full authority to use this important project delivery method, with another 10 having only limited authority. Project management, construction management, operations and maintenance managemen t are increasingly common as outsourced activities, while a smaller but growing number of states use such tools as lane rental, multi-parameter bidding, incentives/disincentives, and warrantees.

One of the ongoing issues in California (and some other states) has been endless controversy over whether it costs less or more than in-house provision to outsource functions like highway design. What GAO found mirrors what a Reason Foundation study ( found eight years ago-“Cost savings do not appear to be an important driver in the trend.” Rather, states outsource primarily to cope with staffing shortages or to ensure timely project delivery.

That was also one of the principal findings of the other report, written by former Utah DOT Director Tom Warne for the California Taxpayers Association and the Infrastructure Delivery Council of California. In producing “A National Assessment of Transportation Strategies and Practices: Lessons for California,” Warne reviewed the experience of nine other states and compared it with that of California. Like GAO and Reason, he found that “Timely delivery drives the decision-making process for selecting which projects stay in-house and which are completed by private-sector engineers.” And “The decision to outsource is not made based on a cost comparison,” because “No state has a robust enough cost-accounting system to accurately determine the true cost of their services and compare that to the known costs of private sector firms.” Please see:

Getting back to the GAO report, there is one dark cloud hanging over this largely positive report. In asking GAO to look into this issue, one of the questions posed by congressional requesters was how do state DOTs ensure the public interest is protected during outsourcing, and is there anything the Federal Highway Administration should be doing about this? GAO tactfully responds that yes, there are potential risks involved, including weaknesses in some state quality assurance programs and also the potential for some conflicts of interest. And while FHWA has identified these risks, “it has not comprehensively addressed how, if at all, it needs to adjust its oversight efforts to protect the public interest” in this regard. GAO also points out that most federal highway money is allocated by formula, not performance, so there is not much scope for direct controls on interna l state DOT management practices. If a future Congress starts mucking around in how states manage their programs, my guess is that this will further fan the flames of those calling for significant devolution of the federal role in surface transportation.

» return to top

Upcoming Conferences

Just a reminder that I don’t have space to list all possible conferences of interest; those listed here are only ones that I or a Reason colleague will be speaking at.

Partnerships in Transit, National Council for Public-Private Partnerships and Federal Transit Administration (3 locations):

  • San Francisco: July 30-31, 2008, Hotel 480
  • Philadelphia: Sept. 17-18, 2008, Sheraton City Center Hotel
  • Dallas: Oct. 22-23, 2008, Radisson Hotel Central Dallas

Details available at

Infrastructure Finance Summit, Infocast, Chicago, Sept. 8-10, 2008 Sutton Place Hotel, Chicago, IL. Details at:

ARTBA’s 20th Annual Public-Private Ventures in Transportation, Washington, DC, Sept. 15-16, 2008, Hilton Washington Hotel. Details at:

» return to top

News Notes

Truckers Support Toll Increase. Despite the trucking industry’s general opposition to toll roads, the Association of Bi-State Motor Carriers has backed a proposal by New Jersey Assemblyman John Wisniewski to increase tolls on the state’s three toll roads, as well as the state gasoline tax. The Motor Carriers’ president, Jeffrey Baker, told Newsday that as long as all the proceeds are used for transportation infrastructure, his group will support the increases. That’s in contrast to Gov. Jon Corzine’s now-dead plan for massive toll increases to bail out the state’s fiscal problems, which the truckers rightly opposed.

Excellent Managed Lanes Video. Florida DOT has released a new video introducing the public to the forthcoming 95 Express Lanes-variably priced managed lanes with express bus service that will replace the existing HOV lanes on I-95 in Miami-Dade and a portion of Broward County. The first northbound section is set to open next month. The video is available in both English and Spanish-language versions. Go to:

TRB Report on Public Opinion re Tolls and Pricing. The outstanding report from TRB’s National Cooperative Highway Research Program showing the recent (past 10 years) growth in public preference for tolls rather than tax increases to fund transportation improvements (that I wrote about in Issue No. 54) is now available on-line. I served on the TRB review committee that oversaw the research, and I’m very enthusiastic about the findings. Go to:

Joint Resolution for Atlanta Managed Lanes. Back in March of this year the Georgia State Transportation Board and the State Road and Tollway Authority signed a joint resolution endorsing managed lane corridors as an important tool for reducing traffic congestion. It commits them to complete ongoing feasibility research and then to work together to develop region-wide governing policy for a system of managed lanes, along with educational outreach activities.

Correction from Last Issue. In last month’s issue I wrote about the latest International Housing Affordability Survey, but did not catch a typo in the URL for downloading the report. The correct URL is:

» return to top

Quotable Quote

“The idea behind focusing on accessibility instead of mobility is that, if cities are designed so that people are close to shops and services, they won’t need to drive as much or as far. The problem with that idea is that consumers rely on a competitive market in retail and services to promote innovation and keep costs low. Consumers who are captive of one or a limited number of stores end up paying higher prices, often for lower-quality goods. Moreover, even in a world with limited energy supplies, there is no guarantee that having local stores within walking distance of residential areas is the optimal pattern. Some experts in the retail industry suggest that higher energy prices will give an advantage to big-box supercenters where people can do all their shopping in one auto trip.”
–Randal O’Toole, in “Roadmap to Gridlock,” Cato Institute, May 27, 2008.

“Why won’t workers cluster around their jobs as they did in the 18th and early 19th century? A number of very forceful reasons:

  • We are not wedded to a job for life any more. The average turn-around in jobs is measured in just a few years. It is expensive to move every time one changes jobs, uprooting one’s family; and self-defeating as well if you may be moving back again soon.
  • About 70% of workers live in households with other workers. Whose job will they live next to?
  • Workers work in much smaller units today, so there is no big factory gate to live next to.

–Alan Pisarski, testimony regarding the future federal role in surface transportation, U.S. Senate Committee on Environment and Public Works, June 25, 2008.

“The ‘out of the suburbs, back to the city’ narrative rests more on anecdote than demographic or economic fact. Yes, high gas prices and rising sub-prime mortgage defaults are hurting some suburban communities, particularly newly built ones on the periphery. But the suburbs remain home to a majority of Americans and a larger proportion of U.S. families-and people aren’t leaving those communities in droves to live in cities. Even with economic growth slowing, many suburbs, exurbs, and smaller towns, especially those whose economies are tied to energy, are continuing to do better than most cities in terms of job creation and population growth.”
–Joel Kotkin, “Suburbia’s Not Dead Yet,” Los Angeles Times, July 6, 2008

» return to top