In this issue:
- Electric cars: this time for real?
- Toll concessions-one size does not fit all
- Cost of crashes vs. costs of congestion
- Where can toll truckways be built?
- Pension fund investment in toll roads
- New data on toll road safety
- News Notes
- Upcoming Conferences
- Quotable Quote
Despite the tall tale told in the 2006 film "Who Killed the Electric Car?" the General Motors EV1 was not ready for prime time. I was one of a group of opinion leaders in Los Angeles invited to take the car for a test drive, on a hot summer day in 1996. While the EV1 was nice and peppy for a short drive near downtown LA, I noted with dismay that the cars available for us to test-drive were all being pre-cooled from an external air conditioning source: the vehicle's own lead-acid battery pack was unable to cope with summer air conditioning loads, in addition to moving the vehicle and its payload the 60-80 miles that one charge would take it. So no, it was not an auto/oil industry conspiracy that led the California Air Resources Board to rescind its 1990 Zero Emission Vehicle (ZEV) mandate that 10% of all cars sold in the state by 2004 must be ZEVs; rather, CARB was simply acknowledging the reality that the technology was not yet there to produce such vehicles.
Recent articles in The Economist (March 8th Technology Quarterly) and MIT's Technology Review (March/April and May/June) recount the remarkable progress in battery technology over the last 10 years. For awhile, it looked as if nickel metal hydride (NIMH) was going to be the replacement for lead-acid batteries-easier to recharge and more energy per pound, but at higher cost. But more recently, engineers have been figuring out new ways to scale up lithium-ion batteries like those we use in our laptops and cellphones. It looks to me as if one or two versions of this approach will prove to be commercially viable for the first successful generation of plug-in hybrids. First-generation hybrids like the Toyota Prius use the battery to supplement a gasoline engine. But plug-in hybrids have enough battery power to handle short trips (typically 30 to 40 miles) en tirely on battery power, needing the gasoline engine only for longer trips. For a typical commuter, such a vehicle can handle the daily round-trip commute, then recharge at night. GM has contracts with two competing lithium-ion battery firms, in hopes that at least one will prove to be good enough to use in its Chevy Volt, scheduled for late 2010 introduction.
These developments spark several thoughts. First, there's a sobering lesson for government agencies such as CARB about the perils of trying to force particular technologies into being. Whenever you see mandates for hydrogen fuel cells or E85 ethanol cars, be very suspicious, because government has a lousy track record in picking winners and losers. Just because the catalytic converter came along in the nick of time to enable automakers to comply with tough federal tailpipe emission standards does not mean that any and every energy or emissions mandate can be met-at all, or at an acceptable cost to consumers.
Second, how successful plug-in hybrids will be in reducing CO2 emissions depends hugely on the source of the electricity used. A table in Technology Review (March/April, p. 30) compares the same vehicle's emissions for a whole variety of power sources. Compared with conventional gasoline propulsion at 452 grams of CO2 per mile, a current-day hybrid would produce 294. But a plug-in hybrid using electricity from conventional coal-fired power plants produces 326; only with coal gasification and CO2 sequestration (a costly proposition) would the car get down to 166 grams/mile. Using conventional natural gas electric power, the plug-in hybrid does 256. But using nuclear electricity, that figure drops to 152-essentially just the CO2 produced by the small amount of gasoline used.
Third, while I'm excited by the whole raft of start-up electric hybrid car companies (like Fisker and Tesla), and would love to see them succeed, I would not put my own money into them. Mass production of reliable, cost-effective automobiles is a very different thing than producing a brilliant prototype, as numerous entrepreneurs such as Malcolm Bricklin and John DeLorean found out the hard way. My money is on GM, Ford, Toyota, and the others that really know mass production.
Finally, one key to the success of plug-in hybrids as a "green" solution to CO2 emissions is how and when they recharge. The ideal is for them to do so at night, when electric utilities' inexpensive base-load generators must keep turning and they have lots of spare, cheap capacity. The worst impact would be if huge numbers of people try to recharge their plug-in hybrids during peak hours, when costly peak-power units have to come on-line to meet the demand for building air conditioning, etc. The answer to this is "congestion pricing" of electricity. And fortunately, the investor-owned utilities in California are gearing up to meet this challenge. The three big ones-PG&E, SCE, and SDG&E-are installing $5 billion worth of smart electric meters between now and 2012. These meters make it easy to charge variable rates, based on the time of day. SDG&E is already of fering half-price night-time rates for plug-in car recharging. (One idea that does not make sense is for helpful governments to install charging stations at places like park & ride lots or workplaces. That charging would take place during peak electricity usage times, precisely when electricity costs the most. Unless they use pay-meters that charge a peak price, such charging stations would be ill-advised.)
So the prospects for practical electric vehicles-as plug-in hybrids-are really looking up.
I'm bemused as I watch the evolving debates over long-term toll concessions in this country. Based on a few high-profile cases, many policy-makers seem to think there is only one form that these long-term agreements between a private consortium and (usually) a state DOT can take. Whether it's the lease of an existing toll road or (more commonly) a long-term deal to design, finance, build, operate, and maintain a brand new toll facility, the impression is that the state holds a competition and the sole deciding factor in selecting the winner (from a set of pre-qualified teams) is the amount of the huge up-front payment they offer. But that is far from the only model, as experience in Europe, Australia, and South America makes clear.
Let's look first at leasing of existing toll roads. John Foote (a senior fellow at Harvard's Kennedy School) pointed out in a talk at the International Bridge, Tunnel & Turnpike Association annual meeting in Vienna last fall that when the French government privatized its remaining state-owned toll operators in 2005, the winning bids were about 12 times EBITDA (earnings before interest, taxation, depreciation and amortization). By contrast, the multiples for the leases of Chicago Skyway and Indiana Toll Road were more like 60 times EBITDA. That difference is a matter of government policy, primarily (1) what sort of toll-escalation formula and (2) how long a concession term the government was willing to agree to in each case. And any government also has the option of choosing to receive some or all of the lease payments over the life of the lease-either as a pre-define d amount or as a percentage of gross or net revenue. The point is that these are policy choices-there is no single model.
When it comes to greenfield (start-up) toll projects, we've hardly scratched the surface in terms of alternative ways to structure concession deals. An excellent overview of alternatives-especially concerning revenue-sharing-was provided last year by Jennifer Mayer of FHWA in "Private Returns, Public Concerns: Addressing Private-Sector Returns in Public-Private Highway Toll Concessions" (Transportation Research Record No. 1996, TRB 2007). She provides examples of how each technique has been used. For example, Ireland has based selection of the winning bidder on the net present value of the proposed up-front fee plus revenue-sharing. In the U.K., recent concessions have included a provision to share any gains from refinancing with the state on a 50/50 basis. She discusses the use in France and Spain of variable-length concessions, in which the agreement terminate s once the concessionaire has achieved a pre-specified internal rate of return.
The latter idea was explained in more detail in an article by Eduardo Engel of Yale and two Chilean colleagues, "A New Approach to Private Roads" (Regulation, Fall 2002) and in a more formal paper in Review of Industrial Organization (Vol. 29, pp. 27-53, 2006). In projects in the U.K., Chile, and Portugal, the present value of revenue (PVR) concession has been used to select the winning bidder and to design the provisions of the concession agreement, especially the length of its term. This type of arrangement tends to reduce the amount of toll revenue needed to finance the project, reduces the need to renegotiate the agreement to deal with outside events such as a major recession, and facilitates an early buyout by the state (by clearly spelling out a procedure for calculating the remaining value of the concession). Because it depends on a pre-set tolli ng regime, however, it does not appear applicable in situations where congestion-priced tolling is called for, such as HOT lanes.
Australia illustrates the transition from early-days concession projects to a more mature market. As Christine Brown of the University of Melbourne explains in "Financing Transport Infrastructure: For Whom the Road Tolls" (Australian Economic Review, Vol. 38, No. 4, pp. 431-438, 2005), in the early Sydney and Melbourne concession toll roads, the government sought to reduce the private sector's risk. For these greenfield urban toll roads, there was no up-front concession fee; instead, annual concession fees were accrued on the company's books until the company reached a target for after-tax internal rate of return, at which point cash payments began. But as the market matured, the fairly standard agreements today call for an up-front payment of 10-20% of project value plus sharing of profits above an agreed baseline level. (That kind of approach may not work for high-risk urban projects or most new inter-city toll roads where traffic levels are not robust.)
The latter is the situation for many inter-city toll roads in Spain. There, the government in 2003 revised the 1967 concessions law to permit a form of traffic-risk sharing, on an optional basis. Companies bidding to develop such a toll road can ask for a subordinated public participation loan (SPPR) of up to 50% of the project's cost; this is in some ways analogous to the U.S. TIFIA or state infrastructure bank loans. Interestingly, the interest rate is variable, directly related to the growth in traffic and toll revenue. So the better the project does, the higher the return to the government-unlike in our TIFIA or SIB loans. Thus far, according to an assessment by Jose Vassallo and Antonio Sanchez-Solino, the process seems to have fostered greater competition and at a low cost to the government: just 5.6% of the $1.8 billion invested in six toll road projects since the new provision went into effect. ("Subordinated Public Participation Loans for Financing Toll Highway Concessions in Spain," Transportation Research Record, No. 1996, RTB 2007).
As you can see, there are many creative ways to structure long-term toll concession agreements. These are early days for U.S. use of this powerful technique. We can learn a lot from the experiences of countries that have been at this far longer than we have.
You probably saw at least one article in March with some variant of the headline, "Automobile Crashes Cost Society Much More than Congestion" (The Urban Transportation Monitor, March 21, 2008). These articles reported somewhat uncritically on a study carried out for the Automobile Association of America by Cambridge Systematics and Prof. Michael D. Meyer. What the study did, competently, was estimate the total annual cost of traffic accidents (deaths, injuries, property damage, etc.) using Federal Highway Administration data. It then compared this total with annual congestion costs as estimated by the Texas Transportation Institute. Since TTI does this only for the largest 85 urban areas, the AAA study used crash costs for these same urban areas in the comparison. The headline numbers were $164.2 billion in crash costs vs. $67.2 billion in congestion costs.
So far, so straightforward. But what does this comparison leave out? Only a month before the above headline about the AAA study appeared on page 2 of The Urban Transportation Monitor, that newsletter had a front-page headline reading "Economists Find Toll of Congestion Is Often Underestimated" (Feb. 8, 2008). That story focused on recent work by Glen Weisbrod and Stephen Fitzroy of Economic Development Research Group in Boston. They point out that most measures of congestion cost, such as TTI's, include only drivers' lost time and wasted fuel due to being stuck in traffic. Their work-including case studies in Vancouver, Chicago, and Portland (OR)-finds much larger costs that congestion imposes on business and a region's economic productivity. An excellent introduction to their work is "Defining the Range of Urban Congestion Impacts on Freight and their Consequen ces for Business Activity," an August 2007 working paper that you can find at www.edrgroup.com. I also highly recommend their Portland and Vancouver case studies, which I've written about in previous newsletters.
While Weisbrod and Fitzroy have not developed an alternative to the TTI nationwide congestion cost estimate, the chief economist of the U.S. DOT has. Jack Wells estimates that congestion costs smaller (non-TTI) cities another $12-13 billion per year, leads to about $13 billion in safety and environmental costs, and adds about $79 billion in productivity losses and goods-movement inefficiencies, for a grand total of $168 billion. That is 2.5 times the widely known TTI figure, and is slightly more than the crash costs total estimated in the AAA report.
It's good to have a plausible estimate of the high cost of crashes on the highways. But in point of fact, the full cost impact of congestion is equally high. Both need to be targeted for major reductions.
Back in 2002, Jose Holguin-Veras, Peter Samuel, and I introduced the concept of value-added toll truckways-separate rights of way built mostly along truck-heavy Interstate corridors aimed at giving much better service to heavy-duty trucking (www.reason.org/ps294.pdf). The idea was to figure out what truckers and shippers want from the highway system that they aren't now getting but might be willing to pay more to get. The benefits we identified were (1) the ability to haul a lot more payload per driver, by using double- and triple-trailer rigs in states where they are banned by federal law, and (2) faster and more reliable trips, compatible with today's just-in-time logistics system.
This idea is now the subject of a four-state feasibility study under the federal Corridors of the Future program, looking at the addition of such truck lanes on I-70 from Kansas City to Columbus, as I reported in Issue No. 49. But there is one major obstacle to implementing the idea, if that study finds it to be technically and economically feasible: the federal freeze on truck sizes and weights. The "LCV freeze" was enacted by Congress as part of ISTEA in 1991, at the behest of railroads and highway safety groups, and still has staunch defenders in Congress. Our best hope for projects like the I-70 toll truckways is that Congress can be persuaded to create a pilot program to try the idea in at least a few corridors, by granting exemptions from the freeze for barrier-separated truck lanes.
At a January FHWA conference on these issues, I asked the agency for clarification on exactly what roadways the LCV freeze applies to. I recently received what appears to be a definitive answer. The freeze applies to the Interstate system, to the National Network for Large Trucks (defined by the states pursuant to the 1982 Surface Transportation Assistance Act), and to roadways added to the National Network since then. It does not apply to any other existing roadways or to new roads that a state may create.
Since the Interstate system was defined based on the demographics and travel patterns of the 1950s, there are many missing links based on today's and tomorrow's needs. For example, two new east-west routes are in the planning stages in Southern California, aimed at getting truck traffic from I-15 to I-5 without having to drop down into congested Los Angeles. Both the High Desert Corridor and a route paralleling SR 58 would be state highways, not Interstates, and neither appears to be part of the National Network.
A separate question has still not been resolved by FHWA, as far as I can tell. If a toll truckway such as that proposed for I-70 or the one linking the Ports of Los Angeles and Long Beach with the Inland Empire were built along the right of way of existing National Network routes but designed in such a way that there was no mingling of traffic whatsoever between the general purpose lanes and the truckway lanes would the truckway be "part of" the National Network highway or simply something built alongside or above it? If that question cannot be definitively resolved by highway-law experts, perhaps it could provide a basis for exemptions to the LCV freeze in the upcoming reauthorization. After all, the safety objection to LCVs has to do solely with their mixing with regular traffic in general purpose lanes. If the toll truckway is a truly separate system, no such mixing w ill be possible, hence the safety concern should go away.
I reported in Issue No. 44 that pension funds, along with insurance companies, are among the institutional investors that are shifting some of their "alternative" investment funds to infrastructure, including toll roads. Sources at Macquarie tell me that their Macquarie Infrastructure Partners fund, with 47% U.S. investors, includes the Midwest Operating Engineers Pension Fund and Mid-Atlantic Carpenters Pension Fund. Giant CalPERS, the largest public pension fund in the country, allocated an initial $2.5 billion to infrastructure investments late last year.
Pension funds can invest in both debt and equity of toll road projects. It is equity investment that they are seeking when they place funds with one or more of the dozens of new infrastructure (equity) investment funds, like those from Macquarie, Goldman Sachs, Carlyle Group, etc. They can also purchase toll road revenue bonds. But here the picture gets interesting. Public employee and union pension funds are tax-exempt entities. So when they look for toll revenue bonds to purchase, it does them no good to invest in the tax-exempt bonds of public-sector toll agencies. Those bonds carry a lower interest rate (usually 1-2 percentage points lower) than taxable toll revenue bonds of comparable investment grade, because interest paid on those public-sector bonds is exempt from federal taxation. But that tax exemption is useless to a pension fund that is not liable for federal taxes; if it buys such a bond, it foregoes the higher interest rates it could be getting via a comparable taxable revenue bond, thereby breaching its fiduciary duty to its pensioners.
This point is so obvious to knowledgeable people that it seems to have been overlooked in public policy discussions over possible investment in toll roads by state pension funds. A number of Texas legislators, for example, have urged entities like the $24 billion Employees Retirement System and the $107 billion Teachers Retirement System to invest in toll roads and other infrastructure. Yet how many of them realize that the only toll roads those funds could responsibly invest in are those developed under long-term concession agreements (called CDAs in Texas) by the private sector? They won't be investing in the tax-exempt revenue bonds of the Harris County Tollway Authority or the North Texas Tollway Authority.
I wonder, too, if this point has dawned on Andy Stern, president of the powerful Service Employees International Union (SEIU). In February, Stern raised the idea of forming an investment pool of money from state retirement systems to invest in U.S. infrastructure projects, "with the idea of keeping the assets under some degree of public control." A Financial Times article reported discussions between New Jersey officials and SEIU about such an infrastructure fund.
Just so that everyone is clear: tax-exempt pension funds do not buy tax-exempt bonds. So the only toll road revenue bonds they can buy are taxable bonds, which are almost exclusively issued by private toll roads. If we want pension funds to have good infrastructure investment opportunities, we need to expand the scope for privatized airports, toll roads, water and wastewater systems, and other infrastructure. Pension funds in Australia and Canada have been doing this for a decade or more.
Over the years I've heard a lot of anecdotal claims that toll roads are safer than non-toll roads. Skeptics should take such a broad claim with a few grains of salt, since toll roads are by definition limited-access road with separation between opposing lanes (like Interstates), and we already know that such roads are safer than ordinary roads.
But now the International Bridge, Tunnel, and Turnpike Association has crunched the numbers and come up with what appear to be fair and realistic comparisons. The research is summarized in the Winter 2008 issue of Tollways, in an article by their research director Jeff Campbell titled "Toll vs. Nontoll: Toll Facilities Are Safer." (www.ibtta.org/files/PDFs/win08_Campbell.pdf)
Campbell surveyed a large number of U.S. toll roads, bridges, and tunnels, compiling data on fatalities and accidents per 100 million vehicle miles of travel, the standard metric. On fatality rates, for all U.S. roads, FHWA figures show 1.47 fatalities per 100 million VMT. For all toll facilities, the comparable figure is 0.50. But that is not the meaningful comparison, as noted previously. For all rural Interstates, the fatality rate is 1.21 and for urban Interstates it is 0.55. Since toll roads are a mix of urban and rural, some kind of weighted average of urban and rural Interstates is the best comparator, but Campbell does not give us that.
But Peter Samuel of Tollroadsnews.com does. Using FHWA data, he developed a weighted average for all three types of non-toll limited access roadways: urban Interstates, rural Interstates, and non-Interstate expressways. That "all-expressways" fatality rate is 0.80. That compares with Campbell's figure of 0.52 for toll roads only (excluding bridges and tunnels). Thus, on an apples-vs.-apples basis, toll roads are 35% safer than non-toll roads. (www.tollroadsnews.com/node/3457).
Metro Naming Rights. We've seen naming rights for sports arenas and convention centers, but now comes the new Dubai Metro. In full page ads in venues like The Economist, Dubai's Roads and Transport Authority is offering Dubai Metro Naming Rights. "You can put your brand on a Dubai Metro Station of your choice, or one of the two lines of the Dubai Metro network." Check it out at www.rta.ae.
Alliance for Toll Interoperability. Following the trail blazed by the InterAgency Group in the northeast and Midwest, with its fully interoperable E-Z Pass electronic toll system, toll providers from the Carolinas, Georgia, Florida, Texas, Louisiana, Texas, Oklahoma, Kansas, and Colorado met in Dallas in February to form the Alliance for Toll Interoperability. Acting chairman is J.J. Eden of the North Carolina Turnpike Authority. He told Tollroadsnews.com that the group's initial focus will be on establishing reciprocity on license plate reading, a key factor in video tolling.
Heat from the Street? The Economist (Dec. 8, 2007) reports a fascinating development from The Netherlands. Dutch construction company Ooms made a deal with the local roads agency to develop a heating/cooling system based on the black asphalt road surface next to their building. In summer, water pipes absorb heat from the asphalt and store it in an underground aquifer, from whence it is pumped to the building in winter, to heat it. After doing that, the water is pumped under the asphalt to help de-ice the road, and then the now-cold water is pumped into a second aquifer where it is kept cold until summer, when it is used to cool the building. Sounds neat-but complicated and possibly expensive. Whether this is just an engineer's toy or a practical solution remains to be seen.
Bay Area Residents Want Better Infrastructure. A new public opinion survey done for the Bay Area Council found that 87% of San Francisco area residents think infrastructure is a serious or very serious problem. And significant majorities are open to public-private partnerships as a way to get more and better infrastructure, with 70% supporting PPPs for transit and 65% supporting PPPs for roads and highways. The findings were released in mid-April.
World Bank PPP Infrastructure Site. The World Bank has launched a new website on infrastructure and the law, designed for public officials, legal experts, and others involved in the planning, design, and legal structuring of infrastructure projects, especially those involving the private sector. The three main sections cover government objectives, legislative and regulatory framework, and agreements. Go to http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTLAWJUSTICE/EXTINFANDLAW
Reason Foundation people are speaking at all four of these upcoming conferences:
"Preserving the American Dream," May 16-18, Omni Houston, Houston, TX
Extensive program on transportation and land-use issues includes my Reason colleagues Sam Staley, Shirley Ybarra, and Adrian Moore. (www.americandreamcoalition.org/pad08.html)
"2008 FICE/FDOT Consultants' Conference: the Future of Florida's Transportation Industry," May 21-22, Hyatt Regency Orlando Airport, Orlando, FL
I'm speaking on HOT Networks, along with a number of national and state transportation experts. (www.fleng.org/seminars.dfm?event_id=316)
"2008 Freeway & Tollway Operations Conference," IBTTA and TRB, June 15-19, Hyatt Regency Bonaventure, Weston, FL
I'm giving a presentation on automated HOT lanes enforcement. (www.2008FTOC.com)
"SMC3 Summer Conference," June 25-27, Seaport Hotel, Boston, MA
I'm speaking on the Infrastructure panel at this logistics-oriented conference. (www.smc3.com)
"As the primary federal funding mechanism for our national highway system, the gas tax is increasingly outdated. When President Eisenhower envisioned our interstate system, he favored a user-pay method of financing its construction and maintenance. Unfortunately, he was limited by the technologies of his day. Now, however, we have exciting new financing mechanisms that are supplementing the gas tax while simultaneously reducing congestion. Through the broad deployment of high-speed open-road tolling technologies coupled with hundreds of billions of dollars of private capital, we can begin eliminating our dependence on a failed gas tax-based transportation model. It's time for our country to embrace a far more efficient, clean, and technology-based approach to charging for road use. This new approach will dramatically improve the quality and performance of our transportat ion systems. It will also give businesses and families the type of predictable and reliable service levels to which they have become accustomed when making phone calls, running the sink, or turning on the lights. We can also eliminate Washington's ability to use our transportation network as its own personal-and political-sandbox.
--DOT Secretary Mary Peters, http://fastlane.dot.gov, May 2, 2008.