Surface Transportation Newsletter #114

Surface Transportation Innovations Newsletter

Surface Transportation Newsletter #114

Conflicting reports on California high-speed rail, rethinking the federal role

In this issue:

Conflicting Reports on California High-Speed Rail

Bent Flyvbjerg of Oxford University has done some of the most important research on the riskiness of transportation megaprojects, both highway and rail. His 2003 book, Megaprojects and Risk (Cambridge University Press) is still unequalled, and he has continued work on this subject since then. Needless to say, I was delighted to receive an email from him in mid-December, enclosing a powerful new journal article, “Quality Control and Due Diligence in Project Management,” from the International Journal of Project Management. The email also noted that the U.S. GAO was going to apply the methodology in this paper to its assessment of how the California High-Speed Rail project has developed and presented its case for this project.

The GAO report, “California High-Speed Passenger Rail: Project Estimates Could Be Improved to Better Inform Future Decisions,” was released in March (GAO-13-304). I read all 63 pages and skimmed the five appendices. Alas, although Flyvbjerg’s journal article was listed as part of the academic literature consulted in the study, there is no indication anywhere in the report of GAO’s team applying any of its ideas. Instead, the GAO report mostly concerns whether CHSRA and its consultants followed conventional best practice in making their projections of costs, ridership, revenue, etc. But according to Flyvbjerg, it is that conventional “best practice” that is the problem!

His journal article draws upon Nobel-Prize-winning research on decision-making under uncertainty, which argues that inaccurate and biased forecasts result from a “systematic fallacy in decision making causing people to underestimate the costs, completion times, and risks of planned actions, [while] overestimate[ing] the benefits of the same actions.” The cure for this, he says, is ensuring an “outside view” of the planned actions that compares the figures for the proposed project with actual outcomes from a range of comparable projects. Due diligence is “a method for consistently taking an outside view of planned actions.” He argues that project proponents’ “forecasts should be made subject to quality control and due diligence by other parties, including banks and independent bodies such as national auditors or independent analysts.”

Flyvbjerg illustrates this methodology with a case study from a European country of a proposed multi-billion-dollar rail project, which he dubs the “A-Train.” In this particular case, the focus of the due diligence was the project’s demand forecast. I don’t have room to describe all eight steps, so I will just include three of them. Benchmarking means comparing the forecast with average and median practice, not best practice. In the A-Train example, he draws on a sample of 62 rail projects of comparable scope, finding that in 53 of these cases the demand forecast was overestimated. Overall, the average project achieved only 59% of forecast demand in the first year of operation. The next step is to compare the confidence level of the project’s forecast with the standard deviation of accuracy in the benchmark cases. The A-Train’s worst-case (downside) demand forecast was 9.1%–compared with an average of 33% for the benchmark cases. And instead of the A-Train’s 5% likelihood of a demand shortfall of 15% or more, the benchmark cases indicate an 80% likelihood.

There’s a lot more, but let me skip ahead to the piece de resistance. That is to compare the forecaster’s track record with actual project experience. In the A-Train case, the global firm responsible for the demand forecast listed credentials from more than 20 previous forecasts for this type of project. “But when asked for evidence of track record for the accuracy of these forecasts, no evidence was provided.” Two of the cited projects were, however, in Flyvbjerg’s database. Project 1 showed a 4th-year demand overestimate of 250%, and Project 2 showed a 150% overestimate of 5th-year demand.

Summing up, Flyvbjerg minces no words about the A-Train demand forecast. Noting that “this is a normal, state-of-the-art demand forecast, like thousands of such forecasts made every year at substantial cost to clients.” But he concludes, “The forecast is garbage, of the type garbage-in/garbage-out. In fact, the forecast is worse than garbage, because it gives the client, investors, and others who use the forecast the impression that they are being informed about future demand when, in fact they are being misinformed. Instead of reducing risk in decision-making, forecasts like this increase risk by systematically misleading decision makers and investors about the real risks involved.”

As I noted previously, the GAO assessed the CHSR project not in this due diligence manner, but instead on whether the Authority and its consultants followed the conventional forecasting methods-the very kind that a decade of Flyvbjerg’s research has shown to be fundamentally flawed. Thus, the crowing by promoters of this project over what they deem its vindication by GAO is misplaced, and the serious critiques by the California Legislative Analyst’s Office, the project’s official Peer Review Group, transportation researchers at UC Berkeley, and several other outside groups remain valid.

One of the very first due diligence reports was carried out by the Reason Foundation prior to the ballot measure authorizing state general obligation bonds for the project in 2008. It questioned the cost estimates, ridership projections, revenue forecasts, trip times, and other factors-and every one of those concerns has been validated by other researchers in the five years since then. The same research team has just published a review of the CHSRA’s 2012 Draft Revised Business Plan. Suffice it to say that the project looks no less risky today than it did in 2008. You can read the executive summary at, and from there you can click on the full report for far greater detail.

? return to top

Rethinking the Federal Role

For the most part, devolving most or all of the federal surface transportation program to the states has been a proposal from conservatives, dating back to the Mack-Kasich bill in the late 1990s that garnered supports from a number of governors and state DOTs in “donor” states (those which typically contribute more in federal fuel tax revenues than they get back in federal highway and transit aid). But the last few months have seen serious calls for devolution from liberals.

The first one on my radar came last year from Harvard urbanologist Edward Glaeser, lamenting the “outsized role” of the federal government. This was followed in January by a Bloomberg opinion piece by former New York City planning chief Rohit Aggarwala, “Want Better Roads? Kill the Gas Tax.” Like many other critics, Aggarwala pointed out that after the Interstate highway system was completed, the federal program lost its national mission and became more of a public works free-for-all, and voters lost confidence in it. Ending the program, he suggested, “would force a serious discussion in each state about how, and how much, to fund roads and transit.” And, he predicted, “Because the bigger states have the biggest needs, this would probably increase investment nationally.”

And last month, in a piece titled “The End of Federal Transportation Funding as We Know It,” Eric Jaffe of Atlantic Cities took the case for reducing or eliminating the federal program quite seriously, citing researchers for and against while pointing out that to some extent, devolution is taking place de-facto, as Congress fails to increase the federal gasoline and diesel tax rates while voters increasingly approve state and metro area transportation funding proposals. He also cites Rob Puentes of Brookings Institution, who has called for a new model in which the feds would still play a funding role, but with states and metro areas taking the lead.

Jaffe’s piece alerted me to an important empirical study by Pengyu Zhu of Boise State University and Jeffrey Brown of Florida State University, published in the journal Transportation‘s first 2013 issue. (“Donor States and Donee States: Investigating Geographic Redistribution of the U.S. Federal-Aid Highway Program, 1974-2008”) Zhu and Brown use extensive data to investigate four possible determinants of how federal highway money gets distributed among the states: based on highway system need, to aid poorer states, due to rural bias and programmatic inertia, and due to political power and pork-barrel incentives. Their detailed regressions reject the first two hypotheses, finding that over this long time period federal highway aid has disproportionately benefitted states with less-extensive highway systems and less highway use, and states with higher per capita incomes. But redistribution does reflect a rural bias and stronger political representation on key congressional committees. To me, this says the case most people have long accepted for geographic redistribution is on very thin ice.

I see this renewed debate over the federal program as very healthy, especially at a time when the federal program has come to depend on significant general-fund bailouts of the Highway Trust Fund, undermining the sensible users-pay/users-benefit principle. It’s also very timely when the entire scope of the federal government needs to be rethought, to bring federal spending in line with federal revenues.

As Adrian Moore and I argued at some length in the 2010 Reason Foundation policy study, “Restoring Trust in the Highway Trust Fund,” this is a time for sorting out which functions of government are properly local, state, and federal. In surface transportation, we argued, facilitating interstate commerce is included in the Constitution as a federal power-and that appears to justify something like the Interstate highway system and possibly other major highways that link states together. The vast array of other programs now funded out of federal highway user taxes make far better sense as state and local responsibilities. ( If all the non-highway programs were eliminated from the Highway Trust Fund, the result would be a 30% increase in available highway funds. That would eliminate the $10-11 billion per year in general fund bailouts, restoring solvency to the Trust Fund.

But narrowing the focus of the Trust Fund to major inter-city highways could also give the federal program a new mission: reconstructing and modernizing the aging Interstate system, which with only 2.5% of total highway lane-miles handles nearly 25% of all vehicle miles of travel (VMT). Since the cost of that program would likely be several trillion dollars, much of the cost will likely need to be financed, and the coming shift from fuel taxes to mileage-based user fees would make that financing far more feasible.

? return to top

Passenger Rail and RRIF

Last September, I wrote about several passenger rail proposals from the private sector, one of which had applied for a $5.5 billion federal loan from the Railroad Rehabilitation and Improvement Financing Program (RRIF). The project, called XpressWest, looks to me like a costly boondoggle that is highly likely to default on any such loan, since its ridership projections (from Victorville to Las Vegas) appear highly exaggerated. This project was the subject of a Reason Foundation due diligence analysis last summer (/wp-content/uploads/2012/08/xpresswest_victorville_las_vegas_train.pdf). At the time the study was released, approval of the loan request by the Federal Railroad Administration was considered imminent, but that approval has still not occurred.

Last month the Chairman of the House Budget Committee, Rep. Paul Ryan (R, WI) and the Ranking Member of the Senate Budget Committee, Sen. Jeff Sessions (R, AL) sent a three-page letter to Transportation Secretary Ray LaHood expressing concerns about the riskiness of the XpressWest project. In addition to citing the concerns raised in the Reason study, they noted that a Congressional Research Service assessment of high-speed rail found that “Few if any HSR lines anywhere in the world have earned enough revenue to cover both their construction and operating costs, even where population density is far greater than anywhere in the United States”-which suggests a high probability of default. The letter closed by urging the Administration to reject the requested loan for XpressWest.

Few people outside of railroading have even heard of RRIF, but those who have may assume that it is similar to the relatively successful TIFIA program for highways and transit. But there are some profound differences. TIFIA is intended to provide only gap financing to projects that (1) gain an investment grade rating on their senior debt, and (2) have a dedicated funding source or sources that can realistically be shown to repay both the senior debt and the subordinated TIFIA loan. As originally enacted, the maximum TIFIA loan amount was for 33% of a project’s budget. MAP-21 unwisely increased this to 49%, but thus far FHWA is sticking with the traditional 33% maximum. (The EU’s new Project Bond Initiative has even more stringent criteria for its loans-no more than 20% of the amount of the project’s senior debt.)

By contrast, RRIF loans can be for up to 100% of a project’s budget, and the description on the FRA website makes no mention of any requirement for an investment-grade bond rating. Since the RRIF program is up for reauthorization this year, it would be wise for Congress to redesign it along the same lines as TIFIA, in the interest of protecting taxpayers from boondoggle projects.

Incidentally, a recent article in the Las Vegas Sun repeatedly mis-characterized RRIF loans as a $35 billion fund “financed by the railroad industry.” Reporter Karoun Demirjian may have picked up this disinformation from Sen. Dean Heller (R, NV), whom she quotes as saying that RRIF is “funded by the railroad industry for this purpose.” That is completely false; it is federal taxpayers who are on the hook for any RRIF loans that default.

The only other non-Amtrak passenger rail project that has applied for a RRIF loan is All Aboard Florida, which is pursuing a higher-speed rail project mostly on existing right of way of parent company Florida East Coast Railway. The amount of the loan request, which was filed on March 15th, was not revealed either by AAF or the FRA. While this project looks far less risky than XpressWest, I would be far more comfortable if RRIF were operating under TIFIA-like safeguards for taxpayers.

And getting back to Las Vegas, it turns out there are two other private-sector proposals to restore passenger service between there and Southern California, of which there has been none since Amtrak pulled the plug on its money-losing Desert Wind train in 1997. A company called Las Vegas Railway Express is working to gain trackage rights from Union Pacific Railroad for its X Train that would link Orange County (at Fullerton) with Las Vegas on a luxurious (but conventional speed) “party train.” And Trains columnist Don Phillips disclosed last month that a start-up company called Pullman Palace Car Co. plans high-end passenger service from Los Angeles Union Station nonstop to Las Vegas. (The company also has plans for dedicated freight service.)

Both of these plans look more viable to me than Desert Xpress, primarily because their starting points in greater Los Angeles are in the heart of massive population centers, rather than in far off Victorville. Their costs are far lower, since they would use existing rail lines and eschew “high-speed” technology-similar to All-Aboard Florida. Neither of these L.A. to Vegas companies has applied for a RRIF loan, and if their numbers actually pencil out, they ought to be able to finance their projects privately.

? return to top

Changed Thinking on Climate Change

This is not an environment newsletter, but since a growing amount of transportation policy stems from concern over global warming, it’s important for transportation people to keep up-to-date on this critically important subject. To that end, you should read two recent articles that suggest important shifts in thinking about this problem.

The first is a three-page article taking up the entire Science and Technology department of the March 30, 2013 edition of The Economist. It reports that despite high temperatures recorded in 2012, the five-year mean global temperature has been flat for a decade. As the magazine notes, “The mismatch between rising greenhouse-gas emissions and not-rising temperatures is among the biggest puzzles in climate science just now.” It goes on from there to a thoughtful discussion of the limitations of current climate models. It also points to several recent studies, by reputable researchers, that come up with lower projected increases in global temperature than “mainstream” climate science as reflected in periodic reports from the Intergovernmental Panel on Climate Change.

The question under review here is climate sensitivity to CO2 increases: “If temperatures are likely to rise by only 2? C in response to a doubling of carbon emissions . . . perhaps the world should seek to adjust to (rather than to stop) the greenhouse gas splurge.” The article goes on to discuss several different approaches to climate modeling and such unresolved questions as the impact of clouds; the IPCC itself notes that “cloud feedback remains the most uncertain radiative feedback in climate models.”

Another surprising article is a major feature in the current (Vol. 116, No 2) issue of MIT’s Technology Review. Editor David Rotman’s article, “A Cheap and Easy Plan to Stop Global Warming,” provides a careful pros-and-cons assessment of Harvard physicist David Keith’s proposal to mitigate global warming by injecting sulfate aerosols at 20 km. altitude over a period of many decades. Once dispersed globally, the aerosols will reflect a portion of the incoming solar radiation back into space, increasing Earth’s albedo and partially offsetting the warming effect due to growing levels of greenhouse gases in the atmosphere. Keith’s proposal falls in the category of geoengineering, and it is considered almost sacrilege by some climate scientists and environmental groups. Rotman’s article gives plenty of space to scientist critics, but ends up concluding that Keith’s proposal is definitely worth a serious test, and that if it works, it might be far more cost-effective (as well as more do-able) than drastic, worldwide reductions in CO2 emissions.

What is notable about both articles is who published them. Over the past decade, both The Economist and Technology Review have mostly presented what the former dubs “mainstream” climate science, while just about ignoring other approaches to modeling and mitigation. Neither publication could remotely be categorized as a “climate skeptic,” but both are seriously interested in improved understanding of climate change and of sensible policy approaches to deal with it.

And that brings me to a third article, in its own way also something of a milestone change. In its March 25th issue, National Review published a thoughtful article by Oren Cass, “The Next Climate Debate,” arguing that “conservatives should accept the science and focus on policy.” It’s an excellent piece, pointing out the absurdly low cost-effectiveness of a number of what I would consider “feel-good” policies that produce miniscule reductions in greenhouse gases (on a global scale) at very high cost. Developing sensible policies for dealing with global warming is the real challenge, in transportation as well as the broader realms of energy use. It’s good to see thoughtful conservatives moving in this direction.

? return to top

News Notes

Reason Privatization Report on Highways. The highways section of Reason Foundation’s Annual Privatization Report was released last week. Researched and written by Robert Poole, it covers state PPP enabling legislation, MAP-21, transportation infrastructure finance, major highway PPP projects, leasing of existing toll roads, and updates on managed lanes and networks. Go to:

Miami “Enclosing” Open-Access Toll Roads. The board of the Miami-Dade Expressway Authority (MDX) voted last month to convert its two busiest toll roads to all-electronic tolling (AET). The new toll system will charge all drivers, rather than only those who go through barrier toll plazas. MDX estimates that under the current system, only 55% of those using the two tollways pay for doing so. The AET system will involve installing 20 new gantries over the next year or so. MDX did a similar conversion of its three smaller tollways several years ago.

Houston as Model 21st-Century Metro Area. Despite many planners’ commitment to increasing density and enhancing traditional central business districts, the actual demographics of urban areas paint a far different picture. In a data-rich article at The Daily Beast, urban geographer Joel Kotkin finds that low-density, multi-centric Houston is far more typical of where Americas metro areas are heading than what is envisioned by “Smart Growth” planners. Whether you like this trend or hate it, you should read Kotkin’s thought-provoking article, “Houston Rising-Why the Next Great American Cities Aren’t What You Think.” (

All National Highways to Have Electronic Toll Collection System by 2014. That headline took me aback when it crossed my screen on March 29th, but then I read the article below it. Turns out the story was from The Economic Times of India, and the announcement was from the country’s Road Transport Minister. Nearly all such highways in India are already toll roads, as are most of those in France and Italy.

Impressive Transit System Revamp in Tallahassee. Last year in this newsletter I wrote about the conversion of the transit system of Florida’s capitol city from a traditional radial (downtown-focused) model to an extensive grid system more in keeping with the locations of employment throughout the metro area. Governing magazine last month published a great overview article on what was done, why it was done, and how it has worked out. Check out “How Tallahassee Overhauled Its Transit Overnight,” by Ryan Holeywell, Governing, March 12, 2013.

New Database on State Highway Finance. AASHTO’s Center for Excellence in Project Finance last month unveiled a new online database on highway finance, developed in cooperation with the National Conference of State Legislatures. The database includes a detailed profile of highway finance in each of the 50 states plus Puerto Rico and the District of Columbia. Go to:

Golden Gate Bridge Goes Cashless. Joining a growing array of tollways that are eliminating cash tolls and implementing all-electronic tolling (AET), the Golden Gate Bridge ended cash tolling late in March. Motorists wishing to pay in cash can do so off-line, at 10 off-road locations with touchscreen kiosks. Of the 30 remaining toll collectors, some retired, others were transferred to other jobs at the toll authority, and the remainder were laid off with severance pay.

Excellent Congestion Pricing Presentation. Stockholm remains the best example, in my view, of introducing cordon pricing to reduce congestion on roadways approaching its city center. At a recent TED conference, Jonas Eliasson gave a short, punchy overview of what was done, how it was done, and what the results have been. Worth watching when you have 10 minutes or so to spare. Go to:

Feedback on Jones Act Article. I received a response to last month’s article in which I criticized the Jones Act, which requires all water-borne cargo between coastwise ports and between the continental United States and Guam, Hawaii, and Puerto Rico to be transported in U.S.-owned, U.S.-built, and U.S.-crewed ships. James Henry, president of the Transportation Institute, defended the Jones Act on grounds of “national defense and a strong domestic maritime industry.” He cited the Navy as favoring the availability of such ships during war, noting that Jones Act ships “transported 90% of all military cargos moved to Afghanistan and Iraq.” He also cited the U.S. jobs created by the domestic maritime industry. I agree that it’s useful to have ships available for call-up during wartime, but there is no need for the Sealift Readiness Program to enroll only Jones Act ships. The comparable program for airlift-the Civil Reserve Aircraft Fleet-works with U.S. airlines, but allows foreign-made aircraft (such as Airbus wide-bodies) to be used.

Kansas Turnpike Saved from Takeover. In a face-saving move, the Kansas legislature junked Gov. Sam Brownback’s proposal to merge the Kansas Turnpike Authority into the state DOT, which I had argued against ( The compromise measure keeps the Turnpike separate but bizarrely makes the state DOT secretary the Turnpike’s chief operating officer, instead of just a board member, for the next three years. For details, go to

Feedback on I-90 Bridge in Seattle. I received a number of emails in response to my article about plans to replace the reversible HOV lanes on the I-90 bridge with light rail. First, I was mis-informed about the lack of replacement capacity for motor vehicles. The WSDOT plan calls for narrowing the lanes and shoulders so as to add one HOV lane in each direction, in addition to the light rail line replacing the two reversible HOV lanes. And the decision to put light rail on the bridge was made by voters in 2008, and I’m told would be politically impossible to reverse. That is unfortunate, since a BRT/HOT solution would cost far less and produce greater mobility benefits than dedicating a 38-ft. wide right of way to light rail.

? return to top

Quotable Quotes

“Forecasters’ unwillingness or inability to document the accuracy of their forecasts runs directly against much-touted policies for evidence-based decision making and management, and it amounts to forecasters effectively asking their clients to buy-both in the sense of believing in and paying for-their product without evidence to substantiate that the product actually works, or has worked in the past. The type of forecasting we are studying here is globally a billion-dollar business. Very few other businesses of this size could get away with not demonstrating the track record and quality of their product. The point is that the forecasting business should also not be allowed to do this, and there are no defensible reasons that they are. It is the task of quality control and due diligence to point to this problem and to make sure the track record is documented to the extent possible.”
-Bent Flyvbjerg, “Quality Control and Due Diligence in Project Management: Getting Decisions Right by Taking the Outside View,” International Journal of Project Management, November 2012

“The Highway Trust Fund no longer can serve as a source of capital for new infrastructure, and funding large capital-intensive projects with current user fee revenues on a pay-as-you-go basis is no longer feasible. Instead, look for the states to assume responsibility for remedial ‘fix-it-first’ activities, and for a shift from funding to financing for multi-year construction megaprojects. This may turn out to be the only practical long-term solution to our transportation funding dilemma.”
-Ken Orski, “Wanted: A Reasoned Approach to Dealing with America’s Infrastructure Needs,” Innovation Briefs, March 16, 2013

“Railroads compete for only a very small share of highway freight. Keep in mind that roughly 60% of highway freight travels no more than 50 miles from a central point. It’s basically local delivery. Most medium-range truck routes of 250 to 500 miles are not competitive for intermodal. Also, long-haul intermodal freight is unloaded at yards outside cities, then trucked to customers who are often in congested areas. Therefore, intermodal does almost nothing for urban congestion. That leaves a fairly small amount of freight that is intermodal-compatible. Railroads would consider a doubling of intermodal traffic to be a massive increase, but it would have little noticeable effect on truck traffic, especially in urban areas.”
-Don Phillips, “The Northeast Corridor and the GOP,” Trains, April 2013

“The U.S.’s 14th place ranking in the World Economic Forum’s infrastructure index scarcely bespeaks a national scandal. Luxembourg and Canada rank just above the U.S. and Austria and Denmark rank just below. None of these countries are exactly slouches in the infrastructure category. Among the 20 largest countries, the U.S. ranks second only to Canada. The World Economic Index also shows that U.S. infrastructure beats the European Union average by a wide margin. How can that be with the high speed rail and gleaming autobahns of the European Union-the envy of our transportation bureaucrats? Consider another hitch. OECD infrastructure experts find that Europe has too much supply of roads and rail relative to demand. Yes, they have trains departing every few minutes, but half empty, and do Germans really need five different autobahns to drive from Munich to Frankfurt? The same OECD experts find that the U.S., Canada, and Australia have built about the amount of infrastructure that fits the demand.”
-Paul Roderick Gregory, “Infrastructure Gap? Look at the Facts. We Spend More than Europe,”, April 1, 2013

“Under almost any imaginable scenario, we are unlikely to see the creation of regions with anything like the dynamic inner cores of successful legacy cities such as New York, Boston, Chicago, or San Francisco. For better or worse, demographic and economic trends suggest our urban destiny lies increasingly with the likes of Houston, Charlotte, Dallas-Ft. Worth, Raleigh, and even Phoenix. The critical reason for this is likely to be missed by those who worship at the altar of density and contemporary planning dogma. These cities grow primarily because they do what cities were designed to do in the first place: help their residents to achieve their aspirations-and that’s why they keep getting bigger and more consequential, in spite of planners who keep ignoring or deploring their ascendance.”
-Joel Kotkin, “Houston Rising-Why the Next Great American Cities Aren’t What You Think,” The Daily Beast, April 8, 2013

? return to top