Last August, the Reason Foundation released a report by Wendell Cox and Adrian Moore critiquing the privately funded XpressWest’s (Xpress) application for a federal loan needed to begin construction. At the time, Streetsblog was still so busy laughing at Reason’s attempts to discredit the Expo Line based on two dudes riding the car its first week of operation, we ignored the report. Besides, the Reason Foundation issuing a report that said that High Speed Rail ridership would be lower than expected or that operations would be costlier than expected isn’t news.
As Angie Schmitt at Streetsblog.net noted earlier in the year, before the report came out, “The Reason Foundation’s “research” on high-speed rail is pretty predictable. We know what this oil industry-backed think tank is going to say before they’ve said it: Ridership will be lower than expected; costs will be higher.”
If constructed, Xpress will be 185 miles of High Speed Rail between Southern California and Las Vegas. The line will run within or adjacent to the I-15 freeway and have no at-grade crossings with vehicle or pedestrian traffic. The non-stop trip between Victorville and Las Vegas could take 80 minutes. Xpress is a private, for-profit venture. Locally it is supported on the California side by conservative County Supervisor Mike Antonovich. On the Nevada side, it is a favorite project of Senator Harry Reid.
Turns out, we made a strategic error. Apparently Republican leadership in the House and Senate don’t realize that a Reason Foundation report on rail is about as credible as tobacco company report on cigarette health or a Disney funded report on a giant talking mouse’s impact on childhood development. Last week, House Budget Committee Chairman Paul Ryan and Sen. Jeff Sessions, the top Republican on the Senate Budget Committee, wrote a letter to Transportation Secretary Ray LaHood that Xpress is wasteful and too risky for taxpayers. They ask USDOT to deny a $5.5 billion loan for the project construction.
Ryan and Sessions based their letter on the Reason report from last August. However, since it’s not good enough to just laugh at them, especially since Ryan is still treated with reverence as a serious thinker by much of the political press. Streetsblog would like to note that attacking Reason’s report is not a blanket endorsement of Xpress, sometimes lampooned on Streetsblog L.A. as the “Gamblin’ Train to Vegas.” There are some concerns with the project planning, none of which were shored up when the CEO for Xpress defended the project to the press this week in the vaguest possible terms.
Now with the disclaimer over, let’s get to work.
The Reason Foundation report gives six reasons that a federal loan is an extremely risky proposition in their “taxpayer risk assessment.” Four of the six concerns have to do with claims that ridership will be lower than expected. The other two are concerns that construction costs will come in higher than expected and that building the rail line might make it harder and more expensive to widen the I-15. The widening of the I-15 is a given in the report. Despite the “libertarian” leanings of the Reason Foundation and their concern about government recouping their money, they are almost always in favor of costly highway expansion even if there is no tolling involved and the government will never get one cent of their investment back.
The attacks on XPress’ ridership figures are some of the larger leaps in the report, but are the underpinnings of all the arguments. On one hand, Cox and Moore get technical, using the “International Average Error Forecast” which somehow shows that ridership on rail projects is over-estimated by at least 39%. There is never a reference given to how this number is calculated. In fact, if you google the term “International Average Error Forecast” you find just a handful of references, all of them related to this report. The authors created a way to devalue transit projections for the report, slapped an official name on it, and never bothered to explain it.
This is the entire explanation of the forecast in the report:
International Average Error Forecast
The International Average Error Forecast (developed for this report) applies the average ridership forecasting discrepancy from the international research to the Project Forecast: Current Data.
Ridership would be 39% less than the Project Forecast: Current Data (53% below the FEIS 2005 Data Forecast). The International Average Error Forecast yields an annual ridership of 4.2 million in the third year.
To be fair, Morten Skou Nicolaisen points out in his doctoral thesis “Forecasts: Fact or Fiction? Uncertainty and Inaccuracy in Transport Project Evaluation” at Aalborg University that Cox’s imaginary number is the same as a conclusion reached by Bent Flyvbjerg, Nils Bruzelius and Werner Rothengatter. Flyvbjerg an academic (who is yes…international!)referenced in the Reason Foundation report. However, Nicolaisen goes on to point out that Cox’s use of said research takes it out of context and mis-uses it.
A recent example is Cox’s (2012) reassessment of the XpressWest project for the Reason Foundation, which is conducted by adjusting ridership figures downwards based on an ‘International Average Error Forecast’ factor without any discussion of why this adjustment might be relevant for the XpressWest project. This adds no useful information and only reduces the ability of the appraisal to function as decision support. I assumed the adjustment factor is based on the results from Flyvbjerg et al. (2005), as the figures are identical. However, the approach used by Cox is far too simple, and not in line with Flyvbjerg’s suggested principles of reference class forecasting. A rail line through a flat area with no grade separation is hardly in the same reference class as metro systems built in developing countries during the 1980s. Having ‘rail’ as a reference class is thus not of much when it is applied so carelessly.
Unfortunately, Cox and Moore fail to take into account the Transportation and Research Devaluing In Statistics Theory (TARDIS) that I just created. TARDIS Theory states that when report authors make up important sounding ways to prove their point but never explain it, you should point out their lack of seriousness and move on.
This is not to say that there are errors made when it comes to ridership projections. However, to state that international high speed rail trains routinely don’t meet ridership expectations, again without showing this to be true, is not intelectually honest. It might appear correct to people that think transit is a waste of money, but in fact many high speed trains do make money. As the California High Speed Rail blog points out in their review of the report.
HSR systems around the world routinely meet their ridership projections, though it takes about five years to reach that level. HSR systems around the world also generate enough revenue to cover the cost of their operations, including the much slower Amtrak Acela. There is no HSR system in the world that has been a failure from the ridership or operating revenue perspective, and as anyone who has been stuck in the horrible Interstate 15 traffic to and from Vegas knows, there is very much a market for this train.
The CAHSR Blog goes on to point out that putting too great a debt burden on a high speed rail project can be problematic, and points to a high speed rail program in Taiwan that is failing to meet its obligations. The difference between Xpress and the Taiwan situation is that by getting a government loan, the payback schedule for Xpress is less agressive. Xpress will have time to build its ridership before the bulk of its payments are due.
The International Error Forecast is one of several ways that Cox and Moore attack the projections. At one point they argue that ridership will be between 3% and 25% of Xpress’ forecast. One of the main reasons is that people will not switch modes from Los Angeles to Las Vegas for part of a trip unless substantial time savings are included. They then show that there are no time savings using Xpress by estimating the travel times for car drivers during off peak periods. In fact, they actually add a congestion buffer to the train trips, but assume free flowing traffic for the car trips in their comparison. Back in August, during a very thorough debunking of the report, Pedestrian Observations referred to this as a “Zinger.” “Completely intellectually dishonest” would be more accurate.
To further debunk the ridership numbers, the authors compare it to the higher-than-average speed ACELA regional train that connects New York City and Washington D.C. on the east coast. The comparison falters for a number of reasons, most notably that the ridership on ACELA is only a fraction of the ridership between the two cities that also include NJ Transit, Maryland Transit Association’s MARC rail, and regional AMTRAK trains.
While ACELA ridership between the two is only gets 2 million riders, leading Reason to wonder how Xpress can expect 9 million, the total ridership in the corridor is much higher. Pedestrian Observed puts it at 8 million just including the AMTRAK service. Adding the regional rail it’s much higher. Xpress will not have such competition in its quest to reach 9 million riders.
Xpress will run between Victorville in Southern California and Las Vegas. This location is the cause of many of the uncertainty the ridership forecast faces and is one of the issues highlighted by Cox and Moore.
Cox and Moore argue that because people will have to drive well over an hour to arrive at Victorville from Downtown Los Angeles or Orange County, that they simply will choose not to switch modes. If the line began in Los Angeles Union Station, or even Palmdale, which has a Metrolink station, more people would flock to it.
This is another curious argument, because Cox and Moore seem to argue that people would be willing to transfer from one mode of transit to another in either Palmdale or Los Angeles, but not willing to transfer from car to transit in Victorville.
While this may be the case, it doesn’t mean that transit riders won’t transfer at the beginning, as evidenced by the full park-and-ride lots in transit stations around the country. Imperically, most people prefer transfers at the end of trips because that is the shortest part of the trip, the so-called “last mile.” The case of a drive to Victorville is different than say a hop from the Expo Line to the Blue Line in Downtown L.A.
While this means that it is quite unlikely that residents from Las Vegas and the surrounding areas are going to take Xpress to access Los Angeles, something Xpress also assumes won’t happen; it doesn’t say much about whether people will take the train to access Las Vegas. Cox and Moore argue that riders prefer transfers at the end, not beginning. The report even sites that 55% of transit riders prefer to transfer at the end of their trip, not the beginning if they must transfer.
A last argument against the ridership comes because the economic situation has changed between when the ridership study was completed in 2005 and today. Specifically, Las Vegas was booming in 2005 and is less booming today. While the national economic downturn has definitely hit Las Vegas’ tourist scene, an economic rebound would likely boost it.
Again, this is not to day that Xpress will meet the ridership it expects. It will be competing with the I-15 and with short-hop air trips. Predicting ridership isn’t an exact science. However, despite its length and serious sounding phrases, this report doesn’t offer any ridership arguments other than to say it doesn’t believe the ridership numbers and then points to research showing that other ridership estimations for completely different projects have been wrong in the past.
It’s not just ridership figures that come under fire by Cox and Moore. Reason also argues that the cost of the project is too low and that the money would be better spent just widening the I-15. It’s true that a road widening would be cheaper, and it’s hard to say that the cost given by Xpress is completely acurate as large mega-projects, such as one with a $6 billion price tag, can have thousands of things go wrong that could drive up the cost. However, that same argument can be given for pretty much any large project, government funded or not.
The report provides no specifics as to why the costs could be higher that are project specific. Cox and Moore point to the work of Flyvbjerg etal that show that over-runs are common in mega-projects. However, again comparing the Xpress project with Boston’s big dig, or high speed rail lines in Europe or Asia doesn’t really do much to show why Xpress wouldn’t meet its construction cost goals. In fact, there are many reasons to believe this line will, or will at least be close, including:
- Xpress does not have to go underground or through mountains. No tunneling;
- Xpress doesn’t have to cross highways, local streets, pedestrian trails, or anything else that requires a signal;
- The project area is mostly flat;
- Labor costs are low because of the aforementioned recession;
- Material costs are low because of the aforementioned recession;
- There is minimal real-estate costs because the project is occurring mostly in an existing right-of-way.
Streetsblog doesn’t claim to be experts in interest rates or federal loan processes, so we’re not counter-urging USDOT to approve the Xpress loan request. However, basing one’s views on the project on a half-baked Reason Foundation report says more about the people conned by Reason than it does the project in question.