Posted on 22 October 2008.
Are the Studies We Rely On Reliable?
We discovered “The Antiplanner,” Randall O’Toole, a few months ago, and ever since we have been publishing selected works by this prolific author and researcher. His findings, carefully documented, contradict important pillars of the conventional wisdom that informs modern urban planning – transportation options in particular. O’Toole’s work deserves as large an audience as possible because his conclusions, if correct, or even partially correct, have profound implications when determining how best to allocate taxpayer funds. If light rail, for example, is not nearly as cost-effective or even fuel efficient as cars and busses, for example, why are we building them?
In Sacramento, California, not only have hundreds of millions of taxpayer dollars already been spent on light rail systems that have done virtually nothing to take traffic off our freeways, but city planners are proposing the downtown streets get ripped up to make room for streetcars. Why on earth would anyone lay tracks onto a street for a streetcar that, unlike a bus, cannot even pull over and get out of traffic during stops? Randall O’Toole has developed compelling data to support what many of us feel in our gut – light rail and streetcars are not solving our transportation challenges.
This feature length investigation by O’Toole compares the benefits of streetcars vs. trolleys, and his conclusion is diametrically opposed to findings in a recent and authoritative study on the topic. Trolleys, busses with wheels and tires that drive around among cars and can, for example, pull out of traffic to make frequent stops, are probably far cheaper than streetcars. Instead of having to rip up the roads and install miles of steel rail for streetcars, you string overhead power lines that provide electricity to the trolleys. One still must wonder why a simple bus – modern and clean and green of course – would not be a far more versatile and cost effective solution than streetcars or trolleys.
Is it nostalgia that makes urban planners so fixated on anything but cars, busses, and roads to meet urban transportation challenges? Is it the thoroughly debatable yet rarely debated notion that cars and busses consume more resources and can never be “clean?” Is the preference for transportation solutions that rely on rail pushed by powerful special interests who love the ongoing pork and patronage such solutions require? Is it motivated by a sincere but misplaced utopian desire to force everyone into communal transportation arrangements? Whatever it is, renewed and vigorous debate on the subject of the car and bus vs. rail options is long overdue. Sometimes rail solutions do make sense, but not nearly as often as we are led to believe.
- Ed Ring
Hypothetically Optimal Transportation – Are the Studies We Rely On Reliable?
by Randall O’Toole, October 22, 2008
|A trolley bus in Vancouver, more
cost-effective than streetcars.
(Photo: Flickr: Jeffrey Beall)
“What is the optimal relationship between land use and transit,” asks Patrick Condon, “and what transit mode would best support this optimum state?”
In his research paper “A Cost Comparison of Transportation Modes,” published in September 2008 by the Foundational Research Bulletin, Condon concludes that cities should invest more in “trams” (streetcars) rather than in long-distance, higher-speed rail systems. Condon is a professor of landscape architecture at the University of British Columbia, where he is also involved in Sustainability by Design, which is trying to create a sustainable “vision” for the Vancouver region.
Condon’s answers to the above questions differ greatly from from the Antiplanner’s. This is partly because Condon bases many of his calculations on hypothetical numbers rather than actual data, and partly because his definition of “optimal” seems to transmogrify from paragraph to paragraph so that, in the end, it means whatever he wants it to mean.
Condon’s previous research shows a regrettable tendency to rely on myth and hearsay rather than actual facts. For example, a 2004 paper on urban design says, “National City Lines, a ‘transit’ company owned outright by GM, Firestone, and Phillips Petroleum was formed to purchase urban streetcar lines, notably in Los Angeles, with the intention of dismantling them. In 1949 GM was convicted of anti trust violations for this practice.” There are so many errors in this statement it is hard to know where to begin.
Start with the incriminating quotation marks around “transit,” which imply that National City Lines was not really in the transit business but in the transit dismantling business. In fact, National City Lines operated more than 60 transit systems between 1920 and the 1960s. General Motors and the other so-called conspirators only owned the company between 1936 and 1949. During that time, only 23 of the transit lines owned by National City replaced their streetcars with buses (and many of them had started dismantling their streetcar lines long before National City bought them).
National City owned only one of the two major transit systems serving Los Angeles, and that system still operated streetcars when National sold it to Los Angeles County in 1958 — it was the county that finally dismantled the streetcars. General Motors was convicted (and fined $5,000) for trying to monopolize the market for buses, but none of the other conspirators were convicted of anything, especially not for trying to dismantle streetcar systems.
The General Motors streetcar conspiracy has been repeatedly debunked by academic researchers who are willing to look at the facts and not just the myth. Condon’s willingness to perpetuate myths and hearsay is further revealed in an 2008 paper called The Case for the Tram: Learning from Portland. The thing I learned from the paper is that Condon doesn’t know much about Portland. He claims that Portland decided to build a streetcar line “for compelling reasons: it was inexpensive and the areas to be served were not dense enough to justify the more expensive MAX light rail system.”
|A modern streetcar – part of the new urban,
politically correct and “sustainable” solution.
(Photo: Flickr: NeiTech)
In fact, Portland’s first streetcar line connected the densest census tracts in the Portland area — Northwest Portland — with downtown. And, at $15 million per track mile plus a close to $2 million per vehicle, the streetcar was far more expensive than buses, which could have traveled through the area far more nimbly than streetcars. It is also worth noting that the streetcar was planned by the city, while the region’s transit agency thought so little of the streetcar route that it had never run bus service in that corridor.
Condon goes on to say that the decision to build the streetcar “was provoked by the electoral defeat” of light rail, which “left the city with only two options: forget transit or build it with their own money.” In fact, Portland decided to build the streetcar line in July, 1997, while the light-rail line was defeated at the polls in November, 1998. (And the transit agency is building the light-rail line — which did not go anywhere near the route of the streetcar — anyway.) Note also that Condon commits the common strategic misrepresentation of conflating “transit” with “rail transit.”
The Antiplanner’s suspicion that Condon relies on Portland official propaganda rather than actual facts is confirmed by his later claim that streetcars promoted urban redevelopment. As the Antiplanner has previously noted, Portland gave developers $665 million in subsidies to build along the streetcar line — something Condon fails to mention. But it is also interesting that Condon’s definition of “optimal” slips from “cost efficiency” in the first part of the Portland paper to “promoting economic development” in the last part.
In contrast, Condon’s more recent paper starts by considering dollar costs in the first paragraph, then shifts in the same paragraph to “sustainability” (which, in context, must have something to do with energy), but then in the second paragraph shifts again to greenhouse gas emissions, and finally in the third paragraph goes back to “long term cost efficiency.”
Condon’s fundamental problem is that you cannot “optimize” multiple variables. To find an optimum, you need to put everything in the same terms. This is what dollars are for: a medium of exchange between different goods. But planners often resist measuring everything in dollars, perhaps because they fear that if they do their preconceived notions will lose out.
In any case, Condon then says he wants to rate transportation choices using “three key sustainability principles”: “shorter trips are better than longer trips,” “low carbon is better than high carbon,” and “choose what is most affordable.” (Although he cites Sustainability by Design for these principles, that site has six principles, not three, and none of them are the same as any of his three.)
Note that his first principle immediately biases the results in favor of trams, which carry people short distances, rather than other forms of transportation that tend to carry people longer distances. Just why is this principle so important, and how are people supposed to apply it? People travel longer distances because the benefits they gain are greater than the added costs of travel. Condon simply ignores these benefits, which are crucial to any attempt to find an optimum.
For example, a major long-term economic trend has been the increasing specialization of work. Many people today have such specialized expertise that the local demand for their products or services could not possibly support them. Should we dispense with such specialists and rely instead on people who can’t do the job as well? Or should we concede that longer distance travel is sometimes worth the cost? And, if the latter, who gets to decide when it is worth it: the traveler or some planner?
In any case, Condon’s analysis of this first principle is skewed by the fact that North American streetcar lines tend to be very short. Based on his assumption that shorter trips are better, he asks: what mode works best for shorter trips? But, really, he is asking: what is the average length trip by mode? Lo and behold, streetcars have the shortest average trip length. That’s because most streetcar lines are short, so you can’t take longer trips. That doesn’t mean that streetcars are better for those short trips.
The longest trip lengths, Condon’s figure shows, are for automobiles. An economist would say that this indicates that autos give people access to more opportunities. But Condon’s strange, shorter-is-better criteria makes autos appear to be the worst choice.
Condon then asks which modes are the most energy- and carbon-efficient per passenger mile. Here he commits a whopper of a strategic misrepresentation by assuming that transit vehicles are, on average, half full, while autos carry, on average, only one or slightly more than one percent.
Both assumptions are wrong. In the U.S., the average car has 1.6 people in it, while U.S. transit vehicles run only about one-sixth full on average. Canadian transit agencies do not publish as detailed statistics as we have in the U.S., so we don’t know what the numbers are for Vancouver, BC, Condon’s target area. However, I suspect they are not much different. Per-capita transit ridership is higher in Vancouver than in comparable American cities, but per-capita vehicle kilometres of transit service is also higher.
Low occupancy rates are inevitable given transit’s fundamental characteristics. First, transit serves many outlying areas, but the vehicles only get full when they approach urban centers. Second, most transit ridership takes place during four to six weekday rush hours, but transit agencies typically offer services for 18 to 20 hours a day, seven days a week.
This means that buses or trains that look full in urban centers at rush hour are relatively empty on other parts of their routes and at other times of the day. The only transit services that have higher occupancy rates are commuter buses and trains that only run during rush hours, and Condon did not include these in his analyses.
For basic energy data, Condon also relies on “Strickland (2008),” but his references do not detail the name of this book or article. Condon probably means this Strickland paper, which is also based on a variety of questionable assumptions and sources.
In contrast to these hypothetical numbers, the United States has fairly precise data on actual energy use per vehicle mile and passenger miles per vehicle mile by mode, all of which were used in the Antiplanner’s paper on this subject. These data show that energy consumption for most modes of transit is not significantly lower than for automobiles.
For example, U.S. data show that buses consume about the same energy, per passenger mile, as SUVs. Buses and SUVs use about a quarter more energy than cars, which are about the same as light rail. Subways and commuter trains are about a quarter more efficient than the average car but much less efficient than the Prius.
These results are a sharp contrast to Condon’s largely hypothetical numbers, which show buses to be much more efficient than a Prius, and all forms of transit to be many times more energy efficient than either cars or SUVs. Once again, Canadian data may vary from the U.S, but unless we see actual numbers (passenger miles and vehicle miles by mode) from Canadian transit agencies, it is foolish to simply assume that Vancouver transit occupancy rates are three times higher than U.S.
Finally, in answer to the affordability criterion, Condon compares the capital and operating costs of various modes including a Prius and an SUV. “For detailed methodology,” he says, see the appendices — but these appendices are not available on line. Because the numbers Condon reports differ so much from actual numbers, I suspect that, like the energy data, his costs rely on hypothetical numbers.
For example, only in the screwy world of urban planners, where light rail is the default solution to just about anything, would streetcars appear to be cost effective. Condon points out that streetcars can carry more people and have longer lifespans than buses, which, he says, balances out their high capital costs.
This is a stretch. Portland’s streetcars have more standing room but only one more seat (41 vs. 40) than the average Portland bus. If they last twice as long and carry twice as many passengers, they are worth four times as much as a bus. Yet Portland paid more than five times as much for each of its streetcars (about $1.9 million) as the cost of a basic, 40-passenger bus (about $354,000 for a 40-foot bus — which typically has 39 to 43 seats — in 2005, several years after Portland bought its streetcars).
Even if the cost of streetcars per seat-year was lower than buses, this ignores the cost of the rails themselves. On top of that, Condon uses the absurd argument that, because streetcars can carry more people, “one tram driver is more than twice as productive per hour than is a diesel bus driver.” But the driver is only a tiny part of the cost of operating rail transit, most of which has to do with maintaining the rails and electrical facilities.
Condon’s other costs are as ridiculous as his energy estimates. In 2006, U.S. drivers spent an average of about 24 cents per passenger mile, including both capital and operating costs. Condon reports capital costs of 45 to 60 cents per passenger mile and operating costs of 60 to 75 cents per passenger mile. Canadian auto and fuel taxes are higher than in the U.S., but not sufficiently high to more than quadruple total costs.
In 2006, U.S. transit agencies spent 56 cents per passenger mile operating light rail. Condon says the cost is about half that. That would be consistent with his hypothetical assumption that occupancy rates are much higher than they really are.
Finally, Condon notes that fuel costs are likely to rise in the next 50 years, which causes the difference in operating costs between streetcars and SUVs to “skyrocket.” However, he fails to consider that automobile fuel efficiencies are certain to increase in the next 50 years as well. Historically, they’ve increased at a steady rate of about 1 to 2 percent per year, while transit energy efficiencies have declined.
Under the Energy Independence and Security Act of 2007, the standard for autos (including SUVs) will increase to an average of 35 miles per gallon by 2020. As new cars replace the existing vehicle fleet, the average auto on the road will be more energy efficient than any mode of rail transit by 2035.
Vancouver trolley bus: more cost-effective than streetcars at reducing greenhouse gas emissions.
Flickr photo by Jeffrey Beall.
Condon also fails to consider the high energy (and greenhouse gas) cost of constructing rail systems. If he were truly interested in reducing greenhouse gases, he would advocate the use of trolley buses, which have all the benefits of his trams without the high energy cost of construction.
Further, if Condon were truly interested in the long-run optimal solution, he would not be so quick to prescribe an inflexible technology that, once installed, is very hard to change. The great thing about autos is that the fleet turns over about every 18 years, so new technologies can quickly be implemented in response to changing needs such as higher energy costs. Rail systems last about 30 years, so if you build one that turns out to be less than optimal, you are pretty much stuck with it for a few decades.
In short, Condon’s analyses make three serious errors. First, his studies of streetcars rely on myth and rewrite history. Second, his comparison of transport modes relies on hypothetical data when real data are available (and very different). Finally, his definition of “optimum” changes so fluidly that he can come to any conclusion he likes (“shorter is better so therefore trams are best”) based on whatever definition he happens to choose.
About the author: Randal O’Toole is the author of Reforming the Forest Service, The Vanishing Automobile and Other Urban Myths, and The Best-Laid Plans, and edits the website The Antiplanner. This article originally was published on The Antiplanner on October 22, 2008, and is republished here with permission.
Email the Editor about this Article