A Free-Market Energy Blog

Urban Rail Transit: On the Wrong Track

By Randal O'Toole -- April 21, 2010

In 2006, Nashville began operating the Music City Star, a commuter train between Lebanon and Nashville. Transit officials brag that this is “the most cost-effective commuter train in the country” because they spent only $41 million to begin service.

To be cost-effective, however, you need more than just a train: that train needs to produce something. The Music City Star carries only about 250 commuters on round trips each day, riders who could easily have been accommodated in a few buses costing less than $3 million. In fact, it would have been less expensive to give each of those commuters a new Toyota Prius every year for the next 30 years than to operate the train.

Since 1970, American cities have spent about $100 billion building new rail transit lines, and virtually all of it has been wasted. Rail transit was rendered obsolete in the 1920s by the development of reliable buses that could go on any streets open to automobile traffic. Since the cost of the streets was shared with autos and trucks, the capital and maintenance requirements for buses are far lower than for rails.

After 1920, some 700 American cities with streetcar or other rail transit systems converted those mostly private systems to buses. By 1970, only eight urban areas still had some form of rail transit. Today, rail transit can be found in more than 30 urban areas, and the number is growing. This turnaround is largely due to perverse incentives in federal transportation funding that rewards transit agencies for selecting high-cost transit solutions when low-cost solutions–usually buses–would work better.

Predictably, the rail construction boom has generated a huge lobby for more federal funding for rail transit. The American Public Transportation Association, whose members include numerous construction and engineering firms, has a $20 million annual budget, which is several times greater than the annual budgets of all of the various highway groups in Washington combined. Most contractors that can build highways can make even more profits building rail lines, so they have no interest in opposing rail.

One of the big arguments for building rail transit is that it will save energy. But, as I found two years ago in a Cato report, the energy saved by using steel wheels instead of rubber tires is offset by three energy costs.

First, for safety reasons, rail cars must be very heavy, typically weighing 50 tons. Since rail cars carry an average of about 25 people, this means an average weight of about 4,000 pounds per person. That’s more than twice the weight-per-passenger ratio of the average automobile.

Second, constructing rail lines consumes a huge amount of energy, and if those lines are lightly used the energy cost per passenger will be much higher than for highways, which are typically much more heavily used. Portland estimated that the energy cost of constructing one of its light-rail lines was 172 times the annual energy savings.

Third, rail transit rarely operates by itself, and instead is typically supplemented by shuttle buses that are supposed to feed passengers into the rail stations. But those buses tend to be very lightly used, while the rail lines themselves have taken passengers away from corridor bus routes that tend to be heavily used. The net result is the energy consumed by a transit system per passenger mile sometimes actually increases after the system opens a new rail line.

Rail advocates love to point to cities such as Portland as examples of success stories. So I gathered data on almost every rail line in the country to see if any of them could be judged successful by any objective criteria. As detailed in a recent Cato Institute report:

  • None of them collect enough fares to cover their operating costs, much less any of their capital costs. New York City subway fares cover just 67 percent of operating costs, more than any other rail system.
  • Almost all of them cost far more than it would have cost to provide equivalent service using buses. The only exceptions were a few commuter-rail systems in dense urban areas including Los Angeles, San Francisco, and Washington.
  • Many cities spent so much money on a few rail lines that they were forced to cut funding to buses, contributing to a net loss in transit riders. Atlanta and Los Angeles, for example, practiced “transit apartheid,” building rail lines into middle-class neighborhoods while cutting bus service to low-income, minority neighborhoods.
  • Even where ridership has increased, it has rarely kept up with auto travel and nowhere has it reversed the trend towards suburbanization of jobs. Transit in Portland, for example, carried 9.8 percent of commuters to work before the city began building light rail; in 2007, with four light-rail lines and a streetcar line, transit carried only 6.5 percent of commuters to work.

Not only is the initial cost of rail transit high, most rail lines cost more to operate than buses and all of them cost far more to maintain. After 30 years, rail infrastructure begins to break down and must be replaced. The Washington DC rail crash that killed 9 people last June took place because that rail system is more than 30 years old and no one has provided any funds to rehabilitate it.

On top of the sheer cost, rail transit usually results in a loss of property rights for many of a region’s landowners. Bus routes can be easily altered in response to changing travel patterns, but rail lines cannot. So rail transit agencies often try to become land-use czars, dictating where development can take place so that new residents and jobs are located near rail stations. They try to restrict growth away from the rail lines while mandating higher-than-marketable densities in rail corridors. Cities often use tax-increment financing–effectively taking money from schools and other urban services–to promote such “transit-oriented developments.”

Residents of regions that open new rail lines can expect to pay higher taxes to fund transit and even more taxes or reduced urban services to fund transit-oriented developments. They can also expect to suffer more traffic congestion as funds that might have been spent relieving congestion are spent on rail instead. Housing costs will increase as restrictions are placed on the kind of housing people want–single-family homes with a yard–while jobs may be lost as businesses move to a less heavily taxed and regulated region.

In all, building rail transit in a city becomes a cancer to normal, sustainable activity, as that rail line becomes a growing burden on legitimate transportation and even the full range of city services itself.


  1. Mark Bulkeley  

    I would agree that the economic arguments against rail are reasonably strong, though I have been a happy user of commuter rail when its been available to me.

    I would have found the article much more compelling if you had (a) discussed the arguments that are typically used for “light” rail (b) differentiated commuting to long-distance freight and (c) discussed the economic failure of not implementing time of use pricing for roads. Seems that (a) could have been easily done (and perhaps you did it implicitly) though (b) is potentially a whole different article.


  2. Ursus  

    Well-written and informative except for the last sentence, which was silly, unnecessary and rather ignorant of virology and medical sciences.


  3. Tránsito ferroviario urbano: en el camino equivocado  

    […] En 2006, Nashville comenzó a funcionar la Música Ciudad de las Estrellas, un tren de pasajeros entre el Líbano y Nashville. Los funcionarios de tránsito presumir que se trata de "la eficaz en tren de cercanías de coste más en el país", ya que sólo gastó 41 millones dólares para iniciar el servicio. Para ser rentable, sin embargo, necesita algo más que un tren: el tren tiene que producir algo . [. . . ] URL del artículo original http://www.masterresource.org/2010/04/urban-rail-transit-not-on-the-right-track/ […]


  4. NormD  


    It would seem like subways are special case as I doubt you could run buses underground. Is there some way to do this?

    Lots of commuter train right-of-ways take space away from roads and impede road travel. Are these inefficiency costs included in your calculations?

    Funds used to build rail are taken from road maintenance and improvement. Are these inefficiency costs included in your calculations?


  5. antiplanner  


    One of my calculations compares rail with buses providing equivalent service. I estimate the number of buses needed to provide such service. Except in Manhattan and possibly Chicago and Boston, there is enough street space to accommodate enough buses to replace the subways. This is particularly true in Washington, DC, which once had exclusive bus lanes in many corridors that were eliminated when the subway was built.

    You are right that both commuter trains and light rail impede road travel. I don’t specifically account for this in every case, but we know from selected cases that the effect is significant. After the Minneapolis light rail opened, auto drivers found it added 20 to 40 minutes to their journeys along the same route.



Leave a Reply