But by 1930, nearly every American family owned an automobile. This competition combined with the Depression to kill off most of the streetcar lines. By 1960, they were found in only a few cities such as Boston, Philadelphia, and San Francisco. It is popularly believed that a conspiracy of auto and oil companies killed the streetcars. In fact, buses have several major advantages over streetcars.
Yet the experience of new light rail lines is that they cost far more and carry far fewer riders than anticipated. According to a report written by Don Pickrell, an analyst for the U.S. Department of Transportation, the projections made when cities decided to build new rail lines in the 1980s almost always overestimated ridership and underestimated both construction and operating costs.
Planners projected that the rail lines would cost about $1 to $2 per rider, including both operating and amortized capital costs. But with higher costs and lower ridership, actual costs came in at about $5 to $15 per rider.
Pickrell emphasizes that "the systematic tendency to overestimate ridership and underestimate capital and operating costs produces a bias toward the choice of capital-intensive transit improvements such as rail lines." City officials might encourage such a bias because rail projects allow them to get more federal money--much of it going to local construction workers--than bus projects.
Light rail doesn't work in most American cities because population densities are too low and most people want to travel in a lot of different directions, not just the direction the tracks go. "I want to live in one place, work in a second place, and play in a third," says Joel Garreau. "If that's what you want, the very least you need technologically is a form of individual transportation that goes where you want to go when you want to go. Trains just don't cut it. Trains require you to go where someone else wants you to go when someone else wants you to go."
Many planners aren't worried about low ridership. They view light rail as a way to attract high-density development. People like to live or work near a light-rail line, even if they rarely use it, because it gives them more options.
The trouble is that using light rail in this way to reduce congestion is self defeating. Suppose you double the population density in a light-rail corridor and that you also double the share of trips using transit from 5 percent to 10 percent. Then you still have an 80 percent increase in automobile traffic, which means you either have a lot of congestion or you have to spend a lot on widening the roads in the corridor. Roads are less congested--and less expensive--if you just keep densities low.
Despite all the hoopla, light rail is ultimately nineteenth century technology that won't solve twenty-first century urban problems. As Garreau says,
"Plan A" for the future of America is to pave the planet to accommodate cars. Everybody knows that's stupid. But the only alternative usually offered, "Plan B," is to return to nineteenth century rail. This involves forcing people to give up their individualism, and to live in apartments that are convenient to a form of mass transportation that requires thousands of people to want to go from the same point A to the same point B at the same time, like in Manhattan. This is in defiance of almost a century's worth of practice that shows that if Americans thought Manhattan was such a wonderful idea, they would have built more places like it, and they have not.