Housing, jobs, and other destinations are so diffused throughout American urban areas that they don’t generate the large numbers of people moving from one point to another that mass transit needs to work.
“Transit worked when American cities were denser,” is the mantra of today’s urban planners. “If we can increase their densities, transit will work again.” Reality is a lot more complicated, and that reality explains why transit can’t work in American urban areas even if their densities increase.
From about 1880 to 1913, transit and cities co-evolved thanks to new technologies that benefited both. The same steam engines that powered commuter and early rapid transit trains also powered downtown factories. The same Bessemer steel that made the rails that streetcars and urban trains rolled upon also provided the structural beams that allowed construction of skyscrapers. The same electric motors that moved electric streetcars also powered electric elevators that gave people quick access to the upper floors of those skyscrapers.
These technologies created monocentric cities by concentrating jobs in urban centers surrounded by residential areas that fed into the centers on transit. Never before in history had cities been like this, yet people today still imagine that cities ought to be monocentric, a myth that drives too much bad policy.
This was transit’s Golden Age, but it was far from perfect. Transit was too expensive for unskilled workers, so they had to live in high-density tenements located within walking distance of downtown factories. To make a profit, rapid transit and streetcar operators used just enough vehicles to carry people but not enough to give them breathing room, at least at rush hour. Many transit lines had been built from the profits of the real estate developments they accessed, and while fares covered operating costs they were insufficient to rehabilitate these lines as they wore out.
Urban and transit evolution parted ways in 1913, when Henry Ford built the first moving assembly line to make his Model Ts. Cheap cars were an obvious threat to transit, but a bigger threat was less visible: unlike steam-powered, belt-driven factories, moving assembly lines required lots of land, so factories moved to the suburbs. When the suburbs refused to be annexed to the cities, monocentric cities became polycentric urban areas.
At least through the 1970s, urban planners and central city officials pretended their cities were still monocentric, and they wrote numerous downtown plans, urban renewal plans, transit plans, commuter-tax plans, and other plans designed to maintain the preeminence of downtown. The construction of the San Francisco BART and Washington Metro systems were among these plans, but were as doomed to fail as all the others.
As both jobs and people left city centers after World War II, most major central cities began to lose population even as their suburbs grew. Since 1950, Buffalo, Cleveland, Detroit, Pittsburgh, and St. Louis have all lost more than half their populations. Cincinnati lost 41 percent; Baltimore 35 percent; Boston and Minneapolis 30 percent; Washington 29 percent; Chicago 25 percent; St. Paul 13 percent; San Francisco and Oakland, 12 percent. Except in New York, one of the few major central cities that had more people in 2000 than 1950, this decentralization greatly reduced transit’s effectiveness.
By the 1980s, planners began to realize that urban areas had become polycentric, and today polycentricity is a fundamental part of the New Urbanism. Too late: cities had changed again with the decline of manufacturing jobs and the growth of service jobs. In 1920, nearly 40 percent of all American jobs were manufacturing, and there was one-and-a-third service jobs per manufacturing job. Today, less than 10 percent of jobs are manufacturing, and there are ten service jobs for every manufacturing job.
Even if they weren’t in city centers, manufacturing jobs were at least concentrated. But service jobs in such fields as health care, education, wholesale and retail trade, and utilities, were diffused throughout urban areas. As noted in Reason #2, less than 30 percent of urban jobs today are located in downtowns or regional or town centers. Other things once concentrated in downtowns, such as shopping, churches, and theaters, also became diffused.
These trends had the least impact on New York, but even in the New York urban area (which includes suburbs in northern New Jersey, southwest Connecticut, and New York state), as opposed to the city itself, transit is pretty marginal. While transit carries 57 percent of New York City commuters to work, it carries just 14 percent of suburban New York commuters.
The diffusion of jobs and other destinations throughout an urban area returned cities to be more what they were like for thousands of years before the late nineteenth century. Simply increasing population densities, as regional governments in California, Oregon, and Washington have done, doesn’t help because those jobs and other destinations remain too diffuse for transit to work for any but a small minority of the population. These changes are not only irreversible, there is no reason why we should want to reverse them.