Christopher Arkills, a key aide to King County Councilman Dwight Pelz wrote a rejoinder to my post on rail. His critique and my full response are here.

But here’s the crux:

Christopher: You build mass transit, not to get folks out of their cars today, but to influence land use patterns over the next 20, 30, 50 years. . . . You tout BRT [bus-rapid transit], HOV, vanpools, and carpools as cost effective solutions. And they are valuable at getting cars off the roads now. But they do NOTHING to stimulate transit dependent development that will allow people to live in dense livable neighborhoods without using a car. Developers simply wont invest in BRT related projects because they lack the permanence of rail.

Were these points true, it would shift the costs and benefits substantially in favor of rail. And I used to believe them to be true. But over the past few years, I’ve been reading empirical analyses of the question. (The best single source, by a pair of avowed advocates for rail financed by the BART in San Francisco, is Transit Villages in the 21st Century by Michael Bernick and Robert Cervero.) And the balance of evidence is that rail does not automatically yield dense, walkable neighborhoods, at least in the North America.

  • During the last half century, the history of rail transit projects on this continent has produced at least as many examples where new rail transit stations or lines made no appreciable impact on residential or job density as there are examples where it did. In the cases where it did apparently spark dense development, there was usually a large additional investment of political capital and public funds in fostering that development. There’s nothing wrong with doing that, but it raises the question whether the rail deserves the credit or the extra effort does.

    There are also many examples of dense, walkable neighborhoods emerging without benefit of rail infrastructure. Most of the urban renaissance in Vancouver, BC, for example, has little to do with SkyTrain. And contrary to your contention, there are many North American examples of dense, walkable, mixed-use neighborhoods developing around bus-transit nodes. Developers in Toronto and Ottawa, for example, are quite content to invest at bus transit nodes.

    The one all-but-universal land-use “good” that comes from urban rail transit investments is strengthened downtowns. That’s an important benefit to factor into public investment decisions. But there are lots of ways to spend (or not spend) tax dollars, several of which strengthen downtowns and encourage walkable neighborhoods.

    When the costs and benefits favor rail, we should buy it; when they don’t, we should not.