Sunday, March 27, 2011

Taking care of itself

We now have 2010 population tallies for all the major U.S. cities. Table 2 in this post lists Wendell Cox's itemization of major U.S. Core Cities. His latest report on head counts shows that most of these places continue to fall behind their suburbs in terms of population growth.

This is old news to some, but perhaps surprising to many who look for (and proclaim) an "urban revival." More than proclaiming, many have worked to put in place a huge number of plans and programs (including public transit) to bring it about.

In his splendid Triumph of the City, Ed Glaeser writes, "Transportation technologies shape our communities and modern sprawl is the child of the automobile." (p. 167)

Here is an even stronger statement from Witold Rybczynski:
“Virtually every technological innovation of the last fifty years has facilitated, if not actually encouraged urban dispersal.” (p. 170)

Yes, agglomeration economies are crucial, but they are available in many places. And suburbs and core cities complement each other. We can network nicely with people who are not down just the street. This is old stuff that bears repeating because no one has yet offered to give me one dollar for every evokation of "regeneration," "renaissance" or "revival" of "the city". Turning back clocks is never simple.

I did not live in New York in the years it was coming apart, but reading Pete Hamill's Piecework gives me an idea how awful it was. New York city's return from those days is amazing. Fred Siegel covers the fall and rise.

But what about the "costs of sprawl?" In my March 12 post, I mentioned that the latest data on commuting indicate that suburban journey-to-work travel times compare favorably with core city times. But those cites referred to all modes of travel and have an apples-and-oranges problem because transit use varies -- and accounts for much slower travel.

The NHTS data make it possible to compare travel via privately-operated-vehicles only, probably the most direct and representative.

In 2001, the average for the "suburbs" was slightly less than for "urban" (NHTS definitions), 20.8 minutes vs. 22 minutes. They had pulled approximately even by 2009, 22.3 minutes for "urban" and 22.8 minutes for "suburbs." The lowest travel time readings in both years were for "second city" (akin to an Edge City), 17.3 minutes in 2001 and 20.9 minutes in 2009.

These are averages, but standard deviations were small in each case; coefficients of variation were always below 1.0.

There are also data for miles instead of minutes. What do they show? Coefficients of variation always greater than one.

Commuters care most about time costs and manage to find job and home location matches that preclude crazy commutes. Planners worry over "jobs-housing" balance (and what they can do about it), but it seems to be taking care of itself.


Ed Glaeser explains his surprise at the New York numbers.


Here is more re New York from Wendell Cox.