Thursday, April 8, 2010

Kotkin Makes Sh!t Up or He Cites Those That Do

Statistics. In the wrong hands they can be dangerous. Fortunately, we have good people like Rob Steuteville of the New Urban News to debunk a statistic that Kotkin uses in his paperweight made of stacked and cut printed paper and then more recently picked up and ran with directly into a wall by David Brooks:
David Brooks’s column in The New York Times April 5, “Relax, We’ll Be Fine,” is a hopeful and thoughtful review of Joel Kotkin’s book, The Next Hundred Million. That is what I thought until I came to the following statement, which Brooks apparently picked up from Kotkin: “For every 10 percent reduction in population density, the odds that people will join a local club rise by 15 percent.”
Steuteville then digs in:
If you read the 2006 study (at the source of the statistic), “Social Interaction and Urban Sprawl,” and look at the numbers, you find that going from Boston to a typical suburb does not net a statistical difference in social interaction of any kind.

Brueckner didn’t find any actual decrease in social interaction or activity related to lower-density locations. What difference he did find, however, was favorable to cities. Yes, the typical person in a city was found to have slightly more social interaction than residents in low-density areas.

But Brueckner also assumed bias in the population. He stipulated, with no apparent research to back it up, and no capacity to observe this bias, that people in cities are of a type that is more prone to social interaction. He put in an equation to counteract this supposed bias, which he called an “unobservable propensity.” In other words, he placed a theoretical thumb on the scales.

Using that method, of course, I could prove anything — that reduction in smoking could reduce your chances of getting cancer, for example.
He then counters with a more recent and at least seemingly more objective study:
Podobnik compared a new community that has a new urbanist design — Orenco Station — to both a conventional suburb and two older city neighborhoods in a longitudinal study. Orenco Station is considerably denser than the conventional suburb, and more carefully designed to encourage walking and facilitate social interaction. People in Orenco Station report that they participate in group activities at nearly double the rate of all of the comparison neighborhoods.

More importantly, perhaps, the Orenco Station residents reported participating in higher-quality activities. In the conventional suburb and the city neighborhoods, the activities were mostly neighborhood watch and homeowners association meetings. In Orenco Station, residents cited book clubs, group dinners, and other informal neighborhood activities. To me, food and literature sound more fulfilling than looking for crime and property violations.

Other studies have come to similar conclusions, including 2009 research that compared new urban developments in Canada to conventional suburban communities. Just like Orenco Station, these new urban Canadian neighborhoods have higher density and are designed with a mix of uses, pedestrian-friendly streets, and appealing public gathering spaces. Surveys in other new urban communities such as Kentlands in Gaithersburg, Maryland, and Celebration in Kissimmee, Florida, have also indicated strong social ties.

None of these studies have shown precisely why new urban places have relatively strong connections among residents — only that this pattern exists. The simplest explanation is that the new urban design, intended to foster social ties, is working. It could also be that people who want to have more social ties choose to live in places that are designed with that goal in mind. Both factors are probably at work.

Lesson: don't believe everything you read. Even this: Joel Kotkin was not born on this Earth and was actually a direct descendant of Xenu sent here to perpetuate the malignant terra firma melanoma of sprawl and destroy the world.