|►◄ Reverse Zone|
Complete List of Posts
Wed, 11 Oct 2006
An interesting article in the May 2006 Quarterly Journal of Economics, regarding where sprawl happened in the US between 1976 and 1992, and what these places had in common.
This paper defines sprawl a little differently than many others: sprawl is a building with undeveloped land around it. Their index uses the amount of undeveloped land in the square kilometre surrounding a building. Their definition of built-up land is also different than most, so they come up with less built-up land than others, about 2% of the U.S. in 1992 as opposed the U.S. government's 3%. They split up the U.S. into 30x30 metre squares, and say it is built up only if there is a building or pavement in it, whereas the U.S. government calls it built-up if the gap between buildings is less than 150 metres. That means if your building is 75 metres from the lot line and you decide to subdivide and build something new, the government calls it infill and this paper calls it greenfield. It also refers to parks as undeveloped urban land, and to highways and parking lots as compact development. Still, if you are going to process that much data, you need simplistic definitions.
Whatever the definitions, what is interesting is what their sprawl index correlates with. One interesting conclusion is that the average "compactness" of development has not changed. As new outer suburbs are developed, the previous generation of sprawling suburbs is intensified. The sorts of places where people live has not changed, but where they work and shop has. Commercial development used to be split between the two extremes of density. It has shifted dramatically to low density. New jobs are far from downtown, bad news for transportation. In fact, the more sprawling employment is, the more sprawling residential development becomes.
The authors tested various theories of the causes of sprawl to see how they correlate with their sprawl index. The most important factors promoting sprawl are dispersed employment, fast growth, the appreciation of undeveloped plots of land, the presence of aquifers allowing individuals to have wells, rugged terrains but no high mountains, temperate climate, land not subject to municipal planning regulations, and the transfer of servicing costs to other taxpayers. The density of roads does not seem to have an effect. The fragmentation of local government is not as important as the existence of unincorporated land with few rules or taxes.
Factors that inhibit sprawl include a mature compact urban core, cities with a transit-oriented history, physical barriers such as mountains.
The most sprawling city according to their index: Pittburgh (up) overaking Atlanta (down). The least sprawling, Miami. Most improved: Phoenix. Portland and Seattle show no improvement but it would be interesting to see what happened after the urban boundaries were imposed.