This post originally appeared on Orphan Road.
Richard Morrill has an article in Crosscut on density:
Studies of the costs of infrastructure and public services show only slight variation with density, with moderately higher costs at very low densities (under 1,500 people per square mile) and at very high densities (over 100,000 per square mile). Lower utilization drives up costs at the low density end, while high costs of construction and maintenance affect highly dense areas. The most effective densities are in the middle range, 5,000 to 15,000 people per square mile, which happens to be where probably more than 90 percent of urban dwellers live.
Compare this with our previous post on density, in which an urban planning firm argued that 50 residents/acre is optimal, at least with respect to per capita energy use. 50 residents/acre translates to about 32,000/sq. mile, which is double the high-end of Morrill’s numbers. Per capita energy use is not the same as “costs of infrastructure and public services,” so it’s not surprising that the numbers should be different.
The thrust of the article is about housing prices, which Morrill says may be inflated due to urban growth restrictions. His proposed alternatives are unclear, though. He wants to get rid of urban growth boundaries in favor of higher density subdivisions, but doesn’t provide any links to the studies, so I can’t really comment one way or another.