The modern metropolis, as critics see it, sends irregular tentacles of low density development snaking through pristine areas, leaving behind large amounts of unused land and under-used infrastructure. To many, the Phoenix metropolitan area epitomizes the problem. Policies to encourage more infill-that is, to locate more development within the urbanized area rather than on its fringes-are advocated as a route to more efficient use of land and existing infrastructure, preservation of open-space, decreased cost of public services, and improved economic and social conditions, as well as to alleviate the general atmosphere of disorder.
If all of the approximately 121 square miles of vacant residential land inside the urban area in 1990 had been developed at the same density as nearby areas, the urban area could have housed almost another half million. If the new housing had been 50 percent denser (measured in population per developed residential acre) the same area could have housed almost another 750,000 persons. Even at the rapid growth rate of the first half of the decade, those figures would be 6 and 9 years of growth respectively, without expanding the urban area beyond its 1990 limits. However, once all vacant land were developed most development would have to occur on the outskirts; sprawl would have been temporarily arrested, then accelerated. Nevertheless, there are a number of reasons why the above scenario was never a practical possibility.
First, cities and utilities confront developers with contradictory price incentives by using a combination of marginal cost pricing in which new users pay for new infrastructure and traditional average cost pricing in which everybody pays. The latter system tends to under-charge for sprawl and thus undercuts a policy of encouraging infill.
Second, city-wide zoning requirements make development on small parcels impractical without variances or some other governmental action. This gives rise to an informal system of negotiations, often involving neighbors, which tends to increase costs, slow infill and reduce density.
Third, even though increasing densities, whether in infill areas or on the fringes, would slow the rate of expansion of the urban area, it appears that most participants in the development process are opposed to or ambivalent about higher density and such a policy seems unlikely to be supported.
Rapid population growth makes expansion of the urban area inevitable regardless of infill rates. Even if all vacant residential land in the 1990 urban area were developed at densities 50 percent greater than nearby neighborhoods, twenty years of population growth at the rapid 1990-95 rate would require developing 232 square miles outside the urban area for housing. Infill can-and very likely does-accommodate some of the population increase in the Phoenix urban area. There are some practical policy options that local governments could use that might increase the rate or density of infill, and thus slow the expansion at the fringes, but unless growth rates decrease significantly some expansion is inevitable.