A few weeks ago, Randall O'Toole (a leading anti-anti-sprawl commentator) and Matthew Yglesias (a Washington-based pundit who primarily writes about politics, but occasionally veers off into planning issues) had an interesting discussion about the extent to which sprawl is a result of land use regulation.(1)
O'Toole argued that zoning does not play a major role in creating sprawl because historically, zoning "was used almost exclusively in areas that were already developed. Those original zones merely reaffirmed the development that was already there. Single-family neighborhoods were zoned for single-family homes; apartments for multi-family; industrial for industry; and so forth." By contrast, zoning in undeveloped areas is more flexible; O'Toole writes that if a developer asks a city or county to rezone for more dense development, the government usually complies.
In other words: if you want to build in the middle of nowhere, you can build what you like- even if you want to build something other than conventional sprawl.
But O'Toole's point is perfectly consistent with the possibility that if you want to build anywhere near any existing neighborhood, you risk running into a brick wall of zoning regulation designed to limit density and cater to "not in my backyard" (NIMBY) hostility to new development.
This difference between developed and undeveloped areas restricts compact development more than it restricts sprawl. Here's why: if you want to build a walkable neighborhood, you're probably going to want to build in a desirable intown or inner-suburban neighborhood, close to public transit. After all, people who value walking to the nearest store are probably more likely to value proximity to transit than people who are just as happy to drive everywhere. So the dominant zoning system means that a landowner can build compact development- but not always where such development is most desirable, i.e. in areas near public transit (which tend to be older, established, heavily-zoned neighborhoods).
The restrictiveness of zoning laws in developed areas affects the location of development as well as its form: it increases the likelihood that developers will prefer to build in the least developed areas in order to avoid NIMBY objections and zoning restrictions that cater to NIMBYism.
O'Toole cites the Maricopa County, Arizona zoning code as an example of developer-friendly zoning. The Maricopa code contains a provision for Planned Area of Development (PAD) districts,(2) which can be more compact than other neighborhoods. But a developer still has to apply to get a parcel rezoned to PAD (3). Where do you think a PAD application is more likely to be approved- in an inner suburb cheek by jowl with existing neighborhoods, or at the edge of the county dozens of miles away? Common sense suggests the latter.
My suspicion is not just a hunch. In 2001, the Urban Land Institute surveyed developers, asking them whether they would build more compactly if government regulation was less restrictive. About 80% of developers responded that they would build more compactly in inner suburbs if government regulation was more permissive, as opposed to less than 40% in rural areas.(4) In other words, developers themselves believe that government regulation limits development in cities and inner suburbs.
In sum, even jurisdictions that are quite permissive towards "greenfield" development may be less permissive towards infill. This bias enourages developers to build in semirural suburbs, and is especially likely to reduce compact development.
(1) O'Toole's points are made at http://www.cato-at-liberty.org/2010/03/18/a-libertarian-view-of-urban-sprawl/ and
http://ti.org/antiplanner/?p=2887 ; For Yglesias' post go to
(3) Id., art. 1001.4
(4) Jonathan Levine, Zoned Out 131(2006). Cities and outer suburbs ranked between those extremes.