The Zoning Straitjacket, Part II

Shane Phillips (at Better Institutions), Dan Keshet (Austin on Your Feet) [update: and Daniel Kay Hertz, at City Notes] both recently authored posts showing how the majority of the area in three large American cities (Seattle, Austin and Chicago) is set aside for single-family detached zoning. Shane estimates the area in Seattle to be around 80%, while Austin, to my eye, looks to be about the same. In general, the zoning maps continue to reflect the land use patterns and planning dogma of the 1920s, with a small, constrained downtown business district hemmed in by single-use residential districts through which snake narrow commercial corridors. This is the same pattern I previously noted in the case of Vancouver, where the zoning map is surprisingly unchanged from the original version drafted in the 1920s (however, Vancouver has only 60% of its zoned area set aside for single-family, and has allowed more density in these areas as well).

Although I've made several arguments against Euclidean zoning before, if a city is to be zoned, I don't think there's anything wrong in and of itself with having detached single-family zones — America's unique contribution to zoning, according to Professor Sonia Hirt — as one of a city's planning tools. The issue seems to be that these zones, and particularly low-density zones, are virtually impervious to change over the decades. Where a city has abundant greenfield land for expansion, this is somewhat less of a pressing concern, but consider a city like Stamford, Connecticut, which exhausted most of its undeveloped territory back in in the 1970s, and is bordered by insular Greenwich and Darien (the latter of which inspired Lisa Prevost's Snob Zones). At left is an image of the zoning map from 1965 (14 years after its initial adoption), and at right is today's map:

Sources: City of Stamford and Ferguson Library Digital Collection.

Although the total number of zones has more than doubled, going from 17 to 39, the major zoning boundaries in the north of town are entirely unchanged. In 1965, as today, approximately 80% of the city was zoned for detached single-family residential. Nonetheless, according to ACS data, only around 40% of the city's housing units consist of single-family detached (SFD) units, which highlights the prevalence of large-lot zoning in the single-family areas (the yellow areas represent 1, 2 and 3-acre minimum lot sizes), and the near-impossibility of adding any further SFD under the current zoning regime.

Population by city proper, from Census data.

For Stamford, this zoning policy, which was less a planning vision than a "paint by numbers" exercise in locking into the place the loose residential patterns built along the non-gridded street networks that were being laid out by land companies as early as the 1930s*, worked to accommodate population growth for a time. While there was still vacant land in the north, the city's population surged from 1950 to 1970, just as other established cities like Bridgeport and New Haven stagnated or lost residents. The exhaustion of this supply of land, paired with an immensely destructive urban renewal project in the early 1970s that virtually obliterated the city's entire downtown (and along with it, much of the city's low-cost attached and multifamily housing stock), resulted in a sudden reversal of fortunes in the 1970s. Most of the gains since 1980 have occurred through infill and densification in the existing downtown and multifamily areas, although this has been a slow process. In the 2000-2010 period, in fact, New Haven grew faster than Stamford for the first time in at least 130 years.

The SFD zones have therefore played a role strongly reminiscent of Smart Growth urban growth boundaries, requiring additional housing units to be built in already-dense, transit-accessible areas, although the establishment of the zones predated Smart Growth ideology by several decades. However, Smart Growth critics like Randal O'Toole and Wendell Cox have never, so far as I know, gone after large-lot SFD zoning for restricting housing supply in geographically-constrained situations like this one [Note: in the comments, NickD points out that Cox has criticized large lot zoning, although often without much specificity]. O'Toole has even gone so far as to condemn upzoning of SFD areas on several occasions on the basis that these zoning changes are out of "character" for the area, and frustrate residents' expectations, observations which inadvertently highlight the growth-restricting role of these areas.

The ideological battle between Smart Growth advocates and the self-appointed defenders of the American dream obscures the real functional purpose of maintaining a reasonably clear break between city and country, which, in the United States at least, could be said to be permitting unencumbered and politically-frictionless urban expansion. The city which, like Stamford, forces new development to spread across large areas at very low density is, in effect, politically closing that area off to urbanization and possibly even modest densification in a way that urban growth boundaries, which can and are periodically adjusted by city and state governments, cannot achieve. Moreover, it makes it uneconomical to produce much additional SFD housing, a result that should be highly distressing to someone like O'Toole. It is an approach which seemingly leaves both sides dissatisfied.

This, if nothing else, seems like a fundamental, if not the only, purpose and challenge of city planning: accommodating population growth in a way that takes into account long-term development prospects and the political difficulty of upzoning low-density SFD areas. In light of this, can a zoning code like Stamford's, with a stated purpose of preserving existing neighborhoods in their 1960s form, and resistant to all but changes in the downtown area, really be called a "planning" document at all? The challenges that Stamford faces are not unique, but typical, and progress on them, as zoning approaches its 100th birthday, remains the exception rather than the rule.

Related link: Vancouver and the Zoning Straitjacket

*Historic aerial maps show that Stamford's urban grid was disintegrating as early as the 1930s, just before its major population boom, and that subsequent rights of way were generally winding and given non-urban descriptors ("Road" rather than "Street"). In other words, it appears that the city ceased taking a prominent role in guiding the framework for urban growth around that time. The zoning code did not cause this change, but it certainly has prevented any future adaptations or retrofits. Chris Bradford noted a similar change in Austin that took place at around the same time.