A pluralistic approach to defining and measuring urban sprawl - Urban Remote Sensing: Monitoring, Synthesis and Modeling in the Urban Environment - page 160

United States, homes and neighborhoods began to appear in the
countryside, soon to be connected with railroads and streetcars
that would catalyze additional demographic movement from city
to suburb (Gillham, 2002).
The process was greatly expedited in the post World War
II mid 20th century, a period that was characterized by spatial
diffusion of residents and activities to the outskirts of urban
centers (Mills and Hamilton, 1994). In the United States, and
to a lesser extent in Europe, the process of suburbanization
(commonly associated with sprawl, but see below) started in
earnest following World War II, with a combination of high
population growth and an inability of city centers to absorb this
growth. 3 A rapidly growing post-war economy, improvements
in technology and a rise in standard of living all contributed
to increasing demand for large-lot, single-family homes on the
outskirts of cities. Concurrently, city centers were in decline
(Batty, Xie and Zhanly, 1999, Golledge and Stimson, 1997). The
rise of the automobile as a predominant form of transportation
facilitated and expanded this process (Glaeser and Kahn, 2004)
and set in motion a positive feedback mechanism: the more car-
dependent society became, the more suburbs held appeal; the
more suburbs proliferated, the more dependent society became
on the automobile.
The actual use of the term ''sprawl'' began in the United
States in the 1950s, and became widely used from the 1960s
(Belser, 1960, Harvey and Clark, 1965, Gans, 1967, Real Estate
Research Corporation, 1974). 4 From the 1970s, the term ''sprawl''
was often accompanied by ''suburbanization,'' although the two
are conceptually unique from one another. Suburbanization
refers to the migration of urban residents to the peripheries or
outside of cities in a metropolitan area in order to establish
new residential neighborhoods (Angotti, 1993). Fishman (1987)
differentiates between English/American suburbanization and
that of continental Europe in that the former was characterized
by middle and upper class residents leaving the cities for green,
low-density homes in the urban periphery, while the latter was
led by industry leaving the cities, followed by the working class.
Sprawl is a broader concept as defined in our introduction
that includes social, demographic and economic characteristics
and a broader diversity of urban spatial development character-
istics of which suburbanization is just one. Other characteristic
development forms include edge cities and exurban development
and also included is the demographic and socioeconomic decline
of urban centers.
While sprawl may be considered a global phenomenon, the
history of sprawl seems largely to have been written in the United
States and to a lesser degree in Europe. As early as the 1920s,
planners in the United States began noting an acceleration of
the rate of loss of open and agricultural land in favor of devel-
opment (Burchell et al ., 1998). The rise of zoning regulations,
which provided the legal foundation for separating land uses, is
considered a major contributor to later sprawl patterns (Gillham,
2002). Later, in the United States, the strong belief in individ-
ual property rights and free markets, along with a distrust of
strong, central government is posited to have had a significant
impact on the shape of sprawling land development patterns. As

Gordon and Richardson suggest (2000), the history of American
movement from cities to suburbs might reasonably be viewed as
people realizing their residential preferences.
In the United States, sprawl, as defined by loss of farms and
open space, was noted by planners as early as 1929 in New York
(Burchell et al ., 1998). Sprawl critics point to Federal zoning
policies from 1922 onward, that gave rise to segregated land use,
whichinturnlaidthefoundationforanautomobile-centered
transportation network (Burchell et al ., 1998). In the 1950s and
1960s in the United States, sprawl terminology began entering
the planning literature, once again emphasizing low density
development and the predominance of automobiles. Leapfrog
development, complemented by the rise of federal highway
system, fed the critique of spatial growth patterns. By 1972,
McKee and Smith (cited in Burchell et al ., 1998) would distill
the definition of sprawl into four forms: (1) very low density
development; (2) ribbon-variety development extending along
access routes; (3) leapfrog development; and (4) a ''haphazard
intermingling of developed and vacant land.''
In 1991, Garreau introduced the concept of ''edge cities''
(Garreau, 1991) as the evolution of non-residential urban cluster
development along junctions of beltways and interstate roads.
Edge cities introduced a new dimension to sprawl, in that it was
not low-density residential development around a single urban
core, but entirely new urban cores developing as satellites to main
cities. Unlike suburbs, edge cities serve all the functions of the
urban core with an emphasis on employment centers. The Euro-
pean analogy to edge cities have been called Functional Urban
Areas (van den Berg et al ., 1982), and in this case, they are con-
sidered a collection of urban communities that together include
residential, employment and recreational centers, developed on
former agricultural land, and within functional proximity of a
major urban center.
The development of edge cities added a new dimension to
thinking about sprawl - the dimension of spatial scale. Now
rather than envisioning only the urban core and sprawled devel-
opment in connection to it, a broader scale of analysis was
needed to consider regional development patterns. Whereas sub-
urbs emphasized sprawl at a municipal level with a particular
emphasis on a decline in building density, at the regional scale,
terms such as satellite towns, edge cities, exurbs, and megalopolis
become relevant to describe spatial broader phenomenon for
which density is only one of many relevant spatial characteris-
tics. Consequently, while density remains the most intuitive and
popular spatial variable for measuring sprawl, the list of variables
becomes longer when considering the multi-scalar dimensions
of sprawl.

12.4 Qualitative dimensions
of sprawl and quantitative
variables for measuring
them

3 Although Jackson (1985) suggests that suburbanization, which is defined as a
situation when peripheral areas develop at a faster pace than central urban areas,
appears as early as 1815 in the US and Britain.
4 For a compilation of early references to sprawl from the early to mid-20th century,
see Hess and colleagues (2001).

In this section, we present quantifiable variables that have been
suggested in the literature and employed empirically to measure
sprawl. We first present several criteria - our own and drawn
from the literature - that a variable measuring sprawl should