Characteristics, Causes, and Effects of Sprawl:

A Literature Review

Reid H. Ewing

Keywords: costs · public services · urban sprawl · pattern of development · Florida · open space ·
subsidies · policy · farmland

Editor’s Note: This literature review was prepared for the Florida Department of Community
Affairs (DCA) as background to the amendment of Florida’s local comprehensive planning rule,
Rule 9J-5. On October 2, 1992, DCA published a proposed amendment that, among other things,
included criteria for the review of plans to ensure that they discourage urban sprawl. The proposed
rule was challenged by the Florida Association of Realtors, Florida Home Builders Association,
Florida Farm Bureau Federation, and other parties. DCA settled with most parties, but not all, and
the matter went to administrative hearing on September 13, 1993. The hearing officer’s order is to
be issued momentarily, and if favorable to DCA, will allow DCA to adopt a final rule by early 1994.

The physical characteristics, causes, and effects of sprawl must be understood before sprawl
can be effectively regulated. Relying on the literature in the field, this paper provides a conceptual
framework against which DCA’s proposed sprawl rule can be judged and upon which the final rule
can rest.

Classic Sprawl Patterns

Sprawl has been equated to the natural expansion of metropolitan areas as population grows
(Sinclair, 1967; Brueckner and Fansler, 1983; Lowry, 1988) and to “haphazard” or unplanned
growth, whatever form it may take (Peiser, 1984; Koenig, 1989). More often, though, sprawl is
defined in terms of “undesirable” land-use patterns—whether scattered development, leapfrog devel-
opment (a type of scattered development that assumes a monocentric city), strip or ribbon develop-
ment, or continuous low-density development. Table 1 indicates which patterns have been labeled
sprawl in the technical literature of the past three decades. Scattered development is probably the
most common archetype, but any “non-compact” development pattern qualifies.
Using archetypes to define sprawl still leaves us short of a working definition. Like obscenity, the
experts may know sprawl when they see it, but that is not good enough for rulemaking. There are
two problems with the archetypes.

R.H. Ewing
National Center for Smart Growth, Preinkert Field House, University of Maryland, College Park, MD 20742 USA
e-mail: [email protected]

Originally Published in 1994 in Environmental and Urban Studies vol 21(2):1–15 519
J.M. Marzluff et al., Urban Ecology,

C Springer 2008
520 R.H. Ewing

Table 1 Sprawl Archetypes

Low-Density Strip Scattered Leapfrog
Development Development Development Development
Whyte (1957) x x
Clawson (1962) x
Lessinger (1962) x x
Boyce (1963) x x
Harvey and Clark (1965) x x x
Bahl (1968) x
McKee and Smith (1972) x x x x
Archer (1973) x x
Real Estate Research x x
Corporation (1974)
Ottensmann (1977) x
Popenoe (1979) x x x
Mills (1981) x x
Gordon and Wong (1985) x
Fischel (1991) x
Heikkila and Peiser (1992) x

First, sprawl is a matter of degree. The line between scattered development and so-called poly-
centric or multinucleated development is a fine one. “At what number of centers polycentrism ceases
and sprawl begins is not clear” (Gordon and Wong, 1985, p. 662). Scattered development is classic
sprawl; it is inefficient from the standpoints of infrastructure and public service provision, personal
travel requirements, and the like. Polycentric development, on the other hand, is more efficient
than even compact, centralized development when metropolitan areas grow beyond a certain size
threshold (Haines, 1986). A polycentric development pattern permits clustering of land uses to
reduce trip lengths without producing the degree of congestion extant in a compact, centralized
pattern (Gordon et al., 1989).
Likewise, the line between leapfrog development and economically efficient “discontinuous
development” is not always clear. Leapfrogging occurs naturally due to variations in terrain (Harvey
and Clark, 1965). New communities nearly always start up just beyond the urban fringe, where large
tracts of land are available at moderate cost (Ewing, 1991). Some sites are necessarily bypassed in
the course of development, awaiting commercial or higher-density residential uses that will become
viable after the surrounding area matures (Ohls and Pines, 1975). “The sprawl of the 1950s is fre-
quently the greatly admired compact urban area of the early 1960s. An important question on sprawl
may be, ‘How long is required for compaction?’ as opposed to whether or not compaction occurs at
all” (Harvey and Clark, 1965, p. 6). Whether leapfrog development is inefficient will depend on how
much land is bypassed, how long it is withheld, how it is ultimately used, and the nature of leapfrog
development (Breslaw, 1990).
Nuances arise with other sprawl archetypes as well. The difference between strip development
and other linear patterns (e.g., Mainstreet USA or transit corridor development) is a matter of degree.
So, too, the difference between low-density urban development, exurban development and rural resi-
dential development. Wherever one draws the line between sprawl and related forms of development
will be subject to challenge unless based on an analysis of impacts. It is the impacts of development
that render development patterns undesirable, not the patterns themselves.
The second problem with the archetypes is that sprawl has multiple dimensions, which are
glossed over in the simple constructs. It is sometimes said that growth management has three
dimensions—density, land use, and time. The same is true of sprawl. Leapfrog development is a
problem only in the time dimension; in terms of ultimate density and land use, leapfrog development
may be relatively efficient. It is known, for example, that infill parcels tend to be developed at higher