Causes of Sprawl - A Portrait from Space

Abstract: We study the extent to which US urban development is
sprawling and what determines differences in sprawl across space. Using
remote sensing data to track the evolution of land use on a grid
of 8.7 billion 30×30 meter cells, we measure sprawl as the amount of
undeveloped land surrounding an average urban dwelling. The extent
of sprawl remained roughly unchanged between 1976 and 1992, although
it varied dramatically across metropolitan areas. Ground water
availability, temperate climate, rugged terrain, decentralized employment,
early public transport infrastructure, uncertainty about metropolitan
growth, and unincorporated land in the urban fringe all increase
sprawl.

Key words: urban sprawl, land development, remote sensing
jel classification: r14, o51

Conclusions
As with many economic and social processes, a true understanding of the implications of urban
sprawl can only come about through the study of both the positive and normative aspects of
the urban development process. Much of the current debate has seen people rushing to address
normative issues without first having a good understanding of the positive aspects. In contrast, in
providing the first detailed description of the process of urban development and its determinants,
our paper is quite clearly focused on improving our understanding of these positive aspects.
To summarize, 1.9 percent of the land area of the United States was developed by 1992. Two thirds
of this developed land was already in urban use around 1976, while the remaining one third
was developed subsequently. Our main findings are concerned with whether development
is sprawling or compact. We measure sprawl as the amount of undeveloped land surrounding
an average urban dwelling. By this measure, commercial development has become somewhat
more sprawling during the study period but the extent of residential sprawl has remained roughly
unchanged between 1976 and 1992. In contrast to this stability over time, the extent of sprawl does
vary dramatically across metropolitan areas.

We study the factors that determine these large differences across metropolitan areas. We find
that sprawl is positively associated with the degree to which employment is dispersed; the reliance
of a city on the automobile over public transport; fast population growth; the value of holding on
to undeveloped plots of land; the ease of drilling a well; rugged terrains and no high mountains;
temperate climate; the percentage of land in the urban fringe not subject to municipal planning
regulations; and low impact of public service financing on local taxpayers.

We are some way from being able to make firm policy recommendations, but our results do
raise some interesting questions for policy in this area. Perhaps the most intriguing issue arises
from the connection between aquifers and sprawl. Often the same aquifer will supply water
both to municipal water systems and to individual private wells. Private incentives may push for
scattered development over the aquifer, where one can sink a well and avoid connection fees to the
municipal supply. However, such development may be costly for others, since concrete, asphalt,
and other nonpermeable materials hinder the replenishment of the aquifer with rain water. In
such a context, raising impact fees may only worsen the problem. This raises the intriguing
possibility that groundwater regulation may provide an important avenue through which policy
makers can influence the form of urban development. Another interesting policy implication arises
from the fact that disparities between municipal and county regulation are important causes of
sprawl. Focus, so far, has been on the fragmented nature of local government, but our results
suggest that harmonization of county and municipal land use regulation may actually play a much
more important role in influencing the form of urban development. Interestingly, while we find
that sprawl is affected by two factors which have received little attention, another (the density of
roads) that has received much more attention seems to have little impact. While more car friendly
cities do experience more sprawl, we find that what really matters is not the density of the road
network on the urban fringe but instead whether the city center was shaped before the advent of
the car. Finally, our results on the transfer share in local revenues suggest that internalizing the
fiscal externalities of new development appears to limit urban sprawl.

Of course, these comments are fairly speculative given the current state of our knowledge.
Further analysis of economic models of development, and of models which incorporate a taste
for landscape features is warranted, and such analysis should form the basis for future policy
recommendations.

from an article by Burchfield†, Overman‡, Puga§ and Turner¶ (2005)

†Neptis Foundation, 50 Park Road, Toronto, Ontario m4w 2n5, Canada (email: This e-mail address is being protected from spambots, you need JavaScript enabled to view it ).

‡Department of Geography and Environment, London School of Economics, Houghton Street, London wc2a 2ae,
United Kingdom (email: This e-mail address is being protected from spambots, you need JavaScript enabled to view it ; website: http://cep.lse.ac.uk/~overman). Also affiliated with
the Centre for Economic Performance at the London School of Economics.

§Department of Economics and crei, Universitat Pompeu Fabra, Ramon Trias Fargas 25–27, 08005 Barcelona, Spain
(email: This e-mail address is being protected from spambots, you need JavaScript enabled to view it ; website: http://diegopuga.org).

¶Department of Economics, University of Toronto, 150 Saint George Street, Toronto, Ontario m5s 3g7, Canada

(email: This e-mail address is being protected from spambots, you need JavaScript enabled to view it ; website: http://www.economics.utoronto.ca/mturner/)