The Fuse

Urban Sprawl and Its Effects on Fuel Consumption

by Hart Schwartz | April 24, 2018

  • Urban sprawl comes in interrelated formats: Physical sprawl and social sprawl
  • Physical sprawl (to be examined in this article) involves a basic decision of “build up or build out,” and can be seen in comparing population growth rates with pace of land development.
  • Social sprawl (to be examined in a follow-up article) underpins physical sprawl. Shrinking households, due to socio-demographic changes, accelerate physical sprawl by creating a need for more land per person.
  • The greater the preference for “build out,” the greater the chance of higher fuel consumption.

Urban sprawl is a key theme that stimulates discussions of urban design, fuel consumption, and vehicle ownership trends. Many writers rail against the evils of sprawl, proclaiming that it holds significant responsibility for America’s petroleum dependence. But what, in fact, is “sprawl”? How does one measure it? How does one know that it exists, and how much of it there is?

This article will deal with the “what is sprawl” question, attempting to clarify a term which does not always lend itself to clear, precise characterization. There are in fact several indices and classifications, each one addressing the same fundamental issue of “build up or build out.” As population grows (with 25 million new Americans being added per decade)—will the built environment (the complex of all manmade structures) expand upward, in terms of taller buildings closer together, or outward, in terms of lower-density development? The “build up or build out” dilemma plays out in multiple dimensions, in that the physical rate of sprawl is heavily influenced by social, cultural, and demographic trends. This article will affect physical sprawl and the following article social sprawl.

Conceptually, fuel consumption rises when vehicle travel increases. Sprawl has a major impact on travel by extending trip distances in urban landscapes as well as creating a need for additional vehicle trips, by undermining the walkability as well as the densities needed for transit. Therefore, understanding the components of the rate of sprawl can help to better understand fuel consumption trends.

Developed land: Basic indicator of physical sprawl

At a very basic level, an increase in the amount of developed land indicates sprawl may have occurred. The United States Department of Agriculture (USDA) has published a nationwide land survey every five years since 1982 (most recent data 2012). Total area of U.S. developed land has increased from 112,383 square miles (71.9 million acres) in 1982 to 178,301 square miles (114.1 million acres) in 2012, an increase of 58.7 percent.


Developed land grows faster than population: “Build Out”

While developed land increased at a rate of 58.6 percent, between 1982 and 2012, the total U.S. population only increased by 35.6 percent during this time, from 231.7 million people in 1982 to 314.1 million people in 2012. The differential growth rate between developed land and population, where land is developed at a faster rate than population grows, is the essence of a sprawling “build-out” dynamic. When population growth induces the additional people to use land faster than the population grows, it means that the fundamental question of “build up or build out” has been answered by “build out.”


Growth in developed land per person: “Build Out” on individual level

The “build out” dynamic can also be expressed on a per-person level. This can be seen in a nationwide average of how much developed land each American “consumes,” by dividing U.S. developed land by U.S. population. The average “land consumption” per person has increased steadily, indicating a sprawl dynamic.


While an increase of 0.08 square miles per 1,000 people, or 0.05 acres per person, over thirty years may not seem like much. However, when multiplied by 82.4 million new Americans in these thirty years, it translates into 6,595 extra square miles or 4.12 extra million acres of land developed in these thirty years that would not have been developed if the lower ratio of 1982 had been maintained. This is the essence of “build out”—each new additional person consumes more land than previously, and thus developed land grows at a faster rate than population.

Urban and Metropolitan Areas

The differing rates of population growth versus land expansion can also be seen in urbanized and metropolitan areas, the land-use types where the vast majority of American population growth and land development have occurred since 1950. For both urbanized and metropolitan areas, land area expanded at a much faster rate than population over the span of 60 years—a clear statistical indication of sprawl.


The same trend can be seen at the level of individual cities, a favorite mention for anti-sprawl groups in their publications. Individual urban/metro areas have experienced diverse rates of sprawl. Examples of low, medium, and extensive sprawl metro areas illustrate the trend in the graphs below.



Cleveland (“high sprawl”) and Minneapolis-St. Paul (“medium sprawl”) exhibit different magnitudes of the same trend, where land development occurs at a faster rate than population growth. As each city built Interstates and automobile-dependent suburbs from the 1950s onward, population grew at a slower rate than the expansion of developed land area. This rate differential occurred even more so in Cleveland than Minneapolis-St. Paul. But both metro areas experienced non-trivial, significant amounts of sprawl, as per the statistical definition.

For both urbanized and metropolitan areas, land area expanded at a much faster rate than population over the span of 60 years—a clear statistical indication of sprawl.

The surprising result of Los Angeles as a “non-sprawl” city reveals several underlying aspects of how to measure sprawl that often go overlooked. The land area of the Los Angeles metro area has actually shrunk by five square miles, from 4,853 to 4,848 (according to the Census Bureau), between 1950 and 2010, even as population nearly tripled, from 4.4 million to 12.8 million. Many urbanists have commented on how Los Angeles has paradoxically become one of the most densely populated places in the United States, even as it contains one of the most highway-oriented, automobile-dependent landscapes. Los Angeles land area has not expanded primarily due to natural barriers—the hills in the north, the desert in the east, and the sea in the west and southwest. Thus, new population must fit into existing boundaries, and density rises.

Land conversion

The case of Los Angeles shows the limitations of measuring sprawl strictly by the relative rates of population growth and land development. A landscape that takes on many of the features commonly thought of as “sprawl” may at the same time not necessarily meet one of the statistical definitions of sprawl. Therefore, additional measures are needed. Another defining feature of sprawl is conversion of other types of land. Environmental groups decry the loss of farmland and other previous uses as new housing subdivisions get built on the fringe of existing developed areas. Statistically, the amount of farmland (and other) conversion has actually been measured, on a national level, in the USDA’s National Resources Inventory.


As developed land has increased, cropland, pastureland, and rangeland have all decreased. That is exactly what anti-sprawl advocates have warned of. The dynamic can be seen even more clearly when expanding the above land-uses to comprehensively include all other land-use types in the United States.


This graph corroborates, arithmetically, that the largest source of land lost to development has indeed been farmland, followed by pasture and range lands. Its comprehensive cross-section at 10-year intervals shows how all most types of land (other than developed, crop, pasture, and range) have remained relatively constant, with only slight increases in federal lands, forest lands, and other rural (the entire “rural” category includes crop, pasture, range, forest, and other rural).

Strikingly, developed land comprises very little of total U.S. land area. The Midwestern, western and southern states tend to contain vast, enormous areas of undeveloped land. The following graph illustrates the ratio of developed land to total U.S. land from 1982 through 2012.


While the proportion of developed land has indeed grown significantly over the past few decades, it still remained only six percent of total U.S. land as of 2012. A propensity toward even further “sprawl” is certainly possible because of the low proportion of developed land to total land and the abundance of land tends to reduce its price (even accounting for the costs of sewers, roads, electricity, and other infrastructure on newly converted land).

“Build Up or Build Out”: Impact on fuel consumption and petroleum dependence

When land is developed at a faster rate than people can populate it, densities decrease, and vehicle trips, trip distances, vehicle ownership, and fuel consumption tend to rise.

When land is developed at a faster rate than people can populate it, densities decrease, and vehicle trips, trip distances, vehicle ownership, and fuel consumption tend to rise. Other articles in The Fuse have shown a strong correlation between decreasing density on the one hand, and increasing fuel consumption, vehicle-miles-traveled (VMT), and vehicle ownership on the other. In a geographic sense, another way of seeing these occurrences is through the lens of “build up or build out.” When land is developed on a “build up” pattern, it means that population growth outpaces land expansion, and density increases. In contrast, on a “build out” pattern, population growth lags behind land expansion, and density goes down. In turn, decreasing density is strongly associated with increased vehicle ownership, more frequent vehicle trips, and longer vehicle trips—all elements which feed into increased fuel consumption.

A look at how social sprawl underpins physical sprawl

The next article in this series will examine a coincident pattern known as “social sprawl.” Examining the “socio-demographics” of the American population over the past several decades—in the form of household sizes, children per family, preference for single-family housing units, and related factors—shows that social and cultural preferences exercise a quiet but profound impact on choices regarding whether to “build up or build out.” It turns out that physical sprawl is underpinned by “social sprawl,” and that the causes of “social sprawl” spring from some unexpected sources.