Urban Sprawl and Public Health

Designing, Planning, and Building for Healthy Communities

ISLAND PRESS

Contents

WHAT IS SPRAWL? WHAT DOES IT HAVE TO DO WITH HEALTH?

* * *

In 1956, the Federal Highway Act set out to "disperse our factories, our stores, our people, in short, to create a revolution in living habits." Within a year, writer and social critic William H. Whyte was already deeply disturbed by what he saw. Highways were allowing cities to expand rapidly into surrounding rural areas. In a short article published in Fortune magazine in January 1958, entitled simply "Urban Sprawl," Whyte observed that "huge patches of once green countryside have been turned into vast, smog-filled deserts that are neither city, suburb, nor country." "It is not merely that the countryside is ever receding," he warned, but "in the great expansion of the metropolitan areas the subdivisions of one city are beginning to meet up with the subdivisions of another."

Nearly a half century later, the term "sprawl" has entered the American vernacular. Originally a reference to a bodily position—"to lie or sit with arms and legs spread out"—the word has more recently assumed a broader meaning: "to spread or develop irregularly." The Vermont Forum on Sprawl (www.vtsprawl.org) offers a succinct definition of sprawl as "dispersed, auto-dependent development outside of compact urban and village centers, along highways, and in rural countryside."

In common use, sprawl has become a pejorative term. It seems to take on a variety of meanings: cheaply and quickly built neighborhoods at the edge of metropolitan areas, architecturally monotonous residential subdivisions, ugly feeder roads lined with strip malls, lifestyles that center around car trips. Critics of sprawl have unleashed a torrent of pungent prose. William H. Whyte, in his original Fortune magazine article, wrote:

Sprawl is bad aesthetics; it is bad economics. Five acres is being made to do the work of one, and do it very poorly. This is bad for the farmers, it is bad for communities, it is bad for industry, it is bad for utilities, it is bad for the railroads, it is bad for the recreation groups, it is bad even for the developers.

Forty years later, a less measured James Kunstler derided sprawl in The Geography of Nowhere as "depressing, brutal, ugly, unhealthy, and spiritually degrading."

In this book, we do not use sprawl as a pejorative term. Instead, we use it as a neutral descriptive term, as convenient shorthand for a complex set of characteristics of towns and cities. Sprawl refers to the way land is used, the way people travel from place to place, and even the way a place "feels." In sprawling metropolitan areas, the city expands outward over large geographic areas, sometimes in a "leapfrog" pattern (see Figure 1-1). Different land uses—residential, commercial, office, recreational, and so on—tend to be separated from each other. Busy arterial roads are lined with commercial strips, accessible only by car, and there is a relative scarcity of both walkable "town center" neighborhoods and public open space. Distances between things are large, which makes walking and biking impractical, and the low density makes mass transit uneconomical. There is a heavy reliance on the automobile, and the road system may provide few direct connections (see Figure 1-2). Oliver Gillham, in The Limitless City, provides a thorough review of various definitions of sprawl, and offers one of his own: "a form of urbanization distinguished by leapfrog patterns of development, commercial strips, low density, separated land uses, automobile dominance, and a minimum of public open space."

Land use and transportation interact to affect many aspects of human activity, well-being, and health. Heavy reliance on the automobile for transportation results in more air pollution, which contributes to respiratory and cardiovascular disease. More driving also means less physical activity, contributing to a national epidemic of overweight and associated diseases. More time on the roads means a greater risk of collisions with other cars and with pedestrians, with associated injuries and deaths. Sprawling cities threaten the quality of drinking water sources and the availability of green spaces. Even mental health and the network of social interactions and trust known as "social capital" may be affected. To come to grips with the health implications of sprawl and to develop better public policy requires, therefore, an understanding of the physical attributes of sprawl and how they affect people.

DEFINING AND MEASURING SPRAWL

"Urban form" refers to the amalgamation of individual elements of the towns and cities in which we live, work, play, and travel: the schools, houses, parking lots, shopping malls, gas stations, post offices, houses of worship, streets, parks, and stadiums, with which we are all familiar. Urban form is partly determined by natural features—the coastlines of Boston and San Francisco, the riverfronts of Pittsburgh and St. Louis, the mountains outside Denver and Salt Lake City. And urban form is partly the result of public and private decisions made over many years, some explicit, others unintended and even unrecognized. Some aspects of urban form, such as regional commuter train systems, exist on a very large scale, whereas others, such as courtyards and sidewalks, are very small and localized. Architects and urban planners have used many concepts to classify this seemingly infinite variety, to allow urban form to be ordered, studied, and understood. Terms such as density, concentration, centrality, diversity, mixed uses, connectivity, and proximity are all used to define and conceptualize urban form.

Sprawl is one kind of urban form (see Figure 1-3). In this book, as we explore the impact of sprawl on human health and well-being, we look to many sources of empirical evidence. To study the relationship between sprawl and health, a general definition of sprawl is not enough. Scientists need a definition that can be operationalized and measured. This allows them to test specific hypotheses about the impact of sprawl on people.

The literature on sprawl offers a wide variety of definitions. A recent review of many of these found "no common definition of sprawl, and relatively few attempts to operationally define it in a manner that would lead to useful comparisons" of metropolitan areas. Some definitions were narrowly oriented to a single metropolitan area such as Los Angeles. Some definitions were historical, based on the planning process that gave rise to a place; some were subjective, based on notions of ugliness; and some were incomplete, measuring only one or a few dimensions of sprawl such as density or land area.

A widely accepted approach to measuring sprawl was proposed by Ewing, Pendall, and Chen. These researchers aimed to incorporate both land use and transportation in their definition and, accordingly, identified four categories for measurement: the strength or vibrancy of activity centers and downtown areas; accessibility of the street network; residential density; and the mix of homes, jobs, and services at the neighborhood level. Each, they maintained, measures a different and important component of urban form; these might be defined as compactness (density), diversity (the mixture of uses over an area), sense of place (strength or vibrancy of activity centers in a region), and connectivity (street network accessibility, meaning how easy it is to get from point to point on the street system). They created a Sprawl Index with data on twenty-two specific measures grouped under the four categories. This showed the most sprawling areas to be in the South and Southeast, with a few in California. The least sprawling areas are in the Northeast, California (San Francisco), and Hawaii (Honolulu).

CORE CONCEPTS: LAND USE AND TRANSPORTATION

In this book, we also take the approach that both land use and transportation are intrinsic to sprawl. We emphasize two core land use concepts, density and land use mix, and two core transportation concepts, automobile dependence and connectivity. (We acknowledge many other important features of urban form, such as whether development is contiguous or leapfrog, the level of architectural variety, and the supply of bicycle paths and sidewalks.) We recognize that sprawl has different meanings on different spatial scales; the most important features of a sprawling metropolitan area are different than the most important features of a residential subdivision (although they are closely related to each other). And we recognize that sprawl is not a single pattern; different places sprawl in different ways.

Land Use: Density and Land Use Mix

Land use patterns determine the degree of proximity between different places. A higher level of proximity means that destinations are close together, and a lower level of proximity means that they are farther apart. The density and variety of uses in a neighborhood, community, or city district largely determine the functional distances that separate the places in which we live, work, and play. Low-proximity levels typify sprawl; there are both fewer destinations and less variety of destinations in sprawling development patterns compared to other types of urban form. This book shows how land use patterns have direct implications for travel behavior.

The density of a place refers to the quantity of people, households, or employment distributed over a unit of area such as an acre, a square kilometer, or a square mile. The relationship of density to travel behavior has been the subject of considerable study in the discipline of urban and regional planning. Higher density is associated with shorter trips, an increased number of trips taken from home, an increase in transportation options ("mode choices"), and reduced vehicle ownership, compared to lower density. Because of its conceptual simplicity and the ease with which it is measured, density is one of the most commonly used measures in planning.

The land use mix is a necessary and important complement to density. Land use mix is a measure of how many types of uses—offices, housing, retail, entertainment, services, and so on—are located in a given area. A high level of land use mix should in theory reduce the need to travel outside of that area to meet one's needs.

Land use mix is relevant over both vertical and horizontal spaces. In older parts of American cities and towns, the vertical mixing of uses was quite common, and it remains the norm in many parts of Europe. Different types of uses, usually retail and housing, are arranged in a single building, typically with retail on the ground floor and housing stacked above it. With the advent of zoning in the first quarter of the twentieth century, however, the vertical mixing of uses was effectively outlawed in most parts of the United States. Horizontal mixing of uses refers to the location of different types of land uses on adjacent or near-adjacent parcels of land. Empirical research has shown that households located in less mixed environments generate longer automobile trips and fewer trips on foot, bicycle, and transit than do similar households located in more mixed use environments.

Transportation: Connectivity and Automobile Dependence

Connectivity refers to how destinations are linked through transportation systems. While the proximity of destinations is central to shaping how people travel, connectivity also has tremendous importance. A poorly connected transportation system can make even nearby destinations functionally far apart. Conversely, a well-connected system can ease travel between destinations by shortening on-the-ground distances. Connectivity is almost always discussed in the context of the street network. Because streets are the primary arteries upon which travel by most modes occurs, they have a central importance in determining travel patterns.

A well-connected street network features many street linkages between trip origins and trip destinations. A poorly connected network has fewer linkages. One way to think of connectivity is to think of how easy it is to "go around the block." Simply put, going around the block becomes much more difficult where streets do not connect. Block size is the area bounded by streets that form its perimeter. The larger the block size, the more difficult it becomes to get to a destination in a reasonably direct path. Connectivity can also be viewed as the number of street intersections scattered across a neighborhood or district. More intersections mean that there are more possible routes between point A and point B. Conversely, poorly connected systems have fewer intersections, offering fewer travel routes, generally implying a less direct and more circuitous route between points A and B.

The street arrangement with the greatest connectivity is the grid pattern, a simple network consisting of regularly intersecting horizontal and vertical streets framing small blocks (see Figure 1-3). Such networks reduce the distances between trip origins and destinations by providing many intersections and, therefore, many possible routes. In contrast, the dendritic street network (the upper part of Figure 1-3) is characterized by fewer streets organized into a hierarchy based upon the amount of traffic each is intended to carry. Like the lifelines in the leaf on a tree, this dendritic form of transportation is highly specialized. At the core of the dendritic system are major arterial roadways, designed to carry cross-regional traffic. Residential streets form the edges of the dendritic network and are designed exclusively for local traffic; to exclude high-speed vehicles, there are very few connections between these streets and arterials. These form the "loop and lollipop" neighborhoods seen in Figure 1-2. As a result, trips to destinations, especially those out of the neighborhood, become more circuitous, and trip lengths increase. Trips to destinations that are nearby in terms of straight-line distance can become long journeys.

Low proximity (reflecting low density and low land use mix) and low connectivity together predict the fourth cardinal feature of sprawl: automobile dependence. In theory, when proximity and connectivity are high, people would be expected to depend less on automobiles, because other modes of travel such as transit and walking are more competitive with automobile travel. In addition, when proximity and connectivity are high, the average automobile trip would be expected to be shorter. Sprawling areas, accordingly, would be expected to feature enormous amounts of driving.

An overview of recent urban development in the United States lends support to this view. During the past few decades, the fastest growing regions have also featured the most extensive road construction, the greatest geographic expansion into exurban areas, and the steepest increases in automobile travel. These trends seem to emerge in tandem; sprawl toward the edges of a metropolitan area is associated with more driving. For instance, from 1982 to 1997, Atlanta added 571,000 acres to its urbanized area, and added approximately 1.3 million people, meaning that the region urbanized approximately 1 acre of land for every two new residents. During the same interval, the number of miles driven per person in the region more than doubled. Figure 1-4 shows the distribution of average vehicle miles traveled (VMT) per day across the Atlanta region, and illustrates that travel distances are greatest on the region's periphery and shortest nearer its center. By 2001, the average Atlantan (including nondrivers) was driving 34 miles each day—a citywide total of 102,000,000 miles, enough to reach from Peachtree Street to the sun and partway back. The city's rush hour had grown to 7.8 hours each day, and the average Atlantan was spending 34 hours per year stuck in congested traffic.