- Shopping Bag ( 0 items )
"Americans' move to the suburbs has many political implications, but that is a subject on which there has been much pontificating but surprisingly little evidence—until now. And along with explicating what has happened and why, Eric Oliver never lets the reader forget why all of this matters, how his results link up with urgent issues of the day and deep concerns of social scientists. Suburbanization is good for participation but bad for democratic decision-making; no other book presents us with quite this conundrum in such a powerful fashion."—Jennifer L. Hochschild, Harvard University
"Democracy in Suburbia is an important contribution to the literature on participation and metropolitan areas, and to modern democratic theory more broadly. Oliver forcefully shows how the spatial, racial, and class divisions written into metropolitan geography weaken democratic vitality."—Margaret Weir, University of California at Berkeley
"The conrtributions of this book are many. Oliver provides sweeping coverage of the various arguments, claims, and anecdotes associated with suburbanization in the United States; offers a systematic empirical analysis of how various dimensions of suburbanization are associated with different levels of participation, and why; and he proposes how the negative consequences of suburbanization might be ameliorated. Particularly notable is Oliver's engaging style: it is direct, intelligent, and masterfully interweaves anecdotal claims and evidence with more systematic statistical evidence."—Jan Leighley, Texas A&M University
"Eric Oliver's book is a welcome departure from unsubstantiated claims, providing a sophisticated empirical analysis of the impact of social contexts and suburbanization on civic engagement. . . . Democracy in Suburbia is an important contribution."—Daniel Rubenson, Political Studies
"Oliver demonstrates that strong communities and large accumulations of social capital can actually undermine democracy. Class and racial divisions must be dealt with by our society. . . . [I]t is at the local level that the most creative solutions are forged and that citizens acquire crucial civic skills and democratic values."—Todd Swanstrom, Review of Politics
"A remarkable book, literally the first one ever published to present definitive evidence on the crucial issue of the impact of the suburbs on American democracy. . . . It virtually forces the reader to step outside old debates and assumptions and examine anew the consequences of America's preference for fragmented governmental arrangements."—Dennis R. Judd, American Political Science Review
COPYRIGHT NOTICE: Published by Princeton University Press and copyrighted, © 2001, by Princeton University Press. All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher, except for reading and browsing via the World Wide Web. Users are not permitted to mount this file on any network servers.
THE RISE OF A SUBURBAN DEMOS
SPRING VALLEY, TEXAS, is a small suburb west of Houston. Like its neighbors Bunker Hill and Hunter's Point, Spring Valley is not particularly well named: it sits on a pancake-flat prairie, has no identifiable water sources, endures a swamplike climate, and, being a few miles from downtown, is more "urban" than a good portion of Houston proper. As a town, Spring Valley does not have much of a municipal identity. It has no main street, no parks, no monuments, no library; for much of its history, it ran city business out of a nondescript office next to a convenience store. Socially, it is very homogeneous, composed mostly of three-bedroom homes on half-acre lots inhabited by white, middle-class families. With its homeowning population largely supportive of its restrictive zoning codes, few issues ever cause controversy. Spring Valley is a very quiet place where residents mostly keep to themselves. In the past few decades, Spring Valley has also become the typical American town.
Over the past half century, a tremendous change has occurred in the types of places Americanscall home. In 1950, most Americans resided in either large cities or small, rural towns. Today, most Americans live somewhere in between; places outside of big cities but still within greater metropolitan areas—places like Spring Valley, places commonly known as suburbs. These suburbs are dissimilar to both their urban and their rural counterparts. Unlike older, central cities, they are often very singular in their social composition and land use—many contain nothing but homes, nothing but white people, or nothing but the affluent. Unlike rural towns, suburban places are highly interconnected with and dependent on a larger metropolis. Whereas people in rural towns typically worked and shopped in the same place, suburbanites often pursue each activity in other locales. Suburbanization has been one of the biggest changes in American society over the past fifty years. It has affected the ways Americans relate to their families, friends, and neighbors, understand local government, and experience community.
Yet, despite the enormity of this suburban transformation, its implications for American democracy are largely unknown. While the social consequences of suburbanization, such as racial segregation and urban sprawl, are well documented, the effects of America's suburban expansion on its basic mechanisms of democratic government are not understood. Take, for example, civic participation. Recent trends suggest a possible negative relationship between suburbanization and political and civic engagement. Over the past four decades, as Americans have been moving to suburbs, they have also become less likely to vote, less attached to political parties, and less trusting of their political institutions.1 Many scholars believe that the past decades have brought not just the erosion of America's civil society but a disturbing loss of community and fellowship among citizens. From Robert Putnam's requiem for bowling leagues to Alan Ehrenhalt's lamentations for the "lost city," from Ray Suarez's remembrances of communities long gone to Thomas Geoghegan's elegy for a public citizenship, a chorus of scholars and journalists have recently pronounced civil society and community in America to be in ill-health.2
In these criticisms, suburbs like Spring Valley have been fingered as likely suspects. Whether it is the shape of their houses, the design of their neighborhoods, or their absence of public spaces, suburbs are routinely accused of stifling the social interaction, sense of membership, and democratic engagement that once existed in America's cities and towns.3 Along with this loss of community, suburbanites allegedly have lost the capacities and incentives to be involved in public affairs. In other words, Americans do not vote, do not trust their government, or do not join the PTA partly because the physical design and social composition of suburbs are keeping them isolated and preoccupied with private concerns. In his exhaustive analysis of the decline of civic engagement in America over the past thirty years, Robert Putnam estimates that suburbanization and sprawl are accountable for about 10 percent of the problem.4 "The suburb," as architects Andres Duany and Elizabeth Plater-Zyberk argue, "is the last word in privatization and spells the end of authentic civic life."5
Like most assertions about the suburbs, however, such claims are without any empirical basis. We have no real evidence on whether suburbanites are less civically engaged than nonsuburbanites or what impact, if any, suburban social environments are having on America's democratic processes. Suburbs like Spring Valley may now be the typical American town, but we have little understanding of how they affect Americans' commitment to their communities or their ability to govern themselves. This absence of knowledge comes largely from three sources.
First, most of us are unclear about what exactly a "suburb" is. Most places that are within a metropolitan area but not part of the central city are counted as suburbs, yet this usage is confusing because it equates places that are quite different in form and composition—for example, wealthy Beverly Hills, eclectic Santa Monica, residential Walnut, and impoverished Compton are all one kind of place (suburb), as distinguished from Los Angeles (city). Given the wide diversity of places that are within metropolitan areas but outside of central cities, such crude taxonomies do more to obfuscate than to clarify the real picture of suburban life. Architectural commentaries are not very useful either. Suburban civic malaise is often attributed to the absence of public spaces, the predominance of single-family homes with garage facades, and the prevalence of private yards; yet none of these studies enumerate how many suburbs actually have these characteristics or whether such traits are unique to suburban areas.6 Indeed, large portions of Los Angeles, Houston, and Orlando have these "suburban" traits, while many suburbs—like Cranbury, New Jersey, or Petaluma, California—more closely resemble traditional small towns. The concept of the suburb has become saddled with so many stereotypes and misconceptions that most people have little understanding of what really distinguishes America's cities and suburbs from each other.7
Second, critics of suburbs have been equally vague about how suburban environments may distort the process of democracy. Most studies of local politics in America are of large cities, with scholars wrangling over whether cities are dominated by a governing elite or subject to more pluralistic political pressures.8 Little research on local politics, however, has focused on suburbs. Meanwhile, other critics who bemoan the loss of "community" or "civil society" rarely specify what these terms mean or why they are important for democratic organization.9 A protean term like "civil society" can include activities as diverse as gathering informally with neighbors and going to the gym,10 and it is not clear that all such activities are either essential or beneficial for democratic governance. As was demonstrated in Weimar Germany, a strong civil society is no guarantee of stable democratic institutions or peaceful coexistence among the citizenry.11 In their preoccupation with vague notions of community, most criticisms of suburbs have largely ignored other essential questions of democratic governance. For instance, do suburbs limit or enhance the ways that citizens can govern themselves? Does suburbanization create any biases in the democratic process? If citizens in suburbs are less civically engaged, does this necessarily undermine their ability to govern themselves? Most critiques of suburbs and community do not address these questions.
Third, after decades of research, no conclusive evidence exists on whether or not suburban environments actually do shape individual civic or political behavior. Most assertions about suburban civic life are based on either pure speculation or case studies of individual places done in the 1950s and 1960s.12 Interestingly, these early works mostly portrayed suburbs as "hotbeds of participation," with the typical suburbanite frantically running from one type of civic activity to another. Some even described suburbs as embodying the democratic ideal. Although these studies provide interesting descriptions of particular communities, they do not reveal whether any differences that may exist between suburban and nonsuburban residents are systematic. In other words, it is impossible from a study of one suburban community to determine whether the activity level in that place is universal to suburbs or something specific to that locale. To draw conclusions about suburbs as a whole, the researcher must examine a wide range of places to see whether consistent differences arise. Unfortunately, the few studies that employ such data (i.e., cross-sectional surveys of large populations) use only crude city/suburb dichotomies or sample from only a few cities.13 Not surprisingly, these studies have found few effects of suburban contexts on civic behavior, leading some to question whether suburbs have any consequence for American democratic life.14
In short, America may be a nation of suburbs and its citizens may be disengaged from civic affairs, but we still have no idea whether these phenomena are related or what their larger democratic consequences may be. Yet scores of academics, journalists, and public commentators continue to assert that Americans have lost a sense of community and civic responsibility, and that suburbs are, somehow, to blame. Are suburbs really affecting the ways Americans interact with their communities? If so, is this a cause for alarm? or are suburbs simply the unfortunate victims of an intellectual and cultural bias? This book seekd to answer these questions.
General Argument Of The Book
In the pages to follow, I will argue that suburbanization is undermining the optimal functioning of America's local democratic institutions. Local government is important primarily because it provides an accessible and small-scale arena for the resolution of social and economic conflict. According to what I term the authentic governance principle, America's municipalities and other local institutions, as instruments of state governments, should function so as to bring together most people within a geographic vicinity to collectively solve problems related to their area. Local political institutions, as democracies, should be organized so as to directly articulate, or maximize the representation of parties to, conflicts within a particular region. Local governments best perform these functions by maximizing citizen input on salient issues for all residents of a community.
At first glance, suburbs hold great promise for meeting the standards of the authentic governance principle. One of the primary by-products of suburbanization is the movement of Americans living in large metropolitan areas into smaller municipal jurisdictions. Today, more urbanized Americans are governed by smaller municipalities than ever before. These smaller local governments allow citizens to come together in more intimate and immediate settings to resolve their political differences. As I show in chapter 2, residents of smaller places are more engaged in community affairs and active in civic life. Learning the practices of compromise, consensus, and organization building among their neighbors, citizens become better skilled in the difficult art of self-governance. Through the growth of these smaller polities, suburbanization promises the cultivation of a richer democratic practice.
Yet the potential benefits of "small-town" government are lost in the economic and racial segregation that suburbs promote. According to an authentic governance principle, municipalities need to adequately encompass the social cleavages and disagreements that occur among people within a particular area. Suburbs often distort this conflict mandate by dividing citizens along class and racial lines. Many suburban governments are constituted solely by people of one class, one race, or one type of land tenure. When municipal borders separate citizens in such ways, social conflicts that once existed among citizens are transformed into conflicts between local governments. This transformation of conflict, as I show in chapter 3 through 6, deters citizens involvement in local civic life.
To elucidate the consequence of this citizen demobilization, I offer a second new concept, civic capacity. The term refers to the extent to which a community's members are engaged in both political and civic activities. In many ways it is akin to the concept of social capital that has recently been popularized by James Coleman and Robert Putnam. Social capital refers to the social connections between individuals that "facilitate action."15 Like notions of human or physical capital, social capital is a resource individuals utilize to achieve their goals. In Putnam's now famous argument, individuals gain social capital primarily by participating in voluntary organizations, an activity that builds networks and norms of reciprocity and trust and leads to greater health, happiness, and well-functioning societies.16 But where social capital is primarily a measure at the individual level, civic capacity refers to communities. Individuals may hold stocks of social capital; communities have civic capacity. Moreover, where Putnam's understanding of social capital is based primarily in voluntary, nongovernmental action (indeed, political participation is an outgrowth of social capital), civic capacity is not so constrained. It refers to all types of civic and political activities, be they softball leagues or political campaigns.
Civic capacity is crucial for sustaining the well-being of America's democracy. In the United States, we ask a lot of our local governments. They must adjudicate between different interests, aggregate information from their constituents, and perform a multitude of functions with little control over productive resources. To meet their social needs and facilitate the process of self-rule, American localities traditionally have relied on the voluntary activities of their residents. In other words, localities have relied upon their civic capacity to maintain the functioning and promote the well-being of society. Just as an economy profits from its unpaid working sector, such as housekeeping and child rearing, so a polity benefits from its unpaid civic sector. Localities with greater civic capacity have more human resources available to identify and prioritize social problems, lobby for governmental solutions, and find alternatives where public resources are unavailable. Democracies with low civic capacity have fewer resources to solve social problems and are more likely to be subject to greater tensions, through riots, corruption, or civil disorder. Democracies with greater civic capacity not only will be more responsive to social problems but will have more citizens offering extrainstitutional solutions, thus providing greater social stability.
Suburbanization, by segregating the population and suppressing citizen involvement in community affairs, is depriving many localities and metropolitan areas of their civic capacity and thus their ability to solve many contemporary social problems. Extreme concentrations of urban poverty, high degrees of racial segregation, and the rampant sprawl of unplanned growth are all predicaments of metropolitan life, the geographic community in which most Americans live. These social ills continue to defy solution partly because of the political divisions between cities and suburbs. At the institutional level, suburban political fragmentation puts local governments in competition with each other and inhibits intermunicipal cooperation.17 And, as I will show in my empirical analysis, suburban segregation demobilizes citizens and deprives metropolitan areas of valuable human resources to address these problems. By encouraging certain residents to "tune out" local politics or to see themselves as different from the greater metropolis, suburban institutions are depriving the metropolitan community of vital civic capacity. Consequently, social problems that require institutional cooperation and active citizen involvement are going unaddressed. Unplanned growth and sprawl and severe economic and racial segregation contribute to high levels of traffic congestion, pollution, and periodic social unrest. These social ills have cost lives, billions of dollars in property damage, and unquantifiable losses in America's quality of life. If such problems are to be solved, the civic capacity of localities must be increased.
How can this be done? Just as the problems of suburban democracy are institutional in origin, so must be their solutions. By institutional change, I do not mean necessarily the form of local government; as I show in chapter 7, replacing council-manager with mayor-council governments or other such reforms will not enhance the civic capacity of a municipality. Rather, the institutional change must be with the way that municipal borders are drawn and land-use decisions are made. The social and economic segregation causing suburban civic withdrawal is the consequence of municipalities' having inordinate power to determine who lives within their borders. If local democracy is to be reinvigorated, the current structure of municipal government needs to be reconfigured. Previous research in both large cities and rural areas demonstrates that when local institutions bring together a variety of perspectives, political solutions that consider all citizen viewpoints can be found.18 This logic needs to be applied to suburbs. Municipalities must be small enough to generate community among residents but socially and economically representative of the greater metropolitan area, so that citizens do not distance their immediate community from that of their greater surroundings. This could serve to make institutions more cooperative and could reintegrate citizens into the public realm. Local institutions need to function as arenas that bring together the diverse elements and interests of the metropolis, not ones that keep them apart. In chapter 8, I will comment more on how this can be done. Let me now offer some clarification of the central ideas of this study.
Defining A Suburb
Ask most people to describe a suburb, and they will probably conjure images of ranch homes, tree-lined streets, and quiet neighborhoods. It is a picture of residential repose and domestic peace, minivans and soccer moms, and daily commutes and weekend barbecues. The reality of suburbanization is, of course, more complex. Suburbanization actually has been a number of different processes of development that have been occurring for over 150 years. Some suburbs started primarily as residential, middle-class communities; others began life as industrial enclaves situated around large employers; and others still were once rural hamlets that are now transformed by shopping malls and housing developments. Some older suburbs have retained their segregated and residential character, while other suburbs have morphed into large commercial or industrial districts. The vast expansion of suburban areas since 1950 has created an enormous variety of places that exist outside of urban areas. Social analysts have coined a host of terms—such as inner-ring and outer-ring suburb, ex-urb, post-suburb, trans-burb, and edge city—in an effort to capture this diversity. The range of places that now fall under the suburban moniker creates a big dilemma for anyone trying to determine what a suburb exactly is.
According to the categorization scheme provided by the U.S. Census Bureau, a suburb could be considered any part of a "metropolitan area" that is not in the central city. A metropolitan area is a major population center composed of a central city of at least 50,000 people and the surrounding county or counties that are densely populated and economically interconnected with the central city. In 1990, as illustrated by map 1.1, there were 329 different metropolitan areas ranging in size from the 18 million people in the greater New York area to 56,735 people in Enid, Oklahoma. These large metropolitan areas contain an enormous variety of cities, towns, townships, villages, and other municipalities, designated by the census as "places." All places that are not central cities within metropolitan areas typically get counted as suburbs.
The census scheme, however, does not offer much assistance for distinguishing suburbs from either central cities or each other. Take the example of New Jersey, a state composed almost entirely of "suburbs." The U.S. Census Bureau has classified the entire state as part of a metropolitan area, yet New Jersey has few dominant cities. Newark is overshadowed by New York City, while Trenton and Camden are eclipsed by nearby Philadelphia. The remainder of New Jersey's municipalities and townships, while meeting the census definition of suburbs, are hardly uniform: Elizabeth and Hoboken are gritty and industrial, Montclair is middle-class and racially diverse; Short Hills is affluent and residential; Princeton has office parks, a university, and shopping malls; Cranbury and Hopewell still retain the flavor of small, rural towns. If all these places are considered suburbs, then how do we distinguish them, not only from cities like Philadelphia and New York, but from each other as well? Aside from their smaller size, middle-class Montclair, affluent Short Hills, and rural Hopewell have little in common that distinguishes them from gritty Elizabeth or academic Princeton.
The observations of architectural commentators are not very useful either. Many critics of suburbs focus on certain environmental characteristics, like uniform single-family homes widely spaced with fenced, private yards; solely residential communities composed of nothing but mass-produced tract homes; or residential developments oriented primarily around maximizing privacy and ease of automotive transportation. Indeed, an entire architectural design movement has arisen in response to the isolating and privatizing characteristics of America's suburbs. "New urbanist" communities such as Seaside, Florida, or Laguna West, California, seek to integrate housing, workplaces, and shopping in new patterns to restore public spaces and rebuild the community that is putatively absent in so many contemporary suburbs.19
Yet not all suburbs share these isolating characteristics—many, like Petaluma, California, or Concord, Massachusetts, were quintessential small, rural towns that were swallowed up by expanding nearby metropolitan areas. Many older suburbs, like New Rochelle, New York, are celebrated for their pro-civic orientation. And not all large cities have neighborhoods that promote social interaction. Significant parts of Houston, Phoenix, and Jacksonville are composed of residential neighborhoods with single homes, large yards, and high fences. While many of these commentators' assertions about the alienation and privatization of suburban designs are provocative and important to consider, most architectural criticisms are simply too vague and unspecified to meaningfully designate places in the contemporary metropolis.
Clearly, a new way of classifying metropolitan places is needed. Toward this end I start with the following deductions. When we address the democratic implications of suburbanization, the most important factor to consider is the distinct political identity of the suburb. Suburbanization is, as Michael Danielson argues, primarily "a political phenomenon."20 At the most basic level, municipal boundaries are what separate central cities from suburbs and suburbs from one another. Spring Valley, Texas, is physically contiguous with neighboring Houston but is a separate social and political community by virtue of its municipal government. Municipal boundaries, by dividing the metropolitan population into distinct political entities, also create communities with particular interests, interests that often compete with those of other municipalities.21 With particular zoning laws and municipal ordinances, suburban governments thus shape the social composition of the community. Town government is also the most primary unit of American democracy: it defines political membership, the agenda of local politics, and the ways people interact. Municipal policies determine who lives in a community, what activities take place, what public issues its residents face, and even what types of public space it contains. When we think of suburbs in terms of their political institutions, all incorporated places within metropolitan areas, from the smallest hamlet to the largest city, can be considered as similar units of analysis. Despite their many differences, giant Houston and tiny Spring Valley are fundamentally comparable as units of democracy. For this study, I will not be comparing suburbs to central cities or analyzing just suburban places; rather, I will be examining all municipalities within a metropolitan area (central cities, suburbs, edge cities, etc.) and counting them as comparable measurement units.
But if all municipalities within a metropolitan area are equivalently similar as units of analysis, then what are the most appropriate characteristics by which these places can be distinguished from one another? This query has no easy answer. In today's diverse metropolis, municipalities can be distinguished by hundreds of traits ranging from their sewage facilities to their street widths. Unfortunately, previous research on America's cities is not very useful in providing a definitive list. Classic theories of urban sociology from Louis Wirth differentiated large cities from rural areas by their size, density, and social heterogeneity.22 Although large central cities resemble each other in these ways, not all suburbs are uniformly small, sparse, and homogeneous. Nor are the observations of contemporary critics useful, because many civically offending traits, such as the absence of public spaces or the presence of privatizing architectural forms, are so difficult to specify. For example, are streets public spaces? If so, then would not wider streets, an often criticized characteristic of the modern suburb, be counted as more public space? Are the negative effects of private yards more important than the positive effects of front porches? Even if one could answer these questions, it is not clear how they would translate into easily quantifiable measures.
To properly distinguish among municipalities in the contemporary metropolis, we need to look beyond simple city/suburb dichotomies and find indicators that are easily measured. In the recent historical development of American metropolitan areas, there are six dominant trends in place differentiation.23 As America suburbanized over the past fifty years, its cities and towns have become increasingly distinguishable by these six characteristics:
1. Population size. The most distinguishing aspect of suburbanization has been the migration of the metropolitan population away from large central cities to smaller and medium-size places. Whereas in 1950 most urbanized Americans lived in large cities of over 100,000 in population, today most people live in smaller places. This diminution in the size of the typical metropolitan place is the very essence of suburbanization. In their most rudimentary form, suburbs represent the fragmentation of metropolitan areas into smaller political units. However, not all suburbs are identically small. Some hold only a few hundred residents, while others, like Garland, Texas, or Livonia, Michigan, contain over 100,000 people. Population size thus distinguishes all places in the contemporary metropolis.
2. Economic composition. With the political fragmentation of the metropolis, American cities have become increasingly distinguished by their affluence and economic composition. Until recently, most American cities contained a wide assortment of social classes and were within a relatively narrow economic range of each other. With suburbanization, however, America's cities have become highly stratified by their affluence. Some communities are desperately poor, with a median household income below $20,000 a year; others are quite affluent, with median household incomes well over $150,000 a year. In this stratification, these places have become distinguishable not just by their wealth but by their economic homogeneity. Many places, like Short Hills, New Jersey, are inhabited almost solely by wealthy people, while older, industrial towns, like Camden, New Jersey, are populated largely by the poor. Affluence thus differentiates not just America's citizens but its cities as well.
3. Racial composition. Although America's cities have always held ethnic neighborhoods, most larger municipalities were still ethnically and racially mixed places. As with affluence, suburbanization has taken the racial divisions that once separated neighborhoods within cities and institutionalized them with municipal boundaries. Today, most African Americans and Latinos living in metropolitan areas are concentrated in a few neighborhoods of central cities or a handful of "minority suburbs," while most whites live in predominantly white suburbs. These racial divisions do not simply mirror economic status. In many metropolitan areas there exist both poor white suburbs, such as Merrionette Park, Illinois, and middle-class suburbs with significant minority populations like Cheverly, Maryland. Race itself has become a distinguishing characteristic of the American city.
4. Land use. For most of its history, America's cities had a combination of residential, industrial, and commercial areas. Because of limited transportation resources most people needed to be close to their work, and most people traveled to downtown areas for shopping and entertainment. But with suburbanization and the expansion of the highway system, the accessibility of transportation made possible by the automobile has also served to differentiate places by their land use. In today's metropolis, places are now composed solely of homes (Spring Valley, Texas), or businesses (Industry, California), or are even noted for their shopping malls (King of Prussia, Pennsylvania). Although a large portion of suburbs are residentially predominant, not all are bedroom communities. Many still retain a mixture of commercial and residential sites. In today's polymorphous metropolis, land use distinguishes suburbs both from central cities and from other suburbs.
5. City age. The rapid expansion of suburban areas in the past thirty years has made city age an important characteristic for distinguishing American cities. Much of this expansion has been part of a larger regional migration to new metropolitan areas in the Sun Belt such as Jacksonville, Houston, Phoenix, and Atlanta. In many of these places, new communities with over 20,000 residents have sprung up within a couple of years. Meanwhile, many older cities and suburbs in the Northeast and Midwest have either ceased growing or lost population. In the ever expanding American metropolis, age has become an increasingly prominent community trait.
6. Political institutions. It is impossible to understand suburbs without looking at their political characteristics. Suburbs are the consequence of political boundaries, and their distinct social composition is the result of municipal practices in zoning ordinances, development, taxation, and land use. But beyond this, suburbs often have distinctive types of electoral systems and institutions of government that can influence citizen activity. Many suburbs have "reform-style" political institutions (e.g., council-manager governments with at-large representative districts andnonpartisan elections), while larger and older cities often elect mayors and council members from specific districts. But while reform governments are more prominent outside of central cities, not all central cities have prereform-style governments, particularly in the Sun Belt and western states. And, as the Republican political machine in Nassau County, New York, demonstrates, not all suburbs are without partisan-style political structures. As with social characteristics, municipal institutions distinguish all cities and suburbs from each other.
Taken together, these six characteristics best distinguish the variety of places that constitute the American metropolis. And while some of these traits are mildly correlated with each other, each represents a distinct dimension of America's recent suburbanization. To gauge the civic consequences of suburbanization, we must consider the effect of each of these traits separately. To understand the democratic consequences of suburbanization, we cannot simply compare places as either city or suburb; rather, we must estimate how a community's size, affluence, racial composition, and the like, distinctly affect the process of self-governance. Once each of these characteristics is examined separately, then we can evaluate them in concert to understand the effects of suburbanization. In other words, in that suburbanization has led to the increasing polarization of America's municipalities along these dimensions, we can understand the cumulative civic consequences of suburban growth by comparing how each of these traits shapes civic involvement. Just as a physician determines the health of an individual by examining a variety of factors—such as temperature, heart rate, and cholesterol level—in concert, so too must we seek to understand democracy in suburbia by examining each of the relevant factors individually and then making a cumulative assessment.
Of course, by focusing on municipalities as the unit of analysis, we lose the ability to examine other contextual areas. In particular, we will not be able to ascertain the civic effects of neighborhoods or architectural design. This is an important drawback. After all, as a whole, New York City can be an anonymous and isolating place, but many of its residents find their neighborhoods quite friendly, personable, and interconnected. The same holds for most suburban places. Indeed, many critics of suburbs focus less attention on suburbs as a whole than on particular characteristics that are common to suburban neighborhoods, such as missing sidewalks or gated streets. Relying on municipalities as units of analysis, we relinquish the ability to know whether these aspects of urban and suburban environments shape civic life. The municipal focus also means that we cannot ascertain the effect of living in the unincorporated outskirts of a metropolis. For many suburbanites, local government is that of a county or a series of special districts. It is unclear how these political arrangements are civically different from an incorporated area. But to examine both neighborhoods and unincorporated places raises a host of measurement and methodological problems like the ones mentioned above. Moreover, local government law rarely gives any rights or powers to neighborhoods.24 Important as they may be for feelings of community, neighborhoods are much less important as arenas for social and political organization. Consequently, in the trade-off between what is clearly identifiable and measurable and what is not, I have chosen to concentrate on the former.
Citizen Participation And Local Democracy In America
Local government is the foundation of American democracy. It is the most visible and immediate unit of government—locally based services such as education, roads, police, and water are the functions of government that have the most impact on citizens' daily lives. It is the most accessible arena of government—whereas the voices of citizens in a nation of 280 million are necessarily faint, in a city of a few thousand they are loud and clear. It is also a vital component of public policy—since the 1960s, many federal programs, ranging from urban redevelopment empowerment zones to the rehabilitation of hazardous waste sites, have been assigned statutory mandates requiring local citizen participation in policy formulation. Similarly, with the devolution of responsibility for the control and administration of other policies, such as TANF (Temporary Assistance for Needy Families)—formerly AFDC (Aid to Families with Dependent Children)—health care, and environmental regulation, local governments have become essential instruments in the maintenance of social welfare. Local government is also the proving ground of a democratic citizenry. Through participating on local boards, joining with neighbors for community actions, or working in city elections, citizens acquire crucial skills and become familiar with the public realm. In short, municipalities are where most citizens are both learning and exercising the arts of self-governance.
Yet despite their significance, our understanding of how well localities function as democracies is incomplete, particularly in an era of suburbanization. Most studies of local democratic government have revolved around a long and exhaustive debate about community power and "who governs" America's cities. Since the 1950s, political scientists and sociologists have factionalized: some view local communities as controlled by a "power elite" of social and economic luminaries, while others view government decisions as subject to pluralistic competition between groups that have varying in different policy arenas.25 Later scholars have tried to reconcile these perspectives by focusing on the land-oriented nature of local politics and the advantages that small groups of economically powerful interests hold for constructing governing "regimes," particularly when they center on the imperatives of economic growth and development.26
Although this debate illuminates much of the logic behind policy making in large cities, it is not particularly useful for exploring the democratic implications of suburbanization. Theorists of community power focus mostly on larger cities. Their analyses typically presume plurality of interests within a community and then work to determine how well those interests are represented within the policy-making process. Yet such explanations are of questionable validity for suburbs, a great many of which are very singular in their social composition or quite limited in their land use. In a community that is composed only of the affluent or homeowners, many presumptions about pluralistic competition between interest groups or the machinations of growth regimes are no longer germane. Few theories of local government consider how systematic differences in the social composition of American localities affect the dynamics of democratic government. In other words, even though American cities are becoming increasingly differentiated by their size, affluence, and racial composition, most existing frameworks for evaluating local democracy cannot explain how this differentiation affects citizens' ability to represent their interests and make their political institutions responsive to their concerns.
An alternative perspective on local politics that takes social segregation into account comes from economics. According to "public choice" theorists, intermunicipal competition in a politically fragmented metropolitan area has changed the nature of traditional democratic organization, with market-driven efficiencies replacing traditional mechanisms of democratic decision making.27 Under this framework, local leaders in politically fragmented metropolitan areas compete for ratables (i.e., those who pay more in taxes than they require in services) and development. Consequently, leaders are under great pressure to tailor services to particular constituencies at a minimum cost. Citizens, in this view, are transformed into consumers, basically shopping among various suburbs for the optimal balance between taxes and services offered. If citizens do not like policies, they can simply "vote with their feet" and move elsewhere. This marketlike competition between suburbs will create more homogeneous polities where citizen interests are easily represented and services delivered with greater efficiency.
But while appealing in its logic and simplicity, the public choice perspective is not very good at explaining the dynamics of democratic politics. To begin with, most economic models are based on very heroic assumptions about citizen/consumer information and behavior. Little evidence exists that citizens are aware of most services offered by their city and neighboring cities, or that they actually calculate these as the primary determinants of their residential choice.28 People may choose to live in a place because it has good schools or less crime, but they are also constrained by many other factors, such as the housing market, proximity to employment, and such personal desires as proximity to family. Moreover, from the perspective of democratic government, the public choice model is inherently biased. The ends of democracy can be achieved through intermunicipal competition only if all citizens have equal resources. In other words, democratic citizens can be consumers only if they all have equal amounts of "shopping" power. Yet not all citizens are equally mobile—affluent people enjoy a much higher degree of residential choice than poor folks and are more desirable as taxable ratables. Consequently, in the municipal market, the affluent have disproportionate power as citizens to dictate public policy. This violates the fundamental principles of democratic politics, which presumes equality among all citizens in representing their preferences. While useful for explaining some of the pressures faced by local leaders and the logic of policy outcomes, the public choice model is not well suited for evaluating the democratic implications of suburbanization.
The inadequacies of existing theories thus require that we step back and reconsider the determinants of local democracy. The United States, like all democracies, faces a continual struggle between the prerequisites of popular rule and the practices of actual governance. Most people understand the term "democracy" in relation to its Greek origins: democracy is literally the sovereignity or role (kratos) of the people (demos). Yet built into this deceptively simple concept are numerous difficult and intractable issues concerning who the demos are, how exactly they are to rule, over what they rule, and to what ends. For instance, are the demos simply a numeric majority or all the citizenry? In a world of diverse and particularized interests, how can a democracy reconcile the demands of minorities holding intense preferences on a few issues and a majority harboring diffuse and limited preferences on all issues? How can a democracy protect the rights of numeric minorities yet still respect the demands of the sovereign majority?
Of course, answering all these questions is not only beyond the scope of this book but would simply reproduce a vast literature within democratic theory that is better explicated elsewhere.29 Therefore, for the purposes of this study, I adopt the simple proposition that a true democracy depends upon all citizens' being able to express their preferences with equal measure. As Robert Dahl would argue, the conundrums of democracy noted above are most effectively resolved not from policies or procedures based on the desires of the majority but from the consent of the largest number of citizens.30 The best corrective for the dynamic tensions inherent in democratic government is maximizing the input of the citizenry. In James Madison's design, by extending the sphere of representation (i.e., increasing the amount of citizen input), democracies can counteract the harmful tendencies of political factions.31 Minority interests can best protect their rights and the majority can best express its preferences when citizens are actively and continually making their wishes known. In other words, democratic government may be rule of the people in principle, but in order for the people to rule in practice, they must regularly articulate their preferences to others and their political institutions.
Therefore, the first step in evaluating the democratic consequences of suburbanization is to examine how citizens articulate and represent their interests to their local governing institutions and fellow citizens. Of course, examining citizen participation alone will not provide a complete description of whether or how suburban governments are performing as democracies. We must consider other factors, such as the structure of local institutions and the fiscal limitations of the localities, when evaluating local political outcomes. Furthermore, as I will discuss in chapter 8, low levels of citizen participation do not necessarily indicate that all citizen preferences are going unmet. Nevertheless, citizen participation is influential in shaping government policy: citizens who make their preferences known are more likely than citizens who are silent to get those preferences represented.32 Participation may not be the only determinant of democratic governance and is not always the most important one; it is, however, the most essential one. Without some level of effective citizen participation in the governing processes, democracy cannot exist. Therefore, by focusing on civic participation, we can evaluate the ways in which suburban citizens are representing their own interests to local governments.
For this study, I focus on the five important types of local civic activity:
1. Voting. Arguably, voting is the most important civic act within any democratic system. It is the most common form of participation, the easiest way that citizens can voice their preferences, and possibly the best means for gauging the overall sentiments of a polity. Voting is also the key mechanism for controlling political leaders, with the reelection mandate ensuring some responsiveness to citizen concerns. Thus the simplest and crudest way of gauging a polity's democratic performance is to see whether or not its citizens are voting.
2. Contacting officials. Although voting may be a crucial mechanism by which citizens control their political leaders, it is a rather blunt instrument for expressing their preferences. Outside of referenda and ballot initiatives, voters rarely determine specific policies and are usually just electing representatives, often with unclear mandates. Nor is it clear that places with low voter turnout, such as the United States or Switzerland, are any less democratic than places where everyone votes. Thus other forms of participation are important to consider. One effective way that citizens express their preferences is through directly contacting local officials. By writing letters, making phone calls, and even scheduling meetings, citizens articulate their particular concerns to greater effect. While contacting may be less common and more difficult than voting, it can convey the intensity and strength of preferences. Elected officials who face reelection and are interested in gauging the sentiments of their constituencies in the absence of information thus pay particular attention to these calls. The extent to which citizen are able to contact their officials is, therefore, some indicator of the polity's responsiveness.
3. Attending community board meetings. In the devolution of policy implementation from the federal to local governments, community board meetings have become an ever more important mechanism of gathering citizen input and ensuring institutional accountability. From school districts to planning commissions, from city councils to mosquito abatement zones, board meetings are vital elements in the administration of local services. In many ways, these meetings are a good venue for citizens to voice their opinions, as most meetings are specifically designed to be receptacles for community opinion. Participation in these meetings therefore is also an indicator of how well citizens control the affairs of their community. Governments with higher meeting attendance are, in theory, places that are also hearing more citizen input on the direction of local policy.
4. Participating in voluntary organizations. In recent years, scholars have begun to focus on the importance and role of associational activity or "civil society" for democratic government. According to many thinkers, the political norms and networks of reciprocity that citizens develop in voluntary organizations are vital for maintaining the health of democracy.33 By associating with neighbors and taking part in organizations, people come to know the issues that shape their lives, acquire techniques for acting collectively, and adopt norms of consensus and compromise necessary for democratic governance. Voluntary public, associational activities are a crucial component of democratic life. In places with a richer associational life, citizens will be able to link more easily with their neighbors, will be informed about local issues, and will express their opinions to local institutions.
5. Working informally with neighbors. Other important civic components of local democracy involve less formalized patterns of association among citizens. Many critics believe that as a by-product of suburbanization the sense of "community" that once existed in America's cities and rural towns has been supplanted by an individualistic and private-regarding orientation. Although the term "community" is somewhat vague, a theme common in these criticisms is that suburbanites are somehow less committed to the localities and less interactive with their neighbors. Such alienation is potentially threatening to a democracy, where citizen preferences are best expressed in concert and where social problems require voluntary collective action. The degree to which citizens work with their neighbors demonstrates their experience of community and their ability to congregate with fellow citizens of their localities.
Taken together, these five civic activities represent a broad cross section of citizen activities that are essential for democratic governance. Of course, one may wonder whether such a list is too broad to permit any generalizations about the civic consequences of suburbanization. Why would a suburban environment affect voting rates in the same way it influences attendance at PTA meetings? Although the five activities may not all be affected by the suburban environment in quite the same way, we must consider a large set of behaviors when considering the civic implications of suburbanization. Like the proverbial blind men trying to describe an elephant, if we focus on any one civic activity, as each blind man senses only one of the elephant's body parts, we may wrongly characterize public, associational activity as a whole: a person may belong to the local Moose Lodge but otherwise be discouraged from local civic activity; a mother may never vote but be quite active in working with neighbors on the PTA; an elderly man may always vote but feel unable to express his more specific preferences to elected officials. Using any single act as a measure, we may misrepresent the whole of suburban civic life.
Conversely, some may wonder about other important civic activities not listed here. For example, churches are important centers for local civic life, serving as places for recruitment and dissemination of local information. Similarly, campaign work, protesting, and contributing to organizations are also ways that citizens express their opinions. All of these behaviors are important to local democracy, but they are not necessarily well suited for this study. Church activity is notoriously difficult to measure: survey respondents often misreport their activity; church involvement is not easily quantified, as church attendance may measure feelings of religious intensity rather than the amount of social contact.34 Campaign work and protesting are among the least common forms of participation, with fewer than 5 percent of the population taking part in either activity, particularly at the local level.35 Given the low rates of activity, it will be hard to find significant differences across different types of places. Similarly, political contributions are largely related to individual-level income, mostly national in direction, and not necessarily a good indicator of local civic involvement. Therefore, for this study, I have limited the analysis to those activities which best represent the interactions among citizens, their political institutions, and their communities.
These activities also provide a rough sketch of the social patterns of residents and how they interact with each other to organize their society. In other words, taken together, these behaviors outline the contours of this slippery and problematic term—"community." Although defining a protean concept like community is beyond the scope of this book, these types of political interaction, particularly the less formal types of political behavior, such as attending organizational meetings or working with neighbors, illustrate the ways that people do interact with their neighbors. Insofar as community is constituted by the interaction of a set of members, comparing civic behaviors across metropolitan social contexts allows us to evaluate how suburbanization has shaped patterns of community in the contemporary United States.
Social Contexts And Civic Participation
Having come to an understanding of how to define suburbs and an appreciation for the need to study local civic participation, we must next determine how the two are related. In other words, how does the size, affluence, or land use of a place shape the way its inhabitants participate in the five civic activities listed above? This question is important to answer. Suburbs may be distinguished by their affluence and residents of nonaffluent places may vote less, but without any theory linking the affluence of a place to its residents' voting behavior, any correlation between the two may be spurious.
Unfortunately, previous research is not very helpful in this regard. Most scholars who have studied social environments generally ignore civic participation.36 For instance, Louis Wirth, the father of modern urban sociology, speculated about the social and psychological effects of urban environments but was generally unconcerned with questions of civil society or local political involvement. The best he could offer was the idea that alienated urbanites compensated for their loneliness by joining clubs and organizations, an assertion that has never been substantiated.37
Similarly, political theorists who have considered the civic implications of social environments, such as Rousseau, Montesquieu, and Dahl, often arrive at contradictory speculations because they do not first postulate why citizens actually participate in civic life. For example, in their book Size and Democracy, Dahl and Tufte reason, "[S]maller democracies provide more opportunity for citizens to participate . . . but, larger democracies provide citizens opportunities to participate in decisions . . . to control the most important aspects of their situation."38 But in making their claims, Dahl and Tufte assume that smaller places provide more opportunities for participation and that these increased opportunities stimulate citizen involvement. Such assumptions, however, are neither based on any general theory of civic involvement nor empirically tested. We do not know whether a city's size really changes opportunities for participation, whether opportunities really do influence involvement, or what other characteristics might affect the relationship between a city's size and its residents' levels of civic involvement.
Contemporary empirical research on how and why citizens participate is not very helpful either for explaining the impact of social environments, because most studies typically view citizens in isolation. Over the past fifty years, political scientists have explained why citizens participate with a host of theories that range from early learned behaviors and unconscious needs, to analyses of individual class position and interests, to "rational choice" models that treat citizens as sequestered utility maximizers participating only when the benefits of their actions outweigh their costs.39 None of these theories really factor environmental effects into their explanations. With the notable exception of voter turnout, most theories of civic participation do not take the social or political environment of the respondent into account.40 Citizens are characterized as atomistic creatures making political choices and decisions largely in a social and institutional vacuum. Despite the near axiom that human behavior is the function of both individual and environmental characteristics, most past research on social contexts and political participation has focused on one to the exclusion of the other.
To understand how suburban environments may shape civic life, we need to start with some basic deductions and form some new hypotheses. The first question we will want to ask is whether civic acts themselves differ across various places. In other words, is casting a ballot, writing a letter to an elected official, or meeting with neighbors a different experience in Boston from what it is Bellevue or Baton Rouge? For the most part, the answer seems to be no. While state laws and local procedures may change the ease of voter registration or allow people to vote by mail, the act of voting itself is probably impervious to the affluence or racial composition of a locality. Similarly, a place's size or land use may affect how difficult it is to meet with neighbors or contact an official, but the act of meeting or contacting is ostensibly the same no matter where one lives.41 In terms of practice, civic activity is roughly the same in Houston as it is in Hoboken.
Therefore, if social contexts are shaping individual civic behavior, they must be doing so indirectly by influencing the determinants of participation. Put differently, a place's racial composition does not make the act of voting any different, but it does change the other factors that influence whether one is likely to vote. What are these other determinants of participation? Here, the voluminous literature within political science offers a wide selection of choices ranging from the incentives to participation to the costs.42 In the most recent and comprehensive study of political participation in the United States, Sidney Verba, Kay Schlozman, and Henry Brady offer what they call the "civic voluntarism" model.43 According to this framework, the major influences on civic participation can be roughly grouped into three categories: skills and resources, interest, and mobilization. People are more likely to participate if they have knowledge of politics or financial means, if they are more psychologically engaged by politics or concerned with political local events, or if they are recruited by others to take part. Verba, Schlozman, and Brady conclude that many individual-level influences identified by other researchers, such as education or age, ultimately can be subsumed under one of these three characteristics. In other words, education is important for political participation because it stimulates political interest, age provides citizens with greater skills and knowledge, and so forth. Thus, according to the civic voluntarism model, if we want to understand why citizens may or may not participate, we need to first determine why they are more interested in public affairs, how they acquire civic skills and resources, or why they are more likely to be mobilized for political action.
When we reflect on the civic voluntarism model (or any other model of civic participation, for that matter), the social and political characteristics of metropolitan places come into consideration. Although this was not explored in the original formulation of the civic voluntarism model, political interest, mobilization, and resources are all affected by a person's social surroundings and can vary in systematic ways. Take, for example, why people are interested in local affairs. One reason usually overlooked by political scientists is the simple fact that politics is a lot more exciting in some places than in others. In cities like New York or Philadelphia, a range of political issues and visible, colorful political candidates enliven local political contests; in many small towns, l ocal politics barely penetrates the public consciousness. Of course, not all small towns are boring—some places have vicious fights between local figures that are the topic of endless town gossip. Nevertheless, political scientists have long asserted that politics is more interesting where there are bigger stakes in question or where larger social cleavages are dividing the community. The same could be said of political mobilization. In some places, people are more familiar with their neighbors or host more local events; in other places, people hardly know their fellow townsfolk. Where such people are socially familiar, neighbors are more likely to talk about politics and recruit others for local activities.44 Even the impact of individual resources on participation is relative to context—where participation is more difficult, the relevance of individual knowledge and skills grows.45
Continuing with the civic voluntarism model, we can identify the linkages between suburbanization and civic participation. I believe that the civic effects of suburban places originate in their local political institutions. Partly this comes from their form of government. Partly this comes from zoning ordinances, taxes, annexation, and other policies that then define the social composition of a community, such as its population size, economic composition, land use, the age of its building structures, and, to a limited extent, its racial composition. These social characteristics, in turn, shape the local political agenda, the social relations between neighbors, and emotional ties of residents to their community. Places that are racially and economically homogeneous are less likely to have political controversies but may have stronger ties between neighbors. Such factors then influence the determinants of civic participation. For instance, as outlined in the civic voluntarism model, local political issues shape interest politics; familiarity between neighbors alters patterns of mobilization and the value of resources. These determinants, in turn, shape the actual civic behaviors, ranging from the informal meeting of neighborhood groups to attending board meetings to voting in local elections. Of course, this causal influence is not entirely a one-way street. Through political action, citizens can change the character of local political institutions and practices. For example, a local organization can galvanize the citizenry to change municipal zoning practices, which then may influence the social composition of a place, and so on. Nevertheless, for the purposes of this book, I will focus primarily on the ways that the social composition of suburbs influences the primary determinants of local civic action.
Data And Measures
The final challenge in examining the democratic consequences of suburbanization is to test the relationships listed above with actual data. In many ways, this presents the biggest challenge to the research. Most existing studies of cities and suburbs are inadequate because they focus only on one particular community.46 While this approach can highlight specific details of policy making within a community, it is inadequate for drawing systematic conclusions across a wide range of places such as suburbs. Indeed, this is a problem with most people's opinions about suburbs: most people have an experience or an opinion about civic life in suburbia, but few know whether their view is shared by others. In order to develop sound empirical conclusions about suburban civic life, we need to sample from a wide number of people across a wide number of places.
Unfortunately, most existing data are not appropriate in the study of suburbs. Most data are either solely on the aggregate level (e.g., county-level demographic statistics or voting patterns) or the individual level (e.g., surveys of individual respondents). This greatly limits whatever conclusions can be drawn about the effects of one level on the other.47 For example, if we make generalizations about suburban behavior by comparing aggregate voting rates in suburbs and cities, we have few ways of knowing what effects are related to the social characteristics of place and what are related to the characteristics of the individual inhabitants. Turn-out in local elections may be higher in a suburb like Spring Valley than in nearby Houston, but with only aggregated data it is impossible to know why: is this because Spring Valley's more affluent and educated citizens are just more likely to vote than Houstonians, or is there something specific about living in Spring Valley that makes people more civically oriented? On the other hand, if we look only at individual-level data as in most surveys, we have no way of gauging the effects of the social context. To appropriately measure contemporary metropolitan social contexts, we need data that measure individual traits and behavior, contain information on the context, and sample from a wide variety of places.
The bulk of my arguments come from a unique series of datasets that merge individual-level survey data with aggregate census data. Most of the individual-level data come from the 1990 Citizen Participation Study (CPS), currently the most comprehensive source of information on the participatory activities of the American public.48 The Citizen Participation Study is a large-scale, two-stage survey of a random sample of Americans conducted in 1989 and 1990. In the first stage, 15,053 Americans were interviewed by telephone about their voluntary and political activities and demographic characteristics. In the second stage, longer, in-person interviews were conducted with 2,517 of the original 15,053 respondents, with more detailed questions regarding their social and political activities. To measure the social context, I extracted information on both city- and metropolitan-level social characteristics from the 1990 Census (Census of Population and Housing 1990) for all of the respondents in these studies whose city of residence could be identified. For the CPS, this provided information on 1,633 different places for the screener data and 822 different places in the follow-up study. Together these data constitute what I will refer to throughout this study as the Citizen Participation/Census (CPC) dataset. I have also supplemented the CPC with data from the 1996 American National Elections Studies (NES). The NES, conducted before and after every congressional election by the survey research center at the University of Michigan, is the preeminent source of data on American political attitudes. It offers a wide range of questions on citizen attitudes toward political institutions, assessments of political efficacy, feelings of trust, and policy opinions. Once again, where each respondent's place of residence could be identified, place-level census information was appended. Combined, these data represent a truly unique opportunity for discerning the effects of suburban environments.
These data are unique because they allow us to overcome the major difficulty in estimating the effects of metropolitan contexts—isolating their effects per se. For example, if we were to compare the simple rates of voter turnout between small and large cities, we would not be able to know which effects are due to a city's size and which are due to the characteristics of the individuals who live in such places. Multilevel data allow for the contextual effects of social environments on individual behavior to be differentiated from the individual-level characteristics of the residents. In other words, these data will allow us to estimate how much variation in people's behavior is attributable to their social surroundings as opposed to their own education, age, income, and the like. Drawing from a national sample, these data also allow for sampling from a wide variety of places. And, using so large a survey as the CPS, most of the analyses are based on a large number of individual cases.
Of course, distinguishing these effects also depends upon using statistical techniques that allow the distinct effects of the social context to be isolated while taking into account other characteristics of both the city and its inhabitants. To meet these ends, I rely on a multivariate regression analysis, which allows for the effects of a particular variable, such as a place's size, to be estimated while taking into account the effects of other variables, such as an individual survey respondent's education, age, race, or length of residence.49 All of the individual-level traits that would be associated with the city-level characteristics can be controlled with the regression techniques, and the distinct effects of the contextual variables can be isolated. The regression analyses include all of the major individual demographic characteristics associated with civic participation: education, income, age, sex, race, length of residence, and homeownership.50
Many contextual models employing aggregated data are subject to intense selection biases that can produce artificial contextual effects.51 By using measures of a social context that are aggregations of the individual-level characteristics of its members, the contextual variable may be incorporating individual processes that relate to the particular behavior that is being analyzed. For example, people in large cities may be less likely to belong to organizations, not because city size discourages their participation, but because they are the types of people not likely to join associations because of some individual-level characteristic. These techniques are limited by the data available, and there may be other individual-level characteristics associated with living in an affluent or all-white suburb that these measures do not take into account. We must be wary of inadvertently capturing unmodeled effects with the contextual data, but, in general, this should not pose a problem. Controlling for the above individual demographic traits is sufficient to eliminate most of the "noise" that the city-level variables might be picking up. Places where self-selection is more problematic will be discussed in greater detail in the chapters to follow.
Finally, the multivariate analyses allow for the several dimensions of metropolitan places to be considered simultaneously. As I noted above, the diversity of suburban places requires examining several aspects of their social composition (i.e., size, affluence, racial composition, etc.) separately. Yet it is also important to consider how each of these factors operates in the context of the others. The multivariate equations allow for the effects of each dimension to be considered while holding the others constant. In other words, when estimating the effects of city size, the multivariate equations also control for the affluence and racial composition of a community. Similarly, when exploring the impact of residential predominance, we will see that most of the effects of living in a "bedroom suburb" disappear once their economic and racial composition is taken into account. Using these techniques, we can then determine which aspects of this complex and multidimensional process of suburbanization are shaping American civic life.
Outline Of The Book
With these concepts in hand, we are now ready to explore the civic implications of suburbanization. The book's central chapters examine how each of the six major characteristics that distinguish places is shaping civic behavior. Chapter 2 explores population size. Residents of smaller municipalities are more active in local politics, more interested in local affairs, and generally more engaged by the democratic process, irrespective of the size of the surrounding metropolis. The remainder of the chapter explores why municipal boundaries are important in an age of metropolitan sprawl, how place size differently affects various demographic groups (particularly men and women), and what civic potential municipal fragmentation holds for metropolitan areas.
Chapter 3 examines the municipal segregation of Americans according to income. The first part of the chapter explores how suburbanization has contributed to municipal economic differentiation. Economic composition not only separates rich suburbs from poor cities but also distinguishes suburbs from one another—most middle-income suburbs have a wide range of income groups, while rich and poor suburbs are more economically homogeneous. Such economic segregation bridles civic vitality. People in economically homogeneous places are less interested in politics and are less active. The remainder of the chapter outlines how this civic withdrawal reflects the changing role of local democratic institutions in segregated suburbs. Democratic government is supposed to provide a place where diverse social groups come together and peacefully resolve their differences. But when social classes are divided by municipal borders and people cease to participate in politics, local governments lose their republican character. Most important, economic segregation limits the range of interests that can get represented.
Chapter 4 is about the civic paradoxes of racial segregation. With suburbanization, an overwhelming number of suburban municipalities have become predominantly white, while most larger cities are racially heterogeneous. This racial bifurcation has mixed implications for democracy in suburbia. For more social or symbolic types of civic behavior, such as voting and organizational participation, racial homogeneity is a stimulant for activity—both whites and blacks who live among more of their own race are more likely to take part in these activities. For more instrumental types of civic action, however, racial segregation deters citizen involvement. People of all races in predominantly white communities are much less likely to work with neighbors, contact officials, or lobby community boards, largely because such places face few of the problems of more urbanized and racially integrated communities. By separating racial groups along municipal boundaries, suburbanization stifles debate around racial issues, effectively demobilizing citizens from public life.
In Chapter 5, I explore the impact of residential land use. Land use is central to local politics, dictating the agenda of public debate, the direction of public resources, and the character of governing regimes. One of the key characteristics of suburbanization has been the creation of cities made up of nothing but homes. In these "bedroom suburbs," zoning codes bar most nonresidential development and restrict the range of local political conflict. Since people in bedroom suburbs must go to other places to work, shop, and even play, they spend less time in their home communities and are less socially connected to their neighbors. Contrary to the stereotype of the alienated suburbanite, the data show that people in "residentially predominant" places are no more disengaged, apathetic, and removed from politics than people in cities with mixed land uses, once the racial and economic composition of the bedroom suburb is taken into account. In other words, at first glance people in bedroom suburbs are less civically involved, but this is only because of the affluence and racial segregation of their communities. Although land use may be an important determinant of policy making in city halls, it is not a crucial factor in shaping the civic actions of the mass public.
Metropolitan expansion has also increased the differentiation of places by their building age. In the past thirty years, "Snow Belt" places like Cleveland and Buffalo have lost population, while "Sun Belt" places like Atlanta, Dallas, and Phoenix have doubled in size. Some researchers have observed high levels of civic activity in new places. New suburbanites are often portrayed as pioneers, making a fresh start and collectively engaged in community building. Other critics, however, believe that the design and architectural forms of many new places stifle spontaneous social contact, or that they lack the history and civic traditions needed to sustain an active community. Chapter 6 tests these arguments and explores the impact of place age on civic participation. People in newer places are generally no less politically active, except in the Sun Belt. Residents of young Sun Belt communities are the least civically engaged of any Americans. Further analysis reveals that these patterns are not the product of reform-style governments in the Sun Belt. The chapter then examines the distinctiveness of southern suburbanization and why the population expansion of the Sun Belt discourages political and civic activism.
Chapter 7 focuses on the role of institutional arrangements for shaping civic participation. Past research finds that residents of cities with reform-style governments (council-manager or commission governments and representatives elected at large) are more likely to vote than people in cities with mayor-council governments and representatives elected in smaller districts.52 None of this research has examined, however, the impact of such arrangements on other types of participation. In general, I find few effects of institutions on any civic acts outside of voting. The form of institutions seems less important for local civic engagement than the political agenda that is determined by their social composition.
In chapter 8, the conclusion, I draw the findings of the previous chapters together to make some generalizations about the civic consequences of suburbanization. By creating smaller political communities within larger metropolitan sprawls, suburbanization offers great promise for nurturing America's civic health. But political fragmentation also encourages higher levels of social segregation, which, in turn, weakens civic capacity. I then speculate on what the ideal contemporary municipality might be. I argue that smaller yet more heterogeneous places, like New Rochelle, New York, should have higher participation rates than their more homogeneous counterparts. The most civically active municipalities are ones that combine the intimacy of a small town and the social diversity of a large city. Civic capacity is maximized when municipalities provide accessible venues for citizens to resolve their most important differences.
The rest of chapter 8 examines why suburban civic withdrawal should be a concern, and proposes alternatives for renewing civic life in suburbia. With suburbanization, local government becomes a mechanism of exclusion rather than a forum for public debate. Struggles over land use and public resources become contests between institutions rather than among citizens, thus undermining an important need for intermunicipal cooperation. As the federal government assigns more responsibility to localities, the solution to many urban problems, such as sprawl and poverty, depends upon active citizens and municipal cohesion. If suburbanization prevents this civic engagement and sets municipalities against each other, the ability to solve these problems is undercut.
Reversing this suburban civic malaise depends upon changing local political institutions. A little-known but important fact is that municipal government is the creature of state government. States, therefore, must bear the responsibility for reworking local institutional arrangements when their municipalities foster dysfunctional civic behavior and inhibit solutions to problems such as poverty and sprawl. While the political constituency for any such changes will be hard to find, policy advocates need to consider alternative ways of governing local areas. In this spirit I return to suggestions made by Woodrow Wilson and Robert Dahl that metropolitan areas should be governed by federations of smaller places. Metropolitan areas do constitute authentic political communities and deserve to be goverened as such. Indeed, the most pressing problems facing our communities, such as racial segregation, concentrations of poverty, and sprawl, are really problems of the metropolis. Creating metropolitan-level governments, as in Portland or Minneapolis, provides citizens with institutions that reflect their urbanized environment. A political system that integrates municipalities into larger metropolitan governments will produce a meaningful context for citizen activity and encourage a renewal of community and local democratic life in America.
List of Maps and Figures ix
List of Tables xiii
Chapter One: The Rise of a Surburban Demos 1
Chapter Two: All Cries Great and Small 33
Chapter Three: Cities of Riches and Squalor 68
Chapter Four: The Civic Paradox of Racial Segreation 99
Chapter Five: A Bedroom Polis 134
Chapter Six: Boomtowns and the Civic Costs of Air-Conditioning 154
Chapter Seven: Reform Governments and Their Aftermath 175
Chapter Eight: Remaking the Democratic Metropolis 187
Appendix A: The Citizen Participation / Cenus Dataset 215
Appendix B: Logistic and OLS Regression Equations for the Figures 220
Appendix C: Testing the Relationship between Civic Participation and "Self-Inertest Rightly Understood" 236